diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,37821 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 200.0,
+  "eval_steps": 500,
+  "global_step": 54000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2127438336610794,
+      "learning_rate": 0.00199962962962963,
+      "loss": 2.6182,
+      "step": 10
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1781795769929886,
+      "learning_rate": 0.0019992592592592592,
+      "loss": 2.5831,
+      "step": 20
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.20666161179542542,
+      "learning_rate": 0.001998888888888889,
+      "loss": 2.5437,
+      "step": 30
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1834264099597931,
+      "learning_rate": 0.001998518518518519,
+      "loss": 2.6241,
+      "step": 40
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1979055106639862,
+      "learning_rate": 0.001998148148148148,
+      "loss": 2.5936,
+      "step": 50
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.21300563216209412,
+      "learning_rate": 0.001997777777777778,
+      "loss": 2.5713,
+      "step": 60
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17894431948661804,
+      "learning_rate": 0.0019974074074074074,
+      "loss": 2.5696,
+      "step": 70
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17115385830402374,
+      "learning_rate": 0.001997037037037037,
+      "loss": 2.5648,
+      "step": 80
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17225870490074158,
+      "learning_rate": 0.0019966666666666666,
+      "loss": 2.6359,
+      "step": 90
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18164905905723572,
+      "learning_rate": 0.0019962962962962964,
+      "loss": 2.5752,
+      "step": 100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16691413521766663,
+      "learning_rate": 0.0019959259259259257,
+      "loss": 2.568,
+      "step": 110
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.169966459274292,
+      "learning_rate": 0.0019955555555555555,
+      "loss": 2.5995,
+      "step": 120
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18239422142505646,
+      "learning_rate": 0.0019951851851851853,
+      "loss": 2.6098,
+      "step": 130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2049872726202011,
+      "learning_rate": 0.0019948148148148147,
+      "loss": 2.6161,
+      "step": 140
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1980140656232834,
+      "learning_rate": 0.0019944444444444445,
+      "loss": 2.5886,
+      "step": 150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1836952418088913,
+      "learning_rate": 0.0019940740740740743,
+      "loss": 2.5801,
+      "step": 160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17529241740703583,
+      "learning_rate": 0.0019937037037037037,
+      "loss": 2.5513,
+      "step": 170
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.15523068606853485,
+      "learning_rate": 0.0019933333333333335,
+      "loss": 2.5621,
+      "step": 180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.167140930891037,
+      "learning_rate": 0.001992962962962963,
+      "loss": 2.5803,
+      "step": 190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.16198833286762238,
+      "learning_rate": 0.0019925925925925927,
+      "loss": 2.5546,
+      "step": 200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.17828112840652466,
+      "learning_rate": 0.001992222222222222,
+      "loss": 2.5633,
+      "step": 210
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.16426187753677368,
+      "learning_rate": 0.001991851851851852,
+      "loss": 2.5959,
+      "step": 220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1646105945110321,
+      "learning_rate": 0.0019914814814814817,
+      "loss": 2.5519,
+      "step": 230
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.15640555322170258,
+      "learning_rate": 0.001991111111111111,
+      "loss": 2.5908,
+      "step": 240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.16206321120262146,
+      "learning_rate": 0.001990740740740741,
+      "loss": 2.5684,
+      "step": 250
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.15909776091575623,
+      "learning_rate": 0.00199037037037037,
+      "loss": 2.6021,
+      "step": 260
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2778194844722748,
+      "learning_rate": 0.00199,
+      "loss": 2.6248,
+      "step": 270
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.16800597310066223,
+      "learning_rate": 0.00198962962962963,
+      "loss": 2.4247,
+      "step": 280
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.16354629397392273,
+      "learning_rate": 0.001989259259259259,
+      "loss": 2.4314,
+      "step": 290
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.18757009506225586,
+      "learning_rate": 0.001988888888888889,
+      "loss": 2.4616,
+      "step": 300
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.16495901346206665,
+      "learning_rate": 0.001988518518518519,
+      "loss": 2.4934,
+      "step": 310
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.172115758061409,
+      "learning_rate": 0.001988148148148148,
+      "loss": 2.4561,
+      "step": 320
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.1990496963262558,
+      "learning_rate": 0.001987777777777778,
+      "loss": 2.4718,
+      "step": 330
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.19888705015182495,
+      "learning_rate": 0.0019874074074074074,
+      "loss": 2.5014,
+      "step": 340
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.20610582828521729,
+      "learning_rate": 0.001987037037037037,
+      "loss": 2.4453,
+      "step": 350
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.19915615022182465,
+      "learning_rate": 0.0019866666666666665,
+      "loss": 2.474,
+      "step": 360
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.1926460564136505,
+      "learning_rate": 0.0019862962962962963,
+      "loss": 2.5039,
+      "step": 370
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.20390328764915466,
+      "learning_rate": 0.0019859259259259257,
+      "loss": 2.5152,
+      "step": 380
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.21577057242393494,
+      "learning_rate": 0.0019855555555555555,
+      "loss": 2.5675,
+      "step": 390
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.20676636695861816,
+      "learning_rate": 0.0019851851851851853,
+      "loss": 2.4926,
+      "step": 400
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.19866999983787537,
+      "learning_rate": 0.0019848148148148147,
+      "loss": 2.4871,
+      "step": 410
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.19230778515338898,
+      "learning_rate": 0.0019844444444444445,
+      "loss": 2.4865,
+      "step": 420
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.19253689050674438,
+      "learning_rate": 0.0019840740740740743,
+      "loss": 2.4906,
+      "step": 430
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.19537055492401123,
+      "learning_rate": 0.0019837037037037037,
+      "loss": 2.5039,
+      "step": 440
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2013939619064331,
+      "learning_rate": 0.0019833333333333335,
+      "loss": 2.5191,
+      "step": 450
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.1931568831205368,
+      "learning_rate": 0.0019829629629629633,
+      "loss": 2.5574,
+      "step": 460
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.18673449754714966,
+      "learning_rate": 0.0019825925925925927,
+      "loss": 2.53,
+      "step": 470
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.1869499236345291,
+      "learning_rate": 0.0019822222222222225,
+      "loss": 2.5651,
+      "step": 480
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.18973563611507416,
+      "learning_rate": 0.001981851851851852,
+      "loss": 2.5289,
+      "step": 490
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.17751389741897583,
+      "learning_rate": 0.0019814814814814816,
+      "loss": 2.4998,
+      "step": 500
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.18833574652671814,
+      "learning_rate": 0.001981111111111111,
+      "loss": 2.539,
+      "step": 510
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.1748533695936203,
+      "learning_rate": 0.001980740740740741,
+      "loss": 2.5488,
+      "step": 520
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.1988702416419983,
+      "learning_rate": 0.00198037037037037,
+      "loss": 2.5585,
+      "step": 530
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3576868772506714,
+      "learning_rate": 0.00198,
+      "loss": 2.5076,
+      "step": 540
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.20755210518836975,
+      "learning_rate": 0.00197962962962963,
+      "loss": 2.3629,
+      "step": 550
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.21379078924655914,
+      "learning_rate": 0.001979259259259259,
+      "loss": 2.3609,
+      "step": 560
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2395298331975937,
+      "learning_rate": 0.001978888888888889,
+      "loss": 2.3526,
+      "step": 570
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.2168968915939331,
+      "learning_rate": 0.0019785185185185188,
+      "loss": 2.366,
+      "step": 580
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.20609894394874573,
+      "learning_rate": 0.001978148148148148,
+      "loss": 2.3096,
+      "step": 590
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.24114735424518585,
+      "learning_rate": 0.001977777777777778,
+      "loss": 2.3611,
+      "step": 600
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.23323725163936615,
+      "learning_rate": 0.0019774074074074078,
+      "loss": 2.3785,
+      "step": 610
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.23909179866313934,
+      "learning_rate": 0.001977037037037037,
+      "loss": 2.4195,
+      "step": 620
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.2208935022354126,
+      "learning_rate": 0.0019766666666666665,
+      "loss": 2.3907,
+      "step": 630
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.21565470099449158,
+      "learning_rate": 0.0019762962962962963,
+      "loss": 2.3982,
+      "step": 640
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.24911703169345856,
+      "learning_rate": 0.001975925925925926,
+      "loss": 2.37,
+      "step": 650
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.2562112510204315,
+      "learning_rate": 0.0019755555555555555,
+      "loss": 2.3608,
+      "step": 660
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.2207982838153839,
+      "learning_rate": 0.0019751851851851853,
+      "loss": 2.4214,
+      "step": 670
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.25064149498939514,
+      "learning_rate": 0.0019748148148148147,
+      "loss": 2.409,
+      "step": 680
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.21787331998348236,
+      "learning_rate": 0.0019744444444444445,
+      "loss": 2.4376,
+      "step": 690
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.23485612869262695,
+      "learning_rate": 0.0019740740740740743,
+      "loss": 2.4168,
+      "step": 700
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.21844473481178284,
+      "learning_rate": 0.0019737037037037036,
+      "loss": 2.4271,
+      "step": 710
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.21845415234565735,
+      "learning_rate": 0.0019733333333333334,
+      "loss": 2.4636,
+      "step": 720
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.2322779893875122,
+      "learning_rate": 0.0019729629629629633,
+      "loss": 2.4407,
+      "step": 730
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.24358735978603363,
+      "learning_rate": 0.0019725925925925926,
+      "loss": 2.4425,
+      "step": 740
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.2179432511329651,
+      "learning_rate": 0.0019722222222222224,
+      "loss": 2.4417,
+      "step": 750
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.2172284722328186,
+      "learning_rate": 0.001971851851851852,
+      "loss": 2.4661,
+      "step": 760
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.23072928190231323,
+      "learning_rate": 0.0019714814814814816,
+      "loss": 2.4636,
+      "step": 770
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.21614006161689758,
+      "learning_rate": 0.001971111111111111,
+      "loss": 2.4556,
+      "step": 780
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.21256545186042786,
+      "learning_rate": 0.001970740740740741,
+      "loss": 2.454,
+      "step": 790
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.21115510165691376,
+      "learning_rate": 0.00197037037037037,
+      "loss": 2.4724,
+      "step": 800
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.532010555267334,
+      "learning_rate": 0.00197,
+      "loss": 2.4652,
+      "step": 810
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.2634429931640625,
+      "learning_rate": 0.0019696296296296298,
+      "loss": 2.2645,
+      "step": 820
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.24721555411815643,
+      "learning_rate": 0.001969259259259259,
+      "loss": 2.2422,
+      "step": 830
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.2868567109107971,
+      "learning_rate": 0.001968888888888889,
+      "loss": 2.2501,
+      "step": 840
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.2567349076271057,
+      "learning_rate": 0.0019685185185185188,
+      "loss": 2.2504,
+      "step": 850
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.28046759963035583,
+      "learning_rate": 0.001968148148148148,
+      "loss": 2.2738,
+      "step": 860
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.26420778036117554,
+      "learning_rate": 0.001967777777777778,
+      "loss": 2.3218,
+      "step": 870
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.2746962308883667,
+      "learning_rate": 0.0019674074074074077,
+      "loss": 2.293,
+      "step": 880
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.2583375871181488,
+      "learning_rate": 0.001967037037037037,
+      "loss": 2.3095,
+      "step": 890
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.26904764771461487,
+      "learning_rate": 0.0019666666666666665,
+      "loss": 2.3102,
+      "step": 900
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.27271828055381775,
+      "learning_rate": 0.0019662962962962963,
+      "loss": 2.2845,
+      "step": 910
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.25372961163520813,
+      "learning_rate": 0.001965925925925926,
+      "loss": 2.3796,
+      "step": 920
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.2448766678571701,
+      "learning_rate": 0.0019655555555555555,
+      "loss": 2.3591,
+      "step": 930
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.24286913871765137,
+      "learning_rate": 0.0019651851851851853,
+      "loss": 2.313,
+      "step": 940
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.3014277219772339,
+      "learning_rate": 0.0019648148148148146,
+      "loss": 2.3547,
+      "step": 950
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.2519340515136719,
+      "learning_rate": 0.0019644444444444444,
+      "loss": 2.3517,
+      "step": 960
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.25753921270370483,
+      "learning_rate": 0.0019640740740740742,
+      "loss": 2.3593,
+      "step": 970
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.25435492396354675,
+      "learning_rate": 0.0019637037037037036,
+      "loss": 2.3535,
+      "step": 980
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.2480631172657013,
+      "learning_rate": 0.0019633333333333334,
+      "loss": 2.3681,
+      "step": 990
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.2389717549085617,
+      "learning_rate": 0.0019629629629629632,
+      "loss": 2.3574,
+      "step": 1000
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.23806588351726532,
+      "learning_rate": 0.0019625925925925926,
+      "loss": 2.3473,
+      "step": 1010
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.25649794936180115,
+      "learning_rate": 0.0019622222222222224,
+      "loss": 2.3862,
+      "step": 1020
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.2556024491786957,
+      "learning_rate": 0.001961851851851852,
+      "loss": 2.3727,
+      "step": 1030
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.23823349177837372,
+      "learning_rate": 0.0019614814814814816,
+      "loss": 2.39,
+      "step": 1040
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.24092094600200653,
+      "learning_rate": 0.001961111111111111,
+      "loss": 2.369,
+      "step": 1050
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.24226288497447968,
+      "learning_rate": 0.0019607407407407408,
+      "loss": 2.387,
+      "step": 1060
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.23601235449314117,
+      "learning_rate": 0.0019603703703703706,
+      "loss": 2.4098,
+      "step": 1070
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.49935537576675415,
+      "learning_rate": 0.00196,
+      "loss": 2.4015,
+      "step": 1080
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.25792741775512695,
+      "learning_rate": 0.0019596296296296297,
+      "loss": 2.1849,
+      "step": 1090
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.2823100686073303,
+      "learning_rate": 0.001959259259259259,
+      "loss": 2.1466,
+      "step": 1100
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.2801184058189392,
+      "learning_rate": 0.001958888888888889,
+      "loss": 2.1778,
+      "step": 1110
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.283256471157074,
+      "learning_rate": 0.0019585185185185187,
+      "loss": 2.1968,
+      "step": 1120
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.3168976306915283,
+      "learning_rate": 0.001958148148148148,
+      "loss": 2.232,
+      "step": 1130
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.34400302171707153,
+      "learning_rate": 0.001957777777777778,
+      "loss": 2.2086,
+      "step": 1140
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.2858618497848511,
+      "learning_rate": 0.0019574074074074077,
+      "loss": 2.2432,
+      "step": 1150
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.286922812461853,
+      "learning_rate": 0.001957037037037037,
+      "loss": 2.2277,
+      "step": 1160
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.3100804090499878,
+      "learning_rate": 0.001956666666666667,
+      "loss": 2.2501,
+      "step": 1170
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.2689295709133148,
+      "learning_rate": 0.0019562962962962963,
+      "loss": 2.2456,
+      "step": 1180
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.2740320563316345,
+      "learning_rate": 0.001955925925925926,
+      "loss": 2.2712,
+      "step": 1190
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.2822621166706085,
+      "learning_rate": 0.0019555555555555554,
+      "loss": 2.2705,
+      "step": 1200
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.2842746376991272,
+      "learning_rate": 0.0019551851851851852,
+      "loss": 2.267,
+      "step": 1210
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.2541738450527191,
+      "learning_rate": 0.0019548148148148146,
+      "loss": 2.2729,
+      "step": 1220
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.28076863288879395,
+      "learning_rate": 0.0019544444444444444,
+      "loss": 2.3218,
+      "step": 1230
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.30424174666404724,
+      "learning_rate": 0.0019540740740740742,
+      "loss": 2.282,
+      "step": 1240
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.28858065605163574,
+      "learning_rate": 0.0019537037037037036,
+      "loss": 2.2981,
+      "step": 1250
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.25463220477104187,
+      "learning_rate": 0.0019533333333333334,
+      "loss": 2.2858,
+      "step": 1260
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.26512467861175537,
+      "learning_rate": 0.001952962962962963,
+      "loss": 2.3391,
+      "step": 1270
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.26096561551094055,
+      "learning_rate": 0.0019525925925925928,
+      "loss": 2.2934,
+      "step": 1280
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.26782888174057007,
+      "learning_rate": 0.0019522222222222224,
+      "loss": 2.2905,
+      "step": 1290
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.28986856341362,
+      "learning_rate": 0.001951851851851852,
+      "loss": 2.3185,
+      "step": 1300
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.25891393423080444,
+      "learning_rate": 0.0019514814814814813,
+      "loss": 2.3583,
+      "step": 1310
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.293237566947937,
+      "learning_rate": 0.0019511111111111111,
+      "loss": 2.3291,
+      "step": 1320
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.275447815656662,
+      "learning_rate": 0.0019507407407407407,
+      "loss": 2.3604,
+      "step": 1330
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.2680230438709259,
+      "learning_rate": 0.0019503703703703703,
+      "loss": 2.3833,
+      "step": 1340
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.5771416425704956,
+      "learning_rate": 0.00195,
+      "loss": 2.3233,
+      "step": 1350
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 0.30692097544670105,
+      "learning_rate": 0.0019496296296296297,
+      "loss": 2.0699,
+      "step": 1360
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 0.2993278205394745,
+      "learning_rate": 0.0019492592592592593,
+      "loss": 2.0887,
+      "step": 1370
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 0.3134785592556,
+      "learning_rate": 0.001948888888888889,
+      "loss": 2.154,
+      "step": 1380
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 0.28041109442710876,
+      "learning_rate": 0.0019485185185185185,
+      "loss": 2.1398,
+      "step": 1390
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 0.33000850677490234,
+      "learning_rate": 0.0019481481481481483,
+      "loss": 2.1235,
+      "step": 1400
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 0.32690125703811646,
+      "learning_rate": 0.0019477777777777779,
+      "loss": 2.1624,
+      "step": 1410
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 0.3051457405090332,
+      "learning_rate": 0.0019474074074074075,
+      "loss": 2.1376,
+      "step": 1420
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 0.30885449051856995,
+      "learning_rate": 0.0019470370370370373,
+      "loss": 2.1909,
+      "step": 1430
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 0.30936262011528015,
+      "learning_rate": 0.0019466666666666669,
+      "loss": 2.2156,
+      "step": 1440
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 0.29614242911338806,
+      "learning_rate": 0.0019462962962962962,
+      "loss": 2.2312,
+      "step": 1450
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 0.3082076907157898,
+      "learning_rate": 0.0019459259259259258,
+      "loss": 2.1844,
+      "step": 1460
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 0.3007611334323883,
+      "learning_rate": 0.0019455555555555556,
+      "loss": 2.2135,
+      "step": 1470
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 0.29505500197410583,
+      "learning_rate": 0.0019451851851851852,
+      "loss": 2.2212,
+      "step": 1480
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 0.2663390040397644,
+      "learning_rate": 0.0019448148148148148,
+      "loss": 2.2553,
+      "step": 1490
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 0.28654712438583374,
+      "learning_rate": 0.0019444444444444444,
+      "loss": 2.2783,
+      "step": 1500
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 0.29491278529167175,
+      "learning_rate": 0.0019440740740740742,
+      "loss": 2.2367,
+      "step": 1510
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 0.31238043308258057,
+      "learning_rate": 0.0019437037037037038,
+      "loss": 2.2897,
+      "step": 1520
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 0.30257317423820496,
+      "learning_rate": 0.0019433333333333334,
+      "loss": 2.2855,
+      "step": 1530
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 0.2944425046443939,
+      "learning_rate": 0.001942962962962963,
+      "loss": 2.242,
+      "step": 1540
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 0.2931998372077942,
+      "learning_rate": 0.0019425925925925928,
+      "loss": 2.2303,
+      "step": 1550
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 0.2865775227546692,
+      "learning_rate": 0.0019422222222222224,
+      "loss": 2.2459,
+      "step": 1560
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 0.28233101963996887,
+      "learning_rate": 0.001941851851851852,
+      "loss": 2.3051,
+      "step": 1570
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 0.2974388599395752,
+      "learning_rate": 0.0019414814814814815,
+      "loss": 2.3188,
+      "step": 1580
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 0.287343293428421,
+      "learning_rate": 0.0019411111111111111,
+      "loss": 2.2731,
+      "step": 1590
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 0.2774428427219391,
+      "learning_rate": 0.0019407407407407407,
+      "loss": 2.308,
+      "step": 1600
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 0.29502561688423157,
+      "learning_rate": 0.0019403703703703703,
+      "loss": 2.3099,
+      "step": 1610
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.7296469807624817,
+      "learning_rate": 0.0019399999999999999,
+      "loss": 2.3111,
+      "step": 1620
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 0.3095107078552246,
+      "learning_rate": 0.0019396296296296297,
+      "loss": 2.0247,
+      "step": 1630
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 0.3360276520252228,
+      "learning_rate": 0.0019392592592592593,
+      "loss": 2.0108,
+      "step": 1640
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 0.2921351194381714,
+      "learning_rate": 0.0019388888888888889,
+      "loss": 2.0452,
+      "step": 1650
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 0.3128146231174469,
+      "learning_rate": 0.0019385185185185187,
+      "loss": 2.0899,
+      "step": 1660
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 0.32681846618652344,
+      "learning_rate": 0.0019381481481481483,
+      "loss": 2.1214,
+      "step": 1670
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 0.3218587040901184,
+      "learning_rate": 0.0019377777777777778,
+      "loss": 2.1333,
+      "step": 1680
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 0.32519564032554626,
+      "learning_rate": 0.0019374074074074074,
+      "loss": 2.1453,
+      "step": 1690
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 0.3455524742603302,
+      "learning_rate": 0.0019370370370370372,
+      "loss": 2.1466,
+      "step": 1700
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 0.3287692666053772,
+      "learning_rate": 0.0019366666666666668,
+      "loss": 2.1663,
+      "step": 1710
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 0.34251290559768677,
+      "learning_rate": 0.0019362962962962964,
+      "loss": 2.1636,
+      "step": 1720
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 0.3383674621582031,
+      "learning_rate": 0.0019359259259259258,
+      "loss": 2.1728,
+      "step": 1730
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 0.29128894209861755,
+      "learning_rate": 0.0019355555555555556,
+      "loss": 2.1815,
+      "step": 1740
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 0.3406875431537628,
+      "learning_rate": 0.0019351851851851852,
+      "loss": 2.228,
+      "step": 1750
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 0.3243136405944824,
+      "learning_rate": 0.0019348148148148148,
+      "loss": 2.1952,
+      "step": 1760
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 0.29255208373069763,
+      "learning_rate": 0.0019344444444444444,
+      "loss": 2.213,
+      "step": 1770
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 0.3365947902202606,
+      "learning_rate": 0.0019340740740740742,
+      "loss": 2.2172,
+      "step": 1780
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 0.3175502121448517,
+      "learning_rate": 0.0019337037037037038,
+      "loss": 2.2419,
+      "step": 1790
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 0.27796781063079834,
+      "learning_rate": 0.0019333333333333333,
+      "loss": 2.1788,
+      "step": 1800
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 0.3237530589103699,
+      "learning_rate": 0.001932962962962963,
+      "loss": 2.1958,
+      "step": 1810
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 0.2936450242996216,
+      "learning_rate": 0.0019325925925925927,
+      "loss": 2.2351,
+      "step": 1820
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 0.3066917657852173,
+      "learning_rate": 0.0019322222222222223,
+      "loss": 2.2373,
+      "step": 1830
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 0.32224395871162415,
+      "learning_rate": 0.001931851851851852,
+      "loss": 2.2343,
+      "step": 1840
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.29986292123794556,
+      "learning_rate": 0.0019314814814814817,
+      "loss": 2.2505,
+      "step": 1850
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 0.29892221093177795,
+      "learning_rate": 0.001931111111111111,
+      "loss": 2.2092,
+      "step": 1860
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 0.2913380265235901,
+      "learning_rate": 0.0019307407407407407,
+      "loss": 2.2298,
+      "step": 1870
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 0.28266847133636475,
+      "learning_rate": 0.0019303703703703703,
+      "loss": 2.2531,
+      "step": 1880
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.5149755477905273,
+      "learning_rate": 0.00193,
+      "loss": 2.2141,
+      "step": 1890
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.34365060925483704,
+      "learning_rate": 0.0019296296296296297,
+      "loss": 2.0277,
+      "step": 1900
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 0.3547360599040985,
+      "learning_rate": 0.0019292592592592593,
+      "loss": 2.0065,
+      "step": 1910
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 0.36707526445388794,
+      "learning_rate": 0.0019288888888888888,
+      "loss": 2.0481,
+      "step": 1920
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 0.3768918812274933,
+      "learning_rate": 0.0019285185185185186,
+      "loss": 2.0431,
+      "step": 1930
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 0.36633971333503723,
+      "learning_rate": 0.0019281481481481482,
+      "loss": 2.0451,
+      "step": 1940
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 0.33258959650993347,
+      "learning_rate": 0.0019277777777777778,
+      "loss": 2.0471,
+      "step": 1950
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 0.31032317876815796,
+      "learning_rate": 0.0019274074074074074,
+      "loss": 2.0915,
+      "step": 1960
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 0.3308986723423004,
+      "learning_rate": 0.0019270370370370372,
+      "loss": 2.1063,
+      "step": 1970
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 0.32878440618515015,
+      "learning_rate": 0.0019266666666666668,
+      "loss": 2.1293,
+      "step": 1980
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 0.3539588153362274,
+      "learning_rate": 0.0019262962962962964,
+      "loss": 2.1302,
+      "step": 1990
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 0.33087241649627686,
+      "learning_rate": 0.0019259259259259258,
+      "loss": 2.1474,
+      "step": 2000
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 0.35217225551605225,
+      "learning_rate": 0.0019255555555555556,
+      "loss": 2.1043,
+      "step": 2010
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.3334253132343292,
+      "learning_rate": 0.0019251851851851852,
+      "loss": 2.1372,
+      "step": 2020
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.32482457160949707,
+      "learning_rate": 0.0019248148148148147,
+      "loss": 2.1861,
+      "step": 2030
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 0.3670805096626282,
+      "learning_rate": 0.0019244444444444443,
+      "loss": 2.1611,
+      "step": 2040
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 0.3149818181991577,
+      "learning_rate": 0.0019240740740740741,
+      "loss": 2.1459,
+      "step": 2050
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 0.3138531446456909,
+      "learning_rate": 0.0019237037037037037,
+      "loss": 2.1299,
+      "step": 2060
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 0.30926617980003357,
+      "learning_rate": 0.0019233333333333333,
+      "loss": 2.1817,
+      "step": 2070
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 0.3508884310722351,
+      "learning_rate": 0.0019229629629629631,
+      "loss": 2.197,
+      "step": 2080
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 0.3276742696762085,
+      "learning_rate": 0.0019225925925925927,
+      "loss": 2.1813,
+      "step": 2090
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 0.31571853160858154,
+      "learning_rate": 0.0019222222222222223,
+      "loss": 2.1706,
+      "step": 2100
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 0.300502210855484,
+      "learning_rate": 0.0019218518518518519,
+      "loss": 2.163,
+      "step": 2110
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 0.32625290751457214,
+      "learning_rate": 0.0019214814814814817,
+      "loss": 2.189,
+      "step": 2120
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 0.28779491782188416,
+      "learning_rate": 0.0019211111111111113,
+      "loss": 2.1737,
+      "step": 2130
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 0.32611775398254395,
+      "learning_rate": 0.0019207407407407407,
+      "loss": 2.2241,
+      "step": 2140
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 0.3058241009712219,
+      "learning_rate": 0.0019203703703703702,
+      "loss": 2.2035,
+      "step": 2150
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.5639318823814392,
+      "learning_rate": 0.00192,
+      "loss": 2.1991,
+      "step": 2160
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 0.34727299213409424,
+      "learning_rate": 0.0019196296296296296,
+      "loss": 1.9146,
+      "step": 2170
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 0.3747304081916809,
+      "learning_rate": 0.0019192592592592592,
+      "loss": 1.9499,
+      "step": 2180
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 0.3590710759162903,
+      "learning_rate": 0.0019188888888888888,
+      "loss": 2.0072,
+      "step": 2190
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 0.413136750459671,
+      "learning_rate": 0.0019185185185185186,
+      "loss": 2.0058,
+      "step": 2200
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.3658411502838135,
+      "learning_rate": 0.0019181481481481482,
+      "loss": 2.0421,
+      "step": 2210
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 0.3345026671886444,
+      "learning_rate": 0.0019177777777777778,
+      "loss": 2.0118,
+      "step": 2220
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 0.3682136833667755,
+      "learning_rate": 0.0019174074074074074,
+      "loss": 2.0397,
+      "step": 2230
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 0.37295275926589966,
+      "learning_rate": 0.0019170370370370372,
+      "loss": 2.0797,
+      "step": 2240
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 0.3477710783481598,
+      "learning_rate": 0.0019166666666666668,
+      "loss": 2.0644,
+      "step": 2250
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 0.33700767159461975,
+      "learning_rate": 0.0019162962962962964,
+      "loss": 2.0802,
+      "step": 2260
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 0.3817288875579834,
+      "learning_rate": 0.0019159259259259262,
+      "loss": 2.1217,
+      "step": 2270
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 0.3545536994934082,
+      "learning_rate": 0.0019155555555555555,
+      "loss": 2.1185,
+      "step": 2280
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 0.39838239550590515,
+      "learning_rate": 0.0019151851851851851,
+      "loss": 2.0784,
+      "step": 2290
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 0.35482680797576904,
+      "learning_rate": 0.0019148148148148147,
+      "loss": 2.0967,
+      "step": 2300
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 0.3527957797050476,
+      "learning_rate": 0.0019144444444444445,
+      "loss": 2.1285,
+      "step": 2310
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 0.33589911460876465,
+      "learning_rate": 0.0019140740740740741,
+      "loss": 2.1375,
+      "step": 2320
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 0.32576510310173035,
+      "learning_rate": 0.0019137037037037037,
+      "loss": 2.1766,
+      "step": 2330
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 0.31789445877075195,
+      "learning_rate": 0.0019133333333333333,
+      "loss": 2.1351,
+      "step": 2340
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 0.3491995632648468,
+      "learning_rate": 0.001912962962962963,
+      "loss": 2.1502,
+      "step": 2350
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 0.33717969059944153,
+      "learning_rate": 0.0019125925925925927,
+      "loss": 2.1529,
+      "step": 2360
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 0.31113317608833313,
+      "learning_rate": 0.0019122222222222223,
+      "loss": 2.1498,
+      "step": 2370
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 0.34073254466056824,
+      "learning_rate": 0.0019118518518518519,
+      "loss": 2.162,
+      "step": 2380
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 0.32688942551612854,
+      "learning_rate": 0.0019114814814814817,
+      "loss": 2.164,
+      "step": 2390
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 0.3454456329345703,
+      "learning_rate": 0.0019111111111111113,
+      "loss": 2.168,
+      "step": 2400
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 0.326792448759079,
+      "learning_rate": 0.0019107407407407408,
+      "loss": 2.1524,
+      "step": 2410
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.32493969798088074,
+      "learning_rate": 0.0019103703703703702,
+      "loss": 2.1982,
+      "step": 2420
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.569231390953064,
+      "learning_rate": 0.00191,
+      "loss": 2.1692,
+      "step": 2430
+    },
+    {
+      "epoch": 9.04,
+      "grad_norm": 0.37039652466773987,
+      "learning_rate": 0.0019096296296296296,
+      "loss": 1.9196,
+      "step": 2440
+    },
+    {
+      "epoch": 9.07,
+      "grad_norm": 0.36663028597831726,
+      "learning_rate": 0.0019092592592592592,
+      "loss": 1.9251,
+      "step": 2450
+    },
+    {
+      "epoch": 9.11,
+      "grad_norm": 0.40725961327552795,
+      "learning_rate": 0.001908888888888889,
+      "loss": 1.9387,
+      "step": 2460
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 0.38818269968032837,
+      "learning_rate": 0.0019085185185185186,
+      "loss": 1.9633,
+      "step": 2470
+    },
+    {
+      "epoch": 9.19,
+      "grad_norm": 0.3413904905319214,
+      "learning_rate": 0.0019081481481481482,
+      "loss": 1.9779,
+      "step": 2480
+    },
+    {
+      "epoch": 9.22,
+      "grad_norm": 0.3483598232269287,
+      "learning_rate": 0.0019077777777777778,
+      "loss": 2.0182,
+      "step": 2490
+    },
+    {
+      "epoch": 9.26,
+      "grad_norm": 0.3658324182033539,
+      "learning_rate": 0.0019074074074074076,
+      "loss": 2.0443,
+      "step": 2500
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 0.3571145534515381,
+      "learning_rate": 0.0019070370370370372,
+      "loss": 2.0266,
+      "step": 2510
+    },
+    {
+      "epoch": 9.33,
+      "grad_norm": 0.3722548186779022,
+      "learning_rate": 0.0019066666666666668,
+      "loss": 2.0479,
+      "step": 2520
+    },
+    {
+      "epoch": 9.37,
+      "grad_norm": 0.39253392815589905,
+      "learning_rate": 0.0019062962962962963,
+      "loss": 2.0347,
+      "step": 2530
+    },
+    {
+      "epoch": 9.41,
+      "grad_norm": 0.36997315287590027,
+      "learning_rate": 0.0019059259259259261,
+      "loss": 2.0454,
+      "step": 2540
+    },
+    {
+      "epoch": 9.44,
+      "grad_norm": 0.3789735734462738,
+      "learning_rate": 0.0019055555555555555,
+      "loss": 2.0611,
+      "step": 2550
+    },
+    {
+      "epoch": 9.48,
+      "grad_norm": 0.3231625556945801,
+      "learning_rate": 0.001905185185185185,
+      "loss": 2.126,
+      "step": 2560
+    },
+    {
+      "epoch": 9.52,
+      "grad_norm": 0.3473018407821655,
+      "learning_rate": 0.0019048148148148147,
+      "loss": 2.0589,
+      "step": 2570
+    },
+    {
+      "epoch": 9.56,
+      "grad_norm": 0.3326464295387268,
+      "learning_rate": 0.0019044444444444445,
+      "loss": 2.1153,
+      "step": 2580
+    },
+    {
+      "epoch": 9.59,
+      "grad_norm": 0.3558256924152374,
+      "learning_rate": 0.001904074074074074,
+      "loss": 2.0861,
+      "step": 2590
+    },
+    {
+      "epoch": 9.63,
+      "grad_norm": 0.35280081629753113,
+      "learning_rate": 0.0019037037037037037,
+      "loss": 2.099,
+      "step": 2600
+    },
+    {
+      "epoch": 9.67,
+      "grad_norm": 0.3500099182128906,
+      "learning_rate": 0.0019033333333333333,
+      "loss": 2.116,
+      "step": 2610
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 0.35508620738983154,
+      "learning_rate": 0.001902962962962963,
+      "loss": 2.1118,
+      "step": 2620
+    },
+    {
+      "epoch": 9.74,
+      "grad_norm": 0.34156134724617004,
+      "learning_rate": 0.0019025925925925927,
+      "loss": 2.1196,
+      "step": 2630
+    },
+    {
+      "epoch": 9.78,
+      "grad_norm": 0.35427629947662354,
+      "learning_rate": 0.0019022222222222222,
+      "loss": 2.1033,
+      "step": 2640
+    },
+    {
+      "epoch": 9.81,
+      "grad_norm": 0.3459043502807617,
+      "learning_rate": 0.0019018518518518518,
+      "loss": 2.124,
+      "step": 2650
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 0.38089868426322937,
+      "learning_rate": 0.0019014814814814816,
+      "loss": 2.1434,
+      "step": 2660
+    },
+    {
+      "epoch": 9.89,
+      "grad_norm": 0.3299236595630646,
+      "learning_rate": 0.0019011111111111112,
+      "loss": 2.1155,
+      "step": 2670
+    },
+    {
+      "epoch": 9.93,
+      "grad_norm": 0.37226101756095886,
+      "learning_rate": 0.0019007407407407408,
+      "loss": 2.1338,
+      "step": 2680
+    },
+    {
+      "epoch": 9.96,
+      "grad_norm": 0.34492045640945435,
+      "learning_rate": 0.0019003703703703704,
+      "loss": 2.1391,
+      "step": 2690
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.718737781047821,
+      "learning_rate": 0.0019,
+      "loss": 2.119,
+      "step": 2700
+    },
+    {
+      "epoch": 10.04,
+      "grad_norm": 0.39133042097091675,
+      "learning_rate": 0.0018996296296296296,
+      "loss": 1.9413,
+      "step": 2710
+    },
+    {
+      "epoch": 10.07,
+      "grad_norm": 0.38486966490745544,
+      "learning_rate": 0.0018992592592592592,
+      "loss": 1.9115,
+      "step": 2720
+    },
+    {
+      "epoch": 10.11,
+      "grad_norm": 0.42015212774276733,
+      "learning_rate": 0.001898888888888889,
+      "loss": 1.9019,
+      "step": 2730
+    },
+    {
+      "epoch": 10.15,
+      "grad_norm": 0.4082113206386566,
+      "learning_rate": 0.0018985185185185186,
+      "loss": 1.9221,
+      "step": 2740
+    },
+    {
+      "epoch": 10.19,
+      "grad_norm": 0.37983325123786926,
+      "learning_rate": 0.0018981481481481482,
+      "loss": 1.942,
+      "step": 2750
+    },
+    {
+      "epoch": 10.22,
+      "grad_norm": 0.37726789712905884,
+      "learning_rate": 0.0018977777777777777,
+      "loss": 1.9894,
+      "step": 2760
+    },
+    {
+      "epoch": 10.26,
+      "grad_norm": 0.3986751139163971,
+      "learning_rate": 0.0018974074074074075,
+      "loss": 2.0071,
+      "step": 2770
+    },
+    {
+      "epoch": 10.3,
+      "grad_norm": 0.37349823117256165,
+      "learning_rate": 0.0018970370370370371,
+      "loss": 1.9686,
+      "step": 2780
+    },
+    {
+      "epoch": 10.33,
+      "grad_norm": 0.37317484617233276,
+      "learning_rate": 0.0018966666666666667,
+      "loss": 2.023,
+      "step": 2790
+    },
+    {
+      "epoch": 10.37,
+      "grad_norm": 0.4159048795700073,
+      "learning_rate": 0.0018962962962962963,
+      "loss": 2.0305,
+      "step": 2800
+    },
+    {
+      "epoch": 10.41,
+      "grad_norm": 0.36336109042167664,
+      "learning_rate": 0.0018959259259259261,
+      "loss": 2.0059,
+      "step": 2810
+    },
+    {
+      "epoch": 10.44,
+      "grad_norm": 0.3658113181591034,
+      "learning_rate": 0.0018955555555555557,
+      "loss": 2.0488,
+      "step": 2820
+    },
+    {
+      "epoch": 10.48,
+      "grad_norm": 0.41594254970550537,
+      "learning_rate": 0.001895185185185185,
+      "loss": 2.0423,
+      "step": 2830
+    },
+    {
+      "epoch": 10.52,
+      "grad_norm": 0.3627219498157501,
+      "learning_rate": 0.0018948148148148147,
+      "loss": 2.0593,
+      "step": 2840
+    },
+    {
+      "epoch": 10.56,
+      "grad_norm": 0.360855370759964,
+      "learning_rate": 0.0018944444444444445,
+      "loss": 2.0557,
+      "step": 2850
+    },
+    {
+      "epoch": 10.59,
+      "grad_norm": 0.3695986866950989,
+      "learning_rate": 0.001894074074074074,
+      "loss": 2.0679,
+      "step": 2860
+    },
+    {
+      "epoch": 10.63,
+      "grad_norm": 0.3898347020149231,
+      "learning_rate": 0.0018937037037037037,
+      "loss": 2.1022,
+      "step": 2870
+    },
+    {
+      "epoch": 10.67,
+      "grad_norm": 0.3742794096469879,
+      "learning_rate": 0.0018933333333333335,
+      "loss": 2.092,
+      "step": 2880
+    },
+    {
+      "epoch": 10.7,
+      "grad_norm": 0.3694680333137512,
+      "learning_rate": 0.001892962962962963,
+      "loss": 2.0915,
+      "step": 2890
+    },
+    {
+      "epoch": 10.74,
+      "grad_norm": 0.3656015992164612,
+      "learning_rate": 0.0018925925925925926,
+      "loss": 2.0832,
+      "step": 2900
+    },
+    {
+      "epoch": 10.78,
+      "grad_norm": 0.3568435609340668,
+      "learning_rate": 0.0018922222222222222,
+      "loss": 2.0981,
+      "step": 2910
+    },
+    {
+      "epoch": 10.81,
+      "grad_norm": 0.3529919981956482,
+      "learning_rate": 0.001891851851851852,
+      "loss": 2.0947,
+      "step": 2920
+    },
+    {
+      "epoch": 10.85,
+      "grad_norm": 0.37247973680496216,
+      "learning_rate": 0.0018914814814814816,
+      "loss": 2.1217,
+      "step": 2930
+    },
+    {
+      "epoch": 10.89,
+      "grad_norm": 0.36205926537513733,
+      "learning_rate": 0.0018911111111111112,
+      "loss": 2.095,
+      "step": 2940
+    },
+    {
+      "epoch": 10.93,
+      "grad_norm": 0.3482193052768707,
+      "learning_rate": 0.0018907407407407408,
+      "loss": 2.0999,
+      "step": 2950
+    },
+    {
+      "epoch": 10.96,
+      "grad_norm": 0.3682381510734558,
+      "learning_rate": 0.0018903703703703706,
+      "loss": 2.097,
+      "step": 2960
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.7415529489517212,
+      "learning_rate": 0.00189,
+      "loss": 2.0817,
+      "step": 2970
+    },
+    {
+      "epoch": 11.04,
+      "grad_norm": 0.3742087781429291,
+      "learning_rate": 0.0018896296296296296,
+      "loss": 1.8732,
+      "step": 2980
+    },
+    {
+      "epoch": 11.07,
+      "grad_norm": 0.42774471640586853,
+      "learning_rate": 0.0018892592592592591,
+      "loss": 1.9006,
+      "step": 2990
+    },
+    {
+      "epoch": 11.11,
+      "grad_norm": 0.4058782458305359,
+      "learning_rate": 0.001888888888888889,
+      "loss": 1.8626,
+      "step": 3000
+    },
+    {
+      "epoch": 11.15,
+      "grad_norm": 0.3706398904323578,
+      "learning_rate": 0.0018885185185185185,
+      "loss": 1.9556,
+      "step": 3010
+    },
+    {
+      "epoch": 11.19,
+      "grad_norm": 0.38733819127082825,
+      "learning_rate": 0.0018881481481481481,
+      "loss": 1.9707,
+      "step": 3020
+    },
+    {
+      "epoch": 11.22,
+      "grad_norm": 0.4044550061225891,
+      "learning_rate": 0.0018877777777777777,
+      "loss": 1.9411,
+      "step": 3030
+    },
+    {
+      "epoch": 11.26,
+      "grad_norm": 0.43712499737739563,
+      "learning_rate": 0.0018874074074074075,
+      "loss": 1.9652,
+      "step": 3040
+    },
+    {
+      "epoch": 11.3,
+      "grad_norm": 0.3710927367210388,
+      "learning_rate": 0.0018870370370370371,
+      "loss": 1.9604,
+      "step": 3050
+    },
+    {
+      "epoch": 11.33,
+      "grad_norm": 0.4272994101047516,
+      "learning_rate": 0.0018866666666666667,
+      "loss": 1.9831,
+      "step": 3060
+    },
+    {
+      "epoch": 11.37,
+      "grad_norm": 0.39658141136169434,
+      "learning_rate": 0.0018862962962962965,
+      "loss": 1.946,
+      "step": 3070
+    },
+    {
+      "epoch": 11.41,
+      "grad_norm": 0.3768123686313629,
+      "learning_rate": 0.001885925925925926,
+      "loss": 2.0135,
+      "step": 3080
+    },
+    {
+      "epoch": 11.44,
+      "grad_norm": 0.405793696641922,
+      "learning_rate": 0.0018855555555555557,
+      "loss": 2.0506,
+      "step": 3090
+    },
+    {
+      "epoch": 11.48,
+      "grad_norm": 0.3863692581653595,
+      "learning_rate": 0.001885185185185185,
+      "loss": 2.0011,
+      "step": 3100
+    },
+    {
+      "epoch": 11.52,
+      "grad_norm": 0.3689294457435608,
+      "learning_rate": 0.0018848148148148149,
+      "loss": 1.9958,
+      "step": 3110
+    },
+    {
+      "epoch": 11.56,
+      "grad_norm": 0.3637295365333557,
+      "learning_rate": 0.0018844444444444444,
+      "loss": 2.0012,
+      "step": 3120
+    },
+    {
+      "epoch": 11.59,
+      "grad_norm": 0.38232067227363586,
+      "learning_rate": 0.001884074074074074,
+      "loss": 2.029,
+      "step": 3130
+    },
+    {
+      "epoch": 11.63,
+      "grad_norm": 0.3879601061344147,
+      "learning_rate": 0.0018837037037037036,
+      "loss": 2.0689,
+      "step": 3140
+    },
+    {
+      "epoch": 11.67,
+      "grad_norm": 0.35336965322494507,
+      "learning_rate": 0.0018833333333333334,
+      "loss": 2.0494,
+      "step": 3150
+    },
+    {
+      "epoch": 11.7,
+      "grad_norm": 0.39037132263183594,
+      "learning_rate": 0.001882962962962963,
+      "loss": 2.0513,
+      "step": 3160
+    },
+    {
+      "epoch": 11.74,
+      "grad_norm": 0.383658766746521,
+      "learning_rate": 0.0018825925925925926,
+      "loss": 2.0459,
+      "step": 3170
+    },
+    {
+      "epoch": 11.78,
+      "grad_norm": 0.3601329028606415,
+      "learning_rate": 0.0018822222222222222,
+      "loss": 2.0795,
+      "step": 3180
+    },
+    {
+      "epoch": 11.81,
+      "grad_norm": 0.36117351055145264,
+      "learning_rate": 0.001881851851851852,
+      "loss": 2.0726,
+      "step": 3190
+    },
+    {
+      "epoch": 11.85,
+      "grad_norm": 0.3798144459724426,
+      "learning_rate": 0.0018814814814814816,
+      "loss": 2.0436,
+      "step": 3200
+    },
+    {
+      "epoch": 11.89,
+      "grad_norm": 0.35812708735466003,
+      "learning_rate": 0.0018811111111111112,
+      "loss": 2.1049,
+      "step": 3210
+    },
+    {
+      "epoch": 11.93,
+      "grad_norm": 0.38053765892982483,
+      "learning_rate": 0.0018807407407407408,
+      "loss": 2.0561,
+      "step": 3220
+    },
+    {
+      "epoch": 11.96,
+      "grad_norm": 0.3914078176021576,
+      "learning_rate": 0.0018803703703703706,
+      "loss": 2.0689,
+      "step": 3230
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.6013017296791077,
+      "learning_rate": 0.00188,
+      "loss": 2.1096,
+      "step": 3240
+    },
+    {
+      "epoch": 12.04,
+      "grad_norm": 0.3909628093242645,
+      "learning_rate": 0.0018796296296296295,
+      "loss": 1.842,
+      "step": 3250
+    },
+    {
+      "epoch": 12.07,
+      "grad_norm": 0.4217047393321991,
+      "learning_rate": 0.0018792592592592591,
+      "loss": 1.8258,
+      "step": 3260
+    },
+    {
+      "epoch": 12.11,
+      "grad_norm": 0.4064037799835205,
+      "learning_rate": 0.001878888888888889,
+      "loss": 1.8367,
+      "step": 3270
+    },
+    {
+      "epoch": 12.15,
+      "grad_norm": 0.4413491189479828,
+      "learning_rate": 0.0018785185185185185,
+      "loss": 1.8863,
+      "step": 3280
+    },
+    {
+      "epoch": 12.19,
+      "grad_norm": 0.4350806474685669,
+      "learning_rate": 0.001878148148148148,
+      "loss": 1.8945,
+      "step": 3290
+    },
+    {
+      "epoch": 12.22,
+      "grad_norm": 0.3987176716327667,
+      "learning_rate": 0.001877777777777778,
+      "loss": 1.9026,
+      "step": 3300
+    },
+    {
+      "epoch": 12.26,
+      "grad_norm": 0.41014865040779114,
+      "learning_rate": 0.0018774074074074075,
+      "loss": 1.9443,
+      "step": 3310
+    },
+    {
+      "epoch": 12.3,
+      "grad_norm": 0.38420772552490234,
+      "learning_rate": 0.001877037037037037,
+      "loss": 1.9551,
+      "step": 3320
+    },
+    {
+      "epoch": 12.33,
+      "grad_norm": 0.41501930356025696,
+      "learning_rate": 0.0018766666666666667,
+      "loss": 1.9956,
+      "step": 3330
+    },
+    {
+      "epoch": 12.37,
+      "grad_norm": 0.386685848236084,
+      "learning_rate": 0.0018762962962962965,
+      "loss": 1.9536,
+      "step": 3340
+    },
+    {
+      "epoch": 12.41,
+      "grad_norm": 0.4385821521282196,
+      "learning_rate": 0.001875925925925926,
+      "loss": 1.9822,
+      "step": 3350
+    },
+    {
+      "epoch": 12.44,
+      "grad_norm": 0.4114198088645935,
+      "learning_rate": 0.0018755555555555557,
+      "loss": 2.0037,
+      "step": 3360
+    },
+    {
+      "epoch": 12.48,
+      "grad_norm": 0.4060194194316864,
+      "learning_rate": 0.0018751851851851852,
+      "loss": 1.9864,
+      "step": 3370
+    },
+    {
+      "epoch": 12.52,
+      "grad_norm": 0.3886150121688843,
+      "learning_rate": 0.0018748148148148148,
+      "loss": 1.9556,
+      "step": 3380
+    },
+    {
+      "epoch": 12.56,
+      "grad_norm": 0.36917322874069214,
+      "learning_rate": 0.0018744444444444444,
+      "loss": 2.0239,
+      "step": 3390
+    },
+    {
+      "epoch": 12.59,
+      "grad_norm": 0.3943292498588562,
+      "learning_rate": 0.001874074074074074,
+      "loss": 2.0027,
+      "step": 3400
+    },
+    {
+      "epoch": 12.63,
+      "grad_norm": 0.3857436180114746,
+      "learning_rate": 0.0018737037037037036,
+      "loss": 1.9949,
+      "step": 3410
+    },
+    {
+      "epoch": 12.67,
+      "grad_norm": 0.4206480383872986,
+      "learning_rate": 0.0018733333333333334,
+      "loss": 2.0477,
+      "step": 3420
+    },
+    {
+      "epoch": 12.7,
+      "grad_norm": 0.3751334249973297,
+      "learning_rate": 0.001872962962962963,
+      "loss": 2.0233,
+      "step": 3430
+    },
+    {
+      "epoch": 12.74,
+      "grad_norm": 0.38285306096076965,
+      "learning_rate": 0.0018725925925925926,
+      "loss": 2.0445,
+      "step": 3440
+    },
+    {
+      "epoch": 12.78,
+      "grad_norm": 0.37726810574531555,
+      "learning_rate": 0.0018722222222222222,
+      "loss": 2.0564,
+      "step": 3450
+    },
+    {
+      "epoch": 12.81,
+      "grad_norm": 0.3757777512073517,
+      "learning_rate": 0.001871851851851852,
+      "loss": 2.0471,
+      "step": 3460
+    },
+    {
+      "epoch": 12.85,
+      "grad_norm": 0.4291413128376007,
+      "learning_rate": 0.0018714814814814816,
+      "loss": 1.9982,
+      "step": 3470
+    },
+    {
+      "epoch": 12.89,
+      "grad_norm": 0.37266525626182556,
+      "learning_rate": 0.0018711111111111112,
+      "loss": 2.0529,
+      "step": 3480
+    },
+    {
+      "epoch": 12.93,
+      "grad_norm": 0.37555840611457825,
+      "learning_rate": 0.001870740740740741,
+      "loss": 2.0596,
+      "step": 3490
+    },
+    {
+      "epoch": 12.96,
+      "grad_norm": 0.38363468647003174,
+      "learning_rate": 0.0018703703703703705,
+      "loss": 2.0597,
+      "step": 3500
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.8527718186378479,
+      "learning_rate": 0.0018700000000000001,
+      "loss": 2.0715,
+      "step": 3510
+    },
+    {
+      "epoch": 13.04,
+      "grad_norm": 0.3870679438114166,
+      "learning_rate": 0.0018696296296296295,
+      "loss": 1.783,
+      "step": 3520
+    },
+    {
+      "epoch": 13.07,
+      "grad_norm": 0.4346376955509186,
+      "learning_rate": 0.0018692592592592593,
+      "loss": 1.8299,
+      "step": 3530
+    },
+    {
+      "epoch": 13.11,
+      "grad_norm": 0.40322786569595337,
+      "learning_rate": 0.001868888888888889,
+      "loss": 1.8313,
+      "step": 3540
+    },
+    {
+      "epoch": 13.15,
+      "grad_norm": 0.43011534214019775,
+      "learning_rate": 0.0018685185185185185,
+      "loss": 1.8661,
+      "step": 3550
+    },
+    {
+      "epoch": 13.19,
+      "grad_norm": 0.40157124400138855,
+      "learning_rate": 0.001868148148148148,
+      "loss": 1.8585,
+      "step": 3560
+    },
+    {
+      "epoch": 13.22,
+      "grad_norm": 0.45722874999046326,
+      "learning_rate": 0.0018677777777777779,
+      "loss": 1.9132,
+      "step": 3570
+    },
+    {
+      "epoch": 13.26,
+      "grad_norm": 0.42185381054878235,
+      "learning_rate": 0.0018674074074074075,
+      "loss": 1.9007,
+      "step": 3580
+    },
+    {
+      "epoch": 13.3,
+      "grad_norm": 0.4317697584629059,
+      "learning_rate": 0.001867037037037037,
+      "loss": 1.9352,
+      "step": 3590
+    },
+    {
+      "epoch": 13.33,
+      "grad_norm": 0.3869325518608093,
+      "learning_rate": 0.0018666666666666666,
+      "loss": 1.9519,
+      "step": 3600
+    },
+    {
+      "epoch": 13.37,
+      "grad_norm": 0.4048895537853241,
+      "learning_rate": 0.0018662962962962965,
+      "loss": 1.9419,
+      "step": 3610
+    },
+    {
+      "epoch": 13.41,
+      "grad_norm": 0.457479864358902,
+      "learning_rate": 0.001865925925925926,
+      "loss": 1.9694,
+      "step": 3620
+    },
+    {
+      "epoch": 13.44,
+      "grad_norm": 0.4000145196914673,
+      "learning_rate": 0.0018655555555555556,
+      "loss": 1.9519,
+      "step": 3630
+    },
+    {
+      "epoch": 13.48,
+      "grad_norm": 0.4235948920249939,
+      "learning_rate": 0.0018651851851851852,
+      "loss": 1.9422,
+      "step": 3640
+    },
+    {
+      "epoch": 13.52,
+      "grad_norm": 0.3991868197917938,
+      "learning_rate": 0.001864814814814815,
+      "loss": 1.9906,
+      "step": 3650
+    },
+    {
+      "epoch": 13.56,
+      "grad_norm": 0.3810255825519562,
+      "learning_rate": 0.0018644444444444444,
+      "loss": 1.9545,
+      "step": 3660
+    },
+    {
+      "epoch": 13.59,
+      "grad_norm": 0.4063350558280945,
+      "learning_rate": 0.001864074074074074,
+      "loss": 1.9896,
+      "step": 3670
+    },
+    {
+      "epoch": 13.63,
+      "grad_norm": 0.39621415734291077,
+      "learning_rate": 0.0018637037037037036,
+      "loss": 2.0098,
+      "step": 3680
+    },
+    {
+      "epoch": 13.67,
+      "grad_norm": 0.4067321717739105,
+      "learning_rate": 0.0018633333333333334,
+      "loss": 1.993,
+      "step": 3690
+    },
+    {
+      "epoch": 13.7,
+      "grad_norm": 0.4024389088153839,
+      "learning_rate": 0.001862962962962963,
+      "loss": 1.9938,
+      "step": 3700
+    },
+    {
+      "epoch": 13.74,
+      "grad_norm": 0.42660489678382874,
+      "learning_rate": 0.0018625925925925926,
+      "loss": 2.0147,
+      "step": 3710
+    },
+    {
+      "epoch": 13.78,
+      "grad_norm": 0.3961524963378906,
+      "learning_rate": 0.0018622222222222224,
+      "loss": 2.0363,
+      "step": 3720
+    },
+    {
+      "epoch": 13.81,
+      "grad_norm": 0.3870633542537689,
+      "learning_rate": 0.001861851851851852,
+      "loss": 2.0368,
+      "step": 3730
+    },
+    {
+      "epoch": 13.85,
+      "grad_norm": 0.3615890443325043,
+      "learning_rate": 0.0018614814814814815,
+      "loss": 1.9881,
+      "step": 3740
+    },
+    {
+      "epoch": 13.89,
+      "grad_norm": 0.3907649517059326,
+      "learning_rate": 0.0018611111111111111,
+      "loss": 2.0409,
+      "step": 3750
+    },
+    {
+      "epoch": 13.93,
+      "grad_norm": 0.3881995677947998,
+      "learning_rate": 0.001860740740740741,
+      "loss": 2.045,
+      "step": 3760
+    },
+    {
+      "epoch": 13.96,
+      "grad_norm": 0.4178624451160431,
+      "learning_rate": 0.0018603703703703705,
+      "loss": 2.0016,
+      "step": 3770
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.7995258569717407,
+      "learning_rate": 0.00186,
+      "loss": 2.0585,
+      "step": 3780
+    },
+    {
+      "epoch": 14.04,
+      "grad_norm": 0.4634484648704529,
+      "learning_rate": 0.0018596296296296295,
+      "loss": 1.8123,
+      "step": 3790
+    },
+    {
+      "epoch": 14.07,
+      "grad_norm": 0.4777580499649048,
+      "learning_rate": 0.0018592592592592593,
+      "loss": 1.8001,
+      "step": 3800
+    },
+    {
+      "epoch": 14.11,
+      "grad_norm": 0.43421947956085205,
+      "learning_rate": 0.0018588888888888889,
+      "loss": 1.8164,
+      "step": 3810
+    },
+    {
+      "epoch": 14.15,
+      "grad_norm": 0.4807451665401459,
+      "learning_rate": 0.0018585185185185185,
+      "loss": 1.8032,
+      "step": 3820
+    },
+    {
+      "epoch": 14.19,
+      "grad_norm": 0.4432723820209503,
+      "learning_rate": 0.001858148148148148,
+      "loss": 1.8449,
+      "step": 3830
+    },
+    {
+      "epoch": 14.22,
+      "grad_norm": 0.44446131587028503,
+      "learning_rate": 0.0018577777777777779,
+      "loss": 1.8489,
+      "step": 3840
+    },
+    {
+      "epoch": 14.26,
+      "grad_norm": 0.4247501790523529,
+      "learning_rate": 0.0018574074074074074,
+      "loss": 1.8724,
+      "step": 3850
+    },
+    {
+      "epoch": 14.3,
+      "grad_norm": 0.4298627972602844,
+      "learning_rate": 0.001857037037037037,
+      "loss": 1.879,
+      "step": 3860
+    },
+    {
+      "epoch": 14.33,
+      "grad_norm": 0.4358549416065216,
+      "learning_rate": 0.0018566666666666666,
+      "loss": 1.9109,
+      "step": 3870
+    },
+    {
+      "epoch": 14.37,
+      "grad_norm": 0.4526694715023041,
+      "learning_rate": 0.0018562962962962964,
+      "loss": 1.9652,
+      "step": 3880
+    },
+    {
+      "epoch": 14.41,
+      "grad_norm": 0.4284061789512634,
+      "learning_rate": 0.001855925925925926,
+      "loss": 1.9399,
+      "step": 3890
+    },
+    {
+      "epoch": 14.44,
+      "grad_norm": 0.4858032763004303,
+      "learning_rate": 0.0018555555555555556,
+      "loss": 1.9424,
+      "step": 3900
+    },
+    {
+      "epoch": 14.48,
+      "grad_norm": 0.4293293356895447,
+      "learning_rate": 0.0018551851851851854,
+      "loss": 1.9548,
+      "step": 3910
+    },
+    {
+      "epoch": 14.52,
+      "grad_norm": 0.4032863676548004,
+      "learning_rate": 0.001854814814814815,
+      "loss": 1.9461,
+      "step": 3920
+    },
+    {
+      "epoch": 14.56,
+      "grad_norm": 0.40789926052093506,
+      "learning_rate": 0.0018544444444444444,
+      "loss": 1.9562,
+      "step": 3930
+    },
+    {
+      "epoch": 14.59,
+      "grad_norm": 0.4035639762878418,
+      "learning_rate": 0.001854074074074074,
+      "loss": 1.9913,
+      "step": 3940
+    },
+    {
+      "epoch": 14.63,
+      "grad_norm": 0.4079335629940033,
+      "learning_rate": 0.0018537037037037038,
+      "loss": 1.9682,
+      "step": 3950
+    },
+    {
+      "epoch": 14.67,
+      "grad_norm": 0.4082600772380829,
+      "learning_rate": 0.0018533333333333334,
+      "loss": 1.9547,
+      "step": 3960
+    },
+    {
+      "epoch": 14.7,
+      "grad_norm": 0.4207163453102112,
+      "learning_rate": 0.001852962962962963,
+      "loss": 1.973,
+      "step": 3970
+    },
+    {
+      "epoch": 14.74,
+      "grad_norm": 0.4074302017688751,
+      "learning_rate": 0.0018525925925925925,
+      "loss": 2.0217,
+      "step": 3980
+    },
+    {
+      "epoch": 14.78,
+      "grad_norm": 0.40997204184532166,
+      "learning_rate": 0.0018522222222222223,
+      "loss": 1.9972,
+      "step": 3990
+    },
+    {
+      "epoch": 14.81,
+      "grad_norm": 0.3812522888183594,
+      "learning_rate": 0.001851851851851852,
+      "loss": 1.9702,
+      "step": 4000
+    },
+    {
+      "epoch": 14.85,
+      "grad_norm": 0.42103034257888794,
+      "learning_rate": 0.0018514814814814815,
+      "loss": 2.0201,
+      "step": 4010
+    },
+    {
+      "epoch": 14.89,
+      "grad_norm": 0.40356218814849854,
+      "learning_rate": 0.001851111111111111,
+      "loss": 2.0157,
+      "step": 4020
+    },
+    {
+      "epoch": 14.93,
+      "grad_norm": 0.4146977365016937,
+      "learning_rate": 0.001850740740740741,
+      "loss": 1.9972,
+      "step": 4030
+    },
+    {
+      "epoch": 14.96,
+      "grad_norm": 0.3836473822593689,
+      "learning_rate": 0.0018503703703703705,
+      "loss": 2.0188,
+      "step": 4040
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.8165007829666138,
+      "learning_rate": 0.00185,
+      "loss": 1.9983,
+      "step": 4050
+    },
+    {
+      "epoch": 15.04,
+      "grad_norm": 0.4440079927444458,
+      "learning_rate": 0.0018496296296296297,
+      "loss": 1.7477,
+      "step": 4060
+    },
+    {
+      "epoch": 15.07,
+      "grad_norm": 0.4328083097934723,
+      "learning_rate": 0.0018492592592592593,
+      "loss": 1.7601,
+      "step": 4070
+    },
+    {
+      "epoch": 15.11,
+      "grad_norm": 0.4315027594566345,
+      "learning_rate": 0.0018488888888888888,
+      "loss": 1.7958,
+      "step": 4080
+    },
+    {
+      "epoch": 15.15,
+      "grad_norm": 0.4953678548336029,
+      "learning_rate": 0.0018485185185185184,
+      "loss": 1.8026,
+      "step": 4090
+    },
+    {
+      "epoch": 15.19,
+      "grad_norm": 0.42505890130996704,
+      "learning_rate": 0.0018481481481481482,
+      "loss": 1.8054,
+      "step": 4100
+    },
+    {
+      "epoch": 15.22,
+      "grad_norm": 0.44902315735816956,
+      "learning_rate": 0.0018477777777777778,
+      "loss": 1.8499,
+      "step": 4110
+    },
+    {
+      "epoch": 15.26,
+      "grad_norm": 0.45028263330459595,
+      "learning_rate": 0.0018474074074074074,
+      "loss": 1.8985,
+      "step": 4120
+    },
+    {
+      "epoch": 15.3,
+      "grad_norm": 0.44876259565353394,
+      "learning_rate": 0.001847037037037037,
+      "loss": 1.8558,
+      "step": 4130
+    },
+    {
+      "epoch": 15.33,
+      "grad_norm": 0.43394652009010315,
+      "learning_rate": 0.0018466666666666668,
+      "loss": 1.8795,
+      "step": 4140
+    },
+    {
+      "epoch": 15.37,
+      "grad_norm": 0.445832222700119,
+      "learning_rate": 0.0018462962962962964,
+      "loss": 1.8889,
+      "step": 4150
+    },
+    {
+      "epoch": 15.41,
+      "grad_norm": 0.46300771832466125,
+      "learning_rate": 0.001845925925925926,
+      "loss": 1.924,
+      "step": 4160
+    },
+    {
+      "epoch": 15.44,
+      "grad_norm": 0.4595911204814911,
+      "learning_rate": 0.0018455555555555556,
+      "loss": 1.9116,
+      "step": 4170
+    },
+    {
+      "epoch": 15.48,
+      "grad_norm": 0.4411725103855133,
+      "learning_rate": 0.0018451851851851854,
+      "loss": 1.9208,
+      "step": 4180
+    },
+    {
+      "epoch": 15.52,
+      "grad_norm": 0.4301586151123047,
+      "learning_rate": 0.001844814814814815,
+      "loss": 1.9635,
+      "step": 4190
+    },
+    {
+      "epoch": 15.56,
+      "grad_norm": 0.4307883679866791,
+      "learning_rate": 0.0018444444444444446,
+      "loss": 1.9143,
+      "step": 4200
+    },
+    {
+      "epoch": 15.59,
+      "grad_norm": 0.43274444341659546,
+      "learning_rate": 0.001844074074074074,
+      "loss": 1.943,
+      "step": 4210
+    },
+    {
+      "epoch": 15.63,
+      "grad_norm": 0.40284019708633423,
+      "learning_rate": 0.0018437037037037037,
+      "loss": 1.9672,
+      "step": 4220
+    },
+    {
+      "epoch": 15.67,
+      "grad_norm": 0.41831302642822266,
+      "learning_rate": 0.0018433333333333333,
+      "loss": 1.985,
+      "step": 4230
+    },
+    {
+      "epoch": 15.7,
+      "grad_norm": 0.39616280794143677,
+      "learning_rate": 0.001842962962962963,
+      "loss": 1.9658,
+      "step": 4240
+    },
+    {
+      "epoch": 15.74,
+      "grad_norm": 0.4373377859592438,
+      "learning_rate": 0.0018425925925925925,
+      "loss": 1.992,
+      "step": 4250
+    },
+    {
+      "epoch": 15.78,
+      "grad_norm": 0.43265146017074585,
+      "learning_rate": 0.0018422222222222223,
+      "loss": 1.9469,
+      "step": 4260
+    },
+    {
+      "epoch": 15.81,
+      "grad_norm": 0.4245070815086365,
+      "learning_rate": 0.001841851851851852,
+      "loss": 1.9895,
+      "step": 4270
+    },
+    {
+      "epoch": 15.85,
+      "grad_norm": 0.4061969220638275,
+      "learning_rate": 0.0018414814814814815,
+      "loss": 1.9768,
+      "step": 4280
+    },
+    {
+      "epoch": 15.89,
+      "grad_norm": 0.41055822372436523,
+      "learning_rate": 0.001841111111111111,
+      "loss": 1.9773,
+      "step": 4290
+    },
+    {
+      "epoch": 15.93,
+      "grad_norm": 0.40104153752326965,
+      "learning_rate": 0.0018407407407407409,
+      "loss": 1.9641,
+      "step": 4300
+    },
+    {
+      "epoch": 15.96,
+      "grad_norm": 0.3944847881793976,
+      "learning_rate": 0.0018403703703703705,
+      "loss": 1.9745,
+      "step": 4310
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 0.8410804867744446,
+      "learning_rate": 0.00184,
+      "loss": 1.9942,
+      "step": 4320
+    },
+    {
+      "epoch": 16.04,
+      "grad_norm": 0.4624631106853485,
+      "learning_rate": 0.0018396296296296299,
+      "loss": 1.7683,
+      "step": 4330
+    },
+    {
+      "epoch": 16.07,
+      "grad_norm": 0.43549641966819763,
+      "learning_rate": 0.0018392592592592592,
+      "loss": 1.7841,
+      "step": 4340
+    },
+    {
+      "epoch": 16.11,
+      "grad_norm": 0.43923529982566833,
+      "learning_rate": 0.0018388888888888888,
+      "loss": 1.8057,
+      "step": 4350
+    },
+    {
+      "epoch": 16.15,
+      "grad_norm": 0.4524279832839966,
+      "learning_rate": 0.0018385185185185184,
+      "loss": 1.7818,
+      "step": 4360
+    },
+    {
+      "epoch": 16.19,
+      "grad_norm": 0.4260402023792267,
+      "learning_rate": 0.0018381481481481482,
+      "loss": 1.8019,
+      "step": 4370
+    },
+    {
+      "epoch": 16.22,
+      "grad_norm": 0.47032997012138367,
+      "learning_rate": 0.0018377777777777778,
+      "loss": 1.7883,
+      "step": 4380
+    },
+    {
+      "epoch": 16.26,
+      "grad_norm": 0.4660649597644806,
+      "learning_rate": 0.0018374074074074074,
+      "loss": 1.819,
+      "step": 4390
+    },
+    {
+      "epoch": 16.3,
+      "grad_norm": 0.4598945081233978,
+      "learning_rate": 0.001837037037037037,
+      "loss": 1.8146,
+      "step": 4400
+    },
+    {
+      "epoch": 16.33,
+      "grad_norm": 0.47421982884407043,
+      "learning_rate": 0.0018366666666666668,
+      "loss": 1.8558,
+      "step": 4410
+    },
+    {
+      "epoch": 16.37,
+      "grad_norm": 0.43668022751808167,
+      "learning_rate": 0.0018362962962962964,
+      "loss": 1.903,
+      "step": 4420
+    },
+    {
+      "epoch": 16.41,
+      "grad_norm": 0.42682287096977234,
+      "learning_rate": 0.001835925925925926,
+      "loss": 1.9004,
+      "step": 4430
+    },
+    {
+      "epoch": 16.44,
+      "grad_norm": 0.46274223923683167,
+      "learning_rate": 0.0018355555555555556,
+      "loss": 1.8872,
+      "step": 4440
+    },
+    {
+      "epoch": 16.48,
+      "grad_norm": 0.43078264594078064,
+      "learning_rate": 0.0018351851851851854,
+      "loss": 1.9173,
+      "step": 4450
+    },
+    {
+      "epoch": 16.52,
+      "grad_norm": 0.4077586233615875,
+      "learning_rate": 0.001834814814814815,
+      "loss": 1.938,
+      "step": 4460
+    },
+    {
+      "epoch": 16.56,
+      "grad_norm": 0.42057204246520996,
+      "learning_rate": 0.0018344444444444445,
+      "loss": 1.9126,
+      "step": 4470
+    },
+    {
+      "epoch": 16.59,
+      "grad_norm": 0.4092435836791992,
+      "learning_rate": 0.001834074074074074,
+      "loss": 1.9148,
+      "step": 4480
+    },
+    {
+      "epoch": 16.63,
+      "grad_norm": 0.44807711243629456,
+      "learning_rate": 0.0018337037037037037,
+      "loss": 1.9139,
+      "step": 4490
+    },
+    {
+      "epoch": 16.67,
+      "grad_norm": 0.4302557408809662,
+      "learning_rate": 0.0018333333333333333,
+      "loss": 1.9476,
+      "step": 4500
+    },
+    {
+      "epoch": 16.7,
+      "grad_norm": 0.41777893900871277,
+      "learning_rate": 0.0018329629629629629,
+      "loss": 1.9249,
+      "step": 4510
+    },
+    {
+      "epoch": 16.74,
+      "grad_norm": 0.44393467903137207,
+      "learning_rate": 0.0018325925925925927,
+      "loss": 1.9524,
+      "step": 4520
+    },
+    {
+      "epoch": 16.78,
+      "grad_norm": 0.4212867319583893,
+      "learning_rate": 0.0018322222222222223,
+      "loss": 1.9233,
+      "step": 4530
+    },
+    {
+      "epoch": 16.81,
+      "grad_norm": 0.3991056978702545,
+      "learning_rate": 0.0018318518518518519,
+      "loss": 1.9497,
+      "step": 4540
+    },
+    {
+      "epoch": 16.85,
+      "grad_norm": 0.3818279504776001,
+      "learning_rate": 0.0018314814814814815,
+      "loss": 1.9828,
+      "step": 4550
+    },
+    {
+      "epoch": 16.89,
+      "grad_norm": 0.4082184135913849,
+      "learning_rate": 0.0018311111111111113,
+      "loss": 1.9505,
+      "step": 4560
+    },
+    {
+      "epoch": 16.93,
+      "grad_norm": 0.40178772807121277,
+      "learning_rate": 0.0018307407407407409,
+      "loss": 1.9886,
+      "step": 4570
+    },
+    {
+      "epoch": 16.96,
+      "grad_norm": 0.4248037040233612,
+      "learning_rate": 0.0018303703703703704,
+      "loss": 1.9412,
+      "step": 4580
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 0.8885304927825928,
+      "learning_rate": 0.00183,
+      "loss": 1.9689,
+      "step": 4590
+    },
+    {
+      "epoch": 17.04,
+      "grad_norm": 0.4501792788505554,
+      "learning_rate": 0.0018296296296296298,
+      "loss": 1.7377,
+      "step": 4600
+    },
+    {
+      "epoch": 17.07,
+      "grad_norm": 0.483440637588501,
+      "learning_rate": 0.0018292592592592594,
+      "loss": 1.7509,
+      "step": 4610
+    },
+    {
+      "epoch": 17.11,
+      "grad_norm": 0.46696922183036804,
+      "learning_rate": 0.0018288888888888888,
+      "loss": 1.7556,
+      "step": 4620
+    },
+    {
+      "epoch": 17.15,
+      "grad_norm": 0.4699177145957947,
+      "learning_rate": 0.0018285185185185184,
+      "loss": 1.7667,
+      "step": 4630
+    },
+    {
+      "epoch": 17.19,
+      "grad_norm": 0.47239282727241516,
+      "learning_rate": 0.0018281481481481482,
+      "loss": 1.7972,
+      "step": 4640
+    },
+    {
+      "epoch": 17.22,
+      "grad_norm": 0.47145387530326843,
+      "learning_rate": 0.0018277777777777778,
+      "loss": 1.8257,
+      "step": 4650
+    },
+    {
+      "epoch": 17.26,
+      "grad_norm": 0.47286316752433777,
+      "learning_rate": 0.0018274074074074074,
+      "loss": 1.8362,
+      "step": 4660
+    },
+    {
+      "epoch": 17.3,
+      "grad_norm": 0.4284460246562958,
+      "learning_rate": 0.001827037037037037,
+      "loss": 1.8442,
+      "step": 4670
+    },
+    {
+      "epoch": 17.33,
+      "grad_norm": 0.44590720534324646,
+      "learning_rate": 0.0018266666666666668,
+      "loss": 1.856,
+      "step": 4680
+    },
+    {
+      "epoch": 17.37,
+      "grad_norm": 0.4675563871860504,
+      "learning_rate": 0.0018262962962962963,
+      "loss": 1.8532,
+      "step": 4690
+    },
+    {
+      "epoch": 17.41,
+      "grad_norm": 0.464427649974823,
+      "learning_rate": 0.001825925925925926,
+      "loss": 1.8839,
+      "step": 4700
+    },
+    {
+      "epoch": 17.44,
+      "grad_norm": 0.4539407193660736,
+      "learning_rate": 0.0018255555555555557,
+      "loss": 1.8841,
+      "step": 4710
+    },
+    {
+      "epoch": 17.48,
+      "grad_norm": 0.4451824724674225,
+      "learning_rate": 0.0018251851851851853,
+      "loss": 1.8744,
+      "step": 4720
+    },
+    {
+      "epoch": 17.52,
+      "grad_norm": 0.43316957354545593,
+      "learning_rate": 0.001824814814814815,
+      "loss": 1.9018,
+      "step": 4730
+    },
+    {
+      "epoch": 17.56,
+      "grad_norm": 0.4676300287246704,
+      "learning_rate": 0.0018244444444444445,
+      "loss": 1.892,
+      "step": 4740
+    },
+    {
+      "epoch": 17.59,
+      "grad_norm": 0.46603700518608093,
+      "learning_rate": 0.0018240740740740743,
+      "loss": 1.8955,
+      "step": 4750
+    },
+    {
+      "epoch": 17.63,
+      "grad_norm": 0.4689687192440033,
+      "learning_rate": 0.0018237037037037037,
+      "loss": 1.9119,
+      "step": 4760
+    },
+    {
+      "epoch": 17.67,
+      "grad_norm": 0.4662397801876068,
+      "learning_rate": 0.0018233333333333333,
+      "loss": 1.9195,
+      "step": 4770
+    },
+    {
+      "epoch": 17.7,
+      "grad_norm": 0.4355107545852661,
+      "learning_rate": 0.0018229629629629629,
+      "loss": 1.9403,
+      "step": 4780
+    },
+    {
+      "epoch": 17.74,
+      "grad_norm": 0.41498440504074097,
+      "learning_rate": 0.0018225925925925927,
+      "loss": 1.9187,
+      "step": 4790
+    },
+    {
+      "epoch": 17.78,
+      "grad_norm": 0.44109728932380676,
+      "learning_rate": 0.0018222222222222223,
+      "loss": 1.9468,
+      "step": 4800
+    },
+    {
+      "epoch": 17.81,
+      "grad_norm": 0.4080424904823303,
+      "learning_rate": 0.0018218518518518518,
+      "loss": 1.9261,
+      "step": 4810
+    },
+    {
+      "epoch": 17.85,
+      "grad_norm": 0.4469340145587921,
+      "learning_rate": 0.0018214814814814814,
+      "loss": 1.9199,
+      "step": 4820
+    },
+    {
+      "epoch": 17.89,
+      "grad_norm": 0.42027363181114197,
+      "learning_rate": 0.0018211111111111112,
+      "loss": 1.9378,
+      "step": 4830
+    },
+    {
+      "epoch": 17.93,
+      "grad_norm": 0.42399153113365173,
+      "learning_rate": 0.0018207407407407408,
+      "loss": 1.9417,
+      "step": 4840
+    },
+    {
+      "epoch": 17.96,
+      "grad_norm": 0.46285194158554077,
+      "learning_rate": 0.0018203703703703704,
+      "loss": 1.944,
+      "step": 4850
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 0.9817782640457153,
+      "learning_rate": 0.00182,
+      "loss": 1.93,
+      "step": 4860
+    },
+    {
+      "epoch": 18.04,
+      "grad_norm": 0.48128384351730347,
+      "learning_rate": 0.0018196296296296298,
+      "loss": 1.6866,
+      "step": 4870
+    },
+    {
+      "epoch": 18.07,
+      "grad_norm": 0.4536389112472534,
+      "learning_rate": 0.0018192592592592594,
+      "loss": 1.6788,
+      "step": 4880
+    },
+    {
+      "epoch": 18.11,
+      "grad_norm": 0.4597219228744507,
+      "learning_rate": 0.001818888888888889,
+      "loss": 1.7284,
+      "step": 4890
+    },
+    {
+      "epoch": 18.15,
+      "grad_norm": 0.47937148809432983,
+      "learning_rate": 0.0018185185185185184,
+      "loss": 1.7616,
+      "step": 4900
+    },
+    {
+      "epoch": 18.19,
+      "grad_norm": 0.4905545711517334,
+      "learning_rate": 0.0018181481481481482,
+      "loss": 1.7659,
+      "step": 4910
+    },
+    {
+      "epoch": 18.22,
+      "grad_norm": 0.45469731092453003,
+      "learning_rate": 0.0018177777777777778,
+      "loss": 1.7832,
+      "step": 4920
+    },
+    {
+      "epoch": 18.26,
+      "grad_norm": 0.5039942860603333,
+      "learning_rate": 0.0018174074074074073,
+      "loss": 1.7958,
+      "step": 4930
+    },
+    {
+      "epoch": 18.3,
+      "grad_norm": 0.45666229724884033,
+      "learning_rate": 0.0018170370370370371,
+      "loss": 1.8288,
+      "step": 4940
+    },
+    {
+      "epoch": 18.33,
+      "grad_norm": 0.4742020070552826,
+      "learning_rate": 0.0018166666666666667,
+      "loss": 1.8486,
+      "step": 4950
+    },
+    {
+      "epoch": 18.37,
+      "grad_norm": 0.48586955666542053,
+      "learning_rate": 0.0018162962962962963,
+      "loss": 1.8158,
+      "step": 4960
+    },
+    {
+      "epoch": 18.41,
+      "grad_norm": 0.44601643085479736,
+      "learning_rate": 0.001815925925925926,
+      "loss": 1.8467,
+      "step": 4970
+    },
+    {
+      "epoch": 18.44,
+      "grad_norm": 0.4457637667655945,
+      "learning_rate": 0.0018155555555555557,
+      "loss": 1.8542,
+      "step": 4980
+    },
+    {
+      "epoch": 18.48,
+      "grad_norm": 0.45777076482772827,
+      "learning_rate": 0.0018151851851851853,
+      "loss": 1.8653,
+      "step": 4990
+    },
+    {
+      "epoch": 18.52,
+      "grad_norm": 0.4599011540412903,
+      "learning_rate": 0.001814814814814815,
+      "loss": 1.8432,
+      "step": 5000
+    },
+    {
+      "epoch": 18.56,
+      "grad_norm": 0.4542379379272461,
+      "learning_rate": 0.0018144444444444445,
+      "loss": 1.9042,
+      "step": 5010
+    },
+    {
+      "epoch": 18.59,
+      "grad_norm": 0.4289243817329407,
+      "learning_rate": 0.0018140740740740743,
+      "loss": 1.8927,
+      "step": 5020
+    },
+    {
+      "epoch": 18.63,
+      "grad_norm": 0.4727242887020111,
+      "learning_rate": 0.0018137037037037037,
+      "loss": 1.8896,
+      "step": 5030
+    },
+    {
+      "epoch": 18.67,
+      "grad_norm": 0.49049800634384155,
+      "learning_rate": 0.0018133333333333332,
+      "loss": 1.8991,
+      "step": 5040
+    },
+    {
+      "epoch": 18.7,
+      "grad_norm": 0.42483386397361755,
+      "learning_rate": 0.0018129629629629628,
+      "loss": 1.8808,
+      "step": 5050
+    },
+    {
+      "epoch": 18.74,
+      "grad_norm": 0.44468677043914795,
+      "learning_rate": 0.0018125925925925926,
+      "loss": 1.9129,
+      "step": 5060
+    },
+    {
+      "epoch": 18.78,
+      "grad_norm": 0.42822033166885376,
+      "learning_rate": 0.0018122222222222222,
+      "loss": 1.9334,
+      "step": 5070
+    },
+    {
+      "epoch": 18.81,
+      "grad_norm": 0.45312270522117615,
+      "learning_rate": 0.0018118518518518518,
+      "loss": 1.9465,
+      "step": 5080
+    },
+    {
+      "epoch": 18.85,
+      "grad_norm": 0.4553111493587494,
+      "learning_rate": 0.0018114814814814814,
+      "loss": 1.8999,
+      "step": 5090
+    },
+    {
+      "epoch": 18.89,
+      "grad_norm": 0.44948896765708923,
+      "learning_rate": 0.0018111111111111112,
+      "loss": 1.9649,
+      "step": 5100
+    },
+    {
+      "epoch": 18.93,
+      "grad_norm": 0.4670604467391968,
+      "learning_rate": 0.0018107407407407408,
+      "loss": 1.9572,
+      "step": 5110
+    },
+    {
+      "epoch": 18.96,
+      "grad_norm": 0.43330103158950806,
+      "learning_rate": 0.0018103703703703704,
+      "loss": 1.9322,
+      "step": 5120
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 0.921258807182312,
+      "learning_rate": 0.0018100000000000002,
+      "loss": 1.9589,
+      "step": 5130
+    },
+    {
+      "epoch": 19.04,
+      "grad_norm": 0.5031507611274719,
+      "learning_rate": 0.0018096296296296298,
+      "loss": 1.6821,
+      "step": 5140
+    },
+    {
+      "epoch": 19.07,
+      "grad_norm": 0.48202696442604065,
+      "learning_rate": 0.0018092592592592594,
+      "loss": 1.7146,
+      "step": 5150
+    },
+    {
+      "epoch": 19.11,
+      "grad_norm": 0.4963012933731079,
+      "learning_rate": 0.001808888888888889,
+      "loss": 1.7043,
+      "step": 5160
+    },
+    {
+      "epoch": 19.15,
+      "grad_norm": 0.4779762327671051,
+      "learning_rate": 0.0018085185185185185,
+      "loss": 1.7551,
+      "step": 5170
+    },
+    {
+      "epoch": 19.19,
+      "grad_norm": 0.4379749298095703,
+      "learning_rate": 0.0018081481481481481,
+      "loss": 1.7755,
+      "step": 5180
+    },
+    {
+      "epoch": 19.22,
+      "grad_norm": 0.49266839027404785,
+      "learning_rate": 0.0018077777777777777,
+      "loss": 1.7835,
+      "step": 5190
+    },
+    {
+      "epoch": 19.26,
+      "grad_norm": 0.4939051568508148,
+      "learning_rate": 0.0018074074074074073,
+      "loss": 1.7979,
+      "step": 5200
+    },
+    {
+      "epoch": 19.3,
+      "grad_norm": 0.4597000181674957,
+      "learning_rate": 0.0018070370370370371,
+      "loss": 1.8011,
+      "step": 5210
+    },
+    {
+      "epoch": 19.33,
+      "grad_norm": 0.47228899598121643,
+      "learning_rate": 0.0018066666666666667,
+      "loss": 1.8119,
+      "step": 5220
+    },
+    {
+      "epoch": 19.37,
+      "grad_norm": 0.4497470259666443,
+      "learning_rate": 0.0018062962962962963,
+      "loss": 1.8311,
+      "step": 5230
+    },
+    {
+      "epoch": 19.41,
+      "grad_norm": 0.46324270963668823,
+      "learning_rate": 0.0018059259259259259,
+      "loss": 1.8474,
+      "step": 5240
+    },
+    {
+      "epoch": 19.44,
+      "grad_norm": 0.4588213264942169,
+      "learning_rate": 0.0018055555555555557,
+      "loss": 1.837,
+      "step": 5250
+    },
+    {
+      "epoch": 19.48,
+      "grad_norm": 0.47411489486694336,
+      "learning_rate": 0.0018051851851851853,
+      "loss": 1.8464,
+      "step": 5260
+    },
+    {
+      "epoch": 19.52,
+      "grad_norm": 0.45634156465530396,
+      "learning_rate": 0.0018048148148148149,
+      "loss": 1.8478,
+      "step": 5270
+    },
+    {
+      "epoch": 19.56,
+      "grad_norm": 0.42189350724220276,
+      "learning_rate": 0.0018044444444444445,
+      "loss": 1.8705,
+      "step": 5280
+    },
+    {
+      "epoch": 19.59,
+      "grad_norm": 0.45213887095451355,
+      "learning_rate": 0.0018040740740740743,
+      "loss": 1.8743,
+      "step": 5290
+    },
+    {
+      "epoch": 19.63,
+      "grad_norm": 0.4699808955192566,
+      "learning_rate": 0.0018037037037037038,
+      "loss": 1.8755,
+      "step": 5300
+    },
+    {
+      "epoch": 19.67,
+      "grad_norm": 0.5012890100479126,
+      "learning_rate": 0.0018033333333333332,
+      "loss": 1.8705,
+      "step": 5310
+    },
+    {
+      "epoch": 19.7,
+      "grad_norm": 0.39823225140571594,
+      "learning_rate": 0.0018029629629629628,
+      "loss": 1.8902,
+      "step": 5320
+    },
+    {
+      "epoch": 19.74,
+      "grad_norm": 0.4447478652000427,
+      "learning_rate": 0.0018025925925925926,
+      "loss": 1.8827,
+      "step": 5330
+    },
+    {
+      "epoch": 19.78,
+      "grad_norm": 0.4571211338043213,
+      "learning_rate": 0.0018022222222222222,
+      "loss": 1.8891,
+      "step": 5340
+    },
+    {
+      "epoch": 19.81,
+      "grad_norm": 0.5060418248176575,
+      "learning_rate": 0.0018018518518518518,
+      "loss": 1.894,
+      "step": 5350
+    },
+    {
+      "epoch": 19.85,
+      "grad_norm": 0.44250112771987915,
+      "learning_rate": 0.0018014814814814816,
+      "loss": 1.9192,
+      "step": 5360
+    },
+    {
+      "epoch": 19.89,
+      "grad_norm": 0.48073089122772217,
+      "learning_rate": 0.0018011111111111112,
+      "loss": 1.9161,
+      "step": 5370
+    },
+    {
+      "epoch": 19.93,
+      "grad_norm": 0.44471997022628784,
+      "learning_rate": 0.0018007407407407408,
+      "loss": 1.8936,
+      "step": 5380
+    },
+    {
+      "epoch": 19.96,
+      "grad_norm": 0.44000044465065,
+      "learning_rate": 0.0018003703703703704,
+      "loss": 1.9215,
+      "step": 5390
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 1.005425214767456,
+      "learning_rate": 0.0018000000000000002,
+      "loss": 1.9003,
+      "step": 5400
+    },
+    {
+      "epoch": 20.04,
+      "grad_norm": 0.48888400197029114,
+      "learning_rate": 0.0017996296296296298,
+      "loss": 1.6326,
+      "step": 5410
+    },
+    {
+      "epoch": 20.07,
+      "grad_norm": 0.5140829682350159,
+      "learning_rate": 0.0017992592592592593,
+      "loss": 1.6589,
+      "step": 5420
+    },
+    {
+      "epoch": 20.11,
+      "grad_norm": 0.5162104964256287,
+      "learning_rate": 0.001798888888888889,
+      "loss": 1.7067,
+      "step": 5430
+    },
+    {
+      "epoch": 20.15,
+      "grad_norm": 0.47877031564712524,
+      "learning_rate": 0.0017985185185185187,
+      "loss": 1.728,
+      "step": 5440
+    },
+    {
+      "epoch": 20.19,
+      "grad_norm": 0.5037879347801208,
+      "learning_rate": 0.0017981481481481481,
+      "loss": 1.7498,
+      "step": 5450
+    },
+    {
+      "epoch": 20.22,
+      "grad_norm": 0.483064740896225,
+      "learning_rate": 0.0017977777777777777,
+      "loss": 1.7433,
+      "step": 5460
+    },
+    {
+      "epoch": 20.26,
+      "grad_norm": 0.5271421670913696,
+      "learning_rate": 0.0017974074074074073,
+      "loss": 1.7561,
+      "step": 5470
+    },
+    {
+      "epoch": 20.3,
+      "grad_norm": 0.4639941155910492,
+      "learning_rate": 0.001797037037037037,
+      "loss": 1.805,
+      "step": 5480
+    },
+    {
+      "epoch": 20.33,
+      "grad_norm": 0.47345638275146484,
+      "learning_rate": 0.0017966666666666667,
+      "loss": 1.837,
+      "step": 5490
+    },
+    {
+      "epoch": 20.37,
+      "grad_norm": 0.5081591606140137,
+      "learning_rate": 0.0017962962962962963,
+      "loss": 1.7829,
+      "step": 5500
+    },
+    {
+      "epoch": 20.41,
+      "grad_norm": 0.4264516532421112,
+      "learning_rate": 0.0017959259259259259,
+      "loss": 1.8165,
+      "step": 5510
+    },
+    {
+      "epoch": 20.44,
+      "grad_norm": 0.4831700623035431,
+      "learning_rate": 0.0017955555555555557,
+      "loss": 1.8385,
+      "step": 5520
+    },
+    {
+      "epoch": 20.48,
+      "grad_norm": 0.4971249997615814,
+      "learning_rate": 0.0017951851851851853,
+      "loss": 1.8195,
+      "step": 5530
+    },
+    {
+      "epoch": 20.52,
+      "grad_norm": 0.4875461161136627,
+      "learning_rate": 0.0017948148148148148,
+      "loss": 1.8137,
+      "step": 5540
+    },
+    {
+      "epoch": 20.56,
+      "grad_norm": 0.47468191385269165,
+      "learning_rate": 0.0017944444444444446,
+      "loss": 1.8548,
+      "step": 5550
+    },
+    {
+      "epoch": 20.59,
+      "grad_norm": 0.4697520136833191,
+      "learning_rate": 0.0017940740740740742,
+      "loss": 1.8786,
+      "step": 5560
+    },
+    {
+      "epoch": 20.63,
+      "grad_norm": 0.4431191384792328,
+      "learning_rate": 0.0017937037037037038,
+      "loss": 1.8743,
+      "step": 5570
+    },
+    {
+      "epoch": 20.67,
+      "grad_norm": 0.46886029839515686,
+      "learning_rate": 0.0017933333333333332,
+      "loss": 1.8492,
+      "step": 5580
+    },
+    {
+      "epoch": 20.7,
+      "grad_norm": 0.46815305948257446,
+      "learning_rate": 0.001792962962962963,
+      "loss": 1.8523,
+      "step": 5590
+    },
+    {
+      "epoch": 20.74,
+      "grad_norm": 0.42444488406181335,
+      "learning_rate": 0.0017925925925925926,
+      "loss": 1.894,
+      "step": 5600
+    },
+    {
+      "epoch": 20.78,
+      "grad_norm": 0.4468463361263275,
+      "learning_rate": 0.0017922222222222222,
+      "loss": 1.8849,
+      "step": 5610
+    },
+    {
+      "epoch": 20.81,
+      "grad_norm": 0.4831134080886841,
+      "learning_rate": 0.0017918518518518518,
+      "loss": 1.917,
+      "step": 5620
+    },
+    {
+      "epoch": 20.85,
+      "grad_norm": 0.4968695342540741,
+      "learning_rate": 0.0017914814814814816,
+      "loss": 1.8697,
+      "step": 5630
+    },
+    {
+      "epoch": 20.89,
+      "grad_norm": 0.4390285015106201,
+      "learning_rate": 0.0017911111111111112,
+      "loss": 1.8904,
+      "step": 5640
+    },
+    {
+      "epoch": 20.93,
+      "grad_norm": 0.47561535239219666,
+      "learning_rate": 0.0017907407407407407,
+      "loss": 1.8959,
+      "step": 5650
+    },
+    {
+      "epoch": 20.96,
+      "grad_norm": 0.45114925503730774,
+      "learning_rate": 0.0017903703703703703,
+      "loss": 1.895,
+      "step": 5660
+    },
+    {
+      "epoch": 21.0,
+      "grad_norm": 0.8482184410095215,
+      "learning_rate": 0.0017900000000000001,
+      "loss": 1.9381,
+      "step": 5670
+    },
+    {
+      "epoch": 21.04,
+      "grad_norm": 0.4956987202167511,
+      "learning_rate": 0.0017896296296296297,
+      "loss": 1.6155,
+      "step": 5680
+    },
+    {
+      "epoch": 21.07,
+      "grad_norm": 0.5676151514053345,
+      "learning_rate": 0.0017892592592592593,
+      "loss": 1.6748,
+      "step": 5690
+    },
+    {
+      "epoch": 21.11,
+      "grad_norm": 0.4920210838317871,
+      "learning_rate": 0.001788888888888889,
+      "loss": 1.7132,
+      "step": 5700
+    },
+    {
+      "epoch": 21.15,
+      "grad_norm": 0.5253728628158569,
+      "learning_rate": 0.0017885185185185187,
+      "loss": 1.6675,
+      "step": 5710
+    },
+    {
+      "epoch": 21.19,
+      "grad_norm": 0.5597649812698364,
+      "learning_rate": 0.001788148148148148,
+      "loss": 1.7294,
+      "step": 5720
+    },
+    {
+      "epoch": 21.22,
+      "grad_norm": 0.5059242248535156,
+      "learning_rate": 0.0017877777777777777,
+      "loss": 1.7161,
+      "step": 5730
+    },
+    {
+      "epoch": 21.26,
+      "grad_norm": 0.5013036131858826,
+      "learning_rate": 0.0017874074074074075,
+      "loss": 1.7621,
+      "step": 5740
+    },
+    {
+      "epoch": 21.3,
+      "grad_norm": 0.4777660071849823,
+      "learning_rate": 0.001787037037037037,
+      "loss": 1.761,
+      "step": 5750
+    },
+    {
+      "epoch": 21.33,
+      "grad_norm": 0.5135553479194641,
+      "learning_rate": 0.0017866666666666667,
+      "loss": 1.7918,
+      "step": 5760
+    },
+    {
+      "epoch": 21.37,
+      "grad_norm": 0.48970091342926025,
+      "learning_rate": 0.0017862962962962962,
+      "loss": 1.7895,
+      "step": 5770
+    },
+    {
+      "epoch": 21.41,
+      "grad_norm": 0.49350282549858093,
+      "learning_rate": 0.001785925925925926,
+      "loss": 1.8056,
+      "step": 5780
+    },
+    {
+      "epoch": 21.44,
+      "grad_norm": 0.49979352951049805,
+      "learning_rate": 0.0017855555555555556,
+      "loss": 1.78,
+      "step": 5790
+    },
+    {
+      "epoch": 21.48,
+      "grad_norm": 0.4719812273979187,
+      "learning_rate": 0.0017851851851851852,
+      "loss": 1.8209,
+      "step": 5800
+    },
+    {
+      "epoch": 21.52,
+      "grad_norm": 0.48076552152633667,
+      "learning_rate": 0.0017848148148148148,
+      "loss": 1.7865,
+      "step": 5810
+    },
+    {
+      "epoch": 21.56,
+      "grad_norm": 0.4831288754940033,
+      "learning_rate": 0.0017844444444444446,
+      "loss": 1.8441,
+      "step": 5820
+    },
+    {
+      "epoch": 21.59,
+      "grad_norm": 0.46618470549583435,
+      "learning_rate": 0.0017840740740740742,
+      "loss": 1.8559,
+      "step": 5830
+    },
+    {
+      "epoch": 21.63,
+      "grad_norm": 0.47354695200920105,
+      "learning_rate": 0.0017837037037037038,
+      "loss": 1.8284,
+      "step": 5840
+    },
+    {
+      "epoch": 21.67,
+      "grad_norm": 0.464826762676239,
+      "learning_rate": 0.0017833333333333334,
+      "loss": 1.8633,
+      "step": 5850
+    },
+    {
+      "epoch": 21.7,
+      "grad_norm": 0.43753430247306824,
+      "learning_rate": 0.001782962962962963,
+      "loss": 1.8726,
+      "step": 5860
+    },
+    {
+      "epoch": 21.74,
+      "grad_norm": 0.4837232232093811,
+      "learning_rate": 0.0017825925925925926,
+      "loss": 1.8374,
+      "step": 5870
+    },
+    {
+      "epoch": 21.78,
+      "grad_norm": 0.47672173380851746,
+      "learning_rate": 0.0017822222222222222,
+      "loss": 1.8707,
+      "step": 5880
+    },
+    {
+      "epoch": 21.81,
+      "grad_norm": 0.48062843084335327,
+      "learning_rate": 0.0017818518518518517,
+      "loss": 1.891,
+      "step": 5890
+    },
+    {
+      "epoch": 21.85,
+      "grad_norm": 0.47505030035972595,
+      "learning_rate": 0.0017814814814814815,
+      "loss": 1.8794,
+      "step": 5900
+    },
+    {
+      "epoch": 21.89,
+      "grad_norm": 0.48005613684654236,
+      "learning_rate": 0.0017811111111111111,
+      "loss": 1.8771,
+      "step": 5910
+    },
+    {
+      "epoch": 21.93,
+      "grad_norm": 0.475238561630249,
+      "learning_rate": 0.0017807407407407407,
+      "loss": 1.8643,
+      "step": 5920
+    },
+    {
+      "epoch": 21.96,
+      "grad_norm": 0.44499918818473816,
+      "learning_rate": 0.0017803703703703703,
+      "loss": 1.9119,
+      "step": 5930
+    },
+    {
+      "epoch": 22.0,
+      "grad_norm": 0.859517514705658,
+      "learning_rate": 0.0017800000000000001,
+      "loss": 1.9101,
+      "step": 5940
+    },
+    {
+      "epoch": 22.04,
+      "grad_norm": 0.47651925683021545,
+      "learning_rate": 0.0017796296296296297,
+      "loss": 1.6509,
+      "step": 5950
+    },
+    {
+      "epoch": 22.07,
+      "grad_norm": 0.5325976014137268,
+      "learning_rate": 0.0017792592592592593,
+      "loss": 1.6124,
+      "step": 5960
+    },
+    {
+      "epoch": 22.11,
+      "grad_norm": 0.5133865475654602,
+      "learning_rate": 0.001778888888888889,
+      "loss": 1.6649,
+      "step": 5970
+    },
+    {
+      "epoch": 22.15,
+      "grad_norm": 0.44351667165756226,
+      "learning_rate": 0.0017785185185185187,
+      "loss": 1.6458,
+      "step": 5980
+    },
+    {
+      "epoch": 22.19,
+      "grad_norm": 0.5221470594406128,
+      "learning_rate": 0.0017781481481481483,
+      "loss": 1.6832,
+      "step": 5990
+    },
+    {
+      "epoch": 22.22,
+      "grad_norm": 0.5310438275337219,
+      "learning_rate": 0.0017777777777777776,
+      "loss": 1.745,
+      "step": 6000
+    },
+    {
+      "epoch": 22.26,
+      "grad_norm": 0.5029959082603455,
+      "learning_rate": 0.0017774074074074075,
+      "loss": 1.7047,
+      "step": 6010
+    },
+    {
+      "epoch": 22.3,
+      "grad_norm": 0.5572134852409363,
+      "learning_rate": 0.001777037037037037,
+      "loss": 1.7418,
+      "step": 6020
+    },
+    {
+      "epoch": 22.33,
+      "grad_norm": 0.46236780285835266,
+      "learning_rate": 0.0017766666666666666,
+      "loss": 1.7839,
+      "step": 6030
+    },
+    {
+      "epoch": 22.37,
+      "grad_norm": 0.4862767457962036,
+      "learning_rate": 0.0017762962962962962,
+      "loss": 1.7749,
+      "step": 6040
+    },
+    {
+      "epoch": 22.41,
+      "grad_norm": 0.46328261494636536,
+      "learning_rate": 0.001775925925925926,
+      "loss": 1.7915,
+      "step": 6050
+    },
+    {
+      "epoch": 22.44,
+      "grad_norm": 0.5205667614936829,
+      "learning_rate": 0.0017755555555555556,
+      "loss": 1.8303,
+      "step": 6060
+    },
+    {
+      "epoch": 22.48,
+      "grad_norm": 0.525134801864624,
+      "learning_rate": 0.0017751851851851852,
+      "loss": 1.7764,
+      "step": 6070
+    },
+    {
+      "epoch": 22.52,
+      "grad_norm": 0.5078030824661255,
+      "learning_rate": 0.0017748148148148148,
+      "loss": 1.8026,
+      "step": 6080
+    },
+    {
+      "epoch": 22.56,
+      "grad_norm": 0.4789275527000427,
+      "learning_rate": 0.0017744444444444446,
+      "loss": 1.8189,
+      "step": 6090
+    },
+    {
+      "epoch": 22.59,
+      "grad_norm": 0.46829742193222046,
+      "learning_rate": 0.0017740740740740742,
+      "loss": 1.8302,
+      "step": 6100
+    },
+    {
+      "epoch": 22.63,
+      "grad_norm": 0.4511633813381195,
+      "learning_rate": 0.0017737037037037038,
+      "loss": 1.8382,
+      "step": 6110
+    },
+    {
+      "epoch": 22.67,
+      "grad_norm": 0.4943828284740448,
+      "learning_rate": 0.0017733333333333334,
+      "loss": 1.8424,
+      "step": 6120
+    },
+    {
+      "epoch": 22.7,
+      "grad_norm": 0.44993260502815247,
+      "learning_rate": 0.0017729629629629632,
+      "loss": 1.8131,
+      "step": 6130
+    },
+    {
+      "epoch": 22.74,
+      "grad_norm": 0.49377501010894775,
+      "learning_rate": 0.0017725925925925925,
+      "loss": 1.8772,
+      "step": 6140
+    },
+    {
+      "epoch": 22.78,
+      "grad_norm": 0.46665674448013306,
+      "learning_rate": 0.0017722222222222221,
+      "loss": 1.8264,
+      "step": 6150
+    },
+    {
+      "epoch": 22.81,
+      "grad_norm": 0.46312686800956726,
+      "learning_rate": 0.001771851851851852,
+      "loss": 1.8625,
+      "step": 6160
+    },
+    {
+      "epoch": 22.85,
+      "grad_norm": 0.4879915714263916,
+      "learning_rate": 0.0017714814814814815,
+      "loss": 1.8923,
+      "step": 6170
+    },
+    {
+      "epoch": 22.89,
+      "grad_norm": 0.4814596474170685,
+      "learning_rate": 0.001771111111111111,
+      "loss": 1.878,
+      "step": 6180
+    },
+    {
+      "epoch": 22.93,
+      "grad_norm": 0.4914632737636566,
+      "learning_rate": 0.0017707407407407407,
+      "loss": 1.8592,
+      "step": 6190
+    },
+    {
+      "epoch": 22.96,
+      "grad_norm": 0.4641704559326172,
+      "learning_rate": 0.0017703703703703705,
+      "loss": 1.8921,
+      "step": 6200
+    },
+    {
+      "epoch": 23.0,
+      "grad_norm": 0.7784886360168457,
+      "learning_rate": 0.00177,
+      "loss": 1.8705,
+      "step": 6210
+    },
+    {
+      "epoch": 23.04,
+      "grad_norm": 0.5497826933860779,
+      "learning_rate": 0.0017696296296296297,
+      "loss": 1.6352,
+      "step": 6220
+    },
+    {
+      "epoch": 23.07,
+      "grad_norm": 0.5098663568496704,
+      "learning_rate": 0.0017692592592592593,
+      "loss": 1.629,
+      "step": 6230
+    },
+    {
+      "epoch": 23.11,
+      "grad_norm": 0.5080466270446777,
+      "learning_rate": 0.001768888888888889,
+      "loss": 1.6593,
+      "step": 6240
+    },
+    {
+      "epoch": 23.15,
+      "grad_norm": 0.4842182993888855,
+      "learning_rate": 0.0017685185185185187,
+      "loss": 1.6483,
+      "step": 6250
+    },
+    {
+      "epoch": 23.19,
+      "grad_norm": 0.5256596207618713,
+      "learning_rate": 0.0017681481481481482,
+      "loss": 1.6967,
+      "step": 6260
+    },
+    {
+      "epoch": 23.22,
+      "grad_norm": 0.5058240294456482,
+      "learning_rate": 0.0017677777777777776,
+      "loss": 1.7019,
+      "step": 6270
+    },
+    {
+      "epoch": 23.26,
+      "grad_norm": 0.5488624572753906,
+      "learning_rate": 0.0017674074074074074,
+      "loss": 1.7202,
+      "step": 6280
+    },
+    {
+      "epoch": 23.3,
+      "grad_norm": 0.5268415212631226,
+      "learning_rate": 0.001767037037037037,
+      "loss": 1.7774,
+      "step": 6290
+    },
+    {
+      "epoch": 23.33,
+      "grad_norm": 0.48284679651260376,
+      "learning_rate": 0.0017666666666666666,
+      "loss": 1.7642,
+      "step": 6300
+    },
+    {
+      "epoch": 23.37,
+      "grad_norm": 0.5055042505264282,
+      "learning_rate": 0.0017662962962962962,
+      "loss": 1.7453,
+      "step": 6310
+    },
+    {
+      "epoch": 23.41,
+      "grad_norm": 0.4818604588508606,
+      "learning_rate": 0.001765925925925926,
+      "loss": 1.7657,
+      "step": 6320
+    },
+    {
+      "epoch": 23.44,
+      "grad_norm": 0.46134889125823975,
+      "learning_rate": 0.0017655555555555556,
+      "loss": 1.7814,
+      "step": 6330
+    },
+    {
+      "epoch": 23.48,
+      "grad_norm": 0.4772760570049286,
+      "learning_rate": 0.0017651851851851852,
+      "loss": 1.7934,
+      "step": 6340
+    },
+    {
+      "epoch": 23.52,
+      "grad_norm": 0.4994891583919525,
+      "learning_rate": 0.001764814814814815,
+      "loss": 1.8017,
+      "step": 6350
+    },
+    {
+      "epoch": 23.56,
+      "grad_norm": 0.4992477297782898,
+      "learning_rate": 0.0017644444444444446,
+      "loss": 1.792,
+      "step": 6360
+    },
+    {
+      "epoch": 23.59,
+      "grad_norm": 0.47414326667785645,
+      "learning_rate": 0.0017640740740740742,
+      "loss": 1.8296,
+      "step": 6370
+    },
+    {
+      "epoch": 23.63,
+      "grad_norm": 0.500939130783081,
+      "learning_rate": 0.0017637037037037037,
+      "loss": 1.8026,
+      "step": 6380
+    },
+    {
+      "epoch": 23.67,
+      "grad_norm": 0.5158637166023254,
+      "learning_rate": 0.0017633333333333335,
+      "loss": 1.8275,
+      "step": 6390
+    },
+    {
+      "epoch": 23.7,
+      "grad_norm": 0.5043916702270508,
+      "learning_rate": 0.0017629629629629631,
+      "loss": 1.8181,
+      "step": 6400
+    },
+    {
+      "epoch": 23.74,
+      "grad_norm": 0.502905547618866,
+      "learning_rate": 0.0017625925925925925,
+      "loss": 1.8219,
+      "step": 6410
+    },
+    {
+      "epoch": 23.78,
+      "grad_norm": 0.4975557029247284,
+      "learning_rate": 0.001762222222222222,
+      "loss": 1.845,
+      "step": 6420
+    },
+    {
+      "epoch": 23.81,
+      "grad_norm": 0.4911981225013733,
+      "learning_rate": 0.001761851851851852,
+      "loss": 1.8564,
+      "step": 6430
+    },
+    {
+      "epoch": 23.85,
+      "grad_norm": 0.4847852289676666,
+      "learning_rate": 0.0017614814814814815,
+      "loss": 1.8248,
+      "step": 6440
+    },
+    {
+      "epoch": 23.89,
+      "grad_norm": 0.4612410366535187,
+      "learning_rate": 0.001761111111111111,
+      "loss": 1.8405,
+      "step": 6450
+    },
+    {
+      "epoch": 23.93,
+      "grad_norm": 0.5118724703788757,
+      "learning_rate": 0.0017607407407407407,
+      "loss": 1.8545,
+      "step": 6460
+    },
+    {
+      "epoch": 23.96,
+      "grad_norm": 0.46121078729629517,
+      "learning_rate": 0.0017603703703703705,
+      "loss": 1.8519,
+      "step": 6470
+    },
+    {
+      "epoch": 24.0,
+      "grad_norm": 0.771907389163971,
+      "learning_rate": 0.00176,
+      "loss": 1.8582,
+      "step": 6480
+    },
+    {
+      "epoch": 24.04,
+      "grad_norm": 0.5113037824630737,
+      "learning_rate": 0.0017596296296296297,
+      "loss": 1.588,
+      "step": 6490
+    },
+    {
+      "epoch": 24.07,
+      "grad_norm": 0.47875529527664185,
+      "learning_rate": 0.0017592592592592592,
+      "loss": 1.6267,
+      "step": 6500
+    },
+    {
+      "epoch": 24.11,
+      "grad_norm": 0.506752610206604,
+      "learning_rate": 0.001758888888888889,
+      "loss": 1.6279,
+      "step": 6510
+    },
+    {
+      "epoch": 24.15,
+      "grad_norm": 0.5407618284225464,
+      "learning_rate": 0.0017585185185185186,
+      "loss": 1.63,
+      "step": 6520
+    },
+    {
+      "epoch": 24.19,
+      "grad_norm": 0.5361552238464355,
+      "learning_rate": 0.0017581481481481482,
+      "loss": 1.6463,
+      "step": 6530
+    },
+    {
+      "epoch": 24.22,
+      "grad_norm": 0.5410870909690857,
+      "learning_rate": 0.001757777777777778,
+      "loss": 1.6656,
+      "step": 6540
+    },
+    {
+      "epoch": 24.26,
+      "grad_norm": 0.5300707817077637,
+      "learning_rate": 0.0017574074074074074,
+      "loss": 1.7206,
+      "step": 6550
+    },
+    {
+      "epoch": 24.3,
+      "grad_norm": 0.5036062598228455,
+      "learning_rate": 0.001757037037037037,
+      "loss": 1.7108,
+      "step": 6560
+    },
+    {
+      "epoch": 24.33,
+      "grad_norm": 0.47783538699150085,
+      "learning_rate": 0.0017566666666666666,
+      "loss": 1.735,
+      "step": 6570
+    },
+    {
+      "epoch": 24.37,
+      "grad_norm": 0.5215782523155212,
+      "learning_rate": 0.0017562962962962964,
+      "loss": 1.7043,
+      "step": 6580
+    },
+    {
+      "epoch": 24.41,
+      "grad_norm": 0.5307631492614746,
+      "learning_rate": 0.001755925925925926,
+      "loss": 1.7816,
+      "step": 6590
+    },
+    {
+      "epoch": 24.44,
+      "grad_norm": 0.5181083083152771,
+      "learning_rate": 0.0017555555555555556,
+      "loss": 1.7592,
+      "step": 6600
+    },
+    {
+      "epoch": 24.48,
+      "grad_norm": 0.521377444267273,
+      "learning_rate": 0.0017551851851851851,
+      "loss": 1.8097,
+      "step": 6610
+    },
+    {
+      "epoch": 24.52,
+      "grad_norm": 0.5385915040969849,
+      "learning_rate": 0.001754814814814815,
+      "loss": 1.7616,
+      "step": 6620
+    },
+    {
+      "epoch": 24.56,
+      "grad_norm": 0.5241475105285645,
+      "learning_rate": 0.0017544444444444445,
+      "loss": 1.7928,
+      "step": 6630
+    },
+    {
+      "epoch": 24.59,
+      "grad_norm": 0.5101531147956848,
+      "learning_rate": 0.0017540740740740741,
+      "loss": 1.7963,
+      "step": 6640
+    },
+    {
+      "epoch": 24.63,
+      "grad_norm": 0.5142462849617004,
+      "learning_rate": 0.0017537037037037037,
+      "loss": 1.803,
+      "step": 6650
+    },
+    {
+      "epoch": 24.67,
+      "grad_norm": 0.5254778265953064,
+      "learning_rate": 0.0017533333333333335,
+      "loss": 1.8004,
+      "step": 6660
+    },
+    {
+      "epoch": 24.7,
+      "grad_norm": 0.5626793503761292,
+      "learning_rate": 0.0017529629629629631,
+      "loss": 1.8089,
+      "step": 6670
+    },
+    {
+      "epoch": 24.74,
+      "grad_norm": 0.5398451089859009,
+      "learning_rate": 0.0017525925925925927,
+      "loss": 1.8282,
+      "step": 6680
+    },
+    {
+      "epoch": 24.78,
+      "grad_norm": 0.5404179096221924,
+      "learning_rate": 0.001752222222222222,
+      "loss": 1.8205,
+      "step": 6690
+    },
+    {
+      "epoch": 24.81,
+      "grad_norm": 0.44393420219421387,
+      "learning_rate": 0.0017518518518518519,
+      "loss": 1.8549,
+      "step": 6700
+    },
+    {
+      "epoch": 24.85,
+      "grad_norm": 0.5345913767814636,
+      "learning_rate": 0.0017514814814814815,
+      "loss": 1.8387,
+      "step": 6710
+    },
+    {
+      "epoch": 24.89,
+      "grad_norm": 0.4756942689418793,
+      "learning_rate": 0.001751111111111111,
+      "loss": 1.8561,
+      "step": 6720
+    },
+    {
+      "epoch": 24.93,
+      "grad_norm": 0.4690358340740204,
+      "learning_rate": 0.0017507407407407406,
+      "loss": 1.8477,
+      "step": 6730
+    },
+    {
+      "epoch": 24.96,
+      "grad_norm": 0.4660181701183319,
+      "learning_rate": 0.0017503703703703704,
+      "loss": 1.8604,
+      "step": 6740
+    },
+    {
+      "epoch": 25.0,
+      "grad_norm": 1.1090754270553589,
+      "learning_rate": 0.00175,
+      "loss": 1.8425,
+      "step": 6750
+    },
+    {
+      "epoch": 25.04,
+      "grad_norm": 0.532478392124176,
+      "learning_rate": 0.0017496296296296296,
+      "loss": 1.5765,
+      "step": 6760
+    },
+    {
+      "epoch": 25.07,
+      "grad_norm": 0.5390456318855286,
+      "learning_rate": 0.0017492592592592594,
+      "loss": 1.6065,
+      "step": 6770
+    },
+    {
+      "epoch": 25.11,
+      "grad_norm": 0.5295791029930115,
+      "learning_rate": 0.001748888888888889,
+      "loss": 1.6317,
+      "step": 6780
+    },
+    {
+      "epoch": 25.15,
+      "grad_norm": 0.5816089510917664,
+      "learning_rate": 0.0017485185185185186,
+      "loss": 1.6569,
+      "step": 6790
+    },
+    {
+      "epoch": 25.19,
+      "grad_norm": 0.5486976504325867,
+      "learning_rate": 0.0017481481481481482,
+      "loss": 1.6866,
+      "step": 6800
+    },
+    {
+      "epoch": 25.22,
+      "grad_norm": 0.5331709384918213,
+      "learning_rate": 0.001747777777777778,
+      "loss": 1.7122,
+      "step": 6810
+    },
+    {
+      "epoch": 25.26,
+      "grad_norm": 0.46710267663002014,
+      "learning_rate": 0.0017474074074074074,
+      "loss": 1.6933,
+      "step": 6820
+    },
+    {
+      "epoch": 25.3,
+      "grad_norm": 0.526747465133667,
+      "learning_rate": 0.001747037037037037,
+      "loss": 1.7145,
+      "step": 6830
+    },
+    {
+      "epoch": 25.33,
+      "grad_norm": 0.5332999229431152,
+      "learning_rate": 0.0017466666666666665,
+      "loss": 1.6894,
+      "step": 6840
+    },
+    {
+      "epoch": 25.37,
+      "grad_norm": 0.5216143131256104,
+      "learning_rate": 0.0017462962962962964,
+      "loss": 1.7418,
+      "step": 6850
+    },
+    {
+      "epoch": 25.41,
+      "grad_norm": 0.5023478865623474,
+      "learning_rate": 0.001745925925925926,
+      "loss": 1.737,
+      "step": 6860
+    },
+    {
+      "epoch": 25.44,
+      "grad_norm": 0.48997578024864197,
+      "learning_rate": 0.0017455555555555555,
+      "loss": 1.744,
+      "step": 6870
+    },
+    {
+      "epoch": 25.48,
+      "grad_norm": 0.49648550152778625,
+      "learning_rate": 0.0017451851851851851,
+      "loss": 1.7504,
+      "step": 6880
+    },
+    {
+      "epoch": 25.52,
+      "grad_norm": 0.5312948822975159,
+      "learning_rate": 0.001744814814814815,
+      "loss": 1.7733,
+      "step": 6890
+    },
+    {
+      "epoch": 25.56,
+      "grad_norm": 0.5200966596603394,
+      "learning_rate": 0.0017444444444444445,
+      "loss": 1.758,
+      "step": 6900
+    },
+    {
+      "epoch": 25.59,
+      "grad_norm": 0.519443929195404,
+      "learning_rate": 0.001744074074074074,
+      "loss": 1.7788,
+      "step": 6910
+    },
+    {
+      "epoch": 25.63,
+      "grad_norm": 0.5155293345451355,
+      "learning_rate": 0.0017437037037037037,
+      "loss": 1.7849,
+      "step": 6920
+    },
+    {
+      "epoch": 25.67,
+      "grad_norm": 0.4716353714466095,
+      "learning_rate": 0.0017433333333333335,
+      "loss": 1.8222,
+      "step": 6930
+    },
+    {
+      "epoch": 25.7,
+      "grad_norm": 0.4919406771659851,
+      "learning_rate": 0.001742962962962963,
+      "loss": 1.8008,
+      "step": 6940
+    },
+    {
+      "epoch": 25.74,
+      "grad_norm": 0.5622554421424866,
+      "learning_rate": 0.0017425925925925927,
+      "loss": 1.8081,
+      "step": 6950
+    },
+    {
+      "epoch": 25.78,
+      "grad_norm": 0.4926491677761078,
+      "learning_rate": 0.001742222222222222,
+      "loss": 1.8174,
+      "step": 6960
+    },
+    {
+      "epoch": 25.81,
+      "grad_norm": 0.4717676043510437,
+      "learning_rate": 0.0017418518518518519,
+      "loss": 1.8291,
+      "step": 6970
+    },
+    {
+      "epoch": 25.85,
+      "grad_norm": 0.4829355776309967,
+      "learning_rate": 0.0017414814814814814,
+      "loss": 1.8098,
+      "step": 6980
+    },
+    {
+      "epoch": 25.89,
+      "grad_norm": 0.5100350379943848,
+      "learning_rate": 0.001741111111111111,
+      "loss": 1.8398,
+      "step": 6990
+    },
+    {
+      "epoch": 25.93,
+      "grad_norm": 0.504905104637146,
+      "learning_rate": 0.0017407407407407408,
+      "loss": 1.8189,
+      "step": 7000
+    },
+    {
+      "epoch": 25.96,
+      "grad_norm": 0.5000597834587097,
+      "learning_rate": 0.0017403703703703704,
+      "loss": 1.8175,
+      "step": 7010
+    },
+    {
+      "epoch": 26.0,
+      "grad_norm": 0.7448571920394897,
+      "learning_rate": 0.00174,
+      "loss": 1.86,
+      "step": 7020
+    },
+    {
+      "epoch": 26.04,
+      "grad_norm": 0.504486620426178,
+      "learning_rate": 0.0017396296296296296,
+      "loss": 1.5784,
+      "step": 7030
+    },
+    {
+      "epoch": 26.07,
+      "grad_norm": 0.5133050084114075,
+      "learning_rate": 0.0017392592592592594,
+      "loss": 1.5856,
+      "step": 7040
+    },
+    {
+      "epoch": 26.11,
+      "grad_norm": 0.49264371395111084,
+      "learning_rate": 0.001738888888888889,
+      "loss": 1.5844,
+      "step": 7050
+    },
+    {
+      "epoch": 26.15,
+      "grad_norm": 0.5140932202339172,
+      "learning_rate": 0.0017385185185185186,
+      "loss": 1.6197,
+      "step": 7060
+    },
+    {
+      "epoch": 26.19,
+      "grad_norm": 0.5410853624343872,
+      "learning_rate": 0.0017381481481481482,
+      "loss": 1.6744,
+      "step": 7070
+    },
+    {
+      "epoch": 26.22,
+      "grad_norm": 0.636168897151947,
+      "learning_rate": 0.001737777777777778,
+      "loss": 1.6568,
+      "step": 7080
+    },
+    {
+      "epoch": 26.26,
+      "grad_norm": 0.5204961895942688,
+      "learning_rate": 0.0017374074074074076,
+      "loss": 1.6965,
+      "step": 7090
+    },
+    {
+      "epoch": 26.3,
+      "grad_norm": 0.512642502784729,
+      "learning_rate": 0.001737037037037037,
+      "loss": 1.691,
+      "step": 7100
+    },
+    {
+      "epoch": 26.33,
+      "grad_norm": 0.5050673484802246,
+      "learning_rate": 0.0017366666666666665,
+      "loss": 1.6974,
+      "step": 7110
+    },
+    {
+      "epoch": 26.37,
+      "grad_norm": 0.5110860466957092,
+      "learning_rate": 0.0017362962962962963,
+      "loss": 1.7239,
+      "step": 7120
+    },
+    {
+      "epoch": 26.41,
+      "grad_norm": 0.5326704382896423,
+      "learning_rate": 0.001735925925925926,
+      "loss": 1.7303,
+      "step": 7130
+    },
+    {
+      "epoch": 26.44,
+      "grad_norm": 0.5366049408912659,
+      "learning_rate": 0.0017355555555555555,
+      "loss": 1.7422,
+      "step": 7140
+    },
+    {
+      "epoch": 26.48,
+      "grad_norm": 0.5421983599662781,
+      "learning_rate": 0.001735185185185185,
+      "loss": 1.7498,
+      "step": 7150
+    },
+    {
+      "epoch": 26.52,
+      "grad_norm": 0.5443106889724731,
+      "learning_rate": 0.001734814814814815,
+      "loss": 1.7636,
+      "step": 7160
+    },
+    {
+      "epoch": 26.56,
+      "grad_norm": 0.5099042654037476,
+      "learning_rate": 0.0017344444444444445,
+      "loss": 1.7464,
+      "step": 7170
+    },
+    {
+      "epoch": 26.59,
+      "grad_norm": 0.5240099430084229,
+      "learning_rate": 0.001734074074074074,
+      "loss": 1.7317,
+      "step": 7180
+    },
+    {
+      "epoch": 26.63,
+      "grad_norm": 0.4959777891635895,
+      "learning_rate": 0.0017337037037037039,
+      "loss": 1.7725,
+      "step": 7190
+    },
+    {
+      "epoch": 26.67,
+      "grad_norm": 0.5010444521903992,
+      "learning_rate": 0.0017333333333333335,
+      "loss": 1.7842,
+      "step": 7200
+    },
+    {
+      "epoch": 26.7,
+      "grad_norm": 0.545404314994812,
+      "learning_rate": 0.001732962962962963,
+      "loss": 1.8179,
+      "step": 7210
+    },
+    {
+      "epoch": 26.74,
+      "grad_norm": 0.5156935453414917,
+      "learning_rate": 0.0017325925925925926,
+      "loss": 1.7742,
+      "step": 7220
+    },
+    {
+      "epoch": 26.78,
+      "grad_norm": 0.518345057964325,
+      "learning_rate": 0.0017322222222222225,
+      "loss": 1.8186,
+      "step": 7230
+    },
+    {
+      "epoch": 26.81,
+      "grad_norm": 0.5182226300239563,
+      "learning_rate": 0.0017318518518518518,
+      "loss": 1.8013,
+      "step": 7240
+    },
+    {
+      "epoch": 26.85,
+      "grad_norm": 0.49864670634269714,
+      "learning_rate": 0.0017314814814814814,
+      "loss": 1.8095,
+      "step": 7250
+    },
+    {
+      "epoch": 26.89,
+      "grad_norm": 0.4981412887573242,
+      "learning_rate": 0.001731111111111111,
+      "loss": 1.7865,
+      "step": 7260
+    },
+    {
+      "epoch": 26.93,
+      "grad_norm": 0.49631133675575256,
+      "learning_rate": 0.0017307407407407408,
+      "loss": 1.8195,
+      "step": 7270
+    },
+    {
+      "epoch": 26.96,
+      "grad_norm": 0.5223541259765625,
+      "learning_rate": 0.0017303703703703704,
+      "loss": 1.815,
+      "step": 7280
+    },
+    {
+      "epoch": 27.0,
+      "grad_norm": 0.8307838439941406,
+      "learning_rate": 0.00173,
+      "loss": 1.8576,
+      "step": 7290
+    },
+    {
+      "epoch": 27.04,
+      "grad_norm": 0.5112069845199585,
+      "learning_rate": 0.0017296296296296296,
+      "loss": 1.5567,
+      "step": 7300
+    },
+    {
+      "epoch": 27.07,
+      "grad_norm": 0.5340370535850525,
+      "learning_rate": 0.0017292592592592594,
+      "loss": 1.5654,
+      "step": 7310
+    },
+    {
+      "epoch": 27.11,
+      "grad_norm": 0.5549976825714111,
+      "learning_rate": 0.001728888888888889,
+      "loss": 1.6092,
+      "step": 7320
+    },
+    {
+      "epoch": 27.15,
+      "grad_norm": 0.5123783349990845,
+      "learning_rate": 0.0017285185185185186,
+      "loss": 1.6213,
+      "step": 7330
+    },
+    {
+      "epoch": 27.19,
+      "grad_norm": 0.5847094058990479,
+      "learning_rate": 0.0017281481481481481,
+      "loss": 1.6237,
+      "step": 7340
+    },
+    {
+      "epoch": 27.22,
+      "grad_norm": 0.5540896654129028,
+      "learning_rate": 0.001727777777777778,
+      "loss": 1.6704,
+      "step": 7350
+    },
+    {
+      "epoch": 27.26,
+      "grad_norm": 0.5649771690368652,
+      "learning_rate": 0.0017274074074074075,
+      "loss": 1.6731,
+      "step": 7360
+    },
+    {
+      "epoch": 27.3,
+      "grad_norm": 0.5565537810325623,
+      "learning_rate": 0.001727037037037037,
+      "loss": 1.6729,
+      "step": 7370
+    },
+    {
+      "epoch": 27.33,
+      "grad_norm": 0.5422229766845703,
+      "learning_rate": 0.0017266666666666667,
+      "loss": 1.686,
+      "step": 7380
+    },
+    {
+      "epoch": 27.37,
+      "grad_norm": 0.4892452657222748,
+      "learning_rate": 0.0017262962962962963,
+      "loss": 1.7129,
+      "step": 7390
+    },
+    {
+      "epoch": 27.41,
+      "grad_norm": 0.5745701789855957,
+      "learning_rate": 0.0017259259259259259,
+      "loss": 1.7245,
+      "step": 7400
+    },
+    {
+      "epoch": 27.44,
+      "grad_norm": 0.5544735789299011,
+      "learning_rate": 0.0017255555555555555,
+      "loss": 1.7314,
+      "step": 7410
+    },
+    {
+      "epoch": 27.48,
+      "grad_norm": 0.5073575973510742,
+      "learning_rate": 0.0017251851851851853,
+      "loss": 1.7392,
+      "step": 7420
+    },
+    {
+      "epoch": 27.52,
+      "grad_norm": 0.5089659690856934,
+      "learning_rate": 0.0017248148148148149,
+      "loss": 1.7203,
+      "step": 7430
+    },
+    {
+      "epoch": 27.56,
+      "grad_norm": 0.51652991771698,
+      "learning_rate": 0.0017244444444444445,
+      "loss": 1.7763,
+      "step": 7440
+    },
+    {
+      "epoch": 27.59,
+      "grad_norm": 0.5411249399185181,
+      "learning_rate": 0.001724074074074074,
+      "loss": 1.7672,
+      "step": 7450
+    },
+    {
+      "epoch": 27.63,
+      "grad_norm": 0.5420709252357483,
+      "learning_rate": 0.0017237037037037039,
+      "loss": 1.7659,
+      "step": 7460
+    },
+    {
+      "epoch": 27.67,
+      "grad_norm": 0.5166801810264587,
+      "learning_rate": 0.0017233333333333334,
+      "loss": 1.7749,
+      "step": 7470
+    },
+    {
+      "epoch": 27.7,
+      "grad_norm": 0.5379444360733032,
+      "learning_rate": 0.001722962962962963,
+      "loss": 1.7884,
+      "step": 7480
+    },
+    {
+      "epoch": 27.74,
+      "grad_norm": 0.5639204978942871,
+      "learning_rate": 0.0017225925925925926,
+      "loss": 1.7857,
+      "step": 7490
+    },
+    {
+      "epoch": 27.78,
+      "grad_norm": 0.5733610987663269,
+      "learning_rate": 0.0017222222222222224,
+      "loss": 1.7809,
+      "step": 7500
+    },
+    {
+      "epoch": 27.81,
+      "grad_norm": 0.5559260845184326,
+      "learning_rate": 0.0017218518518518518,
+      "loss": 1.7841,
+      "step": 7510
+    },
+    {
+      "epoch": 27.85,
+      "grad_norm": 0.5033745169639587,
+      "learning_rate": 0.0017214814814814814,
+      "loss": 1.7862,
+      "step": 7520
+    },
+    {
+      "epoch": 27.89,
+      "grad_norm": 0.4976441264152527,
+      "learning_rate": 0.001721111111111111,
+      "loss": 1.7642,
+      "step": 7530
+    },
+    {
+      "epoch": 27.93,
+      "grad_norm": 0.48872360587120056,
+      "learning_rate": 0.0017207407407407408,
+      "loss": 1.8113,
+      "step": 7540
+    },
+    {
+      "epoch": 27.96,
+      "grad_norm": 0.5321819186210632,
+      "learning_rate": 0.0017203703703703704,
+      "loss": 1.8143,
+      "step": 7550
+    },
+    {
+      "epoch": 28.0,
+      "grad_norm": 1.1776005029678345,
+      "learning_rate": 0.00172,
+      "loss": 1.8321,
+      "step": 7560
+    },
+    {
+      "epoch": 28.04,
+      "grad_norm": 0.5432651042938232,
+      "learning_rate": 0.0017196296296296298,
+      "loss": 1.5738,
+      "step": 7570
+    },
+    {
+      "epoch": 28.07,
+      "grad_norm": 0.512981653213501,
+      "learning_rate": 0.0017192592592592594,
+      "loss": 1.5647,
+      "step": 7580
+    },
+    {
+      "epoch": 28.11,
+      "grad_norm": 0.5530872344970703,
+      "learning_rate": 0.001718888888888889,
+      "loss": 1.5745,
+      "step": 7590
+    },
+    {
+      "epoch": 28.15,
+      "grad_norm": 0.5536510944366455,
+      "learning_rate": 0.0017185185185185185,
+      "loss": 1.603,
+      "step": 7600
+    },
+    {
+      "epoch": 28.19,
+      "grad_norm": 0.5366228222846985,
+      "learning_rate": 0.0017181481481481483,
+      "loss": 1.6581,
+      "step": 7610
+    },
+    {
+      "epoch": 28.22,
+      "grad_norm": 0.631415843963623,
+      "learning_rate": 0.001717777777777778,
+      "loss": 1.6106,
+      "step": 7620
+    },
+    {
+      "epoch": 28.26,
+      "grad_norm": 0.5335027575492859,
+      "learning_rate": 0.0017174074074074075,
+      "loss": 1.6436,
+      "step": 7630
+    },
+    {
+      "epoch": 28.3,
+      "grad_norm": 0.5051824450492859,
+      "learning_rate": 0.001717037037037037,
+      "loss": 1.6577,
+      "step": 7640
+    },
+    {
+      "epoch": 28.33,
+      "grad_norm": 0.5768982768058777,
+      "learning_rate": 0.0017166666666666667,
+      "loss": 1.6765,
+      "step": 7650
+    },
+    {
+      "epoch": 28.37,
+      "grad_norm": 0.5530310869216919,
+      "learning_rate": 0.0017162962962962963,
+      "loss": 1.6831,
+      "step": 7660
+    },
+    {
+      "epoch": 28.41,
+      "grad_norm": 0.5534545183181763,
+      "learning_rate": 0.0017159259259259259,
+      "loss": 1.7275,
+      "step": 7670
+    },
+    {
+      "epoch": 28.44,
+      "grad_norm": 0.5321471095085144,
+      "learning_rate": 0.0017155555555555555,
+      "loss": 1.7241,
+      "step": 7680
+    },
+    {
+      "epoch": 28.48,
+      "grad_norm": 0.5455407500267029,
+      "learning_rate": 0.0017151851851851853,
+      "loss": 1.7278,
+      "step": 7690
+    },
+    {
+      "epoch": 28.52,
+      "grad_norm": 0.5250757336616516,
+      "learning_rate": 0.0017148148148148148,
+      "loss": 1.7389,
+      "step": 7700
+    },
+    {
+      "epoch": 28.56,
+      "grad_norm": 0.5096060037612915,
+      "learning_rate": 0.0017144444444444444,
+      "loss": 1.7569,
+      "step": 7710
+    },
+    {
+      "epoch": 28.59,
+      "grad_norm": 0.5587198138237,
+      "learning_rate": 0.001714074074074074,
+      "loss": 1.7363,
+      "step": 7720
+    },
+    {
+      "epoch": 28.63,
+      "grad_norm": 0.510917067527771,
+      "learning_rate": 0.0017137037037037038,
+      "loss": 1.7823,
+      "step": 7730
+    },
+    {
+      "epoch": 28.67,
+      "grad_norm": 0.5196210741996765,
+      "learning_rate": 0.0017133333333333334,
+      "loss": 1.7751,
+      "step": 7740
+    },
+    {
+      "epoch": 28.7,
+      "grad_norm": 0.5318698883056641,
+      "learning_rate": 0.001712962962962963,
+      "loss": 1.7504,
+      "step": 7750
+    },
+    {
+      "epoch": 28.74,
+      "grad_norm": 0.5652453303337097,
+      "learning_rate": 0.0017125925925925926,
+      "loss": 1.7604,
+      "step": 7760
+    },
+    {
+      "epoch": 28.78,
+      "grad_norm": 0.5259050130844116,
+      "learning_rate": 0.0017122222222222224,
+      "loss": 1.7665,
+      "step": 7770
+    },
+    {
+      "epoch": 28.81,
+      "grad_norm": 0.5152835249900818,
+      "learning_rate": 0.001711851851851852,
+      "loss": 1.7487,
+      "step": 7780
+    },
+    {
+      "epoch": 28.85,
+      "grad_norm": 0.5386359691619873,
+      "learning_rate": 0.0017114814814814814,
+      "loss": 1.8161,
+      "step": 7790
+    },
+    {
+      "epoch": 28.89,
+      "grad_norm": 0.514400064945221,
+      "learning_rate": 0.0017111111111111112,
+      "loss": 1.795,
+      "step": 7800
+    },
+    {
+      "epoch": 28.93,
+      "grad_norm": 0.5309016108512878,
+      "learning_rate": 0.0017107407407407408,
+      "loss": 1.7683,
+      "step": 7810
+    },
+    {
+      "epoch": 28.96,
+      "grad_norm": 0.5700574517250061,
+      "learning_rate": 0.0017103703703703703,
+      "loss": 1.8102,
+      "step": 7820
+    },
+    {
+      "epoch": 29.0,
+      "grad_norm": 1.172451138496399,
+      "learning_rate": 0.00171,
+      "loss": 1.7893,
+      "step": 7830
+    },
+    {
+      "epoch": 29.04,
+      "grad_norm": 0.5429467558860779,
+      "learning_rate": 0.0017096296296296297,
+      "loss": 1.5388,
+      "step": 7840
+    },
+    {
+      "epoch": 29.07,
+      "grad_norm": 0.5805901288986206,
+      "learning_rate": 0.0017092592592592593,
+      "loss": 1.5459,
+      "step": 7850
+    },
+    {
+      "epoch": 29.11,
+      "grad_norm": 0.539729118347168,
+      "learning_rate": 0.001708888888888889,
+      "loss": 1.5985,
+      "step": 7860
+    },
+    {
+      "epoch": 29.15,
+      "grad_norm": 0.5773619413375854,
+      "learning_rate": 0.0017085185185185185,
+      "loss": 1.5833,
+      "step": 7870
+    },
+    {
+      "epoch": 29.19,
+      "grad_norm": 0.564813494682312,
+      "learning_rate": 0.0017081481481481483,
+      "loss": 1.6251,
+      "step": 7880
+    },
+    {
+      "epoch": 29.22,
+      "grad_norm": 0.5339441299438477,
+      "learning_rate": 0.001707777777777778,
+      "loss": 1.6363,
+      "step": 7890
+    },
+    {
+      "epoch": 29.26,
+      "grad_norm": 0.5444590449333191,
+      "learning_rate": 0.0017074074074074075,
+      "loss": 1.653,
+      "step": 7900
+    },
+    {
+      "epoch": 29.3,
+      "grad_norm": 0.5453532934188843,
+      "learning_rate": 0.001707037037037037,
+      "loss": 1.6631,
+      "step": 7910
+    },
+    {
+      "epoch": 29.33,
+      "grad_norm": 0.5633497834205627,
+      "learning_rate": 0.0017066666666666669,
+      "loss": 1.6852,
+      "step": 7920
+    },
+    {
+      "epoch": 29.37,
+      "grad_norm": 0.506697416305542,
+      "learning_rate": 0.0017062962962962963,
+      "loss": 1.6729,
+      "step": 7930
+    },
+    {
+      "epoch": 29.41,
+      "grad_norm": 0.558829665184021,
+      "learning_rate": 0.0017059259259259258,
+      "loss": 1.6851,
+      "step": 7940
+    },
+    {
+      "epoch": 29.44,
+      "grad_norm": 0.5217764377593994,
+      "learning_rate": 0.0017055555555555554,
+      "loss": 1.6967,
+      "step": 7950
+    },
+    {
+      "epoch": 29.48,
+      "grad_norm": 0.5348731875419617,
+      "learning_rate": 0.0017051851851851852,
+      "loss": 1.6994,
+      "step": 7960
+    },
+    {
+      "epoch": 29.52,
+      "grad_norm": 0.5449881553649902,
+      "learning_rate": 0.0017048148148148148,
+      "loss": 1.6906,
+      "step": 7970
+    },
+    {
+      "epoch": 29.56,
+      "grad_norm": 0.5640777349472046,
+      "learning_rate": 0.0017044444444444444,
+      "loss": 1.7228,
+      "step": 7980
+    },
+    {
+      "epoch": 29.59,
+      "grad_norm": 0.5632838606834412,
+      "learning_rate": 0.0017040740740740742,
+      "loss": 1.7475,
+      "step": 7990
+    },
+    {
+      "epoch": 29.63,
+      "grad_norm": 0.514080822467804,
+      "learning_rate": 0.0017037037037037038,
+      "loss": 1.7825,
+      "step": 8000
+    },
+    {
+      "epoch": 29.67,
+      "grad_norm": 0.5150419473648071,
+      "learning_rate": 0.0017033333333333334,
+      "loss": 1.735,
+      "step": 8010
+    },
+    {
+      "epoch": 29.7,
+      "grad_norm": 0.5598604083061218,
+      "learning_rate": 0.001702962962962963,
+      "loss": 1.7627,
+      "step": 8020
+    },
+    {
+      "epoch": 29.74,
+      "grad_norm": 0.5505332350730896,
+      "learning_rate": 0.0017025925925925928,
+      "loss": 1.7434,
+      "step": 8030
+    },
+    {
+      "epoch": 29.78,
+      "grad_norm": 0.5757072567939758,
+      "learning_rate": 0.0017022222222222224,
+      "loss": 1.7673,
+      "step": 8040
+    },
+    {
+      "epoch": 29.81,
+      "grad_norm": 0.5394684672355652,
+      "learning_rate": 0.001701851851851852,
+      "loss": 1.7808,
+      "step": 8050
+    },
+    {
+      "epoch": 29.85,
+      "grad_norm": 0.5051487684249878,
+      "learning_rate": 0.0017014814814814813,
+      "loss": 1.779,
+      "step": 8060
+    },
+    {
+      "epoch": 29.89,
+      "grad_norm": 0.45365673303604126,
+      "learning_rate": 0.0017011111111111111,
+      "loss": 1.7741,
+      "step": 8070
+    },
+    {
+      "epoch": 29.93,
+      "grad_norm": 0.5558077692985535,
+      "learning_rate": 0.0017007407407407407,
+      "loss": 1.7883,
+      "step": 8080
+    },
+    {
+      "epoch": 29.96,
+      "grad_norm": 0.5742713212966919,
+      "learning_rate": 0.0017003703703703703,
+      "loss": 1.7801,
+      "step": 8090
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 0.9552031755447388,
+      "learning_rate": 0.0017,
+      "loss": 1.7876,
+      "step": 8100
+    },
+    {
+      "epoch": 30.04,
+      "grad_norm": 0.5495716333389282,
+      "learning_rate": 0.0016996296296296297,
+      "loss": 1.4899,
+      "step": 8110
+    },
+    {
+      "epoch": 30.07,
+      "grad_norm": 0.5539374947547913,
+      "learning_rate": 0.0016992592592592593,
+      "loss": 1.5593,
+      "step": 8120
+    },
+    {
+      "epoch": 30.11,
+      "grad_norm": 0.6192070841789246,
+      "learning_rate": 0.0016988888888888889,
+      "loss": 1.5557,
+      "step": 8130
+    },
+    {
+      "epoch": 30.15,
+      "grad_norm": 0.5558332204818726,
+      "learning_rate": 0.0016985185185185185,
+      "loss": 1.5884,
+      "step": 8140
+    },
+    {
+      "epoch": 30.19,
+      "grad_norm": 0.5470682978630066,
+      "learning_rate": 0.0016981481481481483,
+      "loss": 1.5894,
+      "step": 8150
+    },
+    {
+      "epoch": 30.22,
+      "grad_norm": 0.547297477722168,
+      "learning_rate": 0.0016977777777777779,
+      "loss": 1.6244,
+      "step": 8160
+    },
+    {
+      "epoch": 30.26,
+      "grad_norm": 0.577240526676178,
+      "learning_rate": 0.0016974074074074075,
+      "loss": 1.6502,
+      "step": 8170
+    },
+    {
+      "epoch": 30.3,
+      "grad_norm": 0.5513033270835876,
+      "learning_rate": 0.0016970370370370373,
+      "loss": 1.6443,
+      "step": 8180
+    },
+    {
+      "epoch": 30.33,
+      "grad_norm": 0.5598578453063965,
+      "learning_rate": 0.0016966666666666669,
+      "loss": 1.6471,
+      "step": 8190
+    },
+    {
+      "epoch": 30.37,
+      "grad_norm": 0.5605131387710571,
+      "learning_rate": 0.0016962962962962962,
+      "loss": 1.6755,
+      "step": 8200
+    },
+    {
+      "epoch": 30.41,
+      "grad_norm": 0.5713827013969421,
+      "learning_rate": 0.0016959259259259258,
+      "loss": 1.6753,
+      "step": 8210
+    },
+    {
+      "epoch": 30.44,
+      "grad_norm": 0.5782068967819214,
+      "learning_rate": 0.0016955555555555556,
+      "loss": 1.7284,
+      "step": 8220
+    },
+    {
+      "epoch": 30.48,
+      "grad_norm": 0.5492303371429443,
+      "learning_rate": 0.0016951851851851852,
+      "loss": 1.6863,
+      "step": 8230
+    },
+    {
+      "epoch": 30.52,
+      "grad_norm": 0.5332817435264587,
+      "learning_rate": 0.0016948148148148148,
+      "loss": 1.702,
+      "step": 8240
+    },
+    {
+      "epoch": 30.56,
+      "grad_norm": 0.5532167553901672,
+      "learning_rate": 0.0016944444444444444,
+      "loss": 1.7317,
+      "step": 8250
+    },
+    {
+      "epoch": 30.59,
+      "grad_norm": 0.566382110118866,
+      "learning_rate": 0.0016940740740740742,
+      "loss": 1.7365,
+      "step": 8260
+    },
+    {
+      "epoch": 30.63,
+      "grad_norm": 0.542638897895813,
+      "learning_rate": 0.0016937037037037038,
+      "loss": 1.7288,
+      "step": 8270
+    },
+    {
+      "epoch": 30.67,
+      "grad_norm": 0.5863054990768433,
+      "learning_rate": 0.0016933333333333334,
+      "loss": 1.7207,
+      "step": 8280
+    },
+    {
+      "epoch": 30.7,
+      "grad_norm": 0.5963066816329956,
+      "learning_rate": 0.001692962962962963,
+      "loss": 1.7264,
+      "step": 8290
+    },
+    {
+      "epoch": 30.74,
+      "grad_norm": 0.5126426219940186,
+      "learning_rate": 0.0016925925925925928,
+      "loss": 1.7056,
+      "step": 8300
+    },
+    {
+      "epoch": 30.78,
+      "grad_norm": 0.5691437125205994,
+      "learning_rate": 0.0016922222222222223,
+      "loss": 1.7685,
+      "step": 8310
+    },
+    {
+      "epoch": 30.81,
+      "grad_norm": 0.4958675801753998,
+      "learning_rate": 0.001691851851851852,
+      "loss": 1.7537,
+      "step": 8320
+    },
+    {
+      "epoch": 30.85,
+      "grad_norm": 0.5244887471199036,
+      "learning_rate": 0.0016914814814814815,
+      "loss": 1.7765,
+      "step": 8330
+    },
+    {
+      "epoch": 30.89,
+      "grad_norm": 0.4961899220943451,
+      "learning_rate": 0.0016911111111111111,
+      "loss": 1.7801,
+      "step": 8340
+    },
+    {
+      "epoch": 30.93,
+      "grad_norm": 0.5399650931358337,
+      "learning_rate": 0.0016907407407407407,
+      "loss": 1.7591,
+      "step": 8350
+    },
+    {
+      "epoch": 30.96,
+      "grad_norm": 0.549034595489502,
+      "learning_rate": 0.0016903703703703703,
+      "loss": 1.7858,
+      "step": 8360
+    },
+    {
+      "epoch": 31.0,
+      "grad_norm": 1.2998181581497192,
+      "learning_rate": 0.0016899999999999999,
+      "loss": 1.8002,
+      "step": 8370
+    },
+    {
+      "epoch": 31.04,
+      "grad_norm": 0.5629531145095825,
+      "learning_rate": 0.0016896296296296297,
+      "loss": 1.5119,
+      "step": 8380
+    },
+    {
+      "epoch": 31.07,
+      "grad_norm": 0.5641016960144043,
+      "learning_rate": 0.0016892592592592593,
+      "loss": 1.5397,
+      "step": 8390
+    },
+    {
+      "epoch": 31.11,
+      "grad_norm": 0.5819301605224609,
+      "learning_rate": 0.0016888888888888889,
+      "loss": 1.5523,
+      "step": 8400
+    },
+    {
+      "epoch": 31.15,
+      "grad_norm": 0.5689669251441956,
+      "learning_rate": 0.0016885185185185187,
+      "loss": 1.5829,
+      "step": 8410
+    },
+    {
+      "epoch": 31.19,
+      "grad_norm": 0.5574280023574829,
+      "learning_rate": 0.0016881481481481483,
+      "loss": 1.5748,
+      "step": 8420
+    },
+    {
+      "epoch": 31.22,
+      "grad_norm": 0.6067005395889282,
+      "learning_rate": 0.0016877777777777778,
+      "loss": 1.6045,
+      "step": 8430
+    },
+    {
+      "epoch": 31.26,
+      "grad_norm": 0.5879372358322144,
+      "learning_rate": 0.0016874074074074074,
+      "loss": 1.6307,
+      "step": 8440
+    },
+    {
+      "epoch": 31.3,
+      "grad_norm": 0.6153671145439148,
+      "learning_rate": 0.0016870370370370372,
+      "loss": 1.6503,
+      "step": 8450
+    },
+    {
+      "epoch": 31.33,
+      "grad_norm": 0.5140969753265381,
+      "learning_rate": 0.0016866666666666668,
+      "loss": 1.6719,
+      "step": 8460
+    },
+    {
+      "epoch": 31.37,
+      "grad_norm": 0.5702369809150696,
+      "learning_rate": 0.0016862962962962964,
+      "loss": 1.6727,
+      "step": 8470
+    },
+    {
+      "epoch": 31.41,
+      "grad_norm": 0.5763192176818848,
+      "learning_rate": 0.0016859259259259258,
+      "loss": 1.6691,
+      "step": 8480
+    },
+    {
+      "epoch": 31.44,
+      "grad_norm": 0.5894213318824768,
+      "learning_rate": 0.0016855555555555556,
+      "loss": 1.69,
+      "step": 8490
+    },
+    {
+      "epoch": 31.48,
+      "grad_norm": 0.5990567207336426,
+      "learning_rate": 0.0016851851851851852,
+      "loss": 1.6665,
+      "step": 8500
+    },
+    {
+      "epoch": 31.52,
+      "grad_norm": 0.5143067240715027,
+      "learning_rate": 0.0016848148148148148,
+      "loss": 1.678,
+      "step": 8510
+    },
+    {
+      "epoch": 31.56,
+      "grad_norm": 0.5345697402954102,
+      "learning_rate": 0.0016844444444444444,
+      "loss": 1.708,
+      "step": 8520
+    },
+    {
+      "epoch": 31.59,
+      "grad_norm": 0.5773447751998901,
+      "learning_rate": 0.0016840740740740742,
+      "loss": 1.7029,
+      "step": 8530
+    },
+    {
+      "epoch": 31.63,
+      "grad_norm": 0.5366548299789429,
+      "learning_rate": 0.0016837037037037038,
+      "loss": 1.7183,
+      "step": 8540
+    },
+    {
+      "epoch": 31.67,
+      "grad_norm": 0.5319424867630005,
+      "learning_rate": 0.0016833333333333333,
+      "loss": 1.7219,
+      "step": 8550
+    },
+    {
+      "epoch": 31.7,
+      "grad_norm": 0.5273670554161072,
+      "learning_rate": 0.001682962962962963,
+      "loss": 1.7296,
+      "step": 8560
+    },
+    {
+      "epoch": 31.74,
+      "grad_norm": 0.5865380764007568,
+      "learning_rate": 0.0016825925925925927,
+      "loss": 1.7321,
+      "step": 8570
+    },
+    {
+      "epoch": 31.78,
+      "grad_norm": 0.5639680624008179,
+      "learning_rate": 0.0016822222222222223,
+      "loss": 1.7335,
+      "step": 8580
+    },
+    {
+      "epoch": 31.81,
+      "grad_norm": 0.5686609148979187,
+      "learning_rate": 0.001681851851851852,
+      "loss": 1.7393,
+      "step": 8590
+    },
+    {
+      "epoch": 31.85,
+      "grad_norm": 0.5990113019943237,
+      "learning_rate": 0.0016814814814814817,
+      "loss": 1.7704,
+      "step": 8600
+    },
+    {
+      "epoch": 31.89,
+      "grad_norm": 0.4920467734336853,
+      "learning_rate": 0.001681111111111111,
+      "loss": 1.7624,
+      "step": 8610
+    },
+    {
+      "epoch": 31.93,
+      "grad_norm": 0.5353578925132751,
+      "learning_rate": 0.0016807407407407407,
+      "loss": 1.7495,
+      "step": 8620
+    },
+    {
+      "epoch": 31.96,
+      "grad_norm": 0.503509521484375,
+      "learning_rate": 0.0016803703703703703,
+      "loss": 1.7685,
+      "step": 8630
+    },
+    {
+      "epoch": 32.0,
+      "grad_norm": 1.060388207435608,
+      "learning_rate": 0.00168,
+      "loss": 1.8034,
+      "step": 8640
+    },
+    {
+      "epoch": 32.04,
+      "grad_norm": 0.5331477522850037,
+      "learning_rate": 0.0016796296296296297,
+      "loss": 1.502,
+      "step": 8650
+    },
+    {
+      "epoch": 32.07,
+      "grad_norm": 0.5541171431541443,
+      "learning_rate": 0.0016792592592592592,
+      "loss": 1.5502,
+      "step": 8660
+    },
+    {
+      "epoch": 32.11,
+      "grad_norm": 0.5719496607780457,
+      "learning_rate": 0.0016788888888888888,
+      "loss": 1.5465,
+      "step": 8670
+    },
+    {
+      "epoch": 32.15,
+      "grad_norm": 0.5539078116416931,
+      "learning_rate": 0.0016785185185185186,
+      "loss": 1.5404,
+      "step": 8680
+    },
+    {
+      "epoch": 32.19,
+      "grad_norm": 0.5710481405258179,
+      "learning_rate": 0.0016781481481481482,
+      "loss": 1.5532,
+      "step": 8690
+    },
+    {
+      "epoch": 32.22,
+      "grad_norm": 0.5715047717094421,
+      "learning_rate": 0.0016777777777777778,
+      "loss": 1.5976,
+      "step": 8700
+    },
+    {
+      "epoch": 32.26,
+      "grad_norm": 0.5798360109329224,
+      "learning_rate": 0.0016774074074074074,
+      "loss": 1.6342,
+      "step": 8710
+    },
+    {
+      "epoch": 32.3,
+      "grad_norm": 0.5830899477005005,
+      "learning_rate": 0.0016770370370370372,
+      "loss": 1.6217,
+      "step": 8720
+    },
+    {
+      "epoch": 32.33,
+      "grad_norm": 0.5676963329315186,
+      "learning_rate": 0.0016766666666666668,
+      "loss": 1.6163,
+      "step": 8730
+    },
+    {
+      "epoch": 32.37,
+      "grad_norm": 0.586168646812439,
+      "learning_rate": 0.0016762962962962964,
+      "loss": 1.6336,
+      "step": 8740
+    },
+    {
+      "epoch": 32.41,
+      "grad_norm": 0.5690294504165649,
+      "learning_rate": 0.0016759259259259258,
+      "loss": 1.6528,
+      "step": 8750
+    },
+    {
+      "epoch": 32.44,
+      "grad_norm": 0.5214428901672363,
+      "learning_rate": 0.0016755555555555556,
+      "loss": 1.6664,
+      "step": 8760
+    },
+    {
+      "epoch": 32.48,
+      "grad_norm": 0.5383786559104919,
+      "learning_rate": 0.0016751851851851852,
+      "loss": 1.6719,
+      "step": 8770
+    },
+    {
+      "epoch": 32.52,
+      "grad_norm": 0.6171760559082031,
+      "learning_rate": 0.0016748148148148147,
+      "loss": 1.6992,
+      "step": 8780
+    },
+    {
+      "epoch": 32.56,
+      "grad_norm": 0.5024405121803284,
+      "learning_rate": 0.0016744444444444443,
+      "loss": 1.6858,
+      "step": 8790
+    },
+    {
+      "epoch": 32.59,
+      "grad_norm": 0.5647597908973694,
+      "learning_rate": 0.0016740740740740741,
+      "loss": 1.6923,
+      "step": 8800
+    },
+    {
+      "epoch": 32.63,
+      "grad_norm": 0.6205365657806396,
+      "learning_rate": 0.0016737037037037037,
+      "loss": 1.7016,
+      "step": 8810
+    },
+    {
+      "epoch": 32.67,
+      "grad_norm": 0.5677077770233154,
+      "learning_rate": 0.0016733333333333333,
+      "loss": 1.7227,
+      "step": 8820
+    },
+    {
+      "epoch": 32.7,
+      "grad_norm": 0.5691731572151184,
+      "learning_rate": 0.0016729629629629631,
+      "loss": 1.6843,
+      "step": 8830
+    },
+    {
+      "epoch": 32.74,
+      "grad_norm": 0.5742966532707214,
+      "learning_rate": 0.0016725925925925927,
+      "loss": 1.7423,
+      "step": 8840
+    },
+    {
+      "epoch": 32.78,
+      "grad_norm": 0.558135986328125,
+      "learning_rate": 0.0016722222222222223,
+      "loss": 1.7457,
+      "step": 8850
+    },
+    {
+      "epoch": 32.81,
+      "grad_norm": 0.5381272435188293,
+      "learning_rate": 0.0016718518518518519,
+      "loss": 1.7229,
+      "step": 8860
+    },
+    {
+      "epoch": 32.85,
+      "grad_norm": 0.53874272108078,
+      "learning_rate": 0.0016714814814814817,
+      "loss": 1.7586,
+      "step": 8870
+    },
+    {
+      "epoch": 32.89,
+      "grad_norm": 0.559933602809906,
+      "learning_rate": 0.0016711111111111113,
+      "loss": 1.7902,
+      "step": 8880
+    },
+    {
+      "epoch": 32.93,
+      "grad_norm": 0.5527853965759277,
+      "learning_rate": 0.0016707407407407406,
+      "loss": 1.7616,
+      "step": 8890
+    },
+    {
+      "epoch": 32.96,
+      "grad_norm": 0.600522518157959,
+      "learning_rate": 0.0016703703703703702,
+      "loss": 1.7471,
+      "step": 8900
+    },
+    {
+      "epoch": 33.0,
+      "grad_norm": 1.1233681440353394,
+      "learning_rate": 0.00167,
+      "loss": 1.7853,
+      "step": 8910
+    },
+    {
+      "epoch": 33.04,
+      "grad_norm": 0.5869570970535278,
+      "learning_rate": 0.0016696296296296296,
+      "loss": 1.5254,
+      "step": 8920
+    },
+    {
+      "epoch": 33.07,
+      "grad_norm": 0.5793870687484741,
+      "learning_rate": 0.0016692592592592592,
+      "loss": 1.52,
+      "step": 8930
+    },
+    {
+      "epoch": 33.11,
+      "grad_norm": 0.5554567575454712,
+      "learning_rate": 0.0016688888888888888,
+      "loss": 1.5477,
+      "step": 8940
+    },
+    {
+      "epoch": 33.15,
+      "grad_norm": 0.5523855686187744,
+      "learning_rate": 0.0016685185185185186,
+      "loss": 1.5726,
+      "step": 8950
+    },
+    {
+      "epoch": 33.19,
+      "grad_norm": 0.5525711178779602,
+      "learning_rate": 0.0016681481481481482,
+      "loss": 1.5859,
+      "step": 8960
+    },
+    {
+      "epoch": 33.22,
+      "grad_norm": 0.593908429145813,
+      "learning_rate": 0.0016677777777777778,
+      "loss": 1.5876,
+      "step": 8970
+    },
+    {
+      "epoch": 33.26,
+      "grad_norm": 0.5583921074867249,
+      "learning_rate": 0.0016674074074074074,
+      "loss": 1.5976,
+      "step": 8980
+    },
+    {
+      "epoch": 33.3,
+      "grad_norm": 0.5856320858001709,
+      "learning_rate": 0.0016670370370370372,
+      "loss": 1.6249,
+      "step": 8990
+    },
+    {
+      "epoch": 33.33,
+      "grad_norm": 0.5795212984085083,
+      "learning_rate": 0.0016666666666666668,
+      "loss": 1.6162,
+      "step": 9000
+    },
+    {
+      "epoch": 33.37,
+      "grad_norm": 0.6002980470657349,
+      "learning_rate": 0.0016662962962962964,
+      "loss": 1.6169,
+      "step": 9010
+    },
+    {
+      "epoch": 33.41,
+      "grad_norm": 0.60965895652771,
+      "learning_rate": 0.0016659259259259262,
+      "loss": 1.6507,
+      "step": 9020
+    },
+    {
+      "epoch": 33.44,
+      "grad_norm": 0.568634569644928,
+      "learning_rate": 0.0016655555555555555,
+      "loss": 1.6778,
+      "step": 9030
+    },
+    {
+      "epoch": 33.48,
+      "grad_norm": 0.5564097166061401,
+      "learning_rate": 0.0016651851851851851,
+      "loss": 1.6964,
+      "step": 9040
+    },
+    {
+      "epoch": 33.52,
+      "grad_norm": 0.5448118448257446,
+      "learning_rate": 0.0016648148148148147,
+      "loss": 1.6412,
+      "step": 9050
+    },
+    {
+      "epoch": 33.56,
+      "grad_norm": 0.5949601531028748,
+      "learning_rate": 0.0016644444444444445,
+      "loss": 1.6647,
+      "step": 9060
+    },
+    {
+      "epoch": 33.59,
+      "grad_norm": 0.5293440222740173,
+      "learning_rate": 0.0016640740740740741,
+      "loss": 1.6952,
+      "step": 9070
+    },
+    {
+      "epoch": 33.63,
+      "grad_norm": 0.6145957112312317,
+      "learning_rate": 0.0016637037037037037,
+      "loss": 1.7021,
+      "step": 9080
+    },
+    {
+      "epoch": 33.67,
+      "grad_norm": 0.5578882098197937,
+      "learning_rate": 0.0016633333333333333,
+      "loss": 1.7205,
+      "step": 9090
+    },
+    {
+      "epoch": 33.7,
+      "grad_norm": 0.5359580516815186,
+      "learning_rate": 0.001662962962962963,
+      "loss": 1.7085,
+      "step": 9100
+    },
+    {
+      "epoch": 33.74,
+      "grad_norm": 0.5765175223350525,
+      "learning_rate": 0.0016625925925925927,
+      "loss": 1.6787,
+      "step": 9110
+    },
+    {
+      "epoch": 33.78,
+      "grad_norm": 0.517185389995575,
+      "learning_rate": 0.0016622222222222223,
+      "loss": 1.7226,
+      "step": 9120
+    },
+    {
+      "epoch": 33.81,
+      "grad_norm": 0.5444340109825134,
+      "learning_rate": 0.0016618518518518519,
+      "loss": 1.7066,
+      "step": 9130
+    },
+    {
+      "epoch": 33.85,
+      "grad_norm": 0.5430245399475098,
+      "learning_rate": 0.0016614814814814817,
+      "loss": 1.6819,
+      "step": 9140
+    },
+    {
+      "epoch": 33.89,
+      "grad_norm": 0.5591315627098083,
+      "learning_rate": 0.0016611111111111113,
+      "loss": 1.7322,
+      "step": 9150
+    },
+    {
+      "epoch": 33.93,
+      "grad_norm": 0.5514317154884338,
+      "learning_rate": 0.0016607407407407408,
+      "loss": 1.7251,
+      "step": 9160
+    },
+    {
+      "epoch": 33.96,
+      "grad_norm": 0.6007348895072937,
+      "learning_rate": 0.0016603703703703702,
+      "loss": 1.7792,
+      "step": 9170
+    },
+    {
+      "epoch": 34.0,
+      "grad_norm": 0.8949159979820251,
+      "learning_rate": 0.00166,
+      "loss": 1.7497,
+      "step": 9180
+    },
+    {
+      "epoch": 34.04,
+      "grad_norm": 0.5425311326980591,
+      "learning_rate": 0.0016596296296296296,
+      "loss": 1.4758,
+      "step": 9190
+    },
+    {
+      "epoch": 34.07,
+      "grad_norm": 0.606648325920105,
+      "learning_rate": 0.0016592592592592592,
+      "loss": 1.5247,
+      "step": 9200
+    },
+    {
+      "epoch": 34.11,
+      "grad_norm": 0.542310357093811,
+      "learning_rate": 0.001658888888888889,
+      "loss": 1.5363,
+      "step": 9210
+    },
+    {
+      "epoch": 34.15,
+      "grad_norm": 0.5714516043663025,
+      "learning_rate": 0.0016585185185185186,
+      "loss": 1.5374,
+      "step": 9220
+    },
+    {
+      "epoch": 34.19,
+      "grad_norm": 0.6230427622795105,
+      "learning_rate": 0.0016581481481481482,
+      "loss": 1.5204,
+      "step": 9230
+    },
+    {
+      "epoch": 34.22,
+      "grad_norm": 0.5695710182189941,
+      "learning_rate": 0.0016577777777777778,
+      "loss": 1.576,
+      "step": 9240
+    },
+    {
+      "epoch": 34.26,
+      "grad_norm": 0.5765951871871948,
+      "learning_rate": 0.0016574074074074076,
+      "loss": 1.5759,
+      "step": 9250
+    },
+    {
+      "epoch": 34.3,
+      "grad_norm": 0.5947939157485962,
+      "learning_rate": 0.0016570370370370372,
+      "loss": 1.608,
+      "step": 9260
+    },
+    {
+      "epoch": 34.33,
+      "grad_norm": 0.607772171497345,
+      "learning_rate": 0.0016566666666666667,
+      "loss": 1.6108,
+      "step": 9270
+    },
+    {
+      "epoch": 34.37,
+      "grad_norm": 0.6625462770462036,
+      "learning_rate": 0.0016562962962962963,
+      "loss": 1.6323,
+      "step": 9280
+    },
+    {
+      "epoch": 34.41,
+      "grad_norm": 0.5714110732078552,
+      "learning_rate": 0.0016559259259259261,
+      "loss": 1.6922,
+      "step": 9290
+    },
+    {
+      "epoch": 34.44,
+      "grad_norm": 0.5694931149482727,
+      "learning_rate": 0.0016555555555555555,
+      "loss": 1.6251,
+      "step": 9300
+    },
+    {
+      "epoch": 34.48,
+      "grad_norm": 0.6467699408531189,
+      "learning_rate": 0.001655185185185185,
+      "loss": 1.6503,
+      "step": 9310
+    },
+    {
+      "epoch": 34.52,
+      "grad_norm": 0.6535106301307678,
+      "learning_rate": 0.0016548148148148147,
+      "loss": 1.6817,
+      "step": 9320
+    },
+    {
+      "epoch": 34.56,
+      "grad_norm": 0.5543297529220581,
+      "learning_rate": 0.0016544444444444445,
+      "loss": 1.6625,
+      "step": 9330
+    },
+    {
+      "epoch": 34.59,
+      "grad_norm": 0.5453930497169495,
+      "learning_rate": 0.001654074074074074,
+      "loss": 1.6692,
+      "step": 9340
+    },
+    {
+      "epoch": 34.63,
+      "grad_norm": 0.5699336528778076,
+      "learning_rate": 0.0016537037037037037,
+      "loss": 1.685,
+      "step": 9350
+    },
+    {
+      "epoch": 34.67,
+      "grad_norm": 0.5972739458084106,
+      "learning_rate": 0.0016533333333333333,
+      "loss": 1.6944,
+      "step": 9360
+    },
+    {
+      "epoch": 34.7,
+      "grad_norm": 0.6316691040992737,
+      "learning_rate": 0.001652962962962963,
+      "loss": 1.6992,
+      "step": 9370
+    },
+    {
+      "epoch": 34.74,
+      "grad_norm": 0.5370445847511292,
+      "learning_rate": 0.0016525925925925927,
+      "loss": 1.7141,
+      "step": 9380
+    },
+    {
+      "epoch": 34.78,
+      "grad_norm": 0.5576089024543762,
+      "learning_rate": 0.0016522222222222222,
+      "loss": 1.7085,
+      "step": 9390
+    },
+    {
+      "epoch": 34.81,
+      "grad_norm": 0.5875248312950134,
+      "learning_rate": 0.0016518518518518518,
+      "loss": 1.7217,
+      "step": 9400
+    },
+    {
+      "epoch": 34.85,
+      "grad_norm": 0.5235693454742432,
+      "learning_rate": 0.0016514814814814816,
+      "loss": 1.7026,
+      "step": 9410
+    },
+    {
+      "epoch": 34.89,
+      "grad_norm": 0.6246321797370911,
+      "learning_rate": 0.0016511111111111112,
+      "loss": 1.7451,
+      "step": 9420
+    },
+    {
+      "epoch": 34.93,
+      "grad_norm": 0.5568575859069824,
+      "learning_rate": 0.0016507407407407408,
+      "loss": 1.7116,
+      "step": 9430
+    },
+    {
+      "epoch": 34.96,
+      "grad_norm": 0.5829682350158691,
+      "learning_rate": 0.0016503703703703704,
+      "loss": 1.7469,
+      "step": 9440
+    },
+    {
+      "epoch": 35.0,
+      "grad_norm": 1.1844964027404785,
+      "learning_rate": 0.00165,
+      "loss": 1.7515,
+      "step": 9450
+    },
+    {
+      "epoch": 35.04,
+      "grad_norm": 0.5359990000724792,
+      "learning_rate": 0.0016496296296296296,
+      "loss": 1.4927,
+      "step": 9460
+    },
+    {
+      "epoch": 35.07,
+      "grad_norm": 0.5757229924201965,
+      "learning_rate": 0.0016492592592592592,
+      "loss": 1.5032,
+      "step": 9470
+    },
+    {
+      "epoch": 35.11,
+      "grad_norm": 0.615017294883728,
+      "learning_rate": 0.001648888888888889,
+      "loss": 1.5318,
+      "step": 9480
+    },
+    {
+      "epoch": 35.15,
+      "grad_norm": 0.540900707244873,
+      "learning_rate": 0.0016485185185185186,
+      "loss": 1.5309,
+      "step": 9490
+    },
+    {
+      "epoch": 35.19,
+      "grad_norm": 0.589608371257782,
+      "learning_rate": 0.0016481481481481482,
+      "loss": 1.5338,
+      "step": 9500
+    },
+    {
+      "epoch": 35.22,
+      "grad_norm": 0.5594486594200134,
+      "learning_rate": 0.0016477777777777777,
+      "loss": 1.5788,
+      "step": 9510
+    },
+    {
+      "epoch": 35.26,
+      "grad_norm": 0.6130346059799194,
+      "learning_rate": 0.0016474074074074075,
+      "loss": 1.612,
+      "step": 9520
+    },
+    {
+      "epoch": 35.3,
+      "grad_norm": 0.5725595355033875,
+      "learning_rate": 0.0016470370370370371,
+      "loss": 1.5903,
+      "step": 9530
+    },
+    {
+      "epoch": 35.33,
+      "grad_norm": 0.5720920562744141,
+      "learning_rate": 0.0016466666666666667,
+      "loss": 1.6188,
+      "step": 9540
+    },
+    {
+      "epoch": 35.37,
+      "grad_norm": 0.5802409648895264,
+      "learning_rate": 0.0016462962962962963,
+      "loss": 1.5857,
+      "step": 9550
+    },
+    {
+      "epoch": 35.41,
+      "grad_norm": 0.6417401432991028,
+      "learning_rate": 0.0016459259259259261,
+      "loss": 1.6185,
+      "step": 9560
+    },
+    {
+      "epoch": 35.44,
+      "grad_norm": 0.6042072772979736,
+      "learning_rate": 0.0016455555555555557,
+      "loss": 1.6666,
+      "step": 9570
+    },
+    {
+      "epoch": 35.48,
+      "grad_norm": 0.6269539594650269,
+      "learning_rate": 0.001645185185185185,
+      "loss": 1.6404,
+      "step": 9580
+    },
+    {
+      "epoch": 35.52,
+      "grad_norm": 0.5669009685516357,
+      "learning_rate": 0.0016448148148148147,
+      "loss": 1.6734,
+      "step": 9590
+    },
+    {
+      "epoch": 35.56,
+      "grad_norm": 0.5852543115615845,
+      "learning_rate": 0.0016444444444444445,
+      "loss": 1.6635,
+      "step": 9600
+    },
+    {
+      "epoch": 35.59,
+      "grad_norm": 0.6072357296943665,
+      "learning_rate": 0.001644074074074074,
+      "loss": 1.6624,
+      "step": 9610
+    },
+    {
+      "epoch": 35.63,
+      "grad_norm": 0.632712721824646,
+      "learning_rate": 0.0016437037037037036,
+      "loss": 1.6884,
+      "step": 9620
+    },
+    {
+      "epoch": 35.67,
+      "grad_norm": 0.5859778523445129,
+      "learning_rate": 0.0016433333333333335,
+      "loss": 1.6693,
+      "step": 9630
+    },
+    {
+      "epoch": 35.7,
+      "grad_norm": 0.5471202731132507,
+      "learning_rate": 0.001642962962962963,
+      "loss": 1.7173,
+      "step": 9640
+    },
+    {
+      "epoch": 35.74,
+      "grad_norm": 0.5577960014343262,
+      "learning_rate": 0.0016425925925925926,
+      "loss": 1.6699,
+      "step": 9650
+    },
+    {
+      "epoch": 35.78,
+      "grad_norm": 0.5657382607460022,
+      "learning_rate": 0.0016422222222222222,
+      "loss": 1.7091,
+      "step": 9660
+    },
+    {
+      "epoch": 35.81,
+      "grad_norm": 0.5789744257926941,
+      "learning_rate": 0.001641851851851852,
+      "loss": 1.6923,
+      "step": 9670
+    },
+    {
+      "epoch": 35.85,
+      "grad_norm": 0.5898125767707825,
+      "learning_rate": 0.0016414814814814816,
+      "loss": 1.6982,
+      "step": 9680
+    },
+    {
+      "epoch": 35.89,
+      "grad_norm": 0.5303668975830078,
+      "learning_rate": 0.0016411111111111112,
+      "loss": 1.722,
+      "step": 9690
+    },
+    {
+      "epoch": 35.93,
+      "grad_norm": 0.5981086492538452,
+      "learning_rate": 0.0016407407407407408,
+      "loss": 1.7036,
+      "step": 9700
+    },
+    {
+      "epoch": 35.96,
+      "grad_norm": 0.5202363729476929,
+      "learning_rate": 0.0016403703703703706,
+      "loss": 1.7071,
+      "step": 9710
+    },
+    {
+      "epoch": 36.0,
+      "grad_norm": 1.0010344982147217,
+      "learning_rate": 0.00164,
+      "loss": 1.7195,
+      "step": 9720
+    },
+    {
+      "epoch": 36.04,
+      "grad_norm": 0.5903262495994568,
+      "learning_rate": 0.0016396296296296296,
+      "loss": 1.4546,
+      "step": 9730
+    },
+    {
+      "epoch": 36.07,
+      "grad_norm": 0.5673249959945679,
+      "learning_rate": 0.0016392592592592591,
+      "loss": 1.4846,
+      "step": 9740
+    },
+    {
+      "epoch": 36.11,
+      "grad_norm": 0.623580813407898,
+      "learning_rate": 0.001638888888888889,
+      "loss": 1.5554,
+      "step": 9750
+    },
+    {
+      "epoch": 36.15,
+      "grad_norm": 0.5951016545295715,
+      "learning_rate": 0.0016385185185185185,
+      "loss": 1.5375,
+      "step": 9760
+    },
+    {
+      "epoch": 36.19,
+      "grad_norm": 0.5958836078643799,
+      "learning_rate": 0.0016381481481481481,
+      "loss": 1.5261,
+      "step": 9770
+    },
+    {
+      "epoch": 36.22,
+      "grad_norm": 0.6242657899856567,
+      "learning_rate": 0.0016377777777777777,
+      "loss": 1.5417,
+      "step": 9780
+    },
+    {
+      "epoch": 36.26,
+      "grad_norm": 0.6229197978973389,
+      "learning_rate": 0.0016374074074074075,
+      "loss": 1.587,
+      "step": 9790
+    },
+    {
+      "epoch": 36.3,
+      "grad_norm": 0.5648860931396484,
+      "learning_rate": 0.001637037037037037,
+      "loss": 1.5499,
+      "step": 9800
+    },
+    {
+      "epoch": 36.33,
+      "grad_norm": 0.5891751646995544,
+      "learning_rate": 0.0016366666666666667,
+      "loss": 1.5852,
+      "step": 9810
+    },
+    {
+      "epoch": 36.37,
+      "grad_norm": 0.8289972543716431,
+      "learning_rate": 0.0016362962962962965,
+      "loss": 1.5974,
+      "step": 9820
+    },
+    {
+      "epoch": 36.41,
+      "grad_norm": 0.6231335401535034,
+      "learning_rate": 0.001635925925925926,
+      "loss": 1.6485,
+      "step": 9830
+    },
+    {
+      "epoch": 36.44,
+      "grad_norm": 0.5463785529136658,
+      "learning_rate": 0.0016355555555555557,
+      "loss": 1.6329,
+      "step": 9840
+    },
+    {
+      "epoch": 36.48,
+      "grad_norm": 0.5732906460762024,
+      "learning_rate": 0.001635185185185185,
+      "loss": 1.6287,
+      "step": 9850
+    },
+    {
+      "epoch": 36.52,
+      "grad_norm": 0.5856965780258179,
+      "learning_rate": 0.0016348148148148149,
+      "loss": 1.6757,
+      "step": 9860
+    },
+    {
+      "epoch": 36.56,
+      "grad_norm": 0.5316782593727112,
+      "learning_rate": 0.0016344444444444444,
+      "loss": 1.6571,
+      "step": 9870
+    },
+    {
+      "epoch": 36.59,
+      "grad_norm": 0.6029806733131409,
+      "learning_rate": 0.001634074074074074,
+      "loss": 1.6276,
+      "step": 9880
+    },
+    {
+      "epoch": 36.63,
+      "grad_norm": 0.5773792862892151,
+      "learning_rate": 0.0016337037037037036,
+      "loss": 1.6763,
+      "step": 9890
+    },
+    {
+      "epoch": 36.67,
+      "grad_norm": 0.5454092621803284,
+      "learning_rate": 0.0016333333333333334,
+      "loss": 1.6742,
+      "step": 9900
+    },
+    {
+      "epoch": 36.7,
+      "grad_norm": 0.560020923614502,
+      "learning_rate": 0.001632962962962963,
+      "loss": 1.6871,
+      "step": 9910
+    },
+    {
+      "epoch": 36.74,
+      "grad_norm": 0.5670827031135559,
+      "learning_rate": 0.0016325925925925926,
+      "loss": 1.6921,
+      "step": 9920
+    },
+    {
+      "epoch": 36.78,
+      "grad_norm": 0.5637957453727722,
+      "learning_rate": 0.0016322222222222222,
+      "loss": 1.6752,
+      "step": 9930
+    },
+    {
+      "epoch": 36.81,
+      "grad_norm": 0.5903841257095337,
+      "learning_rate": 0.001631851851851852,
+      "loss": 1.6807,
+      "step": 9940
+    },
+    {
+      "epoch": 36.85,
+      "grad_norm": 0.5725448727607727,
+      "learning_rate": 0.0016314814814814816,
+      "loss": 1.7029,
+      "step": 9950
+    },
+    {
+      "epoch": 36.89,
+      "grad_norm": 0.554770290851593,
+      "learning_rate": 0.0016311111111111112,
+      "loss": 1.7361,
+      "step": 9960
+    },
+    {
+      "epoch": 36.93,
+      "grad_norm": 0.656335711479187,
+      "learning_rate": 0.0016307407407407408,
+      "loss": 1.7117,
+      "step": 9970
+    },
+    {
+      "epoch": 36.96,
+      "grad_norm": 0.5630307793617249,
+      "learning_rate": 0.0016303703703703706,
+      "loss": 1.6824,
+      "step": 9980
+    },
+    {
+      "epoch": 37.0,
+      "grad_norm": 1.2158167362213135,
+      "learning_rate": 0.00163,
+      "loss": 1.7112,
+      "step": 9990
+    },
+    {
+      "epoch": 37.04,
+      "grad_norm": 0.5833454132080078,
+      "learning_rate": 0.0016296296296296295,
+      "loss": 1.4483,
+      "step": 10000
+    },
+    {
+      "epoch": 37.07,
+      "grad_norm": 0.5893401503562927,
+      "learning_rate": 0.0016292592592592591,
+      "loss": 1.4812,
+      "step": 10010
+    },
+    {
+      "epoch": 37.11,
+      "grad_norm": 0.5967376828193665,
+      "learning_rate": 0.001628888888888889,
+      "loss": 1.5071,
+      "step": 10020
+    },
+    {
+      "epoch": 37.15,
+      "grad_norm": 0.572011411190033,
+      "learning_rate": 0.0016285185185185185,
+      "loss": 1.4945,
+      "step": 10030
+    },
+    {
+      "epoch": 37.19,
+      "grad_norm": 0.6522864699363708,
+      "learning_rate": 0.001628148148148148,
+      "loss": 1.5245,
+      "step": 10040
+    },
+    {
+      "epoch": 37.22,
+      "grad_norm": 0.5800381898880005,
+      "learning_rate": 0.001627777777777778,
+      "loss": 1.5355,
+      "step": 10050
+    },
+    {
+      "epoch": 37.26,
+      "grad_norm": 0.6092601418495178,
+      "learning_rate": 0.0016274074074074075,
+      "loss": 1.5551,
+      "step": 10060
+    },
+    {
+      "epoch": 37.3,
+      "grad_norm": 0.5820883512496948,
+      "learning_rate": 0.001627037037037037,
+      "loss": 1.5907,
+      "step": 10070
+    },
+    {
+      "epoch": 37.33,
+      "grad_norm": 0.6052312850952148,
+      "learning_rate": 0.0016266666666666667,
+      "loss": 1.5893,
+      "step": 10080
+    },
+    {
+      "epoch": 37.37,
+      "grad_norm": 0.6281132698059082,
+      "learning_rate": 0.0016262962962962965,
+      "loss": 1.5962,
+      "step": 10090
+    },
+    {
+      "epoch": 37.41,
+      "grad_norm": 0.6100971102714539,
+      "learning_rate": 0.001625925925925926,
+      "loss": 1.6276,
+      "step": 10100
+    },
+    {
+      "epoch": 37.44,
+      "grad_norm": 0.6440417170524597,
+      "learning_rate": 0.0016255555555555557,
+      "loss": 1.6179,
+      "step": 10110
+    },
+    {
+      "epoch": 37.48,
+      "grad_norm": 0.5597901940345764,
+      "learning_rate": 0.0016251851851851852,
+      "loss": 1.6376,
+      "step": 10120
+    },
+    {
+      "epoch": 37.52,
+      "grad_norm": 0.5792917609214783,
+      "learning_rate": 0.0016248148148148148,
+      "loss": 1.6437,
+      "step": 10130
+    },
+    {
+      "epoch": 37.56,
+      "grad_norm": 0.5647159218788147,
+      "learning_rate": 0.0016244444444444444,
+      "loss": 1.6415,
+      "step": 10140
+    },
+    {
+      "epoch": 37.59,
+      "grad_norm": 0.560454785823822,
+      "learning_rate": 0.001624074074074074,
+      "loss": 1.6694,
+      "step": 10150
+    },
+    {
+      "epoch": 37.63,
+      "grad_norm": 0.5815364718437195,
+      "learning_rate": 0.0016237037037037036,
+      "loss": 1.6779,
+      "step": 10160
+    },
+    {
+      "epoch": 37.67,
+      "grad_norm": 0.5997341871261597,
+      "learning_rate": 0.0016233333333333334,
+      "loss": 1.6526,
+      "step": 10170
+    },
+    {
+      "epoch": 37.7,
+      "grad_norm": 0.6044934988021851,
+      "learning_rate": 0.001622962962962963,
+      "loss": 1.6668,
+      "step": 10180
+    },
+    {
+      "epoch": 37.74,
+      "grad_norm": 0.6119056344032288,
+      "learning_rate": 0.0016225925925925926,
+      "loss": 1.7002,
+      "step": 10190
+    },
+    {
+      "epoch": 37.78,
+      "grad_norm": 0.5937202572822571,
+      "learning_rate": 0.0016222222222222222,
+      "loss": 1.656,
+      "step": 10200
+    },
+    {
+      "epoch": 37.81,
+      "grad_norm": 0.5740966200828552,
+      "learning_rate": 0.001621851851851852,
+      "loss": 1.7033,
+      "step": 10210
+    },
+    {
+      "epoch": 37.85,
+      "grad_norm": 0.5541761517524719,
+      "learning_rate": 0.0016214814814814816,
+      "loss": 1.6999,
+      "step": 10220
+    },
+    {
+      "epoch": 37.89,
+      "grad_norm": 0.5576919913291931,
+      "learning_rate": 0.0016211111111111111,
+      "loss": 1.6883,
+      "step": 10230
+    },
+    {
+      "epoch": 37.93,
+      "grad_norm": 0.5857290625572205,
+      "learning_rate": 0.001620740740740741,
+      "loss": 1.6925,
+      "step": 10240
+    },
+    {
+      "epoch": 37.96,
+      "grad_norm": 0.6111593842506409,
+      "learning_rate": 0.0016203703703703705,
+      "loss": 1.6872,
+      "step": 10250
+    },
+    {
+      "epoch": 38.0,
+      "grad_norm": 1.0599267482757568,
+      "learning_rate": 0.0016200000000000001,
+      "loss": 1.683,
+      "step": 10260
+    },
+    {
+      "epoch": 38.04,
+      "grad_norm": 0.5836325883865356,
+      "learning_rate": 0.0016196296296296295,
+      "loss": 1.4644,
+      "step": 10270
+    },
+    {
+      "epoch": 38.07,
+      "grad_norm": 0.5915476679801941,
+      "learning_rate": 0.0016192592592592593,
+      "loss": 1.4787,
+      "step": 10280
+    },
+    {
+      "epoch": 38.11,
+      "grad_norm": 0.63363116979599,
+      "learning_rate": 0.001618888888888889,
+      "loss": 1.456,
+      "step": 10290
+    },
+    {
+      "epoch": 38.15,
+      "grad_norm": 0.6005881428718567,
+      "learning_rate": 0.0016185185185185185,
+      "loss": 1.4905,
+      "step": 10300
+    },
+    {
+      "epoch": 38.19,
+      "grad_norm": 0.6480348110198975,
+      "learning_rate": 0.001618148148148148,
+      "loss": 1.5057,
+      "step": 10310
+    },
+    {
+      "epoch": 38.22,
+      "grad_norm": 0.5606887936592102,
+      "learning_rate": 0.0016177777777777779,
+      "loss": 1.5551,
+      "step": 10320
+    },
+    {
+      "epoch": 38.26,
+      "grad_norm": 0.5907540917396545,
+      "learning_rate": 0.0016174074074074075,
+      "loss": 1.5862,
+      "step": 10330
+    },
+    {
+      "epoch": 38.3,
+      "grad_norm": 0.5859313011169434,
+      "learning_rate": 0.001617037037037037,
+      "loss": 1.5688,
+      "step": 10340
+    },
+    {
+      "epoch": 38.33,
+      "grad_norm": 0.7137908935546875,
+      "learning_rate": 0.0016166666666666666,
+      "loss": 1.5659,
+      "step": 10350
+    },
+    {
+      "epoch": 38.37,
+      "grad_norm": 0.6301664710044861,
+      "learning_rate": 0.0016162962962962964,
+      "loss": 1.609,
+      "step": 10360
+    },
+    {
+      "epoch": 38.41,
+      "grad_norm": 0.64017254114151,
+      "learning_rate": 0.001615925925925926,
+      "loss": 1.583,
+      "step": 10370
+    },
+    {
+      "epoch": 38.44,
+      "grad_norm": 0.5742957592010498,
+      "learning_rate": 0.0016155555555555556,
+      "loss": 1.6008,
+      "step": 10380
+    },
+    {
+      "epoch": 38.48,
+      "grad_norm": 0.5781733393669128,
+      "learning_rate": 0.0016151851851851852,
+      "loss": 1.6219,
+      "step": 10390
+    },
+    {
+      "epoch": 38.52,
+      "grad_norm": 0.6394752264022827,
+      "learning_rate": 0.001614814814814815,
+      "loss": 1.6126,
+      "step": 10400
+    },
+    {
+      "epoch": 38.56,
+      "grad_norm": 0.5630576610565186,
+      "learning_rate": 0.0016144444444444444,
+      "loss": 1.6292,
+      "step": 10410
+    },
+    {
+      "epoch": 38.59,
+      "grad_norm": 0.6065866947174072,
+      "learning_rate": 0.001614074074074074,
+      "loss": 1.6424,
+      "step": 10420
+    },
+    {
+      "epoch": 38.63,
+      "grad_norm": 0.5589773058891296,
+      "learning_rate": 0.0016137037037037036,
+      "loss": 1.6514,
+      "step": 10430
+    },
+    {
+      "epoch": 38.67,
+      "grad_norm": 0.5719542503356934,
+      "learning_rate": 0.0016133333333333334,
+      "loss": 1.6464,
+      "step": 10440
+    },
+    {
+      "epoch": 38.7,
+      "grad_norm": 0.595180869102478,
+      "learning_rate": 0.001612962962962963,
+      "loss": 1.6467,
+      "step": 10450
+    },
+    {
+      "epoch": 38.74,
+      "grad_norm": 0.575110137462616,
+      "learning_rate": 0.0016125925925925925,
+      "loss": 1.6728,
+      "step": 10460
+    },
+    {
+      "epoch": 38.78,
+      "grad_norm": 0.6241551041603088,
+      "learning_rate": 0.0016122222222222224,
+      "loss": 1.6654,
+      "step": 10470
+    },
+    {
+      "epoch": 38.81,
+      "grad_norm": 0.5750718116760254,
+      "learning_rate": 0.001611851851851852,
+      "loss": 1.6792,
+      "step": 10480
+    },
+    {
+      "epoch": 38.85,
+      "grad_norm": 0.6219273805618286,
+      "learning_rate": 0.0016114814814814815,
+      "loss": 1.6844,
+      "step": 10490
+    },
+    {
+      "epoch": 38.89,
+      "grad_norm": 0.5741206407546997,
+      "learning_rate": 0.0016111111111111111,
+      "loss": 1.696,
+      "step": 10500
+    },
+    {
+      "epoch": 38.93,
+      "grad_norm": 0.5772682428359985,
+      "learning_rate": 0.001610740740740741,
+      "loss": 1.7026,
+      "step": 10510
+    },
+    {
+      "epoch": 38.96,
+      "grad_norm": 0.598931610584259,
+      "learning_rate": 0.0016103703703703705,
+      "loss": 1.6795,
+      "step": 10520
+    },
+    {
+      "epoch": 39.0,
+      "grad_norm": 0.944146454334259,
+      "learning_rate": 0.00161,
+      "loss": 1.6987,
+      "step": 10530
+    },
+    {
+      "epoch": 39.04,
+      "grad_norm": 0.578942596912384,
+      "learning_rate": 0.0016096296296296295,
+      "loss": 1.4329,
+      "step": 10540
+    },
+    {
+      "epoch": 39.07,
+      "grad_norm": 0.6079661250114441,
+      "learning_rate": 0.0016092592592592593,
+      "loss": 1.4834,
+      "step": 10550
+    },
+    {
+      "epoch": 39.11,
+      "grad_norm": 0.6036855578422546,
+      "learning_rate": 0.0016088888888888889,
+      "loss": 1.5066,
+      "step": 10560
+    },
+    {
+      "epoch": 39.15,
+      "grad_norm": 0.620212972164154,
+      "learning_rate": 0.0016085185185185185,
+      "loss": 1.4874,
+      "step": 10570
+    },
+    {
+      "epoch": 39.19,
+      "grad_norm": 0.5837911367416382,
+      "learning_rate": 0.001608148148148148,
+      "loss": 1.5096,
+      "step": 10580
+    },
+    {
+      "epoch": 39.22,
+      "grad_norm": 0.581173837184906,
+      "learning_rate": 0.0016077777777777779,
+      "loss": 1.505,
+      "step": 10590
+    },
+    {
+      "epoch": 39.26,
+      "grad_norm": 0.6712148189544678,
+      "learning_rate": 0.0016074074074074074,
+      "loss": 1.511,
+      "step": 10600
+    },
+    {
+      "epoch": 39.3,
+      "grad_norm": 0.5917540192604065,
+      "learning_rate": 0.001607037037037037,
+      "loss": 1.5772,
+      "step": 10610
+    },
+    {
+      "epoch": 39.33,
+      "grad_norm": 0.6199811100959778,
+      "learning_rate": 0.0016066666666666666,
+      "loss": 1.5893,
+      "step": 10620
+    },
+    {
+      "epoch": 39.37,
+      "grad_norm": 0.5958033800125122,
+      "learning_rate": 0.0016062962962962964,
+      "loss": 1.5871,
+      "step": 10630
+    },
+    {
+      "epoch": 39.41,
+      "grad_norm": 0.578544020652771,
+      "learning_rate": 0.001605925925925926,
+      "loss": 1.592,
+      "step": 10640
+    },
+    {
+      "epoch": 39.44,
+      "grad_norm": 0.6394162178039551,
+      "learning_rate": 0.0016055555555555556,
+      "loss": 1.6051,
+      "step": 10650
+    },
+    {
+      "epoch": 39.48,
+      "grad_norm": 0.6192345023155212,
+      "learning_rate": 0.0016051851851851854,
+      "loss": 1.6229,
+      "step": 10660
+    },
+    {
+      "epoch": 39.52,
+      "grad_norm": 0.5998772382736206,
+      "learning_rate": 0.001604814814814815,
+      "loss": 1.6152,
+      "step": 10670
+    },
+    {
+      "epoch": 39.56,
+      "grad_norm": 0.5876280665397644,
+      "learning_rate": 0.0016044444444444444,
+      "loss": 1.6249,
+      "step": 10680
+    },
+    {
+      "epoch": 39.59,
+      "grad_norm": 0.5959920287132263,
+      "learning_rate": 0.001604074074074074,
+      "loss": 1.6276,
+      "step": 10690
+    },
+    {
+      "epoch": 39.63,
+      "grad_norm": 0.6072013974189758,
+      "learning_rate": 0.0016037037037037038,
+      "loss": 1.6075,
+      "step": 10700
+    },
+    {
+      "epoch": 39.67,
+      "grad_norm": 0.6390188932418823,
+      "learning_rate": 0.0016033333333333333,
+      "loss": 1.6377,
+      "step": 10710
+    },
+    {
+      "epoch": 39.7,
+      "grad_norm": 0.6121541857719421,
+      "learning_rate": 0.001602962962962963,
+      "loss": 1.665,
+      "step": 10720
+    },
+    {
+      "epoch": 39.74,
+      "grad_norm": 0.5736008882522583,
+      "learning_rate": 0.0016025925925925925,
+      "loss": 1.6434,
+      "step": 10730
+    },
+    {
+      "epoch": 39.78,
+      "grad_norm": 0.5839884877204895,
+      "learning_rate": 0.0016022222222222223,
+      "loss": 1.6786,
+      "step": 10740
+    },
+    {
+      "epoch": 39.81,
+      "grad_norm": 0.6004469394683838,
+      "learning_rate": 0.001601851851851852,
+      "loss": 1.6428,
+      "step": 10750
+    },
+    {
+      "epoch": 39.85,
+      "grad_norm": 0.5822156667709351,
+      "learning_rate": 0.0016014814814814815,
+      "loss": 1.6589,
+      "step": 10760
+    },
+    {
+      "epoch": 39.89,
+      "grad_norm": 0.6007235646247864,
+      "learning_rate": 0.001601111111111111,
+      "loss": 1.682,
+      "step": 10770
+    },
+    {
+      "epoch": 39.93,
+      "grad_norm": 0.663915753364563,
+      "learning_rate": 0.001600740740740741,
+      "loss": 1.6864,
+      "step": 10780
+    },
+    {
+      "epoch": 39.96,
+      "grad_norm": 0.5694162249565125,
+      "learning_rate": 0.0016003703703703705,
+      "loss": 1.6844,
+      "step": 10790
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 0.9879453182220459,
+      "learning_rate": 0.0016,
+      "loss": 1.7021,
+      "step": 10800
+    },
+    {
+      "epoch": 40.04,
+      "grad_norm": 0.6228801012039185,
+      "learning_rate": 0.0015996296296296297,
+      "loss": 1.4537,
+      "step": 10810
+    },
+    {
+      "epoch": 40.07,
+      "grad_norm": 0.6306723356246948,
+      "learning_rate": 0.0015992592592592593,
+      "loss": 1.4237,
+      "step": 10820
+    },
+    {
+      "epoch": 40.11,
+      "grad_norm": 0.5553726553916931,
+      "learning_rate": 0.0015988888888888888,
+      "loss": 1.4835,
+      "step": 10830
+    },
+    {
+      "epoch": 40.15,
+      "grad_norm": 0.57485032081604,
+      "learning_rate": 0.0015985185185185184,
+      "loss": 1.4796,
+      "step": 10840
+    },
+    {
+      "epoch": 40.19,
+      "grad_norm": 0.6278731226921082,
+      "learning_rate": 0.0015981481481481482,
+      "loss": 1.4915,
+      "step": 10850
+    },
+    {
+      "epoch": 40.22,
+      "grad_norm": 0.6191769242286682,
+      "learning_rate": 0.0015977777777777778,
+      "loss": 1.509,
+      "step": 10860
+    },
+    {
+      "epoch": 40.26,
+      "grad_norm": 0.593605637550354,
+      "learning_rate": 0.0015974074074074074,
+      "loss": 1.5293,
+      "step": 10870
+    },
+    {
+      "epoch": 40.3,
+      "grad_norm": 0.6278729438781738,
+      "learning_rate": 0.001597037037037037,
+      "loss": 1.5459,
+      "step": 10880
+    },
+    {
+      "epoch": 40.33,
+      "grad_norm": 0.6142277121543884,
+      "learning_rate": 0.0015966666666666668,
+      "loss": 1.5509,
+      "step": 10890
+    },
+    {
+      "epoch": 40.37,
+      "grad_norm": 0.6126216650009155,
+      "learning_rate": 0.0015962962962962964,
+      "loss": 1.5465,
+      "step": 10900
+    },
+    {
+      "epoch": 40.41,
+      "grad_norm": 0.6625998616218567,
+      "learning_rate": 0.001595925925925926,
+      "loss": 1.5905,
+      "step": 10910
+    },
+    {
+      "epoch": 40.44,
+      "grad_norm": 0.6081230044364929,
+      "learning_rate": 0.0015955555555555556,
+      "loss": 1.5928,
+      "step": 10920
+    },
+    {
+      "epoch": 40.48,
+      "grad_norm": 0.5991567969322205,
+      "learning_rate": 0.0015951851851851854,
+      "loss": 1.5605,
+      "step": 10930
+    },
+    {
+      "epoch": 40.52,
+      "grad_norm": 0.6089213490486145,
+      "learning_rate": 0.001594814814814815,
+      "loss": 1.5856,
+      "step": 10940
+    },
+    {
+      "epoch": 40.56,
+      "grad_norm": 0.6468093395233154,
+      "learning_rate": 0.0015944444444444446,
+      "loss": 1.6097,
+      "step": 10950
+    },
+    {
+      "epoch": 40.59,
+      "grad_norm": 0.6019711494445801,
+      "learning_rate": 0.001594074074074074,
+      "loss": 1.613,
+      "step": 10960
+    },
+    {
+      "epoch": 40.63,
+      "grad_norm": 0.6264703869819641,
+      "learning_rate": 0.0015937037037037037,
+      "loss": 1.6485,
+      "step": 10970
+    },
+    {
+      "epoch": 40.67,
+      "grad_norm": 0.5833860635757446,
+      "learning_rate": 0.0015933333333333333,
+      "loss": 1.6455,
+      "step": 10980
+    },
+    {
+      "epoch": 40.7,
+      "grad_norm": 0.6113573908805847,
+      "learning_rate": 0.001592962962962963,
+      "loss": 1.6659,
+      "step": 10990
+    },
+    {
+      "epoch": 40.74,
+      "grad_norm": 0.5929760336875916,
+      "learning_rate": 0.0015925925925925925,
+      "loss": 1.6641,
+      "step": 11000
+    },
+    {
+      "epoch": 40.78,
+      "grad_norm": 0.5363314151763916,
+      "learning_rate": 0.0015922222222222223,
+      "loss": 1.6415,
+      "step": 11010
+    },
+    {
+      "epoch": 40.81,
+      "grad_norm": 0.6000963449478149,
+      "learning_rate": 0.0015918518518518519,
+      "loss": 1.648,
+      "step": 11020
+    },
+    {
+      "epoch": 40.85,
+      "grad_norm": 0.6014578342437744,
+      "learning_rate": 0.0015914814814814815,
+      "loss": 1.692,
+      "step": 11030
+    },
+    {
+      "epoch": 40.89,
+      "grad_norm": 0.5798655152320862,
+      "learning_rate": 0.001591111111111111,
+      "loss": 1.6799,
+      "step": 11040
+    },
+    {
+      "epoch": 40.93,
+      "grad_norm": 0.5583765506744385,
+      "learning_rate": 0.0015907407407407409,
+      "loss": 1.6767,
+      "step": 11050
+    },
+    {
+      "epoch": 40.96,
+      "grad_norm": 0.5604484677314758,
+      "learning_rate": 0.0015903703703703705,
+      "loss": 1.7067,
+      "step": 11060
+    },
+    {
+      "epoch": 41.0,
+      "grad_norm": 1.2024364471435547,
+      "learning_rate": 0.00159,
+      "loss": 1.6975,
+      "step": 11070
+    },
+    {
+      "epoch": 41.04,
+      "grad_norm": 0.6150010824203491,
+      "learning_rate": 0.0015896296296296299,
+      "loss": 1.4372,
+      "step": 11080
+    },
+    {
+      "epoch": 41.07,
+      "grad_norm": 0.605194091796875,
+      "learning_rate": 0.0015892592592592592,
+      "loss": 1.4411,
+      "step": 11090
+    },
+    {
+      "epoch": 41.11,
+      "grad_norm": 0.621233344078064,
+      "learning_rate": 0.0015888888888888888,
+      "loss": 1.4452,
+      "step": 11100
+    },
+    {
+      "epoch": 41.15,
+      "grad_norm": 0.5805311799049377,
+      "learning_rate": 0.0015885185185185184,
+      "loss": 1.478,
+      "step": 11110
+    },
+    {
+      "epoch": 41.19,
+      "grad_norm": 0.5650319457054138,
+      "learning_rate": 0.0015881481481481482,
+      "loss": 1.5149,
+      "step": 11120
+    },
+    {
+      "epoch": 41.22,
+      "grad_norm": 0.6075964570045471,
+      "learning_rate": 0.0015877777777777778,
+      "loss": 1.5134,
+      "step": 11130
+    },
+    {
+      "epoch": 41.26,
+      "grad_norm": 0.6080381870269775,
+      "learning_rate": 0.0015874074074074074,
+      "loss": 1.5133,
+      "step": 11140
+    },
+    {
+      "epoch": 41.3,
+      "grad_norm": 0.5945024490356445,
+      "learning_rate": 0.001587037037037037,
+      "loss": 1.5214,
+      "step": 11150
+    },
+    {
+      "epoch": 41.33,
+      "grad_norm": 0.5842931270599365,
+      "learning_rate": 0.0015866666666666668,
+      "loss": 1.5149,
+      "step": 11160
+    },
+    {
+      "epoch": 41.37,
+      "grad_norm": 0.6802297830581665,
+      "learning_rate": 0.0015862962962962964,
+      "loss": 1.5337,
+      "step": 11170
+    },
+    {
+      "epoch": 41.41,
+      "grad_norm": 0.6695743799209595,
+      "learning_rate": 0.001585925925925926,
+      "loss": 1.5723,
+      "step": 11180
+    },
+    {
+      "epoch": 41.44,
+      "grad_norm": 0.6051291227340698,
+      "learning_rate": 0.0015855555555555555,
+      "loss": 1.5818,
+      "step": 11190
+    },
+    {
+      "epoch": 41.48,
+      "grad_norm": 0.623485803604126,
+      "learning_rate": 0.0015851851851851854,
+      "loss": 1.5934,
+      "step": 11200
+    },
+    {
+      "epoch": 41.52,
+      "grad_norm": 0.5692495703697205,
+      "learning_rate": 0.001584814814814815,
+      "loss": 1.6444,
+      "step": 11210
+    },
+    {
+      "epoch": 41.56,
+      "grad_norm": 0.5983277559280396,
+      "learning_rate": 0.0015844444444444445,
+      "loss": 1.6346,
+      "step": 11220
+    },
+    {
+      "epoch": 41.59,
+      "grad_norm": 0.6303779482841492,
+      "learning_rate": 0.001584074074074074,
+      "loss": 1.6532,
+      "step": 11230
+    },
+    {
+      "epoch": 41.63,
+      "grad_norm": 0.6377682685852051,
+      "learning_rate": 0.0015837037037037037,
+      "loss": 1.639,
+      "step": 11240
+    },
+    {
+      "epoch": 41.67,
+      "grad_norm": 0.5868995189666748,
+      "learning_rate": 0.0015833333333333333,
+      "loss": 1.6179,
+      "step": 11250
+    },
+    {
+      "epoch": 41.7,
+      "grad_norm": 0.645334005355835,
+      "learning_rate": 0.0015829629629629629,
+      "loss": 1.6623,
+      "step": 11260
+    },
+    {
+      "epoch": 41.74,
+      "grad_norm": 0.6188551783561707,
+      "learning_rate": 0.0015825925925925927,
+      "loss": 1.6446,
+      "step": 11270
+    },
+    {
+      "epoch": 41.78,
+      "grad_norm": 0.6530919075012207,
+      "learning_rate": 0.0015822222222222223,
+      "loss": 1.6387,
+      "step": 11280
+    },
+    {
+      "epoch": 41.81,
+      "grad_norm": 0.6173209547996521,
+      "learning_rate": 0.0015818518518518519,
+      "loss": 1.6248,
+      "step": 11290
+    },
+    {
+      "epoch": 41.85,
+      "grad_norm": 0.6064965128898621,
+      "learning_rate": 0.0015814814814814815,
+      "loss": 1.6507,
+      "step": 11300
+    },
+    {
+      "epoch": 41.89,
+      "grad_norm": 0.5441533923149109,
+      "learning_rate": 0.0015811111111111113,
+      "loss": 1.6184,
+      "step": 11310
+    },
+    {
+      "epoch": 41.93,
+      "grad_norm": 0.6284180283546448,
+      "learning_rate": 0.0015807407407407408,
+      "loss": 1.6524,
+      "step": 11320
+    },
+    {
+      "epoch": 41.96,
+      "grad_norm": 0.6238906979560852,
+      "learning_rate": 0.0015803703703703704,
+      "loss": 1.6529,
+      "step": 11330
+    },
+    {
+      "epoch": 42.0,
+      "grad_norm": 1.1080102920532227,
+      "learning_rate": 0.00158,
+      "loss": 1.6716,
+      "step": 11340
+    },
+    {
+      "epoch": 42.04,
+      "grad_norm": 0.6129464507102966,
+      "learning_rate": 0.0015796296296296298,
+      "loss": 1.4015,
+      "step": 11350
+    },
+    {
+      "epoch": 42.07,
+      "grad_norm": 0.6170362234115601,
+      "learning_rate": 0.0015792592592592594,
+      "loss": 1.4541,
+      "step": 11360
+    },
+    {
+      "epoch": 42.11,
+      "grad_norm": 0.6223025321960449,
+      "learning_rate": 0.0015788888888888888,
+      "loss": 1.442,
+      "step": 11370
+    },
+    {
+      "epoch": 42.15,
+      "grad_norm": 0.6103693246841431,
+      "learning_rate": 0.0015785185185185184,
+      "loss": 1.4537,
+      "step": 11380
+    },
+    {
+      "epoch": 42.19,
+      "grad_norm": 0.6398111581802368,
+      "learning_rate": 0.0015781481481481482,
+      "loss": 1.5009,
+      "step": 11390
+    },
+    {
+      "epoch": 42.22,
+      "grad_norm": 0.6270093321800232,
+      "learning_rate": 0.0015777777777777778,
+      "loss": 1.4739,
+      "step": 11400
+    },
+    {
+      "epoch": 42.26,
+      "grad_norm": 0.6488049626350403,
+      "learning_rate": 0.0015774074074074074,
+      "loss": 1.5274,
+      "step": 11410
+    },
+    {
+      "epoch": 42.3,
+      "grad_norm": 0.6524102687835693,
+      "learning_rate": 0.001577037037037037,
+      "loss": 1.5243,
+      "step": 11420
+    },
+    {
+      "epoch": 42.33,
+      "grad_norm": 0.655152440071106,
+      "learning_rate": 0.0015766666666666668,
+      "loss": 1.5228,
+      "step": 11430
+    },
+    {
+      "epoch": 42.37,
+      "grad_norm": 0.675579309463501,
+      "learning_rate": 0.0015762962962962963,
+      "loss": 1.5568,
+      "step": 11440
+    },
+    {
+      "epoch": 42.41,
+      "grad_norm": 0.6234245300292969,
+      "learning_rate": 0.001575925925925926,
+      "loss": 1.6002,
+      "step": 11450
+    },
+    {
+      "epoch": 42.44,
+      "grad_norm": 0.5919743776321411,
+      "learning_rate": 0.0015755555555555557,
+      "loss": 1.5659,
+      "step": 11460
+    },
+    {
+      "epoch": 42.48,
+      "grad_norm": 0.6351736187934875,
+      "learning_rate": 0.0015751851851851853,
+      "loss": 1.5863,
+      "step": 11470
+    },
+    {
+      "epoch": 42.52,
+      "grad_norm": 0.6661149263381958,
+      "learning_rate": 0.001574814814814815,
+      "loss": 1.584,
+      "step": 11480
+    },
+    {
+      "epoch": 42.56,
+      "grad_norm": 0.6044712662696838,
+      "learning_rate": 0.0015744444444444445,
+      "loss": 1.611,
+      "step": 11490
+    },
+    {
+      "epoch": 42.59,
+      "grad_norm": 0.6325424313545227,
+      "learning_rate": 0.0015740740740740743,
+      "loss": 1.597,
+      "step": 11500
+    },
+    {
+      "epoch": 42.63,
+      "grad_norm": 0.5968678593635559,
+      "learning_rate": 0.0015737037037037037,
+      "loss": 1.6211,
+      "step": 11510
+    },
+    {
+      "epoch": 42.67,
+      "grad_norm": 0.60098797082901,
+      "learning_rate": 0.0015733333333333333,
+      "loss": 1.6103,
+      "step": 11520
+    },
+    {
+      "epoch": 42.7,
+      "grad_norm": 0.6495356559753418,
+      "learning_rate": 0.0015729629629629629,
+      "loss": 1.6439,
+      "step": 11530
+    },
+    {
+      "epoch": 42.74,
+      "grad_norm": 0.5992751717567444,
+      "learning_rate": 0.0015725925925925927,
+      "loss": 1.6211,
+      "step": 11540
+    },
+    {
+      "epoch": 42.78,
+      "grad_norm": 0.5891815423965454,
+      "learning_rate": 0.0015722222222222223,
+      "loss": 1.6477,
+      "step": 11550
+    },
+    {
+      "epoch": 42.81,
+      "grad_norm": 0.6189820766448975,
+      "learning_rate": 0.0015718518518518518,
+      "loss": 1.6335,
+      "step": 11560
+    },
+    {
+      "epoch": 42.85,
+      "grad_norm": 0.6549631357192993,
+      "learning_rate": 0.0015714814814814814,
+      "loss": 1.6318,
+      "step": 11570
+    },
+    {
+      "epoch": 42.89,
+      "grad_norm": 0.6061648726463318,
+      "learning_rate": 0.0015711111111111112,
+      "loss": 1.6649,
+      "step": 11580
+    },
+    {
+      "epoch": 42.93,
+      "grad_norm": 0.6047539710998535,
+      "learning_rate": 0.0015707407407407408,
+      "loss": 1.6516,
+      "step": 11590
+    },
+    {
+      "epoch": 42.96,
+      "grad_norm": 0.5719320178031921,
+      "learning_rate": 0.0015703703703703704,
+      "loss": 1.6608,
+      "step": 11600
+    },
+    {
+      "epoch": 43.0,
+      "grad_norm": 1.1914095878601074,
+      "learning_rate": 0.00157,
+      "loss": 1.649,
+      "step": 11610
+    },
+    {
+      "epoch": 43.04,
+      "grad_norm": 0.6697654724121094,
+      "learning_rate": 0.0015696296296296298,
+      "loss": 1.4124,
+      "step": 11620
+    },
+    {
+      "epoch": 43.07,
+      "grad_norm": 0.5861226320266724,
+      "learning_rate": 0.0015692592592592594,
+      "loss": 1.4241,
+      "step": 11630
+    },
+    {
+      "epoch": 43.11,
+      "grad_norm": 0.6160594820976257,
+      "learning_rate": 0.001568888888888889,
+      "loss": 1.4394,
+      "step": 11640
+    },
+    {
+      "epoch": 43.15,
+      "grad_norm": 0.6051038503646851,
+      "learning_rate": 0.0015685185185185184,
+      "loss": 1.469,
+      "step": 11650
+    },
+    {
+      "epoch": 43.19,
+      "grad_norm": 0.6148421764373779,
+      "learning_rate": 0.0015681481481481482,
+      "loss": 1.4866,
+      "step": 11660
+    },
+    {
+      "epoch": 43.22,
+      "grad_norm": 0.6597320437431335,
+      "learning_rate": 0.0015677777777777777,
+      "loss": 1.5145,
+      "step": 11670
+    },
+    {
+      "epoch": 43.26,
+      "grad_norm": 0.6334295868873596,
+      "learning_rate": 0.0015674074074074073,
+      "loss": 1.4995,
+      "step": 11680
+    },
+    {
+      "epoch": 43.3,
+      "grad_norm": 0.6128714680671692,
+      "learning_rate": 0.0015670370370370371,
+      "loss": 1.4923,
+      "step": 11690
+    },
+    {
+      "epoch": 43.33,
+      "grad_norm": 0.6361562609672546,
+      "learning_rate": 0.0015666666666666667,
+      "loss": 1.5172,
+      "step": 11700
+    },
+    {
+      "epoch": 43.37,
+      "grad_norm": 0.615699052810669,
+      "learning_rate": 0.0015662962962962963,
+      "loss": 1.5052,
+      "step": 11710
+    },
+    {
+      "epoch": 43.41,
+      "grad_norm": 0.6639183163642883,
+      "learning_rate": 0.001565925925925926,
+      "loss": 1.5669,
+      "step": 11720
+    },
+    {
+      "epoch": 43.44,
+      "grad_norm": 0.6383649706840515,
+      "learning_rate": 0.0015655555555555557,
+      "loss": 1.5403,
+      "step": 11730
+    },
+    {
+      "epoch": 43.48,
+      "grad_norm": 0.6247989535331726,
+      "learning_rate": 0.0015651851851851853,
+      "loss": 1.5791,
+      "step": 11740
+    },
+    {
+      "epoch": 43.52,
+      "grad_norm": 0.6123582720756531,
+      "learning_rate": 0.0015648148148148149,
+      "loss": 1.5503,
+      "step": 11750
+    },
+    {
+      "epoch": 43.56,
+      "grad_norm": 0.6618642807006836,
+      "learning_rate": 0.0015644444444444445,
+      "loss": 1.5792,
+      "step": 11760
+    },
+    {
+      "epoch": 43.59,
+      "grad_norm": 0.6254457831382751,
+      "learning_rate": 0.0015640740740740743,
+      "loss": 1.6016,
+      "step": 11770
+    },
+    {
+      "epoch": 43.63,
+      "grad_norm": 0.6345099210739136,
+      "learning_rate": 0.0015637037037037037,
+      "loss": 1.5877,
+      "step": 11780
+    },
+    {
+      "epoch": 43.67,
+      "grad_norm": 0.6265460848808289,
+      "learning_rate": 0.0015633333333333332,
+      "loss": 1.6274,
+      "step": 11790
+    },
+    {
+      "epoch": 43.7,
+      "grad_norm": 0.6505123376846313,
+      "learning_rate": 0.0015629629629629628,
+      "loss": 1.6227,
+      "step": 11800
+    },
+    {
+      "epoch": 43.74,
+      "grad_norm": 0.6225472688674927,
+      "learning_rate": 0.0015625925925925926,
+      "loss": 1.6175,
+      "step": 11810
+    },
+    {
+      "epoch": 43.78,
+      "grad_norm": 0.6467319130897522,
+      "learning_rate": 0.0015622222222222222,
+      "loss": 1.6325,
+      "step": 11820
+    },
+    {
+      "epoch": 43.81,
+      "grad_norm": 0.6350839138031006,
+      "learning_rate": 0.0015618518518518518,
+      "loss": 1.6572,
+      "step": 11830
+    },
+    {
+      "epoch": 43.85,
+      "grad_norm": 0.614852249622345,
+      "learning_rate": 0.0015614814814814814,
+      "loss": 1.6469,
+      "step": 11840
+    },
+    {
+      "epoch": 43.89,
+      "grad_norm": 0.6368483901023865,
+      "learning_rate": 0.0015611111111111112,
+      "loss": 1.6346,
+      "step": 11850
+    },
+    {
+      "epoch": 43.93,
+      "grad_norm": 0.6596633195877075,
+      "learning_rate": 0.0015607407407407408,
+      "loss": 1.6669,
+      "step": 11860
+    },
+    {
+      "epoch": 43.96,
+      "grad_norm": 0.5940620303153992,
+      "learning_rate": 0.0015603703703703704,
+      "loss": 1.6529,
+      "step": 11870
+    },
+    {
+      "epoch": 44.0,
+      "grad_norm": 1.0889242887496948,
+      "learning_rate": 0.0015600000000000002,
+      "loss": 1.6417,
+      "step": 11880
+    },
+    {
+      "epoch": 44.04,
+      "grad_norm": 0.6191521883010864,
+      "learning_rate": 0.0015596296296296298,
+      "loss": 1.4271,
+      "step": 11890
+    },
+    {
+      "epoch": 44.07,
+      "grad_norm": 0.6261031031608582,
+      "learning_rate": 0.0015592592592592594,
+      "loss": 1.4191,
+      "step": 11900
+    },
+    {
+      "epoch": 44.11,
+      "grad_norm": 0.6673709154129028,
+      "learning_rate": 0.001558888888888889,
+      "loss": 1.4281,
+      "step": 11910
+    },
+    {
+      "epoch": 44.15,
+      "grad_norm": 0.621164083480835,
+      "learning_rate": 0.0015585185185185185,
+      "loss": 1.4318,
+      "step": 11920
+    },
+    {
+      "epoch": 44.19,
+      "grad_norm": 0.6410382390022278,
+      "learning_rate": 0.0015581481481481481,
+      "loss": 1.498,
+      "step": 11930
+    },
+    {
+      "epoch": 44.22,
+      "grad_norm": 0.6146425604820251,
+      "learning_rate": 0.0015577777777777777,
+      "loss": 1.5023,
+      "step": 11940
+    },
+    {
+      "epoch": 44.26,
+      "grad_norm": 0.6328328251838684,
+      "learning_rate": 0.0015574074074074073,
+      "loss": 1.5026,
+      "step": 11950
+    },
+    {
+      "epoch": 44.3,
+      "grad_norm": 0.677683413028717,
+      "learning_rate": 0.0015570370370370371,
+      "loss": 1.5434,
+      "step": 11960
+    },
+    {
+      "epoch": 44.33,
+      "grad_norm": 0.6298605799674988,
+      "learning_rate": 0.0015566666666666667,
+      "loss": 1.5206,
+      "step": 11970
+    },
+    {
+      "epoch": 44.37,
+      "grad_norm": 0.6399940848350525,
+      "learning_rate": 0.0015562962962962963,
+      "loss": 1.5417,
+      "step": 11980
+    },
+    {
+      "epoch": 44.41,
+      "grad_norm": 0.6250726580619812,
+      "learning_rate": 0.0015559259259259259,
+      "loss": 1.5478,
+      "step": 11990
+    },
+    {
+      "epoch": 44.44,
+      "grad_norm": 0.6349297165870667,
+      "learning_rate": 0.0015555555555555557,
+      "loss": 1.5409,
+      "step": 12000
+    },
+    {
+      "epoch": 44.48,
+      "grad_norm": 0.686031699180603,
+      "learning_rate": 0.0015551851851851853,
+      "loss": 1.5595,
+      "step": 12010
+    },
+    {
+      "epoch": 44.52,
+      "grad_norm": 0.5983814001083374,
+      "learning_rate": 0.0015548148148148149,
+      "loss": 1.5518,
+      "step": 12020
+    },
+    {
+      "epoch": 44.56,
+      "grad_norm": 0.6011641621589661,
+      "learning_rate": 0.0015544444444444444,
+      "loss": 1.5647,
+      "step": 12030
+    },
+    {
+      "epoch": 44.59,
+      "grad_norm": 0.6552057266235352,
+      "learning_rate": 0.0015540740740740743,
+      "loss": 1.5726,
+      "step": 12040
+    },
+    {
+      "epoch": 44.63,
+      "grad_norm": 0.6505006551742554,
+      "learning_rate": 0.0015537037037037038,
+      "loss": 1.5974,
+      "step": 12050
+    },
+    {
+      "epoch": 44.67,
+      "grad_norm": 0.6745777726173401,
+      "learning_rate": 0.0015533333333333332,
+      "loss": 1.6191,
+      "step": 12060
+    },
+    {
+      "epoch": 44.7,
+      "grad_norm": 0.6342930197715759,
+      "learning_rate": 0.0015529629629629628,
+      "loss": 1.5988,
+      "step": 12070
+    },
+    {
+      "epoch": 44.74,
+      "grad_norm": 0.6098774075508118,
+      "learning_rate": 0.0015525925925925926,
+      "loss": 1.611,
+      "step": 12080
+    },
+    {
+      "epoch": 44.78,
+      "grad_norm": 0.7015253901481628,
+      "learning_rate": 0.0015522222222222222,
+      "loss": 1.6142,
+      "step": 12090
+    },
+    {
+      "epoch": 44.81,
+      "grad_norm": 0.5675494074821472,
+      "learning_rate": 0.0015518518518518518,
+      "loss": 1.6014,
+      "step": 12100
+    },
+    {
+      "epoch": 44.85,
+      "grad_norm": 0.6159512996673584,
+      "learning_rate": 0.0015514814814814816,
+      "loss": 1.6346,
+      "step": 12110
+    },
+    {
+      "epoch": 44.89,
+      "grad_norm": 0.6535830497741699,
+      "learning_rate": 0.0015511111111111112,
+      "loss": 1.6159,
+      "step": 12120
+    },
+    {
+      "epoch": 44.93,
+      "grad_norm": 0.6116699576377869,
+      "learning_rate": 0.0015507407407407408,
+      "loss": 1.6558,
+      "step": 12130
+    },
+    {
+      "epoch": 44.96,
+      "grad_norm": 0.6020302176475525,
+      "learning_rate": 0.0015503703703703704,
+      "loss": 1.6704,
+      "step": 12140
+    },
+    {
+      "epoch": 45.0,
+      "grad_norm": 1.1520200967788696,
+      "learning_rate": 0.0015500000000000002,
+      "loss": 1.6353,
+      "step": 12150
+    },
+    {
+      "epoch": 45.04,
+      "grad_norm": 0.6187354326248169,
+      "learning_rate": 0.0015496296296296298,
+      "loss": 1.4284,
+      "step": 12160
+    },
+    {
+      "epoch": 45.07,
+      "grad_norm": 0.6019529104232788,
+      "learning_rate": 0.0015492592592592593,
+      "loss": 1.3639,
+      "step": 12170
+    },
+    {
+      "epoch": 45.11,
+      "grad_norm": 0.6266639232635498,
+      "learning_rate": 0.001548888888888889,
+      "loss": 1.4464,
+      "step": 12180
+    },
+    {
+      "epoch": 45.15,
+      "grad_norm": 0.6570635437965393,
+      "learning_rate": 0.0015485185185185187,
+      "loss": 1.44,
+      "step": 12190
+    },
+    {
+      "epoch": 45.19,
+      "grad_norm": 0.6148927211761475,
+      "learning_rate": 0.001548148148148148,
+      "loss": 1.4382,
+      "step": 12200
+    },
+    {
+      "epoch": 45.22,
+      "grad_norm": 0.6563842296600342,
+      "learning_rate": 0.0015477777777777777,
+      "loss": 1.4584,
+      "step": 12210
+    },
+    {
+      "epoch": 45.26,
+      "grad_norm": 0.6292750239372253,
+      "learning_rate": 0.0015474074074074073,
+      "loss": 1.4918,
+      "step": 12220
+    },
+    {
+      "epoch": 45.3,
+      "grad_norm": 0.6535168886184692,
+      "learning_rate": 0.001547037037037037,
+      "loss": 1.4881,
+      "step": 12230
+    },
+    {
+      "epoch": 45.33,
+      "grad_norm": 0.6435755491256714,
+      "learning_rate": 0.0015466666666666667,
+      "loss": 1.5152,
+      "step": 12240
+    },
+    {
+      "epoch": 45.37,
+      "grad_norm": 0.6139950156211853,
+      "learning_rate": 0.0015462962962962963,
+      "loss": 1.5355,
+      "step": 12250
+    },
+    {
+      "epoch": 45.41,
+      "grad_norm": 0.661777913570404,
+      "learning_rate": 0.0015459259259259259,
+      "loss": 1.5319,
+      "step": 12260
+    },
+    {
+      "epoch": 45.44,
+      "grad_norm": 0.7189764976501465,
+      "learning_rate": 0.0015455555555555557,
+      "loss": 1.5786,
+      "step": 12270
+    },
+    {
+      "epoch": 45.48,
+      "grad_norm": 0.604444146156311,
+      "learning_rate": 0.0015451851851851852,
+      "loss": 1.5659,
+      "step": 12280
+    },
+    {
+      "epoch": 45.52,
+      "grad_norm": 0.6002549529075623,
+      "learning_rate": 0.0015448148148148148,
+      "loss": 1.5513,
+      "step": 12290
+    },
+    {
+      "epoch": 45.56,
+      "grad_norm": 0.6639789342880249,
+      "learning_rate": 0.0015444444444444446,
+      "loss": 1.5771,
+      "step": 12300
+    },
+    {
+      "epoch": 45.59,
+      "grad_norm": 0.6240954399108887,
+      "learning_rate": 0.0015440740740740742,
+      "loss": 1.5577,
+      "step": 12310
+    },
+    {
+      "epoch": 45.63,
+      "grad_norm": 0.6983676552772522,
+      "learning_rate": 0.0015437037037037038,
+      "loss": 1.5728,
+      "step": 12320
+    },
+    {
+      "epoch": 45.67,
+      "grad_norm": 0.6203968524932861,
+      "learning_rate": 0.0015433333333333332,
+      "loss": 1.5723,
+      "step": 12330
+    },
+    {
+      "epoch": 45.7,
+      "grad_norm": 0.6263325810432434,
+      "learning_rate": 0.001542962962962963,
+      "loss": 1.5891,
+      "step": 12340
+    },
+    {
+      "epoch": 45.74,
+      "grad_norm": 0.6136410236358643,
+      "learning_rate": 0.0015425925925925926,
+      "loss": 1.5958,
+      "step": 12350
+    },
+    {
+      "epoch": 45.78,
+      "grad_norm": 0.6326088309288025,
+      "learning_rate": 0.0015422222222222222,
+      "loss": 1.6039,
+      "step": 12360
+    },
+    {
+      "epoch": 45.81,
+      "grad_norm": 0.649049699306488,
+      "learning_rate": 0.0015418518518518518,
+      "loss": 1.6412,
+      "step": 12370
+    },
+    {
+      "epoch": 45.85,
+      "grad_norm": 0.6206584572792053,
+      "learning_rate": 0.0015414814814814816,
+      "loss": 1.6086,
+      "step": 12380
+    },
+    {
+      "epoch": 45.89,
+      "grad_norm": 0.6422502994537354,
+      "learning_rate": 0.0015411111111111112,
+      "loss": 1.6182,
+      "step": 12390
+    },
+    {
+      "epoch": 45.93,
+      "grad_norm": 0.612797737121582,
+      "learning_rate": 0.0015407407407407407,
+      "loss": 1.6429,
+      "step": 12400
+    },
+    {
+      "epoch": 45.96,
+      "grad_norm": 0.652485191822052,
+      "learning_rate": 0.0015403703703703703,
+      "loss": 1.6542,
+      "step": 12410
+    },
+    {
+      "epoch": 46.0,
+      "grad_norm": 1.1145904064178467,
+      "learning_rate": 0.0015400000000000001,
+      "loss": 1.6458,
+      "step": 12420
+    },
+    {
+      "epoch": 46.04,
+      "grad_norm": 0.6264052391052246,
+      "learning_rate": 0.0015396296296296297,
+      "loss": 1.4118,
+      "step": 12430
+    },
+    {
+      "epoch": 46.07,
+      "grad_norm": 0.6265543103218079,
+      "learning_rate": 0.0015392592592592593,
+      "loss": 1.4072,
+      "step": 12440
+    },
+    {
+      "epoch": 46.11,
+      "grad_norm": 0.6246814727783203,
+      "learning_rate": 0.001538888888888889,
+      "loss": 1.4276,
+      "step": 12450
+    },
+    {
+      "epoch": 46.15,
+      "grad_norm": 0.6075099110603333,
+      "learning_rate": 0.0015385185185185187,
+      "loss": 1.4119,
+      "step": 12460
+    },
+    {
+      "epoch": 46.19,
+      "grad_norm": 0.6614938974380493,
+      "learning_rate": 0.001538148148148148,
+      "loss": 1.4499,
+      "step": 12470
+    },
+    {
+      "epoch": 46.22,
+      "grad_norm": 0.6482969522476196,
+      "learning_rate": 0.0015377777777777777,
+      "loss": 1.4508,
+      "step": 12480
+    },
+    {
+      "epoch": 46.26,
+      "grad_norm": 0.61202073097229,
+      "learning_rate": 0.0015374074074074075,
+      "loss": 1.4833,
+      "step": 12490
+    },
+    {
+      "epoch": 46.3,
+      "grad_norm": 0.6306284070014954,
+      "learning_rate": 0.001537037037037037,
+      "loss": 1.5188,
+      "step": 12500
+    },
+    {
+      "epoch": 46.33,
+      "grad_norm": 0.6624696850776672,
+      "learning_rate": 0.0015366666666666666,
+      "loss": 1.4991,
+      "step": 12510
+    },
+    {
+      "epoch": 46.37,
+      "grad_norm": 0.6851840019226074,
+      "learning_rate": 0.0015362962962962962,
+      "loss": 1.5289,
+      "step": 12520
+    },
+    {
+      "epoch": 46.41,
+      "grad_norm": 0.6263905763626099,
+      "learning_rate": 0.001535925925925926,
+      "loss": 1.5316,
+      "step": 12530
+    },
+    {
+      "epoch": 46.44,
+      "grad_norm": 0.588344395160675,
+      "learning_rate": 0.0015355555555555556,
+      "loss": 1.5447,
+      "step": 12540
+    },
+    {
+      "epoch": 46.48,
+      "grad_norm": 0.6301348805427551,
+      "learning_rate": 0.0015351851851851852,
+      "loss": 1.56,
+      "step": 12550
+    },
+    {
+      "epoch": 46.52,
+      "grad_norm": 0.6092745661735535,
+      "learning_rate": 0.0015348148148148148,
+      "loss": 1.5466,
+      "step": 12560
+    },
+    {
+      "epoch": 46.56,
+      "grad_norm": 0.6487380266189575,
+      "learning_rate": 0.0015344444444444446,
+      "loss": 1.5944,
+      "step": 12570
+    },
+    {
+      "epoch": 46.59,
+      "grad_norm": 0.6388218998908997,
+      "learning_rate": 0.0015340740740740742,
+      "loss": 1.5849,
+      "step": 12580
+    },
+    {
+      "epoch": 46.63,
+      "grad_norm": 0.6476507782936096,
+      "learning_rate": 0.0015337037037037038,
+      "loss": 1.541,
+      "step": 12590
+    },
+    {
+      "epoch": 46.67,
+      "grad_norm": 0.622490406036377,
+      "learning_rate": 0.0015333333333333334,
+      "loss": 1.5799,
+      "step": 12600
+    },
+    {
+      "epoch": 46.7,
+      "grad_norm": 0.6342136263847351,
+      "learning_rate": 0.001532962962962963,
+      "loss": 1.5772,
+      "step": 12610
+    },
+    {
+      "epoch": 46.74,
+      "grad_norm": 0.6312634348869324,
+      "learning_rate": 0.0015325925925925926,
+      "loss": 1.5933,
+      "step": 12620
+    },
+    {
+      "epoch": 46.78,
+      "grad_norm": 0.604749858379364,
+      "learning_rate": 0.0015322222222222221,
+      "loss": 1.6059,
+      "step": 12630
+    },
+    {
+      "epoch": 46.81,
+      "grad_norm": 0.6882821917533875,
+      "learning_rate": 0.0015318518518518517,
+      "loss": 1.6056,
+      "step": 12640
+    },
+    {
+      "epoch": 46.85,
+      "grad_norm": 0.656486451625824,
+      "learning_rate": 0.0015314814814814815,
+      "loss": 1.6105,
+      "step": 12650
+    },
+    {
+      "epoch": 46.89,
+      "grad_norm": 0.6267317533493042,
+      "learning_rate": 0.0015311111111111111,
+      "loss": 1.6267,
+      "step": 12660
+    },
+    {
+      "epoch": 46.93,
+      "grad_norm": 0.6135218739509583,
+      "learning_rate": 0.0015307407407407407,
+      "loss": 1.6251,
+      "step": 12670
+    },
+    {
+      "epoch": 46.96,
+      "grad_norm": 0.6487293839454651,
+      "learning_rate": 0.0015303703703703703,
+      "loss": 1.6157,
+      "step": 12680
+    },
+    {
+      "epoch": 47.0,
+      "grad_norm": 1.242255687713623,
+      "learning_rate": 0.0015300000000000001,
+      "loss": 1.6046,
+      "step": 12690
+    },
+    {
+      "epoch": 47.04,
+      "grad_norm": 0.6860459446907043,
+      "learning_rate": 0.0015296296296296297,
+      "loss": 1.3779,
+      "step": 12700
+    },
+    {
+      "epoch": 47.07,
+      "grad_norm": 0.6426923871040344,
+      "learning_rate": 0.0015292592592592593,
+      "loss": 1.4071,
+      "step": 12710
+    },
+    {
+      "epoch": 47.11,
+      "grad_norm": 0.6774601936340332,
+      "learning_rate": 0.001528888888888889,
+      "loss": 1.4033,
+      "step": 12720
+    },
+    {
+      "epoch": 47.15,
+      "grad_norm": 0.6799710392951965,
+      "learning_rate": 0.0015285185185185187,
+      "loss": 1.4038,
+      "step": 12730
+    },
+    {
+      "epoch": 47.19,
+      "grad_norm": 0.6156853437423706,
+      "learning_rate": 0.0015281481481481483,
+      "loss": 1.4496,
+      "step": 12740
+    },
+    {
+      "epoch": 47.22,
+      "grad_norm": 0.6396149396896362,
+      "learning_rate": 0.0015277777777777776,
+      "loss": 1.453,
+      "step": 12750
+    },
+    {
+      "epoch": 47.26,
+      "grad_norm": 0.6574532389640808,
+      "learning_rate": 0.0015274074074074074,
+      "loss": 1.4712,
+      "step": 12760
+    },
+    {
+      "epoch": 47.3,
+      "grad_norm": 0.6416125893592834,
+      "learning_rate": 0.001527037037037037,
+      "loss": 1.4736,
+      "step": 12770
+    },
+    {
+      "epoch": 47.33,
+      "grad_norm": 0.6952414512634277,
+      "learning_rate": 0.0015266666666666666,
+      "loss": 1.4879,
+      "step": 12780
+    },
+    {
+      "epoch": 47.37,
+      "grad_norm": 0.6132145524024963,
+      "learning_rate": 0.0015262962962962962,
+      "loss": 1.5121,
+      "step": 12790
+    },
+    {
+      "epoch": 47.41,
+      "grad_norm": 0.6308304667472839,
+      "learning_rate": 0.001525925925925926,
+      "loss": 1.5141,
+      "step": 12800
+    },
+    {
+      "epoch": 47.44,
+      "grad_norm": 0.6352577209472656,
+      "learning_rate": 0.0015255555555555556,
+      "loss": 1.5348,
+      "step": 12810
+    },
+    {
+      "epoch": 47.48,
+      "grad_norm": 0.624758243560791,
+      "learning_rate": 0.0015251851851851852,
+      "loss": 1.516,
+      "step": 12820
+    },
+    {
+      "epoch": 47.52,
+      "grad_norm": 0.630090594291687,
+      "learning_rate": 0.0015248148148148148,
+      "loss": 1.5644,
+      "step": 12830
+    },
+    {
+      "epoch": 47.56,
+      "grad_norm": 0.6315339803695679,
+      "learning_rate": 0.0015244444444444446,
+      "loss": 1.575,
+      "step": 12840
+    },
+    {
+      "epoch": 47.59,
+      "grad_norm": 0.6333447098731995,
+      "learning_rate": 0.0015240740740740742,
+      "loss": 1.5693,
+      "step": 12850
+    },
+    {
+      "epoch": 47.63,
+      "grad_norm": 0.6272197365760803,
+      "learning_rate": 0.0015237037037037038,
+      "loss": 1.5711,
+      "step": 12860
+    },
+    {
+      "epoch": 47.67,
+      "grad_norm": 0.6363614797592163,
+      "learning_rate": 0.0015233333333333334,
+      "loss": 1.5668,
+      "step": 12870
+    },
+    {
+      "epoch": 47.7,
+      "grad_norm": 0.6725128293037415,
+      "learning_rate": 0.0015229629629629632,
+      "loss": 1.5753,
+      "step": 12880
+    },
+    {
+      "epoch": 47.74,
+      "grad_norm": 0.6107433438301086,
+      "learning_rate": 0.0015225925925925925,
+      "loss": 1.5887,
+      "step": 12890
+    },
+    {
+      "epoch": 47.78,
+      "grad_norm": 0.6690382361412048,
+      "learning_rate": 0.0015222222222222221,
+      "loss": 1.6191,
+      "step": 12900
+    },
+    {
+      "epoch": 47.81,
+      "grad_norm": 0.617404043674469,
+      "learning_rate": 0.001521851851851852,
+      "loss": 1.5842,
+      "step": 12910
+    },
+    {
+      "epoch": 47.85,
+      "grad_norm": 0.6330074667930603,
+      "learning_rate": 0.0015214814814814815,
+      "loss": 1.6184,
+      "step": 12920
+    },
+    {
+      "epoch": 47.89,
+      "grad_norm": 0.6891388297080994,
+      "learning_rate": 0.001521111111111111,
+      "loss": 1.6195,
+      "step": 12930
+    },
+    {
+      "epoch": 47.93,
+      "grad_norm": 0.61796635389328,
+      "learning_rate": 0.0015207407407407407,
+      "loss": 1.6086,
+      "step": 12940
+    },
+    {
+      "epoch": 47.96,
+      "grad_norm": 0.6967139840126038,
+      "learning_rate": 0.0015203703703703705,
+      "loss": 1.613,
+      "step": 12950
+    },
+    {
+      "epoch": 48.0,
+      "grad_norm": 1.0406641960144043,
+      "learning_rate": 0.00152,
+      "loss": 1.6041,
+      "step": 12960
+    },
+    {
+      "epoch": 48.04,
+      "grad_norm": 0.6598649024963379,
+      "learning_rate": 0.0015196296296296297,
+      "loss": 1.3771,
+      "step": 12970
+    },
+    {
+      "epoch": 48.07,
+      "grad_norm": 0.6702351570129395,
+      "learning_rate": 0.0015192592592592593,
+      "loss": 1.3665,
+      "step": 12980
+    },
+    {
+      "epoch": 48.11,
+      "grad_norm": 0.6731034517288208,
+      "learning_rate": 0.001518888888888889,
+      "loss": 1.3969,
+      "step": 12990
+    },
+    {
+      "epoch": 48.15,
+      "grad_norm": 0.6232157349586487,
+      "learning_rate": 0.0015185185185185187,
+      "loss": 1.4008,
+      "step": 13000
+    },
+    {
+      "epoch": 48.19,
+      "grad_norm": 0.6420234441757202,
+      "learning_rate": 0.0015181481481481482,
+      "loss": 1.4486,
+      "step": 13010
+    },
+    {
+      "epoch": 48.22,
+      "grad_norm": 0.6267896890640259,
+      "learning_rate": 0.0015177777777777776,
+      "loss": 1.4501,
+      "step": 13020
+    },
+    {
+      "epoch": 48.26,
+      "grad_norm": 0.6549137234687805,
+      "learning_rate": 0.0015174074074074074,
+      "loss": 1.4482,
+      "step": 13030
+    },
+    {
+      "epoch": 48.3,
+      "grad_norm": 0.7054146528244019,
+      "learning_rate": 0.001517037037037037,
+      "loss": 1.4779,
+      "step": 13040
+    },
+    {
+      "epoch": 48.33,
+      "grad_norm": 0.6636216044425964,
+      "learning_rate": 0.0015166666666666666,
+      "loss": 1.4828,
+      "step": 13050
+    },
+    {
+      "epoch": 48.37,
+      "grad_norm": 0.6946861743927002,
+      "learning_rate": 0.0015162962962962962,
+      "loss": 1.4607,
+      "step": 13060
+    },
+    {
+      "epoch": 48.41,
+      "grad_norm": 0.6364338994026184,
+      "learning_rate": 0.001515925925925926,
+      "loss": 1.5354,
+      "step": 13070
+    },
+    {
+      "epoch": 48.44,
+      "grad_norm": 0.6550884246826172,
+      "learning_rate": 0.0015155555555555556,
+      "loss": 1.5186,
+      "step": 13080
+    },
+    {
+      "epoch": 48.48,
+      "grad_norm": 0.671321451663971,
+      "learning_rate": 0.0015151851851851852,
+      "loss": 1.5092,
+      "step": 13090
+    },
+    {
+      "epoch": 48.52,
+      "grad_norm": 0.6748003959655762,
+      "learning_rate": 0.001514814814814815,
+      "loss": 1.5204,
+      "step": 13100
+    },
+    {
+      "epoch": 48.56,
+      "grad_norm": 0.7088521718978882,
+      "learning_rate": 0.0015144444444444446,
+      "loss": 1.5215,
+      "step": 13110
+    },
+    {
+      "epoch": 48.59,
+      "grad_norm": 0.7026909589767456,
+      "learning_rate": 0.0015140740740740742,
+      "loss": 1.5686,
+      "step": 13120
+    },
+    {
+      "epoch": 48.63,
+      "grad_norm": 0.5976001620292664,
+      "learning_rate": 0.0015137037037037037,
+      "loss": 1.5663,
+      "step": 13130
+    },
+    {
+      "epoch": 48.67,
+      "grad_norm": 0.651646077632904,
+      "learning_rate": 0.0015133333333333335,
+      "loss": 1.5718,
+      "step": 13140
+    },
+    {
+      "epoch": 48.7,
+      "grad_norm": 0.664674699306488,
+      "learning_rate": 0.0015129629629629631,
+      "loss": 1.5884,
+      "step": 13150
+    },
+    {
+      "epoch": 48.74,
+      "grad_norm": 0.6666598320007324,
+      "learning_rate": 0.0015125925925925925,
+      "loss": 1.5799,
+      "step": 13160
+    },
+    {
+      "epoch": 48.78,
+      "grad_norm": 0.6095290780067444,
+      "learning_rate": 0.001512222222222222,
+      "loss": 1.5815,
+      "step": 13170
+    },
+    {
+      "epoch": 48.81,
+      "grad_norm": 0.6643814444541931,
+      "learning_rate": 0.001511851851851852,
+      "loss": 1.6082,
+      "step": 13180
+    },
+    {
+      "epoch": 48.85,
+      "grad_norm": 0.6361218094825745,
+      "learning_rate": 0.0015114814814814815,
+      "loss": 1.6041,
+      "step": 13190
+    },
+    {
+      "epoch": 48.89,
+      "grad_norm": 0.6308338642120361,
+      "learning_rate": 0.001511111111111111,
+      "loss": 1.5788,
+      "step": 13200
+    },
+    {
+      "epoch": 48.93,
+      "grad_norm": 0.6339203119277954,
+      "learning_rate": 0.0015107407407407407,
+      "loss": 1.6344,
+      "step": 13210
+    },
+    {
+      "epoch": 48.96,
+      "grad_norm": 0.6878967881202698,
+      "learning_rate": 0.0015103703703703705,
+      "loss": 1.6565,
+      "step": 13220
+    },
+    {
+      "epoch": 49.0,
+      "grad_norm": 1.5557820796966553,
+      "learning_rate": 0.00151,
+      "loss": 1.5785,
+      "step": 13230
+    },
+    {
+      "epoch": 49.04,
+      "grad_norm": 0.6234275102615356,
+      "learning_rate": 0.0015096296296296296,
+      "loss": 1.3406,
+      "step": 13240
+    },
+    {
+      "epoch": 49.07,
+      "grad_norm": 0.6727663278579712,
+      "learning_rate": 0.0015092592592592592,
+      "loss": 1.3492,
+      "step": 13250
+    },
+    {
+      "epoch": 49.11,
+      "grad_norm": 0.6145593523979187,
+      "learning_rate": 0.001508888888888889,
+      "loss": 1.3564,
+      "step": 13260
+    },
+    {
+      "epoch": 49.15,
+      "grad_norm": 0.7036429047584534,
+      "learning_rate": 0.0015085185185185186,
+      "loss": 1.378,
+      "step": 13270
+    },
+    {
+      "epoch": 49.19,
+      "grad_norm": 0.6589803695678711,
+      "learning_rate": 0.0015081481481481482,
+      "loss": 1.4108,
+      "step": 13280
+    },
+    {
+      "epoch": 49.22,
+      "grad_norm": 0.6215453743934631,
+      "learning_rate": 0.001507777777777778,
+      "loss": 1.4335,
+      "step": 13290
+    },
+    {
+      "epoch": 49.26,
+      "grad_norm": 0.6513487100601196,
+      "learning_rate": 0.0015074074074074074,
+      "loss": 1.4713,
+      "step": 13300
+    },
+    {
+      "epoch": 49.3,
+      "grad_norm": 0.6613112688064575,
+      "learning_rate": 0.001507037037037037,
+      "loss": 1.4948,
+      "step": 13310
+    },
+    {
+      "epoch": 49.33,
+      "grad_norm": 0.6666290760040283,
+      "learning_rate": 0.0015066666666666666,
+      "loss": 1.4837,
+      "step": 13320
+    },
+    {
+      "epoch": 49.37,
+      "grad_norm": 0.6554436087608337,
+      "learning_rate": 0.0015062962962962964,
+      "loss": 1.4909,
+      "step": 13330
+    },
+    {
+      "epoch": 49.41,
+      "grad_norm": 0.635359525680542,
+      "learning_rate": 0.001505925925925926,
+      "loss": 1.4803,
+      "step": 13340
+    },
+    {
+      "epoch": 49.44,
+      "grad_norm": 0.6472628712654114,
+      "learning_rate": 0.0015055555555555556,
+      "loss": 1.5059,
+      "step": 13350
+    },
+    {
+      "epoch": 49.48,
+      "grad_norm": 0.6820836663246155,
+      "learning_rate": 0.0015051851851851851,
+      "loss": 1.5074,
+      "step": 13360
+    },
+    {
+      "epoch": 49.52,
+      "grad_norm": 0.6269857287406921,
+      "learning_rate": 0.001504814814814815,
+      "loss": 1.5613,
+      "step": 13370
+    },
+    {
+      "epoch": 49.56,
+      "grad_norm": 0.6553345322608948,
+      "learning_rate": 0.0015044444444444445,
+      "loss": 1.5285,
+      "step": 13380
+    },
+    {
+      "epoch": 49.59,
+      "grad_norm": 0.7129311561584473,
+      "learning_rate": 0.0015040740740740741,
+      "loss": 1.5444,
+      "step": 13390
+    },
+    {
+      "epoch": 49.63,
+      "grad_norm": 0.6317949891090393,
+      "learning_rate": 0.0015037037037037037,
+      "loss": 1.5464,
+      "step": 13400
+    },
+    {
+      "epoch": 49.67,
+      "grad_norm": 0.6542783975601196,
+      "learning_rate": 0.0015033333333333335,
+      "loss": 1.5649,
+      "step": 13410
+    },
+    {
+      "epoch": 49.7,
+      "grad_norm": 0.645422101020813,
+      "learning_rate": 0.001502962962962963,
+      "loss": 1.5832,
+      "step": 13420
+    },
+    {
+      "epoch": 49.74,
+      "grad_norm": 0.6377230286598206,
+      "learning_rate": 0.0015025925925925927,
+      "loss": 1.6086,
+      "step": 13430
+    },
+    {
+      "epoch": 49.78,
+      "grad_norm": 0.6812024712562561,
+      "learning_rate": 0.001502222222222222,
+      "loss": 1.5837,
+      "step": 13440
+    },
+    {
+      "epoch": 49.81,
+      "grad_norm": 0.666580855846405,
+      "learning_rate": 0.0015018518518518519,
+      "loss": 1.5813,
+      "step": 13450
+    },
+    {
+      "epoch": 49.85,
+      "grad_norm": 0.7089651226997375,
+      "learning_rate": 0.0015014814814814815,
+      "loss": 1.5876,
+      "step": 13460
+    },
+    {
+      "epoch": 49.89,
+      "grad_norm": 0.6948536038398743,
+      "learning_rate": 0.001501111111111111,
+      "loss": 1.5941,
+      "step": 13470
+    },
+    {
+      "epoch": 49.93,
+      "grad_norm": 0.6506425738334656,
+      "learning_rate": 0.0015007407407407406,
+      "loss": 1.5814,
+      "step": 13480
+    },
+    {
+      "epoch": 49.96,
+      "grad_norm": 0.6823866963386536,
+      "learning_rate": 0.0015003703703703704,
+      "loss": 1.6191,
+      "step": 13490
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 1.197648286819458,
+      "learning_rate": 0.0015,
+      "loss": 1.6013,
+      "step": 13500
+    },
+    {
+      "epoch": 50.04,
+      "grad_norm": 0.6759616732597351,
+      "learning_rate": 0.0014996296296296296,
+      "loss": 1.3345,
+      "step": 13510
+    },
+    {
+      "epoch": 50.07,
+      "grad_norm": 0.6753141283988953,
+      "learning_rate": 0.0014992592592592594,
+      "loss": 1.3395,
+      "step": 13520
+    },
+    {
+      "epoch": 50.11,
+      "grad_norm": 0.6210965514183044,
+      "learning_rate": 0.001498888888888889,
+      "loss": 1.3955,
+      "step": 13530
+    },
+    {
+      "epoch": 50.15,
+      "grad_norm": 0.6821613907814026,
+      "learning_rate": 0.0014985185185185186,
+      "loss": 1.4198,
+      "step": 13540
+    },
+    {
+      "epoch": 50.19,
+      "grad_norm": 0.7067998647689819,
+      "learning_rate": 0.0014981481481481482,
+      "loss": 1.4207,
+      "step": 13550
+    },
+    {
+      "epoch": 50.22,
+      "grad_norm": 0.6579769849777222,
+      "learning_rate": 0.001497777777777778,
+      "loss": 1.4274,
+      "step": 13560
+    },
+    {
+      "epoch": 50.26,
+      "grad_norm": 0.666738748550415,
+      "learning_rate": 0.0014974074074074074,
+      "loss": 1.4537,
+      "step": 13570
+    },
+    {
+      "epoch": 50.3,
+      "grad_norm": 0.7158663868904114,
+      "learning_rate": 0.001497037037037037,
+      "loss": 1.476,
+      "step": 13580
+    },
+    {
+      "epoch": 50.33,
+      "grad_norm": 0.6293248534202576,
+      "learning_rate": 0.0014966666666666665,
+      "loss": 1.476,
+      "step": 13590
+    },
+    {
+      "epoch": 50.37,
+      "grad_norm": 0.7110118865966797,
+      "learning_rate": 0.0014962962962962963,
+      "loss": 1.5088,
+      "step": 13600
+    },
+    {
+      "epoch": 50.41,
+      "grad_norm": 0.6568295955657959,
+      "learning_rate": 0.001495925925925926,
+      "loss": 1.4764,
+      "step": 13610
+    },
+    {
+      "epoch": 50.44,
+      "grad_norm": 0.6455584764480591,
+      "learning_rate": 0.0014955555555555555,
+      "loss": 1.4821,
+      "step": 13620
+    },
+    {
+      "epoch": 50.48,
+      "grad_norm": 0.6564566493034363,
+      "learning_rate": 0.0014951851851851851,
+      "loss": 1.4976,
+      "step": 13630
+    },
+    {
+      "epoch": 50.52,
+      "grad_norm": 0.6496867537498474,
+      "learning_rate": 0.001494814814814815,
+      "loss": 1.5238,
+      "step": 13640
+    },
+    {
+      "epoch": 50.56,
+      "grad_norm": 0.6984143853187561,
+      "learning_rate": 0.0014944444444444445,
+      "loss": 1.5487,
+      "step": 13650
+    },
+    {
+      "epoch": 50.59,
+      "grad_norm": 0.6508861184120178,
+      "learning_rate": 0.001494074074074074,
+      "loss": 1.5372,
+      "step": 13660
+    },
+    {
+      "epoch": 50.63,
+      "grad_norm": 0.670586109161377,
+      "learning_rate": 0.0014937037037037037,
+      "loss": 1.5353,
+      "step": 13670
+    },
+    {
+      "epoch": 50.67,
+      "grad_norm": 0.6368412971496582,
+      "learning_rate": 0.0014933333333333335,
+      "loss": 1.5751,
+      "step": 13680
+    },
+    {
+      "epoch": 50.7,
+      "grad_norm": 0.6989796757698059,
+      "learning_rate": 0.001492962962962963,
+      "loss": 1.5406,
+      "step": 13690
+    },
+    {
+      "epoch": 50.74,
+      "grad_norm": 0.6799941658973694,
+      "learning_rate": 0.0014925925925925927,
+      "loss": 1.5596,
+      "step": 13700
+    },
+    {
+      "epoch": 50.78,
+      "grad_norm": 0.6923843622207642,
+      "learning_rate": 0.001492222222222222,
+      "loss": 1.5707,
+      "step": 13710
+    },
+    {
+      "epoch": 50.81,
+      "grad_norm": 0.6336764097213745,
+      "learning_rate": 0.0014918518518518518,
+      "loss": 1.5833,
+      "step": 13720
+    },
+    {
+      "epoch": 50.85,
+      "grad_norm": 0.6471168398857117,
+      "learning_rate": 0.0014914814814814814,
+      "loss": 1.5799,
+      "step": 13730
+    },
+    {
+      "epoch": 50.89,
+      "grad_norm": 0.6990411877632141,
+      "learning_rate": 0.001491111111111111,
+      "loss": 1.5772,
+      "step": 13740
+    },
+    {
+      "epoch": 50.93,
+      "grad_norm": 0.629309356212616,
+      "learning_rate": 0.0014907407407407408,
+      "loss": 1.6318,
+      "step": 13750
+    },
+    {
+      "epoch": 50.96,
+      "grad_norm": 0.6218940615653992,
+      "learning_rate": 0.0014903703703703704,
+      "loss": 1.6012,
+      "step": 13760
+    },
+    {
+      "epoch": 51.0,
+      "grad_norm": 1.1239666938781738,
+      "learning_rate": 0.00149,
+      "loss": 1.5763,
+      "step": 13770
+    },
+    {
+      "epoch": 51.04,
+      "grad_norm": 0.6716301441192627,
+      "learning_rate": 0.0014896296296296296,
+      "loss": 1.3228,
+      "step": 13780
+    },
+    {
+      "epoch": 51.07,
+      "grad_norm": 0.6490216851234436,
+      "learning_rate": 0.0014892592592592594,
+      "loss": 1.3701,
+      "step": 13790
+    },
+    {
+      "epoch": 51.11,
+      "grad_norm": 0.6851688623428345,
+      "learning_rate": 0.001488888888888889,
+      "loss": 1.3892,
+      "step": 13800
+    },
+    {
+      "epoch": 51.15,
+      "grad_norm": 0.6652293801307678,
+      "learning_rate": 0.0014885185185185186,
+      "loss": 1.3798,
+      "step": 13810
+    },
+    {
+      "epoch": 51.19,
+      "grad_norm": 0.6425241231918335,
+      "learning_rate": 0.0014881481481481482,
+      "loss": 1.4337,
+      "step": 13820
+    },
+    {
+      "epoch": 51.22,
+      "grad_norm": 0.6521861553192139,
+      "learning_rate": 0.001487777777777778,
+      "loss": 1.4287,
+      "step": 13830
+    },
+    {
+      "epoch": 51.26,
+      "grad_norm": 0.6515931487083435,
+      "learning_rate": 0.0014874074074074076,
+      "loss": 1.4422,
+      "step": 13840
+    },
+    {
+      "epoch": 51.3,
+      "grad_norm": 0.6525137424468994,
+      "learning_rate": 0.001487037037037037,
+      "loss": 1.4593,
+      "step": 13850
+    },
+    {
+      "epoch": 51.33,
+      "grad_norm": 0.615257203578949,
+      "learning_rate": 0.0014866666666666665,
+      "loss": 1.4458,
+      "step": 13860
+    },
+    {
+      "epoch": 51.37,
+      "grad_norm": 0.7334824800491333,
+      "learning_rate": 0.0014862962962962963,
+      "loss": 1.4622,
+      "step": 13870
+    },
+    {
+      "epoch": 51.41,
+      "grad_norm": 0.6889981031417847,
+      "learning_rate": 0.001485925925925926,
+      "loss": 1.5,
+      "step": 13880
+    },
+    {
+      "epoch": 51.44,
+      "grad_norm": 0.6559149026870728,
+      "learning_rate": 0.0014855555555555555,
+      "loss": 1.4843,
+      "step": 13890
+    },
+    {
+      "epoch": 51.48,
+      "grad_norm": 0.6670100688934326,
+      "learning_rate": 0.001485185185185185,
+      "loss": 1.5008,
+      "step": 13900
+    },
+    {
+      "epoch": 51.52,
+      "grad_norm": 0.6622434258460999,
+      "learning_rate": 0.001484814814814815,
+      "loss": 1.5177,
+      "step": 13910
+    },
+    {
+      "epoch": 51.56,
+      "grad_norm": 0.6639595627784729,
+      "learning_rate": 0.0014844444444444445,
+      "loss": 1.5215,
+      "step": 13920
+    },
+    {
+      "epoch": 51.59,
+      "grad_norm": 0.6925778388977051,
+      "learning_rate": 0.001484074074074074,
+      "loss": 1.5358,
+      "step": 13930
+    },
+    {
+      "epoch": 51.63,
+      "grad_norm": 0.6861134171485901,
+      "learning_rate": 0.0014837037037037039,
+      "loss": 1.566,
+      "step": 13940
+    },
+    {
+      "epoch": 51.67,
+      "grad_norm": 0.6408432126045227,
+      "learning_rate": 0.0014833333333333335,
+      "loss": 1.5612,
+      "step": 13950
+    },
+    {
+      "epoch": 51.7,
+      "grad_norm": 0.679660439491272,
+      "learning_rate": 0.001482962962962963,
+      "loss": 1.558,
+      "step": 13960
+    },
+    {
+      "epoch": 51.74,
+      "grad_norm": 0.6439672708511353,
+      "learning_rate": 0.0014825925925925926,
+      "loss": 1.5728,
+      "step": 13970
+    },
+    {
+      "epoch": 51.78,
+      "grad_norm": 0.6197627782821655,
+      "learning_rate": 0.0014822222222222224,
+      "loss": 1.5413,
+      "step": 13980
+    },
+    {
+      "epoch": 51.81,
+      "grad_norm": 0.6940560340881348,
+      "learning_rate": 0.0014818518518518518,
+      "loss": 1.5327,
+      "step": 13990
+    },
+    {
+      "epoch": 51.85,
+      "grad_norm": 0.6489192843437195,
+      "learning_rate": 0.0014814814814814814,
+      "loss": 1.5833,
+      "step": 14000
+    },
+    {
+      "epoch": 51.89,
+      "grad_norm": 0.7189292907714844,
+      "learning_rate": 0.001481111111111111,
+      "loss": 1.5644,
+      "step": 14010
+    },
+    {
+      "epoch": 51.93,
+      "grad_norm": 0.7071329951286316,
+      "learning_rate": 0.0014807407407407408,
+      "loss": 1.5661,
+      "step": 14020
+    },
+    {
+      "epoch": 51.96,
+      "grad_norm": 0.6607621908187866,
+      "learning_rate": 0.0014803703703703704,
+      "loss": 1.5743,
+      "step": 14030
+    },
+    {
+      "epoch": 52.0,
+      "grad_norm": 1.4042985439300537,
+      "learning_rate": 0.00148,
+      "loss": 1.5808,
+      "step": 14040
+    },
+    {
+      "epoch": 52.04,
+      "grad_norm": 0.625298798084259,
+      "learning_rate": 0.0014796296296296296,
+      "loss": 1.3482,
+      "step": 14050
+    },
+    {
+      "epoch": 52.07,
+      "grad_norm": 0.6287106871604919,
+      "learning_rate": 0.0014792592592592594,
+      "loss": 1.3591,
+      "step": 14060
+    },
+    {
+      "epoch": 52.11,
+      "grad_norm": 0.6880126595497131,
+      "learning_rate": 0.001478888888888889,
+      "loss": 1.3634,
+      "step": 14070
+    },
+    {
+      "epoch": 52.15,
+      "grad_norm": 0.7166730761528015,
+      "learning_rate": 0.0014785185185185185,
+      "loss": 1.3566,
+      "step": 14080
+    },
+    {
+      "epoch": 52.19,
+      "grad_norm": 0.7044410109519958,
+      "learning_rate": 0.0014781481481481481,
+      "loss": 1.3779,
+      "step": 14090
+    },
+    {
+      "epoch": 52.22,
+      "grad_norm": 0.7081303000450134,
+      "learning_rate": 0.001477777777777778,
+      "loss": 1.4112,
+      "step": 14100
+    },
+    {
+      "epoch": 52.26,
+      "grad_norm": 0.7148960828781128,
+      "learning_rate": 0.0014774074074074075,
+      "loss": 1.4383,
+      "step": 14110
+    },
+    {
+      "epoch": 52.3,
+      "grad_norm": 0.722838282585144,
+      "learning_rate": 0.001477037037037037,
+      "loss": 1.4278,
+      "step": 14120
+    },
+    {
+      "epoch": 52.33,
+      "grad_norm": 0.6913270354270935,
+      "learning_rate": 0.0014766666666666667,
+      "loss": 1.4577,
+      "step": 14130
+    },
+    {
+      "epoch": 52.37,
+      "grad_norm": 0.6681016087532043,
+      "learning_rate": 0.0014762962962962963,
+      "loss": 1.4674,
+      "step": 14140
+    },
+    {
+      "epoch": 52.41,
+      "grad_norm": 0.7427762746810913,
+      "learning_rate": 0.0014759259259259259,
+      "loss": 1.4696,
+      "step": 14150
+    },
+    {
+      "epoch": 52.44,
+      "grad_norm": 0.744890034198761,
+      "learning_rate": 0.0014755555555555555,
+      "loss": 1.5042,
+      "step": 14160
+    },
+    {
+      "epoch": 52.48,
+      "grad_norm": 0.7112950682640076,
+      "learning_rate": 0.0014751851851851853,
+      "loss": 1.4885,
+      "step": 14170
+    },
+    {
+      "epoch": 52.52,
+      "grad_norm": 0.6994560956954956,
+      "learning_rate": 0.0014748148148148149,
+      "loss": 1.527,
+      "step": 14180
+    },
+    {
+      "epoch": 52.56,
+      "grad_norm": 0.6381304264068604,
+      "learning_rate": 0.0014744444444444445,
+      "loss": 1.5282,
+      "step": 14190
+    },
+    {
+      "epoch": 52.59,
+      "grad_norm": 0.6289535164833069,
+      "learning_rate": 0.001474074074074074,
+      "loss": 1.5379,
+      "step": 14200
+    },
+    {
+      "epoch": 52.63,
+      "grad_norm": 0.7018424868583679,
+      "learning_rate": 0.0014737037037037039,
+      "loss": 1.5422,
+      "step": 14210
+    },
+    {
+      "epoch": 52.67,
+      "grad_norm": 0.6578280329704285,
+      "learning_rate": 0.0014733333333333334,
+      "loss": 1.5505,
+      "step": 14220
+    },
+    {
+      "epoch": 52.7,
+      "grad_norm": 0.6513351202011108,
+      "learning_rate": 0.001472962962962963,
+      "loss": 1.533,
+      "step": 14230
+    },
+    {
+      "epoch": 52.74,
+      "grad_norm": 0.6813905835151672,
+      "learning_rate": 0.0014725925925925926,
+      "loss": 1.5365,
+      "step": 14240
+    },
+    {
+      "epoch": 52.78,
+      "grad_norm": 0.7050051093101501,
+      "learning_rate": 0.0014722222222222224,
+      "loss": 1.5358,
+      "step": 14250
+    },
+    {
+      "epoch": 52.81,
+      "grad_norm": 0.6147084832191467,
+      "learning_rate": 0.0014718518518518518,
+      "loss": 1.5515,
+      "step": 14260
+    },
+    {
+      "epoch": 52.85,
+      "grad_norm": 0.6574007868766785,
+      "learning_rate": 0.0014714814814814814,
+      "loss": 1.5587,
+      "step": 14270
+    },
+    {
+      "epoch": 52.89,
+      "grad_norm": 0.74244225025177,
+      "learning_rate": 0.001471111111111111,
+      "loss": 1.5827,
+      "step": 14280
+    },
+    {
+      "epoch": 52.93,
+      "grad_norm": 0.6948452591896057,
+      "learning_rate": 0.0014707407407407408,
+      "loss": 1.5675,
+      "step": 14290
+    },
+    {
+      "epoch": 52.96,
+      "grad_norm": 0.6986340880393982,
+      "learning_rate": 0.0014703703703703704,
+      "loss": 1.5783,
+      "step": 14300
+    },
+    {
+      "epoch": 53.0,
+      "grad_norm": 1.0962846279144287,
+      "learning_rate": 0.00147,
+      "loss": 1.5808,
+      "step": 14310
+    },
+    {
+      "epoch": 53.04,
+      "grad_norm": 0.6880574822425842,
+      "learning_rate": 0.0014696296296296298,
+      "loss": 1.3356,
+      "step": 14320
+    },
+    {
+      "epoch": 53.07,
+      "grad_norm": 0.6791065335273743,
+      "learning_rate": 0.0014692592592592593,
+      "loss": 1.3505,
+      "step": 14330
+    },
+    {
+      "epoch": 53.11,
+      "grad_norm": 0.6797129511833191,
+      "learning_rate": 0.001468888888888889,
+      "loss": 1.3387,
+      "step": 14340
+    },
+    {
+      "epoch": 53.15,
+      "grad_norm": 0.6955825686454773,
+      "learning_rate": 0.0014685185185185185,
+      "loss": 1.3706,
+      "step": 14350
+    },
+    {
+      "epoch": 53.19,
+      "grad_norm": 0.6407482028007507,
+      "learning_rate": 0.0014681481481481483,
+      "loss": 1.3851,
+      "step": 14360
+    },
+    {
+      "epoch": 53.22,
+      "grad_norm": 0.7131444215774536,
+      "learning_rate": 0.001467777777777778,
+      "loss": 1.4148,
+      "step": 14370
+    },
+    {
+      "epoch": 53.26,
+      "grad_norm": 0.7091857194900513,
+      "learning_rate": 0.0014674074074074075,
+      "loss": 1.438,
+      "step": 14380
+    },
+    {
+      "epoch": 53.3,
+      "grad_norm": 0.6437903046607971,
+      "learning_rate": 0.001467037037037037,
+      "loss": 1.4216,
+      "step": 14390
+    },
+    {
+      "epoch": 53.33,
+      "grad_norm": 0.6952517628669739,
+      "learning_rate": 0.0014666666666666667,
+      "loss": 1.4454,
+      "step": 14400
+    },
+    {
+      "epoch": 53.37,
+      "grad_norm": 0.6853577494621277,
+      "learning_rate": 0.0014662962962962963,
+      "loss": 1.4695,
+      "step": 14410
+    },
+    {
+      "epoch": 53.41,
+      "grad_norm": 0.683782696723938,
+      "learning_rate": 0.0014659259259259259,
+      "loss": 1.4711,
+      "step": 14420
+    },
+    {
+      "epoch": 53.44,
+      "grad_norm": 0.6749939918518066,
+      "learning_rate": 0.0014655555555555554,
+      "loss": 1.4837,
+      "step": 14430
+    },
+    {
+      "epoch": 53.48,
+      "grad_norm": 0.6679930090904236,
+      "learning_rate": 0.0014651851851851853,
+      "loss": 1.4844,
+      "step": 14440
+    },
+    {
+      "epoch": 53.52,
+      "grad_norm": 0.7036862373352051,
+      "learning_rate": 0.0014648148148148148,
+      "loss": 1.478,
+      "step": 14450
+    },
+    {
+      "epoch": 53.56,
+      "grad_norm": 0.6320893168449402,
+      "learning_rate": 0.0014644444444444444,
+      "loss": 1.5058,
+      "step": 14460
+    },
+    {
+      "epoch": 53.59,
+      "grad_norm": 0.6774411797523499,
+      "learning_rate": 0.001464074074074074,
+      "loss": 1.5123,
+      "step": 14470
+    },
+    {
+      "epoch": 53.63,
+      "grad_norm": 0.7401627898216248,
+      "learning_rate": 0.0014637037037037038,
+      "loss": 1.5269,
+      "step": 14480
+    },
+    {
+      "epoch": 53.67,
+      "grad_norm": 0.6697722673416138,
+      "learning_rate": 0.0014633333333333334,
+      "loss": 1.5253,
+      "step": 14490
+    },
+    {
+      "epoch": 53.7,
+      "grad_norm": 0.7028366923332214,
+      "learning_rate": 0.001462962962962963,
+      "loss": 1.5125,
+      "step": 14500
+    },
+    {
+      "epoch": 53.74,
+      "grad_norm": 0.6694210171699524,
+      "learning_rate": 0.0014625925925925926,
+      "loss": 1.5426,
+      "step": 14510
+    },
+    {
+      "epoch": 53.78,
+      "grad_norm": 0.6523129343986511,
+      "learning_rate": 0.0014622222222222224,
+      "loss": 1.5194,
+      "step": 14520
+    },
+    {
+      "epoch": 53.81,
+      "grad_norm": 0.6751163005828857,
+      "learning_rate": 0.001461851851851852,
+      "loss": 1.5516,
+      "step": 14530
+    },
+    {
+      "epoch": 53.85,
+      "grad_norm": 0.6803255081176758,
+      "learning_rate": 0.0014614814814814814,
+      "loss": 1.5671,
+      "step": 14540
+    },
+    {
+      "epoch": 53.89,
+      "grad_norm": 0.7053388357162476,
+      "learning_rate": 0.0014611111111111112,
+      "loss": 1.5692,
+      "step": 14550
+    },
+    {
+      "epoch": 53.93,
+      "grad_norm": 0.6673864722251892,
+      "learning_rate": 0.0014607407407407407,
+      "loss": 1.581,
+      "step": 14560
+    },
+    {
+      "epoch": 53.96,
+      "grad_norm": 0.6886114478111267,
+      "learning_rate": 0.0014603703703703703,
+      "loss": 1.5639,
+      "step": 14570
+    },
+    {
+      "epoch": 54.0,
+      "grad_norm": 1.3442867994308472,
+      "learning_rate": 0.00146,
+      "loss": 1.5796,
+      "step": 14580
+    },
+    {
+      "epoch": 54.04,
+      "grad_norm": 0.669252336025238,
+      "learning_rate": 0.0014596296296296297,
+      "loss": 1.3192,
+      "step": 14590
+    },
+    {
+      "epoch": 54.07,
+      "grad_norm": 0.6675966382026672,
+      "learning_rate": 0.0014592592592592593,
+      "loss": 1.353,
+      "step": 14600
+    },
+    {
+      "epoch": 54.11,
+      "grad_norm": 0.7156364321708679,
+      "learning_rate": 0.001458888888888889,
+      "loss": 1.3499,
+      "step": 14610
+    },
+    {
+      "epoch": 54.15,
+      "grad_norm": 0.7250548005104065,
+      "learning_rate": 0.0014585185185185185,
+      "loss": 1.3749,
+      "step": 14620
+    },
+    {
+      "epoch": 54.19,
+      "grad_norm": 0.6688981652259827,
+      "learning_rate": 0.0014581481481481483,
+      "loss": 1.3711,
+      "step": 14630
+    },
+    {
+      "epoch": 54.22,
+      "grad_norm": 0.7125627398490906,
+      "learning_rate": 0.0014577777777777779,
+      "loss": 1.3894,
+      "step": 14640
+    },
+    {
+      "epoch": 54.26,
+      "grad_norm": 0.6876236796379089,
+      "learning_rate": 0.0014574074074074075,
+      "loss": 1.3956,
+      "step": 14650
+    },
+    {
+      "epoch": 54.3,
+      "grad_norm": 0.7455447912216187,
+      "learning_rate": 0.001457037037037037,
+      "loss": 1.4206,
+      "step": 14660
+    },
+    {
+      "epoch": 54.33,
+      "grad_norm": 0.7188883423805237,
+      "learning_rate": 0.0014566666666666669,
+      "loss": 1.418,
+      "step": 14670
+    },
+    {
+      "epoch": 54.37,
+      "grad_norm": 0.6575316190719604,
+      "learning_rate": 0.0014562962962962962,
+      "loss": 1.4615,
+      "step": 14680
+    },
+    {
+      "epoch": 54.41,
+      "grad_norm": 0.6895075440406799,
+      "learning_rate": 0.0014559259259259258,
+      "loss": 1.4702,
+      "step": 14690
+    },
+    {
+      "epoch": 54.44,
+      "grad_norm": 0.7062181830406189,
+      "learning_rate": 0.0014555555555555554,
+      "loss": 1.5067,
+      "step": 14700
+    },
+    {
+      "epoch": 54.48,
+      "grad_norm": 0.7163558602333069,
+      "learning_rate": 0.0014551851851851852,
+      "loss": 1.4755,
+      "step": 14710
+    },
+    {
+      "epoch": 54.52,
+      "grad_norm": 0.644522488117218,
+      "learning_rate": 0.0014548148148148148,
+      "loss": 1.4929,
+      "step": 14720
+    },
+    {
+      "epoch": 54.56,
+      "grad_norm": 0.7128614187240601,
+      "learning_rate": 0.0014544444444444444,
+      "loss": 1.5137,
+      "step": 14730
+    },
+    {
+      "epoch": 54.59,
+      "grad_norm": 0.6865217685699463,
+      "learning_rate": 0.0014540740740740742,
+      "loss": 1.4744,
+      "step": 14740
+    },
+    {
+      "epoch": 54.63,
+      "grad_norm": 0.6782078742980957,
+      "learning_rate": 0.0014537037037037038,
+      "loss": 1.5026,
+      "step": 14750
+    },
+    {
+      "epoch": 54.67,
+      "grad_norm": 0.6496673226356506,
+      "learning_rate": 0.0014533333333333334,
+      "loss": 1.5087,
+      "step": 14760
+    },
+    {
+      "epoch": 54.7,
+      "grad_norm": 0.7069528102874756,
+      "learning_rate": 0.001452962962962963,
+      "loss": 1.5458,
+      "step": 14770
+    },
+    {
+      "epoch": 54.74,
+      "grad_norm": 0.6799771785736084,
+      "learning_rate": 0.0014525925925925928,
+      "loss": 1.5454,
+      "step": 14780
+    },
+    {
+      "epoch": 54.78,
+      "grad_norm": 0.6214430928230286,
+      "learning_rate": 0.0014522222222222224,
+      "loss": 1.5295,
+      "step": 14790
+    },
+    {
+      "epoch": 54.81,
+      "grad_norm": 0.6988822221755981,
+      "learning_rate": 0.001451851851851852,
+      "loss": 1.5521,
+      "step": 14800
+    },
+    {
+      "epoch": 54.85,
+      "grad_norm": 0.7073298096656799,
+      "learning_rate": 0.0014514814814814813,
+      "loss": 1.5719,
+      "step": 14810
+    },
+    {
+      "epoch": 54.89,
+      "grad_norm": 0.6836145520210266,
+      "learning_rate": 0.0014511111111111111,
+      "loss": 1.5707,
+      "step": 14820
+    },
+    {
+      "epoch": 54.93,
+      "grad_norm": 0.6364758014678955,
+      "learning_rate": 0.0014507407407407407,
+      "loss": 1.5669,
+      "step": 14830
+    },
+    {
+      "epoch": 54.96,
+      "grad_norm": 0.7107850909233093,
+      "learning_rate": 0.0014503703703703703,
+      "loss": 1.5315,
+      "step": 14840
+    },
+    {
+      "epoch": 55.0,
+      "grad_norm": 1.1641267538070679,
+      "learning_rate": 0.00145,
+      "loss": 1.5796,
+      "step": 14850
+    },
+    {
+      "epoch": 55.04,
+      "grad_norm": 0.7165002822875977,
+      "learning_rate": 0.0014496296296296297,
+      "loss": 1.3244,
+      "step": 14860
+    },
+    {
+      "epoch": 55.07,
+      "grad_norm": 0.7347918152809143,
+      "learning_rate": 0.0014492592592592593,
+      "loss": 1.3162,
+      "step": 14870
+    },
+    {
+      "epoch": 55.11,
+      "grad_norm": 0.623955488204956,
+      "learning_rate": 0.0014488888888888889,
+      "loss": 1.3492,
+      "step": 14880
+    },
+    {
+      "epoch": 55.15,
+      "grad_norm": 0.6705119609832764,
+      "learning_rate": 0.0014485185185185185,
+      "loss": 1.3307,
+      "step": 14890
+    },
+    {
+      "epoch": 55.19,
+      "grad_norm": 0.6863866448402405,
+      "learning_rate": 0.0014481481481481483,
+      "loss": 1.4045,
+      "step": 14900
+    },
+    {
+      "epoch": 55.22,
+      "grad_norm": 0.727873682975769,
+      "learning_rate": 0.0014477777777777779,
+      "loss": 1.3976,
+      "step": 14910
+    },
+    {
+      "epoch": 55.26,
+      "grad_norm": 0.6954161524772644,
+      "learning_rate": 0.0014474074074074075,
+      "loss": 1.4039,
+      "step": 14920
+    },
+    {
+      "epoch": 55.3,
+      "grad_norm": 0.6985371112823486,
+      "learning_rate": 0.0014470370370370373,
+      "loss": 1.4561,
+      "step": 14930
+    },
+    {
+      "epoch": 55.33,
+      "grad_norm": 0.6702157258987427,
+      "learning_rate": 0.0014466666666666668,
+      "loss": 1.4516,
+      "step": 14940
+    },
+    {
+      "epoch": 55.37,
+      "grad_norm": 0.6751401424407959,
+      "learning_rate": 0.0014462962962962962,
+      "loss": 1.4431,
+      "step": 14950
+    },
+    {
+      "epoch": 55.41,
+      "grad_norm": 0.6904053092002869,
+      "learning_rate": 0.0014459259259259258,
+      "loss": 1.472,
+      "step": 14960
+    },
+    {
+      "epoch": 55.44,
+      "grad_norm": 0.6829718351364136,
+      "learning_rate": 0.0014455555555555556,
+      "loss": 1.4178,
+      "step": 14970
+    },
+    {
+      "epoch": 55.48,
+      "grad_norm": 0.7579282522201538,
+      "learning_rate": 0.0014451851851851852,
+      "loss": 1.4848,
+      "step": 14980
+    },
+    {
+      "epoch": 55.52,
+      "grad_norm": 0.6680785417556763,
+      "learning_rate": 0.0014448148148148148,
+      "loss": 1.4719,
+      "step": 14990
+    },
+    {
+      "epoch": 55.56,
+      "grad_norm": 0.7036536335945129,
+      "learning_rate": 0.0014444444444444444,
+      "loss": 1.4953,
+      "step": 15000
+    },
+    {
+      "epoch": 55.59,
+      "grad_norm": 0.6639661192893982,
+      "learning_rate": 0.0014440740740740742,
+      "loss": 1.5032,
+      "step": 15010
+    },
+    {
+      "epoch": 55.63,
+      "grad_norm": 0.6836636662483215,
+      "learning_rate": 0.0014437037037037038,
+      "loss": 1.4788,
+      "step": 15020
+    },
+    {
+      "epoch": 55.67,
+      "grad_norm": 0.7251687049865723,
+      "learning_rate": 0.0014433333333333334,
+      "loss": 1.5103,
+      "step": 15030
+    },
+    {
+      "epoch": 55.7,
+      "grad_norm": 0.7114871144294739,
+      "learning_rate": 0.001442962962962963,
+      "loss": 1.4929,
+      "step": 15040
+    },
+    {
+      "epoch": 55.74,
+      "grad_norm": 0.6300182342529297,
+      "learning_rate": 0.0014425925925925928,
+      "loss": 1.537,
+      "step": 15050
+    },
+    {
+      "epoch": 55.78,
+      "grad_norm": 0.6930983662605286,
+      "learning_rate": 0.0014422222222222223,
+      "loss": 1.5433,
+      "step": 15060
+    },
+    {
+      "epoch": 55.81,
+      "grad_norm": 0.7037534713745117,
+      "learning_rate": 0.001441851851851852,
+      "loss": 1.5347,
+      "step": 15070
+    },
+    {
+      "epoch": 55.85,
+      "grad_norm": 0.6565731763839722,
+      "learning_rate": 0.0014414814814814815,
+      "loss": 1.5356,
+      "step": 15080
+    },
+    {
+      "epoch": 55.89,
+      "grad_norm": 0.6927472949028015,
+      "learning_rate": 0.001441111111111111,
+      "loss": 1.5415,
+      "step": 15090
+    },
+    {
+      "epoch": 55.93,
+      "grad_norm": 0.6429632306098938,
+      "learning_rate": 0.0014407407407407407,
+      "loss": 1.5364,
+      "step": 15100
+    },
+    {
+      "epoch": 55.96,
+      "grad_norm": 0.6984735727310181,
+      "learning_rate": 0.0014403703703703703,
+      "loss": 1.5524,
+      "step": 15110
+    },
+    {
+      "epoch": 56.0,
+      "grad_norm": 1.1884384155273438,
+      "learning_rate": 0.0014399999999999999,
+      "loss": 1.5601,
+      "step": 15120
+    },
+    {
+      "epoch": 56.04,
+      "grad_norm": 0.7488458752632141,
+      "learning_rate": 0.0014396296296296297,
+      "loss": 1.3167,
+      "step": 15130
+    },
+    {
+      "epoch": 56.07,
+      "grad_norm": 0.6538281440734863,
+      "learning_rate": 0.0014392592592592593,
+      "loss": 1.3421,
+      "step": 15140
+    },
+    {
+      "epoch": 56.11,
+      "grad_norm": 0.711005449295044,
+      "learning_rate": 0.0014388888888888889,
+      "loss": 1.3511,
+      "step": 15150
+    },
+    {
+      "epoch": 56.15,
+      "grad_norm": 0.656951904296875,
+      "learning_rate": 0.0014385185185185187,
+      "loss": 1.357,
+      "step": 15160
+    },
+    {
+      "epoch": 56.19,
+      "grad_norm": 0.7439039945602417,
+      "learning_rate": 0.0014381481481481483,
+      "loss": 1.3834,
+      "step": 15170
+    },
+    {
+      "epoch": 56.22,
+      "grad_norm": 0.6957058906555176,
+      "learning_rate": 0.0014377777777777778,
+      "loss": 1.3929,
+      "step": 15180
+    },
+    {
+      "epoch": 56.26,
+      "grad_norm": 0.6609116792678833,
+      "learning_rate": 0.0014374074074074074,
+      "loss": 1.3745,
+      "step": 15190
+    },
+    {
+      "epoch": 56.3,
+      "grad_norm": 0.7077909708023071,
+      "learning_rate": 0.0014370370370370372,
+      "loss": 1.4118,
+      "step": 15200
+    },
+    {
+      "epoch": 56.33,
+      "grad_norm": 0.68143230676651,
+      "learning_rate": 0.0014366666666666668,
+      "loss": 1.4364,
+      "step": 15210
+    },
+    {
+      "epoch": 56.37,
+      "grad_norm": 0.6971974968910217,
+      "learning_rate": 0.0014362962962962964,
+      "loss": 1.4552,
+      "step": 15220
+    },
+    {
+      "epoch": 56.41,
+      "grad_norm": 0.6728135347366333,
+      "learning_rate": 0.0014359259259259258,
+      "loss": 1.4186,
+      "step": 15230
+    },
+    {
+      "epoch": 56.44,
+      "grad_norm": 0.7541175484657288,
+      "learning_rate": 0.0014355555555555556,
+      "loss": 1.4607,
+      "step": 15240
+    },
+    {
+      "epoch": 56.48,
+      "grad_norm": 0.7734980583190918,
+      "learning_rate": 0.0014351851851851852,
+      "loss": 1.4602,
+      "step": 15250
+    },
+    {
+      "epoch": 56.52,
+      "grad_norm": 0.7637688517570496,
+      "learning_rate": 0.0014348148148148148,
+      "loss": 1.4519,
+      "step": 15260
+    },
+    {
+      "epoch": 56.56,
+      "grad_norm": 0.701262354850769,
+      "learning_rate": 0.0014344444444444444,
+      "loss": 1.4738,
+      "step": 15270
+    },
+    {
+      "epoch": 56.59,
+      "grad_norm": 0.7055422067642212,
+      "learning_rate": 0.0014340740740740742,
+      "loss": 1.4991,
+      "step": 15280
+    },
+    {
+      "epoch": 56.63,
+      "grad_norm": 0.6790987253189087,
+      "learning_rate": 0.0014337037037037037,
+      "loss": 1.4904,
+      "step": 15290
+    },
+    {
+      "epoch": 56.67,
+      "grad_norm": 0.6938120126724243,
+      "learning_rate": 0.0014333333333333333,
+      "loss": 1.5177,
+      "step": 15300
+    },
+    {
+      "epoch": 56.7,
+      "grad_norm": 0.6558261513710022,
+      "learning_rate": 0.001432962962962963,
+      "loss": 1.4946,
+      "step": 15310
+    },
+    {
+      "epoch": 56.74,
+      "grad_norm": 0.7213912010192871,
+      "learning_rate": 0.0014325925925925927,
+      "loss": 1.5161,
+      "step": 15320
+    },
+    {
+      "epoch": 56.78,
+      "grad_norm": 0.6817654967308044,
+      "learning_rate": 0.0014322222222222223,
+      "loss": 1.5097,
+      "step": 15330
+    },
+    {
+      "epoch": 56.81,
+      "grad_norm": 0.669110894203186,
+      "learning_rate": 0.001431851851851852,
+      "loss": 1.5273,
+      "step": 15340
+    },
+    {
+      "epoch": 56.85,
+      "grad_norm": 0.6431531310081482,
+      "learning_rate": 0.0014314814814814817,
+      "loss": 1.5635,
+      "step": 15350
+    },
+    {
+      "epoch": 56.89,
+      "grad_norm": 0.7609390020370483,
+      "learning_rate": 0.001431111111111111,
+      "loss": 1.5422,
+      "step": 15360
+    },
+    {
+      "epoch": 56.93,
+      "grad_norm": 0.7726841568946838,
+      "learning_rate": 0.0014307407407407407,
+      "loss": 1.5369,
+      "step": 15370
+    },
+    {
+      "epoch": 56.96,
+      "grad_norm": 0.6646488308906555,
+      "learning_rate": 0.0014303703703703703,
+      "loss": 1.5354,
+      "step": 15380
+    },
+    {
+      "epoch": 57.0,
+      "grad_norm": 1.5030916929244995,
+      "learning_rate": 0.00143,
+      "loss": 1.5473,
+      "step": 15390
+    },
+    {
+      "epoch": 57.04,
+      "grad_norm": 0.7018294930458069,
+      "learning_rate": 0.0014296296296296297,
+      "loss": 1.3131,
+      "step": 15400
+    },
+    {
+      "epoch": 57.07,
+      "grad_norm": 0.6784168481826782,
+      "learning_rate": 0.0014292592592592592,
+      "loss": 1.3124,
+      "step": 15410
+    },
+    {
+      "epoch": 57.11,
+      "grad_norm": 0.702157735824585,
+      "learning_rate": 0.0014288888888888888,
+      "loss": 1.3219,
+      "step": 15420
+    },
+    {
+      "epoch": 57.15,
+      "grad_norm": 0.6698986291885376,
+      "learning_rate": 0.0014285185185185186,
+      "loss": 1.3259,
+      "step": 15430
+    },
+    {
+      "epoch": 57.19,
+      "grad_norm": 0.7673943638801575,
+      "learning_rate": 0.0014281481481481482,
+      "loss": 1.3273,
+      "step": 15440
+    },
+    {
+      "epoch": 57.22,
+      "grad_norm": 0.6996111869812012,
+      "learning_rate": 0.0014277777777777778,
+      "loss": 1.3506,
+      "step": 15450
+    },
+    {
+      "epoch": 57.26,
+      "grad_norm": 0.7185860276222229,
+      "learning_rate": 0.0014274074074074074,
+      "loss": 1.4138,
+      "step": 15460
+    },
+    {
+      "epoch": 57.3,
+      "grad_norm": 0.6754123568534851,
+      "learning_rate": 0.0014270370370370372,
+      "loss": 1.4118,
+      "step": 15470
+    },
+    {
+      "epoch": 57.33,
+      "grad_norm": 0.7364366054534912,
+      "learning_rate": 0.0014266666666666668,
+      "loss": 1.4377,
+      "step": 15480
+    },
+    {
+      "epoch": 57.37,
+      "grad_norm": 0.691130518913269,
+      "learning_rate": 0.0014262962962962964,
+      "loss": 1.4192,
+      "step": 15490
+    },
+    {
+      "epoch": 57.41,
+      "grad_norm": 0.703258216381073,
+      "learning_rate": 0.0014259259259259258,
+      "loss": 1.4592,
+      "step": 15500
+    },
+    {
+      "epoch": 57.44,
+      "grad_norm": 0.6907350420951843,
+      "learning_rate": 0.0014255555555555556,
+      "loss": 1.4591,
+      "step": 15510
+    },
+    {
+      "epoch": 57.48,
+      "grad_norm": 0.6878598928451538,
+      "learning_rate": 0.0014251851851851851,
+      "loss": 1.4409,
+      "step": 15520
+    },
+    {
+      "epoch": 57.52,
+      "grad_norm": 0.694089412689209,
+      "learning_rate": 0.0014248148148148147,
+      "loss": 1.4814,
+      "step": 15530
+    },
+    {
+      "epoch": 57.56,
+      "grad_norm": 0.677599310874939,
+      "learning_rate": 0.0014244444444444443,
+      "loss": 1.4811,
+      "step": 15540
+    },
+    {
+      "epoch": 57.59,
+      "grad_norm": 0.7053378224372864,
+      "learning_rate": 0.0014240740740740741,
+      "loss": 1.5,
+      "step": 15550
+    },
+    {
+      "epoch": 57.63,
+      "grad_norm": 0.7823730707168579,
+      "learning_rate": 0.0014237037037037037,
+      "loss": 1.5035,
+      "step": 15560
+    },
+    {
+      "epoch": 57.67,
+      "grad_norm": 0.7754201292991638,
+      "learning_rate": 0.0014233333333333333,
+      "loss": 1.491,
+      "step": 15570
+    },
+    {
+      "epoch": 57.7,
+      "grad_norm": 0.742430567741394,
+      "learning_rate": 0.0014229629629629631,
+      "loss": 1.4948,
+      "step": 15580
+    },
+    {
+      "epoch": 57.74,
+      "grad_norm": 0.7142267823219299,
+      "learning_rate": 0.0014225925925925927,
+      "loss": 1.4721,
+      "step": 15590
+    },
+    {
+      "epoch": 57.78,
+      "grad_norm": 0.6918465495109558,
+      "learning_rate": 0.0014222222222222223,
+      "loss": 1.5362,
+      "step": 15600
+    },
+    {
+      "epoch": 57.81,
+      "grad_norm": 0.7011120915412903,
+      "learning_rate": 0.0014218518518518519,
+      "loss": 1.5254,
+      "step": 15610
+    },
+    {
+      "epoch": 57.85,
+      "grad_norm": 0.6877245903015137,
+      "learning_rate": 0.0014214814814814817,
+      "loss": 1.4972,
+      "step": 15620
+    },
+    {
+      "epoch": 57.89,
+      "grad_norm": 0.6386919021606445,
+      "learning_rate": 0.0014211111111111113,
+      "loss": 1.5306,
+      "step": 15630
+    },
+    {
+      "epoch": 57.93,
+      "grad_norm": 0.7149176597595215,
+      "learning_rate": 0.0014207407407407406,
+      "loss": 1.5251,
+      "step": 15640
+    },
+    {
+      "epoch": 57.96,
+      "grad_norm": 0.6852824091911316,
+      "learning_rate": 0.0014203703703703702,
+      "loss": 1.5594,
+      "step": 15650
+    },
+    {
+      "epoch": 58.0,
+      "grad_norm": 1.1446104049682617,
+      "learning_rate": 0.00142,
+      "loss": 1.5287,
+      "step": 15660
+    },
+    {
+      "epoch": 58.04,
+      "grad_norm": 0.7290907502174377,
+      "learning_rate": 0.0014196296296296296,
+      "loss": 1.2803,
+      "step": 15670
+    },
+    {
+      "epoch": 58.07,
+      "grad_norm": 0.6596196889877319,
+      "learning_rate": 0.0014192592592592592,
+      "loss": 1.3151,
+      "step": 15680
+    },
+    {
+      "epoch": 58.11,
+      "grad_norm": 0.6732152104377747,
+      "learning_rate": 0.0014188888888888888,
+      "loss": 1.3397,
+      "step": 15690
+    },
+    {
+      "epoch": 58.15,
+      "grad_norm": 0.7772537469863892,
+      "learning_rate": 0.0014185185185185186,
+      "loss": 1.3527,
+      "step": 15700
+    },
+    {
+      "epoch": 58.19,
+      "grad_norm": 0.7557883858680725,
+      "learning_rate": 0.0014181481481481482,
+      "loss": 1.3314,
+      "step": 15710
+    },
+    {
+      "epoch": 58.22,
+      "grad_norm": 0.7082036733627319,
+      "learning_rate": 0.0014177777777777778,
+      "loss": 1.3414,
+      "step": 15720
+    },
+    {
+      "epoch": 58.26,
+      "grad_norm": 0.773300290107727,
+      "learning_rate": 0.0014174074074074074,
+      "loss": 1.3594,
+      "step": 15730
+    },
+    {
+      "epoch": 58.3,
+      "grad_norm": 0.6985162496566772,
+      "learning_rate": 0.0014170370370370372,
+      "loss": 1.3939,
+      "step": 15740
+    },
+    {
+      "epoch": 58.33,
+      "grad_norm": 0.8249867558479309,
+      "learning_rate": 0.0014166666666666668,
+      "loss": 1.4084,
+      "step": 15750
+    },
+    {
+      "epoch": 58.37,
+      "grad_norm": 0.6576926112174988,
+      "learning_rate": 0.0014162962962962964,
+      "loss": 1.4324,
+      "step": 15760
+    },
+    {
+      "epoch": 58.41,
+      "grad_norm": 0.7057226896286011,
+      "learning_rate": 0.0014159259259259262,
+      "loss": 1.4467,
+      "step": 15770
+    },
+    {
+      "epoch": 58.44,
+      "grad_norm": 0.6919978260993958,
+      "learning_rate": 0.0014155555555555555,
+      "loss": 1.4659,
+      "step": 15780
+    },
+    {
+      "epoch": 58.48,
+      "grad_norm": 0.7028883695602417,
+      "learning_rate": 0.0014151851851851851,
+      "loss": 1.4111,
+      "step": 15790
+    },
+    {
+      "epoch": 58.52,
+      "grad_norm": 0.7400588989257812,
+      "learning_rate": 0.0014148148148148147,
+      "loss": 1.4447,
+      "step": 15800
+    },
+    {
+      "epoch": 58.56,
+      "grad_norm": 0.6596423983573914,
+      "learning_rate": 0.0014144444444444445,
+      "loss": 1.4724,
+      "step": 15810
+    },
+    {
+      "epoch": 58.59,
+      "grad_norm": 0.6897073984146118,
+      "learning_rate": 0.001414074074074074,
+      "loss": 1.4845,
+      "step": 15820
+    },
+    {
+      "epoch": 58.63,
+      "grad_norm": 0.7325702905654907,
+      "learning_rate": 0.0014137037037037037,
+      "loss": 1.4635,
+      "step": 15830
+    },
+    {
+      "epoch": 58.67,
+      "grad_norm": 0.6876033544540405,
+      "learning_rate": 0.0014133333333333333,
+      "loss": 1.5037,
+      "step": 15840
+    },
+    {
+      "epoch": 58.7,
+      "grad_norm": 0.7476497888565063,
+      "learning_rate": 0.001412962962962963,
+      "loss": 1.5093,
+      "step": 15850
+    },
+    {
+      "epoch": 58.74,
+      "grad_norm": 0.7862426042556763,
+      "learning_rate": 0.0014125925925925927,
+      "loss": 1.4884,
+      "step": 15860
+    },
+    {
+      "epoch": 58.78,
+      "grad_norm": 0.6755629777908325,
+      "learning_rate": 0.0014122222222222223,
+      "loss": 1.5251,
+      "step": 15870
+    },
+    {
+      "epoch": 58.81,
+      "grad_norm": 0.6785138845443726,
+      "learning_rate": 0.0014118518518518519,
+      "loss": 1.4881,
+      "step": 15880
+    },
+    {
+      "epoch": 58.85,
+      "grad_norm": 0.6602843999862671,
+      "learning_rate": 0.0014114814814814817,
+      "loss": 1.4935,
+      "step": 15890
+    },
+    {
+      "epoch": 58.89,
+      "grad_norm": 0.6840865015983582,
+      "learning_rate": 0.0014111111111111112,
+      "loss": 1.5311,
+      "step": 15900
+    },
+    {
+      "epoch": 58.93,
+      "grad_norm": 0.7246437668800354,
+      "learning_rate": 0.0014107407407407408,
+      "loss": 1.5321,
+      "step": 15910
+    },
+    {
+      "epoch": 58.96,
+      "grad_norm": 0.7134275436401367,
+      "learning_rate": 0.0014103703703703702,
+      "loss": 1.5445,
+      "step": 15920
+    },
+    {
+      "epoch": 59.0,
+      "grad_norm": 1.0462781190872192,
+      "learning_rate": 0.00141,
+      "loss": 1.5421,
+      "step": 15930
+    },
+    {
+      "epoch": 59.04,
+      "grad_norm": 0.7037992477416992,
+      "learning_rate": 0.0014096296296296296,
+      "loss": 1.2916,
+      "step": 15940
+    },
+    {
+      "epoch": 59.07,
+      "grad_norm": 0.7080190777778625,
+      "learning_rate": 0.0014092592592592592,
+      "loss": 1.2838,
+      "step": 15950
+    },
+    {
+      "epoch": 59.11,
+      "grad_norm": 0.6890296936035156,
+      "learning_rate": 0.001408888888888889,
+      "loss": 1.3009,
+      "step": 15960
+    },
+    {
+      "epoch": 59.15,
+      "grad_norm": 0.7467537522315979,
+      "learning_rate": 0.0014085185185185186,
+      "loss": 1.3518,
+      "step": 15970
+    },
+    {
+      "epoch": 59.19,
+      "grad_norm": 0.6879865527153015,
+      "learning_rate": 0.0014081481481481482,
+      "loss": 1.3118,
+      "step": 15980
+    },
+    {
+      "epoch": 59.22,
+      "grad_norm": 0.7323064208030701,
+      "learning_rate": 0.0014077777777777778,
+      "loss": 1.3767,
+      "step": 15990
+    },
+    {
+      "epoch": 59.26,
+      "grad_norm": 0.7376371622085571,
+      "learning_rate": 0.0014074074074074076,
+      "loss": 1.3924,
+      "step": 16000
+    },
+    {
+      "epoch": 59.3,
+      "grad_norm": 0.7222793102264404,
+      "learning_rate": 0.0014070370370370372,
+      "loss": 1.3829,
+      "step": 16010
+    },
+    {
+      "epoch": 59.33,
+      "grad_norm": 0.7288392186164856,
+      "learning_rate": 0.0014066666666666667,
+      "loss": 1.4339,
+      "step": 16020
+    },
+    {
+      "epoch": 59.37,
+      "grad_norm": 0.7137278318405151,
+      "learning_rate": 0.0014062962962962963,
+      "loss": 1.4027,
+      "step": 16030
+    },
+    {
+      "epoch": 59.41,
+      "grad_norm": 0.7457606792449951,
+      "learning_rate": 0.0014059259259259261,
+      "loss": 1.4154,
+      "step": 16040
+    },
+    {
+      "epoch": 59.44,
+      "grad_norm": 0.7324532866477966,
+      "learning_rate": 0.0014055555555555555,
+      "loss": 1.4614,
+      "step": 16050
+    },
+    {
+      "epoch": 59.48,
+      "grad_norm": 0.6916303634643555,
+      "learning_rate": 0.001405185185185185,
+      "loss": 1.3899,
+      "step": 16060
+    },
+    {
+      "epoch": 59.52,
+      "grad_norm": 0.7010741233825684,
+      "learning_rate": 0.0014048148148148147,
+      "loss": 1.4602,
+      "step": 16070
+    },
+    {
+      "epoch": 59.56,
+      "grad_norm": 0.7448731064796448,
+      "learning_rate": 0.0014044444444444445,
+      "loss": 1.4703,
+      "step": 16080
+    },
+    {
+      "epoch": 59.59,
+      "grad_norm": 0.6861727833747864,
+      "learning_rate": 0.001404074074074074,
+      "loss": 1.4591,
+      "step": 16090
+    },
+    {
+      "epoch": 59.63,
+      "grad_norm": 0.7314042448997498,
+      "learning_rate": 0.0014037037037037037,
+      "loss": 1.4722,
+      "step": 16100
+    },
+    {
+      "epoch": 59.67,
+      "grad_norm": 0.656716525554657,
+      "learning_rate": 0.0014033333333333333,
+      "loss": 1.4766,
+      "step": 16110
+    },
+    {
+      "epoch": 59.7,
+      "grad_norm": 0.7024694085121155,
+      "learning_rate": 0.001402962962962963,
+      "loss": 1.5067,
+      "step": 16120
+    },
+    {
+      "epoch": 59.74,
+      "grad_norm": 0.7215801477432251,
+      "learning_rate": 0.0014025925925925926,
+      "loss": 1.4881,
+      "step": 16130
+    },
+    {
+      "epoch": 59.78,
+      "grad_norm": 0.6713820695877075,
+      "learning_rate": 0.0014022222222222222,
+      "loss": 1.5,
+      "step": 16140
+    },
+    {
+      "epoch": 59.81,
+      "grad_norm": 0.6949275732040405,
+      "learning_rate": 0.0014018518518518518,
+      "loss": 1.5227,
+      "step": 16150
+    },
+    {
+      "epoch": 59.85,
+      "grad_norm": 0.720103919506073,
+      "learning_rate": 0.0014014814814814816,
+      "loss": 1.5081,
+      "step": 16160
+    },
+    {
+      "epoch": 59.89,
+      "grad_norm": 0.7111461162567139,
+      "learning_rate": 0.0014011111111111112,
+      "loss": 1.4886,
+      "step": 16170
+    },
+    {
+      "epoch": 59.93,
+      "grad_norm": 0.7205804586410522,
+      "learning_rate": 0.0014007407407407408,
+      "loss": 1.5198,
+      "step": 16180
+    },
+    {
+      "epoch": 59.96,
+      "grad_norm": 0.6634063124656677,
+      "learning_rate": 0.0014003703703703704,
+      "loss": 1.5365,
+      "step": 16190
+    },
+    {
+      "epoch": 60.0,
+      "grad_norm": 1.360110878944397,
+      "learning_rate": 0.0014,
+      "loss": 1.5187,
+      "step": 16200
+    },
+    {
+      "epoch": 60.04,
+      "grad_norm": 0.6875532865524292,
+      "learning_rate": 0.0013996296296296296,
+      "loss": 1.2746,
+      "step": 16210
+    },
+    {
+      "epoch": 60.07,
+      "grad_norm": 0.7562872767448425,
+      "learning_rate": 0.0013992592592592592,
+      "loss": 1.2724,
+      "step": 16220
+    },
+    {
+      "epoch": 60.11,
+      "grad_norm": 0.7381096482276917,
+      "learning_rate": 0.001398888888888889,
+      "loss": 1.3072,
+      "step": 16230
+    },
+    {
+      "epoch": 60.15,
+      "grad_norm": 0.7193554043769836,
+      "learning_rate": 0.0013985185185185186,
+      "loss": 1.3413,
+      "step": 16240
+    },
+    {
+      "epoch": 60.19,
+      "grad_norm": 0.6620252728462219,
+      "learning_rate": 0.0013981481481481481,
+      "loss": 1.35,
+      "step": 16250
+    },
+    {
+      "epoch": 60.22,
+      "grad_norm": 0.7208623886108398,
+      "learning_rate": 0.0013977777777777777,
+      "loss": 1.3402,
+      "step": 16260
+    },
+    {
+      "epoch": 60.26,
+      "grad_norm": 0.7059289813041687,
+      "learning_rate": 0.0013974074074074075,
+      "loss": 1.3602,
+      "step": 16270
+    },
+    {
+      "epoch": 60.3,
+      "grad_norm": 0.6971426010131836,
+      "learning_rate": 0.0013970370370370371,
+      "loss": 1.3834,
+      "step": 16280
+    },
+    {
+      "epoch": 60.33,
+      "grad_norm": 0.6854264140129089,
+      "learning_rate": 0.0013966666666666667,
+      "loss": 1.3635,
+      "step": 16290
+    },
+    {
+      "epoch": 60.37,
+      "grad_norm": 0.7690728306770325,
+      "learning_rate": 0.0013962962962962963,
+      "loss": 1.3863,
+      "step": 16300
+    },
+    {
+      "epoch": 60.41,
+      "grad_norm": 0.7954300045967102,
+      "learning_rate": 0.0013959259259259261,
+      "loss": 1.4147,
+      "step": 16310
+    },
+    {
+      "epoch": 60.44,
+      "grad_norm": 0.7556511163711548,
+      "learning_rate": 0.0013955555555555557,
+      "loss": 1.4286,
+      "step": 16320
+    },
+    {
+      "epoch": 60.48,
+      "grad_norm": 0.6833049058914185,
+      "learning_rate": 0.001395185185185185,
+      "loss": 1.4271,
+      "step": 16330
+    },
+    {
+      "epoch": 60.52,
+      "grad_norm": 0.7242010831832886,
+      "learning_rate": 0.0013948148148148147,
+      "loss": 1.4355,
+      "step": 16340
+    },
+    {
+      "epoch": 60.56,
+      "grad_norm": 0.7006980776786804,
+      "learning_rate": 0.0013944444444444445,
+      "loss": 1.4785,
+      "step": 16350
+    },
+    {
+      "epoch": 60.59,
+      "grad_norm": 0.7061491012573242,
+      "learning_rate": 0.001394074074074074,
+      "loss": 1.4504,
+      "step": 16360
+    },
+    {
+      "epoch": 60.63,
+      "grad_norm": 0.710753858089447,
+      "learning_rate": 0.0013937037037037036,
+      "loss": 1.4822,
+      "step": 16370
+    },
+    {
+      "epoch": 60.67,
+      "grad_norm": 0.7644755840301514,
+      "learning_rate": 0.0013933333333333334,
+      "loss": 1.457,
+      "step": 16380
+    },
+    {
+      "epoch": 60.7,
+      "grad_norm": 0.7168658971786499,
+      "learning_rate": 0.001392962962962963,
+      "loss": 1.5014,
+      "step": 16390
+    },
+    {
+      "epoch": 60.74,
+      "grad_norm": 0.7056949138641357,
+      "learning_rate": 0.0013925925925925926,
+      "loss": 1.5104,
+      "step": 16400
+    },
+    {
+      "epoch": 60.78,
+      "grad_norm": 0.6816049814224243,
+      "learning_rate": 0.0013922222222222222,
+      "loss": 1.485,
+      "step": 16410
+    },
+    {
+      "epoch": 60.81,
+      "grad_norm": 0.7482391595840454,
+      "learning_rate": 0.001391851851851852,
+      "loss": 1.4858,
+      "step": 16420
+    },
+    {
+      "epoch": 60.85,
+      "grad_norm": 0.7468370795249939,
+      "learning_rate": 0.0013914814814814816,
+      "loss": 1.4805,
+      "step": 16430
+    },
+    {
+      "epoch": 60.89,
+      "grad_norm": 0.7351725697517395,
+      "learning_rate": 0.0013911111111111112,
+      "loss": 1.521,
+      "step": 16440
+    },
+    {
+      "epoch": 60.93,
+      "grad_norm": 0.7151160836219788,
+      "learning_rate": 0.0013907407407407408,
+      "loss": 1.4889,
+      "step": 16450
+    },
+    {
+      "epoch": 60.96,
+      "grad_norm": 0.7012620568275452,
+      "learning_rate": 0.0013903703703703706,
+      "loss": 1.5615,
+      "step": 16460
+    },
+    {
+      "epoch": 61.0,
+      "grad_norm": 1.633631944656372,
+      "learning_rate": 0.00139,
+      "loss": 1.5143,
+      "step": 16470
+    },
+    {
+      "epoch": 61.04,
+      "grad_norm": 0.7052876353263855,
+      "learning_rate": 0.0013896296296296295,
+      "loss": 1.2774,
+      "step": 16480
+    },
+    {
+      "epoch": 61.07,
+      "grad_norm": 0.6705044507980347,
+      "learning_rate": 0.0013892592592592591,
+      "loss": 1.2894,
+      "step": 16490
+    },
+    {
+      "epoch": 61.11,
+      "grad_norm": 0.746444046497345,
+      "learning_rate": 0.001388888888888889,
+      "loss": 1.274,
+      "step": 16500
+    },
+    {
+      "epoch": 61.15,
+      "grad_norm": 0.6993348002433777,
+      "learning_rate": 0.0013885185185185185,
+      "loss": 1.3287,
+      "step": 16510
+    },
+    {
+      "epoch": 61.19,
+      "grad_norm": 0.7455359697341919,
+      "learning_rate": 0.0013881481481481481,
+      "loss": 1.327,
+      "step": 16520
+    },
+    {
+      "epoch": 61.22,
+      "grad_norm": 0.6975919604301453,
+      "learning_rate": 0.0013877777777777777,
+      "loss": 1.3357,
+      "step": 16530
+    },
+    {
+      "epoch": 61.26,
+      "grad_norm": 0.7186405062675476,
+      "learning_rate": 0.0013874074074074075,
+      "loss": 1.3618,
+      "step": 16540
+    },
+    {
+      "epoch": 61.3,
+      "grad_norm": 0.7653625011444092,
+      "learning_rate": 0.001387037037037037,
+      "loss": 1.3658,
+      "step": 16550
+    },
+    {
+      "epoch": 61.33,
+      "grad_norm": 0.7147905230522156,
+      "learning_rate": 0.0013866666666666667,
+      "loss": 1.3879,
+      "step": 16560
+    },
+    {
+      "epoch": 61.37,
+      "grad_norm": 0.760710597038269,
+      "learning_rate": 0.0013862962962962965,
+      "loss": 1.3938,
+      "step": 16570
+    },
+    {
+      "epoch": 61.41,
+      "grad_norm": 0.6825000047683716,
+      "learning_rate": 0.001385925925925926,
+      "loss": 1.4133,
+      "step": 16580
+    },
+    {
+      "epoch": 61.44,
+      "grad_norm": 0.7693556547164917,
+      "learning_rate": 0.0013855555555555557,
+      "loss": 1.4312,
+      "step": 16590
+    },
+    {
+      "epoch": 61.48,
+      "grad_norm": 0.7311041355133057,
+      "learning_rate": 0.001385185185185185,
+      "loss": 1.4205,
+      "step": 16600
+    },
+    {
+      "epoch": 61.52,
+      "grad_norm": 0.7689315676689148,
+      "learning_rate": 0.0013848148148148148,
+      "loss": 1.4454,
+      "step": 16610
+    },
+    {
+      "epoch": 61.56,
+      "grad_norm": 0.7419061064720154,
+      "learning_rate": 0.0013844444444444444,
+      "loss": 1.4293,
+      "step": 16620
+    },
+    {
+      "epoch": 61.59,
+      "grad_norm": 0.6940130591392517,
+      "learning_rate": 0.001384074074074074,
+      "loss": 1.4663,
+      "step": 16630
+    },
+    {
+      "epoch": 61.63,
+      "grad_norm": 0.721855640411377,
+      "learning_rate": 0.0013837037037037036,
+      "loss": 1.4526,
+      "step": 16640
+    },
+    {
+      "epoch": 61.67,
+      "grad_norm": 0.6498404741287231,
+      "learning_rate": 0.0013833333333333334,
+      "loss": 1.4941,
+      "step": 16650
+    },
+    {
+      "epoch": 61.7,
+      "grad_norm": 0.7447950839996338,
+      "learning_rate": 0.001382962962962963,
+      "loss": 1.484,
+      "step": 16660
+    },
+    {
+      "epoch": 61.74,
+      "grad_norm": 0.6579961776733398,
+      "learning_rate": 0.0013825925925925926,
+      "loss": 1.4905,
+      "step": 16670
+    },
+    {
+      "epoch": 61.78,
+      "grad_norm": 0.70018070936203,
+      "learning_rate": 0.0013822222222222222,
+      "loss": 1.4693,
+      "step": 16680
+    },
+    {
+      "epoch": 61.81,
+      "grad_norm": 0.7552095055580139,
+      "learning_rate": 0.001381851851851852,
+      "loss": 1.5029,
+      "step": 16690
+    },
+    {
+      "epoch": 61.85,
+      "grad_norm": 0.7254689335823059,
+      "learning_rate": 0.0013814814814814816,
+      "loss": 1.4784,
+      "step": 16700
+    },
+    {
+      "epoch": 61.89,
+      "grad_norm": 0.6803253889083862,
+      "learning_rate": 0.0013811111111111112,
+      "loss": 1.4925,
+      "step": 16710
+    },
+    {
+      "epoch": 61.93,
+      "grad_norm": 0.6918293237686157,
+      "learning_rate": 0.0013807407407407408,
+      "loss": 1.5135,
+      "step": 16720
+    },
+    {
+      "epoch": 61.96,
+      "grad_norm": 0.7400352358818054,
+      "learning_rate": 0.0013803703703703706,
+      "loss": 1.4999,
+      "step": 16730
+    },
+    {
+      "epoch": 62.0,
+      "grad_norm": 1.246131181716919,
+      "learning_rate": 0.00138,
+      "loss": 1.5048,
+      "step": 16740
+    },
+    {
+      "epoch": 62.04,
+      "grad_norm": 0.7409060597419739,
+      "learning_rate": 0.0013796296296296295,
+      "loss": 1.2478,
+      "step": 16750
+    },
+    {
+      "epoch": 62.07,
+      "grad_norm": 0.7132163047790527,
+      "learning_rate": 0.0013792592592592591,
+      "loss": 1.2838,
+      "step": 16760
+    },
+    {
+      "epoch": 62.11,
+      "grad_norm": 0.7276562452316284,
+      "learning_rate": 0.001378888888888889,
+      "loss": 1.2673,
+      "step": 16770
+    },
+    {
+      "epoch": 62.15,
+      "grad_norm": 0.7163761854171753,
+      "learning_rate": 0.0013785185185185185,
+      "loss": 1.3045,
+      "step": 16780
+    },
+    {
+      "epoch": 62.19,
+      "grad_norm": 0.7683485746383667,
+      "learning_rate": 0.001378148148148148,
+      "loss": 1.3284,
+      "step": 16790
+    },
+    {
+      "epoch": 62.22,
+      "grad_norm": 0.678407609462738,
+      "learning_rate": 0.001377777777777778,
+      "loss": 1.363,
+      "step": 16800
+    },
+    {
+      "epoch": 62.26,
+      "grad_norm": 0.652581512928009,
+      "learning_rate": 0.0013774074074074075,
+      "loss": 1.3628,
+      "step": 16810
+    },
+    {
+      "epoch": 62.3,
+      "grad_norm": 0.7575969696044922,
+      "learning_rate": 0.001377037037037037,
+      "loss": 1.3731,
+      "step": 16820
+    },
+    {
+      "epoch": 62.33,
+      "grad_norm": 0.7279390096664429,
+      "learning_rate": 0.0013766666666666667,
+      "loss": 1.3444,
+      "step": 16830
+    },
+    {
+      "epoch": 62.37,
+      "grad_norm": 0.7434338331222534,
+      "learning_rate": 0.0013762962962962965,
+      "loss": 1.3699,
+      "step": 16840
+    },
+    {
+      "epoch": 62.41,
+      "grad_norm": 0.7671130895614624,
+      "learning_rate": 0.001375925925925926,
+      "loss": 1.3842,
+      "step": 16850
+    },
+    {
+      "epoch": 62.44,
+      "grad_norm": 0.775354266166687,
+      "learning_rate": 0.0013755555555555556,
+      "loss": 1.4063,
+      "step": 16860
+    },
+    {
+      "epoch": 62.48,
+      "grad_norm": 0.7300680875778198,
+      "learning_rate": 0.0013751851851851852,
+      "loss": 1.4227,
+      "step": 16870
+    },
+    {
+      "epoch": 62.52,
+      "grad_norm": 0.6965778470039368,
+      "learning_rate": 0.0013748148148148148,
+      "loss": 1.4352,
+      "step": 16880
+    },
+    {
+      "epoch": 62.56,
+      "grad_norm": 0.7088303565979004,
+      "learning_rate": 0.0013744444444444444,
+      "loss": 1.4484,
+      "step": 16890
+    },
+    {
+      "epoch": 62.59,
+      "grad_norm": 0.6896482706069946,
+      "learning_rate": 0.001374074074074074,
+      "loss": 1.46,
+      "step": 16900
+    },
+    {
+      "epoch": 62.63,
+      "grad_norm": 0.7502049803733826,
+      "learning_rate": 0.0013737037037037036,
+      "loss": 1.4443,
+      "step": 16910
+    },
+    {
+      "epoch": 62.67,
+      "grad_norm": 0.7610279321670532,
+      "learning_rate": 0.0013733333333333334,
+      "loss": 1.4562,
+      "step": 16920
+    },
+    {
+      "epoch": 62.7,
+      "grad_norm": 0.7035161852836609,
+      "learning_rate": 0.001372962962962963,
+      "loss": 1.4789,
+      "step": 16930
+    },
+    {
+      "epoch": 62.74,
+      "grad_norm": 0.635873556137085,
+      "learning_rate": 0.0013725925925925926,
+      "loss": 1.4699,
+      "step": 16940
+    },
+    {
+      "epoch": 62.78,
+      "grad_norm": 0.7646804451942444,
+      "learning_rate": 0.0013722222222222222,
+      "loss": 1.4785,
+      "step": 16950
+    },
+    {
+      "epoch": 62.81,
+      "grad_norm": 0.6760735511779785,
+      "learning_rate": 0.001371851851851852,
+      "loss": 1.4836,
+      "step": 16960
+    },
+    {
+      "epoch": 62.85,
+      "grad_norm": 0.7437437772750854,
+      "learning_rate": 0.0013714814814814816,
+      "loss": 1.4912,
+      "step": 16970
+    },
+    {
+      "epoch": 62.89,
+      "grad_norm": 0.7423031330108643,
+      "learning_rate": 0.0013711111111111111,
+      "loss": 1.4699,
+      "step": 16980
+    },
+    {
+      "epoch": 62.93,
+      "grad_norm": 0.7733505368232727,
+      "learning_rate": 0.001370740740740741,
+      "loss": 1.5027,
+      "step": 16990
+    },
+    {
+      "epoch": 62.96,
+      "grad_norm": 0.7638499140739441,
+      "learning_rate": 0.0013703703703703705,
+      "loss": 1.4934,
+      "step": 17000
+    },
+    {
+      "epoch": 63.0,
+      "grad_norm": 1.214098334312439,
+      "learning_rate": 0.0013700000000000001,
+      "loss": 1.5209,
+      "step": 17010
+    },
+    {
+      "epoch": 63.04,
+      "grad_norm": 0.7190075516700745,
+      "learning_rate": 0.0013696296296296295,
+      "loss": 1.2687,
+      "step": 17020
+    },
+    {
+      "epoch": 63.07,
+      "grad_norm": 0.7111839652061462,
+      "learning_rate": 0.0013692592592592593,
+      "loss": 1.261,
+      "step": 17030
+    },
+    {
+      "epoch": 63.11,
+      "grad_norm": 0.7525911331176758,
+      "learning_rate": 0.0013688888888888889,
+      "loss": 1.2793,
+      "step": 17040
+    },
+    {
+      "epoch": 63.15,
+      "grad_norm": 0.720476508140564,
+      "learning_rate": 0.0013685185185185185,
+      "loss": 1.2787,
+      "step": 17050
+    },
+    {
+      "epoch": 63.19,
+      "grad_norm": 0.7493318915367126,
+      "learning_rate": 0.001368148148148148,
+      "loss": 1.3218,
+      "step": 17060
+    },
+    {
+      "epoch": 63.22,
+      "grad_norm": 0.7084575891494751,
+      "learning_rate": 0.0013677777777777779,
+      "loss": 1.3391,
+      "step": 17070
+    },
+    {
+      "epoch": 63.26,
+      "grad_norm": 0.8328226208686829,
+      "learning_rate": 0.0013674074074074075,
+      "loss": 1.3533,
+      "step": 17080
+    },
+    {
+      "epoch": 63.3,
+      "grad_norm": 0.7640653848648071,
+      "learning_rate": 0.001367037037037037,
+      "loss": 1.3232,
+      "step": 17090
+    },
+    {
+      "epoch": 63.33,
+      "grad_norm": 0.773227334022522,
+      "learning_rate": 0.0013666666666666666,
+      "loss": 1.381,
+      "step": 17100
+    },
+    {
+      "epoch": 63.37,
+      "grad_norm": 0.6970317363739014,
+      "learning_rate": 0.0013662962962962964,
+      "loss": 1.3851,
+      "step": 17110
+    },
+    {
+      "epoch": 63.41,
+      "grad_norm": 0.7326919436454773,
+      "learning_rate": 0.001365925925925926,
+      "loss": 1.3854,
+      "step": 17120
+    },
+    {
+      "epoch": 63.44,
+      "grad_norm": 0.8480170965194702,
+      "learning_rate": 0.0013655555555555556,
+      "loss": 1.4173,
+      "step": 17130
+    },
+    {
+      "epoch": 63.48,
+      "grad_norm": 0.7686656713485718,
+      "learning_rate": 0.0013651851851851852,
+      "loss": 1.3834,
+      "step": 17140
+    },
+    {
+      "epoch": 63.52,
+      "grad_norm": 0.705543577671051,
+      "learning_rate": 0.001364814814814815,
+      "loss": 1.4647,
+      "step": 17150
+    },
+    {
+      "epoch": 63.56,
+      "grad_norm": 0.7954389452934265,
+      "learning_rate": 0.0013644444444444444,
+      "loss": 1.4571,
+      "step": 17160
+    },
+    {
+      "epoch": 63.59,
+      "grad_norm": 0.7180191874504089,
+      "learning_rate": 0.001364074074074074,
+      "loss": 1.4375,
+      "step": 17170
+    },
+    {
+      "epoch": 63.63,
+      "grad_norm": 0.6709361672401428,
+      "learning_rate": 0.0013637037037037036,
+      "loss": 1.4573,
+      "step": 17180
+    },
+    {
+      "epoch": 63.67,
+      "grad_norm": 0.7622054815292358,
+      "learning_rate": 0.0013633333333333334,
+      "loss": 1.4532,
+      "step": 17190
+    },
+    {
+      "epoch": 63.7,
+      "grad_norm": 0.7002084851264954,
+      "learning_rate": 0.001362962962962963,
+      "loss": 1.4217,
+      "step": 17200
+    },
+    {
+      "epoch": 63.74,
+      "grad_norm": 0.7226837277412415,
+      "learning_rate": 0.0013625925925925925,
+      "loss": 1.4612,
+      "step": 17210
+    },
+    {
+      "epoch": 63.78,
+      "grad_norm": 0.6710451245307922,
+      "learning_rate": 0.0013622222222222223,
+      "loss": 1.491,
+      "step": 17220
+    },
+    {
+      "epoch": 63.81,
+      "grad_norm": 0.8155320286750793,
+      "learning_rate": 0.001361851851851852,
+      "loss": 1.4587,
+      "step": 17230
+    },
+    {
+      "epoch": 63.85,
+      "grad_norm": 0.73924320936203,
+      "learning_rate": 0.0013614814814814815,
+      "loss": 1.4874,
+      "step": 17240
+    },
+    {
+      "epoch": 63.89,
+      "grad_norm": 0.7509787082672119,
+      "learning_rate": 0.0013611111111111111,
+      "loss": 1.4736,
+      "step": 17250
+    },
+    {
+      "epoch": 63.93,
+      "grad_norm": 0.7284494042396545,
+      "learning_rate": 0.001360740740740741,
+      "loss": 1.4753,
+      "step": 17260
+    },
+    {
+      "epoch": 63.96,
+      "grad_norm": 0.6677215695381165,
+      "learning_rate": 0.0013603703703703705,
+      "loss": 1.5084,
+      "step": 17270
+    },
+    {
+      "epoch": 64.0,
+      "grad_norm": 1.1213937997817993,
+      "learning_rate": 0.00136,
+      "loss": 1.5176,
+      "step": 17280
+    },
+    {
+      "epoch": 64.04,
+      "grad_norm": 0.7577294111251831,
+      "learning_rate": 0.0013596296296296295,
+      "loss": 1.2681,
+      "step": 17290
+    },
+    {
+      "epoch": 64.07,
+      "grad_norm": 0.7758843898773193,
+      "learning_rate": 0.0013592592592592593,
+      "loss": 1.2588,
+      "step": 17300
+    },
+    {
+      "epoch": 64.11,
+      "grad_norm": 0.71478271484375,
+      "learning_rate": 0.0013588888888888889,
+      "loss": 1.2865,
+      "step": 17310
+    },
+    {
+      "epoch": 64.15,
+      "grad_norm": 0.7606163024902344,
+      "learning_rate": 0.0013585185185185185,
+      "loss": 1.302,
+      "step": 17320
+    },
+    {
+      "epoch": 64.19,
+      "grad_norm": 0.7798721194267273,
+      "learning_rate": 0.001358148148148148,
+      "loss": 1.3156,
+      "step": 17330
+    },
+    {
+      "epoch": 64.22,
+      "grad_norm": 0.756719708442688,
+      "learning_rate": 0.0013577777777777778,
+      "loss": 1.3245,
+      "step": 17340
+    },
+    {
+      "epoch": 64.26,
+      "grad_norm": 0.7324285507202148,
+      "learning_rate": 0.0013574074074074074,
+      "loss": 1.3406,
+      "step": 17350
+    },
+    {
+      "epoch": 64.3,
+      "grad_norm": 0.7444923520088196,
+      "learning_rate": 0.001357037037037037,
+      "loss": 1.3676,
+      "step": 17360
+    },
+    {
+      "epoch": 64.33,
+      "grad_norm": 0.7438486814498901,
+      "learning_rate": 0.0013566666666666666,
+      "loss": 1.3631,
+      "step": 17370
+    },
+    {
+      "epoch": 64.37,
+      "grad_norm": 0.7301568984985352,
+      "learning_rate": 0.0013562962962962964,
+      "loss": 1.3979,
+      "step": 17380
+    },
+    {
+      "epoch": 64.41,
+      "grad_norm": 0.7605223655700684,
+      "learning_rate": 0.001355925925925926,
+      "loss": 1.3797,
+      "step": 17390
+    },
+    {
+      "epoch": 64.44,
+      "grad_norm": 0.7399788498878479,
+      "learning_rate": 0.0013555555555555556,
+      "loss": 1.3976,
+      "step": 17400
+    },
+    {
+      "epoch": 64.48,
+      "grad_norm": 0.7360973954200745,
+      "learning_rate": 0.0013551851851851854,
+      "loss": 1.4176,
+      "step": 17410
+    },
+    {
+      "epoch": 64.52,
+      "grad_norm": 0.7445151805877686,
+      "learning_rate": 0.001354814814814815,
+      "loss": 1.4123,
+      "step": 17420
+    },
+    {
+      "epoch": 64.56,
+      "grad_norm": 0.7263211011886597,
+      "learning_rate": 0.0013544444444444444,
+      "loss": 1.403,
+      "step": 17430
+    },
+    {
+      "epoch": 64.59,
+      "grad_norm": 0.7512028813362122,
+      "learning_rate": 0.001354074074074074,
+      "loss": 1.4177,
+      "step": 17440
+    },
+    {
+      "epoch": 64.63,
+      "grad_norm": 0.7566067576408386,
+      "learning_rate": 0.0013537037037037038,
+      "loss": 1.4249,
+      "step": 17450
+    },
+    {
+      "epoch": 64.67,
+      "grad_norm": 0.7968330383300781,
+      "learning_rate": 0.0013533333333333333,
+      "loss": 1.4237,
+      "step": 17460
+    },
+    {
+      "epoch": 64.7,
+      "grad_norm": 0.7006673216819763,
+      "learning_rate": 0.001352962962962963,
+      "loss": 1.4591,
+      "step": 17470
+    },
+    {
+      "epoch": 64.74,
+      "grad_norm": 0.7665920257568359,
+      "learning_rate": 0.0013525925925925925,
+      "loss": 1.4532,
+      "step": 17480
+    },
+    {
+      "epoch": 64.78,
+      "grad_norm": 0.7508714199066162,
+      "learning_rate": 0.0013522222222222223,
+      "loss": 1.474,
+      "step": 17490
+    },
+    {
+      "epoch": 64.81,
+      "grad_norm": 0.7315797209739685,
+      "learning_rate": 0.001351851851851852,
+      "loss": 1.439,
+      "step": 17500
+    },
+    {
+      "epoch": 64.85,
+      "grad_norm": 0.8568592071533203,
+      "learning_rate": 0.0013514814814814815,
+      "loss": 1.4562,
+      "step": 17510
+    },
+    {
+      "epoch": 64.89,
+      "grad_norm": 0.7268970012664795,
+      "learning_rate": 0.001351111111111111,
+      "loss": 1.5033,
+      "step": 17520
+    },
+    {
+      "epoch": 64.93,
+      "grad_norm": 0.7254437804222107,
+      "learning_rate": 0.001350740740740741,
+      "loss": 1.4433,
+      "step": 17530
+    },
+    {
+      "epoch": 64.96,
+      "grad_norm": 0.7384520769119263,
+      "learning_rate": 0.0013503703703703705,
+      "loss": 1.4822,
+      "step": 17540
+    },
+    {
+      "epoch": 65.0,
+      "grad_norm": 1.2170757055282593,
+      "learning_rate": 0.00135,
+      "loss": 1.4843,
+      "step": 17550
+    },
+    {
+      "epoch": 65.04,
+      "grad_norm": 0.715018630027771,
+      "learning_rate": 0.0013496296296296297,
+      "loss": 1.2211,
+      "step": 17560
+    },
+    {
+      "epoch": 65.07,
+      "grad_norm": 0.8384785652160645,
+      "learning_rate": 0.0013492592592592592,
+      "loss": 1.2132,
+      "step": 17570
+    },
+    {
+      "epoch": 65.11,
+      "grad_norm": 0.6869712471961975,
+      "learning_rate": 0.0013488888888888888,
+      "loss": 1.2573,
+      "step": 17580
+    },
+    {
+      "epoch": 65.15,
+      "grad_norm": 0.7423787117004395,
+      "learning_rate": 0.0013485185185185184,
+      "loss": 1.2947,
+      "step": 17590
+    },
+    {
+      "epoch": 65.19,
+      "grad_norm": 0.7109069228172302,
+      "learning_rate": 0.0013481481481481482,
+      "loss": 1.3025,
+      "step": 17600
+    },
+    {
+      "epoch": 65.22,
+      "grad_norm": 0.7798488736152649,
+      "learning_rate": 0.0013477777777777778,
+      "loss": 1.3109,
+      "step": 17610
+    },
+    {
+      "epoch": 65.26,
+      "grad_norm": 0.7772756218910217,
+      "learning_rate": 0.0013474074074074074,
+      "loss": 1.3323,
+      "step": 17620
+    },
+    {
+      "epoch": 65.3,
+      "grad_norm": 0.7582665681838989,
+      "learning_rate": 0.001347037037037037,
+      "loss": 1.3301,
+      "step": 17630
+    },
+    {
+      "epoch": 65.33,
+      "grad_norm": 0.7635788917541504,
+      "learning_rate": 0.0013466666666666668,
+      "loss": 1.3534,
+      "step": 17640
+    },
+    {
+      "epoch": 65.37,
+      "grad_norm": 0.7907354235649109,
+      "learning_rate": 0.0013462962962962964,
+      "loss": 1.3917,
+      "step": 17650
+    },
+    {
+      "epoch": 65.41,
+      "grad_norm": 0.7538552284240723,
+      "learning_rate": 0.001345925925925926,
+      "loss": 1.3934,
+      "step": 17660
+    },
+    {
+      "epoch": 65.44,
+      "grad_norm": 0.7302442193031311,
+      "learning_rate": 0.0013455555555555556,
+      "loss": 1.3849,
+      "step": 17670
+    },
+    {
+      "epoch": 65.48,
+      "grad_norm": 0.7725008130073547,
+      "learning_rate": 0.0013451851851851854,
+      "loss": 1.3807,
+      "step": 17680
+    },
+    {
+      "epoch": 65.52,
+      "grad_norm": 0.7958641648292542,
+      "learning_rate": 0.001344814814814815,
+      "loss": 1.4248,
+      "step": 17690
+    },
+    {
+      "epoch": 65.56,
+      "grad_norm": 0.770406186580658,
+      "learning_rate": 0.0013444444444444445,
+      "loss": 1.4143,
+      "step": 17700
+    },
+    {
+      "epoch": 65.59,
+      "grad_norm": 0.6958473920822144,
+      "learning_rate": 0.001344074074074074,
+      "loss": 1.4358,
+      "step": 17710
+    },
+    {
+      "epoch": 65.63,
+      "grad_norm": 0.7644444108009338,
+      "learning_rate": 0.0013437037037037037,
+      "loss": 1.4147,
+      "step": 17720
+    },
+    {
+      "epoch": 65.67,
+      "grad_norm": 0.7265663743019104,
+      "learning_rate": 0.0013433333333333333,
+      "loss": 1.4309,
+      "step": 17730
+    },
+    {
+      "epoch": 65.7,
+      "grad_norm": 0.7758519649505615,
+      "learning_rate": 0.001342962962962963,
+      "loss": 1.4262,
+      "step": 17740
+    },
+    {
+      "epoch": 65.74,
+      "grad_norm": 0.7135465741157532,
+      "learning_rate": 0.0013425925925925925,
+      "loss": 1.4435,
+      "step": 17750
+    },
+    {
+      "epoch": 65.78,
+      "grad_norm": 0.7321381568908691,
+      "learning_rate": 0.0013422222222222223,
+      "loss": 1.4586,
+      "step": 17760
+    },
+    {
+      "epoch": 65.81,
+      "grad_norm": 0.7740125060081482,
+      "learning_rate": 0.0013418518518518519,
+      "loss": 1.4454,
+      "step": 17770
+    },
+    {
+      "epoch": 65.85,
+      "grad_norm": 0.714545488357544,
+      "learning_rate": 0.0013414814814814815,
+      "loss": 1.4606,
+      "step": 17780
+    },
+    {
+      "epoch": 65.89,
+      "grad_norm": 0.7659265995025635,
+      "learning_rate": 0.001341111111111111,
+      "loss": 1.4924,
+      "step": 17790
+    },
+    {
+      "epoch": 65.93,
+      "grad_norm": 0.6884202361106873,
+      "learning_rate": 0.0013407407407407409,
+      "loss": 1.502,
+      "step": 17800
+    },
+    {
+      "epoch": 65.96,
+      "grad_norm": 0.8415507078170776,
+      "learning_rate": 0.0013403703703703705,
+      "loss": 1.4725,
+      "step": 17810
+    },
+    {
+      "epoch": 66.0,
+      "grad_norm": 1.2466375827789307,
+      "learning_rate": 0.00134,
+      "loss": 1.4451,
+      "step": 17820
+    },
+    {
+      "epoch": 66.04,
+      "grad_norm": 0.7439112663269043,
+      "learning_rate": 0.0013396296296296299,
+      "loss": 1.256,
+      "step": 17830
+    },
+    {
+      "epoch": 66.07,
+      "grad_norm": 0.8170116543769836,
+      "learning_rate": 0.0013392592592592592,
+      "loss": 1.2327,
+      "step": 17840
+    },
+    {
+      "epoch": 66.11,
+      "grad_norm": 0.707555890083313,
+      "learning_rate": 0.0013388888888888888,
+      "loss": 1.2809,
+      "step": 17850
+    },
+    {
+      "epoch": 66.15,
+      "grad_norm": 0.6814079284667969,
+      "learning_rate": 0.0013385185185185184,
+      "loss": 1.276,
+      "step": 17860
+    },
+    {
+      "epoch": 66.19,
+      "grad_norm": 0.7349382042884827,
+      "learning_rate": 0.0013381481481481482,
+      "loss": 1.2884,
+      "step": 17870
+    },
+    {
+      "epoch": 66.22,
+      "grad_norm": 0.7153812646865845,
+      "learning_rate": 0.0013377777777777778,
+      "loss": 1.322,
+      "step": 17880
+    },
+    {
+      "epoch": 66.26,
+      "grad_norm": 0.7231837511062622,
+      "learning_rate": 0.0013374074074074074,
+      "loss": 1.299,
+      "step": 17890
+    },
+    {
+      "epoch": 66.3,
+      "grad_norm": 0.7510600686073303,
+      "learning_rate": 0.001337037037037037,
+      "loss": 1.3416,
+      "step": 17900
+    },
+    {
+      "epoch": 66.33,
+      "grad_norm": 0.7508620023727417,
+      "learning_rate": 0.0013366666666666668,
+      "loss": 1.3809,
+      "step": 17910
+    },
+    {
+      "epoch": 66.37,
+      "grad_norm": 0.7234875559806824,
+      "learning_rate": 0.0013362962962962964,
+      "loss": 1.3698,
+      "step": 17920
+    },
+    {
+      "epoch": 66.41,
+      "grad_norm": 0.731718897819519,
+      "learning_rate": 0.001335925925925926,
+      "loss": 1.3692,
+      "step": 17930
+    },
+    {
+      "epoch": 66.44,
+      "grad_norm": 0.7301697731018066,
+      "learning_rate": 0.0013355555555555555,
+      "loss": 1.3878,
+      "step": 17940
+    },
+    {
+      "epoch": 66.48,
+      "grad_norm": 0.7749512195587158,
+      "learning_rate": 0.0013351851851851853,
+      "loss": 1.3832,
+      "step": 17950
+    },
+    {
+      "epoch": 66.52,
+      "grad_norm": 0.8031456470489502,
+      "learning_rate": 0.001334814814814815,
+      "loss": 1.4165,
+      "step": 17960
+    },
+    {
+      "epoch": 66.56,
+      "grad_norm": 0.7439578771591187,
+      "learning_rate": 0.0013344444444444445,
+      "loss": 1.3794,
+      "step": 17970
+    },
+    {
+      "epoch": 66.59,
+      "grad_norm": 0.7619519233703613,
+      "learning_rate": 0.001334074074074074,
+      "loss": 1.4354,
+      "step": 17980
+    },
+    {
+      "epoch": 66.63,
+      "grad_norm": 0.7329776883125305,
+      "learning_rate": 0.0013337037037037037,
+      "loss": 1.4155,
+      "step": 17990
+    },
+    {
+      "epoch": 66.67,
+      "grad_norm": 0.7692404389381409,
+      "learning_rate": 0.0013333333333333333,
+      "loss": 1.4214,
+      "step": 18000
+    },
+    {
+      "epoch": 66.7,
+      "grad_norm": 0.7139493227005005,
+      "learning_rate": 0.0013329629629629629,
+      "loss": 1.4516,
+      "step": 18010
+    },
+    {
+      "epoch": 66.74,
+      "grad_norm": 0.7538207173347473,
+      "learning_rate": 0.0013325925925925927,
+      "loss": 1.432,
+      "step": 18020
+    },
+    {
+      "epoch": 66.78,
+      "grad_norm": 0.7339174747467041,
+      "learning_rate": 0.0013322222222222223,
+      "loss": 1.4441,
+      "step": 18030
+    },
+    {
+      "epoch": 66.81,
+      "grad_norm": 0.7614271640777588,
+      "learning_rate": 0.0013318518518518519,
+      "loss": 1.4407,
+      "step": 18040
+    },
+    {
+      "epoch": 66.85,
+      "grad_norm": 0.7265300750732422,
+      "learning_rate": 0.0013314814814814814,
+      "loss": 1.4497,
+      "step": 18050
+    },
+    {
+      "epoch": 66.89,
+      "grad_norm": 0.7015870213508606,
+      "learning_rate": 0.0013311111111111113,
+      "loss": 1.463,
+      "step": 18060
+    },
+    {
+      "epoch": 66.93,
+      "grad_norm": 0.736568033695221,
+      "learning_rate": 0.0013307407407407408,
+      "loss": 1.4459,
+      "step": 18070
+    },
+    {
+      "epoch": 66.96,
+      "grad_norm": 0.8021724820137024,
+      "learning_rate": 0.0013303703703703704,
+      "loss": 1.4521,
+      "step": 18080
+    },
+    {
+      "epoch": 67.0,
+      "grad_norm": 1.27347993850708,
+      "learning_rate": 0.00133,
+      "loss": 1.4864,
+      "step": 18090
+    },
+    {
+      "epoch": 67.04,
+      "grad_norm": 0.7476965188980103,
+      "learning_rate": 0.0013296296296296298,
+      "loss": 1.2337,
+      "step": 18100
+    },
+    {
+      "epoch": 67.07,
+      "grad_norm": 0.7044015526771545,
+      "learning_rate": 0.0013292592592592594,
+      "loss": 1.2397,
+      "step": 18110
+    },
+    {
+      "epoch": 67.11,
+      "grad_norm": 0.7922229170799255,
+      "learning_rate": 0.0013288888888888888,
+      "loss": 1.256,
+      "step": 18120
+    },
+    {
+      "epoch": 67.15,
+      "grad_norm": 0.7829376459121704,
+      "learning_rate": 0.0013285185185185184,
+      "loss": 1.2582,
+      "step": 18130
+    },
+    {
+      "epoch": 67.19,
+      "grad_norm": 0.7793470621109009,
+      "learning_rate": 0.0013281481481481482,
+      "loss": 1.2719,
+      "step": 18140
+    },
+    {
+      "epoch": 67.22,
+      "grad_norm": 0.727636456489563,
+      "learning_rate": 0.0013277777777777778,
+      "loss": 1.3313,
+      "step": 18150
+    },
+    {
+      "epoch": 67.26,
+      "grad_norm": 0.7546897530555725,
+      "learning_rate": 0.0013274074074074074,
+      "loss": 1.312,
+      "step": 18160
+    },
+    {
+      "epoch": 67.3,
+      "grad_norm": 0.6836079955101013,
+      "learning_rate": 0.001327037037037037,
+      "loss": 1.3276,
+      "step": 18170
+    },
+    {
+      "epoch": 67.33,
+      "grad_norm": 0.802686333656311,
+      "learning_rate": 0.0013266666666666667,
+      "loss": 1.3376,
+      "step": 18180
+    },
+    {
+      "epoch": 67.37,
+      "grad_norm": 0.8498095870018005,
+      "learning_rate": 0.0013262962962962963,
+      "loss": 1.3466,
+      "step": 18190
+    },
+    {
+      "epoch": 67.41,
+      "grad_norm": 0.7194114327430725,
+      "learning_rate": 0.001325925925925926,
+      "loss": 1.3611,
+      "step": 18200
+    },
+    {
+      "epoch": 67.44,
+      "grad_norm": 0.7203561663627625,
+      "learning_rate": 0.0013255555555555557,
+      "loss": 1.3319,
+      "step": 18210
+    },
+    {
+      "epoch": 67.48,
+      "grad_norm": 0.7772254943847656,
+      "learning_rate": 0.0013251851851851853,
+      "loss": 1.4026,
+      "step": 18220
+    },
+    {
+      "epoch": 67.52,
+      "grad_norm": 0.7512720823287964,
+      "learning_rate": 0.001324814814814815,
+      "loss": 1.3907,
+      "step": 18230
+    },
+    {
+      "epoch": 67.56,
+      "grad_norm": 0.7536607980728149,
+      "learning_rate": 0.0013244444444444445,
+      "loss": 1.4007,
+      "step": 18240
+    },
+    {
+      "epoch": 67.59,
+      "grad_norm": 0.7336398363113403,
+      "learning_rate": 0.0013240740740740743,
+      "loss": 1.3965,
+      "step": 18250
+    },
+    {
+      "epoch": 67.63,
+      "grad_norm": 0.7424538135528564,
+      "learning_rate": 0.0013237037037037037,
+      "loss": 1.4303,
+      "step": 18260
+    },
+    {
+      "epoch": 67.67,
+      "grad_norm": 0.7541590332984924,
+      "learning_rate": 0.0013233333333333333,
+      "loss": 1.419,
+      "step": 18270
+    },
+    {
+      "epoch": 67.7,
+      "grad_norm": 0.7302483320236206,
+      "learning_rate": 0.0013229629629629629,
+      "loss": 1.4236,
+      "step": 18280
+    },
+    {
+      "epoch": 67.74,
+      "grad_norm": 0.7059903740882874,
+      "learning_rate": 0.0013225925925925927,
+      "loss": 1.4488,
+      "step": 18290
+    },
+    {
+      "epoch": 67.78,
+      "grad_norm": 0.7474135756492615,
+      "learning_rate": 0.0013222222222222222,
+      "loss": 1.4424,
+      "step": 18300
+    },
+    {
+      "epoch": 67.81,
+      "grad_norm": 0.7122178077697754,
+      "learning_rate": 0.0013218518518518518,
+      "loss": 1.4612,
+      "step": 18310
+    },
+    {
+      "epoch": 67.85,
+      "grad_norm": 0.7674033641815186,
+      "learning_rate": 0.0013214814814814814,
+      "loss": 1.4424,
+      "step": 18320
+    },
+    {
+      "epoch": 67.89,
+      "grad_norm": 0.7650808095932007,
+      "learning_rate": 0.0013211111111111112,
+      "loss": 1.4694,
+      "step": 18330
+    },
+    {
+      "epoch": 67.93,
+      "grad_norm": 0.7181142568588257,
+      "learning_rate": 0.0013207407407407408,
+      "loss": 1.4476,
+      "step": 18340
+    },
+    {
+      "epoch": 67.96,
+      "grad_norm": 0.7057814598083496,
+      "learning_rate": 0.0013203703703703704,
+      "loss": 1.4677,
+      "step": 18350
+    },
+    {
+      "epoch": 68.0,
+      "grad_norm": 1.2915374040603638,
+      "learning_rate": 0.00132,
+      "loss": 1.4835,
+      "step": 18360
+    },
+    {
+      "epoch": 68.04,
+      "grad_norm": 0.709320068359375,
+      "learning_rate": 0.0013196296296296298,
+      "loss": 1.1855,
+      "step": 18370
+    },
+    {
+      "epoch": 68.07,
+      "grad_norm": 0.7228713035583496,
+      "learning_rate": 0.0013192592592592594,
+      "loss": 1.2257,
+      "step": 18380
+    },
+    {
+      "epoch": 68.11,
+      "grad_norm": 0.7062861919403076,
+      "learning_rate": 0.001318888888888889,
+      "loss": 1.2602,
+      "step": 18390
+    },
+    {
+      "epoch": 68.15,
+      "grad_norm": 0.7732377052307129,
+      "learning_rate": 0.0013185185185185183,
+      "loss": 1.2496,
+      "step": 18400
+    },
+    {
+      "epoch": 68.19,
+      "grad_norm": 0.7629472017288208,
+      "learning_rate": 0.0013181481481481482,
+      "loss": 1.2419,
+      "step": 18410
+    },
+    {
+      "epoch": 68.22,
+      "grad_norm": 0.7411864399909973,
+      "learning_rate": 0.0013177777777777777,
+      "loss": 1.2923,
+      "step": 18420
+    },
+    {
+      "epoch": 68.26,
+      "grad_norm": 0.7572233080863953,
+      "learning_rate": 0.0013174074074074073,
+      "loss": 1.308,
+      "step": 18430
+    },
+    {
+      "epoch": 68.3,
+      "grad_norm": 0.7793846726417542,
+      "learning_rate": 0.0013170370370370371,
+      "loss": 1.3047,
+      "step": 18440
+    },
+    {
+      "epoch": 68.33,
+      "grad_norm": 0.8033546805381775,
+      "learning_rate": 0.0013166666666666667,
+      "loss": 1.3386,
+      "step": 18450
+    },
+    {
+      "epoch": 68.37,
+      "grad_norm": 0.7773953676223755,
+      "learning_rate": 0.0013162962962962963,
+      "loss": 1.3362,
+      "step": 18460
+    },
+    {
+      "epoch": 68.41,
+      "grad_norm": 0.7706006169319153,
+      "learning_rate": 0.001315925925925926,
+      "loss": 1.3677,
+      "step": 18470
+    },
+    {
+      "epoch": 68.44,
+      "grad_norm": 0.751693844795227,
+      "learning_rate": 0.0013155555555555557,
+      "loss": 1.3489,
+      "step": 18480
+    },
+    {
+      "epoch": 68.48,
+      "grad_norm": 0.7412254810333252,
+      "learning_rate": 0.0013151851851851853,
+      "loss": 1.3638,
+      "step": 18490
+    },
+    {
+      "epoch": 68.52,
+      "grad_norm": 0.704195499420166,
+      "learning_rate": 0.0013148148148148149,
+      "loss": 1.3764,
+      "step": 18500
+    },
+    {
+      "epoch": 68.56,
+      "grad_norm": 0.8066453337669373,
+      "learning_rate": 0.0013144444444444445,
+      "loss": 1.4044,
+      "step": 18510
+    },
+    {
+      "epoch": 68.59,
+      "grad_norm": 0.789995014667511,
+      "learning_rate": 0.0013140740740740743,
+      "loss": 1.424,
+      "step": 18520
+    },
+    {
+      "epoch": 68.63,
+      "grad_norm": 0.7190313339233398,
+      "learning_rate": 0.0013137037037037036,
+      "loss": 1.3813,
+      "step": 18530
+    },
+    {
+      "epoch": 68.67,
+      "grad_norm": 0.8326443433761597,
+      "learning_rate": 0.0013133333333333332,
+      "loss": 1.4548,
+      "step": 18540
+    },
+    {
+      "epoch": 68.7,
+      "grad_norm": 0.7189662456512451,
+      "learning_rate": 0.0013129629629629628,
+      "loss": 1.4213,
+      "step": 18550
+    },
+    {
+      "epoch": 68.74,
+      "grad_norm": 0.7524738907814026,
+      "learning_rate": 0.0013125925925925926,
+      "loss": 1.4392,
+      "step": 18560
+    },
+    {
+      "epoch": 68.78,
+      "grad_norm": 0.7662227749824524,
+      "learning_rate": 0.0013122222222222222,
+      "loss": 1.4195,
+      "step": 18570
+    },
+    {
+      "epoch": 68.81,
+      "grad_norm": 0.7383051514625549,
+      "learning_rate": 0.0013118518518518518,
+      "loss": 1.4304,
+      "step": 18580
+    },
+    {
+      "epoch": 68.85,
+      "grad_norm": 0.750335693359375,
+      "learning_rate": 0.0013114814814814814,
+      "loss": 1.4339,
+      "step": 18590
+    },
+    {
+      "epoch": 68.89,
+      "grad_norm": 0.6686143279075623,
+      "learning_rate": 0.0013111111111111112,
+      "loss": 1.4295,
+      "step": 18600
+    },
+    {
+      "epoch": 68.93,
+      "grad_norm": 0.7729794979095459,
+      "learning_rate": 0.0013107407407407408,
+      "loss": 1.468,
+      "step": 18610
+    },
+    {
+      "epoch": 68.96,
+      "grad_norm": 0.7255796194076538,
+      "learning_rate": 0.0013103703703703704,
+      "loss": 1.4713,
+      "step": 18620
+    },
+    {
+      "epoch": 69.0,
+      "grad_norm": 1.3524179458618164,
+      "learning_rate": 0.0013100000000000002,
+      "loss": 1.4484,
+      "step": 18630
+    },
+    {
+      "epoch": 69.04,
+      "grad_norm": 0.7976313233375549,
+      "learning_rate": 0.0013096296296296298,
+      "loss": 1.2158,
+      "step": 18640
+    },
+    {
+      "epoch": 69.07,
+      "grad_norm": 0.7665860652923584,
+      "learning_rate": 0.0013092592592592594,
+      "loss": 1.2247,
+      "step": 18650
+    },
+    {
+      "epoch": 69.11,
+      "grad_norm": 0.7670441269874573,
+      "learning_rate": 0.001308888888888889,
+      "loss": 1.2384,
+      "step": 18660
+    },
+    {
+      "epoch": 69.15,
+      "grad_norm": 0.8014783263206482,
+      "learning_rate": 0.0013085185185185185,
+      "loss": 1.2996,
+      "step": 18670
+    },
+    {
+      "epoch": 69.19,
+      "grad_norm": 0.7495048642158508,
+      "learning_rate": 0.0013081481481481481,
+      "loss": 1.2836,
+      "step": 18680
+    },
+    {
+      "epoch": 69.22,
+      "grad_norm": 0.7853214740753174,
+      "learning_rate": 0.0013077777777777777,
+      "loss": 1.2771,
+      "step": 18690
+    },
+    {
+      "epoch": 69.26,
+      "grad_norm": 0.69195157289505,
+      "learning_rate": 0.0013074074074074073,
+      "loss": 1.2946,
+      "step": 18700
+    },
+    {
+      "epoch": 69.3,
+      "grad_norm": 0.7509171962738037,
+      "learning_rate": 0.001307037037037037,
+      "loss": 1.3141,
+      "step": 18710
+    },
+    {
+      "epoch": 69.33,
+      "grad_norm": 0.7930737733840942,
+      "learning_rate": 0.0013066666666666667,
+      "loss": 1.2937,
+      "step": 18720
+    },
+    {
+      "epoch": 69.37,
+      "grad_norm": 0.7614037990570068,
+      "learning_rate": 0.0013062962962962963,
+      "loss": 1.3544,
+      "step": 18730
+    },
+    {
+      "epoch": 69.41,
+      "grad_norm": 0.7565775513648987,
+      "learning_rate": 0.0013059259259259259,
+      "loss": 1.3547,
+      "step": 18740
+    },
+    {
+      "epoch": 69.44,
+      "grad_norm": 0.7282271385192871,
+      "learning_rate": 0.0013055555555555557,
+      "loss": 1.3461,
+      "step": 18750
+    },
+    {
+      "epoch": 69.48,
+      "grad_norm": 0.717829704284668,
+      "learning_rate": 0.0013051851851851853,
+      "loss": 1.3662,
+      "step": 18760
+    },
+    {
+      "epoch": 69.52,
+      "grad_norm": 0.697563111782074,
+      "learning_rate": 0.0013048148148148149,
+      "loss": 1.3676,
+      "step": 18770
+    },
+    {
+      "epoch": 69.56,
+      "grad_norm": 0.8086504936218262,
+      "learning_rate": 0.0013044444444444444,
+      "loss": 1.384,
+      "step": 18780
+    },
+    {
+      "epoch": 69.59,
+      "grad_norm": 0.754641056060791,
+      "learning_rate": 0.0013040740740740742,
+      "loss": 1.4019,
+      "step": 18790
+    },
+    {
+      "epoch": 69.63,
+      "grad_norm": 0.7806912660598755,
+      "learning_rate": 0.0013037037037037038,
+      "loss": 1.4061,
+      "step": 18800
+    },
+    {
+      "epoch": 69.67,
+      "grad_norm": 0.7380501627922058,
+      "learning_rate": 0.0013033333333333332,
+      "loss": 1.4203,
+      "step": 18810
+    },
+    {
+      "epoch": 69.7,
+      "grad_norm": 0.8265412449836731,
+      "learning_rate": 0.0013029629629629628,
+      "loss": 1.3984,
+      "step": 18820
+    },
+    {
+      "epoch": 69.74,
+      "grad_norm": 0.7108408808708191,
+      "learning_rate": 0.0013025925925925926,
+      "loss": 1.4126,
+      "step": 18830
+    },
+    {
+      "epoch": 69.78,
+      "grad_norm": 0.7676126956939697,
+      "learning_rate": 0.0013022222222222222,
+      "loss": 1.3999,
+      "step": 18840
+    },
+    {
+      "epoch": 69.81,
+      "grad_norm": 0.8513116836547852,
+      "learning_rate": 0.0013018518518518518,
+      "loss": 1.4224,
+      "step": 18850
+    },
+    {
+      "epoch": 69.85,
+      "grad_norm": 0.7177796363830566,
+      "learning_rate": 0.0013014814814814816,
+      "loss": 1.4688,
+      "step": 18860
+    },
+    {
+      "epoch": 69.89,
+      "grad_norm": 0.7571403384208679,
+      "learning_rate": 0.0013011111111111112,
+      "loss": 1.4465,
+      "step": 18870
+    },
+    {
+      "epoch": 69.93,
+      "grad_norm": 0.7542866468429565,
+      "learning_rate": 0.0013007407407407408,
+      "loss": 1.4452,
+      "step": 18880
+    },
+    {
+      "epoch": 69.96,
+      "grad_norm": 0.7163647413253784,
+      "learning_rate": 0.0013003703703703704,
+      "loss": 1.4489,
+      "step": 18890
+    },
+    {
+      "epoch": 70.0,
+      "grad_norm": 1.3719452619552612,
+      "learning_rate": 0.0013000000000000002,
+      "loss": 1.4496,
+      "step": 18900
+    },
+    {
+      "epoch": 70.04,
+      "grad_norm": 0.7573015689849854,
+      "learning_rate": 0.0012996296296296297,
+      "loss": 1.216,
+      "step": 18910
+    },
+    {
+      "epoch": 70.07,
+      "grad_norm": 0.7383880019187927,
+      "learning_rate": 0.0012992592592592593,
+      "loss": 1.1958,
+      "step": 18920
+    },
+    {
+      "epoch": 70.11,
+      "grad_norm": 0.7401066422462463,
+      "learning_rate": 0.001298888888888889,
+      "loss": 1.2191,
+      "step": 18930
+    },
+    {
+      "epoch": 70.15,
+      "grad_norm": 0.7491777539253235,
+      "learning_rate": 0.0012985185185185187,
+      "loss": 1.257,
+      "step": 18940
+    },
+    {
+      "epoch": 70.19,
+      "grad_norm": 0.7398974895477295,
+      "learning_rate": 0.001298148148148148,
+      "loss": 1.2388,
+      "step": 18950
+    },
+    {
+      "epoch": 70.22,
+      "grad_norm": 0.7718586921691895,
+      "learning_rate": 0.0012977777777777777,
+      "loss": 1.3012,
+      "step": 18960
+    },
+    {
+      "epoch": 70.26,
+      "grad_norm": 0.7411701083183289,
+      "learning_rate": 0.0012974074074074073,
+      "loss": 1.2783,
+      "step": 18970
+    },
+    {
+      "epoch": 70.3,
+      "grad_norm": 0.764370858669281,
+      "learning_rate": 0.001297037037037037,
+      "loss": 1.2849,
+      "step": 18980
+    },
+    {
+      "epoch": 70.33,
+      "grad_norm": 0.7457252740859985,
+      "learning_rate": 0.0012966666666666667,
+      "loss": 1.3388,
+      "step": 18990
+    },
+    {
+      "epoch": 70.37,
+      "grad_norm": 0.736430287361145,
+      "learning_rate": 0.0012962962962962963,
+      "loss": 1.3276,
+      "step": 19000
+    },
+    {
+      "epoch": 70.41,
+      "grad_norm": 0.8024626970291138,
+      "learning_rate": 0.0012959259259259258,
+      "loss": 1.3272,
+      "step": 19010
+    },
+    {
+      "epoch": 70.44,
+      "grad_norm": 0.8038630485534668,
+      "learning_rate": 0.0012955555555555557,
+      "loss": 1.3592,
+      "step": 19020
+    },
+    {
+      "epoch": 70.48,
+      "grad_norm": 0.7494184970855713,
+      "learning_rate": 0.0012951851851851852,
+      "loss": 1.3458,
+      "step": 19030
+    },
+    {
+      "epoch": 70.52,
+      "grad_norm": 0.7740994095802307,
+      "learning_rate": 0.0012948148148148148,
+      "loss": 1.3552,
+      "step": 19040
+    },
+    {
+      "epoch": 70.56,
+      "grad_norm": 0.8263460993766785,
+      "learning_rate": 0.0012944444444444446,
+      "loss": 1.365,
+      "step": 19050
+    },
+    {
+      "epoch": 70.59,
+      "grad_norm": 0.7671233415603638,
+      "learning_rate": 0.0012940740740740742,
+      "loss": 1.3681,
+      "step": 19060
+    },
+    {
+      "epoch": 70.63,
+      "grad_norm": 0.7569007277488708,
+      "learning_rate": 0.0012937037037037038,
+      "loss": 1.3722,
+      "step": 19070
+    },
+    {
+      "epoch": 70.67,
+      "grad_norm": 0.748662531375885,
+      "learning_rate": 0.0012933333333333332,
+      "loss": 1.391,
+      "step": 19080
+    },
+    {
+      "epoch": 70.7,
+      "grad_norm": 0.759195864200592,
+      "learning_rate": 0.001292962962962963,
+      "loss": 1.3899,
+      "step": 19090
+    },
+    {
+      "epoch": 70.74,
+      "grad_norm": 0.7581557631492615,
+      "learning_rate": 0.0012925925925925926,
+      "loss": 1.4389,
+      "step": 19100
+    },
+    {
+      "epoch": 70.78,
+      "grad_norm": 0.7946545481681824,
+      "learning_rate": 0.0012922222222222222,
+      "loss": 1.4117,
+      "step": 19110
+    },
+    {
+      "epoch": 70.81,
+      "grad_norm": 0.7786641120910645,
+      "learning_rate": 0.0012918518518518518,
+      "loss": 1.4132,
+      "step": 19120
+    },
+    {
+      "epoch": 70.85,
+      "grad_norm": 0.8186221718788147,
+      "learning_rate": 0.0012914814814814816,
+      "loss": 1.44,
+      "step": 19130
+    },
+    {
+      "epoch": 70.89,
+      "grad_norm": 0.7414263486862183,
+      "learning_rate": 0.0012911111111111111,
+      "loss": 1.4461,
+      "step": 19140
+    },
+    {
+      "epoch": 70.93,
+      "grad_norm": 0.7749980092048645,
+      "learning_rate": 0.0012907407407407407,
+      "loss": 1.4314,
+      "step": 19150
+    },
+    {
+      "epoch": 70.96,
+      "grad_norm": 0.7037963271141052,
+      "learning_rate": 0.0012903703703703703,
+      "loss": 1.4367,
+      "step": 19160
+    },
+    {
+      "epoch": 71.0,
+      "grad_norm": 1.8344924449920654,
+      "learning_rate": 0.0012900000000000001,
+      "loss": 1.4445,
+      "step": 19170
+    },
+    {
+      "epoch": 71.04,
+      "grad_norm": 0.7934958338737488,
+      "learning_rate": 0.0012896296296296297,
+      "loss": 1.2252,
+      "step": 19180
+    },
+    {
+      "epoch": 71.07,
+      "grad_norm": 0.724882185459137,
+      "learning_rate": 0.0012892592592592593,
+      "loss": 1.231,
+      "step": 19190
+    },
+    {
+      "epoch": 71.11,
+      "grad_norm": 0.6996666789054871,
+      "learning_rate": 0.001288888888888889,
+      "loss": 1.2123,
+      "step": 19200
+    },
+    {
+      "epoch": 71.15,
+      "grad_norm": 0.7373991012573242,
+      "learning_rate": 0.0012885185185185187,
+      "loss": 1.2471,
+      "step": 19210
+    },
+    {
+      "epoch": 71.19,
+      "grad_norm": 0.8212770223617554,
+      "learning_rate": 0.001288148148148148,
+      "loss": 1.2238,
+      "step": 19220
+    },
+    {
+      "epoch": 71.22,
+      "grad_norm": 0.8789666891098022,
+      "learning_rate": 0.0012877777777777777,
+      "loss": 1.2717,
+      "step": 19230
+    },
+    {
+      "epoch": 71.26,
+      "grad_norm": 0.7911102771759033,
+      "learning_rate": 0.0012874074074074075,
+      "loss": 1.2665,
+      "step": 19240
+    },
+    {
+      "epoch": 71.3,
+      "grad_norm": 0.7351455688476562,
+      "learning_rate": 0.001287037037037037,
+      "loss": 1.3006,
+      "step": 19250
+    },
+    {
+      "epoch": 71.33,
+      "grad_norm": 0.7648353576660156,
+      "learning_rate": 0.0012866666666666666,
+      "loss": 1.3268,
+      "step": 19260
+    },
+    {
+      "epoch": 71.37,
+      "grad_norm": 0.7373409271240234,
+      "learning_rate": 0.0012862962962962962,
+      "loss": 1.3222,
+      "step": 19270
+    },
+    {
+      "epoch": 71.41,
+      "grad_norm": 0.7790569067001343,
+      "learning_rate": 0.001285925925925926,
+      "loss": 1.3216,
+      "step": 19280
+    },
+    {
+      "epoch": 71.44,
+      "grad_norm": 0.7489818930625916,
+      "learning_rate": 0.0012855555555555556,
+      "loss": 1.3577,
+      "step": 19290
+    },
+    {
+      "epoch": 71.48,
+      "grad_norm": 0.8069300055503845,
+      "learning_rate": 0.0012851851851851852,
+      "loss": 1.331,
+      "step": 19300
+    },
+    {
+      "epoch": 71.52,
+      "grad_norm": 0.8932879567146301,
+      "learning_rate": 0.0012848148148148148,
+      "loss": 1.3861,
+      "step": 19310
+    },
+    {
+      "epoch": 71.56,
+      "grad_norm": 0.803302526473999,
+      "learning_rate": 0.0012844444444444446,
+      "loss": 1.3809,
+      "step": 19320
+    },
+    {
+      "epoch": 71.59,
+      "grad_norm": 0.8255770802497864,
+      "learning_rate": 0.0012840740740740742,
+      "loss": 1.3624,
+      "step": 19330
+    },
+    {
+      "epoch": 71.63,
+      "grad_norm": 0.8509892225265503,
+      "learning_rate": 0.0012837037037037038,
+      "loss": 1.3956,
+      "step": 19340
+    },
+    {
+      "epoch": 71.67,
+      "grad_norm": 0.7588201761245728,
+      "learning_rate": 0.0012833333333333334,
+      "loss": 1.3989,
+      "step": 19350
+    },
+    {
+      "epoch": 71.7,
+      "grad_norm": 0.7393040657043457,
+      "learning_rate": 0.001282962962962963,
+      "loss": 1.3604,
+      "step": 19360
+    },
+    {
+      "epoch": 71.74,
+      "grad_norm": 0.7771015167236328,
+      "learning_rate": 0.0012825925925925926,
+      "loss": 1.399,
+      "step": 19370
+    },
+    {
+      "epoch": 71.78,
+      "grad_norm": 0.8176193237304688,
+      "learning_rate": 0.0012822222222222221,
+      "loss": 1.4179,
+      "step": 19380
+    },
+    {
+      "epoch": 71.81,
+      "grad_norm": 0.8000503778457642,
+      "learning_rate": 0.0012818518518518517,
+      "loss": 1.425,
+      "step": 19390
+    },
+    {
+      "epoch": 71.85,
+      "grad_norm": 0.7793208956718445,
+      "learning_rate": 0.0012814814814814815,
+      "loss": 1.4215,
+      "step": 19400
+    },
+    {
+      "epoch": 71.89,
+      "grad_norm": 0.7359727025032043,
+      "learning_rate": 0.0012811111111111111,
+      "loss": 1.4261,
+      "step": 19410
+    },
+    {
+      "epoch": 71.93,
+      "grad_norm": 0.7612905502319336,
+      "learning_rate": 0.0012807407407407407,
+      "loss": 1.4268,
+      "step": 19420
+    },
+    {
+      "epoch": 71.96,
+      "grad_norm": 0.7255610823631287,
+      "learning_rate": 0.0012803703703703703,
+      "loss": 1.4153,
+      "step": 19430
+    },
+    {
+      "epoch": 72.0,
+      "grad_norm": 1.6871055364608765,
+      "learning_rate": 0.00128,
+      "loss": 1.4457,
+      "step": 19440
+    },
+    {
+      "epoch": 72.04,
+      "grad_norm": 0.7359239459037781,
+      "learning_rate": 0.0012796296296296297,
+      "loss": 1.1798,
+      "step": 19450
+    },
+    {
+      "epoch": 72.07,
+      "grad_norm": 0.7320151925086975,
+      "learning_rate": 0.0012792592592592593,
+      "loss": 1.1813,
+      "step": 19460
+    },
+    {
+      "epoch": 72.11,
+      "grad_norm": 0.866115927696228,
+      "learning_rate": 0.001278888888888889,
+      "loss": 1.2234,
+      "step": 19470
+    },
+    {
+      "epoch": 72.15,
+      "grad_norm": 0.7804827690124512,
+      "learning_rate": 0.0012785185185185187,
+      "loss": 1.224,
+      "step": 19480
+    },
+    {
+      "epoch": 72.19,
+      "grad_norm": 0.7378130555152893,
+      "learning_rate": 0.0012781481481481483,
+      "loss": 1.2452,
+      "step": 19490
+    },
+    {
+      "epoch": 72.22,
+      "grad_norm": 0.8071990609169006,
+      "learning_rate": 0.0012777777777777776,
+      "loss": 1.2677,
+      "step": 19500
+    },
+    {
+      "epoch": 72.26,
+      "grad_norm": 0.7027687430381775,
+      "learning_rate": 0.0012774074074074074,
+      "loss": 1.2874,
+      "step": 19510
+    },
+    {
+      "epoch": 72.3,
+      "grad_norm": 0.8697335720062256,
+      "learning_rate": 0.001277037037037037,
+      "loss": 1.2834,
+      "step": 19520
+    },
+    {
+      "epoch": 72.33,
+      "grad_norm": 0.7541685104370117,
+      "learning_rate": 0.0012766666666666666,
+      "loss": 1.3215,
+      "step": 19530
+    },
+    {
+      "epoch": 72.37,
+      "grad_norm": 0.7561916708946228,
+      "learning_rate": 0.0012762962962962962,
+      "loss": 1.3078,
+      "step": 19540
+    },
+    {
+      "epoch": 72.41,
+      "grad_norm": 0.8127706050872803,
+      "learning_rate": 0.001275925925925926,
+      "loss": 1.3113,
+      "step": 19550
+    },
+    {
+      "epoch": 72.44,
+      "grad_norm": 0.7472105622291565,
+      "learning_rate": 0.0012755555555555556,
+      "loss": 1.3324,
+      "step": 19560
+    },
+    {
+      "epoch": 72.48,
+      "grad_norm": 0.7457278370857239,
+      "learning_rate": 0.0012751851851851852,
+      "loss": 1.3743,
+      "step": 19570
+    },
+    {
+      "epoch": 72.52,
+      "grad_norm": 0.8131669759750366,
+      "learning_rate": 0.0012748148148148148,
+      "loss": 1.3736,
+      "step": 19580
+    },
+    {
+      "epoch": 72.56,
+      "grad_norm": 0.757375955581665,
+      "learning_rate": 0.0012744444444444446,
+      "loss": 1.3808,
+      "step": 19590
+    },
+    {
+      "epoch": 72.59,
+      "grad_norm": 0.779960572719574,
+      "learning_rate": 0.0012740740740740742,
+      "loss": 1.4005,
+      "step": 19600
+    },
+    {
+      "epoch": 72.63,
+      "grad_norm": 0.7514369487762451,
+      "learning_rate": 0.0012737037037037038,
+      "loss": 1.3723,
+      "step": 19610
+    },
+    {
+      "epoch": 72.67,
+      "grad_norm": 0.8244257569313049,
+      "learning_rate": 0.0012733333333333333,
+      "loss": 1.3948,
+      "step": 19620
+    },
+    {
+      "epoch": 72.7,
+      "grad_norm": 0.8477365970611572,
+      "learning_rate": 0.0012729629629629632,
+      "loss": 1.3758,
+      "step": 19630
+    },
+    {
+      "epoch": 72.74,
+      "grad_norm": 0.7021734118461609,
+      "learning_rate": 0.0012725925925925925,
+      "loss": 1.3878,
+      "step": 19640
+    },
+    {
+      "epoch": 72.78,
+      "grad_norm": 0.7419242858886719,
+      "learning_rate": 0.0012722222222222221,
+      "loss": 1.3916,
+      "step": 19650
+    },
+    {
+      "epoch": 72.81,
+      "grad_norm": 0.7143440246582031,
+      "learning_rate": 0.001271851851851852,
+      "loss": 1.3839,
+      "step": 19660
+    },
+    {
+      "epoch": 72.85,
+      "grad_norm": 0.7648002505302429,
+      "learning_rate": 0.0012714814814814815,
+      "loss": 1.3988,
+      "step": 19670
+    },
+    {
+      "epoch": 72.89,
+      "grad_norm": 0.7516924738883972,
+      "learning_rate": 0.001271111111111111,
+      "loss": 1.4166,
+      "step": 19680
+    },
+    {
+      "epoch": 72.93,
+      "grad_norm": 0.7532698512077332,
+      "learning_rate": 0.0012707407407407407,
+      "loss": 1.4171,
+      "step": 19690
+    },
+    {
+      "epoch": 72.96,
+      "grad_norm": 0.7403442859649658,
+      "learning_rate": 0.0012703703703703705,
+      "loss": 1.4313,
+      "step": 19700
+    },
+    {
+      "epoch": 73.0,
+      "grad_norm": 1.462038278579712,
+      "learning_rate": 0.00127,
+      "loss": 1.4292,
+      "step": 19710
+    },
+    {
+      "epoch": 73.04,
+      "grad_norm": 0.7673774361610413,
+      "learning_rate": 0.0012696296296296297,
+      "loss": 1.1586,
+      "step": 19720
+    },
+    {
+      "epoch": 73.07,
+      "grad_norm": 0.75072181224823,
+      "learning_rate": 0.0012692592592592593,
+      "loss": 1.222,
+      "step": 19730
+    },
+    {
+      "epoch": 73.11,
+      "grad_norm": 0.7945419549942017,
+      "learning_rate": 0.001268888888888889,
+      "loss": 1.2124,
+      "step": 19740
+    },
+    {
+      "epoch": 73.15,
+      "grad_norm": 0.7756657600402832,
+      "learning_rate": 0.0012685185185185186,
+      "loss": 1.212,
+      "step": 19750
+    },
+    {
+      "epoch": 73.19,
+      "grad_norm": 0.760955274105072,
+      "learning_rate": 0.0012681481481481482,
+      "loss": 1.2436,
+      "step": 19760
+    },
+    {
+      "epoch": 73.22,
+      "grad_norm": 0.7394781708717346,
+      "learning_rate": 0.0012677777777777776,
+      "loss": 1.2815,
+      "step": 19770
+    },
+    {
+      "epoch": 73.26,
+      "grad_norm": 0.8197513818740845,
+      "learning_rate": 0.0012674074074074074,
+      "loss": 1.2559,
+      "step": 19780
+    },
+    {
+      "epoch": 73.3,
+      "grad_norm": 0.8319255113601685,
+      "learning_rate": 0.001267037037037037,
+      "loss": 1.2566,
+      "step": 19790
+    },
+    {
+      "epoch": 73.33,
+      "grad_norm": 0.8381143808364868,
+      "learning_rate": 0.0012666666666666666,
+      "loss": 1.2853,
+      "step": 19800
+    },
+    {
+      "epoch": 73.37,
+      "grad_norm": 0.7372511625289917,
+      "learning_rate": 0.0012662962962962962,
+      "loss": 1.3118,
+      "step": 19810
+    },
+    {
+      "epoch": 73.41,
+      "grad_norm": 0.7255342602729797,
+      "learning_rate": 0.001265925925925926,
+      "loss": 1.3363,
+      "step": 19820
+    },
+    {
+      "epoch": 73.44,
+      "grad_norm": 0.7761967182159424,
+      "learning_rate": 0.0012655555555555556,
+      "loss": 1.3206,
+      "step": 19830
+    },
+    {
+      "epoch": 73.48,
+      "grad_norm": 0.788222074508667,
+      "learning_rate": 0.0012651851851851852,
+      "loss": 1.3595,
+      "step": 19840
+    },
+    {
+      "epoch": 73.52,
+      "grad_norm": 0.8446279168128967,
+      "learning_rate": 0.001264814814814815,
+      "loss": 1.34,
+      "step": 19850
+    },
+    {
+      "epoch": 73.56,
+      "grad_norm": 0.7769805788993835,
+      "learning_rate": 0.0012644444444444446,
+      "loss": 1.3626,
+      "step": 19860
+    },
+    {
+      "epoch": 73.59,
+      "grad_norm": 0.8095614314079285,
+      "learning_rate": 0.0012640740740740741,
+      "loss": 1.3667,
+      "step": 19870
+    },
+    {
+      "epoch": 73.63,
+      "grad_norm": 0.7476008534431458,
+      "learning_rate": 0.0012637037037037037,
+      "loss": 1.3414,
+      "step": 19880
+    },
+    {
+      "epoch": 73.67,
+      "grad_norm": 0.8041726350784302,
+      "learning_rate": 0.0012633333333333335,
+      "loss": 1.3698,
+      "step": 19890
+    },
+    {
+      "epoch": 73.7,
+      "grad_norm": 0.7896828651428223,
+      "learning_rate": 0.0012629629629629631,
+      "loss": 1.3821,
+      "step": 19900
+    },
+    {
+      "epoch": 73.74,
+      "grad_norm": 0.763286828994751,
+      "learning_rate": 0.0012625925925925925,
+      "loss": 1.3715,
+      "step": 19910
+    },
+    {
+      "epoch": 73.78,
+      "grad_norm": 0.7226396203041077,
+      "learning_rate": 0.001262222222222222,
+      "loss": 1.3837,
+      "step": 19920
+    },
+    {
+      "epoch": 73.81,
+      "grad_norm": 0.8018871545791626,
+      "learning_rate": 0.001261851851851852,
+      "loss": 1.4263,
+      "step": 19930
+    },
+    {
+      "epoch": 73.85,
+      "grad_norm": 0.7554224729537964,
+      "learning_rate": 0.0012614814814814815,
+      "loss": 1.4306,
+      "step": 19940
+    },
+    {
+      "epoch": 73.89,
+      "grad_norm": 0.780648946762085,
+      "learning_rate": 0.001261111111111111,
+      "loss": 1.4078,
+      "step": 19950
+    },
+    {
+      "epoch": 73.93,
+      "grad_norm": 0.7898496389389038,
+      "learning_rate": 0.0012607407407407407,
+      "loss": 1.4097,
+      "step": 19960
+    },
+    {
+      "epoch": 73.96,
+      "grad_norm": 0.7502182126045227,
+      "learning_rate": 0.0012603703703703705,
+      "loss": 1.3955,
+      "step": 19970
+    },
+    {
+      "epoch": 74.0,
+      "grad_norm": 1.3677808046340942,
+      "learning_rate": 0.00126,
+      "loss": 1.44,
+      "step": 19980
+    },
+    {
+      "epoch": 74.04,
+      "grad_norm": 0.8023880124092102,
+      "learning_rate": 0.0012596296296296296,
+      "loss": 1.1658,
+      "step": 19990
+    },
+    {
+      "epoch": 74.07,
+      "grad_norm": 0.8229208588600159,
+      "learning_rate": 0.0012592592592592592,
+      "loss": 1.205,
+      "step": 20000
+    },
+    {
+      "epoch": 74.11,
+      "grad_norm": 0.7845343351364136,
+      "learning_rate": 0.001258888888888889,
+      "loss": 1.2099,
+      "step": 20010
+    },
+    {
+      "epoch": 74.15,
+      "grad_norm": 0.7770639657974243,
+      "learning_rate": 0.0012585185185185186,
+      "loss": 1.2291,
+      "step": 20020
+    },
+    {
+      "epoch": 74.19,
+      "grad_norm": 0.7614319920539856,
+      "learning_rate": 0.0012581481481481482,
+      "loss": 1.2247,
+      "step": 20030
+    },
+    {
+      "epoch": 74.22,
+      "grad_norm": 0.808239758014679,
+      "learning_rate": 0.001257777777777778,
+      "loss": 1.244,
+      "step": 20040
+    },
+    {
+      "epoch": 74.26,
+      "grad_norm": 0.7549338340759277,
+      "learning_rate": 0.0012574074074074074,
+      "loss": 1.2564,
+      "step": 20050
+    },
+    {
+      "epoch": 74.3,
+      "grad_norm": 0.8833985328674316,
+      "learning_rate": 0.001257037037037037,
+      "loss": 1.2776,
+      "step": 20060
+    },
+    {
+      "epoch": 74.33,
+      "grad_norm": 0.7880859971046448,
+      "learning_rate": 0.0012566666666666666,
+      "loss": 1.3141,
+      "step": 20070
+    },
+    {
+      "epoch": 74.37,
+      "grad_norm": 0.8000026941299438,
+      "learning_rate": 0.0012562962962962964,
+      "loss": 1.2965,
+      "step": 20080
+    },
+    {
+      "epoch": 74.41,
+      "grad_norm": 0.7329573035240173,
+      "learning_rate": 0.001255925925925926,
+      "loss": 1.3052,
+      "step": 20090
+    },
+    {
+      "epoch": 74.44,
+      "grad_norm": 0.7924298048019409,
+      "learning_rate": 0.0012555555555555555,
+      "loss": 1.3378,
+      "step": 20100
+    },
+    {
+      "epoch": 74.48,
+      "grad_norm": 0.7924289703369141,
+      "learning_rate": 0.0012551851851851851,
+      "loss": 1.3316,
+      "step": 20110
+    },
+    {
+      "epoch": 74.52,
+      "grad_norm": 0.8606981635093689,
+      "learning_rate": 0.001254814814814815,
+      "loss": 1.322,
+      "step": 20120
+    },
+    {
+      "epoch": 74.56,
+      "grad_norm": 0.8466325998306274,
+      "learning_rate": 0.0012544444444444445,
+      "loss": 1.3403,
+      "step": 20130
+    },
+    {
+      "epoch": 74.59,
+      "grad_norm": 0.7616990804672241,
+      "learning_rate": 0.0012540740740740741,
+      "loss": 1.3709,
+      "step": 20140
+    },
+    {
+      "epoch": 74.63,
+      "grad_norm": 0.8217259645462036,
+      "learning_rate": 0.0012537037037037037,
+      "loss": 1.3643,
+      "step": 20150
+    },
+    {
+      "epoch": 74.67,
+      "grad_norm": 0.8023414611816406,
+      "learning_rate": 0.0012533333333333335,
+      "loss": 1.3635,
+      "step": 20160
+    },
+    {
+      "epoch": 74.7,
+      "grad_norm": 0.7349652051925659,
+      "learning_rate": 0.001252962962962963,
+      "loss": 1.3707,
+      "step": 20170
+    },
+    {
+      "epoch": 74.74,
+      "grad_norm": 0.7817662358283997,
+      "learning_rate": 0.0012525925925925927,
+      "loss": 1.4089,
+      "step": 20180
+    },
+    {
+      "epoch": 74.78,
+      "grad_norm": 0.819149374961853,
+      "learning_rate": 0.001252222222222222,
+      "loss": 1.3633,
+      "step": 20190
+    },
+    {
+      "epoch": 74.81,
+      "grad_norm": 0.811070442199707,
+      "learning_rate": 0.0012518518518518519,
+      "loss": 1.3932,
+      "step": 20200
+    },
+    {
+      "epoch": 74.85,
+      "grad_norm": 0.7167898416519165,
+      "learning_rate": 0.0012514814814814815,
+      "loss": 1.412,
+      "step": 20210
+    },
+    {
+      "epoch": 74.89,
+      "grad_norm": 0.6857497692108154,
+      "learning_rate": 0.001251111111111111,
+      "loss": 1.3975,
+      "step": 20220
+    },
+    {
+      "epoch": 74.93,
+      "grad_norm": 0.7601510882377625,
+      "learning_rate": 0.0012507407407407406,
+      "loss": 1.3972,
+      "step": 20230
+    },
+    {
+      "epoch": 74.96,
+      "grad_norm": 0.7811948657035828,
+      "learning_rate": 0.0012503703703703704,
+      "loss": 1.4251,
+      "step": 20240
+    },
+    {
+      "epoch": 75.0,
+      "grad_norm": 1.9018360376358032,
+      "learning_rate": 0.00125,
+      "loss": 1.3991,
+      "step": 20250
+    },
+    {
+      "epoch": 75.04,
+      "grad_norm": 0.7820109724998474,
+      "learning_rate": 0.0012496296296296296,
+      "loss": 1.1763,
+      "step": 20260
+    },
+    {
+      "epoch": 75.07,
+      "grad_norm": 0.799656867980957,
+      "learning_rate": 0.0012492592592592594,
+      "loss": 1.1977,
+      "step": 20270
+    },
+    {
+      "epoch": 75.11,
+      "grad_norm": 0.7293298840522766,
+      "learning_rate": 0.001248888888888889,
+      "loss": 1.1909,
+      "step": 20280
+    },
+    {
+      "epoch": 75.15,
+      "grad_norm": 0.762310266494751,
+      "learning_rate": 0.0012485185185185186,
+      "loss": 1.1957,
+      "step": 20290
+    },
+    {
+      "epoch": 75.19,
+      "grad_norm": 0.8296578526496887,
+      "learning_rate": 0.0012481481481481482,
+      "loss": 1.2131,
+      "step": 20300
+    },
+    {
+      "epoch": 75.22,
+      "grad_norm": 0.7799403071403503,
+      "learning_rate": 0.001247777777777778,
+      "loss": 1.2428,
+      "step": 20310
+    },
+    {
+      "epoch": 75.26,
+      "grad_norm": 0.8273761868476868,
+      "learning_rate": 0.0012474074074074074,
+      "loss": 1.2384,
+      "step": 20320
+    },
+    {
+      "epoch": 75.3,
+      "grad_norm": 0.7832826375961304,
+      "learning_rate": 0.001247037037037037,
+      "loss": 1.2835,
+      "step": 20330
+    },
+    {
+      "epoch": 75.33,
+      "grad_norm": 0.7677596807479858,
+      "learning_rate": 0.0012466666666666665,
+      "loss": 1.2618,
+      "step": 20340
+    },
+    {
+      "epoch": 75.37,
+      "grad_norm": 0.7945502400398254,
+      "learning_rate": 0.0012462962962962963,
+      "loss": 1.2838,
+      "step": 20350
+    },
+    {
+      "epoch": 75.41,
+      "grad_norm": 0.7328624725341797,
+      "learning_rate": 0.001245925925925926,
+      "loss": 1.3011,
+      "step": 20360
+    },
+    {
+      "epoch": 75.44,
+      "grad_norm": 0.8257105946540833,
+      "learning_rate": 0.0012455555555555555,
+      "loss": 1.3188,
+      "step": 20370
+    },
+    {
+      "epoch": 75.48,
+      "grad_norm": 0.8209635615348816,
+      "learning_rate": 0.0012451851851851851,
+      "loss": 1.3072,
+      "step": 20380
+    },
+    {
+      "epoch": 75.52,
+      "grad_norm": 0.8422302603721619,
+      "learning_rate": 0.001244814814814815,
+      "loss": 1.3266,
+      "step": 20390
+    },
+    {
+      "epoch": 75.56,
+      "grad_norm": 0.8323236107826233,
+      "learning_rate": 0.0012444444444444445,
+      "loss": 1.3249,
+      "step": 20400
+    },
+    {
+      "epoch": 75.59,
+      "grad_norm": 0.7966023683547974,
+      "learning_rate": 0.001244074074074074,
+      "loss": 1.3378,
+      "step": 20410
+    },
+    {
+      "epoch": 75.63,
+      "grad_norm": 0.7316237688064575,
+      "learning_rate": 0.0012437037037037037,
+      "loss": 1.3489,
+      "step": 20420
+    },
+    {
+      "epoch": 75.67,
+      "grad_norm": 0.8376904726028442,
+      "learning_rate": 0.0012433333333333335,
+      "loss": 1.3569,
+      "step": 20430
+    },
+    {
+      "epoch": 75.7,
+      "grad_norm": 0.799994945526123,
+      "learning_rate": 0.001242962962962963,
+      "loss": 1.3984,
+      "step": 20440
+    },
+    {
+      "epoch": 75.74,
+      "grad_norm": 0.7840830087661743,
+      "learning_rate": 0.0012425925925925927,
+      "loss": 1.3776,
+      "step": 20450
+    },
+    {
+      "epoch": 75.78,
+      "grad_norm": 0.7834445834159851,
+      "learning_rate": 0.001242222222222222,
+      "loss": 1.383,
+      "step": 20460
+    },
+    {
+      "epoch": 75.81,
+      "grad_norm": 0.8726301193237305,
+      "learning_rate": 0.0012418518518518518,
+      "loss": 1.3965,
+      "step": 20470
+    },
+    {
+      "epoch": 75.85,
+      "grad_norm": 0.8033871054649353,
+      "learning_rate": 0.0012414814814814814,
+      "loss": 1.386,
+      "step": 20480
+    },
+    {
+      "epoch": 75.89,
+      "grad_norm": 0.8240883946418762,
+      "learning_rate": 0.001241111111111111,
+      "loss": 1.3963,
+      "step": 20490
+    },
+    {
+      "epoch": 75.93,
+      "grad_norm": 0.7927721738815308,
+      "learning_rate": 0.0012407407407407408,
+      "loss": 1.4127,
+      "step": 20500
+    },
+    {
+      "epoch": 75.96,
+      "grad_norm": 0.8034483194351196,
+      "learning_rate": 0.0012403703703703704,
+      "loss": 1.3915,
+      "step": 20510
+    },
+    {
+      "epoch": 76.0,
+      "grad_norm": 1.4726250171661377,
+      "learning_rate": 0.00124,
+      "loss": 1.4116,
+      "step": 20520
+    },
+    {
+      "epoch": 76.04,
+      "grad_norm": 0.775287389755249,
+      "learning_rate": 0.0012396296296296296,
+      "loss": 1.1529,
+      "step": 20530
+    },
+    {
+      "epoch": 76.07,
+      "grad_norm": 0.7530531287193298,
+      "learning_rate": 0.0012392592592592594,
+      "loss": 1.1796,
+      "step": 20540
+    },
+    {
+      "epoch": 76.11,
+      "grad_norm": 0.8483404517173767,
+      "learning_rate": 0.001238888888888889,
+      "loss": 1.1751,
+      "step": 20550
+    },
+    {
+      "epoch": 76.15,
+      "grad_norm": 0.8471804857254028,
+      "learning_rate": 0.0012385185185185186,
+      "loss": 1.1828,
+      "step": 20560
+    },
+    {
+      "epoch": 76.19,
+      "grad_norm": 0.7711694836616516,
+      "learning_rate": 0.0012381481481481482,
+      "loss": 1.2362,
+      "step": 20570
+    },
+    {
+      "epoch": 76.22,
+      "grad_norm": 0.8558454513549805,
+      "learning_rate": 0.001237777777777778,
+      "loss": 1.2296,
+      "step": 20580
+    },
+    {
+      "epoch": 76.26,
+      "grad_norm": 0.7980576753616333,
+      "learning_rate": 0.0012374074074074076,
+      "loss": 1.2518,
+      "step": 20590
+    },
+    {
+      "epoch": 76.3,
+      "grad_norm": 0.7709694504737854,
+      "learning_rate": 0.001237037037037037,
+      "loss": 1.2723,
+      "step": 20600
+    },
+    {
+      "epoch": 76.33,
+      "grad_norm": 0.7979551553726196,
+      "learning_rate": 0.0012366666666666665,
+      "loss": 1.2802,
+      "step": 20610
+    },
+    {
+      "epoch": 76.37,
+      "grad_norm": 0.7723931670188904,
+      "learning_rate": 0.0012362962962962963,
+      "loss": 1.2805,
+      "step": 20620
+    },
+    {
+      "epoch": 76.41,
+      "grad_norm": 0.7951886653900146,
+      "learning_rate": 0.001235925925925926,
+      "loss": 1.3112,
+      "step": 20630
+    },
+    {
+      "epoch": 76.44,
+      "grad_norm": 0.8366179466247559,
+      "learning_rate": 0.0012355555555555555,
+      "loss": 1.3134,
+      "step": 20640
+    },
+    {
+      "epoch": 76.48,
+      "grad_norm": 0.759067177772522,
+      "learning_rate": 0.001235185185185185,
+      "loss": 1.3315,
+      "step": 20650
+    },
+    {
+      "epoch": 76.52,
+      "grad_norm": 0.7688968777656555,
+      "learning_rate": 0.0012348148148148149,
+      "loss": 1.3062,
+      "step": 20660
+    },
+    {
+      "epoch": 76.56,
+      "grad_norm": 0.7929026484489441,
+      "learning_rate": 0.0012344444444444445,
+      "loss": 1.3243,
+      "step": 20670
+    },
+    {
+      "epoch": 76.59,
+      "grad_norm": 0.7894778251647949,
+      "learning_rate": 0.001234074074074074,
+      "loss": 1.3411,
+      "step": 20680
+    },
+    {
+      "epoch": 76.63,
+      "grad_norm": 0.7940474152565002,
+      "learning_rate": 0.0012337037037037039,
+      "loss": 1.3307,
+      "step": 20690
+    },
+    {
+      "epoch": 76.67,
+      "grad_norm": 0.7494774460792542,
+      "learning_rate": 0.0012333333333333335,
+      "loss": 1.3671,
+      "step": 20700
+    },
+    {
+      "epoch": 76.7,
+      "grad_norm": 0.8801915049552917,
+      "learning_rate": 0.001232962962962963,
+      "loss": 1.3683,
+      "step": 20710
+    },
+    {
+      "epoch": 76.74,
+      "grad_norm": 0.8223617076873779,
+      "learning_rate": 0.0012325925925925926,
+      "loss": 1.3709,
+      "step": 20720
+    },
+    {
+      "epoch": 76.78,
+      "grad_norm": 0.8362076878547668,
+      "learning_rate": 0.0012322222222222224,
+      "loss": 1.3642,
+      "step": 20730
+    },
+    {
+      "epoch": 76.81,
+      "grad_norm": 0.8495985865592957,
+      "learning_rate": 0.0012318518518518518,
+      "loss": 1.3719,
+      "step": 20740
+    },
+    {
+      "epoch": 76.85,
+      "grad_norm": 0.8225275278091431,
+      "learning_rate": 0.0012314814814814814,
+      "loss": 1.3903,
+      "step": 20750
+    },
+    {
+      "epoch": 76.89,
+      "grad_norm": 0.7975557446479797,
+      "learning_rate": 0.001231111111111111,
+      "loss": 1.4067,
+      "step": 20760
+    },
+    {
+      "epoch": 76.93,
+      "grad_norm": 0.8219361901283264,
+      "learning_rate": 0.0012307407407407408,
+      "loss": 1.4093,
+      "step": 20770
+    },
+    {
+      "epoch": 76.96,
+      "grad_norm": 0.8116274476051331,
+      "learning_rate": 0.0012303703703703704,
+      "loss": 1.4051,
+      "step": 20780
+    },
+    {
+      "epoch": 77.0,
+      "grad_norm": 1.761422872543335,
+      "learning_rate": 0.00123,
+      "loss": 1.3978,
+      "step": 20790
+    },
+    {
+      "epoch": 77.04,
+      "grad_norm": 0.7580262422561646,
+      "learning_rate": 0.0012296296296296296,
+      "loss": 1.1828,
+      "step": 20800
+    },
+    {
+      "epoch": 77.07,
+      "grad_norm": 0.7770316004753113,
+      "learning_rate": 0.0012292592592592594,
+      "loss": 1.1474,
+      "step": 20810
+    },
+    {
+      "epoch": 77.11,
+      "grad_norm": 0.7733827233314514,
+      "learning_rate": 0.001228888888888889,
+      "loss": 1.1756,
+      "step": 20820
+    },
+    {
+      "epoch": 77.15,
+      "grad_norm": 0.8824875354766846,
+      "learning_rate": 0.0012285185185185185,
+      "loss": 1.213,
+      "step": 20830
+    },
+    {
+      "epoch": 77.19,
+      "grad_norm": 0.8360825777053833,
+      "learning_rate": 0.0012281481481481481,
+      "loss": 1.2194,
+      "step": 20840
+    },
+    {
+      "epoch": 77.22,
+      "grad_norm": 0.790859580039978,
+      "learning_rate": 0.001227777777777778,
+      "loss": 1.2376,
+      "step": 20850
+    },
+    {
+      "epoch": 77.26,
+      "grad_norm": 0.8717347383499146,
+      "learning_rate": 0.0012274074074074075,
+      "loss": 1.2512,
+      "step": 20860
+    },
+    {
+      "epoch": 77.3,
+      "grad_norm": 0.7890602350234985,
+      "learning_rate": 0.001227037037037037,
+      "loss": 1.2397,
+      "step": 20870
+    },
+    {
+      "epoch": 77.33,
+      "grad_norm": 0.8594658970832825,
+      "learning_rate": 0.0012266666666666667,
+      "loss": 1.2704,
+      "step": 20880
+    },
+    {
+      "epoch": 77.37,
+      "grad_norm": 0.8138068914413452,
+      "learning_rate": 0.0012262962962962963,
+      "loss": 1.2879,
+      "step": 20890
+    },
+    {
+      "epoch": 77.41,
+      "grad_norm": 0.8145601749420166,
+      "learning_rate": 0.0012259259259259259,
+      "loss": 1.3026,
+      "step": 20900
+    },
+    {
+      "epoch": 77.44,
+      "grad_norm": 0.8223507404327393,
+      "learning_rate": 0.0012255555555555555,
+      "loss": 1.3116,
+      "step": 20910
+    },
+    {
+      "epoch": 77.48,
+      "grad_norm": 0.8235211372375488,
+      "learning_rate": 0.0012251851851851853,
+      "loss": 1.3071,
+      "step": 20920
+    },
+    {
+      "epoch": 77.52,
+      "grad_norm": 0.7376195788383484,
+      "learning_rate": 0.0012248148148148149,
+      "loss": 1.3193,
+      "step": 20930
+    },
+    {
+      "epoch": 77.56,
+      "grad_norm": 0.7939662337303162,
+      "learning_rate": 0.0012244444444444445,
+      "loss": 1.3261,
+      "step": 20940
+    },
+    {
+      "epoch": 77.59,
+      "grad_norm": 0.8232659101486206,
+      "learning_rate": 0.001224074074074074,
+      "loss": 1.3285,
+      "step": 20950
+    },
+    {
+      "epoch": 77.63,
+      "grad_norm": 0.772993266582489,
+      "learning_rate": 0.0012237037037037038,
+      "loss": 1.3368,
+      "step": 20960
+    },
+    {
+      "epoch": 77.67,
+      "grad_norm": 0.8146404027938843,
+      "learning_rate": 0.0012233333333333334,
+      "loss": 1.3365,
+      "step": 20970
+    },
+    {
+      "epoch": 77.7,
+      "grad_norm": 0.7433125972747803,
+      "learning_rate": 0.001222962962962963,
+      "loss": 1.3431,
+      "step": 20980
+    },
+    {
+      "epoch": 77.74,
+      "grad_norm": 0.8123118281364441,
+      "learning_rate": 0.0012225925925925926,
+      "loss": 1.3616,
+      "step": 20990
+    },
+    {
+      "epoch": 77.78,
+      "grad_norm": 0.7485589385032654,
+      "learning_rate": 0.0012222222222222224,
+      "loss": 1.3594,
+      "step": 21000
+    },
+    {
+      "epoch": 77.81,
+      "grad_norm": 0.8744904398918152,
+      "learning_rate": 0.0012218518518518518,
+      "loss": 1.3389,
+      "step": 21010
+    },
+    {
+      "epoch": 77.85,
+      "grad_norm": 0.813572883605957,
+      "learning_rate": 0.0012214814814814814,
+      "loss": 1.3902,
+      "step": 21020
+    },
+    {
+      "epoch": 77.89,
+      "grad_norm": 0.8318865895271301,
+      "learning_rate": 0.001221111111111111,
+      "loss": 1.3729,
+      "step": 21030
+    },
+    {
+      "epoch": 77.93,
+      "grad_norm": 0.7886088490486145,
+      "learning_rate": 0.0012207407407407408,
+      "loss": 1.3966,
+      "step": 21040
+    },
+    {
+      "epoch": 77.96,
+      "grad_norm": 0.7813388109207153,
+      "learning_rate": 0.0012203703703703704,
+      "loss": 1.3769,
+      "step": 21050
+    },
+    {
+      "epoch": 78.0,
+      "grad_norm": 1.5267912149429321,
+      "learning_rate": 0.00122,
+      "loss": 1.408,
+      "step": 21060
+    },
+    {
+      "epoch": 78.04,
+      "grad_norm": 0.7442604899406433,
+      "learning_rate": 0.0012196296296296298,
+      "loss": 1.1388,
+      "step": 21070
+    },
+    {
+      "epoch": 78.07,
+      "grad_norm": 0.7419655919075012,
+      "learning_rate": 0.0012192592592592593,
+      "loss": 1.1539,
+      "step": 21080
+    },
+    {
+      "epoch": 78.11,
+      "grad_norm": 0.8307773470878601,
+      "learning_rate": 0.001218888888888889,
+      "loss": 1.1785,
+      "step": 21090
+    },
+    {
+      "epoch": 78.15,
+      "grad_norm": 0.7722887992858887,
+      "learning_rate": 0.0012185185185185185,
+      "loss": 1.1858,
+      "step": 21100
+    },
+    {
+      "epoch": 78.19,
+      "grad_norm": 0.7633727192878723,
+      "learning_rate": 0.0012181481481481483,
+      "loss": 1.1952,
+      "step": 21110
+    },
+    {
+      "epoch": 78.22,
+      "grad_norm": 0.82599937915802,
+      "learning_rate": 0.001217777777777778,
+      "loss": 1.2302,
+      "step": 21120
+    },
+    {
+      "epoch": 78.26,
+      "grad_norm": 0.8898715376853943,
+      "learning_rate": 0.0012174074074074075,
+      "loss": 1.222,
+      "step": 21130
+    },
+    {
+      "epoch": 78.3,
+      "grad_norm": 0.8221423625946045,
+      "learning_rate": 0.001217037037037037,
+      "loss": 1.2533,
+      "step": 21140
+    },
+    {
+      "epoch": 78.33,
+      "grad_norm": 0.8007521629333496,
+      "learning_rate": 0.0012166666666666667,
+      "loss": 1.2614,
+      "step": 21150
+    },
+    {
+      "epoch": 78.37,
+      "grad_norm": 0.7668291926383972,
+      "learning_rate": 0.0012162962962962963,
+      "loss": 1.262,
+      "step": 21160
+    },
+    {
+      "epoch": 78.41,
+      "grad_norm": 0.8179112672805786,
+      "learning_rate": 0.0012159259259259259,
+      "loss": 1.2717,
+      "step": 21170
+    },
+    {
+      "epoch": 78.44,
+      "grad_norm": 0.8355836272239685,
+      "learning_rate": 0.0012155555555555554,
+      "loss": 1.2962,
+      "step": 21180
+    },
+    {
+      "epoch": 78.48,
+      "grad_norm": 0.7732020020484924,
+      "learning_rate": 0.0012151851851851852,
+      "loss": 1.322,
+      "step": 21190
+    },
+    {
+      "epoch": 78.52,
+      "grad_norm": 0.8221396207809448,
+      "learning_rate": 0.0012148148148148148,
+      "loss": 1.2944,
+      "step": 21200
+    },
+    {
+      "epoch": 78.56,
+      "grad_norm": 0.8312259316444397,
+      "learning_rate": 0.0012144444444444444,
+      "loss": 1.3004,
+      "step": 21210
+    },
+    {
+      "epoch": 78.59,
+      "grad_norm": 0.8105566501617432,
+      "learning_rate": 0.001214074074074074,
+      "loss": 1.2993,
+      "step": 21220
+    },
+    {
+      "epoch": 78.63,
+      "grad_norm": 0.8554087281227112,
+      "learning_rate": 0.0012137037037037038,
+      "loss": 1.3287,
+      "step": 21230
+    },
+    {
+      "epoch": 78.67,
+      "grad_norm": 0.8251190185546875,
+      "learning_rate": 0.0012133333333333334,
+      "loss": 1.3235,
+      "step": 21240
+    },
+    {
+      "epoch": 78.7,
+      "grad_norm": 0.8064236640930176,
+      "learning_rate": 0.001212962962962963,
+      "loss": 1.3465,
+      "step": 21250
+    },
+    {
+      "epoch": 78.74,
+      "grad_norm": 0.75641930103302,
+      "learning_rate": 0.0012125925925925926,
+      "loss": 1.3483,
+      "step": 21260
+    },
+    {
+      "epoch": 78.78,
+      "grad_norm": 0.7814801335334778,
+      "learning_rate": 0.0012122222222222224,
+      "loss": 1.3438,
+      "step": 21270
+    },
+    {
+      "epoch": 78.81,
+      "grad_norm": 0.8190613389015198,
+      "learning_rate": 0.001211851851851852,
+      "loss": 1.3722,
+      "step": 21280
+    },
+    {
+      "epoch": 78.85,
+      "grad_norm": 0.8852077722549438,
+      "learning_rate": 0.0012114814814814814,
+      "loss": 1.3676,
+      "step": 21290
+    },
+    {
+      "epoch": 78.89,
+      "grad_norm": 0.8047147393226624,
+      "learning_rate": 0.0012111111111111112,
+      "loss": 1.3839,
+      "step": 21300
+    },
+    {
+      "epoch": 78.93,
+      "grad_norm": 0.7829154133796692,
+      "learning_rate": 0.0012107407407407407,
+      "loss": 1.3801,
+      "step": 21310
+    },
+    {
+      "epoch": 78.96,
+      "grad_norm": 0.7903611063957214,
+      "learning_rate": 0.0012103703703703703,
+      "loss": 1.409,
+      "step": 21320
+    },
+    {
+      "epoch": 79.0,
+      "grad_norm": 1.498512864112854,
+      "learning_rate": 0.00121,
+      "loss": 1.3902,
+      "step": 21330
+    },
+    {
+      "epoch": 79.04,
+      "grad_norm": 0.8592652082443237,
+      "learning_rate": 0.0012096296296296297,
+      "loss": 1.1362,
+      "step": 21340
+    },
+    {
+      "epoch": 79.07,
+      "grad_norm": 0.8371859192848206,
+      "learning_rate": 0.0012092592592592593,
+      "loss": 1.1476,
+      "step": 21350
+    },
+    {
+      "epoch": 79.11,
+      "grad_norm": 0.8219607472419739,
+      "learning_rate": 0.001208888888888889,
+      "loss": 1.1573,
+      "step": 21360
+    },
+    {
+      "epoch": 79.15,
+      "grad_norm": 0.7551726698875427,
+      "learning_rate": 0.0012085185185185185,
+      "loss": 1.1812,
+      "step": 21370
+    },
+    {
+      "epoch": 79.19,
+      "grad_norm": 0.774641215801239,
+      "learning_rate": 0.0012081481481481483,
+      "loss": 1.199,
+      "step": 21380
+    },
+    {
+      "epoch": 79.22,
+      "grad_norm": 0.8316019773483276,
+      "learning_rate": 0.0012077777777777779,
+      "loss": 1.23,
+      "step": 21390
+    },
+    {
+      "epoch": 79.26,
+      "grad_norm": 0.8777797222137451,
+      "learning_rate": 0.0012074074074074075,
+      "loss": 1.2181,
+      "step": 21400
+    },
+    {
+      "epoch": 79.3,
+      "grad_norm": 0.8278775215148926,
+      "learning_rate": 0.001207037037037037,
+      "loss": 1.2521,
+      "step": 21410
+    },
+    {
+      "epoch": 79.33,
+      "grad_norm": 0.8261604309082031,
+      "learning_rate": 0.0012066666666666669,
+      "loss": 1.2502,
+      "step": 21420
+    },
+    {
+      "epoch": 79.37,
+      "grad_norm": 0.784346342086792,
+      "learning_rate": 0.0012062962962962962,
+      "loss": 1.2532,
+      "step": 21430
+    },
+    {
+      "epoch": 79.41,
+      "grad_norm": 0.7922409772872925,
+      "learning_rate": 0.0012059259259259258,
+      "loss": 1.2805,
+      "step": 21440
+    },
+    {
+      "epoch": 79.44,
+      "grad_norm": 0.8692464232444763,
+      "learning_rate": 0.0012055555555555554,
+      "loss": 1.2817,
+      "step": 21450
+    },
+    {
+      "epoch": 79.48,
+      "grad_norm": 0.8120541572570801,
+      "learning_rate": 0.0012051851851851852,
+      "loss": 1.3181,
+      "step": 21460
+    },
+    {
+      "epoch": 79.52,
+      "grad_norm": 0.8435699939727783,
+      "learning_rate": 0.0012048148148148148,
+      "loss": 1.3104,
+      "step": 21470
+    },
+    {
+      "epoch": 79.56,
+      "grad_norm": 0.8154937028884888,
+      "learning_rate": 0.0012044444444444444,
+      "loss": 1.333,
+      "step": 21480
+    },
+    {
+      "epoch": 79.59,
+      "grad_norm": 0.852403461933136,
+      "learning_rate": 0.0012040740740740742,
+      "loss": 1.304,
+      "step": 21490
+    },
+    {
+      "epoch": 79.63,
+      "grad_norm": 0.8190922141075134,
+      "learning_rate": 0.0012037037037037038,
+      "loss": 1.3346,
+      "step": 21500
+    },
+    {
+      "epoch": 79.67,
+      "grad_norm": 0.8322864770889282,
+      "learning_rate": 0.0012033333333333334,
+      "loss": 1.3163,
+      "step": 21510
+    },
+    {
+      "epoch": 79.7,
+      "grad_norm": 0.8163490295410156,
+      "learning_rate": 0.001202962962962963,
+      "loss": 1.342,
+      "step": 21520
+    },
+    {
+      "epoch": 79.74,
+      "grad_norm": 0.8012131452560425,
+      "learning_rate": 0.0012025925925925928,
+      "loss": 1.336,
+      "step": 21530
+    },
+    {
+      "epoch": 79.78,
+      "grad_norm": 0.8321338295936584,
+      "learning_rate": 0.0012022222222222224,
+      "loss": 1.3515,
+      "step": 21540
+    },
+    {
+      "epoch": 79.81,
+      "grad_norm": 0.772287130355835,
+      "learning_rate": 0.001201851851851852,
+      "loss": 1.3789,
+      "step": 21550
+    },
+    {
+      "epoch": 79.85,
+      "grad_norm": 0.7578256130218506,
+      "learning_rate": 0.0012014814814814813,
+      "loss": 1.3435,
+      "step": 21560
+    },
+    {
+      "epoch": 79.89,
+      "grad_norm": 0.7756651043891907,
+      "learning_rate": 0.0012011111111111111,
+      "loss": 1.3559,
+      "step": 21570
+    },
+    {
+      "epoch": 79.93,
+      "grad_norm": 0.7970812916755676,
+      "learning_rate": 0.0012007407407407407,
+      "loss": 1.3538,
+      "step": 21580
+    },
+    {
+      "epoch": 79.96,
+      "grad_norm": 0.8378322124481201,
+      "learning_rate": 0.0012003703703703703,
+      "loss": 1.3555,
+      "step": 21590
+    },
+    {
+      "epoch": 80.0,
+      "grad_norm": 1.7049368619918823,
+      "learning_rate": 0.0012,
+      "loss": 1.3829,
+      "step": 21600
+    },
+    {
+      "epoch": 80.04,
+      "grad_norm": 0.7638210654258728,
+      "learning_rate": 0.0011996296296296297,
+      "loss": 1.1375,
+      "step": 21610
+    },
+    {
+      "epoch": 80.07,
+      "grad_norm": 0.7893775701522827,
+      "learning_rate": 0.0011992592592592593,
+      "loss": 1.1598,
+      "step": 21620
+    },
+    {
+      "epoch": 80.11,
+      "grad_norm": 0.7928591370582581,
+      "learning_rate": 0.0011988888888888889,
+      "loss": 1.1812,
+      "step": 21630
+    },
+    {
+      "epoch": 80.15,
+      "grad_norm": 0.7909486293792725,
+      "learning_rate": 0.0011985185185185185,
+      "loss": 1.156,
+      "step": 21640
+    },
+    {
+      "epoch": 80.19,
+      "grad_norm": 0.7979161143302917,
+      "learning_rate": 0.0011981481481481483,
+      "loss": 1.2035,
+      "step": 21650
+    },
+    {
+      "epoch": 80.22,
+      "grad_norm": 0.7922539114952087,
+      "learning_rate": 0.0011977777777777779,
+      "loss": 1.2042,
+      "step": 21660
+    },
+    {
+      "epoch": 80.26,
+      "grad_norm": 0.7434383630752563,
+      "learning_rate": 0.0011974074074074074,
+      "loss": 1.2278,
+      "step": 21670
+    },
+    {
+      "epoch": 80.3,
+      "grad_norm": 0.814395010471344,
+      "learning_rate": 0.0011970370370370373,
+      "loss": 1.2216,
+      "step": 21680
+    },
+    {
+      "epoch": 80.33,
+      "grad_norm": 0.77228182554245,
+      "learning_rate": 0.0011966666666666668,
+      "loss": 1.2295,
+      "step": 21690
+    },
+    {
+      "epoch": 80.37,
+      "grad_norm": 0.8123186230659485,
+      "learning_rate": 0.0011962962962962962,
+      "loss": 1.2581,
+      "step": 21700
+    },
+    {
+      "epoch": 80.41,
+      "grad_norm": 0.7760822176933289,
+      "learning_rate": 0.0011959259259259258,
+      "loss": 1.2721,
+      "step": 21710
+    },
+    {
+      "epoch": 80.44,
+      "grad_norm": 0.8242724537849426,
+      "learning_rate": 0.0011955555555555556,
+      "loss": 1.245,
+      "step": 21720
+    },
+    {
+      "epoch": 80.48,
+      "grad_norm": 0.823330819606781,
+      "learning_rate": 0.0011951851851851852,
+      "loss": 1.2789,
+      "step": 21730
+    },
+    {
+      "epoch": 80.52,
+      "grad_norm": 0.758445143699646,
+      "learning_rate": 0.0011948148148148148,
+      "loss": 1.276,
+      "step": 21740
+    },
+    {
+      "epoch": 80.56,
+      "grad_norm": 0.8475410342216492,
+      "learning_rate": 0.0011944444444444444,
+      "loss": 1.3156,
+      "step": 21750
+    },
+    {
+      "epoch": 80.59,
+      "grad_norm": 0.8873695731163025,
+      "learning_rate": 0.0011940740740740742,
+      "loss": 1.3206,
+      "step": 21760
+    },
+    {
+      "epoch": 80.63,
+      "grad_norm": 0.7427895665168762,
+      "learning_rate": 0.0011937037037037038,
+      "loss": 1.3145,
+      "step": 21770
+    },
+    {
+      "epoch": 80.67,
+      "grad_norm": 0.7813780307769775,
+      "learning_rate": 0.0011933333333333334,
+      "loss": 1.3108,
+      "step": 21780
+    },
+    {
+      "epoch": 80.7,
+      "grad_norm": 0.8349199891090393,
+      "learning_rate": 0.001192962962962963,
+      "loss": 1.3172,
+      "step": 21790
+    },
+    {
+      "epoch": 80.74,
+      "grad_norm": 0.8662471175193787,
+      "learning_rate": 0.0011925925925925927,
+      "loss": 1.3509,
+      "step": 21800
+    },
+    {
+      "epoch": 80.78,
+      "grad_norm": 0.7870455980300903,
+      "learning_rate": 0.0011922222222222223,
+      "loss": 1.345,
+      "step": 21810
+    },
+    {
+      "epoch": 80.81,
+      "grad_norm": 0.8087645769119263,
+      "learning_rate": 0.001191851851851852,
+      "loss": 1.34,
+      "step": 21820
+    },
+    {
+      "epoch": 80.85,
+      "grad_norm": 0.8210803866386414,
+      "learning_rate": 0.0011914814814814815,
+      "loss": 1.3496,
+      "step": 21830
+    },
+    {
+      "epoch": 80.89,
+      "grad_norm": 0.7690105438232422,
+      "learning_rate": 0.001191111111111111,
+      "loss": 1.3824,
+      "step": 21840
+    },
+    {
+      "epoch": 80.93,
+      "grad_norm": 0.7893791794776917,
+      "learning_rate": 0.0011907407407407407,
+      "loss": 1.3591,
+      "step": 21850
+    },
+    {
+      "epoch": 80.96,
+      "grad_norm": 0.7507551312446594,
+      "learning_rate": 0.0011903703703703703,
+      "loss": 1.3562,
+      "step": 21860
+    },
+    {
+      "epoch": 81.0,
+      "grad_norm": 1.2659991979599,
+      "learning_rate": 0.0011899999999999999,
+      "loss": 1.3887,
+      "step": 21870
+    },
+    {
+      "epoch": 81.04,
+      "grad_norm": 0.7406259179115295,
+      "learning_rate": 0.0011896296296296297,
+      "loss": 1.1216,
+      "step": 21880
+    },
+    {
+      "epoch": 81.07,
+      "grad_norm": 0.8164389133453369,
+      "learning_rate": 0.0011892592592592593,
+      "loss": 1.1308,
+      "step": 21890
+    },
+    {
+      "epoch": 81.11,
+      "grad_norm": 0.7737013101577759,
+      "learning_rate": 0.0011888888888888889,
+      "loss": 1.1673,
+      "step": 21900
+    },
+    {
+      "epoch": 81.15,
+      "grad_norm": 0.7928569912910461,
+      "learning_rate": 0.0011885185185185187,
+      "loss": 1.1678,
+      "step": 21910
+    },
+    {
+      "epoch": 81.19,
+      "grad_norm": 0.8306735754013062,
+      "learning_rate": 0.0011881481481481482,
+      "loss": 1.1786,
+      "step": 21920
+    },
+    {
+      "epoch": 81.22,
+      "grad_norm": 0.7711634635925293,
+      "learning_rate": 0.0011877777777777778,
+      "loss": 1.2144,
+      "step": 21930
+    },
+    {
+      "epoch": 81.26,
+      "grad_norm": 0.8662676811218262,
+      "learning_rate": 0.0011874074074074074,
+      "loss": 1.1809,
+      "step": 21940
+    },
+    {
+      "epoch": 81.3,
+      "grad_norm": 0.8031598329544067,
+      "learning_rate": 0.0011870370370370372,
+      "loss": 1.2241,
+      "step": 21950
+    },
+    {
+      "epoch": 81.33,
+      "grad_norm": 0.797000527381897,
+      "learning_rate": 0.0011866666666666668,
+      "loss": 1.2302,
+      "step": 21960
+    },
+    {
+      "epoch": 81.37,
+      "grad_norm": 0.9229769706726074,
+      "learning_rate": 0.0011862962962962964,
+      "loss": 1.2381,
+      "step": 21970
+    },
+    {
+      "epoch": 81.41,
+      "grad_norm": 0.9003340601921082,
+      "learning_rate": 0.0011859259259259258,
+      "loss": 1.2554,
+      "step": 21980
+    },
+    {
+      "epoch": 81.44,
+      "grad_norm": 0.798082172870636,
+      "learning_rate": 0.0011855555555555556,
+      "loss": 1.2481,
+      "step": 21990
+    },
+    {
+      "epoch": 81.48,
+      "grad_norm": 0.8588927984237671,
+      "learning_rate": 0.0011851851851851852,
+      "loss": 1.2658,
+      "step": 22000
+    },
+    {
+      "epoch": 81.52,
+      "grad_norm": 0.846261203289032,
+      "learning_rate": 0.0011848148148148148,
+      "loss": 1.2796,
+      "step": 22010
+    },
+    {
+      "epoch": 81.56,
+      "grad_norm": 0.86174076795578,
+      "learning_rate": 0.0011844444444444443,
+      "loss": 1.3243,
+      "step": 22020
+    },
+    {
+      "epoch": 81.59,
+      "grad_norm": 0.7928067445755005,
+      "learning_rate": 0.0011840740740740742,
+      "loss": 1.2979,
+      "step": 22030
+    },
+    {
+      "epoch": 81.63,
+      "grad_norm": 0.8082354068756104,
+      "learning_rate": 0.0011837037037037037,
+      "loss": 1.311,
+      "step": 22040
+    },
+    {
+      "epoch": 81.67,
+      "grad_norm": 0.8235210180282593,
+      "learning_rate": 0.0011833333333333333,
+      "loss": 1.3244,
+      "step": 22050
+    },
+    {
+      "epoch": 81.7,
+      "grad_norm": 0.8794702887535095,
+      "learning_rate": 0.001182962962962963,
+      "loss": 1.305,
+      "step": 22060
+    },
+    {
+      "epoch": 81.74,
+      "grad_norm": 0.7763599753379822,
+      "learning_rate": 0.0011825925925925927,
+      "loss": 1.3389,
+      "step": 22070
+    },
+    {
+      "epoch": 81.78,
+      "grad_norm": 0.8223667740821838,
+      "learning_rate": 0.0011822222222222223,
+      "loss": 1.3309,
+      "step": 22080
+    },
+    {
+      "epoch": 81.81,
+      "grad_norm": 0.8330878615379333,
+      "learning_rate": 0.001181851851851852,
+      "loss": 1.3424,
+      "step": 22090
+    },
+    {
+      "epoch": 81.85,
+      "grad_norm": 0.8290947079658508,
+      "learning_rate": 0.0011814814814814817,
+      "loss": 1.3654,
+      "step": 22100
+    },
+    {
+      "epoch": 81.89,
+      "grad_norm": 0.7992547750473022,
+      "learning_rate": 0.001181111111111111,
+      "loss": 1.3522,
+      "step": 22110
+    },
+    {
+      "epoch": 81.93,
+      "grad_norm": 0.795511782169342,
+      "learning_rate": 0.0011807407407407407,
+      "loss": 1.3555,
+      "step": 22120
+    },
+    {
+      "epoch": 81.96,
+      "grad_norm": 0.8905317187309265,
+      "learning_rate": 0.0011803703703703703,
+      "loss": 1.3828,
+      "step": 22130
+    },
+    {
+      "epoch": 82.0,
+      "grad_norm": 1.376718521118164,
+      "learning_rate": 0.00118,
+      "loss": 1.3494,
+      "step": 22140
+    },
+    {
+      "epoch": 82.04,
+      "grad_norm": 0.8007773160934448,
+      "learning_rate": 0.0011796296296296296,
+      "loss": 1.1338,
+      "step": 22150
+    },
+    {
+      "epoch": 82.07,
+      "grad_norm": 0.7765668630599976,
+      "learning_rate": 0.0011792592592592592,
+      "loss": 1.1248,
+      "step": 22160
+    },
+    {
+      "epoch": 82.11,
+      "grad_norm": 0.8055347800254822,
+      "learning_rate": 0.0011788888888888888,
+      "loss": 1.1597,
+      "step": 22170
+    },
+    {
+      "epoch": 82.15,
+      "grad_norm": 0.8667762279510498,
+      "learning_rate": 0.0011785185185185186,
+      "loss": 1.1803,
+      "step": 22180
+    },
+    {
+      "epoch": 82.19,
+      "grad_norm": 0.8008133769035339,
+      "learning_rate": 0.0011781481481481482,
+      "loss": 1.1927,
+      "step": 22190
+    },
+    {
+      "epoch": 82.22,
+      "grad_norm": 0.8304885625839233,
+      "learning_rate": 0.0011777777777777778,
+      "loss": 1.1807,
+      "step": 22200
+    },
+    {
+      "epoch": 82.26,
+      "grad_norm": 0.8007597327232361,
+      "learning_rate": 0.0011774074074074074,
+      "loss": 1.2058,
+      "step": 22210
+    },
+    {
+      "epoch": 82.3,
+      "grad_norm": 0.8569217920303345,
+      "learning_rate": 0.0011770370370370372,
+      "loss": 1.2233,
+      "step": 22220
+    },
+    {
+      "epoch": 82.33,
+      "grad_norm": 0.8300482034683228,
+      "learning_rate": 0.0011766666666666668,
+      "loss": 1.2384,
+      "step": 22230
+    },
+    {
+      "epoch": 82.37,
+      "grad_norm": 0.8271434903144836,
+      "learning_rate": 0.0011762962962962964,
+      "loss": 1.2295,
+      "step": 22240
+    },
+    {
+      "epoch": 82.41,
+      "grad_norm": 0.7994027733802795,
+      "learning_rate": 0.0011759259259259257,
+      "loss": 1.235,
+      "step": 22250
+    },
+    {
+      "epoch": 82.44,
+      "grad_norm": 0.7732693552970886,
+      "learning_rate": 0.0011755555555555556,
+      "loss": 1.2754,
+      "step": 22260
+    },
+    {
+      "epoch": 82.48,
+      "grad_norm": 0.8843140602111816,
+      "learning_rate": 0.0011751851851851851,
+      "loss": 1.2734,
+      "step": 22270
+    },
+    {
+      "epoch": 82.52,
+      "grad_norm": 0.7856510281562805,
+      "learning_rate": 0.0011748148148148147,
+      "loss": 1.2753,
+      "step": 22280
+    },
+    {
+      "epoch": 82.56,
+      "grad_norm": 0.851887583732605,
+      "learning_rate": 0.0011744444444444443,
+      "loss": 1.2706,
+      "step": 22290
+    },
+    {
+      "epoch": 82.59,
+      "grad_norm": 0.8285180926322937,
+      "learning_rate": 0.0011740740740740741,
+      "loss": 1.2877,
+      "step": 22300
+    },
+    {
+      "epoch": 82.63,
+      "grad_norm": 0.8394416570663452,
+      "learning_rate": 0.0011737037037037037,
+      "loss": 1.2833,
+      "step": 22310
+    },
+    {
+      "epoch": 82.67,
+      "grad_norm": 0.8958646059036255,
+      "learning_rate": 0.0011733333333333333,
+      "loss": 1.311,
+      "step": 22320
+    },
+    {
+      "epoch": 82.7,
+      "grad_norm": 0.8369964361190796,
+      "learning_rate": 0.001172962962962963,
+      "loss": 1.3369,
+      "step": 22330
+    },
+    {
+      "epoch": 82.74,
+      "grad_norm": 0.8263498544692993,
+      "learning_rate": 0.0011725925925925927,
+      "loss": 1.3226,
+      "step": 22340
+    },
+    {
+      "epoch": 82.78,
+      "grad_norm": 0.7990745306015015,
+      "learning_rate": 0.0011722222222222223,
+      "loss": 1.3485,
+      "step": 22350
+    },
+    {
+      "epoch": 82.81,
+      "grad_norm": 0.7908133864402771,
+      "learning_rate": 0.0011718518518518519,
+      "loss": 1.3262,
+      "step": 22360
+    },
+    {
+      "epoch": 82.85,
+      "grad_norm": 0.8254690170288086,
+      "learning_rate": 0.0011714814814814817,
+      "loss": 1.3466,
+      "step": 22370
+    },
+    {
+      "epoch": 82.89,
+      "grad_norm": 0.8180753588676453,
+      "learning_rate": 0.0011711111111111113,
+      "loss": 1.3433,
+      "step": 22380
+    },
+    {
+      "epoch": 82.93,
+      "grad_norm": 0.8374766707420349,
+      "learning_rate": 0.0011707407407407406,
+      "loss": 1.3298,
+      "step": 22390
+    },
+    {
+      "epoch": 82.96,
+      "grad_norm": 0.8209193348884583,
+      "learning_rate": 0.0011703703703703702,
+      "loss": 1.3422,
+      "step": 22400
+    },
+    {
+      "epoch": 83.0,
+      "grad_norm": 1.5476839542388916,
+      "learning_rate": 0.00117,
+      "loss": 1.356,
+      "step": 22410
+    },
+    {
+      "epoch": 83.04,
+      "grad_norm": 0.9864056706428528,
+      "learning_rate": 0.0011696296296296296,
+      "loss": 1.1234,
+      "step": 22420
+    },
+    {
+      "epoch": 83.07,
+      "grad_norm": 0.847768247127533,
+      "learning_rate": 0.0011692592592592592,
+      "loss": 1.1251,
+      "step": 22430
+    },
+    {
+      "epoch": 83.11,
+      "grad_norm": 0.8513081073760986,
+      "learning_rate": 0.0011688888888888888,
+      "loss": 1.132,
+      "step": 22440
+    },
+    {
+      "epoch": 83.15,
+      "grad_norm": 0.7776345610618591,
+      "learning_rate": 0.0011685185185185186,
+      "loss": 1.1227,
+      "step": 22450
+    },
+    {
+      "epoch": 83.19,
+      "grad_norm": 0.8347729444503784,
+      "learning_rate": 0.0011681481481481482,
+      "loss": 1.178,
+      "step": 22460
+    },
+    {
+      "epoch": 83.22,
+      "grad_norm": 0.8397579193115234,
+      "learning_rate": 0.0011677777777777778,
+      "loss": 1.1669,
+      "step": 22470
+    },
+    {
+      "epoch": 83.26,
+      "grad_norm": 0.8121944665908813,
+      "learning_rate": 0.0011674074074074074,
+      "loss": 1.1843,
+      "step": 22480
+    },
+    {
+      "epoch": 83.3,
+      "grad_norm": 0.8163562417030334,
+      "learning_rate": 0.0011670370370370372,
+      "loss": 1.2016,
+      "step": 22490
+    },
+    {
+      "epoch": 83.33,
+      "grad_norm": 0.7758229970932007,
+      "learning_rate": 0.0011666666666666668,
+      "loss": 1.2412,
+      "step": 22500
+    },
+    {
+      "epoch": 83.37,
+      "grad_norm": 0.8472750186920166,
+      "learning_rate": 0.0011662962962962964,
+      "loss": 1.2395,
+      "step": 22510
+    },
+    {
+      "epoch": 83.41,
+      "grad_norm": 0.802095353603363,
+      "learning_rate": 0.0011659259259259262,
+      "loss": 1.2375,
+      "step": 22520
+    },
+    {
+      "epoch": 83.44,
+      "grad_norm": 0.8349648714065552,
+      "learning_rate": 0.0011655555555555555,
+      "loss": 1.2775,
+      "step": 22530
+    },
+    {
+      "epoch": 83.48,
+      "grad_norm": 0.8553137183189392,
+      "learning_rate": 0.0011651851851851851,
+      "loss": 1.2516,
+      "step": 22540
+    },
+    {
+      "epoch": 83.52,
+      "grad_norm": 0.836848795413971,
+      "learning_rate": 0.0011648148148148147,
+      "loss": 1.2481,
+      "step": 22550
+    },
+    {
+      "epoch": 83.56,
+      "grad_norm": 0.7819425463676453,
+      "learning_rate": 0.0011644444444444445,
+      "loss": 1.2914,
+      "step": 22560
+    },
+    {
+      "epoch": 83.59,
+      "grad_norm": 0.8485893607139587,
+      "learning_rate": 0.001164074074074074,
+      "loss": 1.3089,
+      "step": 22570
+    },
+    {
+      "epoch": 83.63,
+      "grad_norm": 0.8001742362976074,
+      "learning_rate": 0.0011637037037037037,
+      "loss": 1.3022,
+      "step": 22580
+    },
+    {
+      "epoch": 83.67,
+      "grad_norm": 0.7983673810958862,
+      "learning_rate": 0.0011633333333333333,
+      "loss": 1.2967,
+      "step": 22590
+    },
+    {
+      "epoch": 83.7,
+      "grad_norm": 0.8425639867782593,
+      "learning_rate": 0.001162962962962963,
+      "loss": 1.3126,
+      "step": 22600
+    },
+    {
+      "epoch": 83.74,
+      "grad_norm": 0.8058281540870667,
+      "learning_rate": 0.0011625925925925927,
+      "loss": 1.3247,
+      "step": 22610
+    },
+    {
+      "epoch": 83.78,
+      "grad_norm": 0.8269002437591553,
+      "learning_rate": 0.0011622222222222223,
+      "loss": 1.3237,
+      "step": 22620
+    },
+    {
+      "epoch": 83.81,
+      "grad_norm": 0.8307337164878845,
+      "learning_rate": 0.0011618518518518518,
+      "loss": 1.3169,
+      "step": 22630
+    },
+    {
+      "epoch": 83.85,
+      "grad_norm": 0.8474606275558472,
+      "learning_rate": 0.0011614814814814817,
+      "loss": 1.3301,
+      "step": 22640
+    },
+    {
+      "epoch": 83.89,
+      "grad_norm": 0.7992864847183228,
+      "learning_rate": 0.0011611111111111112,
+      "loss": 1.3275,
+      "step": 22650
+    },
+    {
+      "epoch": 83.93,
+      "grad_norm": 0.9049937129020691,
+      "learning_rate": 0.0011607407407407408,
+      "loss": 1.3411,
+      "step": 22660
+    },
+    {
+      "epoch": 83.96,
+      "grad_norm": 0.8497620224952698,
+      "learning_rate": 0.0011603703703703702,
+      "loss": 1.343,
+      "step": 22670
+    },
+    {
+      "epoch": 84.0,
+      "grad_norm": 1.48526132106781,
+      "learning_rate": 0.00116,
+      "loss": 1.3499,
+      "step": 22680
+    },
+    {
+      "epoch": 84.04,
+      "grad_norm": 0.8892446756362915,
+      "learning_rate": 0.0011596296296296296,
+      "loss": 1.1125,
+      "step": 22690
+    },
+    {
+      "epoch": 84.07,
+      "grad_norm": 0.7651263475418091,
+      "learning_rate": 0.0011592592592592592,
+      "loss": 1.0929,
+      "step": 22700
+    },
+    {
+      "epoch": 84.11,
+      "grad_norm": 0.8239293098449707,
+      "learning_rate": 0.001158888888888889,
+      "loss": 1.1195,
+      "step": 22710
+    },
+    {
+      "epoch": 84.15,
+      "grad_norm": 0.8643993139266968,
+      "learning_rate": 0.0011585185185185186,
+      "loss": 1.1429,
+      "step": 22720
+    },
+    {
+      "epoch": 84.19,
+      "grad_norm": 0.806343138217926,
+      "learning_rate": 0.0011581481481481482,
+      "loss": 1.1518,
+      "step": 22730
+    },
+    {
+      "epoch": 84.22,
+      "grad_norm": 0.9068303108215332,
+      "learning_rate": 0.0011577777777777778,
+      "loss": 1.1607,
+      "step": 22740
+    },
+    {
+      "epoch": 84.26,
+      "grad_norm": 0.8175188899040222,
+      "learning_rate": 0.0011574074074074076,
+      "loss": 1.2107,
+      "step": 22750
+    },
+    {
+      "epoch": 84.3,
+      "grad_norm": 0.8616414070129395,
+      "learning_rate": 0.0011570370370370371,
+      "loss": 1.209,
+      "step": 22760
+    },
+    {
+      "epoch": 84.33,
+      "grad_norm": 0.8178994655609131,
+      "learning_rate": 0.0011566666666666667,
+      "loss": 1.2123,
+      "step": 22770
+    },
+    {
+      "epoch": 84.37,
+      "grad_norm": 0.831458568572998,
+      "learning_rate": 0.0011562962962962963,
+      "loss": 1.221,
+      "step": 22780
+    },
+    {
+      "epoch": 84.41,
+      "grad_norm": 0.7806716561317444,
+      "learning_rate": 0.0011559259259259261,
+      "loss": 1.2491,
+      "step": 22790
+    },
+    {
+      "epoch": 84.44,
+      "grad_norm": 0.816102147102356,
+      "learning_rate": 0.0011555555555555555,
+      "loss": 1.2511,
+      "step": 22800
+    },
+    {
+      "epoch": 84.48,
+      "grad_norm": 0.824440062046051,
+      "learning_rate": 0.001155185185185185,
+      "loss": 1.2438,
+      "step": 22810
+    },
+    {
+      "epoch": 84.52,
+      "grad_norm": 0.8962615728378296,
+      "learning_rate": 0.0011548148148148147,
+      "loss": 1.2523,
+      "step": 22820
+    },
+    {
+      "epoch": 84.56,
+      "grad_norm": 0.7982916831970215,
+      "learning_rate": 0.0011544444444444445,
+      "loss": 1.2795,
+      "step": 22830
+    },
+    {
+      "epoch": 84.59,
+      "grad_norm": 0.9069809913635254,
+      "learning_rate": 0.001154074074074074,
+      "loss": 1.2509,
+      "step": 22840
+    },
+    {
+      "epoch": 84.63,
+      "grad_norm": 0.8316929936408997,
+      "learning_rate": 0.0011537037037037037,
+      "loss": 1.3157,
+      "step": 22850
+    },
+    {
+      "epoch": 84.67,
+      "grad_norm": 0.8130622506141663,
+      "learning_rate": 0.0011533333333333333,
+      "loss": 1.2828,
+      "step": 22860
+    },
+    {
+      "epoch": 84.7,
+      "grad_norm": 0.8644458651542664,
+      "learning_rate": 0.001152962962962963,
+      "loss": 1.3089,
+      "step": 22870
+    },
+    {
+      "epoch": 84.74,
+      "grad_norm": 0.8116170763969421,
+      "learning_rate": 0.0011525925925925926,
+      "loss": 1.3101,
+      "step": 22880
+    },
+    {
+      "epoch": 84.78,
+      "grad_norm": 0.7835599184036255,
+      "learning_rate": 0.0011522222222222222,
+      "loss": 1.2817,
+      "step": 22890
+    },
+    {
+      "epoch": 84.81,
+      "grad_norm": 0.9315503239631653,
+      "learning_rate": 0.0011518518518518518,
+      "loss": 1.3308,
+      "step": 22900
+    },
+    {
+      "epoch": 84.85,
+      "grad_norm": 0.823969304561615,
+      "learning_rate": 0.0011514814814814816,
+      "loss": 1.3179,
+      "step": 22910
+    },
+    {
+      "epoch": 84.89,
+      "grad_norm": 0.8243674635887146,
+      "learning_rate": 0.0011511111111111112,
+      "loss": 1.3323,
+      "step": 22920
+    },
+    {
+      "epoch": 84.93,
+      "grad_norm": 0.8511907458305359,
+      "learning_rate": 0.0011507407407407408,
+      "loss": 1.3621,
+      "step": 22930
+    },
+    {
+      "epoch": 84.96,
+      "grad_norm": 0.8778294324874878,
+      "learning_rate": 0.0011503703703703704,
+      "loss": 1.3481,
+      "step": 22940
+    },
+    {
+      "epoch": 85.0,
+      "grad_norm": 1.4845128059387207,
+      "learning_rate": 0.00115,
+      "loss": 1.3523,
+      "step": 22950
+    },
+    {
+      "epoch": 85.04,
+      "grad_norm": 0.7813113927841187,
+      "learning_rate": 0.0011496296296296296,
+      "loss": 1.1324,
+      "step": 22960
+    },
+    {
+      "epoch": 85.07,
+      "grad_norm": 0.881803035736084,
+      "learning_rate": 0.0011492592592592592,
+      "loss": 1.1028,
+      "step": 22970
+    },
+    {
+      "epoch": 85.11,
+      "grad_norm": 0.8418541550636292,
+      "learning_rate": 0.001148888888888889,
+      "loss": 1.1163,
+      "step": 22980
+    },
+    {
+      "epoch": 85.15,
+      "grad_norm": 0.8543863296508789,
+      "learning_rate": 0.0011485185185185186,
+      "loss": 1.1579,
+      "step": 22990
+    },
+    {
+      "epoch": 85.19,
+      "grad_norm": 0.9971126317977905,
+      "learning_rate": 0.0011481481481481481,
+      "loss": 1.1375,
+      "step": 23000
+    },
+    {
+      "epoch": 85.22,
+      "grad_norm": 0.90798419713974,
+      "learning_rate": 0.0011477777777777777,
+      "loss": 1.1714,
+      "step": 23010
+    },
+    {
+      "epoch": 85.26,
+      "grad_norm": 0.8441275954246521,
+      "learning_rate": 0.0011474074074074075,
+      "loss": 1.1726,
+      "step": 23020
+    },
+    {
+      "epoch": 85.3,
+      "grad_norm": 0.8868027925491333,
+      "learning_rate": 0.0011470370370370371,
+      "loss": 1.2162,
+      "step": 23030
+    },
+    {
+      "epoch": 85.33,
+      "grad_norm": 0.832936704158783,
+      "learning_rate": 0.0011466666666666667,
+      "loss": 1.2021,
+      "step": 23040
+    },
+    {
+      "epoch": 85.37,
+      "grad_norm": 0.8773447871208191,
+      "learning_rate": 0.0011462962962962963,
+      "loss": 1.2186,
+      "step": 23050
+    },
+    {
+      "epoch": 85.41,
+      "grad_norm": 0.89570552110672,
+      "learning_rate": 0.001145925925925926,
+      "loss": 1.2462,
+      "step": 23060
+    },
+    {
+      "epoch": 85.44,
+      "grad_norm": 0.8472298979759216,
+      "learning_rate": 0.0011455555555555557,
+      "loss": 1.225,
+      "step": 23070
+    },
+    {
+      "epoch": 85.48,
+      "grad_norm": 0.8481177687644958,
+      "learning_rate": 0.001145185185185185,
+      "loss": 1.261,
+      "step": 23080
+    },
+    {
+      "epoch": 85.52,
+      "grad_norm": 0.8076756596565247,
+      "learning_rate": 0.0011448148148148147,
+      "loss": 1.2411,
+      "step": 23090
+    },
+    {
+      "epoch": 85.56,
+      "grad_norm": 0.8581802845001221,
+      "learning_rate": 0.0011444444444444445,
+      "loss": 1.2508,
+      "step": 23100
+    },
+    {
+      "epoch": 85.59,
+      "grad_norm": 0.8750701546669006,
+      "learning_rate": 0.001144074074074074,
+      "loss": 1.24,
+      "step": 23110
+    },
+    {
+      "epoch": 85.63,
+      "grad_norm": 0.9328594207763672,
+      "learning_rate": 0.0011437037037037036,
+      "loss": 1.2903,
+      "step": 23120
+    },
+    {
+      "epoch": 85.67,
+      "grad_norm": 0.850226104259491,
+      "learning_rate": 0.0011433333333333334,
+      "loss": 1.2963,
+      "step": 23130
+    },
+    {
+      "epoch": 85.7,
+      "grad_norm": 0.81463223695755,
+      "learning_rate": 0.001142962962962963,
+      "loss": 1.3,
+      "step": 23140
+    },
+    {
+      "epoch": 85.74,
+      "grad_norm": 0.7953822016716003,
+      "learning_rate": 0.0011425925925925926,
+      "loss": 1.2952,
+      "step": 23150
+    },
+    {
+      "epoch": 85.78,
+      "grad_norm": 0.8868482112884521,
+      "learning_rate": 0.0011422222222222222,
+      "loss": 1.3008,
+      "step": 23160
+    },
+    {
+      "epoch": 85.81,
+      "grad_norm": 0.8748762607574463,
+      "learning_rate": 0.001141851851851852,
+      "loss": 1.3137,
+      "step": 23170
+    },
+    {
+      "epoch": 85.85,
+      "grad_norm": 0.8665784001350403,
+      "learning_rate": 0.0011414814814814816,
+      "loss": 1.3211,
+      "step": 23180
+    },
+    {
+      "epoch": 85.89,
+      "grad_norm": 0.8576427698135376,
+      "learning_rate": 0.0011411111111111112,
+      "loss": 1.3432,
+      "step": 23190
+    },
+    {
+      "epoch": 85.93,
+      "grad_norm": 0.8850252032279968,
+      "learning_rate": 0.0011407407407407408,
+      "loss": 1.3025,
+      "step": 23200
+    },
+    {
+      "epoch": 85.96,
+      "grad_norm": 0.8282588720321655,
+      "learning_rate": 0.0011403703703703706,
+      "loss": 1.3204,
+      "step": 23210
+    },
+    {
+      "epoch": 86.0,
+      "grad_norm": 1.6549211740493774,
+      "learning_rate": 0.00114,
+      "loss": 1.3311,
+      "step": 23220
+    },
+    {
+      "epoch": 86.04,
+      "grad_norm": 0.8455803394317627,
+      "learning_rate": 0.0011396296296296295,
+      "loss": 1.1147,
+      "step": 23230
+    },
+    {
+      "epoch": 86.07,
+      "grad_norm": 0.8316532969474792,
+      "learning_rate": 0.0011392592592592591,
+      "loss": 1.0876,
+      "step": 23240
+    },
+    {
+      "epoch": 86.11,
+      "grad_norm": 0.8999677300453186,
+      "learning_rate": 0.001138888888888889,
+      "loss": 1.1153,
+      "step": 23250
+    },
+    {
+      "epoch": 86.15,
+      "grad_norm": 0.8368595242500305,
+      "learning_rate": 0.0011385185185185185,
+      "loss": 1.1155,
+      "step": 23260
+    },
+    {
+      "epoch": 86.19,
+      "grad_norm": 0.8152023553848267,
+      "learning_rate": 0.0011381481481481481,
+      "loss": 1.1636,
+      "step": 23270
+    },
+    {
+      "epoch": 86.22,
+      "grad_norm": 0.851401150226593,
+      "learning_rate": 0.0011377777777777777,
+      "loss": 1.1503,
+      "step": 23280
+    },
+    {
+      "epoch": 86.26,
+      "grad_norm": 0.8852841854095459,
+      "learning_rate": 0.0011374074074074075,
+      "loss": 1.1782,
+      "step": 23290
+    },
+    {
+      "epoch": 86.3,
+      "grad_norm": 0.8195506930351257,
+      "learning_rate": 0.001137037037037037,
+      "loss": 1.1948,
+      "step": 23300
+    },
+    {
+      "epoch": 86.33,
+      "grad_norm": 0.9015066623687744,
+      "learning_rate": 0.0011366666666666667,
+      "loss": 1.2219,
+      "step": 23310
+    },
+    {
+      "epoch": 86.37,
+      "grad_norm": 0.9026146531105042,
+      "learning_rate": 0.0011362962962962965,
+      "loss": 1.2189,
+      "step": 23320
+    },
+    {
+      "epoch": 86.41,
+      "grad_norm": 0.8678985238075256,
+      "learning_rate": 0.001135925925925926,
+      "loss": 1.1943,
+      "step": 23330
+    },
+    {
+      "epoch": 86.44,
+      "grad_norm": 0.7977427244186401,
+      "learning_rate": 0.0011355555555555557,
+      "loss": 1.2379,
+      "step": 23340
+    },
+    {
+      "epoch": 86.48,
+      "grad_norm": 0.8538101315498352,
+      "learning_rate": 0.001135185185185185,
+      "loss": 1.2376,
+      "step": 23350
+    },
+    {
+      "epoch": 86.52,
+      "grad_norm": 0.8959237337112427,
+      "learning_rate": 0.0011348148148148148,
+      "loss": 1.2397,
+      "step": 23360
+    },
+    {
+      "epoch": 86.56,
+      "grad_norm": 0.7996075749397278,
+      "learning_rate": 0.0011344444444444444,
+      "loss": 1.2342,
+      "step": 23370
+    },
+    {
+      "epoch": 86.59,
+      "grad_norm": 0.8509365320205688,
+      "learning_rate": 0.001134074074074074,
+      "loss": 1.2578,
+      "step": 23380
+    },
+    {
+      "epoch": 86.63,
+      "grad_norm": 0.8598107099533081,
+      "learning_rate": 0.0011337037037037036,
+      "loss": 1.2738,
+      "step": 23390
+    },
+    {
+      "epoch": 86.67,
+      "grad_norm": 0.830723762512207,
+      "learning_rate": 0.0011333333333333334,
+      "loss": 1.2829,
+      "step": 23400
+    },
+    {
+      "epoch": 86.7,
+      "grad_norm": 0.8874588012695312,
+      "learning_rate": 0.001132962962962963,
+      "loss": 1.297,
+      "step": 23410
+    },
+    {
+      "epoch": 86.74,
+      "grad_norm": 0.8866403698921204,
+      "learning_rate": 0.0011325925925925926,
+      "loss": 1.273,
+      "step": 23420
+    },
+    {
+      "epoch": 86.78,
+      "grad_norm": 0.8580409288406372,
+      "learning_rate": 0.0011322222222222222,
+      "loss": 1.2863,
+      "step": 23430
+    },
+    {
+      "epoch": 86.81,
+      "grad_norm": 0.8711990714073181,
+      "learning_rate": 0.001131851851851852,
+      "loss": 1.3363,
+      "step": 23440
+    },
+    {
+      "epoch": 86.85,
+      "grad_norm": 0.876951277256012,
+      "learning_rate": 0.0011314814814814816,
+      "loss": 1.2952,
+      "step": 23450
+    },
+    {
+      "epoch": 86.89,
+      "grad_norm": 0.8068513870239258,
+      "learning_rate": 0.0011311111111111112,
+      "loss": 1.2992,
+      "step": 23460
+    },
+    {
+      "epoch": 86.93,
+      "grad_norm": 0.8394823670387268,
+      "learning_rate": 0.0011307407407407408,
+      "loss": 1.3135,
+      "step": 23470
+    },
+    {
+      "epoch": 86.96,
+      "grad_norm": 0.8134591579437256,
+      "learning_rate": 0.0011303703703703706,
+      "loss": 1.3362,
+      "step": 23480
+    },
+    {
+      "epoch": 87.0,
+      "grad_norm": 1.8034298419952393,
+      "learning_rate": 0.00113,
+      "loss": 1.3274,
+      "step": 23490
+    },
+    {
+      "epoch": 87.04,
+      "grad_norm": 0.7800376415252686,
+      "learning_rate": 0.0011296296296296295,
+      "loss": 1.095,
+      "step": 23500
+    },
+    {
+      "epoch": 87.07,
+      "grad_norm": 0.7570058703422546,
+      "learning_rate": 0.001129259259259259,
+      "loss": 1.0836,
+      "step": 23510
+    },
+    {
+      "epoch": 87.11,
+      "grad_norm": 0.8313126564025879,
+      "learning_rate": 0.001128888888888889,
+      "loss": 1.1297,
+      "step": 23520
+    },
+    {
+      "epoch": 87.15,
+      "grad_norm": 0.8045477271080017,
+      "learning_rate": 0.0011285185185185185,
+      "loss": 1.119,
+      "step": 23530
+    },
+    {
+      "epoch": 87.19,
+      "grad_norm": 0.8991451859474182,
+      "learning_rate": 0.001128148148148148,
+      "loss": 1.1516,
+      "step": 23540
+    },
+    {
+      "epoch": 87.22,
+      "grad_norm": 0.8472118377685547,
+      "learning_rate": 0.001127777777777778,
+      "loss": 1.1534,
+      "step": 23550
+    },
+    {
+      "epoch": 87.26,
+      "grad_norm": 0.8414618372917175,
+      "learning_rate": 0.0011274074074074075,
+      "loss": 1.1633,
+      "step": 23560
+    },
+    {
+      "epoch": 87.3,
+      "grad_norm": 0.820631206035614,
+      "learning_rate": 0.001127037037037037,
+      "loss": 1.1679,
+      "step": 23570
+    },
+    {
+      "epoch": 87.33,
+      "grad_norm": 0.8632519841194153,
+      "learning_rate": 0.0011266666666666667,
+      "loss": 1.2005,
+      "step": 23580
+    },
+    {
+      "epoch": 87.37,
+      "grad_norm": 0.896757185459137,
+      "learning_rate": 0.0011262962962962965,
+      "loss": 1.201,
+      "step": 23590
+    },
+    {
+      "epoch": 87.41,
+      "grad_norm": 0.815828263759613,
+      "learning_rate": 0.001125925925925926,
+      "loss": 1.236,
+      "step": 23600
+    },
+    {
+      "epoch": 87.44,
+      "grad_norm": 0.8946236371994019,
+      "learning_rate": 0.0011255555555555556,
+      "loss": 1.2244,
+      "step": 23610
+    },
+    {
+      "epoch": 87.48,
+      "grad_norm": 0.8468908667564392,
+      "learning_rate": 0.0011251851851851852,
+      "loss": 1.237,
+      "step": 23620
+    },
+    {
+      "epoch": 87.52,
+      "grad_norm": 0.8856787085533142,
+      "learning_rate": 0.0011248148148148148,
+      "loss": 1.2426,
+      "step": 23630
+    },
+    {
+      "epoch": 87.56,
+      "grad_norm": 0.9098236560821533,
+      "learning_rate": 0.0011244444444444444,
+      "loss": 1.2613,
+      "step": 23640
+    },
+    {
+      "epoch": 87.59,
+      "grad_norm": 0.8589096665382385,
+      "learning_rate": 0.001124074074074074,
+      "loss": 1.2704,
+      "step": 23650
+    },
+    {
+      "epoch": 87.63,
+      "grad_norm": 0.9214912056922913,
+      "learning_rate": 0.0011237037037037036,
+      "loss": 1.243,
+      "step": 23660
+    },
+    {
+      "epoch": 87.67,
+      "grad_norm": 0.8244339227676392,
+      "learning_rate": 0.0011233333333333334,
+      "loss": 1.2651,
+      "step": 23670
+    },
+    {
+      "epoch": 87.7,
+      "grad_norm": 0.892946183681488,
+      "learning_rate": 0.001122962962962963,
+      "loss": 1.3073,
+      "step": 23680
+    },
+    {
+      "epoch": 87.74,
+      "grad_norm": 0.9119607210159302,
+      "learning_rate": 0.0011225925925925926,
+      "loss": 1.2674,
+      "step": 23690
+    },
+    {
+      "epoch": 87.78,
+      "grad_norm": 0.7816550731658936,
+      "learning_rate": 0.0011222222222222222,
+      "loss": 1.2668,
+      "step": 23700
+    },
+    {
+      "epoch": 87.81,
+      "grad_norm": 0.8281494379043579,
+      "learning_rate": 0.001121851851851852,
+      "loss": 1.2764,
+      "step": 23710
+    },
+    {
+      "epoch": 87.85,
+      "grad_norm": 0.8064820170402527,
+      "learning_rate": 0.0011214814814814815,
+      "loss": 1.3038,
+      "step": 23720
+    },
+    {
+      "epoch": 87.89,
+      "grad_norm": 0.8691319823265076,
+      "learning_rate": 0.0011211111111111111,
+      "loss": 1.3299,
+      "step": 23730
+    },
+    {
+      "epoch": 87.93,
+      "grad_norm": 0.8435444831848145,
+      "learning_rate": 0.001120740740740741,
+      "loss": 1.292,
+      "step": 23740
+    },
+    {
+      "epoch": 87.96,
+      "grad_norm": 0.8803039789199829,
+      "learning_rate": 0.0011203703703703705,
+      "loss": 1.3089,
+      "step": 23750
+    },
+    {
+      "epoch": 88.0,
+      "grad_norm": 1.5175323486328125,
+      "learning_rate": 0.0011200000000000001,
+      "loss": 1.3126,
+      "step": 23760
+    },
+    {
+      "epoch": 88.04,
+      "grad_norm": 0.807668149471283,
+      "learning_rate": 0.0011196296296296295,
+      "loss": 1.0896,
+      "step": 23770
+    },
+    {
+      "epoch": 88.07,
+      "grad_norm": 0.8427616953849792,
+      "learning_rate": 0.0011192592592592593,
+      "loss": 1.0739,
+      "step": 23780
+    },
+    {
+      "epoch": 88.11,
+      "grad_norm": 0.8259497880935669,
+      "learning_rate": 0.0011188888888888889,
+      "loss": 1.0804,
+      "step": 23790
+    },
+    {
+      "epoch": 88.15,
+      "grad_norm": 0.8123213052749634,
+      "learning_rate": 0.0011185185185185185,
+      "loss": 1.1136,
+      "step": 23800
+    },
+    {
+      "epoch": 88.19,
+      "grad_norm": 0.8309470415115356,
+      "learning_rate": 0.001118148148148148,
+      "loss": 1.109,
+      "step": 23810
+    },
+    {
+      "epoch": 88.22,
+      "grad_norm": 0.7814146876335144,
+      "learning_rate": 0.0011177777777777779,
+      "loss": 1.1466,
+      "step": 23820
+    },
+    {
+      "epoch": 88.26,
+      "grad_norm": 0.9524590969085693,
+      "learning_rate": 0.0011174074074074075,
+      "loss": 1.1338,
+      "step": 23830
+    },
+    {
+      "epoch": 88.3,
+      "grad_norm": 0.8634507656097412,
+      "learning_rate": 0.001117037037037037,
+      "loss": 1.183,
+      "step": 23840
+    },
+    {
+      "epoch": 88.33,
+      "grad_norm": 0.7970640659332275,
+      "learning_rate": 0.0011166666666666666,
+      "loss": 1.2017,
+      "step": 23850
+    },
+    {
+      "epoch": 88.37,
+      "grad_norm": 0.8886352777481079,
+      "learning_rate": 0.0011162962962962964,
+      "loss": 1.1993,
+      "step": 23860
+    },
+    {
+      "epoch": 88.41,
+      "grad_norm": 0.8125959038734436,
+      "learning_rate": 0.001115925925925926,
+      "loss": 1.1988,
+      "step": 23870
+    },
+    {
+      "epoch": 88.44,
+      "grad_norm": 0.8432615995407104,
+      "learning_rate": 0.0011155555555555556,
+      "loss": 1.213,
+      "step": 23880
+    },
+    {
+      "epoch": 88.48,
+      "grad_norm": 0.8891323804855347,
+      "learning_rate": 0.0011151851851851852,
+      "loss": 1.2315,
+      "step": 23890
+    },
+    {
+      "epoch": 88.52,
+      "grad_norm": 0.8381022214889526,
+      "learning_rate": 0.001114814814814815,
+      "loss": 1.2435,
+      "step": 23900
+    },
+    {
+      "epoch": 88.56,
+      "grad_norm": 0.8534558415412903,
+      "learning_rate": 0.0011144444444444444,
+      "loss": 1.2576,
+      "step": 23910
+    },
+    {
+      "epoch": 88.59,
+      "grad_norm": 0.8542389273643494,
+      "learning_rate": 0.001114074074074074,
+      "loss": 1.2542,
+      "step": 23920
+    },
+    {
+      "epoch": 88.63,
+      "grad_norm": 0.8446402549743652,
+      "learning_rate": 0.0011137037037037036,
+      "loss": 1.2542,
+      "step": 23930
+    },
+    {
+      "epoch": 88.67,
+      "grad_norm": 0.9329958558082581,
+      "learning_rate": 0.0011133333333333334,
+      "loss": 1.2625,
+      "step": 23940
+    },
+    {
+      "epoch": 88.7,
+      "grad_norm": 0.8419949412345886,
+      "learning_rate": 0.001112962962962963,
+      "loss": 1.2744,
+      "step": 23950
+    },
+    {
+      "epoch": 88.74,
+      "grad_norm": 0.8235114812850952,
+      "learning_rate": 0.0011125925925925925,
+      "loss": 1.2628,
+      "step": 23960
+    },
+    {
+      "epoch": 88.78,
+      "grad_norm": 0.8560745120048523,
+      "learning_rate": 0.0011122222222222223,
+      "loss": 1.2809,
+      "step": 23970
+    },
+    {
+      "epoch": 88.81,
+      "grad_norm": 0.8349394202232361,
+      "learning_rate": 0.001111851851851852,
+      "loss": 1.2943,
+      "step": 23980
+    },
+    {
+      "epoch": 88.85,
+      "grad_norm": 0.9164478778839111,
+      "learning_rate": 0.0011114814814814815,
+      "loss": 1.3001,
+      "step": 23990
+    },
+    {
+      "epoch": 88.89,
+      "grad_norm": 0.9206733107566833,
+      "learning_rate": 0.0011111111111111111,
+      "loss": 1.2985,
+      "step": 24000
+    },
+    {
+      "epoch": 88.93,
+      "grad_norm": 0.8984618186950684,
+      "learning_rate": 0.001110740740740741,
+      "loss": 1.2869,
+      "step": 24010
+    },
+    {
+      "epoch": 88.96,
+      "grad_norm": 0.8212258219718933,
+      "learning_rate": 0.0011103703703703705,
+      "loss": 1.3071,
+      "step": 24020
+    },
+    {
+      "epoch": 89.0,
+      "grad_norm": 1.393999457359314,
+      "learning_rate": 0.00111,
+      "loss": 1.3113,
+      "step": 24030
+    },
+    {
+      "epoch": 89.04,
+      "grad_norm": 0.8700305223464966,
+      "learning_rate": 0.0011096296296296295,
+      "loss": 1.1065,
+      "step": 24040
+    },
+    {
+      "epoch": 89.07,
+      "grad_norm": 0.769532322883606,
+      "learning_rate": 0.0011092592592592593,
+      "loss": 1.0946,
+      "step": 24050
+    },
+    {
+      "epoch": 89.11,
+      "grad_norm": 0.8377566933631897,
+      "learning_rate": 0.0011088888888888889,
+      "loss": 1.0966,
+      "step": 24060
+    },
+    {
+      "epoch": 89.15,
+      "grad_norm": 0.8229069113731384,
+      "learning_rate": 0.0011085185185185184,
+      "loss": 1.1145,
+      "step": 24070
+    },
+    {
+      "epoch": 89.19,
+      "grad_norm": 0.8501279354095459,
+      "learning_rate": 0.001108148148148148,
+      "loss": 1.11,
+      "step": 24080
+    },
+    {
+      "epoch": 89.22,
+      "grad_norm": 0.8882164359092712,
+      "learning_rate": 0.0011077777777777778,
+      "loss": 1.1364,
+      "step": 24090
+    },
+    {
+      "epoch": 89.26,
+      "grad_norm": 0.8216584324836731,
+      "learning_rate": 0.0011074074074074074,
+      "loss": 1.1462,
+      "step": 24100
+    },
+    {
+      "epoch": 89.3,
+      "grad_norm": 0.944375216960907,
+      "learning_rate": 0.001107037037037037,
+      "loss": 1.166,
+      "step": 24110
+    },
+    {
+      "epoch": 89.33,
+      "grad_norm": 0.9312400221824646,
+      "learning_rate": 0.0011066666666666666,
+      "loss": 1.2047,
+      "step": 24120
+    },
+    {
+      "epoch": 89.37,
+      "grad_norm": 0.8870893716812134,
+      "learning_rate": 0.0011062962962962964,
+      "loss": 1.1928,
+      "step": 24130
+    },
+    {
+      "epoch": 89.41,
+      "grad_norm": 0.8170095682144165,
+      "learning_rate": 0.001105925925925926,
+      "loss": 1.1968,
+      "step": 24140
+    },
+    {
+      "epoch": 89.44,
+      "grad_norm": 0.8027873635292053,
+      "learning_rate": 0.0011055555555555556,
+      "loss": 1.1993,
+      "step": 24150
+    },
+    {
+      "epoch": 89.48,
+      "grad_norm": 0.8569116592407227,
+      "learning_rate": 0.0011051851851851854,
+      "loss": 1.2012,
+      "step": 24160
+    },
+    {
+      "epoch": 89.52,
+      "grad_norm": 0.8904064893722534,
+      "learning_rate": 0.001104814814814815,
+      "loss": 1.2071,
+      "step": 24170
+    },
+    {
+      "epoch": 89.56,
+      "grad_norm": 0.9111688733100891,
+      "learning_rate": 0.0011044444444444444,
+      "loss": 1.2494,
+      "step": 24180
+    },
+    {
+      "epoch": 89.59,
+      "grad_norm": 0.8626366853713989,
+      "learning_rate": 0.001104074074074074,
+      "loss": 1.2422,
+      "step": 24190
+    },
+    {
+      "epoch": 89.63,
+      "grad_norm": 0.8405423164367676,
+      "learning_rate": 0.0011037037037037037,
+      "loss": 1.2314,
+      "step": 24200
+    },
+    {
+      "epoch": 89.67,
+      "grad_norm": 0.9029109477996826,
+      "learning_rate": 0.0011033333333333333,
+      "loss": 1.2691,
+      "step": 24210
+    },
+    {
+      "epoch": 89.7,
+      "grad_norm": 0.8996674418449402,
+      "learning_rate": 0.001102962962962963,
+      "loss": 1.2561,
+      "step": 24220
+    },
+    {
+      "epoch": 89.74,
+      "grad_norm": 0.9102544784545898,
+      "learning_rate": 0.0011025925925925925,
+      "loss": 1.26,
+      "step": 24230
+    },
+    {
+      "epoch": 89.78,
+      "grad_norm": 0.9108865857124329,
+      "learning_rate": 0.0011022222222222223,
+      "loss": 1.275,
+      "step": 24240
+    },
+    {
+      "epoch": 89.81,
+      "grad_norm": 0.8576236963272095,
+      "learning_rate": 0.001101851851851852,
+      "loss": 1.2678,
+      "step": 24250
+    },
+    {
+      "epoch": 89.85,
+      "grad_norm": 0.8501005172729492,
+      "learning_rate": 0.0011014814814814815,
+      "loss": 1.2909,
+      "step": 24260
+    },
+    {
+      "epoch": 89.89,
+      "grad_norm": 0.8619106411933899,
+      "learning_rate": 0.001101111111111111,
+      "loss": 1.2885,
+      "step": 24270
+    },
+    {
+      "epoch": 89.93,
+      "grad_norm": 0.8372892141342163,
+      "learning_rate": 0.0011007407407407409,
+      "loss": 1.3103,
+      "step": 24280
+    },
+    {
+      "epoch": 89.96,
+      "grad_norm": 0.860112190246582,
+      "learning_rate": 0.0011003703703703705,
+      "loss": 1.2871,
+      "step": 24290
+    },
+    {
+      "epoch": 90.0,
+      "grad_norm": 1.5859699249267578,
+      "learning_rate": 0.0011,
+      "loss": 1.3098,
+      "step": 24300
+    },
+    {
+      "epoch": 90.04,
+      "grad_norm": 0.8784357309341431,
+      "learning_rate": 0.0010996296296296297,
+      "loss": 1.0776,
+      "step": 24310
+    },
+    {
+      "epoch": 90.07,
+      "grad_norm": 0.8911356329917908,
+      "learning_rate": 0.0010992592592592592,
+      "loss": 1.1031,
+      "step": 24320
+    },
+    {
+      "epoch": 90.11,
+      "grad_norm": 0.7874282002449036,
+      "learning_rate": 0.0010988888888888888,
+      "loss": 1.1044,
+      "step": 24330
+    },
+    {
+      "epoch": 90.15,
+      "grad_norm": 0.8270366787910461,
+      "learning_rate": 0.0010985185185185184,
+      "loss": 1.0937,
+      "step": 24340
+    },
+    {
+      "epoch": 90.19,
+      "grad_norm": 0.8908728361129761,
+      "learning_rate": 0.0010981481481481482,
+      "loss": 1.1221,
+      "step": 24350
+    },
+    {
+      "epoch": 90.22,
+      "grad_norm": 0.9187735915184021,
+      "learning_rate": 0.0010977777777777778,
+      "loss": 1.141,
+      "step": 24360
+    },
+    {
+      "epoch": 90.26,
+      "grad_norm": 0.9003711342811584,
+      "learning_rate": 0.0010974074074074074,
+      "loss": 1.1368,
+      "step": 24370
+    },
+    {
+      "epoch": 90.3,
+      "grad_norm": 0.8380781412124634,
+      "learning_rate": 0.001097037037037037,
+      "loss": 1.149,
+      "step": 24380
+    },
+    {
+      "epoch": 90.33,
+      "grad_norm": 0.8692864775657654,
+      "learning_rate": 0.0010966666666666668,
+      "loss": 1.1744,
+      "step": 24390
+    },
+    {
+      "epoch": 90.37,
+      "grad_norm": 0.8086166977882385,
+      "learning_rate": 0.0010962962962962964,
+      "loss": 1.1666,
+      "step": 24400
+    },
+    {
+      "epoch": 90.41,
+      "grad_norm": 0.9610259532928467,
+      "learning_rate": 0.001095925925925926,
+      "loss": 1.1919,
+      "step": 24410
+    },
+    {
+      "epoch": 90.44,
+      "grad_norm": 0.9051281809806824,
+      "learning_rate": 0.0010955555555555556,
+      "loss": 1.206,
+      "step": 24420
+    },
+    {
+      "epoch": 90.48,
+      "grad_norm": 0.8761212825775146,
+      "learning_rate": 0.0010951851851851854,
+      "loss": 1.1962,
+      "step": 24430
+    },
+    {
+      "epoch": 90.52,
+      "grad_norm": 0.8568070530891418,
+      "learning_rate": 0.001094814814814815,
+      "loss": 1.2054,
+      "step": 24440
+    },
+    {
+      "epoch": 90.56,
+      "grad_norm": 0.8986058235168457,
+      "learning_rate": 0.0010944444444444445,
+      "loss": 1.1984,
+      "step": 24450
+    },
+    {
+      "epoch": 90.59,
+      "grad_norm": 0.8418546319007874,
+      "learning_rate": 0.001094074074074074,
+      "loss": 1.2403,
+      "step": 24460
+    },
+    {
+      "epoch": 90.63,
+      "grad_norm": 0.8225176930427551,
+      "learning_rate": 0.0010937037037037037,
+      "loss": 1.2449,
+      "step": 24470
+    },
+    {
+      "epoch": 90.67,
+      "grad_norm": 0.9044831991195679,
+      "learning_rate": 0.0010933333333333333,
+      "loss": 1.2398,
+      "step": 24480
+    },
+    {
+      "epoch": 90.7,
+      "grad_norm": 0.9028358459472656,
+      "learning_rate": 0.001092962962962963,
+      "loss": 1.2816,
+      "step": 24490
+    },
+    {
+      "epoch": 90.74,
+      "grad_norm": 0.8567687273025513,
+      "learning_rate": 0.0010925925925925925,
+      "loss": 1.244,
+      "step": 24500
+    },
+    {
+      "epoch": 90.78,
+      "grad_norm": 0.8195727467536926,
+      "learning_rate": 0.0010922222222222223,
+      "loss": 1.2528,
+      "step": 24510
+    },
+    {
+      "epoch": 90.81,
+      "grad_norm": 0.9707487225532532,
+      "learning_rate": 0.0010918518518518519,
+      "loss": 1.2705,
+      "step": 24520
+    },
+    {
+      "epoch": 90.85,
+      "grad_norm": 0.894263505935669,
+      "learning_rate": 0.0010914814814814815,
+      "loss": 1.2607,
+      "step": 24530
+    },
+    {
+      "epoch": 90.89,
+      "grad_norm": 0.8730370402336121,
+      "learning_rate": 0.001091111111111111,
+      "loss": 1.2851,
+      "step": 24540
+    },
+    {
+      "epoch": 90.93,
+      "grad_norm": 0.8939180374145508,
+      "learning_rate": 0.0010907407407407409,
+      "loss": 1.2997,
+      "step": 24550
+    },
+    {
+      "epoch": 90.96,
+      "grad_norm": 0.8519564867019653,
+      "learning_rate": 0.0010903703703703705,
+      "loss": 1.296,
+      "step": 24560
+    },
+    {
+      "epoch": 91.0,
+      "grad_norm": 1.6609879732131958,
+      "learning_rate": 0.00109,
+      "loss": 1.3057,
+      "step": 24570
+    },
+    {
+      "epoch": 91.04,
+      "grad_norm": 0.8076825141906738,
+      "learning_rate": 0.0010896296296296298,
+      "loss": 1.0765,
+      "step": 24580
+    },
+    {
+      "epoch": 91.07,
+      "grad_norm": 0.844336748123169,
+      "learning_rate": 0.0010892592592592592,
+      "loss": 1.0612,
+      "step": 24590
+    },
+    {
+      "epoch": 91.11,
+      "grad_norm": 0.8776166439056396,
+      "learning_rate": 0.0010888888888888888,
+      "loss": 1.0979,
+      "step": 24600
+    },
+    {
+      "epoch": 91.15,
+      "grad_norm": 0.8715381622314453,
+      "learning_rate": 0.0010885185185185184,
+      "loss": 1.1177,
+      "step": 24610
+    },
+    {
+      "epoch": 91.19,
+      "grad_norm": 0.8251622319221497,
+      "learning_rate": 0.0010881481481481482,
+      "loss": 1.1033,
+      "step": 24620
+    },
+    {
+      "epoch": 91.22,
+      "grad_norm": 0.8432281613349915,
+      "learning_rate": 0.0010877777777777778,
+      "loss": 1.1156,
+      "step": 24630
+    },
+    {
+      "epoch": 91.26,
+      "grad_norm": 0.8701592683792114,
+      "learning_rate": 0.0010874074074074074,
+      "loss": 1.1421,
+      "step": 24640
+    },
+    {
+      "epoch": 91.3,
+      "grad_norm": 0.9401283860206604,
+      "learning_rate": 0.001087037037037037,
+      "loss": 1.1513,
+      "step": 24650
+    },
+    {
+      "epoch": 91.33,
+      "grad_norm": 0.8380709290504456,
+      "learning_rate": 0.0010866666666666668,
+      "loss": 1.1442,
+      "step": 24660
+    },
+    {
+      "epoch": 91.37,
+      "grad_norm": 0.8541663289070129,
+      "learning_rate": 0.0010862962962962964,
+      "loss": 1.166,
+      "step": 24670
+    },
+    {
+      "epoch": 91.41,
+      "grad_norm": 0.869760274887085,
+      "learning_rate": 0.001085925925925926,
+      "loss": 1.1895,
+      "step": 24680
+    },
+    {
+      "epoch": 91.44,
+      "grad_norm": 0.8138124346733093,
+      "learning_rate": 0.0010855555555555555,
+      "loss": 1.2083,
+      "step": 24690
+    },
+    {
+      "epoch": 91.48,
+      "grad_norm": 0.9111208915710449,
+      "learning_rate": 0.0010851851851851853,
+      "loss": 1.2269,
+      "step": 24700
+    },
+    {
+      "epoch": 91.52,
+      "grad_norm": 0.8547767996788025,
+      "learning_rate": 0.001084814814814815,
+      "loss": 1.1998,
+      "step": 24710
+    },
+    {
+      "epoch": 91.56,
+      "grad_norm": 0.8412507772445679,
+      "learning_rate": 0.0010844444444444445,
+      "loss": 1.1929,
+      "step": 24720
+    },
+    {
+      "epoch": 91.59,
+      "grad_norm": 0.8674251437187195,
+      "learning_rate": 0.0010840740740740739,
+      "loss": 1.2064,
+      "step": 24730
+    },
+    {
+      "epoch": 91.63,
+      "grad_norm": 0.9130268692970276,
+      "learning_rate": 0.0010837037037037037,
+      "loss": 1.2473,
+      "step": 24740
+    },
+    {
+      "epoch": 91.67,
+      "grad_norm": 0.9256370067596436,
+      "learning_rate": 0.0010833333333333333,
+      "loss": 1.2374,
+      "step": 24750
+    },
+    {
+      "epoch": 91.7,
+      "grad_norm": 0.8963067531585693,
+      "learning_rate": 0.0010829629629629629,
+      "loss": 1.2553,
+      "step": 24760
+    },
+    {
+      "epoch": 91.74,
+      "grad_norm": 0.8186632394790649,
+      "learning_rate": 0.0010825925925925927,
+      "loss": 1.2617,
+      "step": 24770
+    },
+    {
+      "epoch": 91.78,
+      "grad_norm": 0.8365766406059265,
+      "learning_rate": 0.0010822222222222223,
+      "loss": 1.2506,
+      "step": 24780
+    },
+    {
+      "epoch": 91.81,
+      "grad_norm": 0.8591589331626892,
+      "learning_rate": 0.0010818518518518519,
+      "loss": 1.2742,
+      "step": 24790
+    },
+    {
+      "epoch": 91.85,
+      "grad_norm": 0.8758793473243713,
+      "learning_rate": 0.0010814814814814814,
+      "loss": 1.2634,
+      "step": 24800
+    },
+    {
+      "epoch": 91.89,
+      "grad_norm": 0.8638857007026672,
+      "learning_rate": 0.0010811111111111112,
+      "loss": 1.2615,
+      "step": 24810
+    },
+    {
+      "epoch": 91.93,
+      "grad_norm": 0.8489501476287842,
+      "learning_rate": 0.0010807407407407408,
+      "loss": 1.2734,
+      "step": 24820
+    },
+    {
+      "epoch": 91.96,
+      "grad_norm": 0.8600893616676331,
+      "learning_rate": 0.0010803703703703704,
+      "loss": 1.29,
+      "step": 24830
+    },
+    {
+      "epoch": 92.0,
+      "grad_norm": 1.6491703987121582,
+      "learning_rate": 0.00108,
+      "loss": 1.2892,
+      "step": 24840
+    },
+    {
+      "epoch": 92.04,
+      "grad_norm": 0.8379200100898743,
+      "learning_rate": 0.0010796296296296298,
+      "loss": 1.048,
+      "step": 24850
+    },
+    {
+      "epoch": 92.07,
+      "grad_norm": 0.9475078582763672,
+      "learning_rate": 0.0010792592592592594,
+      "loss": 1.0562,
+      "step": 24860
+    },
+    {
+      "epoch": 92.11,
+      "grad_norm": 0.9267643094062805,
+      "learning_rate": 0.0010788888888888888,
+      "loss": 1.0861,
+      "step": 24870
+    },
+    {
+      "epoch": 92.15,
+      "grad_norm": 0.8084042072296143,
+      "learning_rate": 0.0010785185185185184,
+      "loss": 1.0767,
+      "step": 24880
+    },
+    {
+      "epoch": 92.19,
+      "grad_norm": 0.8792321681976318,
+      "learning_rate": 0.0010781481481481482,
+      "loss": 1.1191,
+      "step": 24890
+    },
+    {
+      "epoch": 92.22,
+      "grad_norm": 0.8480071425437927,
+      "learning_rate": 0.0010777777777777778,
+      "loss": 1.1152,
+      "step": 24900
+    },
+    {
+      "epoch": 92.26,
+      "grad_norm": 0.8491085767745972,
+      "learning_rate": 0.0010774074074074074,
+      "loss": 1.1557,
+      "step": 24910
+    },
+    {
+      "epoch": 92.3,
+      "grad_norm": 0.9552561640739441,
+      "learning_rate": 0.001077037037037037,
+      "loss": 1.131,
+      "step": 24920
+    },
+    {
+      "epoch": 92.33,
+      "grad_norm": 0.9046753644943237,
+      "learning_rate": 0.0010766666666666667,
+      "loss": 1.1554,
+      "step": 24930
+    },
+    {
+      "epoch": 92.37,
+      "grad_norm": 0.9224265217781067,
+      "learning_rate": 0.0010762962962962963,
+      "loss": 1.1562,
+      "step": 24940
+    },
+    {
+      "epoch": 92.41,
+      "grad_norm": 0.8900438547134399,
+      "learning_rate": 0.001075925925925926,
+      "loss": 1.1793,
+      "step": 24950
+    },
+    {
+      "epoch": 92.44,
+      "grad_norm": 0.8760910034179688,
+      "learning_rate": 0.0010755555555555557,
+      "loss": 1.1719,
+      "step": 24960
+    },
+    {
+      "epoch": 92.48,
+      "grad_norm": 0.9153395891189575,
+      "learning_rate": 0.0010751851851851853,
+      "loss": 1.185,
+      "step": 24970
+    },
+    {
+      "epoch": 92.52,
+      "grad_norm": 0.854253888130188,
+      "learning_rate": 0.001074814814814815,
+      "loss": 1.197,
+      "step": 24980
+    },
+    {
+      "epoch": 92.56,
+      "grad_norm": 0.8273108601570129,
+      "learning_rate": 0.0010744444444444445,
+      "loss": 1.201,
+      "step": 24990
+    },
+    {
+      "epoch": 92.59,
+      "grad_norm": 0.8812609314918518,
+      "learning_rate": 0.0010740740740740743,
+      "loss": 1.2097,
+      "step": 25000
+    },
+    {
+      "epoch": 92.63,
+      "grad_norm": 0.8414286971092224,
+      "learning_rate": 0.0010737037037037037,
+      "loss": 1.2252,
+      "step": 25010
+    },
+    {
+      "epoch": 92.67,
+      "grad_norm": 0.8828840851783752,
+      "learning_rate": 0.0010733333333333333,
+      "loss": 1.224,
+      "step": 25020
+    },
+    {
+      "epoch": 92.7,
+      "grad_norm": 0.8736853003501892,
+      "learning_rate": 0.0010729629629629628,
+      "loss": 1.2576,
+      "step": 25030
+    },
+    {
+      "epoch": 92.74,
+      "grad_norm": 0.8438439965248108,
+      "learning_rate": 0.0010725925925925927,
+      "loss": 1.2458,
+      "step": 25040
+    },
+    {
+      "epoch": 92.78,
+      "grad_norm": 0.8953974843025208,
+      "learning_rate": 0.0010722222222222222,
+      "loss": 1.2583,
+      "step": 25050
+    },
+    {
+      "epoch": 92.81,
+      "grad_norm": 0.8555558323860168,
+      "learning_rate": 0.0010718518518518518,
+      "loss": 1.2383,
+      "step": 25060
+    },
+    {
+      "epoch": 92.85,
+      "grad_norm": 0.8639121651649475,
+      "learning_rate": 0.0010714814814814814,
+      "loss": 1.264,
+      "step": 25070
+    },
+    {
+      "epoch": 92.89,
+      "grad_norm": 0.8091748952865601,
+      "learning_rate": 0.0010711111111111112,
+      "loss": 1.2607,
+      "step": 25080
+    },
+    {
+      "epoch": 92.93,
+      "grad_norm": 0.8656150102615356,
+      "learning_rate": 0.0010707407407407408,
+      "loss": 1.2705,
+      "step": 25090
+    },
+    {
+      "epoch": 92.96,
+      "grad_norm": 0.8873560428619385,
+      "learning_rate": 0.0010703703703703704,
+      "loss": 1.3017,
+      "step": 25100
+    },
+    {
+      "epoch": 93.0,
+      "grad_norm": 1.5208163261413574,
+      "learning_rate": 0.00107,
+      "loss": 1.2864,
+      "step": 25110
+    },
+    {
+      "epoch": 93.04,
+      "grad_norm": 0.8099291324615479,
+      "learning_rate": 0.0010696296296296298,
+      "loss": 1.0601,
+      "step": 25120
+    },
+    {
+      "epoch": 93.07,
+      "grad_norm": 0.8671590685844421,
+      "learning_rate": 0.0010692592592592594,
+      "loss": 1.0655,
+      "step": 25130
+    },
+    {
+      "epoch": 93.11,
+      "grad_norm": 0.8976333737373352,
+      "learning_rate": 0.001068888888888889,
+      "loss": 1.0471,
+      "step": 25140
+    },
+    {
+      "epoch": 93.15,
+      "grad_norm": 0.8625257015228271,
+      "learning_rate": 0.0010685185185185183,
+      "loss": 1.0606,
+      "step": 25150
+    },
+    {
+      "epoch": 93.19,
+      "grad_norm": 0.8490420579910278,
+      "learning_rate": 0.0010681481481481481,
+      "loss": 1.0883,
+      "step": 25160
+    },
+    {
+      "epoch": 93.22,
+      "grad_norm": 0.9137793779373169,
+      "learning_rate": 0.0010677777777777777,
+      "loss": 1.1154,
+      "step": 25170
+    },
+    {
+      "epoch": 93.26,
+      "grad_norm": 0.8413072228431702,
+      "learning_rate": 0.0010674074074074073,
+      "loss": 1.1194,
+      "step": 25180
+    },
+    {
+      "epoch": 93.3,
+      "grad_norm": 0.8381540179252625,
+      "learning_rate": 0.0010670370370370371,
+      "loss": 1.1144,
+      "step": 25190
+    },
+    {
+      "epoch": 93.33,
+      "grad_norm": 0.8788276314735413,
+      "learning_rate": 0.0010666666666666667,
+      "loss": 1.1448,
+      "step": 25200
+    },
+    {
+      "epoch": 93.37,
+      "grad_norm": 0.8643991947174072,
+      "learning_rate": 0.0010662962962962963,
+      "loss": 1.136,
+      "step": 25210
+    },
+    {
+      "epoch": 93.41,
+      "grad_norm": 0.8980141878128052,
+      "learning_rate": 0.001065925925925926,
+      "loss": 1.1751,
+      "step": 25220
+    },
+    {
+      "epoch": 93.44,
+      "grad_norm": 0.9232836365699768,
+      "learning_rate": 0.0010655555555555557,
+      "loss": 1.1715,
+      "step": 25230
+    },
+    {
+      "epoch": 93.48,
+      "grad_norm": 0.9864527583122253,
+      "learning_rate": 0.0010651851851851853,
+      "loss": 1.17,
+      "step": 25240
+    },
+    {
+      "epoch": 93.52,
+      "grad_norm": 0.9622303247451782,
+      "learning_rate": 0.0010648148148148149,
+      "loss": 1.2191,
+      "step": 25250
+    },
+    {
+      "epoch": 93.56,
+      "grad_norm": 0.9083154797554016,
+      "learning_rate": 0.0010644444444444445,
+      "loss": 1.2056,
+      "step": 25260
+    },
+    {
+      "epoch": 93.59,
+      "grad_norm": 0.8920983076095581,
+      "learning_rate": 0.0010640740740740743,
+      "loss": 1.2426,
+      "step": 25270
+    },
+    {
+      "epoch": 93.63,
+      "grad_norm": 0.8751955628395081,
+      "learning_rate": 0.0010637037037037036,
+      "loss": 1.2283,
+      "step": 25280
+    },
+    {
+      "epoch": 93.67,
+      "grad_norm": 0.9048447608947754,
+      "learning_rate": 0.0010633333333333332,
+      "loss": 1.2235,
+      "step": 25290
+    },
+    {
+      "epoch": 93.7,
+      "grad_norm": 0.8830949068069458,
+      "learning_rate": 0.0010629629629629628,
+      "loss": 1.237,
+      "step": 25300
+    },
+    {
+      "epoch": 93.74,
+      "grad_norm": 0.9014407992362976,
+      "learning_rate": 0.0010625925925925926,
+      "loss": 1.2593,
+      "step": 25310
+    },
+    {
+      "epoch": 93.78,
+      "grad_norm": 0.8940639495849609,
+      "learning_rate": 0.0010622222222222222,
+      "loss": 1.2289,
+      "step": 25320
+    },
+    {
+      "epoch": 93.81,
+      "grad_norm": 0.8975448608398438,
+      "learning_rate": 0.0010618518518518518,
+      "loss": 1.2498,
+      "step": 25330
+    },
+    {
+      "epoch": 93.85,
+      "grad_norm": 0.8936808705329895,
+      "learning_rate": 0.0010614814814814814,
+      "loss": 1.2903,
+      "step": 25340
+    },
+    {
+      "epoch": 93.89,
+      "grad_norm": 0.9473136067390442,
+      "learning_rate": 0.0010611111111111112,
+      "loss": 1.2497,
+      "step": 25350
+    },
+    {
+      "epoch": 93.93,
+      "grad_norm": 0.892784595489502,
+      "learning_rate": 0.0010607407407407408,
+      "loss": 1.2611,
+      "step": 25360
+    },
+    {
+      "epoch": 93.96,
+      "grad_norm": 0.8802827000617981,
+      "learning_rate": 0.0010603703703703704,
+      "loss": 1.2424,
+      "step": 25370
+    },
+    {
+      "epoch": 94.0,
+      "grad_norm": 1.943390130996704,
+      "learning_rate": 0.0010600000000000002,
+      "loss": 1.2733,
+      "step": 25380
+    },
+    {
+      "epoch": 94.04,
+      "grad_norm": 0.859840989112854,
+      "learning_rate": 0.0010596296296296298,
+      "loss": 1.0207,
+      "step": 25390
+    },
+    {
+      "epoch": 94.07,
+      "grad_norm": 0.8180606961250305,
+      "learning_rate": 0.0010592592592592594,
+      "loss": 1.0581,
+      "step": 25400
+    },
+    {
+      "epoch": 94.11,
+      "grad_norm": 0.8518406748771667,
+      "learning_rate": 0.001058888888888889,
+      "loss": 1.095,
+      "step": 25410
+    },
+    {
+      "epoch": 94.15,
+      "grad_norm": 0.8576958179473877,
+      "learning_rate": 0.0010585185185185185,
+      "loss": 1.0672,
+      "step": 25420
+    },
+    {
+      "epoch": 94.19,
+      "grad_norm": 0.8683270812034607,
+      "learning_rate": 0.0010581481481481481,
+      "loss": 1.0929,
+      "step": 25430
+    },
+    {
+      "epoch": 94.22,
+      "grad_norm": 0.8180840611457825,
+      "learning_rate": 0.0010577777777777777,
+      "loss": 1.1359,
+      "step": 25440
+    },
+    {
+      "epoch": 94.26,
+      "grad_norm": 0.8479984402656555,
+      "learning_rate": 0.0010574074074074073,
+      "loss": 1.1059,
+      "step": 25450
+    },
+    {
+      "epoch": 94.3,
+      "grad_norm": 0.8408753871917725,
+      "learning_rate": 0.001057037037037037,
+      "loss": 1.0997,
+      "step": 25460
+    },
+    {
+      "epoch": 94.33,
+      "grad_norm": 0.8361555337905884,
+      "learning_rate": 0.0010566666666666667,
+      "loss": 1.1254,
+      "step": 25470
+    },
+    {
+      "epoch": 94.37,
+      "grad_norm": 0.8891112804412842,
+      "learning_rate": 0.0010562962962962963,
+      "loss": 1.1493,
+      "step": 25480
+    },
+    {
+      "epoch": 94.41,
+      "grad_norm": 0.9582691788673401,
+      "learning_rate": 0.0010559259259259259,
+      "loss": 1.1596,
+      "step": 25490
+    },
+    {
+      "epoch": 94.44,
+      "grad_norm": 0.9022079706192017,
+      "learning_rate": 0.0010555555555555557,
+      "loss": 1.16,
+      "step": 25500
+    },
+    {
+      "epoch": 94.48,
+      "grad_norm": 0.8625105619430542,
+      "learning_rate": 0.0010551851851851853,
+      "loss": 1.1839,
+      "step": 25510
+    },
+    {
+      "epoch": 94.52,
+      "grad_norm": 0.8581701517105103,
+      "learning_rate": 0.0010548148148148149,
+      "loss": 1.1875,
+      "step": 25520
+    },
+    {
+      "epoch": 94.56,
+      "grad_norm": 0.9323828220367432,
+      "learning_rate": 0.0010544444444444444,
+      "loss": 1.2022,
+      "step": 25530
+    },
+    {
+      "epoch": 94.59,
+      "grad_norm": 0.8733723163604736,
+      "learning_rate": 0.0010540740740740742,
+      "loss": 1.2044,
+      "step": 25540
+    },
+    {
+      "epoch": 94.63,
+      "grad_norm": 0.8635187745094299,
+      "learning_rate": 0.0010537037037037038,
+      "loss": 1.1975,
+      "step": 25550
+    },
+    {
+      "epoch": 94.67,
+      "grad_norm": 0.8991508483886719,
+      "learning_rate": 0.0010533333333333332,
+      "loss": 1.1968,
+      "step": 25560
+    },
+    {
+      "epoch": 94.7,
+      "grad_norm": 0.9222977757453918,
+      "learning_rate": 0.0010529629629629628,
+      "loss": 1.2285,
+      "step": 25570
+    },
+    {
+      "epoch": 94.74,
+      "grad_norm": 0.8603134155273438,
+      "learning_rate": 0.0010525925925925926,
+      "loss": 1.2376,
+      "step": 25580
+    },
+    {
+      "epoch": 94.78,
+      "grad_norm": 0.9961068630218506,
+      "learning_rate": 0.0010522222222222222,
+      "loss": 1.2233,
+      "step": 25590
+    },
+    {
+      "epoch": 94.81,
+      "grad_norm": 0.8663851022720337,
+      "learning_rate": 0.0010518518518518518,
+      "loss": 1.2645,
+      "step": 25600
+    },
+    {
+      "epoch": 94.85,
+      "grad_norm": 0.9062612056732178,
+      "learning_rate": 0.0010514814814814816,
+      "loss": 1.2363,
+      "step": 25610
+    },
+    {
+      "epoch": 94.89,
+      "grad_norm": 0.9243030548095703,
+      "learning_rate": 0.0010511111111111112,
+      "loss": 1.2508,
+      "step": 25620
+    },
+    {
+      "epoch": 94.93,
+      "grad_norm": 0.8614787459373474,
+      "learning_rate": 0.0010507407407407408,
+      "loss": 1.2322,
+      "step": 25630
+    },
+    {
+      "epoch": 94.96,
+      "grad_norm": 0.9229607582092285,
+      "learning_rate": 0.0010503703703703703,
+      "loss": 1.2798,
+      "step": 25640
+    },
+    {
+      "epoch": 95.0,
+      "grad_norm": 1.8464992046356201,
+      "learning_rate": 0.0010500000000000002,
+      "loss": 1.2714,
+      "step": 25650
+    },
+    {
+      "epoch": 95.04,
+      "grad_norm": 0.7462485432624817,
+      "learning_rate": 0.0010496296296296297,
+      "loss": 1.0297,
+      "step": 25660
+    },
+    {
+      "epoch": 95.07,
+      "grad_norm": 0.8666638731956482,
+      "learning_rate": 0.0010492592592592593,
+      "loss": 1.0276,
+      "step": 25670
+    },
+    {
+      "epoch": 95.11,
+      "grad_norm": 0.9033090472221375,
+      "learning_rate": 0.001048888888888889,
+      "loss": 1.061,
+      "step": 25680
+    },
+    {
+      "epoch": 95.15,
+      "grad_norm": 0.8938956260681152,
+      "learning_rate": 0.0010485185185185187,
+      "loss": 1.077,
+      "step": 25690
+    },
+    {
+      "epoch": 95.19,
+      "grad_norm": 0.9139565229415894,
+      "learning_rate": 0.001048148148148148,
+      "loss": 1.0753,
+      "step": 25700
+    },
+    {
+      "epoch": 95.22,
+      "grad_norm": 0.8628860712051392,
+      "learning_rate": 0.0010477777777777777,
+      "loss": 1.085,
+      "step": 25710
+    },
+    {
+      "epoch": 95.26,
+      "grad_norm": 0.908026397228241,
+      "learning_rate": 0.0010474074074074073,
+      "loss": 1.1087,
+      "step": 25720
+    },
+    {
+      "epoch": 95.3,
+      "grad_norm": 0.8440184593200684,
+      "learning_rate": 0.001047037037037037,
+      "loss": 1.1133,
+      "step": 25730
+    },
+    {
+      "epoch": 95.33,
+      "grad_norm": 0.9705326557159424,
+      "learning_rate": 0.0010466666666666667,
+      "loss": 1.135,
+      "step": 25740
+    },
+    {
+      "epoch": 95.37,
+      "grad_norm": 0.8961520195007324,
+      "learning_rate": 0.0010462962962962963,
+      "loss": 1.161,
+      "step": 25750
+    },
+    {
+      "epoch": 95.41,
+      "grad_norm": 0.8462703227996826,
+      "learning_rate": 0.0010459259259259258,
+      "loss": 1.1503,
+      "step": 25760
+    },
+    {
+      "epoch": 95.44,
+      "grad_norm": 0.8871561288833618,
+      "learning_rate": 0.0010455555555555556,
+      "loss": 1.1721,
+      "step": 25770
+    },
+    {
+      "epoch": 95.48,
+      "grad_norm": 0.9002634882926941,
+      "learning_rate": 0.0010451851851851852,
+      "loss": 1.169,
+      "step": 25780
+    },
+    {
+      "epoch": 95.52,
+      "grad_norm": 0.9232921600341797,
+      "learning_rate": 0.0010448148148148148,
+      "loss": 1.1717,
+      "step": 25790
+    },
+    {
+      "epoch": 95.56,
+      "grad_norm": 0.8582503795623779,
+      "learning_rate": 0.0010444444444444446,
+      "loss": 1.1827,
+      "step": 25800
+    },
+    {
+      "epoch": 95.59,
+      "grad_norm": 0.8893172144889832,
+      "learning_rate": 0.0010440740740740742,
+      "loss": 1.2161,
+      "step": 25810
+    },
+    {
+      "epoch": 95.63,
+      "grad_norm": 0.9103512167930603,
+      "learning_rate": 0.0010437037037037038,
+      "loss": 1.2138,
+      "step": 25820
+    },
+    {
+      "epoch": 95.67,
+      "grad_norm": 0.8459279537200928,
+      "learning_rate": 0.0010433333333333332,
+      "loss": 1.2184,
+      "step": 25830
+    },
+    {
+      "epoch": 95.7,
+      "grad_norm": 0.9495593309402466,
+      "learning_rate": 0.001042962962962963,
+      "loss": 1.1986,
+      "step": 25840
+    },
+    {
+      "epoch": 95.74,
+      "grad_norm": 0.8854154348373413,
+      "learning_rate": 0.0010425925925925926,
+      "loss": 1.2264,
+      "step": 25850
+    },
+    {
+      "epoch": 95.78,
+      "grad_norm": 0.9533488154411316,
+      "learning_rate": 0.0010422222222222222,
+      "loss": 1.2086,
+      "step": 25860
+    },
+    {
+      "epoch": 95.81,
+      "grad_norm": 0.9248517751693726,
+      "learning_rate": 0.0010418518518518517,
+      "loss": 1.2536,
+      "step": 25870
+    },
+    {
+      "epoch": 95.85,
+      "grad_norm": 0.8576090931892395,
+      "learning_rate": 0.0010414814814814816,
+      "loss": 1.2604,
+      "step": 25880
+    },
+    {
+      "epoch": 95.89,
+      "grad_norm": 0.89030522108078,
+      "learning_rate": 0.0010411111111111111,
+      "loss": 1.2545,
+      "step": 25890
+    },
+    {
+      "epoch": 95.93,
+      "grad_norm": 0.9573636651039124,
+      "learning_rate": 0.0010407407407407407,
+      "loss": 1.2565,
+      "step": 25900
+    },
+    {
+      "epoch": 95.96,
+      "grad_norm": 0.8615505695343018,
+      "learning_rate": 0.0010403703703703703,
+      "loss": 1.2782,
+      "step": 25910
+    },
+    {
+      "epoch": 96.0,
+      "grad_norm": 1.4731565713882446,
+      "learning_rate": 0.0010400000000000001,
+      "loss": 1.2516,
+      "step": 25920
+    },
+    {
+      "epoch": 96.04,
+      "grad_norm": 0.8369557857513428,
+      "learning_rate": 0.0010396296296296297,
+      "loss": 1.0401,
+      "step": 25930
+    },
+    {
+      "epoch": 96.07,
+      "grad_norm": 0.9119203090667725,
+      "learning_rate": 0.0010392592592592593,
+      "loss": 1.0128,
+      "step": 25940
+    },
+    {
+      "epoch": 96.11,
+      "grad_norm": 0.8330380320549011,
+      "learning_rate": 0.0010388888888888889,
+      "loss": 1.0457,
+      "step": 25950
+    },
+    {
+      "epoch": 96.15,
+      "grad_norm": 0.901683509349823,
+      "learning_rate": 0.0010385185185185187,
+      "loss": 1.0575,
+      "step": 25960
+    },
+    {
+      "epoch": 96.19,
+      "grad_norm": 0.8877632021903992,
+      "learning_rate": 0.001038148148148148,
+      "loss": 1.0487,
+      "step": 25970
+    },
+    {
+      "epoch": 96.22,
+      "grad_norm": 0.9112820625305176,
+      "learning_rate": 0.0010377777777777777,
+      "loss": 1.0781,
+      "step": 25980
+    },
+    {
+      "epoch": 96.26,
+      "grad_norm": 0.9182432889938354,
+      "learning_rate": 0.0010374074074074075,
+      "loss": 1.1201,
+      "step": 25990
+    },
+    {
+      "epoch": 96.3,
+      "grad_norm": 0.9750530123710632,
+      "learning_rate": 0.001037037037037037,
+      "loss": 1.1322,
+      "step": 26000
+    },
+    {
+      "epoch": 96.33,
+      "grad_norm": 0.9546353220939636,
+      "learning_rate": 0.0010366666666666666,
+      "loss": 1.124,
+      "step": 26010
+    },
+    {
+      "epoch": 96.37,
+      "grad_norm": 0.8506656885147095,
+      "learning_rate": 0.0010362962962962962,
+      "loss": 1.1454,
+      "step": 26020
+    },
+    {
+      "epoch": 96.41,
+      "grad_norm": 0.8744280338287354,
+      "learning_rate": 0.001035925925925926,
+      "loss": 1.1397,
+      "step": 26030
+    },
+    {
+      "epoch": 96.44,
+      "grad_norm": 0.8825117349624634,
+      "learning_rate": 0.0010355555555555556,
+      "loss": 1.1655,
+      "step": 26040
+    },
+    {
+      "epoch": 96.48,
+      "grad_norm": 0.9138257503509521,
+      "learning_rate": 0.0010351851851851852,
+      "loss": 1.1494,
+      "step": 26050
+    },
+    {
+      "epoch": 96.52,
+      "grad_norm": 0.91966313123703,
+      "learning_rate": 0.0010348148148148148,
+      "loss": 1.1891,
+      "step": 26060
+    },
+    {
+      "epoch": 96.56,
+      "grad_norm": 0.9052321314811707,
+      "learning_rate": 0.0010344444444444446,
+      "loss": 1.1694,
+      "step": 26070
+    },
+    {
+      "epoch": 96.59,
+      "grad_norm": 0.87230384349823,
+      "learning_rate": 0.0010340740740740742,
+      "loss": 1.1635,
+      "step": 26080
+    },
+    {
+      "epoch": 96.63,
+      "grad_norm": 0.8561148643493652,
+      "learning_rate": 0.0010337037037037038,
+      "loss": 1.1816,
+      "step": 26090
+    },
+    {
+      "epoch": 96.67,
+      "grad_norm": 0.9230003952980042,
+      "learning_rate": 0.0010333333333333334,
+      "loss": 1.1944,
+      "step": 26100
+    },
+    {
+      "epoch": 96.7,
+      "grad_norm": 0.9017607569694519,
+      "learning_rate": 0.001032962962962963,
+      "loss": 1.1991,
+      "step": 26110
+    },
+    {
+      "epoch": 96.74,
+      "grad_norm": 0.9002452492713928,
+      "learning_rate": 0.0010325925925925925,
+      "loss": 1.2403,
+      "step": 26120
+    },
+    {
+      "epoch": 96.78,
+      "grad_norm": 0.9941157102584839,
+      "learning_rate": 0.0010322222222222221,
+      "loss": 1.2229,
+      "step": 26130
+    },
+    {
+      "epoch": 96.81,
+      "grad_norm": 0.9742600917816162,
+      "learning_rate": 0.0010318518518518517,
+      "loss": 1.2228,
+      "step": 26140
+    },
+    {
+      "epoch": 96.85,
+      "grad_norm": 0.9086616635322571,
+      "learning_rate": 0.0010314814814814815,
+      "loss": 1.2287,
+      "step": 26150
+    },
+    {
+      "epoch": 96.89,
+      "grad_norm": 0.9041900634765625,
+      "learning_rate": 0.0010311111111111111,
+      "loss": 1.2324,
+      "step": 26160
+    },
+    {
+      "epoch": 96.93,
+      "grad_norm": 0.9248058199882507,
+      "learning_rate": 0.0010307407407407407,
+      "loss": 1.269,
+      "step": 26170
+    },
+    {
+      "epoch": 96.96,
+      "grad_norm": 0.928729772567749,
+      "learning_rate": 0.0010303703703703703,
+      "loss": 1.2477,
+      "step": 26180
+    },
+    {
+      "epoch": 97.0,
+      "grad_norm": 1.5879347324371338,
+      "learning_rate": 0.00103,
+      "loss": 1.2538,
+      "step": 26190
+    },
+    {
+      "epoch": 97.04,
+      "grad_norm": 0.9306336045265198,
+      "learning_rate": 0.0010296296296296297,
+      "loss": 1.0265,
+      "step": 26200
+    },
+    {
+      "epoch": 97.07,
+      "grad_norm": 0.8678556680679321,
+      "learning_rate": 0.0010292592592592593,
+      "loss": 1.0392,
+      "step": 26210
+    },
+    {
+      "epoch": 97.11,
+      "grad_norm": 0.8407970070838928,
+      "learning_rate": 0.001028888888888889,
+      "loss": 1.0319,
+      "step": 26220
+    },
+    {
+      "epoch": 97.15,
+      "grad_norm": 0.9070879817008972,
+      "learning_rate": 0.0010285185185185187,
+      "loss": 1.0491,
+      "step": 26230
+    },
+    {
+      "epoch": 97.19,
+      "grad_norm": 0.9178571701049805,
+      "learning_rate": 0.0010281481481481483,
+      "loss": 1.0906,
+      "step": 26240
+    },
+    {
+      "epoch": 97.22,
+      "grad_norm": 0.8487735986709595,
+      "learning_rate": 0.0010277777777777776,
+      "loss": 1.0685,
+      "step": 26250
+    },
+    {
+      "epoch": 97.26,
+      "grad_norm": 0.9132791757583618,
+      "learning_rate": 0.0010274074074074074,
+      "loss": 1.0635,
+      "step": 26260
+    },
+    {
+      "epoch": 97.3,
+      "grad_norm": 0.838406503200531,
+      "learning_rate": 0.001027037037037037,
+      "loss": 1.111,
+      "step": 26270
+    },
+    {
+      "epoch": 97.33,
+      "grad_norm": 0.9344068169593811,
+      "learning_rate": 0.0010266666666666666,
+      "loss": 1.1066,
+      "step": 26280
+    },
+    {
+      "epoch": 97.37,
+      "grad_norm": 0.9382205605506897,
+      "learning_rate": 0.0010262962962962962,
+      "loss": 1.1355,
+      "step": 26290
+    },
+    {
+      "epoch": 97.41,
+      "grad_norm": 0.873230516910553,
+      "learning_rate": 0.001025925925925926,
+      "loss": 1.1282,
+      "step": 26300
+    },
+    {
+      "epoch": 97.44,
+      "grad_norm": 0.8769469857215881,
+      "learning_rate": 0.0010255555555555556,
+      "loss": 1.1348,
+      "step": 26310
+    },
+    {
+      "epoch": 97.48,
+      "grad_norm": 0.9158292412757874,
+      "learning_rate": 0.0010251851851851852,
+      "loss": 1.162,
+      "step": 26320
+    },
+    {
+      "epoch": 97.52,
+      "grad_norm": 0.9506550431251526,
+      "learning_rate": 0.0010248148148148148,
+      "loss": 1.1757,
+      "step": 26330
+    },
+    {
+      "epoch": 97.56,
+      "grad_norm": 0.8787921071052551,
+      "learning_rate": 0.0010244444444444446,
+      "loss": 1.182,
+      "step": 26340
+    },
+    {
+      "epoch": 97.59,
+      "grad_norm": 0.8608382940292358,
+      "learning_rate": 0.0010240740740740742,
+      "loss": 1.1883,
+      "step": 26350
+    },
+    {
+      "epoch": 97.63,
+      "grad_norm": 0.8862053751945496,
+      "learning_rate": 0.0010237037037037038,
+      "loss": 1.1919,
+      "step": 26360
+    },
+    {
+      "epoch": 97.67,
+      "grad_norm": 0.8187049031257629,
+      "learning_rate": 0.0010233333333333333,
+      "loss": 1.1939,
+      "step": 26370
+    },
+    {
+      "epoch": 97.7,
+      "grad_norm": 0.8783975839614868,
+      "learning_rate": 0.0010229629629629631,
+      "loss": 1.1777,
+      "step": 26380
+    },
+    {
+      "epoch": 97.74,
+      "grad_norm": 0.8841999173164368,
+      "learning_rate": 0.0010225925925925925,
+      "loss": 1.2191,
+      "step": 26390
+    },
+    {
+      "epoch": 97.78,
+      "grad_norm": 0.970224142074585,
+      "learning_rate": 0.0010222222222222221,
+      "loss": 1.2244,
+      "step": 26400
+    },
+    {
+      "epoch": 97.81,
+      "grad_norm": 0.8861240148544312,
+      "learning_rate": 0.001021851851851852,
+      "loss": 1.2127,
+      "step": 26410
+    },
+    {
+      "epoch": 97.85,
+      "grad_norm": 0.9190117716789246,
+      "learning_rate": 0.0010214814814814815,
+      "loss": 1.2305,
+      "step": 26420
+    },
+    {
+      "epoch": 97.89,
+      "grad_norm": 0.8924251794815063,
+      "learning_rate": 0.001021111111111111,
+      "loss": 1.2419,
+      "step": 26430
+    },
+    {
+      "epoch": 97.93,
+      "grad_norm": 0.9394065737724304,
+      "learning_rate": 0.0010207407407407407,
+      "loss": 1.2266,
+      "step": 26440
+    },
+    {
+      "epoch": 97.96,
+      "grad_norm": 0.9103227257728577,
+      "learning_rate": 0.0010203703703703705,
+      "loss": 1.2484,
+      "step": 26450
+    },
+    {
+      "epoch": 98.0,
+      "grad_norm": 1.6072869300842285,
+      "learning_rate": 0.00102,
+      "loss": 1.2523,
+      "step": 26460
+    },
+    {
+      "epoch": 98.04,
+      "grad_norm": 0.8078530430793762,
+      "learning_rate": 0.0010196296296296297,
+      "loss": 1.0086,
+      "step": 26470
+    },
+    {
+      "epoch": 98.07,
+      "grad_norm": 0.9085745811462402,
+      "learning_rate": 0.0010192592592592593,
+      "loss": 1.0258,
+      "step": 26480
+    },
+    {
+      "epoch": 98.11,
+      "grad_norm": 0.8509937524795532,
+      "learning_rate": 0.001018888888888889,
+      "loss": 1.0416,
+      "step": 26490
+    },
+    {
+      "epoch": 98.15,
+      "grad_norm": 0.9164654612541199,
+      "learning_rate": 0.0010185185185185186,
+      "loss": 1.0102,
+      "step": 26500
+    },
+    {
+      "epoch": 98.19,
+      "grad_norm": 0.8645424842834473,
+      "learning_rate": 0.0010181481481481482,
+      "loss": 1.0746,
+      "step": 26510
+    },
+    {
+      "epoch": 98.22,
+      "grad_norm": 0.850975751876831,
+      "learning_rate": 0.0010177777777777776,
+      "loss": 1.0675,
+      "step": 26520
+    },
+    {
+      "epoch": 98.26,
+      "grad_norm": 0.922157883644104,
+      "learning_rate": 0.0010174074074074074,
+      "loss": 1.1008,
+      "step": 26530
+    },
+    {
+      "epoch": 98.3,
+      "grad_norm": 0.8732514381408691,
+      "learning_rate": 0.001017037037037037,
+      "loss": 1.1038,
+      "step": 26540
+    },
+    {
+      "epoch": 98.33,
+      "grad_norm": 0.8597608804702759,
+      "learning_rate": 0.0010166666666666666,
+      "loss": 1.0974,
+      "step": 26550
+    },
+    {
+      "epoch": 98.37,
+      "grad_norm": 0.8999730944633484,
+      "learning_rate": 0.0010162962962962962,
+      "loss": 1.0949,
+      "step": 26560
+    },
+    {
+      "epoch": 98.41,
+      "grad_norm": 0.8790624141693115,
+      "learning_rate": 0.001015925925925926,
+      "loss": 1.1276,
+      "step": 26570
+    },
+    {
+      "epoch": 98.44,
+      "grad_norm": 0.9210389256477356,
+      "learning_rate": 0.0010155555555555556,
+      "loss": 1.1287,
+      "step": 26580
+    },
+    {
+      "epoch": 98.48,
+      "grad_norm": 1.0357301235198975,
+      "learning_rate": 0.0010151851851851852,
+      "loss": 1.1505,
+      "step": 26590
+    },
+    {
+      "epoch": 98.52,
+      "grad_norm": 0.8954038619995117,
+      "learning_rate": 0.001014814814814815,
+      "loss": 1.156,
+      "step": 26600
+    },
+    {
+      "epoch": 98.56,
+      "grad_norm": 0.9502341747283936,
+      "learning_rate": 0.0010144444444444446,
+      "loss": 1.1765,
+      "step": 26610
+    },
+    {
+      "epoch": 98.59,
+      "grad_norm": 0.9129772186279297,
+      "learning_rate": 0.0010140740740740741,
+      "loss": 1.1711,
+      "step": 26620
+    },
+    {
+      "epoch": 98.63,
+      "grad_norm": 0.9075857996940613,
+      "learning_rate": 0.0010137037037037037,
+      "loss": 1.1851,
+      "step": 26630
+    },
+    {
+      "epoch": 98.67,
+      "grad_norm": 0.8680686950683594,
+      "learning_rate": 0.0010133333333333335,
+      "loss": 1.1878,
+      "step": 26640
+    },
+    {
+      "epoch": 98.7,
+      "grad_norm": 0.8635008931159973,
+      "learning_rate": 0.0010129629629629631,
+      "loss": 1.2095,
+      "step": 26650
+    },
+    {
+      "epoch": 98.74,
+      "grad_norm": 0.871293306350708,
+      "learning_rate": 0.0010125925925925925,
+      "loss": 1.1938,
+      "step": 26660
+    },
+    {
+      "epoch": 98.78,
+      "grad_norm": 0.8697920441627502,
+      "learning_rate": 0.001012222222222222,
+      "loss": 1.1918,
+      "step": 26670
+    },
+    {
+      "epoch": 98.81,
+      "grad_norm": 0.9331200122833252,
+      "learning_rate": 0.0010118518518518519,
+      "loss": 1.228,
+      "step": 26680
+    },
+    {
+      "epoch": 98.85,
+      "grad_norm": 0.9641883969306946,
+      "learning_rate": 0.0010114814814814815,
+      "loss": 1.2206,
+      "step": 26690
+    },
+    {
+      "epoch": 98.89,
+      "grad_norm": 0.8545400500297546,
+      "learning_rate": 0.001011111111111111,
+      "loss": 1.2681,
+      "step": 26700
+    },
+    {
+      "epoch": 98.93,
+      "grad_norm": 0.9419219493865967,
+      "learning_rate": 0.0010107407407407407,
+      "loss": 1.218,
+      "step": 26710
+    },
+    {
+      "epoch": 98.96,
+      "grad_norm": 0.9317721724510193,
+      "learning_rate": 0.0010103703703703705,
+      "loss": 1.2097,
+      "step": 26720
+    },
+    {
+      "epoch": 99.0,
+      "grad_norm": 1.9032236337661743,
+      "learning_rate": 0.00101,
+      "loss": 1.2093,
+      "step": 26730
+    },
+    {
+      "epoch": 99.04,
+      "grad_norm": 0.873572051525116,
+      "learning_rate": 0.0010096296296296296,
+      "loss": 1.0002,
+      "step": 26740
+    },
+    {
+      "epoch": 99.07,
+      "grad_norm": 0.8706053495407104,
+      "learning_rate": 0.0010092592592592592,
+      "loss": 0.9963,
+      "step": 26750
+    },
+    {
+      "epoch": 99.11,
+      "grad_norm": 0.881637454032898,
+      "learning_rate": 0.001008888888888889,
+      "loss": 1.0306,
+      "step": 26760
+    },
+    {
+      "epoch": 99.15,
+      "grad_norm": 0.9505958557128906,
+      "learning_rate": 0.0010085185185185186,
+      "loss": 1.0336,
+      "step": 26770
+    },
+    {
+      "epoch": 99.19,
+      "grad_norm": 0.8608532547950745,
+      "learning_rate": 0.0010081481481481482,
+      "loss": 1.0318,
+      "step": 26780
+    },
+    {
+      "epoch": 99.22,
+      "grad_norm": 0.9403489232063293,
+      "learning_rate": 0.001007777777777778,
+      "loss": 1.0933,
+      "step": 26790
+    },
+    {
+      "epoch": 99.26,
+      "grad_norm": 0.9129186868667603,
+      "learning_rate": 0.0010074074074074074,
+      "loss": 1.0923,
+      "step": 26800
+    },
+    {
+      "epoch": 99.3,
+      "grad_norm": 0.9360917210578918,
+      "learning_rate": 0.001007037037037037,
+      "loss": 1.116,
+      "step": 26810
+    },
+    {
+      "epoch": 99.33,
+      "grad_norm": 0.8772351741790771,
+      "learning_rate": 0.0010066666666666666,
+      "loss": 1.1082,
+      "step": 26820
+    },
+    {
+      "epoch": 99.37,
+      "grad_norm": 0.8974486589431763,
+      "learning_rate": 0.0010062962962962964,
+      "loss": 1.1225,
+      "step": 26830
+    },
+    {
+      "epoch": 99.41,
+      "grad_norm": 0.9104517698287964,
+      "learning_rate": 0.001005925925925926,
+      "loss": 1.1213,
+      "step": 26840
+    },
+    {
+      "epoch": 99.44,
+      "grad_norm": 0.9942017793655396,
+      "learning_rate": 0.0010055555555555555,
+      "loss": 1.1117,
+      "step": 26850
+    },
+    {
+      "epoch": 99.48,
+      "grad_norm": 0.8957462906837463,
+      "learning_rate": 0.0010051851851851851,
+      "loss": 1.1415,
+      "step": 26860
+    },
+    {
+      "epoch": 99.52,
+      "grad_norm": 0.9015626907348633,
+      "learning_rate": 0.001004814814814815,
+      "loss": 1.1527,
+      "step": 26870
+    },
+    {
+      "epoch": 99.56,
+      "grad_norm": 0.8942874670028687,
+      "learning_rate": 0.0010044444444444445,
+      "loss": 1.1544,
+      "step": 26880
+    },
+    {
+      "epoch": 99.59,
+      "grad_norm": 0.9468600749969482,
+      "learning_rate": 0.0010040740740740741,
+      "loss": 1.145,
+      "step": 26890
+    },
+    {
+      "epoch": 99.63,
+      "grad_norm": 0.9078618884086609,
+      "learning_rate": 0.0010037037037037037,
+      "loss": 1.1799,
+      "step": 26900
+    },
+    {
+      "epoch": 99.67,
+      "grad_norm": 0.9247041940689087,
+      "learning_rate": 0.0010033333333333335,
+      "loss": 1.1948,
+      "step": 26910
+    },
+    {
+      "epoch": 99.7,
+      "grad_norm": 0.9799883961677551,
+      "learning_rate": 0.001002962962962963,
+      "loss": 1.1786,
+      "step": 26920
+    },
+    {
+      "epoch": 99.74,
+      "grad_norm": 0.8741432428359985,
+      "learning_rate": 0.0010025925925925927,
+      "loss": 1.1948,
+      "step": 26930
+    },
+    {
+      "epoch": 99.78,
+      "grad_norm": 0.9404388666152954,
+      "learning_rate": 0.001002222222222222,
+      "loss": 1.1932,
+      "step": 26940
+    },
+    {
+      "epoch": 99.81,
+      "grad_norm": 0.9425485134124756,
+      "learning_rate": 0.0010018518518518519,
+      "loss": 1.2049,
+      "step": 26950
+    },
+    {
+      "epoch": 99.85,
+      "grad_norm": 0.9867810010910034,
+      "learning_rate": 0.0010014814814814814,
+      "loss": 1.2078,
+      "step": 26960
+    },
+    {
+      "epoch": 99.89,
+      "grad_norm": 0.9349437952041626,
+      "learning_rate": 0.001001111111111111,
+      "loss": 1.2086,
+      "step": 26970
+    },
+    {
+      "epoch": 99.93,
+      "grad_norm": 0.8289681673049927,
+      "learning_rate": 0.0010007407407407406,
+      "loss": 1.2475,
+      "step": 26980
+    },
+    {
+      "epoch": 99.96,
+      "grad_norm": 0.9117506146430969,
+      "learning_rate": 0.0010003703703703704,
+      "loss": 1.2344,
+      "step": 26990
+    },
+    {
+      "epoch": 100.0,
+      "grad_norm": 1.7707023620605469,
+      "learning_rate": 0.001,
+      "loss": 1.2213,
+      "step": 27000
+    },
+    {
+      "epoch": 100.04,
+      "grad_norm": 0.8565974235534668,
+      "learning_rate": 0.0009996296296296296,
+      "loss": 0.9891,
+      "step": 27010
+    },
+    {
+      "epoch": 100.07,
+      "grad_norm": 0.8615112900733948,
+      "learning_rate": 0.0009992592592592594,
+      "loss": 1.0083,
+      "step": 27020
+    },
+    {
+      "epoch": 100.11,
+      "grad_norm": 0.9136719703674316,
+      "learning_rate": 0.000998888888888889,
+      "loss": 0.9975,
+      "step": 27030
+    },
+    {
+      "epoch": 100.15,
+      "grad_norm": 0.8897717595100403,
+      "learning_rate": 0.0009985185185185186,
+      "loss": 1.0088,
+      "step": 27040
+    },
+    {
+      "epoch": 100.19,
+      "grad_norm": 0.8684051036834717,
+      "learning_rate": 0.0009981481481481482,
+      "loss": 1.0412,
+      "step": 27050
+    },
+    {
+      "epoch": 100.22,
+      "grad_norm": 1.019988775253296,
+      "learning_rate": 0.0009977777777777778,
+      "loss": 1.0733,
+      "step": 27060
+    },
+    {
+      "epoch": 100.26,
+      "grad_norm": 0.8850982189178467,
+      "learning_rate": 0.0009974074074074074,
+      "loss": 1.0866,
+      "step": 27070
+    },
+    {
+      "epoch": 100.3,
+      "grad_norm": 0.8889622688293457,
+      "learning_rate": 0.0009970370370370372,
+      "loss": 1.068,
+      "step": 27080
+    },
+    {
+      "epoch": 100.33,
+      "grad_norm": 0.9454230070114136,
+      "learning_rate": 0.0009966666666666668,
+      "loss": 1.0957,
+      "step": 27090
+    },
+    {
+      "epoch": 100.37,
+      "grad_norm": 0.9413827657699585,
+      "learning_rate": 0.0009962962962962963,
+      "loss": 1.1336,
+      "step": 27100
+    },
+    {
+      "epoch": 100.41,
+      "grad_norm": 0.8245477676391602,
+      "learning_rate": 0.000995925925925926,
+      "loss": 1.1296,
+      "step": 27110
+    },
+    {
+      "epoch": 100.44,
+      "grad_norm": 0.9790675640106201,
+      "learning_rate": 0.0009955555555555555,
+      "loss": 1.1404,
+      "step": 27120
+    },
+    {
+      "epoch": 100.48,
+      "grad_norm": 0.9440389275550842,
+      "learning_rate": 0.000995185185185185,
+      "loss": 1.1168,
+      "step": 27130
+    },
+    {
+      "epoch": 100.52,
+      "grad_norm": 0.9235543012619019,
+      "learning_rate": 0.000994814814814815,
+      "loss": 1.1552,
+      "step": 27140
+    },
+    {
+      "epoch": 100.56,
+      "grad_norm": 0.9291632771492004,
+      "learning_rate": 0.0009944444444444445,
+      "loss": 1.156,
+      "step": 27150
+    },
+    {
+      "epoch": 100.59,
+      "grad_norm": 0.912441074848175,
+      "learning_rate": 0.000994074074074074,
+      "loss": 1.1397,
+      "step": 27160
+    },
+    {
+      "epoch": 100.63,
+      "grad_norm": 0.9047914147377014,
+      "learning_rate": 0.0009937037037037037,
+      "loss": 1.1661,
+      "step": 27170
+    },
+    {
+      "epoch": 100.67,
+      "grad_norm": 0.862169623374939,
+      "learning_rate": 0.0009933333333333333,
+      "loss": 1.1737,
+      "step": 27180
+    },
+    {
+      "epoch": 100.7,
+      "grad_norm": 1.00532066822052,
+      "learning_rate": 0.0009929629629629629,
+      "loss": 1.172,
+      "step": 27190
+    },
+    {
+      "epoch": 100.74,
+      "grad_norm": 0.9798426628112793,
+      "learning_rate": 0.0009925925925925927,
+      "loss": 1.1825,
+      "step": 27200
+    },
+    {
+      "epoch": 100.78,
+      "grad_norm": 0.9061884880065918,
+      "learning_rate": 0.0009922222222222222,
+      "loss": 1.1892,
+      "step": 27210
+    },
+    {
+      "epoch": 100.81,
+      "grad_norm": 0.9554376006126404,
+      "learning_rate": 0.0009918518518518518,
+      "loss": 1.1927,
+      "step": 27220
+    },
+    {
+      "epoch": 100.85,
+      "grad_norm": 0.8725953698158264,
+      "learning_rate": 0.0009914814814814816,
+      "loss": 1.1914,
+      "step": 27230
+    },
+    {
+      "epoch": 100.89,
+      "grad_norm": 0.8509664535522461,
+      "learning_rate": 0.0009911111111111112,
+      "loss": 1.1964,
+      "step": 27240
+    },
+    {
+      "epoch": 100.93,
+      "grad_norm": 0.9445471167564392,
+      "learning_rate": 0.0009907407407407408,
+      "loss": 1.2348,
+      "step": 27250
+    },
+    {
+      "epoch": 100.96,
+      "grad_norm": 0.8974488973617554,
+      "learning_rate": 0.0009903703703703704,
+      "loss": 1.2092,
+      "step": 27260
+    },
+    {
+      "epoch": 101.0,
+      "grad_norm": 1.5322643518447876,
+      "learning_rate": 0.00099,
+      "loss": 1.2047,
+      "step": 27270
+    },
+    {
+      "epoch": 101.04,
+      "grad_norm": 0.8989846706390381,
+      "learning_rate": 0.0009896296296296296,
+      "loss": 0.989,
+      "step": 27280
+    },
+    {
+      "epoch": 101.07,
+      "grad_norm": 0.9590046405792236,
+      "learning_rate": 0.0009892592592592594,
+      "loss": 1.0178,
+      "step": 27290
+    },
+    {
+      "epoch": 101.11,
+      "grad_norm": 0.8980274796485901,
+      "learning_rate": 0.000988888888888889,
+      "loss": 1.0279,
+      "step": 27300
+    },
+    {
+      "epoch": 101.15,
+      "grad_norm": 0.9266800880432129,
+      "learning_rate": 0.0009885185185185186,
+      "loss": 1.0291,
+      "step": 27310
+    },
+    {
+      "epoch": 101.19,
+      "grad_norm": 0.8878993391990662,
+      "learning_rate": 0.0009881481481481482,
+      "loss": 1.0202,
+      "step": 27320
+    },
+    {
+      "epoch": 101.22,
+      "grad_norm": 0.9423307776451111,
+      "learning_rate": 0.0009877777777777777,
+      "loss": 1.0479,
+      "step": 27330
+    },
+    {
+      "epoch": 101.26,
+      "grad_norm": 0.911369800567627,
+      "learning_rate": 0.0009874074074074073,
+      "loss": 1.0598,
+      "step": 27340
+    },
+    {
+      "epoch": 101.3,
+      "grad_norm": 0.842523455619812,
+      "learning_rate": 0.0009870370370370371,
+      "loss": 1.0575,
+      "step": 27350
+    },
+    {
+      "epoch": 101.33,
+      "grad_norm": 0.8670434355735779,
+      "learning_rate": 0.0009866666666666667,
+      "loss": 1.086,
+      "step": 27360
+    },
+    {
+      "epoch": 101.37,
+      "grad_norm": 0.9488222599029541,
+      "learning_rate": 0.0009862962962962963,
+      "loss": 1.108,
+      "step": 27370
+    },
+    {
+      "epoch": 101.41,
+      "grad_norm": 0.935550332069397,
+      "learning_rate": 0.000985925925925926,
+      "loss": 1.0789,
+      "step": 27380
+    },
+    {
+      "epoch": 101.44,
+      "grad_norm": 0.930512011051178,
+      "learning_rate": 0.0009855555555555555,
+      "loss": 1.1293,
+      "step": 27390
+    },
+    {
+      "epoch": 101.48,
+      "grad_norm": 0.9525163769721985,
+      "learning_rate": 0.000985185185185185,
+      "loss": 1.1127,
+      "step": 27400
+    },
+    {
+      "epoch": 101.52,
+      "grad_norm": 0.9982643723487854,
+      "learning_rate": 0.0009848148148148149,
+      "loss": 1.1436,
+      "step": 27410
+    },
+    {
+      "epoch": 101.56,
+      "grad_norm": 0.8970288634300232,
+      "learning_rate": 0.0009844444444444445,
+      "loss": 1.11,
+      "step": 27420
+    },
+    {
+      "epoch": 101.59,
+      "grad_norm": 0.9926027655601501,
+      "learning_rate": 0.000984074074074074,
+      "loss": 1.1894,
+      "step": 27430
+    },
+    {
+      "epoch": 101.63,
+      "grad_norm": 0.9675959348678589,
+      "learning_rate": 0.0009837037037037039,
+      "loss": 1.1559,
+      "step": 27440
+    },
+    {
+      "epoch": 101.67,
+      "grad_norm": 0.9283835887908936,
+      "learning_rate": 0.0009833333333333332,
+      "loss": 1.1515,
+      "step": 27450
+    },
+    {
+      "epoch": 101.7,
+      "grad_norm": 0.9680076241493225,
+      "learning_rate": 0.000982962962962963,
+      "loss": 1.1891,
+      "step": 27460
+    },
+    {
+      "epoch": 101.74,
+      "grad_norm": 0.9401657581329346,
+      "learning_rate": 0.0009825925925925926,
+      "loss": 1.1778,
+      "step": 27470
+    },
+    {
+      "epoch": 101.78,
+      "grad_norm": 0.9478588104248047,
+      "learning_rate": 0.0009822222222222222,
+      "loss": 1.1805,
+      "step": 27480
+    },
+    {
+      "epoch": 101.81,
+      "grad_norm": 0.8912003636360168,
+      "learning_rate": 0.0009818518518518518,
+      "loss": 1.1929,
+      "step": 27490
+    },
+    {
+      "epoch": 101.85,
+      "grad_norm": 0.9238876700401306,
+      "learning_rate": 0.0009814814814814816,
+      "loss": 1.1934,
+      "step": 27500
+    },
+    {
+      "epoch": 101.89,
+      "grad_norm": 0.9057316184043884,
+      "learning_rate": 0.0009811111111111112,
+      "loss": 1.1807,
+      "step": 27510
+    },
+    {
+      "epoch": 101.93,
+      "grad_norm": 0.8781848549842834,
+      "learning_rate": 0.0009807407407407408,
+      "loss": 1.227,
+      "step": 27520
+    },
+    {
+      "epoch": 101.96,
+      "grad_norm": 0.9485657811164856,
+      "learning_rate": 0.0009803703703703704,
+      "loss": 1.227,
+      "step": 27530
+    },
+    {
+      "epoch": 102.0,
+      "grad_norm": 1.7745193243026733,
+      "learning_rate": 0.00098,
+      "loss": 1.2126,
+      "step": 27540
+    },
+    {
+      "epoch": 102.04,
+      "grad_norm": 0.8811674118041992,
+      "learning_rate": 0.0009796296296296296,
+      "loss": 0.9722,
+      "step": 27550
+    },
+    {
+      "epoch": 102.07,
+      "grad_norm": 0.9530802369117737,
+      "learning_rate": 0.0009792592592592594,
+      "loss": 0.9783,
+      "step": 27560
+    },
+    {
+      "epoch": 102.11,
+      "grad_norm": 0.8173848986625671,
+      "learning_rate": 0.000978888888888889,
+      "loss": 1.0037,
+      "step": 27570
+    },
+    {
+      "epoch": 102.15,
+      "grad_norm": 0.95224529504776,
+      "learning_rate": 0.0009785185185185185,
+      "loss": 1.0521,
+      "step": 27580
+    },
+    {
+      "epoch": 102.19,
+      "grad_norm": 0.8620432615280151,
+      "learning_rate": 0.0009781481481481481,
+      "loss": 1.0357,
+      "step": 27590
+    },
+    {
+      "epoch": 102.22,
+      "grad_norm": 0.922075092792511,
+      "learning_rate": 0.0009777777777777777,
+      "loss": 1.0585,
+      "step": 27600
+    },
+    {
+      "epoch": 102.26,
+      "grad_norm": 0.8372809290885925,
+      "learning_rate": 0.0009774074074074073,
+      "loss": 1.0536,
+      "step": 27610
+    },
+    {
+      "epoch": 102.3,
+      "grad_norm": 0.9226149320602417,
+      "learning_rate": 0.0009770370370370371,
+      "loss": 1.0465,
+      "step": 27620
+    },
+    {
+      "epoch": 102.33,
+      "grad_norm": 1.0217820405960083,
+      "learning_rate": 0.0009766666666666667,
+      "loss": 1.0734,
+      "step": 27630
+    },
+    {
+      "epoch": 102.37,
+      "grad_norm": 0.8591089248657227,
+      "learning_rate": 0.0009762962962962964,
+      "loss": 1.0942,
+      "step": 27640
+    },
+    {
+      "epoch": 102.41,
+      "grad_norm": 0.9037002325057983,
+      "learning_rate": 0.000975925925925926,
+      "loss": 1.1171,
+      "step": 27650
+    },
+    {
+      "epoch": 102.44,
+      "grad_norm": 0.9767083525657654,
+      "learning_rate": 0.0009755555555555556,
+      "loss": 1.1325,
+      "step": 27660
+    },
+    {
+      "epoch": 102.48,
+      "grad_norm": 0.8895253539085388,
+      "learning_rate": 0.0009751851851851852,
+      "loss": 1.1,
+      "step": 27670
+    },
+    {
+      "epoch": 102.52,
+      "grad_norm": 0.9272596836090088,
+      "learning_rate": 0.0009748148148148149,
+      "loss": 1.1255,
+      "step": 27680
+    },
+    {
+      "epoch": 102.56,
+      "grad_norm": 0.9447619915008545,
+      "learning_rate": 0.0009744444444444444,
+      "loss": 1.1298,
+      "step": 27690
+    },
+    {
+      "epoch": 102.59,
+      "grad_norm": 0.8245322704315186,
+      "learning_rate": 0.0009740740740740741,
+      "loss": 1.135,
+      "step": 27700
+    },
+    {
+      "epoch": 102.63,
+      "grad_norm": 0.8659734129905701,
+      "learning_rate": 0.0009737037037037037,
+      "loss": 1.1282,
+      "step": 27710
+    },
+    {
+      "epoch": 102.67,
+      "grad_norm": 0.8939563035964966,
+      "learning_rate": 0.0009733333333333334,
+      "loss": 1.1602,
+      "step": 27720
+    },
+    {
+      "epoch": 102.7,
+      "grad_norm": 0.9657849073410034,
+      "learning_rate": 0.0009729629629629629,
+      "loss": 1.1643,
+      "step": 27730
+    },
+    {
+      "epoch": 102.74,
+      "grad_norm": 0.9996271133422852,
+      "learning_rate": 0.0009725925925925926,
+      "loss": 1.1761,
+      "step": 27740
+    },
+    {
+      "epoch": 102.78,
+      "grad_norm": 0.8725650310516357,
+      "learning_rate": 0.0009722222222222222,
+      "loss": 1.1291,
+      "step": 27750
+    },
+    {
+      "epoch": 102.81,
+      "grad_norm": 0.9499808549880981,
+      "learning_rate": 0.0009718518518518519,
+      "loss": 1.1882,
+      "step": 27760
+    },
+    {
+      "epoch": 102.85,
+      "grad_norm": 0.9083239436149597,
+      "learning_rate": 0.0009714814814814815,
+      "loss": 1.1722,
+      "step": 27770
+    },
+    {
+      "epoch": 102.89,
+      "grad_norm": 0.9593604803085327,
+      "learning_rate": 0.0009711111111111112,
+      "loss": 1.1936,
+      "step": 27780
+    },
+    {
+      "epoch": 102.93,
+      "grad_norm": 0.9239686131477356,
+      "learning_rate": 0.0009707407407407408,
+      "loss": 1.2229,
+      "step": 27790
+    },
+    {
+      "epoch": 102.96,
+      "grad_norm": 0.9181411862373352,
+      "learning_rate": 0.0009703703703703704,
+      "loss": 1.1673,
+      "step": 27800
+    },
+    {
+      "epoch": 103.0,
+      "grad_norm": 1.701149582862854,
+      "learning_rate": 0.0009699999999999999,
+      "loss": 1.225,
+      "step": 27810
+    },
+    {
+      "epoch": 103.04,
+      "grad_norm": 0.8860218524932861,
+      "learning_rate": 0.0009696296296296296,
+      "loss": 0.9796,
+      "step": 27820
+    },
+    {
+      "epoch": 103.07,
+      "grad_norm": 0.9513492584228516,
+      "learning_rate": 0.0009692592592592593,
+      "loss": 0.9803,
+      "step": 27830
+    },
+    {
+      "epoch": 103.11,
+      "grad_norm": 0.8623229265213013,
+      "learning_rate": 0.0009688888888888889,
+      "loss": 1.0065,
+      "step": 27840
+    },
+    {
+      "epoch": 103.15,
+      "grad_norm": 0.9298003911972046,
+      "learning_rate": 0.0009685185185185186,
+      "loss": 1.0145,
+      "step": 27850
+    },
+    {
+      "epoch": 103.19,
+      "grad_norm": 0.8768572211265564,
+      "learning_rate": 0.0009681481481481482,
+      "loss": 1.0327,
+      "step": 27860
+    },
+    {
+      "epoch": 103.22,
+      "grad_norm": 0.9230377078056335,
+      "learning_rate": 0.0009677777777777778,
+      "loss": 1.0383,
+      "step": 27870
+    },
+    {
+      "epoch": 103.26,
+      "grad_norm": 0.8653807640075684,
+      "learning_rate": 0.0009674074074074074,
+      "loss": 1.0602,
+      "step": 27880
+    },
+    {
+      "epoch": 103.3,
+      "grad_norm": 0.918269157409668,
+      "learning_rate": 0.0009670370370370371,
+      "loss": 1.0644,
+      "step": 27890
+    },
+    {
+      "epoch": 103.33,
+      "grad_norm": 0.940070629119873,
+      "learning_rate": 0.0009666666666666667,
+      "loss": 1.0769,
+      "step": 27900
+    },
+    {
+      "epoch": 103.37,
+      "grad_norm": 0.9915195107460022,
+      "learning_rate": 0.0009662962962962964,
+      "loss": 1.0904,
+      "step": 27910
+    },
+    {
+      "epoch": 103.41,
+      "grad_norm": 0.9446269869804382,
+      "learning_rate": 0.000965925925925926,
+      "loss": 1.0902,
+      "step": 27920
+    },
+    {
+      "epoch": 103.44,
+      "grad_norm": 0.9780631065368652,
+      "learning_rate": 0.0009655555555555555,
+      "loss": 1.1287,
+      "step": 27930
+    },
+    {
+      "epoch": 103.48,
+      "grad_norm": 1.0551878213882446,
+      "learning_rate": 0.0009651851851851851,
+      "loss": 1.103,
+      "step": 27940
+    },
+    {
+      "epoch": 103.52,
+      "grad_norm": 0.9577930569648743,
+      "learning_rate": 0.0009648148148148148,
+      "loss": 1.0911,
+      "step": 27950
+    },
+    {
+      "epoch": 103.56,
+      "grad_norm": 0.8572773933410645,
+      "learning_rate": 0.0009644444444444444,
+      "loss": 1.1244,
+      "step": 27960
+    },
+    {
+      "epoch": 103.59,
+      "grad_norm": 0.8974463939666748,
+      "learning_rate": 0.0009640740740740741,
+      "loss": 1.1524,
+      "step": 27970
+    },
+    {
+      "epoch": 103.63,
+      "grad_norm": 0.911722719669342,
+      "learning_rate": 0.0009637037037037037,
+      "loss": 1.1478,
+      "step": 27980
+    },
+    {
+      "epoch": 103.67,
+      "grad_norm": 0.9285465478897095,
+      "learning_rate": 0.0009633333333333334,
+      "loss": 1.1461,
+      "step": 27990
+    },
+    {
+      "epoch": 103.7,
+      "grad_norm": 0.9444763660430908,
+      "learning_rate": 0.0009629629629629629,
+      "loss": 1.1529,
+      "step": 28000
+    },
+    {
+      "epoch": 103.74,
+      "grad_norm": 0.897024929523468,
+      "learning_rate": 0.0009625925925925926,
+      "loss": 1.1455,
+      "step": 28010
+    },
+    {
+      "epoch": 103.78,
+      "grad_norm": 0.9245248436927795,
+      "learning_rate": 0.0009622222222222222,
+      "loss": 1.172,
+      "step": 28020
+    },
+    {
+      "epoch": 103.81,
+      "grad_norm": 0.9002532958984375,
+      "learning_rate": 0.0009618518518518519,
+      "loss": 1.167,
+      "step": 28030
+    },
+    {
+      "epoch": 103.85,
+      "grad_norm": 0.938520610332489,
+      "learning_rate": 0.0009614814814814816,
+      "loss": 1.1713,
+      "step": 28040
+    },
+    {
+      "epoch": 103.89,
+      "grad_norm": 0.9639129042625427,
+      "learning_rate": 0.0009611111111111112,
+      "loss": 1.181,
+      "step": 28050
+    },
+    {
+      "epoch": 103.93,
+      "grad_norm": 0.9609660506248474,
+      "learning_rate": 0.0009607407407407408,
+      "loss": 1.1948,
+      "step": 28060
+    },
+    {
+      "epoch": 103.96,
+      "grad_norm": 0.9552947878837585,
+      "learning_rate": 0.0009603703703703703,
+      "loss": 1.2001,
+      "step": 28070
+    },
+    {
+      "epoch": 104.0,
+      "grad_norm": 1.452337384223938,
+      "learning_rate": 0.00096,
+      "loss": 1.1939,
+      "step": 28080
+    },
+    {
+      "epoch": 104.04,
+      "grad_norm": 0.8623477220535278,
+      "learning_rate": 0.0009596296296296296,
+      "loss": 0.9715,
+      "step": 28090
+    },
+    {
+      "epoch": 104.07,
+      "grad_norm": 0.8784454464912415,
+      "learning_rate": 0.0009592592592592593,
+      "loss": 0.9805,
+      "step": 28100
+    },
+    {
+      "epoch": 104.11,
+      "grad_norm": 0.939306914806366,
+      "learning_rate": 0.0009588888888888889,
+      "loss": 0.9661,
+      "step": 28110
+    },
+    {
+      "epoch": 104.15,
+      "grad_norm": 0.9585001468658447,
+      "learning_rate": 0.0009585185185185186,
+      "loss": 1.0202,
+      "step": 28120
+    },
+    {
+      "epoch": 104.19,
+      "grad_norm": 0.9838183522224426,
+      "learning_rate": 0.0009581481481481482,
+      "loss": 1.0172,
+      "step": 28130
+    },
+    {
+      "epoch": 104.22,
+      "grad_norm": 0.9469972252845764,
+      "learning_rate": 0.0009577777777777778,
+      "loss": 1.0398,
+      "step": 28140
+    },
+    {
+      "epoch": 104.26,
+      "grad_norm": 0.933664083480835,
+      "learning_rate": 0.0009574074074074074,
+      "loss": 1.0533,
+      "step": 28150
+    },
+    {
+      "epoch": 104.3,
+      "grad_norm": 0.97636479139328,
+      "learning_rate": 0.0009570370370370371,
+      "loss": 1.0557,
+      "step": 28160
+    },
+    {
+      "epoch": 104.33,
+      "grad_norm": 0.8865147829055786,
+      "learning_rate": 0.0009566666666666666,
+      "loss": 1.0592,
+      "step": 28170
+    },
+    {
+      "epoch": 104.37,
+      "grad_norm": 0.9205927848815918,
+      "learning_rate": 0.0009562962962962963,
+      "loss": 1.0823,
+      "step": 28180
+    },
+    {
+      "epoch": 104.41,
+      "grad_norm": 0.9018583297729492,
+      "learning_rate": 0.0009559259259259259,
+      "loss": 1.0743,
+      "step": 28190
+    },
+    {
+      "epoch": 104.44,
+      "grad_norm": 1.000645637512207,
+      "learning_rate": 0.0009555555555555556,
+      "loss": 1.1016,
+      "step": 28200
+    },
+    {
+      "epoch": 104.48,
+      "grad_norm": 0.9386842250823975,
+      "learning_rate": 0.0009551851851851851,
+      "loss": 1.0779,
+      "step": 28210
+    },
+    {
+      "epoch": 104.52,
+      "grad_norm": 0.8640486001968384,
+      "learning_rate": 0.0009548148148148148,
+      "loss": 1.0997,
+      "step": 28220
+    },
+    {
+      "epoch": 104.56,
+      "grad_norm": 0.9148708581924438,
+      "learning_rate": 0.0009544444444444445,
+      "loss": 1.1083,
+      "step": 28230
+    },
+    {
+      "epoch": 104.59,
+      "grad_norm": 0.8763410449028015,
+      "learning_rate": 0.0009540740740740741,
+      "loss": 1.1069,
+      "step": 28240
+    },
+    {
+      "epoch": 104.63,
+      "grad_norm": 0.9024518728256226,
+      "learning_rate": 0.0009537037037037038,
+      "loss": 1.0925,
+      "step": 28250
+    },
+    {
+      "epoch": 104.67,
+      "grad_norm": 1.0220943689346313,
+      "learning_rate": 0.0009533333333333334,
+      "loss": 1.1489,
+      "step": 28260
+    },
+    {
+      "epoch": 104.7,
+      "grad_norm": 0.9682163596153259,
+      "learning_rate": 0.0009529629629629631,
+      "loss": 1.1572,
+      "step": 28270
+    },
+    {
+      "epoch": 104.74,
+      "grad_norm": 0.9184615612030029,
+      "learning_rate": 0.0009525925925925926,
+      "loss": 1.1555,
+      "step": 28280
+    },
+    {
+      "epoch": 104.78,
+      "grad_norm": 0.9440074563026428,
+      "learning_rate": 0.0009522222222222223,
+      "loss": 1.1587,
+      "step": 28290
+    },
+    {
+      "epoch": 104.81,
+      "grad_norm": 0.9604818224906921,
+      "learning_rate": 0.0009518518518518518,
+      "loss": 1.16,
+      "step": 28300
+    },
+    {
+      "epoch": 104.85,
+      "grad_norm": 0.8982845544815063,
+      "learning_rate": 0.0009514814814814815,
+      "loss": 1.1673,
+      "step": 28310
+    },
+    {
+      "epoch": 104.89,
+      "grad_norm": 0.8533664345741272,
+      "learning_rate": 0.0009511111111111111,
+      "loss": 1.1979,
+      "step": 28320
+    },
+    {
+      "epoch": 104.93,
+      "grad_norm": 0.9376116394996643,
+      "learning_rate": 0.0009507407407407408,
+      "loss": 1.2014,
+      "step": 28330
+    },
+    {
+      "epoch": 104.96,
+      "grad_norm": 0.9965450763702393,
+      "learning_rate": 0.0009503703703703704,
+      "loss": 1.1756,
+      "step": 28340
+    },
+    {
+      "epoch": 105.0,
+      "grad_norm": 1.7779862880706787,
+      "learning_rate": 0.00095,
+      "loss": 1.1981,
+      "step": 28350
+    },
+    {
+      "epoch": 105.04,
+      "grad_norm": 0.9287167191505432,
+      "learning_rate": 0.0009496296296296296,
+      "loss": 0.9635,
+      "step": 28360
+    },
+    {
+      "epoch": 105.07,
+      "grad_norm": 0.8633070588111877,
+      "learning_rate": 0.0009492592592592593,
+      "loss": 0.9735,
+      "step": 28370
+    },
+    {
+      "epoch": 105.11,
+      "grad_norm": 0.9253849983215332,
+      "learning_rate": 0.0009488888888888889,
+      "loss": 0.9967,
+      "step": 28380
+    },
+    {
+      "epoch": 105.15,
+      "grad_norm": 0.8466000556945801,
+      "learning_rate": 0.0009485185185185186,
+      "loss": 0.9589,
+      "step": 28390
+    },
+    {
+      "epoch": 105.19,
+      "grad_norm": 0.8577874302864075,
+      "learning_rate": 0.0009481481481481482,
+      "loss": 0.9805,
+      "step": 28400
+    },
+    {
+      "epoch": 105.22,
+      "grad_norm": 0.9374727010726929,
+      "learning_rate": 0.0009477777777777779,
+      "loss": 1.0293,
+      "step": 28410
+    },
+    {
+      "epoch": 105.26,
+      "grad_norm": 1.04707932472229,
+      "learning_rate": 0.0009474074074074073,
+      "loss": 1.0373,
+      "step": 28420
+    },
+    {
+      "epoch": 105.3,
+      "grad_norm": 0.9615543484687805,
+      "learning_rate": 0.000947037037037037,
+      "loss": 1.0446,
+      "step": 28430
+    },
+    {
+      "epoch": 105.33,
+      "grad_norm": 0.9235749244689941,
+      "learning_rate": 0.0009466666666666667,
+      "loss": 1.0288,
+      "step": 28440
+    },
+    {
+      "epoch": 105.37,
+      "grad_norm": 0.9541982412338257,
+      "learning_rate": 0.0009462962962962963,
+      "loss": 1.0546,
+      "step": 28450
+    },
+    {
+      "epoch": 105.41,
+      "grad_norm": 0.891716718673706,
+      "learning_rate": 0.000945925925925926,
+      "loss": 1.095,
+      "step": 28460
+    },
+    {
+      "epoch": 105.44,
+      "grad_norm": 0.9158815145492554,
+      "learning_rate": 0.0009455555555555556,
+      "loss": 1.0881,
+      "step": 28470
+    },
+    {
+      "epoch": 105.48,
+      "grad_norm": 1.0235891342163086,
+      "learning_rate": 0.0009451851851851853,
+      "loss": 1.1174,
+      "step": 28480
+    },
+    {
+      "epoch": 105.52,
+      "grad_norm": 0.9569323658943176,
+      "learning_rate": 0.0009448148148148148,
+      "loss": 1.1238,
+      "step": 28490
+    },
+    {
+      "epoch": 105.56,
+      "grad_norm": 1.079743504524231,
+      "learning_rate": 0.0009444444444444445,
+      "loss": 1.1312,
+      "step": 28500
+    },
+    {
+      "epoch": 105.59,
+      "grad_norm": 0.929822564125061,
+      "learning_rate": 0.0009440740740740741,
+      "loss": 1.106,
+      "step": 28510
+    },
+    {
+      "epoch": 105.63,
+      "grad_norm": 0.9850989580154419,
+      "learning_rate": 0.0009437037037037038,
+      "loss": 1.1194,
+      "step": 28520
+    },
+    {
+      "epoch": 105.67,
+      "grad_norm": 0.919579029083252,
+      "learning_rate": 0.0009433333333333334,
+      "loss": 1.1395,
+      "step": 28530
+    },
+    {
+      "epoch": 105.7,
+      "grad_norm": 0.9597387313842773,
+      "learning_rate": 0.000942962962962963,
+      "loss": 1.1527,
+      "step": 28540
+    },
+    {
+      "epoch": 105.74,
+      "grad_norm": 0.9359583854675293,
+      "learning_rate": 0.0009425925925925925,
+      "loss": 1.1315,
+      "step": 28550
+    },
+    {
+      "epoch": 105.78,
+      "grad_norm": 1.0015721321105957,
+      "learning_rate": 0.0009422222222222222,
+      "loss": 1.1733,
+      "step": 28560
+    },
+    {
+      "epoch": 105.81,
+      "grad_norm": 0.9597902894020081,
+      "learning_rate": 0.0009418518518518518,
+      "loss": 1.1285,
+      "step": 28570
+    },
+    {
+      "epoch": 105.85,
+      "grad_norm": 0.9335023760795593,
+      "learning_rate": 0.0009414814814814815,
+      "loss": 1.1722,
+      "step": 28580
+    },
+    {
+      "epoch": 105.89,
+      "grad_norm": 0.9263126254081726,
+      "learning_rate": 0.0009411111111111111,
+      "loss": 1.164,
+      "step": 28590
+    },
+    {
+      "epoch": 105.93,
+      "grad_norm": 0.9743094444274902,
+      "learning_rate": 0.0009407407407407408,
+      "loss": 1.1819,
+      "step": 28600
+    },
+    {
+      "epoch": 105.96,
+      "grad_norm": 0.8798518776893616,
+      "learning_rate": 0.0009403703703703704,
+      "loss": 1.1831,
+      "step": 28610
+    },
+    {
+      "epoch": 106.0,
+      "grad_norm": 2.3925769329071045,
+      "learning_rate": 0.00094,
+      "loss": 1.1827,
+      "step": 28620
+    },
+    {
+      "epoch": 106.04,
+      "grad_norm": 0.9399033188819885,
+      "learning_rate": 0.0009396296296296296,
+      "loss": 0.966,
+      "step": 28630
+    },
+    {
+      "epoch": 106.07,
+      "grad_norm": 0.9358739852905273,
+      "learning_rate": 0.0009392592592592593,
+      "loss": 0.9811,
+      "step": 28640
+    },
+    {
+      "epoch": 106.11,
+      "grad_norm": 0.9023323059082031,
+      "learning_rate": 0.000938888888888889,
+      "loss": 0.9646,
+      "step": 28650
+    },
+    {
+      "epoch": 106.15,
+      "grad_norm": 0.8303003907203674,
+      "learning_rate": 0.0009385185185185185,
+      "loss": 0.9887,
+      "step": 28660
+    },
+    {
+      "epoch": 106.19,
+      "grad_norm": 0.8875153064727783,
+      "learning_rate": 0.0009381481481481482,
+      "loss": 0.9814,
+      "step": 28670
+    },
+    {
+      "epoch": 106.22,
+      "grad_norm": 0.8730666041374207,
+      "learning_rate": 0.0009377777777777778,
+      "loss": 0.9953,
+      "step": 28680
+    },
+    {
+      "epoch": 106.26,
+      "grad_norm": 0.979469895362854,
+      "learning_rate": 0.0009374074074074074,
+      "loss": 1.018,
+      "step": 28690
+    },
+    {
+      "epoch": 106.3,
+      "grad_norm": 1.0179818868637085,
+      "learning_rate": 0.000937037037037037,
+      "loss": 1.0454,
+      "step": 28700
+    },
+    {
+      "epoch": 106.33,
+      "grad_norm": 1.0186294317245483,
+      "learning_rate": 0.0009366666666666667,
+      "loss": 1.053,
+      "step": 28710
+    },
+    {
+      "epoch": 106.37,
+      "grad_norm": 0.8819359540939331,
+      "learning_rate": 0.0009362962962962963,
+      "loss": 1.0552,
+      "step": 28720
+    },
+    {
+      "epoch": 106.41,
+      "grad_norm": 0.8499149084091187,
+      "learning_rate": 0.000935925925925926,
+      "loss": 1.077,
+      "step": 28730
+    },
+    {
+      "epoch": 106.44,
+      "grad_norm": 0.9276627898216248,
+      "learning_rate": 0.0009355555555555556,
+      "loss": 1.0989,
+      "step": 28740
+    },
+    {
+      "epoch": 106.48,
+      "grad_norm": 0.9268879890441895,
+      "learning_rate": 0.0009351851851851853,
+      "loss": 1.0761,
+      "step": 28750
+    },
+    {
+      "epoch": 106.52,
+      "grad_norm": 0.9042457342147827,
+      "learning_rate": 0.0009348148148148148,
+      "loss": 1.0841,
+      "step": 28760
+    },
+    {
+      "epoch": 106.56,
+      "grad_norm": 0.9719822406768799,
+      "learning_rate": 0.0009344444444444444,
+      "loss": 1.0817,
+      "step": 28770
+    },
+    {
+      "epoch": 106.59,
+      "grad_norm": 0.9261794090270996,
+      "learning_rate": 0.000934074074074074,
+      "loss": 1.0902,
+      "step": 28780
+    },
+    {
+      "epoch": 106.63,
+      "grad_norm": 0.9322009086608887,
+      "learning_rate": 0.0009337037037037037,
+      "loss": 1.1328,
+      "step": 28790
+    },
+    {
+      "epoch": 106.67,
+      "grad_norm": 0.9948487877845764,
+      "learning_rate": 0.0009333333333333333,
+      "loss": 1.1484,
+      "step": 28800
+    },
+    {
+      "epoch": 106.7,
+      "grad_norm": 1.0035136938095093,
+      "learning_rate": 0.000932962962962963,
+      "loss": 1.1586,
+      "step": 28810
+    },
+    {
+      "epoch": 106.74,
+      "grad_norm": 1.028941035270691,
+      "learning_rate": 0.0009325925925925926,
+      "loss": 1.1316,
+      "step": 28820
+    },
+    {
+      "epoch": 106.78,
+      "grad_norm": 0.9474858641624451,
+      "learning_rate": 0.0009322222222222222,
+      "loss": 1.1288,
+      "step": 28830
+    },
+    {
+      "epoch": 106.81,
+      "grad_norm": 0.9305292963981628,
+      "learning_rate": 0.0009318518518518518,
+      "loss": 1.119,
+      "step": 28840
+    },
+    {
+      "epoch": 106.85,
+      "grad_norm": 0.919070303440094,
+      "learning_rate": 0.0009314814814814815,
+      "loss": 1.1667,
+      "step": 28850
+    },
+    {
+      "epoch": 106.89,
+      "grad_norm": 0.9206348657608032,
+      "learning_rate": 0.0009311111111111112,
+      "loss": 1.1655,
+      "step": 28860
+    },
+    {
+      "epoch": 106.93,
+      "grad_norm": 1.0312327146530151,
+      "learning_rate": 0.0009307407407407408,
+      "loss": 1.1592,
+      "step": 28870
+    },
+    {
+      "epoch": 106.96,
+      "grad_norm": 0.9701687693595886,
+      "learning_rate": 0.0009303703703703705,
+      "loss": 1.1635,
+      "step": 28880
+    },
+    {
+      "epoch": 107.0,
+      "grad_norm": 1.8190052509307861,
+      "learning_rate": 0.00093,
+      "loss": 1.178,
+      "step": 28890
+    },
+    {
+      "epoch": 107.04,
+      "grad_norm": 0.9251551628112793,
+      "learning_rate": 0.0009296296296296296,
+      "loss": 0.9336,
+      "step": 28900
+    },
+    {
+      "epoch": 107.07,
+      "grad_norm": 0.8757014274597168,
+      "learning_rate": 0.0009292592592592592,
+      "loss": 0.9469,
+      "step": 28910
+    },
+    {
+      "epoch": 107.11,
+      "grad_norm": 0.9648277163505554,
+      "learning_rate": 0.0009288888888888889,
+      "loss": 0.9738,
+      "step": 28920
+    },
+    {
+      "epoch": 107.15,
+      "grad_norm": 0.938714325428009,
+      "learning_rate": 0.0009285185185185185,
+      "loss": 0.9906,
+      "step": 28930
+    },
+    {
+      "epoch": 107.19,
+      "grad_norm": 0.9022919535636902,
+      "learning_rate": 0.0009281481481481482,
+      "loss": 1.01,
+      "step": 28940
+    },
+    {
+      "epoch": 107.22,
+      "grad_norm": 0.917122483253479,
+      "learning_rate": 0.0009277777777777778,
+      "loss": 1.0086,
+      "step": 28950
+    },
+    {
+      "epoch": 107.26,
+      "grad_norm": 0.9072996377944946,
+      "learning_rate": 0.0009274074074074075,
+      "loss": 1.0098,
+      "step": 28960
+    },
+    {
+      "epoch": 107.3,
+      "grad_norm": 0.9521521925926208,
+      "learning_rate": 0.000927037037037037,
+      "loss": 1.0383,
+      "step": 28970
+    },
+    {
+      "epoch": 107.33,
+      "grad_norm": 0.9218719601631165,
+      "learning_rate": 0.0009266666666666667,
+      "loss": 1.0318,
+      "step": 28980
+    },
+    {
+      "epoch": 107.37,
+      "grad_norm": 0.9192262291908264,
+      "learning_rate": 0.0009262962962962963,
+      "loss": 1.0327,
+      "step": 28990
+    },
+    {
+      "epoch": 107.41,
+      "grad_norm": 0.9420268535614014,
+      "learning_rate": 0.000925925925925926,
+      "loss": 1.0612,
+      "step": 29000
+    },
+    {
+      "epoch": 107.44,
+      "grad_norm": 0.9131075143814087,
+      "learning_rate": 0.0009255555555555555,
+      "loss": 1.0599,
+      "step": 29010
+    },
+    {
+      "epoch": 107.48,
+      "grad_norm": 0.9630795121192932,
+      "learning_rate": 0.0009251851851851852,
+      "loss": 1.0725,
+      "step": 29020
+    },
+    {
+      "epoch": 107.52,
+      "grad_norm": 0.8927704691886902,
+      "learning_rate": 0.0009248148148148148,
+      "loss": 1.0808,
+      "step": 29030
+    },
+    {
+      "epoch": 107.56,
+      "grad_norm": 0.9514608979225159,
+      "learning_rate": 0.0009244444444444444,
+      "loss": 1.098,
+      "step": 29040
+    },
+    {
+      "epoch": 107.59,
+      "grad_norm": 0.9280536770820618,
+      "learning_rate": 0.0009240740740740741,
+      "loss": 1.1015,
+      "step": 29050
+    },
+    {
+      "epoch": 107.63,
+      "grad_norm": 1.0405588150024414,
+      "learning_rate": 0.0009237037037037037,
+      "loss": 1.1082,
+      "step": 29060
+    },
+    {
+      "epoch": 107.67,
+      "grad_norm": 0.8975964188575745,
+      "learning_rate": 0.0009233333333333334,
+      "loss": 1.1108,
+      "step": 29070
+    },
+    {
+      "epoch": 107.7,
+      "grad_norm": 1.017134666442871,
+      "learning_rate": 0.000922962962962963,
+      "loss": 1.1371,
+      "step": 29080
+    },
+    {
+      "epoch": 107.74,
+      "grad_norm": 0.9074270129203796,
+      "learning_rate": 0.0009225925925925927,
+      "loss": 1.1241,
+      "step": 29090
+    },
+    {
+      "epoch": 107.78,
+      "grad_norm": 1.0147840976715088,
+      "learning_rate": 0.0009222222222222223,
+      "loss": 1.1267,
+      "step": 29100
+    },
+    {
+      "epoch": 107.81,
+      "grad_norm": 0.9706998467445374,
+      "learning_rate": 0.0009218518518518519,
+      "loss": 1.1578,
+      "step": 29110
+    },
+    {
+      "epoch": 107.85,
+      "grad_norm": 0.9495412111282349,
+      "learning_rate": 0.0009214814814814815,
+      "loss": 1.1471,
+      "step": 29120
+    },
+    {
+      "epoch": 107.89,
+      "grad_norm": 0.8967522382736206,
+      "learning_rate": 0.0009211111111111112,
+      "loss": 1.167,
+      "step": 29130
+    },
+    {
+      "epoch": 107.93,
+      "grad_norm": 0.9381629228591919,
+      "learning_rate": 0.0009207407407407407,
+      "loss": 1.1626,
+      "step": 29140
+    },
+    {
+      "epoch": 107.96,
+      "grad_norm": 0.9799057841300964,
+      "learning_rate": 0.0009203703703703704,
+      "loss": 1.1653,
+      "step": 29150
+    },
+    {
+      "epoch": 108.0,
+      "grad_norm": 1.7408097982406616,
+      "learning_rate": 0.00092,
+      "loss": 1.1384,
+      "step": 29160
+    },
+    {
+      "epoch": 108.04,
+      "grad_norm": 0.8933138251304626,
+      "learning_rate": 0.0009196296296296296,
+      "loss": 0.9233,
+      "step": 29170
+    },
+    {
+      "epoch": 108.07,
+      "grad_norm": 0.838168203830719,
+      "learning_rate": 0.0009192592592592592,
+      "loss": 0.9545,
+      "step": 29180
+    },
+    {
+      "epoch": 108.11,
+      "grad_norm": 0.9422211050987244,
+      "learning_rate": 0.0009188888888888889,
+      "loss": 0.9547,
+      "step": 29190
+    },
+    {
+      "epoch": 108.15,
+      "grad_norm": 0.8847689628601074,
+      "learning_rate": 0.0009185185185185185,
+      "loss": 0.9664,
+      "step": 29200
+    },
+    {
+      "epoch": 108.19,
+      "grad_norm": 0.9079718589782715,
+      "learning_rate": 0.0009181481481481482,
+      "loss": 0.9592,
+      "step": 29210
+    },
+    {
+      "epoch": 108.22,
+      "grad_norm": 0.9272188544273376,
+      "learning_rate": 0.0009177777777777778,
+      "loss": 0.9965,
+      "step": 29220
+    },
+    {
+      "epoch": 108.26,
+      "grad_norm": 0.8760783076286316,
+      "learning_rate": 0.0009174074074074075,
+      "loss": 1.0044,
+      "step": 29230
+    },
+    {
+      "epoch": 108.3,
+      "grad_norm": 0.9839217066764832,
+      "learning_rate": 0.000917037037037037,
+      "loss": 1.0261,
+      "step": 29240
+    },
+    {
+      "epoch": 108.33,
+      "grad_norm": 0.8970228433609009,
+      "learning_rate": 0.0009166666666666666,
+      "loss": 1.0533,
+      "step": 29250
+    },
+    {
+      "epoch": 108.37,
+      "grad_norm": 0.9859808087348938,
+      "learning_rate": 0.0009162962962962963,
+      "loss": 1.0375,
+      "step": 29260
+    },
+    {
+      "epoch": 108.41,
+      "grad_norm": 0.9420928955078125,
+      "learning_rate": 0.0009159259259259259,
+      "loss": 1.064,
+      "step": 29270
+    },
+    {
+      "epoch": 108.44,
+      "grad_norm": 0.9270564317703247,
+      "learning_rate": 0.0009155555555555556,
+      "loss": 1.0486,
+      "step": 29280
+    },
+    {
+      "epoch": 108.48,
+      "grad_norm": 0.9782211184501648,
+      "learning_rate": 0.0009151851851851852,
+      "loss": 1.0586,
+      "step": 29290
+    },
+    {
+      "epoch": 108.52,
+      "grad_norm": 0.9622315764427185,
+      "learning_rate": 0.0009148148148148149,
+      "loss": 1.0872,
+      "step": 29300
+    },
+    {
+      "epoch": 108.56,
+      "grad_norm": 1.0131900310516357,
+      "learning_rate": 0.0009144444444444444,
+      "loss": 1.0958,
+      "step": 29310
+    },
+    {
+      "epoch": 108.59,
+      "grad_norm": 1.0280399322509766,
+      "learning_rate": 0.0009140740740740741,
+      "loss": 1.0959,
+      "step": 29320
+    },
+    {
+      "epoch": 108.63,
+      "grad_norm": 0.9791631698608398,
+      "learning_rate": 0.0009137037037037037,
+      "loss": 1.123,
+      "step": 29330
+    },
+    {
+      "epoch": 108.67,
+      "grad_norm": 1.0855169296264648,
+      "learning_rate": 0.0009133333333333334,
+      "loss": 1.1025,
+      "step": 29340
+    },
+    {
+      "epoch": 108.7,
+      "grad_norm": 0.9479579925537109,
+      "learning_rate": 0.000912962962962963,
+      "loss": 1.1195,
+      "step": 29350
+    },
+    {
+      "epoch": 108.74,
+      "grad_norm": 0.9171406626701355,
+      "learning_rate": 0.0009125925925925927,
+      "loss": 1.1071,
+      "step": 29360
+    },
+    {
+      "epoch": 108.78,
+      "grad_norm": 0.9832227230072021,
+      "learning_rate": 0.0009122222222222223,
+      "loss": 1.1309,
+      "step": 29370
+    },
+    {
+      "epoch": 108.81,
+      "grad_norm": 1.0037956237792969,
+      "learning_rate": 0.0009118518518518518,
+      "loss": 1.1322,
+      "step": 29380
+    },
+    {
+      "epoch": 108.85,
+      "grad_norm": 1.0154857635498047,
+      "learning_rate": 0.0009114814814814814,
+      "loss": 1.1363,
+      "step": 29390
+    },
+    {
+      "epoch": 108.89,
+      "grad_norm": 0.9449747204780579,
+      "learning_rate": 0.0009111111111111111,
+      "loss": 1.1613,
+      "step": 29400
+    },
+    {
+      "epoch": 108.93,
+      "grad_norm": 0.9238153696060181,
+      "learning_rate": 0.0009107407407407407,
+      "loss": 1.1424,
+      "step": 29410
+    },
+    {
+      "epoch": 108.96,
+      "grad_norm": 0.9596208930015564,
+      "learning_rate": 0.0009103703703703704,
+      "loss": 1.1441,
+      "step": 29420
+    },
+    {
+      "epoch": 109.0,
+      "grad_norm": 1.8561451435089111,
+      "learning_rate": 0.00091,
+      "loss": 1.1483,
+      "step": 29430
+    },
+    {
+      "epoch": 109.04,
+      "grad_norm": 0.9528045654296875,
+      "learning_rate": 0.0009096296296296297,
+      "loss": 0.9567,
+      "step": 29440
+    },
+    {
+      "epoch": 109.07,
+      "grad_norm": 0.9137498736381531,
+      "learning_rate": 0.0009092592592592592,
+      "loss": 0.9376,
+      "step": 29450
+    },
+    {
+      "epoch": 109.11,
+      "grad_norm": 0.8889475464820862,
+      "learning_rate": 0.0009088888888888889,
+      "loss": 0.9589,
+      "step": 29460
+    },
+    {
+      "epoch": 109.15,
+      "grad_norm": 0.9228796362876892,
+      "learning_rate": 0.0009085185185185186,
+      "loss": 0.9657,
+      "step": 29470
+    },
+    {
+      "epoch": 109.19,
+      "grad_norm": 0.9670528173446655,
+      "learning_rate": 0.0009081481481481482,
+      "loss": 0.9863,
+      "step": 29480
+    },
+    {
+      "epoch": 109.22,
+      "grad_norm": 1.0339789390563965,
+      "learning_rate": 0.0009077777777777779,
+      "loss": 0.9995,
+      "step": 29490
+    },
+    {
+      "epoch": 109.26,
+      "grad_norm": 0.9455500245094299,
+      "learning_rate": 0.0009074074074074074,
+      "loss": 1.0002,
+      "step": 29500
+    },
+    {
+      "epoch": 109.3,
+      "grad_norm": 0.9690384268760681,
+      "learning_rate": 0.0009070370370370371,
+      "loss": 1.0194,
+      "step": 29510
+    },
+    {
+      "epoch": 109.33,
+      "grad_norm": 0.9900230765342712,
+      "learning_rate": 0.0009066666666666666,
+      "loss": 1.0351,
+      "step": 29520
+    },
+    {
+      "epoch": 109.37,
+      "grad_norm": 1.0021939277648926,
+      "learning_rate": 0.0009062962962962963,
+      "loss": 1.0486,
+      "step": 29530
+    },
+    {
+      "epoch": 109.41,
+      "grad_norm": 0.9087638854980469,
+      "learning_rate": 0.0009059259259259259,
+      "loss": 1.0461,
+      "step": 29540
+    },
+    {
+      "epoch": 109.44,
+      "grad_norm": 0.8851718306541443,
+      "learning_rate": 0.0009055555555555556,
+      "loss": 1.0577,
+      "step": 29550
+    },
+    {
+      "epoch": 109.48,
+      "grad_norm": 0.9451716542243958,
+      "learning_rate": 0.0009051851851851852,
+      "loss": 1.0504,
+      "step": 29560
+    },
+    {
+      "epoch": 109.52,
+      "grad_norm": 0.9180352091789246,
+      "learning_rate": 0.0009048148148148149,
+      "loss": 1.0896,
+      "step": 29570
+    },
+    {
+      "epoch": 109.56,
+      "grad_norm": 0.9067238569259644,
+      "learning_rate": 0.0009044444444444445,
+      "loss": 1.0742,
+      "step": 29580
+    },
+    {
+      "epoch": 109.59,
+      "grad_norm": 0.9143742918968201,
+      "learning_rate": 0.0009040740740740741,
+      "loss": 1.0828,
+      "step": 29590
+    },
+    {
+      "epoch": 109.63,
+      "grad_norm": 0.95147305727005,
+      "learning_rate": 0.0009037037037037037,
+      "loss": 1.0835,
+      "step": 29600
+    },
+    {
+      "epoch": 109.67,
+      "grad_norm": 0.9197315573692322,
+      "learning_rate": 0.0009033333333333334,
+      "loss": 1.1058,
+      "step": 29610
+    },
+    {
+      "epoch": 109.7,
+      "grad_norm": 1.0208014249801636,
+      "learning_rate": 0.0009029629629629629,
+      "loss": 1.1029,
+      "step": 29620
+    },
+    {
+      "epoch": 109.74,
+      "grad_norm": 0.9763293266296387,
+      "learning_rate": 0.0009025925925925926,
+      "loss": 1.0799,
+      "step": 29630
+    },
+    {
+      "epoch": 109.78,
+      "grad_norm": 0.9239047169685364,
+      "learning_rate": 0.0009022222222222222,
+      "loss": 1.1195,
+      "step": 29640
+    },
+    {
+      "epoch": 109.81,
+      "grad_norm": 0.9763393998146057,
+      "learning_rate": 0.0009018518518518519,
+      "loss": 1.1211,
+      "step": 29650
+    },
+    {
+      "epoch": 109.85,
+      "grad_norm": 0.9521705508232117,
+      "learning_rate": 0.0009014814814814814,
+      "loss": 1.1382,
+      "step": 29660
+    },
+    {
+      "epoch": 109.89,
+      "grad_norm": 0.9943112134933472,
+      "learning_rate": 0.0009011111111111111,
+      "loss": 1.1134,
+      "step": 29670
+    },
+    {
+      "epoch": 109.93,
+      "grad_norm": 0.9353739023208618,
+      "learning_rate": 0.0009007407407407408,
+      "loss": 1.1501,
+      "step": 29680
+    },
+    {
+      "epoch": 109.96,
+      "grad_norm": 0.9845727682113647,
+      "learning_rate": 0.0009003703703703704,
+      "loss": 1.1209,
+      "step": 29690
+    },
+    {
+      "epoch": 110.0,
+      "grad_norm": 1.6984339952468872,
+      "learning_rate": 0.0009000000000000001,
+      "loss": 1.1666,
+      "step": 29700
+    },
+    {
+      "epoch": 110.04,
+      "grad_norm": 0.9160362482070923,
+      "learning_rate": 0.0008996296296296297,
+      "loss": 0.9263,
+      "step": 29710
+    },
+    {
+      "epoch": 110.07,
+      "grad_norm": 0.908966600894928,
+      "learning_rate": 0.0008992592592592594,
+      "loss": 0.9175,
+      "step": 29720
+    },
+    {
+      "epoch": 110.11,
+      "grad_norm": 0.9472144246101379,
+      "learning_rate": 0.0008988888888888888,
+      "loss": 0.935,
+      "step": 29730
+    },
+    {
+      "epoch": 110.15,
+      "grad_norm": 0.9157045483589172,
+      "learning_rate": 0.0008985185185185185,
+      "loss": 0.9417,
+      "step": 29740
+    },
+    {
+      "epoch": 110.19,
+      "grad_norm": 0.9928398728370667,
+      "learning_rate": 0.0008981481481481481,
+      "loss": 0.9652,
+      "step": 29750
+    },
+    {
+      "epoch": 110.22,
+      "grad_norm": 0.9916333556175232,
+      "learning_rate": 0.0008977777777777778,
+      "loss": 1.0001,
+      "step": 29760
+    },
+    {
+      "epoch": 110.26,
+      "grad_norm": 0.9936506152153015,
+      "learning_rate": 0.0008974074074074074,
+      "loss": 0.9845,
+      "step": 29770
+    },
+    {
+      "epoch": 110.3,
+      "grad_norm": 0.9548264145851135,
+      "learning_rate": 0.0008970370370370371,
+      "loss": 0.9968,
+      "step": 29780
+    },
+    {
+      "epoch": 110.33,
+      "grad_norm": 0.9424086213111877,
+      "learning_rate": 0.0008966666666666666,
+      "loss": 0.9978,
+      "step": 29790
+    },
+    {
+      "epoch": 110.37,
+      "grad_norm": 0.8982470035552979,
+      "learning_rate": 0.0008962962962962963,
+      "loss": 1.0418,
+      "step": 29800
+    },
+    {
+      "epoch": 110.41,
+      "grad_norm": 0.9713009595870972,
+      "learning_rate": 0.0008959259259259259,
+      "loss": 1.051,
+      "step": 29810
+    },
+    {
+      "epoch": 110.44,
+      "grad_norm": 0.9629430174827576,
+      "learning_rate": 0.0008955555555555556,
+      "loss": 1.0485,
+      "step": 29820
+    },
+    {
+      "epoch": 110.48,
+      "grad_norm": 1.0035597085952759,
+      "learning_rate": 0.0008951851851851852,
+      "loss": 1.0479,
+      "step": 29830
+    },
+    {
+      "epoch": 110.52,
+      "grad_norm": 1.0095973014831543,
+      "learning_rate": 0.0008948148148148149,
+      "loss": 1.074,
+      "step": 29840
+    },
+    {
+      "epoch": 110.56,
+      "grad_norm": 1.0107605457305908,
+      "learning_rate": 0.0008944444444444445,
+      "loss": 1.0728,
+      "step": 29850
+    },
+    {
+      "epoch": 110.59,
+      "grad_norm": 0.9929389357566833,
+      "learning_rate": 0.000894074074074074,
+      "loss": 1.0744,
+      "step": 29860
+    },
+    {
+      "epoch": 110.63,
+      "grad_norm": 0.9642514586448669,
+      "learning_rate": 0.0008937037037037037,
+      "loss": 1.1,
+      "step": 29870
+    },
+    {
+      "epoch": 110.67,
+      "grad_norm": 0.9914767742156982,
+      "learning_rate": 0.0008933333333333333,
+      "loss": 1.0979,
+      "step": 29880
+    },
+    {
+      "epoch": 110.7,
+      "grad_norm": 0.9608526825904846,
+      "learning_rate": 0.000892962962962963,
+      "loss": 1.104,
+      "step": 29890
+    },
+    {
+      "epoch": 110.74,
+      "grad_norm": 0.9871772527694702,
+      "learning_rate": 0.0008925925925925926,
+      "loss": 1.1148,
+      "step": 29900
+    },
+    {
+      "epoch": 110.78,
+      "grad_norm": 0.9967049956321716,
+      "learning_rate": 0.0008922222222222223,
+      "loss": 1.1284,
+      "step": 29910
+    },
+    {
+      "epoch": 110.81,
+      "grad_norm": 1.0077149868011475,
+      "learning_rate": 0.0008918518518518519,
+      "loss": 1.1268,
+      "step": 29920
+    },
+    {
+      "epoch": 110.85,
+      "grad_norm": 1.011237382888794,
+      "learning_rate": 0.0008914814814814815,
+      "loss": 1.1051,
+      "step": 29930
+    },
+    {
+      "epoch": 110.89,
+      "grad_norm": 0.973310112953186,
+      "learning_rate": 0.0008911111111111111,
+      "loss": 1.1182,
+      "step": 29940
+    },
+    {
+      "epoch": 110.93,
+      "grad_norm": 0.9287843108177185,
+      "learning_rate": 0.0008907407407407408,
+      "loss": 1.1358,
+      "step": 29950
+    },
+    {
+      "epoch": 110.96,
+      "grad_norm": 0.9778017997741699,
+      "learning_rate": 0.0008903703703703704,
+      "loss": 1.1474,
+      "step": 29960
+    },
+    {
+      "epoch": 111.0,
+      "grad_norm": 1.8783273696899414,
+      "learning_rate": 0.0008900000000000001,
+      "loss": 1.127,
+      "step": 29970
+    },
+    {
+      "epoch": 111.04,
+      "grad_norm": 0.9425782561302185,
+      "learning_rate": 0.0008896296296296296,
+      "loss": 0.9075,
+      "step": 29980
+    },
+    {
+      "epoch": 111.07,
+      "grad_norm": 0.9134290814399719,
+      "learning_rate": 0.0008892592592592593,
+      "loss": 0.934,
+      "step": 29990
+    },
+    {
+      "epoch": 111.11,
+      "grad_norm": 0.9246062636375427,
+      "learning_rate": 0.0008888888888888888,
+      "loss": 0.9395,
+      "step": 30000
+    },
+    {
+      "epoch": 111.15,
+      "grad_norm": 0.9228861331939697,
+      "learning_rate": 0.0008885185185185185,
+      "loss": 0.9561,
+      "step": 30010
+    },
+    {
+      "epoch": 111.19,
+      "grad_norm": 0.9056755304336548,
+      "learning_rate": 0.0008881481481481481,
+      "loss": 0.9332,
+      "step": 30020
+    },
+    {
+      "epoch": 111.22,
+      "grad_norm": 0.8870821595191956,
+      "learning_rate": 0.0008877777777777778,
+      "loss": 0.9757,
+      "step": 30030
+    },
+    {
+      "epoch": 111.26,
+      "grad_norm": 0.989964485168457,
+      "learning_rate": 0.0008874074074074074,
+      "loss": 0.9914,
+      "step": 30040
+    },
+    {
+      "epoch": 111.3,
+      "grad_norm": 0.9611079692840576,
+      "learning_rate": 0.0008870370370370371,
+      "loss": 0.9961,
+      "step": 30050
+    },
+    {
+      "epoch": 111.33,
+      "grad_norm": 0.9657167792320251,
+      "learning_rate": 0.0008866666666666667,
+      "loss": 0.9881,
+      "step": 30060
+    },
+    {
+      "epoch": 111.37,
+      "grad_norm": 0.9873797297477722,
+      "learning_rate": 0.0008862962962962963,
+      "loss": 1.0194,
+      "step": 30070
+    },
+    {
+      "epoch": 111.41,
+      "grad_norm": 0.9685762524604797,
+      "learning_rate": 0.000885925925925926,
+      "loss": 1.0231,
+      "step": 30080
+    },
+    {
+      "epoch": 111.44,
+      "grad_norm": 0.9828411340713501,
+      "learning_rate": 0.0008855555555555556,
+      "loss": 1.0538,
+      "step": 30090
+    },
+    {
+      "epoch": 111.48,
+      "grad_norm": 0.9536802172660828,
+      "learning_rate": 0.0008851851851851853,
+      "loss": 1.0479,
+      "step": 30100
+    },
+    {
+      "epoch": 111.52,
+      "grad_norm": 0.8964811563491821,
+      "learning_rate": 0.0008848148148148148,
+      "loss": 1.0633,
+      "step": 30110
+    },
+    {
+      "epoch": 111.56,
+      "grad_norm": 0.9257550239562988,
+      "learning_rate": 0.0008844444444444445,
+      "loss": 1.056,
+      "step": 30120
+    },
+    {
+      "epoch": 111.59,
+      "grad_norm": 0.9024585485458374,
+      "learning_rate": 0.0008840740740740741,
+      "loss": 1.0587,
+      "step": 30130
+    },
+    {
+      "epoch": 111.63,
+      "grad_norm": 1.0489718914031982,
+      "learning_rate": 0.0008837037037037037,
+      "loss": 1.0819,
+      "step": 30140
+    },
+    {
+      "epoch": 111.67,
+      "grad_norm": 0.9898716807365417,
+      "learning_rate": 0.0008833333333333333,
+      "loss": 1.0785,
+      "step": 30150
+    },
+    {
+      "epoch": 111.7,
+      "grad_norm": 0.9546767473220825,
+      "learning_rate": 0.000882962962962963,
+      "loss": 1.1157,
+      "step": 30160
+    },
+    {
+      "epoch": 111.74,
+      "grad_norm": 0.9728699922561646,
+      "learning_rate": 0.0008825925925925926,
+      "loss": 1.0798,
+      "step": 30170
+    },
+    {
+      "epoch": 111.78,
+      "grad_norm": 0.9877276420593262,
+      "learning_rate": 0.0008822222222222223,
+      "loss": 1.1088,
+      "step": 30180
+    },
+    {
+      "epoch": 111.81,
+      "grad_norm": 0.9211771488189697,
+      "learning_rate": 0.0008818518518518519,
+      "loss": 1.117,
+      "step": 30190
+    },
+    {
+      "epoch": 111.85,
+      "grad_norm": 1.0157147645950317,
+      "learning_rate": 0.0008814814814814816,
+      "loss": 1.1102,
+      "step": 30200
+    },
+    {
+      "epoch": 111.89,
+      "grad_norm": 1.092393398284912,
+      "learning_rate": 0.000881111111111111,
+      "loss": 1.1293,
+      "step": 30210
+    },
+    {
+      "epoch": 111.93,
+      "grad_norm": 0.9899991750717163,
+      "learning_rate": 0.0008807407407407407,
+      "loss": 1.1227,
+      "step": 30220
+    },
+    {
+      "epoch": 111.96,
+      "grad_norm": 0.9688059091567993,
+      "learning_rate": 0.0008803703703703703,
+      "loss": 1.1286,
+      "step": 30230
+    },
+    {
+      "epoch": 112.0,
+      "grad_norm": 1.6932052373886108,
+      "learning_rate": 0.00088,
+      "loss": 1.1382,
+      "step": 30240
+    },
+    {
+      "epoch": 112.04,
+      "grad_norm": 0.9143195748329163,
+      "learning_rate": 0.0008796296296296296,
+      "loss": 0.9058,
+      "step": 30250
+    },
+    {
+      "epoch": 112.07,
+      "grad_norm": 0.990135908126831,
+      "learning_rate": 0.0008792592592592593,
+      "loss": 0.9267,
+      "step": 30260
+    },
+    {
+      "epoch": 112.11,
+      "grad_norm": 0.9176962375640869,
+      "learning_rate": 0.000878888888888889,
+      "loss": 0.917,
+      "step": 30270
+    },
+    {
+      "epoch": 112.15,
+      "grad_norm": 0.9764485359191895,
+      "learning_rate": 0.0008785185185185185,
+      "loss": 0.9068,
+      "step": 30280
+    },
+    {
+      "epoch": 112.19,
+      "grad_norm": 0.9197412133216858,
+      "learning_rate": 0.0008781481481481482,
+      "loss": 0.9584,
+      "step": 30290
+    },
+    {
+      "epoch": 112.22,
+      "grad_norm": 1.0627151727676392,
+      "learning_rate": 0.0008777777777777778,
+      "loss": 0.9894,
+      "step": 30300
+    },
+    {
+      "epoch": 112.26,
+      "grad_norm": 0.9765280485153198,
+      "learning_rate": 0.0008774074074074075,
+      "loss": 0.9884,
+      "step": 30310
+    },
+    {
+      "epoch": 112.3,
+      "grad_norm": 0.9730259776115417,
+      "learning_rate": 0.0008770370370370371,
+      "loss": 0.9733,
+      "step": 30320
+    },
+    {
+      "epoch": 112.33,
+      "grad_norm": 0.9327222108840942,
+      "learning_rate": 0.0008766666666666668,
+      "loss": 1.016,
+      "step": 30330
+    },
+    {
+      "epoch": 112.37,
+      "grad_norm": 0.9560900330543518,
+      "learning_rate": 0.0008762962962962964,
+      "loss": 1.0147,
+      "step": 30340
+    },
+    {
+      "epoch": 112.41,
+      "grad_norm": 0.9207103252410889,
+      "learning_rate": 0.0008759259259259259,
+      "loss": 1.003,
+      "step": 30350
+    },
+    {
+      "epoch": 112.44,
+      "grad_norm": 1.0215859413146973,
+      "learning_rate": 0.0008755555555555555,
+      "loss": 1.0212,
+      "step": 30360
+    },
+    {
+      "epoch": 112.48,
+      "grad_norm": 0.9427470564842224,
+      "learning_rate": 0.0008751851851851852,
+      "loss": 1.0495,
+      "step": 30370
+    },
+    {
+      "epoch": 112.52,
+      "grad_norm": 1.003307819366455,
+      "learning_rate": 0.0008748148148148148,
+      "loss": 1.0227,
+      "step": 30380
+    },
+    {
+      "epoch": 112.56,
+      "grad_norm": 1.093425989151001,
+      "learning_rate": 0.0008744444444444445,
+      "loss": 1.0566,
+      "step": 30390
+    },
+    {
+      "epoch": 112.59,
+      "grad_norm": 0.9261013269424438,
+      "learning_rate": 0.0008740740740740741,
+      "loss": 1.0736,
+      "step": 30400
+    },
+    {
+      "epoch": 112.63,
+      "grad_norm": 0.9942784309387207,
+      "learning_rate": 0.0008737037037037037,
+      "loss": 1.0688,
+      "step": 30410
+    },
+    {
+      "epoch": 112.67,
+      "grad_norm": 0.9616612792015076,
+      "learning_rate": 0.0008733333333333333,
+      "loss": 1.0768,
+      "step": 30420
+    },
+    {
+      "epoch": 112.7,
+      "grad_norm": 0.9479864239692688,
+      "learning_rate": 0.000872962962962963,
+      "loss": 1.1053,
+      "step": 30430
+    },
+    {
+      "epoch": 112.74,
+      "grad_norm": 1.0143781900405884,
+      "learning_rate": 0.0008725925925925926,
+      "loss": 1.0862,
+      "step": 30440
+    },
+    {
+      "epoch": 112.78,
+      "grad_norm": 1.0063732862472534,
+      "learning_rate": 0.0008722222222222223,
+      "loss": 1.114,
+      "step": 30450
+    },
+    {
+      "epoch": 112.81,
+      "grad_norm": 0.9845717549324036,
+      "learning_rate": 0.0008718518518518518,
+      "loss": 1.079,
+      "step": 30460
+    },
+    {
+      "epoch": 112.85,
+      "grad_norm": 0.9684877991676331,
+      "learning_rate": 0.0008714814814814815,
+      "loss": 1.1111,
+      "step": 30470
+    },
+    {
+      "epoch": 112.89,
+      "grad_norm": 0.9328610301017761,
+      "learning_rate": 0.000871111111111111,
+      "loss": 1.1305,
+      "step": 30480
+    },
+    {
+      "epoch": 112.93,
+      "grad_norm": 0.9969044327735901,
+      "learning_rate": 0.0008707407407407407,
+      "loss": 1.1102,
+      "step": 30490
+    },
+    {
+      "epoch": 112.96,
+      "grad_norm": 1.059865117073059,
+      "learning_rate": 0.0008703703703703704,
+      "loss": 1.1202,
+      "step": 30500
+    },
+    {
+      "epoch": 113.0,
+      "grad_norm": 2.0708632469177246,
+      "learning_rate": 0.00087,
+      "loss": 1.1286,
+      "step": 30510
+    },
+    {
+      "epoch": 113.04,
+      "grad_norm": 0.9408242702484131,
+      "learning_rate": 0.0008696296296296297,
+      "loss": 0.8969,
+      "step": 30520
+    },
+    {
+      "epoch": 113.07,
+      "grad_norm": 0.9614538550376892,
+      "learning_rate": 0.0008692592592592593,
+      "loss": 0.9021,
+      "step": 30530
+    },
+    {
+      "epoch": 113.11,
+      "grad_norm": 0.9190179109573364,
+      "learning_rate": 0.000868888888888889,
+      "loss": 0.9108,
+      "step": 30540
+    },
+    {
+      "epoch": 113.15,
+      "grad_norm": 0.9450747966766357,
+      "learning_rate": 0.0008685185185185185,
+      "loss": 0.9466,
+      "step": 30550
+    },
+    {
+      "epoch": 113.19,
+      "grad_norm": 0.9533342123031616,
+      "learning_rate": 0.0008681481481481482,
+      "loss": 0.9656,
+      "step": 30560
+    },
+    {
+      "epoch": 113.22,
+      "grad_norm": 0.9292343258857727,
+      "learning_rate": 0.0008677777777777778,
+      "loss": 0.9565,
+      "step": 30570
+    },
+    {
+      "epoch": 113.26,
+      "grad_norm": 0.9235038757324219,
+      "learning_rate": 0.0008674074074074074,
+      "loss": 0.9754,
+      "step": 30580
+    },
+    {
+      "epoch": 113.3,
+      "grad_norm": 1.0145610570907593,
+      "learning_rate": 0.000867037037037037,
+      "loss": 0.9838,
+      "step": 30590
+    },
+    {
+      "epoch": 113.33,
+      "grad_norm": 0.9950644373893738,
+      "learning_rate": 0.0008666666666666667,
+      "loss": 0.9909,
+      "step": 30600
+    },
+    {
+      "epoch": 113.37,
+      "grad_norm": 1.0521272420883179,
+      "learning_rate": 0.0008662962962962963,
+      "loss": 0.9892,
+      "step": 30610
+    },
+    {
+      "epoch": 113.41,
+      "grad_norm": 0.9194403886795044,
+      "learning_rate": 0.0008659259259259259,
+      "loss": 1.0081,
+      "step": 30620
+    },
+    {
+      "epoch": 113.44,
+      "grad_norm": 0.9850278496742249,
+      "learning_rate": 0.0008655555555555555,
+      "loss": 1.0406,
+      "step": 30630
+    },
+    {
+      "epoch": 113.48,
+      "grad_norm": 0.971551239490509,
+      "learning_rate": 0.0008651851851851852,
+      "loss": 1.0378,
+      "step": 30640
+    },
+    {
+      "epoch": 113.52,
+      "grad_norm": 0.9850895404815674,
+      "learning_rate": 0.0008648148148148148,
+      "loss": 1.0316,
+      "step": 30650
+    },
+    {
+      "epoch": 113.56,
+      "grad_norm": 0.9480496048927307,
+      "learning_rate": 0.0008644444444444445,
+      "loss": 1.0412,
+      "step": 30660
+    },
+    {
+      "epoch": 113.59,
+      "grad_norm": 1.0179239511489868,
+      "learning_rate": 0.0008640740740740741,
+      "loss": 1.0276,
+      "step": 30670
+    },
+    {
+      "epoch": 113.63,
+      "grad_norm": 0.9929103255271912,
+      "learning_rate": 0.0008637037037037038,
+      "loss": 1.0659,
+      "step": 30680
+    },
+    {
+      "epoch": 113.67,
+      "grad_norm": 1.006871223449707,
+      "learning_rate": 0.0008633333333333334,
+      "loss": 1.0808,
+      "step": 30690
+    },
+    {
+      "epoch": 113.7,
+      "grad_norm": 0.990407407283783,
+      "learning_rate": 0.0008629629629629629,
+      "loss": 1.0747,
+      "step": 30700
+    },
+    {
+      "epoch": 113.74,
+      "grad_norm": 1.0033669471740723,
+      "learning_rate": 0.0008625925925925926,
+      "loss": 1.0571,
+      "step": 30710
+    },
+    {
+      "epoch": 113.78,
+      "grad_norm": 0.9885395169258118,
+      "learning_rate": 0.0008622222222222222,
+      "loss": 1.0932,
+      "step": 30720
+    },
+    {
+      "epoch": 113.81,
+      "grad_norm": 0.9462271928787231,
+      "learning_rate": 0.0008618518518518519,
+      "loss": 1.0848,
+      "step": 30730
+    },
+    {
+      "epoch": 113.85,
+      "grad_norm": 0.9825132489204407,
+      "learning_rate": 0.0008614814814814815,
+      "loss": 1.1212,
+      "step": 30740
+    },
+    {
+      "epoch": 113.89,
+      "grad_norm": 0.9224004149436951,
+      "learning_rate": 0.0008611111111111112,
+      "loss": 1.1188,
+      "step": 30750
+    },
+    {
+      "epoch": 113.93,
+      "grad_norm": 1.0247914791107178,
+      "learning_rate": 0.0008607407407407407,
+      "loss": 1.1017,
+      "step": 30760
+    },
+    {
+      "epoch": 113.96,
+      "grad_norm": 1.016513705253601,
+      "learning_rate": 0.0008603703703703704,
+      "loss": 1.1242,
+      "step": 30770
+    },
+    {
+      "epoch": 114.0,
+      "grad_norm": 1.9813746213912964,
+      "learning_rate": 0.00086,
+      "loss": 1.1129,
+      "step": 30780
+    },
+    {
+      "epoch": 114.04,
+      "grad_norm": 0.9882152676582336,
+      "learning_rate": 0.0008596296296296297,
+      "loss": 0.9152,
+      "step": 30790
+    },
+    {
+      "epoch": 114.07,
+      "grad_norm": 0.9574514031410217,
+      "learning_rate": 0.0008592592592592593,
+      "loss": 0.9013,
+      "step": 30800
+    },
+    {
+      "epoch": 114.11,
+      "grad_norm": 0.9351339936256409,
+      "learning_rate": 0.000858888888888889,
+      "loss": 0.9075,
+      "step": 30810
+    },
+    {
+      "epoch": 114.15,
+      "grad_norm": 0.9363587498664856,
+      "learning_rate": 0.0008585185185185185,
+      "loss": 0.9144,
+      "step": 30820
+    },
+    {
+      "epoch": 114.19,
+      "grad_norm": 0.9973554015159607,
+      "learning_rate": 0.0008581481481481481,
+      "loss": 0.9334,
+      "step": 30830
+    },
+    {
+      "epoch": 114.22,
+      "grad_norm": 1.1108694076538086,
+      "learning_rate": 0.0008577777777777777,
+      "loss": 0.9683,
+      "step": 30840
+    },
+    {
+      "epoch": 114.26,
+      "grad_norm": 1.0352013111114502,
+      "learning_rate": 0.0008574074074074074,
+      "loss": 0.9817,
+      "step": 30850
+    },
+    {
+      "epoch": 114.3,
+      "grad_norm": 0.9350374341011047,
+      "learning_rate": 0.000857037037037037,
+      "loss": 0.9652,
+      "step": 30860
+    },
+    {
+      "epoch": 114.33,
+      "grad_norm": 0.942579984664917,
+      "learning_rate": 0.0008566666666666667,
+      "loss": 0.9818,
+      "step": 30870
+    },
+    {
+      "epoch": 114.37,
+      "grad_norm": 1.0011614561080933,
+      "learning_rate": 0.0008562962962962963,
+      "loss": 0.9991,
+      "step": 30880
+    },
+    {
+      "epoch": 114.41,
+      "grad_norm": 0.9837273955345154,
+      "learning_rate": 0.000855925925925926,
+      "loss": 0.9953,
+      "step": 30890
+    },
+    {
+      "epoch": 114.44,
+      "grad_norm": 0.9239174127578735,
+      "learning_rate": 0.0008555555555555556,
+      "loss": 1.0251,
+      "step": 30900
+    },
+    {
+      "epoch": 114.48,
+      "grad_norm": 0.9788716435432434,
+      "learning_rate": 0.0008551851851851852,
+      "loss": 1.011,
+      "step": 30910
+    },
+    {
+      "epoch": 114.52,
+      "grad_norm": 1.0957821607589722,
+      "learning_rate": 0.0008548148148148149,
+      "loss": 1.0122,
+      "step": 30920
+    },
+    {
+      "epoch": 114.56,
+      "grad_norm": 1.0088284015655518,
+      "learning_rate": 0.0008544444444444445,
+      "loss": 1.0364,
+      "step": 30930
+    },
+    {
+      "epoch": 114.59,
+      "grad_norm": 1.0810710191726685,
+      "learning_rate": 0.0008540740740740742,
+      "loss": 1.0542,
+      "step": 30940
+    },
+    {
+      "epoch": 114.63,
+      "grad_norm": 1.0543538331985474,
+      "learning_rate": 0.0008537037037037037,
+      "loss": 1.0374,
+      "step": 30950
+    },
+    {
+      "epoch": 114.67,
+      "grad_norm": 1.0714164972305298,
+      "learning_rate": 0.0008533333333333334,
+      "loss": 1.0632,
+      "step": 30960
+    },
+    {
+      "epoch": 114.7,
+      "grad_norm": 0.9550819993019104,
+      "learning_rate": 0.0008529629629629629,
+      "loss": 1.0696,
+      "step": 30970
+    },
+    {
+      "epoch": 114.74,
+      "grad_norm": 1.0119558572769165,
+      "learning_rate": 0.0008525925925925926,
+      "loss": 1.0603,
+      "step": 30980
+    },
+    {
+      "epoch": 114.78,
+      "grad_norm": 0.9348704814910889,
+      "learning_rate": 0.0008522222222222222,
+      "loss": 1.1039,
+      "step": 30990
+    },
+    {
+      "epoch": 114.81,
+      "grad_norm": 1.025181770324707,
+      "learning_rate": 0.0008518518518518519,
+      "loss": 1.0769,
+      "step": 31000
+    },
+    {
+      "epoch": 114.85,
+      "grad_norm": 1.011375904083252,
+      "learning_rate": 0.0008514814814814815,
+      "loss": 1.077,
+      "step": 31010
+    },
+    {
+      "epoch": 114.89,
+      "grad_norm": 0.9699951410293579,
+      "learning_rate": 0.0008511111111111112,
+      "loss": 1.1129,
+      "step": 31020
+    },
+    {
+      "epoch": 114.93,
+      "grad_norm": 0.9796174168586731,
+      "learning_rate": 0.0008507407407407407,
+      "loss": 1.1002,
+      "step": 31030
+    },
+    {
+      "epoch": 114.96,
+      "grad_norm": 0.9660859704017639,
+      "learning_rate": 0.0008503703703703704,
+      "loss": 1.0966,
+      "step": 31040
+    },
+    {
+      "epoch": 115.0,
+      "grad_norm": 1.6068373918533325,
+      "learning_rate": 0.00085,
+      "loss": 1.0997,
+      "step": 31050
+    },
+    {
+      "epoch": 115.04,
+      "grad_norm": 0.9463887214660645,
+      "learning_rate": 0.0008496296296296296,
+      "loss": 0.8809,
+      "step": 31060
+    },
+    {
+      "epoch": 115.07,
+      "grad_norm": 0.9858732223510742,
+      "learning_rate": 0.0008492592592592592,
+      "loss": 0.8918,
+      "step": 31070
+    },
+    {
+      "epoch": 115.11,
+      "grad_norm": 1.081920862197876,
+      "learning_rate": 0.0008488888888888889,
+      "loss": 0.9019,
+      "step": 31080
+    },
+    {
+      "epoch": 115.15,
+      "grad_norm": 0.914100170135498,
+      "learning_rate": 0.0008485185185185186,
+      "loss": 0.9338,
+      "step": 31090
+    },
+    {
+      "epoch": 115.19,
+      "grad_norm": 0.9406884908676147,
+      "learning_rate": 0.0008481481481481481,
+      "loss": 0.924,
+      "step": 31100
+    },
+    {
+      "epoch": 115.22,
+      "grad_norm": 1.0523425340652466,
+      "learning_rate": 0.0008477777777777778,
+      "loss": 0.9472,
+      "step": 31110
+    },
+    {
+      "epoch": 115.26,
+      "grad_norm": 1.0571317672729492,
+      "learning_rate": 0.0008474074074074074,
+      "loss": 0.9532,
+      "step": 31120
+    },
+    {
+      "epoch": 115.3,
+      "grad_norm": 1.1222143173217773,
+      "learning_rate": 0.0008470370370370371,
+      "loss": 0.973,
+      "step": 31130
+    },
+    {
+      "epoch": 115.33,
+      "grad_norm": 1.019164800643921,
+      "learning_rate": 0.0008466666666666667,
+      "loss": 0.9732,
+      "step": 31140
+    },
+    {
+      "epoch": 115.37,
+      "grad_norm": 0.9789102077484131,
+      "learning_rate": 0.0008462962962962964,
+      "loss": 0.983,
+      "step": 31150
+    },
+    {
+      "epoch": 115.41,
+      "grad_norm": 1.0151702165603638,
+      "learning_rate": 0.000845925925925926,
+      "loss": 1.0162,
+      "step": 31160
+    },
+    {
+      "epoch": 115.44,
+      "grad_norm": 1.0261796712875366,
+      "learning_rate": 0.0008455555555555556,
+      "loss": 0.9956,
+      "step": 31170
+    },
+    {
+      "epoch": 115.48,
+      "grad_norm": 1.0111148357391357,
+      "learning_rate": 0.0008451851851851851,
+      "loss": 1.0375,
+      "step": 31180
+    },
+    {
+      "epoch": 115.52,
+      "grad_norm": 0.9380441904067993,
+      "learning_rate": 0.0008448148148148148,
+      "loss": 1.0244,
+      "step": 31190
+    },
+    {
+      "epoch": 115.56,
+      "grad_norm": 0.9684368371963501,
+      "learning_rate": 0.0008444444444444444,
+      "loss": 1.0179,
+      "step": 31200
+    },
+    {
+      "epoch": 115.59,
+      "grad_norm": 1.1318254470825195,
+      "learning_rate": 0.0008440740740740741,
+      "loss": 1.0402,
+      "step": 31210
+    },
+    {
+      "epoch": 115.63,
+      "grad_norm": 0.9918487071990967,
+      "learning_rate": 0.0008437037037037037,
+      "loss": 1.0473,
+      "step": 31220
+    },
+    {
+      "epoch": 115.67,
+      "grad_norm": 1.001694917678833,
+      "learning_rate": 0.0008433333333333334,
+      "loss": 1.0424,
+      "step": 31230
+    },
+    {
+      "epoch": 115.7,
+      "grad_norm": 0.9491502046585083,
+      "learning_rate": 0.0008429629629629629,
+      "loss": 1.0667,
+      "step": 31240
+    },
+    {
+      "epoch": 115.74,
+      "grad_norm": 1.0236819982528687,
+      "learning_rate": 0.0008425925925925926,
+      "loss": 1.0565,
+      "step": 31250
+    },
+    {
+      "epoch": 115.78,
+      "grad_norm": 1.0704491138458252,
+      "learning_rate": 0.0008422222222222222,
+      "loss": 1.0906,
+      "step": 31260
+    },
+    {
+      "epoch": 115.81,
+      "grad_norm": 1.1185559034347534,
+      "learning_rate": 0.0008418518518518519,
+      "loss": 1.0476,
+      "step": 31270
+    },
+    {
+      "epoch": 115.85,
+      "grad_norm": 1.0499783754348755,
+      "learning_rate": 0.0008414814814814815,
+      "loss": 1.0939,
+      "step": 31280
+    },
+    {
+      "epoch": 115.89,
+      "grad_norm": 0.9921557307243347,
+      "learning_rate": 0.0008411111111111112,
+      "loss": 1.0777,
+      "step": 31290
+    },
+    {
+      "epoch": 115.93,
+      "grad_norm": 1.0125623941421509,
+      "learning_rate": 0.0008407407407407409,
+      "loss": 1.0939,
+      "step": 31300
+    },
+    {
+      "epoch": 115.96,
+      "grad_norm": 0.985906720161438,
+      "learning_rate": 0.0008403703703703703,
+      "loss": 1.0973,
+      "step": 31310
+    },
+    {
+      "epoch": 116.0,
+      "grad_norm": 2.2523534297943115,
+      "learning_rate": 0.00084,
+      "loss": 1.0789,
+      "step": 31320
+    },
+    {
+      "epoch": 116.04,
+      "grad_norm": 0.9431299567222595,
+      "learning_rate": 0.0008396296296296296,
+      "loss": 0.8982,
+      "step": 31330
+    },
+    {
+      "epoch": 116.07,
+      "grad_norm": 0.9881550669670105,
+      "learning_rate": 0.0008392592592592593,
+      "loss": 0.8887,
+      "step": 31340
+    },
+    {
+      "epoch": 116.11,
+      "grad_norm": 0.985056459903717,
+      "learning_rate": 0.0008388888888888889,
+      "loss": 0.8935,
+      "step": 31350
+    },
+    {
+      "epoch": 116.15,
+      "grad_norm": 0.9958218336105347,
+      "learning_rate": 0.0008385185185185186,
+      "loss": 0.9165,
+      "step": 31360
+    },
+    {
+      "epoch": 116.19,
+      "grad_norm": 0.9456029534339905,
+      "learning_rate": 0.0008381481481481482,
+      "loss": 0.922,
+      "step": 31370
+    },
+    {
+      "epoch": 116.22,
+      "grad_norm": 1.0164287090301514,
+      "learning_rate": 0.0008377777777777778,
+      "loss": 0.9428,
+      "step": 31380
+    },
+    {
+      "epoch": 116.26,
+      "grad_norm": 0.9778671264648438,
+      "learning_rate": 0.0008374074074074074,
+      "loss": 0.9509,
+      "step": 31390
+    },
+    {
+      "epoch": 116.3,
+      "grad_norm": 1.0401966571807861,
+      "learning_rate": 0.0008370370370370371,
+      "loss": 0.9524,
+      "step": 31400
+    },
+    {
+      "epoch": 116.33,
+      "grad_norm": 0.9463452100753784,
+      "learning_rate": 0.0008366666666666667,
+      "loss": 0.9865,
+      "step": 31410
+    },
+    {
+      "epoch": 116.37,
+      "grad_norm": 1.0005884170532227,
+      "learning_rate": 0.0008362962962962964,
+      "loss": 0.9834,
+      "step": 31420
+    },
+    {
+      "epoch": 116.41,
+      "grad_norm": 1.0348414182662964,
+      "learning_rate": 0.0008359259259259259,
+      "loss": 0.9834,
+      "step": 31430
+    },
+    {
+      "epoch": 116.44,
+      "grad_norm": 1.0169605016708374,
+      "learning_rate": 0.0008355555555555556,
+      "loss": 0.9884,
+      "step": 31440
+    },
+    {
+      "epoch": 116.48,
+      "grad_norm": 1.0272061824798584,
+      "learning_rate": 0.0008351851851851851,
+      "loss": 1.0093,
+      "step": 31450
+    },
+    {
+      "epoch": 116.52,
+      "grad_norm": 1.0215003490447998,
+      "learning_rate": 0.0008348148148148148,
+      "loss": 1.0236,
+      "step": 31460
+    },
+    {
+      "epoch": 116.56,
+      "grad_norm": 0.9531969428062439,
+      "learning_rate": 0.0008344444444444444,
+      "loss": 1.0269,
+      "step": 31470
+    },
+    {
+      "epoch": 116.59,
+      "grad_norm": 0.9995023012161255,
+      "learning_rate": 0.0008340740740740741,
+      "loss": 1.0146,
+      "step": 31480
+    },
+    {
+      "epoch": 116.63,
+      "grad_norm": 0.9878466129302979,
+      "learning_rate": 0.0008337037037037037,
+      "loss": 1.0294,
+      "step": 31490
+    },
+    {
+      "epoch": 116.67,
+      "grad_norm": 0.9585810303688049,
+      "learning_rate": 0.0008333333333333334,
+      "loss": 1.0438,
+      "step": 31500
+    },
+    {
+      "epoch": 116.7,
+      "grad_norm": 1.0456923246383667,
+      "learning_rate": 0.0008329629629629631,
+      "loss": 1.0627,
+      "step": 31510
+    },
+    {
+      "epoch": 116.74,
+      "grad_norm": 0.9957906603813171,
+      "learning_rate": 0.0008325925925925926,
+      "loss": 1.0584,
+      "step": 31520
+    },
+    {
+      "epoch": 116.78,
+      "grad_norm": 1.0360143184661865,
+      "learning_rate": 0.0008322222222222223,
+      "loss": 1.0573,
+      "step": 31530
+    },
+    {
+      "epoch": 116.81,
+      "grad_norm": 0.935924768447876,
+      "learning_rate": 0.0008318518518518518,
+      "loss": 1.0462,
+      "step": 31540
+    },
+    {
+      "epoch": 116.85,
+      "grad_norm": 1.0759156942367554,
+      "learning_rate": 0.0008314814814814815,
+      "loss": 1.0702,
+      "step": 31550
+    },
+    {
+      "epoch": 116.89,
+      "grad_norm": 0.9487184286117554,
+      "learning_rate": 0.0008311111111111111,
+      "loss": 1.0916,
+      "step": 31560
+    },
+    {
+      "epoch": 116.93,
+      "grad_norm": 0.9857691526412964,
+      "learning_rate": 0.0008307407407407408,
+      "loss": 1.0904,
+      "step": 31570
+    },
+    {
+      "epoch": 116.96,
+      "grad_norm": 0.9965876936912537,
+      "learning_rate": 0.0008303703703703704,
+      "loss": 1.0766,
+      "step": 31580
+    },
+    {
+      "epoch": 117.0,
+      "grad_norm": 1.8307380676269531,
+      "learning_rate": 0.00083,
+      "loss": 1.092,
+      "step": 31590
+    },
+    {
+      "epoch": 117.04,
+      "grad_norm": 0.9478293061256409,
+      "learning_rate": 0.0008296296296296296,
+      "loss": 0.8833,
+      "step": 31600
+    },
+    {
+      "epoch": 117.07,
+      "grad_norm": 0.9572622776031494,
+      "learning_rate": 0.0008292592592592593,
+      "loss": 0.869,
+      "step": 31610
+    },
+    {
+      "epoch": 117.11,
+      "grad_norm": 0.9900893568992615,
+      "learning_rate": 0.0008288888888888889,
+      "loss": 0.9014,
+      "step": 31620
+    },
+    {
+      "epoch": 117.15,
+      "grad_norm": 0.9924251437187195,
+      "learning_rate": 0.0008285185185185186,
+      "loss": 0.9139,
+      "step": 31630
+    },
+    {
+      "epoch": 117.19,
+      "grad_norm": 1.012449026107788,
+      "learning_rate": 0.0008281481481481482,
+      "loss": 0.9283,
+      "step": 31640
+    },
+    {
+      "epoch": 117.22,
+      "grad_norm": 0.9699010252952576,
+      "learning_rate": 0.0008277777777777778,
+      "loss": 0.9421,
+      "step": 31650
+    },
+    {
+      "epoch": 117.26,
+      "grad_norm": 0.9940907955169678,
+      "learning_rate": 0.0008274074074074073,
+      "loss": 0.9248,
+      "step": 31660
+    },
+    {
+      "epoch": 117.3,
+      "grad_norm": 1.0201570987701416,
+      "learning_rate": 0.000827037037037037,
+      "loss": 0.9485,
+      "step": 31670
+    },
+    {
+      "epoch": 117.33,
+      "grad_norm": 0.9949468970298767,
+      "learning_rate": 0.0008266666666666666,
+      "loss": 0.9616,
+      "step": 31680
+    },
+    {
+      "epoch": 117.37,
+      "grad_norm": 1.0526440143585205,
+      "learning_rate": 0.0008262962962962963,
+      "loss": 0.977,
+      "step": 31690
+    },
+    {
+      "epoch": 117.41,
+      "grad_norm": 1.0306730270385742,
+      "learning_rate": 0.0008259259259259259,
+      "loss": 0.9835,
+      "step": 31700
+    },
+    {
+      "epoch": 117.44,
+      "grad_norm": 1.047856330871582,
+      "learning_rate": 0.0008255555555555556,
+      "loss": 0.9882,
+      "step": 31710
+    },
+    {
+      "epoch": 117.48,
+      "grad_norm": 1.017426609992981,
+      "learning_rate": 0.0008251851851851852,
+      "loss": 1.0126,
+      "step": 31720
+    },
+    {
+      "epoch": 117.52,
+      "grad_norm": 1.0520775318145752,
+      "learning_rate": 0.0008248148148148148,
+      "loss": 0.9974,
+      "step": 31730
+    },
+    {
+      "epoch": 117.56,
+      "grad_norm": 1.0695947408676147,
+      "learning_rate": 0.0008244444444444445,
+      "loss": 1.0,
+      "step": 31740
+    },
+    {
+      "epoch": 117.59,
+      "grad_norm": 1.0761319398880005,
+      "learning_rate": 0.0008240740740740741,
+      "loss": 1.0147,
+      "step": 31750
+    },
+    {
+      "epoch": 117.63,
+      "grad_norm": 0.9545421004295349,
+      "learning_rate": 0.0008237037037037038,
+      "loss": 1.0315,
+      "step": 31760
+    },
+    {
+      "epoch": 117.67,
+      "grad_norm": 1.005906105041504,
+      "learning_rate": 0.0008233333333333334,
+      "loss": 1.0242,
+      "step": 31770
+    },
+    {
+      "epoch": 117.7,
+      "grad_norm": 0.9745824933052063,
+      "learning_rate": 0.0008229629629629631,
+      "loss": 1.0309,
+      "step": 31780
+    },
+    {
+      "epoch": 117.74,
+      "grad_norm": 1.0264394283294678,
+      "learning_rate": 0.0008225925925925925,
+      "loss": 1.0684,
+      "step": 31790
+    },
+    {
+      "epoch": 117.78,
+      "grad_norm": 1.0310300588607788,
+      "learning_rate": 0.0008222222222222222,
+      "loss": 1.0371,
+      "step": 31800
+    },
+    {
+      "epoch": 117.81,
+      "grad_norm": 1.017672061920166,
+      "learning_rate": 0.0008218518518518518,
+      "loss": 1.0531,
+      "step": 31810
+    },
+    {
+      "epoch": 117.85,
+      "grad_norm": 1.05331552028656,
+      "learning_rate": 0.0008214814814814815,
+      "loss": 1.0585,
+      "step": 31820
+    },
+    {
+      "epoch": 117.89,
+      "grad_norm": 0.970564603805542,
+      "learning_rate": 0.0008211111111111111,
+      "loss": 1.0757,
+      "step": 31830
+    },
+    {
+      "epoch": 117.93,
+      "grad_norm": 0.9827378392219543,
+      "learning_rate": 0.0008207407407407408,
+      "loss": 1.0806,
+      "step": 31840
+    },
+    {
+      "epoch": 117.96,
+      "grad_norm": 1.0662480592727661,
+      "learning_rate": 0.0008203703703703704,
+      "loss": 1.063,
+      "step": 31850
+    },
+    {
+      "epoch": 118.0,
+      "grad_norm": 1.6402631998062134,
+      "learning_rate": 0.00082,
+      "loss": 1.0978,
+      "step": 31860
+    },
+    {
+      "epoch": 118.04,
+      "grad_norm": 1.0310686826705933,
+      "learning_rate": 0.0008196296296296296,
+      "loss": 0.8643,
+      "step": 31870
+    },
+    {
+      "epoch": 118.07,
+      "grad_norm": 0.9705514907836914,
+      "learning_rate": 0.0008192592592592593,
+      "loss": 0.8804,
+      "step": 31880
+    },
+    {
+      "epoch": 118.11,
+      "grad_norm": 1.0193184614181519,
+      "learning_rate": 0.0008188888888888889,
+      "loss": 0.9038,
+      "step": 31890
+    },
+    {
+      "epoch": 118.15,
+      "grad_norm": 0.9885033965110779,
+      "learning_rate": 0.0008185185185185186,
+      "loss": 0.9113,
+      "step": 31900
+    },
+    {
+      "epoch": 118.19,
+      "grad_norm": 1.0788235664367676,
+      "learning_rate": 0.0008181481481481483,
+      "loss": 0.8971,
+      "step": 31910
+    },
+    {
+      "epoch": 118.22,
+      "grad_norm": 1.086966633796692,
+      "learning_rate": 0.0008177777777777778,
+      "loss": 0.9177,
+      "step": 31920
+    },
+    {
+      "epoch": 118.26,
+      "grad_norm": 1.0231404304504395,
+      "learning_rate": 0.0008174074074074074,
+      "loss": 0.9102,
+      "step": 31930
+    },
+    {
+      "epoch": 118.3,
+      "grad_norm": 0.9642085433006287,
+      "learning_rate": 0.000817037037037037,
+      "loss": 0.9613,
+      "step": 31940
+    },
+    {
+      "epoch": 118.33,
+      "grad_norm": 0.9916780591011047,
+      "learning_rate": 0.0008166666666666667,
+      "loss": 0.955,
+      "step": 31950
+    },
+    {
+      "epoch": 118.37,
+      "grad_norm": 1.0243855714797974,
+      "learning_rate": 0.0008162962962962963,
+      "loss": 0.9787,
+      "step": 31960
+    },
+    {
+      "epoch": 118.41,
+      "grad_norm": 1.0688339471817017,
+      "learning_rate": 0.000815925925925926,
+      "loss": 0.9504,
+      "step": 31970
+    },
+    {
+      "epoch": 118.44,
+      "grad_norm": 1.0077600479125977,
+      "learning_rate": 0.0008155555555555556,
+      "loss": 0.972,
+      "step": 31980
+    },
+    {
+      "epoch": 118.48,
+      "grad_norm": 1.018144965171814,
+      "learning_rate": 0.0008151851851851853,
+      "loss": 0.9725,
+      "step": 31990
+    },
+    {
+      "epoch": 118.52,
+      "grad_norm": 0.9333834052085876,
+      "learning_rate": 0.0008148148148148148,
+      "loss": 1.0076,
+      "step": 32000
+    },
+    {
+      "epoch": 118.56,
+      "grad_norm": 0.9722175002098083,
+      "learning_rate": 0.0008144444444444445,
+      "loss": 1.0176,
+      "step": 32010
+    },
+    {
+      "epoch": 118.59,
+      "grad_norm": 1.027160882949829,
+      "learning_rate": 0.000814074074074074,
+      "loss": 1.0272,
+      "step": 32020
+    },
+    {
+      "epoch": 118.63,
+      "grad_norm": 0.9696611762046814,
+      "learning_rate": 0.0008137037037037037,
+      "loss": 1.0268,
+      "step": 32030
+    },
+    {
+      "epoch": 118.67,
+      "grad_norm": 0.994503378868103,
+      "learning_rate": 0.0008133333333333333,
+      "loss": 1.0403,
+      "step": 32040
+    },
+    {
+      "epoch": 118.7,
+      "grad_norm": 1.0111019611358643,
+      "learning_rate": 0.000812962962962963,
+      "loss": 1.0298,
+      "step": 32050
+    },
+    {
+      "epoch": 118.74,
+      "grad_norm": 0.9856045842170715,
+      "learning_rate": 0.0008125925925925926,
+      "loss": 1.02,
+      "step": 32060
+    },
+    {
+      "epoch": 118.78,
+      "grad_norm": 1.0325449705123901,
+      "learning_rate": 0.0008122222222222222,
+      "loss": 1.0528,
+      "step": 32070
+    },
+    {
+      "epoch": 118.81,
+      "grad_norm": 1.0795844793319702,
+      "learning_rate": 0.0008118518518518518,
+      "loss": 1.0516,
+      "step": 32080
+    },
+    {
+      "epoch": 118.85,
+      "grad_norm": 0.968748152256012,
+      "learning_rate": 0.0008114814814814815,
+      "loss": 1.0355,
+      "step": 32090
+    },
+    {
+      "epoch": 118.89,
+      "grad_norm": 1.0011259317398071,
+      "learning_rate": 0.0008111111111111111,
+      "loss": 1.0581,
+      "step": 32100
+    },
+    {
+      "epoch": 118.93,
+      "grad_norm": 1.0687892436981201,
+      "learning_rate": 0.0008107407407407408,
+      "loss": 1.0671,
+      "step": 32110
+    },
+    {
+      "epoch": 118.96,
+      "grad_norm": 1.000532865524292,
+      "learning_rate": 0.0008103703703703705,
+      "loss": 1.074,
+      "step": 32120
+    },
+    {
+      "epoch": 119.0,
+      "grad_norm": 1.6631872653961182,
+      "learning_rate": 0.0008100000000000001,
+      "loss": 1.0843,
+      "step": 32130
+    },
+    {
+      "epoch": 119.04,
+      "grad_norm": 0.9295353889465332,
+      "learning_rate": 0.0008096296296296297,
+      "loss": 0.8693,
+      "step": 32140
+    },
+    {
+      "epoch": 119.07,
+      "grad_norm": 0.9665623307228088,
+      "learning_rate": 0.0008092592592592592,
+      "loss": 0.8577,
+      "step": 32150
+    },
+    {
+      "epoch": 119.11,
+      "grad_norm": 1.050161361694336,
+      "learning_rate": 0.0008088888888888889,
+      "loss": 0.8593,
+      "step": 32160
+    },
+    {
+      "epoch": 119.15,
+      "grad_norm": 0.9540622234344482,
+      "learning_rate": 0.0008085185185185185,
+      "loss": 0.8811,
+      "step": 32170
+    },
+    {
+      "epoch": 119.19,
+      "grad_norm": 0.9798749685287476,
+      "learning_rate": 0.0008081481481481482,
+      "loss": 0.9113,
+      "step": 32180
+    },
+    {
+      "epoch": 119.22,
+      "grad_norm": 0.9423299431800842,
+      "learning_rate": 0.0008077777777777778,
+      "loss": 0.9449,
+      "step": 32190
+    },
+    {
+      "epoch": 119.26,
+      "grad_norm": 0.8942784070968628,
+      "learning_rate": 0.0008074074074074075,
+      "loss": 0.9102,
+      "step": 32200
+    },
+    {
+      "epoch": 119.3,
+      "grad_norm": 0.9834209084510803,
+      "learning_rate": 0.000807037037037037,
+      "loss": 0.9457,
+      "step": 32210
+    },
+    {
+      "epoch": 119.33,
+      "grad_norm": 1.0610178709030151,
+      "learning_rate": 0.0008066666666666667,
+      "loss": 0.9484,
+      "step": 32220
+    },
+    {
+      "epoch": 119.37,
+      "grad_norm": 0.979119598865509,
+      "learning_rate": 0.0008062962962962963,
+      "loss": 0.9453,
+      "step": 32230
+    },
+    {
+      "epoch": 119.41,
+      "grad_norm": 0.9742558598518372,
+      "learning_rate": 0.000805925925925926,
+      "loss": 0.9711,
+      "step": 32240
+    },
+    {
+      "epoch": 119.44,
+      "grad_norm": 0.9968915581703186,
+      "learning_rate": 0.0008055555555555556,
+      "loss": 1.0053,
+      "step": 32250
+    },
+    {
+      "epoch": 119.48,
+      "grad_norm": 1.0201388597488403,
+      "learning_rate": 0.0008051851851851853,
+      "loss": 0.9567,
+      "step": 32260
+    },
+    {
+      "epoch": 119.52,
+      "grad_norm": 1.1031317710876465,
+      "learning_rate": 0.0008048148148148147,
+      "loss": 0.9918,
+      "step": 32270
+    },
+    {
+      "epoch": 119.56,
+      "grad_norm": 0.9733071327209473,
+      "learning_rate": 0.0008044444444444444,
+      "loss": 1.0015,
+      "step": 32280
+    },
+    {
+      "epoch": 119.59,
+      "grad_norm": 0.9740049242973328,
+      "learning_rate": 0.000804074074074074,
+      "loss": 1.0096,
+      "step": 32290
+    },
+    {
+      "epoch": 119.63,
+      "grad_norm": 1.0093234777450562,
+      "learning_rate": 0.0008037037037037037,
+      "loss": 1.0235,
+      "step": 32300
+    },
+    {
+      "epoch": 119.67,
+      "grad_norm": 0.9918221831321716,
+      "learning_rate": 0.0008033333333333333,
+      "loss": 1.0267,
+      "step": 32310
+    },
+    {
+      "epoch": 119.7,
+      "grad_norm": 1.023028016090393,
+      "learning_rate": 0.000802962962962963,
+      "loss": 1.0352,
+      "step": 32320
+    },
+    {
+      "epoch": 119.74,
+      "grad_norm": 1.0290321111679077,
+      "learning_rate": 0.0008025925925925927,
+      "loss": 1.0427,
+      "step": 32330
+    },
+    {
+      "epoch": 119.78,
+      "grad_norm": 0.9687529802322388,
+      "learning_rate": 0.0008022222222222222,
+      "loss": 1.0386,
+      "step": 32340
+    },
+    {
+      "epoch": 119.81,
+      "grad_norm": 1.0421777963638306,
+      "learning_rate": 0.0008018518518518519,
+      "loss": 1.0382,
+      "step": 32350
+    },
+    {
+      "epoch": 119.85,
+      "grad_norm": 1.0451891422271729,
+      "learning_rate": 0.0008014814814814815,
+      "loss": 1.0428,
+      "step": 32360
+    },
+    {
+      "epoch": 119.89,
+      "grad_norm": 1.085588812828064,
+      "learning_rate": 0.0008011111111111112,
+      "loss": 1.0628,
+      "step": 32370
+    },
+    {
+      "epoch": 119.93,
+      "grad_norm": 1.0221836566925049,
+      "learning_rate": 0.0008007407407407408,
+      "loss": 1.051,
+      "step": 32380
+    },
+    {
+      "epoch": 119.96,
+      "grad_norm": 1.0491361618041992,
+      "learning_rate": 0.0008003703703703704,
+      "loss": 1.0423,
+      "step": 32390
+    },
+    {
+      "epoch": 120.0,
+      "grad_norm": 1.8813464641571045,
+      "learning_rate": 0.0008,
+      "loss": 1.0748,
+      "step": 32400
+    },
+    {
+      "epoch": 120.04,
+      "grad_norm": 0.8277899026870728,
+      "learning_rate": 0.0007996296296296296,
+      "loss": 0.8565,
+      "step": 32410
+    },
+    {
+      "epoch": 120.07,
+      "grad_norm": 0.9462749361991882,
+      "learning_rate": 0.0007992592592592592,
+      "loss": 0.8509,
+      "step": 32420
+    },
+    {
+      "epoch": 120.11,
+      "grad_norm": 0.9435334801673889,
+      "learning_rate": 0.0007988888888888889,
+      "loss": 0.8491,
+      "step": 32430
+    },
+    {
+      "epoch": 120.15,
+      "grad_norm": 1.0299818515777588,
+      "learning_rate": 0.0007985185185185185,
+      "loss": 0.885,
+      "step": 32440
+    },
+    {
+      "epoch": 120.19,
+      "grad_norm": 0.9899550080299377,
+      "learning_rate": 0.0007981481481481482,
+      "loss": 0.8785,
+      "step": 32450
+    },
+    {
+      "epoch": 120.22,
+      "grad_norm": 1.0667580366134644,
+      "learning_rate": 0.0007977777777777778,
+      "loss": 0.9158,
+      "step": 32460
+    },
+    {
+      "epoch": 120.26,
+      "grad_norm": 0.9750791788101196,
+      "learning_rate": 0.0007974074074074075,
+      "loss": 0.9294,
+      "step": 32470
+    },
+    {
+      "epoch": 120.3,
+      "grad_norm": 1.0431256294250488,
+      "learning_rate": 0.000797037037037037,
+      "loss": 0.9302,
+      "step": 32480
+    },
+    {
+      "epoch": 120.33,
+      "grad_norm": 1.0735743045806885,
+      "learning_rate": 0.0007966666666666667,
+      "loss": 0.9377,
+      "step": 32490
+    },
+    {
+      "epoch": 120.37,
+      "grad_norm": 0.9974359273910522,
+      "learning_rate": 0.0007962962962962962,
+      "loss": 0.9559,
+      "step": 32500
+    },
+    {
+      "epoch": 120.41,
+      "grad_norm": 1.0248149633407593,
+      "learning_rate": 0.0007959259259259259,
+      "loss": 0.9594,
+      "step": 32510
+    },
+    {
+      "epoch": 120.44,
+      "grad_norm": 0.953029215335846,
+      "learning_rate": 0.0007955555555555555,
+      "loss": 0.9662,
+      "step": 32520
+    },
+    {
+      "epoch": 120.48,
+      "grad_norm": 0.9739720821380615,
+      "learning_rate": 0.0007951851851851852,
+      "loss": 0.965,
+      "step": 32530
+    },
+    {
+      "epoch": 120.52,
+      "grad_norm": 0.9954352378845215,
+      "learning_rate": 0.0007948148148148149,
+      "loss": 0.9945,
+      "step": 32540
+    },
+    {
+      "epoch": 120.56,
+      "grad_norm": 1.062570333480835,
+      "learning_rate": 0.0007944444444444444,
+      "loss": 0.9869,
+      "step": 32550
+    },
+    {
+      "epoch": 120.59,
+      "grad_norm": 0.9684034585952759,
+      "learning_rate": 0.0007940740740740741,
+      "loss": 1.019,
+      "step": 32560
+    },
+    {
+      "epoch": 120.63,
+      "grad_norm": 0.9902993440628052,
+      "learning_rate": 0.0007937037037037037,
+      "loss": 0.9838,
+      "step": 32570
+    },
+    {
+      "epoch": 120.67,
+      "grad_norm": 0.9986051917076111,
+      "learning_rate": 0.0007933333333333334,
+      "loss": 1.0076,
+      "step": 32580
+    },
+    {
+      "epoch": 120.7,
+      "grad_norm": 1.0536465644836426,
+      "learning_rate": 0.000792962962962963,
+      "loss": 1.0074,
+      "step": 32590
+    },
+    {
+      "epoch": 120.74,
+      "grad_norm": 1.090319275856018,
+      "learning_rate": 0.0007925925925925927,
+      "loss": 1.0298,
+      "step": 32600
+    },
+    {
+      "epoch": 120.78,
+      "grad_norm": 0.9828684329986572,
+      "learning_rate": 0.0007922222222222223,
+      "loss": 1.0508,
+      "step": 32610
+    },
+    {
+      "epoch": 120.81,
+      "grad_norm": 0.9842421412467957,
+      "learning_rate": 0.0007918518518518519,
+      "loss": 1.0391,
+      "step": 32620
+    },
+    {
+      "epoch": 120.85,
+      "grad_norm": 1.0293996334075928,
+      "learning_rate": 0.0007914814814814814,
+      "loss": 1.0497,
+      "step": 32630
+    },
+    {
+      "epoch": 120.89,
+      "grad_norm": 1.0321550369262695,
+      "learning_rate": 0.0007911111111111111,
+      "loss": 1.0525,
+      "step": 32640
+    },
+    {
+      "epoch": 120.93,
+      "grad_norm": 1.0024299621582031,
+      "learning_rate": 0.0007907407407407407,
+      "loss": 1.0228,
+      "step": 32650
+    },
+    {
+      "epoch": 120.96,
+      "grad_norm": 0.9814883470535278,
+      "learning_rate": 0.0007903703703703704,
+      "loss": 1.0406,
+      "step": 32660
+    },
+    {
+      "epoch": 121.0,
+      "grad_norm": 1.9922078847885132,
+      "learning_rate": 0.00079,
+      "loss": 1.0307,
+      "step": 32670
+    },
+    {
+      "epoch": 121.04,
+      "grad_norm": 1.0609437227249146,
+      "learning_rate": 0.0007896296296296297,
+      "loss": 0.8377,
+      "step": 32680
+    },
+    {
+      "epoch": 121.07,
+      "grad_norm": 0.9742987751960754,
+      "learning_rate": 0.0007892592592592592,
+      "loss": 0.8462,
+      "step": 32690
+    },
+    {
+      "epoch": 121.11,
+      "grad_norm": 0.940413773059845,
+      "learning_rate": 0.0007888888888888889,
+      "loss": 0.8666,
+      "step": 32700
+    },
+    {
+      "epoch": 121.15,
+      "grad_norm": 0.9640865325927734,
+      "learning_rate": 0.0007885185185185185,
+      "loss": 0.8805,
+      "step": 32710
+    },
+    {
+      "epoch": 121.19,
+      "grad_norm": 1.014676809310913,
+      "learning_rate": 0.0007881481481481482,
+      "loss": 0.878,
+      "step": 32720
+    },
+    {
+      "epoch": 121.22,
+      "grad_norm": 0.9674298167228699,
+      "learning_rate": 0.0007877777777777779,
+      "loss": 0.8901,
+      "step": 32730
+    },
+    {
+      "epoch": 121.26,
+      "grad_norm": 1.026246190071106,
+      "learning_rate": 0.0007874074074074075,
+      "loss": 0.9114,
+      "step": 32740
+    },
+    {
+      "epoch": 121.3,
+      "grad_norm": 1.008767008781433,
+      "learning_rate": 0.0007870370370370372,
+      "loss": 0.9209,
+      "step": 32750
+    },
+    {
+      "epoch": 121.33,
+      "grad_norm": 0.9800826907157898,
+      "learning_rate": 0.0007866666666666666,
+      "loss": 0.9407,
+      "step": 32760
+    },
+    {
+      "epoch": 121.37,
+      "grad_norm": 1.0001041889190674,
+      "learning_rate": 0.0007862962962962963,
+      "loss": 0.9482,
+      "step": 32770
+    },
+    {
+      "epoch": 121.41,
+      "grad_norm": 0.9887999296188354,
+      "learning_rate": 0.0007859259259259259,
+      "loss": 0.9462,
+      "step": 32780
+    },
+    {
+      "epoch": 121.44,
+      "grad_norm": 0.9700247049331665,
+      "learning_rate": 0.0007855555555555556,
+      "loss": 0.9602,
+      "step": 32790
+    },
+    {
+      "epoch": 121.48,
+      "grad_norm": 1.0221569538116455,
+      "learning_rate": 0.0007851851851851852,
+      "loss": 0.9525,
+      "step": 32800
+    },
+    {
+      "epoch": 121.52,
+      "grad_norm": 1.010459065437317,
+      "learning_rate": 0.0007848148148148149,
+      "loss": 0.9699,
+      "step": 32810
+    },
+    {
+      "epoch": 121.56,
+      "grad_norm": 0.9542679786682129,
+      "learning_rate": 0.0007844444444444445,
+      "loss": 0.9942,
+      "step": 32820
+    },
+    {
+      "epoch": 121.59,
+      "grad_norm": 1.0301337242126465,
+      "learning_rate": 0.0007840740740740741,
+      "loss": 0.9792,
+      "step": 32830
+    },
+    {
+      "epoch": 121.63,
+      "grad_norm": 1.0495370626449585,
+      "learning_rate": 0.0007837037037037037,
+      "loss": 0.9873,
+      "step": 32840
+    },
+    {
+      "epoch": 121.67,
+      "grad_norm": 0.9977912902832031,
+      "learning_rate": 0.0007833333333333334,
+      "loss": 0.9885,
+      "step": 32850
+    },
+    {
+      "epoch": 121.7,
+      "grad_norm": 1.0673377513885498,
+      "learning_rate": 0.000782962962962963,
+      "loss": 1.006,
+      "step": 32860
+    },
+    {
+      "epoch": 121.74,
+      "grad_norm": 0.8986328840255737,
+      "learning_rate": 0.0007825925925925926,
+      "loss": 1.028,
+      "step": 32870
+    },
+    {
+      "epoch": 121.78,
+      "grad_norm": 0.9932073354721069,
+      "learning_rate": 0.0007822222222222222,
+      "loss": 1.0151,
+      "step": 32880
+    },
+    {
+      "epoch": 121.81,
+      "grad_norm": 0.9659121036529541,
+      "learning_rate": 0.0007818518518518518,
+      "loss": 1.0158,
+      "step": 32890
+    },
+    {
+      "epoch": 121.85,
+      "grad_norm": 1.0076465606689453,
+      "learning_rate": 0.0007814814814814814,
+      "loss": 1.0401,
+      "step": 32900
+    },
+    {
+      "epoch": 121.89,
+      "grad_norm": 0.9795430898666382,
+      "learning_rate": 0.0007811111111111111,
+      "loss": 1.051,
+      "step": 32910
+    },
+    {
+      "epoch": 121.93,
+      "grad_norm": 1.0545165538787842,
+      "learning_rate": 0.0007807407407407407,
+      "loss": 1.0493,
+      "step": 32920
+    },
+    {
+      "epoch": 121.96,
+      "grad_norm": 1.021371603012085,
+      "learning_rate": 0.0007803703703703704,
+      "loss": 1.035,
+      "step": 32930
+    },
+    {
+      "epoch": 122.0,
+      "grad_norm": 1.738236904144287,
+      "learning_rate": 0.0007800000000000001,
+      "loss": 1.0769,
+      "step": 32940
+    },
+    {
+      "epoch": 122.04,
+      "grad_norm": 0.9248874187469482,
+      "learning_rate": 0.0007796296296296297,
+      "loss": 0.8293,
+      "step": 32950
+    },
+    {
+      "epoch": 122.07,
+      "grad_norm": 0.9658506512641907,
+      "learning_rate": 0.0007792592592592593,
+      "loss": 0.829,
+      "step": 32960
+    },
+    {
+      "epoch": 122.11,
+      "grad_norm": 0.9685617089271545,
+      "learning_rate": 0.0007788888888888889,
+      "loss": 0.8443,
+      "step": 32970
+    },
+    {
+      "epoch": 122.15,
+      "grad_norm": 1.0525528192520142,
+      "learning_rate": 0.0007785185185185186,
+      "loss": 0.8754,
+      "step": 32980
+    },
+    {
+      "epoch": 122.19,
+      "grad_norm": 0.9952585101127625,
+      "learning_rate": 0.0007781481481481481,
+      "loss": 0.8739,
+      "step": 32990
+    },
+    {
+      "epoch": 122.22,
+      "grad_norm": 0.993129312992096,
+      "learning_rate": 0.0007777777777777778,
+      "loss": 0.8789,
+      "step": 33000
+    },
+    {
+      "epoch": 122.26,
+      "grad_norm": 0.9832338094711304,
+      "learning_rate": 0.0007774074074074074,
+      "loss": 0.9105,
+      "step": 33010
+    },
+    {
+      "epoch": 122.3,
+      "grad_norm": 0.9596456289291382,
+      "learning_rate": 0.0007770370370370371,
+      "loss": 0.9036,
+      "step": 33020
+    },
+    {
+      "epoch": 122.33,
+      "grad_norm": 0.9873278141021729,
+      "learning_rate": 0.0007766666666666666,
+      "loss": 0.9134,
+      "step": 33030
+    },
+    {
+      "epoch": 122.37,
+      "grad_norm": 0.9594837427139282,
+      "learning_rate": 0.0007762962962962963,
+      "loss": 0.9609,
+      "step": 33040
+    },
+    {
+      "epoch": 122.41,
+      "grad_norm": 1.0137587785720825,
+      "learning_rate": 0.0007759259259259259,
+      "loss": 0.9351,
+      "step": 33050
+    },
+    {
+      "epoch": 122.44,
+      "grad_norm": 0.9831103086471558,
+      "learning_rate": 0.0007755555555555556,
+      "loss": 0.9798,
+      "step": 33060
+    },
+    {
+      "epoch": 122.48,
+      "grad_norm": 1.0451667308807373,
+      "learning_rate": 0.0007751851851851852,
+      "loss": 0.9436,
+      "step": 33070
+    },
+    {
+      "epoch": 122.52,
+      "grad_norm": 1.0881359577178955,
+      "learning_rate": 0.0007748148148148149,
+      "loss": 0.9354,
+      "step": 33080
+    },
+    {
+      "epoch": 122.56,
+      "grad_norm": 0.9605501890182495,
+      "learning_rate": 0.0007744444444444445,
+      "loss": 0.9794,
+      "step": 33090
+    },
+    {
+      "epoch": 122.59,
+      "grad_norm": 1.031227707862854,
+      "learning_rate": 0.000774074074074074,
+      "loss": 0.9752,
+      "step": 33100
+    },
+    {
+      "epoch": 122.63,
+      "grad_norm": 0.9974759817123413,
+      "learning_rate": 0.0007737037037037036,
+      "loss": 1.0123,
+      "step": 33110
+    },
+    {
+      "epoch": 122.67,
+      "grad_norm": 1.045825481414795,
+      "learning_rate": 0.0007733333333333333,
+      "loss": 0.9922,
+      "step": 33120
+    },
+    {
+      "epoch": 122.7,
+      "grad_norm": 1.1240839958190918,
+      "learning_rate": 0.0007729629629629629,
+      "loss": 1.0032,
+      "step": 33130
+    },
+    {
+      "epoch": 122.74,
+      "grad_norm": 1.0099389553070068,
+      "learning_rate": 0.0007725925925925926,
+      "loss": 1.0178,
+      "step": 33140
+    },
+    {
+      "epoch": 122.78,
+      "grad_norm": 1.0204278230667114,
+      "learning_rate": 0.0007722222222222223,
+      "loss": 1.0352,
+      "step": 33150
+    },
+    {
+      "epoch": 122.81,
+      "grad_norm": 0.9793760776519775,
+      "learning_rate": 0.0007718518518518519,
+      "loss": 1.0066,
+      "step": 33160
+    },
+    {
+      "epoch": 122.85,
+      "grad_norm": 1.0178688764572144,
+      "learning_rate": 0.0007714814814814815,
+      "loss": 1.0268,
+      "step": 33170
+    },
+    {
+      "epoch": 122.89,
+      "grad_norm": 1.049781084060669,
+      "learning_rate": 0.0007711111111111111,
+      "loss": 1.0268,
+      "step": 33180
+    },
+    {
+      "epoch": 122.93,
+      "grad_norm": 1.0570471286773682,
+      "learning_rate": 0.0007707407407407408,
+      "loss": 1.0192,
+      "step": 33190
+    },
+    {
+      "epoch": 122.96,
+      "grad_norm": 1.07374107837677,
+      "learning_rate": 0.0007703703703703704,
+      "loss": 1.0386,
+      "step": 33200
+    },
+    {
+      "epoch": 123.0,
+      "grad_norm": 1.6337435245513916,
+      "learning_rate": 0.0007700000000000001,
+      "loss": 1.0267,
+      "step": 33210
+    },
+    {
+      "epoch": 123.04,
+      "grad_norm": 0.9590765833854675,
+      "learning_rate": 0.0007696296296296297,
+      "loss": 0.8187,
+      "step": 33220
+    },
+    {
+      "epoch": 123.07,
+      "grad_norm": 0.9379592537879944,
+      "learning_rate": 0.0007692592592592594,
+      "loss": 0.8466,
+      "step": 33230
+    },
+    {
+      "epoch": 123.11,
+      "grad_norm": 0.9660229086875916,
+      "learning_rate": 0.0007688888888888888,
+      "loss": 0.8508,
+      "step": 33240
+    },
+    {
+      "epoch": 123.15,
+      "grad_norm": 1.0518168210983276,
+      "learning_rate": 0.0007685185185185185,
+      "loss": 0.8775,
+      "step": 33250
+    },
+    {
+      "epoch": 123.19,
+      "grad_norm": 0.9530587792396545,
+      "learning_rate": 0.0007681481481481481,
+      "loss": 0.8695,
+      "step": 33260
+    },
+    {
+      "epoch": 123.22,
+      "grad_norm": 1.0211882591247559,
+      "learning_rate": 0.0007677777777777778,
+      "loss": 0.8615,
+      "step": 33270
+    },
+    {
+      "epoch": 123.26,
+      "grad_norm": 1.0455659627914429,
+      "learning_rate": 0.0007674074074074074,
+      "loss": 0.8946,
+      "step": 33280
+    },
+    {
+      "epoch": 123.3,
+      "grad_norm": 1.0107338428497314,
+      "learning_rate": 0.0007670370370370371,
+      "loss": 0.9007,
+      "step": 33290
+    },
+    {
+      "epoch": 123.33,
+      "grad_norm": 1.074458122253418,
+      "learning_rate": 0.0007666666666666667,
+      "loss": 0.9061,
+      "step": 33300
+    },
+    {
+      "epoch": 123.37,
+      "grad_norm": 1.048465371131897,
+      "learning_rate": 0.0007662962962962963,
+      "loss": 0.9312,
+      "step": 33310
+    },
+    {
+      "epoch": 123.41,
+      "grad_norm": 1.0410969257354736,
+      "learning_rate": 0.0007659259259259259,
+      "loss": 0.9258,
+      "step": 33320
+    },
+    {
+      "epoch": 123.44,
+      "grad_norm": 1.0864170789718628,
+      "learning_rate": 0.0007655555555555556,
+      "loss": 0.9429,
+      "step": 33330
+    },
+    {
+      "epoch": 123.48,
+      "grad_norm": 0.9920828342437744,
+      "learning_rate": 0.0007651851851851852,
+      "loss": 0.941,
+      "step": 33340
+    },
+    {
+      "epoch": 123.52,
+      "grad_norm": 1.063452959060669,
+      "learning_rate": 0.0007648148148148148,
+      "loss": 0.9541,
+      "step": 33350
+    },
+    {
+      "epoch": 123.56,
+      "grad_norm": 0.9817026853561401,
+      "learning_rate": 0.0007644444444444445,
+      "loss": 0.9551,
+      "step": 33360
+    },
+    {
+      "epoch": 123.59,
+      "grad_norm": 0.9735251069068909,
+      "learning_rate": 0.0007640740740740741,
+      "loss": 0.9821,
+      "step": 33370
+    },
+    {
+      "epoch": 123.63,
+      "grad_norm": 1.1082037687301636,
+      "learning_rate": 0.0007637037037037037,
+      "loss": 0.9739,
+      "step": 33380
+    },
+    {
+      "epoch": 123.67,
+      "grad_norm": 1.0017740726470947,
+      "learning_rate": 0.0007633333333333333,
+      "loss": 1.0015,
+      "step": 33390
+    },
+    {
+      "epoch": 123.7,
+      "grad_norm": 1.0714524984359741,
+      "learning_rate": 0.000762962962962963,
+      "loss": 1.0068,
+      "step": 33400
+    },
+    {
+      "epoch": 123.74,
+      "grad_norm": 1.0269007682800293,
+      "learning_rate": 0.0007625925925925926,
+      "loss": 0.9995,
+      "step": 33410
+    },
+    {
+      "epoch": 123.78,
+      "grad_norm": 0.9959019422531128,
+      "learning_rate": 0.0007622222222222223,
+      "loss": 0.9977,
+      "step": 33420
+    },
+    {
+      "epoch": 123.81,
+      "grad_norm": 1.0212533473968506,
+      "learning_rate": 0.0007618518518518519,
+      "loss": 1.0152,
+      "step": 33430
+    },
+    {
+      "epoch": 123.85,
+      "grad_norm": 1.0144740343093872,
+      "learning_rate": 0.0007614814814814816,
+      "loss": 1.0319,
+      "step": 33440
+    },
+    {
+      "epoch": 123.89,
+      "grad_norm": 1.0200554132461548,
+      "learning_rate": 0.0007611111111111111,
+      "loss": 0.9977,
+      "step": 33450
+    },
+    {
+      "epoch": 123.93,
+      "grad_norm": 1.09546959400177,
+      "learning_rate": 0.0007607407407407408,
+      "loss": 1.0041,
+      "step": 33460
+    },
+    {
+      "epoch": 123.96,
+      "grad_norm": 1.0374383926391602,
+      "learning_rate": 0.0007603703703703703,
+      "loss": 1.0376,
+      "step": 33470
+    },
+    {
+      "epoch": 124.0,
+      "grad_norm": 1.8594433069229126,
+      "learning_rate": 0.00076,
+      "loss": 1.0343,
+      "step": 33480
+    },
+    {
+      "epoch": 124.04,
+      "grad_norm": 0.9433583617210388,
+      "learning_rate": 0.0007596296296296296,
+      "loss": 0.8191,
+      "step": 33490
+    },
+    {
+      "epoch": 124.07,
+      "grad_norm": 0.9556878805160522,
+      "learning_rate": 0.0007592592592592593,
+      "loss": 0.8319,
+      "step": 33500
+    },
+    {
+      "epoch": 124.11,
+      "grad_norm": 1.0316696166992188,
+      "learning_rate": 0.0007588888888888888,
+      "loss": 0.832,
+      "step": 33510
+    },
+    {
+      "epoch": 124.15,
+      "grad_norm": 1.0155198574066162,
+      "learning_rate": 0.0007585185185185185,
+      "loss": 0.8381,
+      "step": 33520
+    },
+    {
+      "epoch": 124.19,
+      "grad_norm": 0.9573029279708862,
+      "learning_rate": 0.0007581481481481481,
+      "loss": 0.8658,
+      "step": 33530
+    },
+    {
+      "epoch": 124.22,
+      "grad_norm": 1.0463179349899292,
+      "learning_rate": 0.0007577777777777778,
+      "loss": 0.8555,
+      "step": 33540
+    },
+    {
+      "epoch": 124.26,
+      "grad_norm": 0.9773217439651489,
+      "learning_rate": 0.0007574074074074075,
+      "loss": 0.8756,
+      "step": 33550
+    },
+    {
+      "epoch": 124.3,
+      "grad_norm": 1.0205600261688232,
+      "learning_rate": 0.0007570370370370371,
+      "loss": 0.893,
+      "step": 33560
+    },
+    {
+      "epoch": 124.33,
+      "grad_norm": 1.0751043558120728,
+      "learning_rate": 0.0007566666666666668,
+      "loss": 0.9006,
+      "step": 33570
+    },
+    {
+      "epoch": 124.37,
+      "grad_norm": 1.078392744064331,
+      "learning_rate": 0.0007562962962962963,
+      "loss": 0.9209,
+      "step": 33580
+    },
+    {
+      "epoch": 124.41,
+      "grad_norm": 1.0406211614608765,
+      "learning_rate": 0.000755925925925926,
+      "loss": 0.9654,
+      "step": 33590
+    },
+    {
+      "epoch": 124.44,
+      "grad_norm": 1.0287216901779175,
+      "learning_rate": 0.0007555555555555555,
+      "loss": 0.942,
+      "step": 33600
+    },
+    {
+      "epoch": 124.48,
+      "grad_norm": 1.0419343709945679,
+      "learning_rate": 0.0007551851851851852,
+      "loss": 0.9506,
+      "step": 33610
+    },
+    {
+      "epoch": 124.52,
+      "grad_norm": 1.0294045209884644,
+      "learning_rate": 0.0007548148148148148,
+      "loss": 0.9267,
+      "step": 33620
+    },
+    {
+      "epoch": 124.56,
+      "grad_norm": 1.050539255142212,
+      "learning_rate": 0.0007544444444444445,
+      "loss": 0.9569,
+      "step": 33630
+    },
+    {
+      "epoch": 124.59,
+      "grad_norm": 1.0169801712036133,
+      "learning_rate": 0.0007540740740740741,
+      "loss": 0.964,
+      "step": 33640
+    },
+    {
+      "epoch": 124.63,
+      "grad_norm": 1.0593849420547485,
+      "learning_rate": 0.0007537037037037037,
+      "loss": 0.9614,
+      "step": 33650
+    },
+    {
+      "epoch": 124.67,
+      "grad_norm": 1.0408095121383667,
+      "learning_rate": 0.0007533333333333333,
+      "loss": 0.9934,
+      "step": 33660
+    },
+    {
+      "epoch": 124.7,
+      "grad_norm": 1.070684790611267,
+      "learning_rate": 0.000752962962962963,
+      "loss": 0.9856,
+      "step": 33670
+    },
+    {
+      "epoch": 124.74,
+      "grad_norm": 0.9906821250915527,
+      "learning_rate": 0.0007525925925925926,
+      "loss": 0.9733,
+      "step": 33680
+    },
+    {
+      "epoch": 124.78,
+      "grad_norm": 1.0587712526321411,
+      "learning_rate": 0.0007522222222222223,
+      "loss": 0.997,
+      "step": 33690
+    },
+    {
+      "epoch": 124.81,
+      "grad_norm": 1.0874089002609253,
+      "learning_rate": 0.0007518518518518519,
+      "loss": 1.008,
+      "step": 33700
+    },
+    {
+      "epoch": 124.85,
+      "grad_norm": 1.0352959632873535,
+      "learning_rate": 0.0007514814814814816,
+      "loss": 1.0052,
+      "step": 33710
+    },
+    {
+      "epoch": 124.89,
+      "grad_norm": 1.0280036926269531,
+      "learning_rate": 0.000751111111111111,
+      "loss": 1.0301,
+      "step": 33720
+    },
+    {
+      "epoch": 124.93,
+      "grad_norm": 1.1919608116149902,
+      "learning_rate": 0.0007507407407407407,
+      "loss": 1.0094,
+      "step": 33730
+    },
+    {
+      "epoch": 124.96,
+      "grad_norm": 1.0236761569976807,
+      "learning_rate": 0.0007503703703703703,
+      "loss": 1.0026,
+      "step": 33740
+    },
+    {
+      "epoch": 125.0,
+      "grad_norm": 1.8308442831039429,
+      "learning_rate": 0.00075,
+      "loss": 1.0376,
+      "step": 33750
+    },
+    {
+      "epoch": 125.04,
+      "grad_norm": 1.008743166923523,
+      "learning_rate": 0.0007496296296296297,
+      "loss": 0.8049,
+      "step": 33760
+    },
+    {
+      "epoch": 125.07,
+      "grad_norm": 0.9807414412498474,
+      "learning_rate": 0.0007492592592592593,
+      "loss": 0.8093,
+      "step": 33770
+    },
+    {
+      "epoch": 125.11,
+      "grad_norm": 1.0613797903060913,
+      "learning_rate": 0.000748888888888889,
+      "loss": 0.8376,
+      "step": 33780
+    },
+    {
+      "epoch": 125.15,
+      "grad_norm": 0.9105537533760071,
+      "learning_rate": 0.0007485185185185185,
+      "loss": 0.8249,
+      "step": 33790
+    },
+    {
+      "epoch": 125.19,
+      "grad_norm": 0.9461547136306763,
+      "learning_rate": 0.0007481481481481482,
+      "loss": 0.8464,
+      "step": 33800
+    },
+    {
+      "epoch": 125.22,
+      "grad_norm": 0.9608639478683472,
+      "learning_rate": 0.0007477777777777778,
+      "loss": 0.8709,
+      "step": 33810
+    },
+    {
+      "epoch": 125.26,
+      "grad_norm": 1.0083810091018677,
+      "learning_rate": 0.0007474074074074075,
+      "loss": 0.8711,
+      "step": 33820
+    },
+    {
+      "epoch": 125.3,
+      "grad_norm": 1.0758371353149414,
+      "learning_rate": 0.000747037037037037,
+      "loss": 0.874,
+      "step": 33830
+    },
+    {
+      "epoch": 125.33,
+      "grad_norm": 1.0421946048736572,
+      "learning_rate": 0.0007466666666666667,
+      "loss": 0.9086,
+      "step": 33840
+    },
+    {
+      "epoch": 125.37,
+      "grad_norm": 1.081777572631836,
+      "learning_rate": 0.0007462962962962963,
+      "loss": 0.9225,
+      "step": 33850
+    },
+    {
+      "epoch": 125.41,
+      "grad_norm": 1.0842175483703613,
+      "learning_rate": 0.0007459259259259259,
+      "loss": 0.9046,
+      "step": 33860
+    },
+    {
+      "epoch": 125.44,
+      "grad_norm": 0.9907249808311462,
+      "learning_rate": 0.0007455555555555555,
+      "loss": 0.9335,
+      "step": 33870
+    },
+    {
+      "epoch": 125.48,
+      "grad_norm": 1.116257667541504,
+      "learning_rate": 0.0007451851851851852,
+      "loss": 0.9632,
+      "step": 33880
+    },
+    {
+      "epoch": 125.52,
+      "grad_norm": 1.0658822059631348,
+      "learning_rate": 0.0007448148148148148,
+      "loss": 0.9315,
+      "step": 33890
+    },
+    {
+      "epoch": 125.56,
+      "grad_norm": 1.0912448167800903,
+      "learning_rate": 0.0007444444444444445,
+      "loss": 0.9481,
+      "step": 33900
+    },
+    {
+      "epoch": 125.59,
+      "grad_norm": 1.0179263353347778,
+      "learning_rate": 0.0007440740740740741,
+      "loss": 0.9667,
+      "step": 33910
+    },
+    {
+      "epoch": 125.63,
+      "grad_norm": 0.9753648638725281,
+      "learning_rate": 0.0007437037037037038,
+      "loss": 0.986,
+      "step": 33920
+    },
+    {
+      "epoch": 125.67,
+      "grad_norm": 1.0527098178863525,
+      "learning_rate": 0.0007433333333333333,
+      "loss": 0.9526,
+      "step": 33930
+    },
+    {
+      "epoch": 125.7,
+      "grad_norm": 1.0270941257476807,
+      "learning_rate": 0.000742962962962963,
+      "loss": 0.9615,
+      "step": 33940
+    },
+    {
+      "epoch": 125.74,
+      "grad_norm": 1.0542274713516235,
+      "learning_rate": 0.0007425925925925925,
+      "loss": 0.9825,
+      "step": 33950
+    },
+    {
+      "epoch": 125.78,
+      "grad_norm": 1.0598552227020264,
+      "learning_rate": 0.0007422222222222222,
+      "loss": 0.9979,
+      "step": 33960
+    },
+    {
+      "epoch": 125.81,
+      "grad_norm": 1.0630998611450195,
+      "learning_rate": 0.0007418518518518519,
+      "loss": 0.9916,
+      "step": 33970
+    },
+    {
+      "epoch": 125.85,
+      "grad_norm": 1.0409631729125977,
+      "learning_rate": 0.0007414814814814815,
+      "loss": 0.999,
+      "step": 33980
+    },
+    {
+      "epoch": 125.89,
+      "grad_norm": 1.0184717178344727,
+      "learning_rate": 0.0007411111111111112,
+      "loss": 1.0104,
+      "step": 33990
+    },
+    {
+      "epoch": 125.93,
+      "grad_norm": 1.0719102621078491,
+      "learning_rate": 0.0007407407407407407,
+      "loss": 0.9791,
+      "step": 34000
+    },
+    {
+      "epoch": 125.96,
+      "grad_norm": 1.1049607992172241,
+      "learning_rate": 0.0007403703703703704,
+      "loss": 1.0184,
+      "step": 34010
+    },
+    {
+      "epoch": 126.0,
+      "grad_norm": 2.2022151947021484,
+      "learning_rate": 0.00074,
+      "loss": 1.0025,
+      "step": 34020
+    },
+    {
+      "epoch": 126.04,
+      "grad_norm": 0.9834014773368835,
+      "learning_rate": 0.0007396296296296297,
+      "loss": 0.8303,
+      "step": 34030
+    },
+    {
+      "epoch": 126.07,
+      "grad_norm": 0.9932088851928711,
+      "learning_rate": 0.0007392592592592593,
+      "loss": 0.805,
+      "step": 34040
+    },
+    {
+      "epoch": 126.11,
+      "grad_norm": 1.0134624242782593,
+      "learning_rate": 0.000738888888888889,
+      "loss": 0.8157,
+      "step": 34050
+    },
+    {
+      "epoch": 126.15,
+      "grad_norm": 0.9468620419502258,
+      "learning_rate": 0.0007385185185185185,
+      "loss": 0.8269,
+      "step": 34060
+    },
+    {
+      "epoch": 126.19,
+      "grad_norm": 1.043798565864563,
+      "learning_rate": 0.0007381481481481481,
+      "loss": 0.8467,
+      "step": 34070
+    },
+    {
+      "epoch": 126.22,
+      "grad_norm": 0.9726648926734924,
+      "learning_rate": 0.0007377777777777777,
+      "loss": 0.8784,
+      "step": 34080
+    },
+    {
+      "epoch": 126.26,
+      "grad_norm": 0.96148282289505,
+      "learning_rate": 0.0007374074074074074,
+      "loss": 0.8735,
+      "step": 34090
+    },
+    {
+      "epoch": 126.3,
+      "grad_norm": 1.0156971216201782,
+      "learning_rate": 0.000737037037037037,
+      "loss": 0.8906,
+      "step": 34100
+    },
+    {
+      "epoch": 126.33,
+      "grad_norm": 1.028420090675354,
+      "learning_rate": 0.0007366666666666667,
+      "loss": 0.8966,
+      "step": 34110
+    },
+    {
+      "epoch": 126.37,
+      "grad_norm": 1.0544856786727905,
+      "learning_rate": 0.0007362962962962963,
+      "loss": 0.8922,
+      "step": 34120
+    },
+    {
+      "epoch": 126.41,
+      "grad_norm": 1.127189040184021,
+      "learning_rate": 0.0007359259259259259,
+      "loss": 0.9044,
+      "step": 34130
+    },
+    {
+      "epoch": 126.44,
+      "grad_norm": 0.9666948914527893,
+      "learning_rate": 0.0007355555555555555,
+      "loss": 0.9138,
+      "step": 34140
+    },
+    {
+      "epoch": 126.48,
+      "grad_norm": 1.0822830200195312,
+      "learning_rate": 0.0007351851851851852,
+      "loss": 0.926,
+      "step": 34150
+    },
+    {
+      "epoch": 126.52,
+      "grad_norm": 1.027675747871399,
+      "learning_rate": 0.0007348148148148149,
+      "loss": 0.9362,
+      "step": 34160
+    },
+    {
+      "epoch": 126.56,
+      "grad_norm": 1.0344058275222778,
+      "learning_rate": 0.0007344444444444445,
+      "loss": 0.948,
+      "step": 34170
+    },
+    {
+      "epoch": 126.59,
+      "grad_norm": 0.994404673576355,
+      "learning_rate": 0.0007340740740740742,
+      "loss": 0.9551,
+      "step": 34180
+    },
+    {
+      "epoch": 126.63,
+      "grad_norm": 1.0416563749313354,
+      "learning_rate": 0.0007337037037037038,
+      "loss": 0.9298,
+      "step": 34190
+    },
+    {
+      "epoch": 126.67,
+      "grad_norm": 0.9902626872062683,
+      "learning_rate": 0.0007333333333333333,
+      "loss": 0.9507,
+      "step": 34200
+    },
+    {
+      "epoch": 126.7,
+      "grad_norm": 1.032486915588379,
+      "learning_rate": 0.0007329629629629629,
+      "loss": 0.9601,
+      "step": 34210
+    },
+    {
+      "epoch": 126.74,
+      "grad_norm": 0.974919855594635,
+      "learning_rate": 0.0007325925925925926,
+      "loss": 0.9625,
+      "step": 34220
+    },
+    {
+      "epoch": 126.78,
+      "grad_norm": 1.0024913549423218,
+      "learning_rate": 0.0007322222222222222,
+      "loss": 0.9646,
+      "step": 34230
+    },
+    {
+      "epoch": 126.81,
+      "grad_norm": 1.0604099035263062,
+      "learning_rate": 0.0007318518518518519,
+      "loss": 0.9898,
+      "step": 34240
+    },
+    {
+      "epoch": 126.85,
+      "grad_norm": 1.0008738040924072,
+      "learning_rate": 0.0007314814814814815,
+      "loss": 0.9755,
+      "step": 34250
+    },
+    {
+      "epoch": 126.89,
+      "grad_norm": 1.154698133468628,
+      "learning_rate": 0.0007311111111111112,
+      "loss": 0.9874,
+      "step": 34260
+    },
+    {
+      "epoch": 126.93,
+      "grad_norm": 0.9528394937515259,
+      "learning_rate": 0.0007307407407407407,
+      "loss": 1.0169,
+      "step": 34270
+    },
+    {
+      "epoch": 126.96,
+      "grad_norm": 1.0628700256347656,
+      "learning_rate": 0.0007303703703703704,
+      "loss": 1.0054,
+      "step": 34280
+    },
+    {
+      "epoch": 127.0,
+      "grad_norm": 2.0968518257141113,
+      "learning_rate": 0.00073,
+      "loss": 1.0023,
+      "step": 34290
+    },
+    {
+      "epoch": 127.04,
+      "grad_norm": 1.0144600868225098,
+      "learning_rate": 0.0007296296296296297,
+      "loss": 0.7879,
+      "step": 34300
+    },
+    {
+      "epoch": 127.07,
+      "grad_norm": 1.0181728601455688,
+      "learning_rate": 0.0007292592592592592,
+      "loss": 0.8311,
+      "step": 34310
+    },
+    {
+      "epoch": 127.11,
+      "grad_norm": 0.9796633720397949,
+      "learning_rate": 0.0007288888888888889,
+      "loss": 0.8241,
+      "step": 34320
+    },
+    {
+      "epoch": 127.15,
+      "grad_norm": 1.0185998678207397,
+      "learning_rate": 0.0007285185185185185,
+      "loss": 0.8255,
+      "step": 34330
+    },
+    {
+      "epoch": 127.19,
+      "grad_norm": 1.039688229560852,
+      "learning_rate": 0.0007281481481481481,
+      "loss": 0.87,
+      "step": 34340
+    },
+    {
+      "epoch": 127.22,
+      "grad_norm": 0.9951873421669006,
+      "learning_rate": 0.0007277777777777777,
+      "loss": 0.8566,
+      "step": 34350
+    },
+    {
+      "epoch": 127.26,
+      "grad_norm": 0.9923152923583984,
+      "learning_rate": 0.0007274074074074074,
+      "loss": 0.894,
+      "step": 34360
+    },
+    {
+      "epoch": 127.3,
+      "grad_norm": 1.0109975337982178,
+      "learning_rate": 0.0007270370370370371,
+      "loss": 0.847,
+      "step": 34370
+    },
+    {
+      "epoch": 127.33,
+      "grad_norm": 1.042435884475708,
+      "learning_rate": 0.0007266666666666667,
+      "loss": 0.893,
+      "step": 34380
+    },
+    {
+      "epoch": 127.37,
+      "grad_norm": 1.0469590425491333,
+      "learning_rate": 0.0007262962962962964,
+      "loss": 0.9133,
+      "step": 34390
+    },
+    {
+      "epoch": 127.41,
+      "grad_norm": 0.952255129814148,
+      "learning_rate": 0.000725925925925926,
+      "loss": 0.889,
+      "step": 34400
+    },
+    {
+      "epoch": 127.44,
+      "grad_norm": 1.045935869216919,
+      "learning_rate": 0.0007255555555555556,
+      "loss": 0.9128,
+      "step": 34410
+    },
+    {
+      "epoch": 127.48,
+      "grad_norm": 1.0044687986373901,
+      "learning_rate": 0.0007251851851851852,
+      "loss": 0.8965,
+      "step": 34420
+    },
+    {
+      "epoch": 127.52,
+      "grad_norm": 1.0032331943511963,
+      "learning_rate": 0.0007248148148148149,
+      "loss": 0.9286,
+      "step": 34430
+    },
+    {
+      "epoch": 127.56,
+      "grad_norm": 1.060608148574829,
+      "learning_rate": 0.0007244444444444444,
+      "loss": 0.9221,
+      "step": 34440
+    },
+    {
+      "epoch": 127.59,
+      "grad_norm": 1.0088783502578735,
+      "learning_rate": 0.0007240740740740741,
+      "loss": 0.9412,
+      "step": 34450
+    },
+    {
+      "epoch": 127.63,
+      "grad_norm": 1.036992073059082,
+      "learning_rate": 0.0007237037037037037,
+      "loss": 0.9274,
+      "step": 34460
+    },
+    {
+      "epoch": 127.67,
+      "grad_norm": 1.015704870223999,
+      "learning_rate": 0.0007233333333333334,
+      "loss": 0.9389,
+      "step": 34470
+    },
+    {
+      "epoch": 127.7,
+      "grad_norm": 1.0427706241607666,
+      "learning_rate": 0.0007229629629629629,
+      "loss": 0.9613,
+      "step": 34480
+    },
+    {
+      "epoch": 127.74,
+      "grad_norm": 1.0360188484191895,
+      "learning_rate": 0.0007225925925925926,
+      "loss": 0.9712,
+      "step": 34490
+    },
+    {
+      "epoch": 127.78,
+      "grad_norm": 1.0515973567962646,
+      "learning_rate": 0.0007222222222222222,
+      "loss": 0.9528,
+      "step": 34500
+    },
+    {
+      "epoch": 127.81,
+      "grad_norm": 1.0636948347091675,
+      "learning_rate": 0.0007218518518518519,
+      "loss": 0.9715,
+      "step": 34510
+    },
+    {
+      "epoch": 127.85,
+      "grad_norm": 1.0268125534057617,
+      "learning_rate": 0.0007214814814814815,
+      "loss": 0.9949,
+      "step": 34520
+    },
+    {
+      "epoch": 127.89,
+      "grad_norm": 1.0717504024505615,
+      "learning_rate": 0.0007211111111111112,
+      "loss": 0.9876,
+      "step": 34530
+    },
+    {
+      "epoch": 127.93,
+      "grad_norm": 1.1457937955856323,
+      "learning_rate": 0.0007207407407407408,
+      "loss": 0.9975,
+      "step": 34540
+    },
+    {
+      "epoch": 127.96,
+      "grad_norm": 1.0468859672546387,
+      "learning_rate": 0.0007203703703703703,
+      "loss": 0.999,
+      "step": 34550
+    },
+    {
+      "epoch": 128.0,
+      "grad_norm": 1.6671861410140991,
+      "learning_rate": 0.0007199999999999999,
+      "loss": 0.9965,
+      "step": 34560
+    },
+    {
+      "epoch": 128.04,
+      "grad_norm": 0.9380455017089844,
+      "learning_rate": 0.0007196296296296296,
+      "loss": 0.7957,
+      "step": 34570
+    },
+    {
+      "epoch": 128.07,
+      "grad_norm": 1.0969723463058472,
+      "learning_rate": 0.0007192592592592593,
+      "loss": 0.8217,
+      "step": 34580
+    },
+    {
+      "epoch": 128.11,
+      "grad_norm": 1.0519886016845703,
+      "learning_rate": 0.0007188888888888889,
+      "loss": 0.8212,
+      "step": 34590
+    },
+    {
+      "epoch": 128.15,
+      "grad_norm": 1.0497735738754272,
+      "learning_rate": 0.0007185185185185186,
+      "loss": 0.79,
+      "step": 34600
+    },
+    {
+      "epoch": 128.19,
+      "grad_norm": 0.9826639890670776,
+      "learning_rate": 0.0007181481481481482,
+      "loss": 0.8176,
+      "step": 34610
+    },
+    {
+      "epoch": 128.22,
+      "grad_norm": 0.9700071811676025,
+      "learning_rate": 0.0007177777777777778,
+      "loss": 0.8425,
+      "step": 34620
+    },
+    {
+      "epoch": 128.26,
+      "grad_norm": 1.027782678604126,
+      "learning_rate": 0.0007174074074074074,
+      "loss": 0.8592,
+      "step": 34630
+    },
+    {
+      "epoch": 128.3,
+      "grad_norm": 1.0365031957626343,
+      "learning_rate": 0.0007170370370370371,
+      "loss": 0.8615,
+      "step": 34640
+    },
+    {
+      "epoch": 128.33,
+      "grad_norm": 1.0428645610809326,
+      "learning_rate": 0.0007166666666666667,
+      "loss": 0.878,
+      "step": 34650
+    },
+    {
+      "epoch": 128.37,
+      "grad_norm": 1.0856304168701172,
+      "learning_rate": 0.0007162962962962964,
+      "loss": 0.8684,
+      "step": 34660
+    },
+    {
+      "epoch": 128.41,
+      "grad_norm": 1.0055755376815796,
+      "learning_rate": 0.000715925925925926,
+      "loss": 0.8736,
+      "step": 34670
+    },
+    {
+      "epoch": 128.44,
+      "grad_norm": 1.0482579469680786,
+      "learning_rate": 0.0007155555555555555,
+      "loss": 0.8844,
+      "step": 34680
+    },
+    {
+      "epoch": 128.48,
+      "grad_norm": 1.0333335399627686,
+      "learning_rate": 0.0007151851851851851,
+      "loss": 0.8991,
+      "step": 34690
+    },
+    {
+      "epoch": 128.52,
+      "grad_norm": 1.0611454248428345,
+      "learning_rate": 0.0007148148148148148,
+      "loss": 0.9079,
+      "step": 34700
+    },
+    {
+      "epoch": 128.56,
+      "grad_norm": 1.0447732210159302,
+      "learning_rate": 0.0007144444444444444,
+      "loss": 0.948,
+      "step": 34710
+    },
+    {
+      "epoch": 128.59,
+      "grad_norm": 0.9698881506919861,
+      "learning_rate": 0.0007140740740740741,
+      "loss": 0.935,
+      "step": 34720
+    },
+    {
+      "epoch": 128.63,
+      "grad_norm": 1.0098685026168823,
+      "learning_rate": 0.0007137037037037037,
+      "loss": 0.9341,
+      "step": 34730
+    },
+    {
+      "epoch": 128.67,
+      "grad_norm": 1.0363351106643677,
+      "learning_rate": 0.0007133333333333334,
+      "loss": 0.9442,
+      "step": 34740
+    },
+    {
+      "epoch": 128.7,
+      "grad_norm": 1.1238548755645752,
+      "learning_rate": 0.0007129629629629629,
+      "loss": 0.9577,
+      "step": 34750
+    },
+    {
+      "epoch": 128.74,
+      "grad_norm": 1.1039295196533203,
+      "learning_rate": 0.0007125925925925926,
+      "loss": 0.9572,
+      "step": 34760
+    },
+    {
+      "epoch": 128.78,
+      "grad_norm": 1.0145635604858398,
+      "learning_rate": 0.0007122222222222222,
+      "loss": 0.9531,
+      "step": 34770
+    },
+    {
+      "epoch": 128.81,
+      "grad_norm": 1.1182947158813477,
+      "learning_rate": 0.0007118518518518519,
+      "loss": 0.9555,
+      "step": 34780
+    },
+    {
+      "epoch": 128.85,
+      "grad_norm": 1.1297719478607178,
+      "learning_rate": 0.0007114814814814816,
+      "loss": 0.9767,
+      "step": 34790
+    },
+    {
+      "epoch": 128.89,
+      "grad_norm": 1.0659856796264648,
+      "learning_rate": 0.0007111111111111111,
+      "loss": 0.9844,
+      "step": 34800
+    },
+    {
+      "epoch": 128.93,
+      "grad_norm": 1.1524667739868164,
+      "learning_rate": 0.0007107407407407408,
+      "loss": 1.002,
+      "step": 34810
+    },
+    {
+      "epoch": 128.96,
+      "grad_norm": 1.0247752666473389,
+      "learning_rate": 0.0007103703703703703,
+      "loss": 0.9854,
+      "step": 34820
+    },
+    {
+      "epoch": 129.0,
+      "grad_norm": 1.7865345478057861,
+      "learning_rate": 0.00071,
+      "loss": 0.994,
+      "step": 34830
+    },
+    {
+      "epoch": 129.04,
+      "grad_norm": 0.9138088822364807,
+      "learning_rate": 0.0007096296296296296,
+      "loss": 0.7795,
+      "step": 34840
+    },
+    {
+      "epoch": 129.07,
+      "grad_norm": 1.0572221279144287,
+      "learning_rate": 0.0007092592592592593,
+      "loss": 0.7784,
+      "step": 34850
+    },
+    {
+      "epoch": 129.11,
+      "grad_norm": 0.9977957010269165,
+      "learning_rate": 0.0007088888888888889,
+      "loss": 0.8109,
+      "step": 34860
+    },
+    {
+      "epoch": 129.15,
+      "grad_norm": 0.9493637681007385,
+      "learning_rate": 0.0007085185185185186,
+      "loss": 0.8413,
+      "step": 34870
+    },
+    {
+      "epoch": 129.19,
+      "grad_norm": 0.9977490305900574,
+      "learning_rate": 0.0007081481481481482,
+      "loss": 0.8216,
+      "step": 34880
+    },
+    {
+      "epoch": 129.22,
+      "grad_norm": 1.0183244943618774,
+      "learning_rate": 0.0007077777777777778,
+      "loss": 0.8501,
+      "step": 34890
+    },
+    {
+      "epoch": 129.26,
+      "grad_norm": 0.9969905018806458,
+      "learning_rate": 0.0007074074074074074,
+      "loss": 0.8457,
+      "step": 34900
+    },
+    {
+      "epoch": 129.3,
+      "grad_norm": 0.9821087121963501,
+      "learning_rate": 0.000707037037037037,
+      "loss": 0.843,
+      "step": 34910
+    },
+    {
+      "epoch": 129.33,
+      "grad_norm": 1.037099838256836,
+      "learning_rate": 0.0007066666666666666,
+      "loss": 0.874,
+      "step": 34920
+    },
+    {
+      "epoch": 129.37,
+      "grad_norm": 1.114404559135437,
+      "learning_rate": 0.0007062962962962963,
+      "loss": 0.8413,
+      "step": 34930
+    },
+    {
+      "epoch": 129.41,
+      "grad_norm": 1.0478968620300293,
+      "learning_rate": 0.0007059259259259259,
+      "loss": 0.8771,
+      "step": 34940
+    },
+    {
+      "epoch": 129.44,
+      "grad_norm": 1.1025892496109009,
+      "learning_rate": 0.0007055555555555556,
+      "loss": 0.8864,
+      "step": 34950
+    },
+    {
+      "epoch": 129.48,
+      "grad_norm": 1.0482041835784912,
+      "learning_rate": 0.0007051851851851851,
+      "loss": 0.9107,
+      "step": 34960
+    },
+    {
+      "epoch": 129.52,
+      "grad_norm": 1.0418429374694824,
+      "learning_rate": 0.0007048148148148148,
+      "loss": 0.9103,
+      "step": 34970
+    },
+    {
+      "epoch": 129.56,
+      "grad_norm": 1.1483935117721558,
+      "learning_rate": 0.0007044444444444445,
+      "loss": 0.9035,
+      "step": 34980
+    },
+    {
+      "epoch": 129.59,
+      "grad_norm": 1.089949131011963,
+      "learning_rate": 0.0007040740740740741,
+      "loss": 0.9068,
+      "step": 34990
+    },
+    {
+      "epoch": 129.63,
+      "grad_norm": 1.081794261932373,
+      "learning_rate": 0.0007037037037037038,
+      "loss": 0.9361,
+      "step": 35000
+    },
+    {
+      "epoch": 129.67,
+      "grad_norm": 1.0955454111099243,
+      "learning_rate": 0.0007033333333333334,
+      "loss": 0.933,
+      "step": 35010
+    },
+    {
+      "epoch": 129.7,
+      "grad_norm": 1.129664659500122,
+      "learning_rate": 0.0007029629629629631,
+      "loss": 0.9376,
+      "step": 35020
+    },
+    {
+      "epoch": 129.74,
+      "grad_norm": 1.086525559425354,
+      "learning_rate": 0.0007025925925925925,
+      "loss": 0.9453,
+      "step": 35030
+    },
+    {
+      "epoch": 129.78,
+      "grad_norm": 0.9380663633346558,
+      "learning_rate": 0.0007022222222222222,
+      "loss": 0.9583,
+      "step": 35040
+    },
+    {
+      "epoch": 129.81,
+      "grad_norm": 1.0038551092147827,
+      "learning_rate": 0.0007018518518518518,
+      "loss": 0.9623,
+      "step": 35050
+    },
+    {
+      "epoch": 129.85,
+      "grad_norm": 1.0711705684661865,
+      "learning_rate": 0.0007014814814814815,
+      "loss": 0.9572,
+      "step": 35060
+    },
+    {
+      "epoch": 129.89,
+      "grad_norm": 1.134537696838379,
+      "learning_rate": 0.0007011111111111111,
+      "loss": 0.9686,
+      "step": 35070
+    },
+    {
+      "epoch": 129.93,
+      "grad_norm": 1.0236729383468628,
+      "learning_rate": 0.0007007407407407408,
+      "loss": 0.9866,
+      "step": 35080
+    },
+    {
+      "epoch": 129.96,
+      "grad_norm": 1.0966039896011353,
+      "learning_rate": 0.0007003703703703704,
+      "loss": 0.9742,
+      "step": 35090
+    },
+    {
+      "epoch": 130.0,
+      "grad_norm": 1.7335938215255737,
+      "learning_rate": 0.0007,
+      "loss": 0.9832,
+      "step": 35100
+    },
+    {
+      "epoch": 130.04,
+      "grad_norm": 0.9678661823272705,
+      "learning_rate": 0.0006996296296296296,
+      "loss": 0.7645,
+      "step": 35110
+    },
+    {
+      "epoch": 130.07,
+      "grad_norm": 0.9778347015380859,
+      "learning_rate": 0.0006992592592592593,
+      "loss": 0.7927,
+      "step": 35120
+    },
+    {
+      "epoch": 130.11,
+      "grad_norm": 1.0727163553237915,
+      "learning_rate": 0.0006988888888888889,
+      "loss": 0.8129,
+      "step": 35130
+    },
+    {
+      "epoch": 130.15,
+      "grad_norm": 1.0289082527160645,
+      "learning_rate": 0.0006985185185185186,
+      "loss": 0.8036,
+      "step": 35140
+    },
+    {
+      "epoch": 130.19,
+      "grad_norm": 1.00894033908844,
+      "learning_rate": 0.0006981481481481482,
+      "loss": 0.8088,
+      "step": 35150
+    },
+    {
+      "epoch": 130.22,
+      "grad_norm": 1.071042776107788,
+      "learning_rate": 0.0006977777777777778,
+      "loss": 0.8325,
+      "step": 35160
+    },
+    {
+      "epoch": 130.26,
+      "grad_norm": 1.1067564487457275,
+      "learning_rate": 0.0006974074074074073,
+      "loss": 0.8388,
+      "step": 35170
+    },
+    {
+      "epoch": 130.3,
+      "grad_norm": 1.0250170230865479,
+      "learning_rate": 0.000697037037037037,
+      "loss": 0.825,
+      "step": 35180
+    },
+    {
+      "epoch": 130.33,
+      "grad_norm": 0.9799749255180359,
+      "learning_rate": 0.0006966666666666667,
+      "loss": 0.8384,
+      "step": 35190
+    },
+    {
+      "epoch": 130.37,
+      "grad_norm": 1.0186576843261719,
+      "learning_rate": 0.0006962962962962963,
+      "loss": 0.8747,
+      "step": 35200
+    },
+    {
+      "epoch": 130.41,
+      "grad_norm": 1.0256609916687012,
+      "learning_rate": 0.000695925925925926,
+      "loss": 0.8986,
+      "step": 35210
+    },
+    {
+      "epoch": 130.44,
+      "grad_norm": 1.1367439031600952,
+      "learning_rate": 0.0006955555555555556,
+      "loss": 0.8918,
+      "step": 35220
+    },
+    {
+      "epoch": 130.48,
+      "grad_norm": 1.050878643989563,
+      "learning_rate": 0.0006951851851851853,
+      "loss": 0.8694,
+      "step": 35230
+    },
+    {
+      "epoch": 130.52,
+      "grad_norm": 1.0891170501708984,
+      "learning_rate": 0.0006948148148148148,
+      "loss": 0.8837,
+      "step": 35240
+    },
+    {
+      "epoch": 130.56,
+      "grad_norm": 1.0573313236236572,
+      "learning_rate": 0.0006944444444444445,
+      "loss": 0.9004,
+      "step": 35250
+    },
+    {
+      "epoch": 130.59,
+      "grad_norm": 1.0353542566299438,
+      "learning_rate": 0.0006940740740740741,
+      "loss": 0.922,
+      "step": 35260
+    },
+    {
+      "epoch": 130.63,
+      "grad_norm": 1.1184134483337402,
+      "learning_rate": 0.0006937037037037038,
+      "loss": 0.924,
+      "step": 35270
+    },
+    {
+      "epoch": 130.67,
+      "grad_norm": 1.0716382265090942,
+      "learning_rate": 0.0006933333333333333,
+      "loss": 0.9273,
+      "step": 35280
+    },
+    {
+      "epoch": 130.7,
+      "grad_norm": 1.0608584880828857,
+      "learning_rate": 0.000692962962962963,
+      "loss": 0.9439,
+      "step": 35290
+    },
+    {
+      "epoch": 130.74,
+      "grad_norm": 1.103045105934143,
+      "learning_rate": 0.0006925925925925925,
+      "loss": 0.9234,
+      "step": 35300
+    },
+    {
+      "epoch": 130.78,
+      "grad_norm": 0.9750839471817017,
+      "learning_rate": 0.0006922222222222222,
+      "loss": 0.926,
+      "step": 35310
+    },
+    {
+      "epoch": 130.81,
+      "grad_norm": 1.008241057395935,
+      "learning_rate": 0.0006918518518518518,
+      "loss": 0.9397,
+      "step": 35320
+    },
+    {
+      "epoch": 130.85,
+      "grad_norm": 1.0332655906677246,
+      "learning_rate": 0.0006914814814814815,
+      "loss": 0.9596,
+      "step": 35330
+    },
+    {
+      "epoch": 130.89,
+      "grad_norm": 1.1135740280151367,
+      "learning_rate": 0.0006911111111111111,
+      "loss": 0.9674,
+      "step": 35340
+    },
+    {
+      "epoch": 130.93,
+      "grad_norm": 1.0512356758117676,
+      "learning_rate": 0.0006907407407407408,
+      "loss": 0.963,
+      "step": 35350
+    },
+    {
+      "epoch": 130.96,
+      "grad_norm": 1.0693310499191284,
+      "learning_rate": 0.0006903703703703704,
+      "loss": 0.9622,
+      "step": 35360
+    },
+    {
+      "epoch": 131.0,
+      "grad_norm": 2.0000877380371094,
+      "learning_rate": 0.00069,
+      "loss": 0.9982,
+      "step": 35370
+    },
+    {
+      "epoch": 131.04,
+      "grad_norm": 1.1226000785827637,
+      "learning_rate": 0.0006896296296296296,
+      "loss": 0.7572,
+      "step": 35380
+    },
+    {
+      "epoch": 131.07,
+      "grad_norm": 1.0355254411697388,
+      "learning_rate": 0.0006892592592592593,
+      "loss": 0.7667,
+      "step": 35390
+    },
+    {
+      "epoch": 131.11,
+      "grad_norm": 0.9900449514389038,
+      "learning_rate": 0.000688888888888889,
+      "loss": 0.7738,
+      "step": 35400
+    },
+    {
+      "epoch": 131.15,
+      "grad_norm": 0.9880581498146057,
+      "learning_rate": 0.0006885185185185185,
+      "loss": 0.7815,
+      "step": 35410
+    },
+    {
+      "epoch": 131.19,
+      "grad_norm": 1.0864583253860474,
+      "learning_rate": 0.0006881481481481482,
+      "loss": 0.7919,
+      "step": 35420
+    },
+    {
+      "epoch": 131.22,
+      "grad_norm": 0.9945310354232788,
+      "learning_rate": 0.0006877777777777778,
+      "loss": 0.8051,
+      "step": 35430
+    },
+    {
+      "epoch": 131.26,
+      "grad_norm": 1.0401781797409058,
+      "learning_rate": 0.0006874074074074074,
+      "loss": 0.8267,
+      "step": 35440
+    },
+    {
+      "epoch": 131.3,
+      "grad_norm": 1.097060203552246,
+      "learning_rate": 0.000687037037037037,
+      "loss": 0.8195,
+      "step": 35450
+    },
+    {
+      "epoch": 131.33,
+      "grad_norm": 1.0526179075241089,
+      "learning_rate": 0.0006866666666666667,
+      "loss": 0.8823,
+      "step": 35460
+    },
+    {
+      "epoch": 131.37,
+      "grad_norm": 1.0602213144302368,
+      "learning_rate": 0.0006862962962962963,
+      "loss": 0.8842,
+      "step": 35470
+    },
+    {
+      "epoch": 131.41,
+      "grad_norm": 1.0021635293960571,
+      "learning_rate": 0.000685925925925926,
+      "loss": 0.8676,
+      "step": 35480
+    },
+    {
+      "epoch": 131.44,
+      "grad_norm": 1.0285704135894775,
+      "learning_rate": 0.0006855555555555556,
+      "loss": 0.882,
+      "step": 35490
+    },
+    {
+      "epoch": 131.48,
+      "grad_norm": 1.020806074142456,
+      "learning_rate": 0.0006851851851851853,
+      "loss": 0.924,
+      "step": 35500
+    },
+    {
+      "epoch": 131.52,
+      "grad_norm": 1.071965217590332,
+      "learning_rate": 0.0006848148148148147,
+      "loss": 0.8934,
+      "step": 35510
+    },
+    {
+      "epoch": 131.56,
+      "grad_norm": 1.0444443225860596,
+      "learning_rate": 0.0006844444444444444,
+      "loss": 0.9157,
+      "step": 35520
+    },
+    {
+      "epoch": 131.59,
+      "grad_norm": 1.0531617403030396,
+      "learning_rate": 0.000684074074074074,
+      "loss": 0.8889,
+      "step": 35530
+    },
+    {
+      "epoch": 131.63,
+      "grad_norm": 1.088879942893982,
+      "learning_rate": 0.0006837037037037037,
+      "loss": 0.8971,
+      "step": 35540
+    },
+    {
+      "epoch": 131.67,
+      "grad_norm": 1.0203979015350342,
+      "learning_rate": 0.0006833333333333333,
+      "loss": 0.9109,
+      "step": 35550
+    },
+    {
+      "epoch": 131.7,
+      "grad_norm": 1.16692054271698,
+      "learning_rate": 0.000682962962962963,
+      "loss": 0.8981,
+      "step": 35560
+    },
+    {
+      "epoch": 131.74,
+      "grad_norm": 1.067527174949646,
+      "learning_rate": 0.0006825925925925926,
+      "loss": 0.9372,
+      "step": 35570
+    },
+    {
+      "epoch": 131.78,
+      "grad_norm": 1.068550944328308,
+      "learning_rate": 0.0006822222222222222,
+      "loss": 0.9395,
+      "step": 35580
+    },
+    {
+      "epoch": 131.81,
+      "grad_norm": 1.1175237894058228,
+      "learning_rate": 0.0006818518518518518,
+      "loss": 0.943,
+      "step": 35590
+    },
+    {
+      "epoch": 131.85,
+      "grad_norm": 1.0925233364105225,
+      "learning_rate": 0.0006814814814814815,
+      "loss": 0.9499,
+      "step": 35600
+    },
+    {
+      "epoch": 131.89,
+      "grad_norm": 1.0584123134613037,
+      "learning_rate": 0.0006811111111111112,
+      "loss": 0.9339,
+      "step": 35610
+    },
+    {
+      "epoch": 131.93,
+      "grad_norm": 1.250260591506958,
+      "learning_rate": 0.0006807407407407408,
+      "loss": 0.9787,
+      "step": 35620
+    },
+    {
+      "epoch": 131.96,
+      "grad_norm": 1.0699496269226074,
+      "learning_rate": 0.0006803703703703705,
+      "loss": 0.965,
+      "step": 35630
+    },
+    {
+      "epoch": 132.0,
+      "grad_norm": 1.947646975517273,
+      "learning_rate": 0.00068,
+      "loss": 0.9718,
+      "step": 35640
+    },
+    {
+      "epoch": 132.04,
+      "grad_norm": 1.0208699703216553,
+      "learning_rate": 0.0006796296296296296,
+      "loss": 0.7436,
+      "step": 35650
+    },
+    {
+      "epoch": 132.07,
+      "grad_norm": 1.0202369689941406,
+      "learning_rate": 0.0006792592592592592,
+      "loss": 0.7711,
+      "step": 35660
+    },
+    {
+      "epoch": 132.11,
+      "grad_norm": 1.000402569770813,
+      "learning_rate": 0.0006788888888888889,
+      "loss": 0.7827,
+      "step": 35670
+    },
+    {
+      "epoch": 132.15,
+      "grad_norm": 0.9595361948013306,
+      "learning_rate": 0.0006785185185185185,
+      "loss": 0.7905,
+      "step": 35680
+    },
+    {
+      "epoch": 132.19,
+      "grad_norm": 1.0671266317367554,
+      "learning_rate": 0.0006781481481481482,
+      "loss": 0.8049,
+      "step": 35690
+    },
+    {
+      "epoch": 132.22,
+      "grad_norm": 1.0414221286773682,
+      "learning_rate": 0.0006777777777777778,
+      "loss": 0.8074,
+      "step": 35700
+    },
+    {
+      "epoch": 132.26,
+      "grad_norm": 1.0275877714157104,
+      "learning_rate": 0.0006774074074074075,
+      "loss": 0.8105,
+      "step": 35710
+    },
+    {
+      "epoch": 132.3,
+      "grad_norm": 1.0704302787780762,
+      "learning_rate": 0.000677037037037037,
+      "loss": 0.8278,
+      "step": 35720
+    },
+    {
+      "epoch": 132.33,
+      "grad_norm": 1.0970571041107178,
+      "learning_rate": 0.0006766666666666667,
+      "loss": 0.8678,
+      "step": 35730
+    },
+    {
+      "epoch": 132.37,
+      "grad_norm": 1.033130168914795,
+      "learning_rate": 0.0006762962962962963,
+      "loss": 0.8492,
+      "step": 35740
+    },
+    {
+      "epoch": 132.41,
+      "grad_norm": 1.0589123964309692,
+      "learning_rate": 0.000675925925925926,
+      "loss": 0.8804,
+      "step": 35750
+    },
+    {
+      "epoch": 132.44,
+      "grad_norm": 1.0511099100112915,
+      "learning_rate": 0.0006755555555555555,
+      "loss": 0.869,
+      "step": 35760
+    },
+    {
+      "epoch": 132.48,
+      "grad_norm": 1.077571153640747,
+      "learning_rate": 0.0006751851851851852,
+      "loss": 0.863,
+      "step": 35770
+    },
+    {
+      "epoch": 132.52,
+      "grad_norm": 1.062020182609558,
+      "learning_rate": 0.0006748148148148148,
+      "loss": 0.8803,
+      "step": 35780
+    },
+    {
+      "epoch": 132.56,
+      "grad_norm": 1.0527825355529785,
+      "learning_rate": 0.0006744444444444444,
+      "loss": 0.9096,
+      "step": 35790
+    },
+    {
+      "epoch": 132.59,
+      "grad_norm": 1.0730772018432617,
+      "learning_rate": 0.0006740740740740741,
+      "loss": 0.886,
+      "step": 35800
+    },
+    {
+      "epoch": 132.63,
+      "grad_norm": 1.022678017616272,
+      "learning_rate": 0.0006737037037037037,
+      "loss": 0.9041,
+      "step": 35810
+    },
+    {
+      "epoch": 132.67,
+      "grad_norm": 1.1468278169631958,
+      "learning_rate": 0.0006733333333333334,
+      "loss": 0.9103,
+      "step": 35820
+    },
+    {
+      "epoch": 132.7,
+      "grad_norm": 1.0276800394058228,
+      "learning_rate": 0.000672962962962963,
+      "loss": 0.899,
+      "step": 35830
+    },
+    {
+      "epoch": 132.74,
+      "grad_norm": 1.0555604696273804,
+      "learning_rate": 0.0006725925925925927,
+      "loss": 0.9011,
+      "step": 35840
+    },
+    {
+      "epoch": 132.78,
+      "grad_norm": 1.079404354095459,
+      "learning_rate": 0.0006722222222222223,
+      "loss": 0.9099,
+      "step": 35850
+    },
+    {
+      "epoch": 132.81,
+      "grad_norm": 1.0517371892929077,
+      "learning_rate": 0.0006718518518518519,
+      "loss": 0.9586,
+      "step": 35860
+    },
+    {
+      "epoch": 132.85,
+      "grad_norm": 1.044086217880249,
+      "learning_rate": 0.0006714814814814815,
+      "loss": 0.9198,
+      "step": 35870
+    },
+    {
+      "epoch": 132.89,
+      "grad_norm": 1.0771526098251343,
+      "learning_rate": 0.0006711111111111111,
+      "loss": 0.953,
+      "step": 35880
+    },
+    {
+      "epoch": 132.93,
+      "grad_norm": 1.1051455736160278,
+      "learning_rate": 0.0006707407407407407,
+      "loss": 0.9572,
+      "step": 35890
+    },
+    {
+      "epoch": 132.96,
+      "grad_norm": 1.1288875341415405,
+      "learning_rate": 0.0006703703703703704,
+      "loss": 0.9377,
+      "step": 35900
+    },
+    {
+      "epoch": 133.0,
+      "grad_norm": 1.9927622079849243,
+      "learning_rate": 0.00067,
+      "loss": 0.9691,
+      "step": 35910
+    },
+    {
+      "epoch": 133.04,
+      "grad_norm": 0.9369173049926758,
+      "learning_rate": 0.0006696296296296296,
+      "loss": 0.7519,
+      "step": 35920
+    },
+    {
+      "epoch": 133.07,
+      "grad_norm": 0.9838842153549194,
+      "learning_rate": 0.0006692592592592592,
+      "loss": 0.745,
+      "step": 35930
+    },
+    {
+      "epoch": 133.11,
+      "grad_norm": 1.00220787525177,
+      "learning_rate": 0.0006688888888888889,
+      "loss": 0.7842,
+      "step": 35940
+    },
+    {
+      "epoch": 133.15,
+      "grad_norm": 1.036676287651062,
+      "learning_rate": 0.0006685185185185185,
+      "loss": 0.7638,
+      "step": 35950
+    },
+    {
+      "epoch": 133.19,
+      "grad_norm": 1.0177491903305054,
+      "learning_rate": 0.0006681481481481482,
+      "loss": 0.7857,
+      "step": 35960
+    },
+    {
+      "epoch": 133.22,
+      "grad_norm": 1.0435868501663208,
+      "learning_rate": 0.0006677777777777778,
+      "loss": 0.799,
+      "step": 35970
+    },
+    {
+      "epoch": 133.26,
+      "grad_norm": 1.0663994550704956,
+      "learning_rate": 0.0006674074074074075,
+      "loss": 0.8105,
+      "step": 35980
+    },
+    {
+      "epoch": 133.3,
+      "grad_norm": 1.102872610092163,
+      "learning_rate": 0.000667037037037037,
+      "loss": 0.8117,
+      "step": 35990
+    },
+    {
+      "epoch": 133.33,
+      "grad_norm": 1.109534502029419,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 0.8529,
+      "step": 36000
+    },
+    {
+      "epoch": 133.37,
+      "grad_norm": 1.137069582939148,
+      "learning_rate": 0.0006662962962962963,
+      "loss": 0.8337,
+      "step": 36010
+    },
+    {
+      "epoch": 133.41,
+      "grad_norm": 1.0792121887207031,
+      "learning_rate": 0.0006659259259259259,
+      "loss": 0.8385,
+      "step": 36020
+    },
+    {
+      "epoch": 133.44,
+      "grad_norm": 1.0801819562911987,
+      "learning_rate": 0.0006655555555555556,
+      "loss": 0.8567,
+      "step": 36030
+    },
+    {
+      "epoch": 133.48,
+      "grad_norm": 1.0681891441345215,
+      "learning_rate": 0.0006651851851851852,
+      "loss": 0.8468,
+      "step": 36040
+    },
+    {
+      "epoch": 133.52,
+      "grad_norm": 1.0811008214950562,
+      "learning_rate": 0.0006648148148148149,
+      "loss": 0.872,
+      "step": 36050
+    },
+    {
+      "epoch": 133.56,
+      "grad_norm": 1.0719008445739746,
+      "learning_rate": 0.0006644444444444444,
+      "loss": 0.8814,
+      "step": 36060
+    },
+    {
+      "epoch": 133.59,
+      "grad_norm": 1.0063518285751343,
+      "learning_rate": 0.0006640740740740741,
+      "loss": 0.8777,
+      "step": 36070
+    },
+    {
+      "epoch": 133.63,
+      "grad_norm": 1.0991443395614624,
+      "learning_rate": 0.0006637037037037037,
+      "loss": 0.8806,
+      "step": 36080
+    },
+    {
+      "epoch": 133.67,
+      "grad_norm": 1.1498854160308838,
+      "learning_rate": 0.0006633333333333334,
+      "loss": 0.9285,
+      "step": 36090
+    },
+    {
+      "epoch": 133.7,
+      "grad_norm": 1.061549425125122,
+      "learning_rate": 0.000662962962962963,
+      "loss": 0.9263,
+      "step": 36100
+    },
+    {
+      "epoch": 133.74,
+      "grad_norm": 1.0795164108276367,
+      "learning_rate": 0.0006625925925925927,
+      "loss": 0.9244,
+      "step": 36110
+    },
+    {
+      "epoch": 133.78,
+      "grad_norm": 1.093467116355896,
+      "learning_rate": 0.0006622222222222222,
+      "loss": 0.9297,
+      "step": 36120
+    },
+    {
+      "epoch": 133.81,
+      "grad_norm": 1.2153167724609375,
+      "learning_rate": 0.0006618518518518518,
+      "loss": 0.9381,
+      "step": 36130
+    },
+    {
+      "epoch": 133.85,
+      "grad_norm": 1.0366630554199219,
+      "learning_rate": 0.0006614814814814814,
+      "loss": 0.9309,
+      "step": 36140
+    },
+    {
+      "epoch": 133.89,
+      "grad_norm": 1.1849980354309082,
+      "learning_rate": 0.0006611111111111111,
+      "loss": 0.9296,
+      "step": 36150
+    },
+    {
+      "epoch": 133.93,
+      "grad_norm": 1.1084529161453247,
+      "learning_rate": 0.0006607407407407407,
+      "loss": 0.9504,
+      "step": 36160
+    },
+    {
+      "epoch": 133.96,
+      "grad_norm": 1.0794782638549805,
+      "learning_rate": 0.0006603703703703704,
+      "loss": 0.9344,
+      "step": 36170
+    },
+    {
+      "epoch": 134.0,
+      "grad_norm": 1.9394818544387817,
+      "learning_rate": 0.00066,
+      "loss": 0.9645,
+      "step": 36180
+    },
+    {
+      "epoch": 134.04,
+      "grad_norm": 1.1422300338745117,
+      "learning_rate": 0.0006596296296296297,
+      "loss": 0.7624,
+      "step": 36190
+    },
+    {
+      "epoch": 134.07,
+      "grad_norm": 1.0502318143844604,
+      "learning_rate": 0.0006592592592592592,
+      "loss": 0.745,
+      "step": 36200
+    },
+    {
+      "epoch": 134.11,
+      "grad_norm": 0.9741052389144897,
+      "learning_rate": 0.0006588888888888889,
+      "loss": 0.7667,
+      "step": 36210
+    },
+    {
+      "epoch": 134.15,
+      "grad_norm": 1.0162742137908936,
+      "learning_rate": 0.0006585185185185186,
+      "loss": 0.7539,
+      "step": 36220
+    },
+    {
+      "epoch": 134.19,
+      "grad_norm": 1.1134952306747437,
+      "learning_rate": 0.0006581481481481482,
+      "loss": 0.7852,
+      "step": 36230
+    },
+    {
+      "epoch": 134.22,
+      "grad_norm": 1.068518877029419,
+      "learning_rate": 0.0006577777777777779,
+      "loss": 0.8275,
+      "step": 36240
+    },
+    {
+      "epoch": 134.26,
+      "grad_norm": 1.0402024984359741,
+      "learning_rate": 0.0006574074074074074,
+      "loss": 0.8097,
+      "step": 36250
+    },
+    {
+      "epoch": 134.3,
+      "grad_norm": 1.0597301721572876,
+      "learning_rate": 0.0006570370370370371,
+      "loss": 0.8198,
+      "step": 36260
+    },
+    {
+      "epoch": 134.33,
+      "grad_norm": 1.1437287330627441,
+      "learning_rate": 0.0006566666666666666,
+      "loss": 0.8308,
+      "step": 36270
+    },
+    {
+      "epoch": 134.37,
+      "grad_norm": 1.0142033100128174,
+      "learning_rate": 0.0006562962962962963,
+      "loss": 0.8471,
+      "step": 36280
+    },
+    {
+      "epoch": 134.41,
+      "grad_norm": 1.10916268825531,
+      "learning_rate": 0.0006559259259259259,
+      "loss": 0.8532,
+      "step": 36290
+    },
+    {
+      "epoch": 134.44,
+      "grad_norm": 1.0391181707382202,
+      "learning_rate": 0.0006555555555555556,
+      "loss": 0.855,
+      "step": 36300
+    },
+    {
+      "epoch": 134.48,
+      "grad_norm": 1.0528442859649658,
+      "learning_rate": 0.0006551851851851852,
+      "loss": 0.8665,
+      "step": 36310
+    },
+    {
+      "epoch": 134.52,
+      "grad_norm": 1.035757064819336,
+      "learning_rate": 0.0006548148148148149,
+      "loss": 0.8671,
+      "step": 36320
+    },
+    {
+      "epoch": 134.56,
+      "grad_norm": 1.064133644104004,
+      "learning_rate": 0.0006544444444444445,
+      "loss": 0.8735,
+      "step": 36330
+    },
+    {
+      "epoch": 134.59,
+      "grad_norm": 1.0806165933609009,
+      "learning_rate": 0.0006540740740740741,
+      "loss": 0.8761,
+      "step": 36340
+    },
+    {
+      "epoch": 134.63,
+      "grad_norm": 1.0308071374893188,
+      "learning_rate": 0.0006537037037037037,
+      "loss": 0.8762,
+      "step": 36350
+    },
+    {
+      "epoch": 134.67,
+      "grad_norm": 1.1964023113250732,
+      "learning_rate": 0.0006533333333333333,
+      "loss": 0.8909,
+      "step": 36360
+    },
+    {
+      "epoch": 134.7,
+      "grad_norm": 1.065822958946228,
+      "learning_rate": 0.0006529629629629629,
+      "loss": 0.9089,
+      "step": 36370
+    },
+    {
+      "epoch": 134.74,
+      "grad_norm": 1.0590929985046387,
+      "learning_rate": 0.0006525925925925926,
+      "loss": 0.8952,
+      "step": 36380
+    },
+    {
+      "epoch": 134.78,
+      "grad_norm": 1.0043385028839111,
+      "learning_rate": 0.0006522222222222222,
+      "loss": 0.8984,
+      "step": 36390
+    },
+    {
+      "epoch": 134.81,
+      "grad_norm": 1.0728081464767456,
+      "learning_rate": 0.0006518518518518519,
+      "loss": 0.9091,
+      "step": 36400
+    },
+    {
+      "epoch": 134.85,
+      "grad_norm": 1.0429506301879883,
+      "learning_rate": 0.0006514814814814814,
+      "loss": 0.9294,
+      "step": 36410
+    },
+    {
+      "epoch": 134.89,
+      "grad_norm": 1.0224993228912354,
+      "learning_rate": 0.0006511111111111111,
+      "loss": 0.9178,
+      "step": 36420
+    },
+    {
+      "epoch": 134.93,
+      "grad_norm": 1.0892304182052612,
+      "learning_rate": 0.0006507407407407408,
+      "loss": 0.9257,
+      "step": 36430
+    },
+    {
+      "epoch": 134.96,
+      "grad_norm": 1.1427512168884277,
+      "learning_rate": 0.0006503703703703704,
+      "loss": 0.914,
+      "step": 36440
+    },
+    {
+      "epoch": 135.0,
+      "grad_norm": 1.7460381984710693,
+      "learning_rate": 0.0006500000000000001,
+      "loss": 0.921,
+      "step": 36450
+    },
+    {
+      "epoch": 135.04,
+      "grad_norm": 1.1062042713165283,
+      "learning_rate": 0.0006496296296296297,
+      "loss": 0.7485,
+      "step": 36460
+    },
+    {
+      "epoch": 135.07,
+      "grad_norm": 0.9701430201530457,
+      "learning_rate": 0.0006492592592592594,
+      "loss": 0.7171,
+      "step": 36470
+    },
+    {
+      "epoch": 135.11,
+      "grad_norm": 0.9689076542854309,
+      "learning_rate": 0.0006488888888888888,
+      "loss": 0.7617,
+      "step": 36480
+    },
+    {
+      "epoch": 135.15,
+      "grad_norm": 1.0703787803649902,
+      "learning_rate": 0.0006485185185185185,
+      "loss": 0.7717,
+      "step": 36490
+    },
+    {
+      "epoch": 135.19,
+      "grad_norm": 1.0255299806594849,
+      "learning_rate": 0.0006481481481481481,
+      "loss": 0.7933,
+      "step": 36500
+    },
+    {
+      "epoch": 135.22,
+      "grad_norm": 1.018778681755066,
+      "learning_rate": 0.0006477777777777778,
+      "loss": 0.8015,
+      "step": 36510
+    },
+    {
+      "epoch": 135.26,
+      "grad_norm": 1.0683023929595947,
+      "learning_rate": 0.0006474074074074074,
+      "loss": 0.8079,
+      "step": 36520
+    },
+    {
+      "epoch": 135.3,
+      "grad_norm": 1.0992554426193237,
+      "learning_rate": 0.0006470370370370371,
+      "loss": 0.8185,
+      "step": 36530
+    },
+    {
+      "epoch": 135.33,
+      "grad_norm": 1.0614908933639526,
+      "learning_rate": 0.0006466666666666666,
+      "loss": 0.8257,
+      "step": 36540
+    },
+    {
+      "epoch": 135.37,
+      "grad_norm": 1.0698963403701782,
+      "learning_rate": 0.0006462962962962963,
+      "loss": 0.825,
+      "step": 36550
+    },
+    {
+      "epoch": 135.41,
+      "grad_norm": 1.0651202201843262,
+      "learning_rate": 0.0006459259259259259,
+      "loss": 0.8319,
+      "step": 36560
+    },
+    {
+      "epoch": 135.44,
+      "grad_norm": 1.0973514318466187,
+      "learning_rate": 0.0006455555555555556,
+      "loss": 0.8525,
+      "step": 36570
+    },
+    {
+      "epoch": 135.48,
+      "grad_norm": 1.115716814994812,
+      "learning_rate": 0.0006451851851851852,
+      "loss": 0.8737,
+      "step": 36580
+    },
+    {
+      "epoch": 135.52,
+      "grad_norm": 1.034720540046692,
+      "learning_rate": 0.0006448148148148149,
+      "loss": 0.8434,
+      "step": 36590
+    },
+    {
+      "epoch": 135.56,
+      "grad_norm": 1.1401606798171997,
+      "learning_rate": 0.0006444444444444444,
+      "loss": 0.8523,
+      "step": 36600
+    },
+    {
+      "epoch": 135.59,
+      "grad_norm": 1.0871201753616333,
+      "learning_rate": 0.000644074074074074,
+      "loss": 0.8677,
+      "step": 36610
+    },
+    {
+      "epoch": 135.63,
+      "grad_norm": 1.0409997701644897,
+      "learning_rate": 0.0006437037037037037,
+      "loss": 0.8639,
+      "step": 36620
+    },
+    {
+      "epoch": 135.67,
+      "grad_norm": 1.1030999422073364,
+      "learning_rate": 0.0006433333333333333,
+      "loss": 0.8791,
+      "step": 36630
+    },
+    {
+      "epoch": 135.7,
+      "grad_norm": 1.0416744947433472,
+      "learning_rate": 0.000642962962962963,
+      "loss": 0.8811,
+      "step": 36640
+    },
+    {
+      "epoch": 135.74,
+      "grad_norm": 1.1191257238388062,
+      "learning_rate": 0.0006425925925925926,
+      "loss": 0.8853,
+      "step": 36650
+    },
+    {
+      "epoch": 135.78,
+      "grad_norm": 1.0980498790740967,
+      "learning_rate": 0.0006422222222222223,
+      "loss": 0.8938,
+      "step": 36660
+    },
+    {
+      "epoch": 135.81,
+      "grad_norm": 1.062381625175476,
+      "learning_rate": 0.0006418518518518519,
+      "loss": 0.8988,
+      "step": 36670
+    },
+    {
+      "epoch": 135.85,
+      "grad_norm": 1.1369181871414185,
+      "learning_rate": 0.0006414814814814815,
+      "loss": 0.8878,
+      "step": 36680
+    },
+    {
+      "epoch": 135.89,
+      "grad_norm": 1.0846474170684814,
+      "learning_rate": 0.0006411111111111111,
+      "loss": 0.9399,
+      "step": 36690
+    },
+    {
+      "epoch": 135.93,
+      "grad_norm": 1.0928053855895996,
+      "learning_rate": 0.0006407407407407408,
+      "loss": 0.9173,
+      "step": 36700
+    },
+    {
+      "epoch": 135.96,
+      "grad_norm": 1.0402101278305054,
+      "learning_rate": 0.0006403703703703704,
+      "loss": 0.9408,
+      "step": 36710
+    },
+    {
+      "epoch": 136.0,
+      "grad_norm": 2.5675206184387207,
+      "learning_rate": 0.00064,
+      "loss": 0.9232,
+      "step": 36720
+    },
+    {
+      "epoch": 136.04,
+      "grad_norm": 0.9730777740478516,
+      "learning_rate": 0.0006396296296296296,
+      "loss": 0.7415,
+      "step": 36730
+    },
+    {
+      "epoch": 136.07,
+      "grad_norm": 0.9571381211280823,
+      "learning_rate": 0.0006392592592592593,
+      "loss": 0.749,
+      "step": 36740
+    },
+    {
+      "epoch": 136.11,
+      "grad_norm": 1.0041247606277466,
+      "learning_rate": 0.0006388888888888888,
+      "loss": 0.7541,
+      "step": 36750
+    },
+    {
+      "epoch": 136.15,
+      "grad_norm": 1.0825740098953247,
+      "learning_rate": 0.0006385185185185185,
+      "loss": 0.7562,
+      "step": 36760
+    },
+    {
+      "epoch": 136.19,
+      "grad_norm": 0.9825723767280579,
+      "learning_rate": 0.0006381481481481481,
+      "loss": 0.7826,
+      "step": 36770
+    },
+    {
+      "epoch": 136.22,
+      "grad_norm": 1.0305911302566528,
+      "learning_rate": 0.0006377777777777778,
+      "loss": 0.7847,
+      "step": 36780
+    },
+    {
+      "epoch": 136.26,
+      "grad_norm": 1.0714701414108276,
+      "learning_rate": 0.0006374074074074074,
+      "loss": 0.7714,
+      "step": 36790
+    },
+    {
+      "epoch": 136.3,
+      "grad_norm": 1.0887693166732788,
+      "learning_rate": 0.0006370370370370371,
+      "loss": 0.8174,
+      "step": 36800
+    },
+    {
+      "epoch": 136.33,
+      "grad_norm": 1.063496470451355,
+      "learning_rate": 0.0006366666666666667,
+      "loss": 0.8019,
+      "step": 36810
+    },
+    {
+      "epoch": 136.37,
+      "grad_norm": 1.1221165657043457,
+      "learning_rate": 0.0006362962962962963,
+      "loss": 0.8354,
+      "step": 36820
+    },
+    {
+      "epoch": 136.41,
+      "grad_norm": 1.0420963764190674,
+      "learning_rate": 0.000635925925925926,
+      "loss": 0.838,
+      "step": 36830
+    },
+    {
+      "epoch": 136.44,
+      "grad_norm": 0.9873172640800476,
+      "learning_rate": 0.0006355555555555555,
+      "loss": 0.8367,
+      "step": 36840
+    },
+    {
+      "epoch": 136.48,
+      "grad_norm": 1.1159552335739136,
+      "learning_rate": 0.0006351851851851852,
+      "loss": 0.8414,
+      "step": 36850
+    },
+    {
+      "epoch": 136.52,
+      "grad_norm": 1.1081640720367432,
+      "learning_rate": 0.0006348148148148148,
+      "loss": 0.8466,
+      "step": 36860
+    },
+    {
+      "epoch": 136.56,
+      "grad_norm": 1.1748673915863037,
+      "learning_rate": 0.0006344444444444445,
+      "loss": 0.8403,
+      "step": 36870
+    },
+    {
+      "epoch": 136.59,
+      "grad_norm": 1.0808719396591187,
+      "learning_rate": 0.0006340740740740741,
+      "loss": 0.8689,
+      "step": 36880
+    },
+    {
+      "epoch": 136.63,
+      "grad_norm": 1.0514870882034302,
+      "learning_rate": 0.0006337037037037037,
+      "loss": 0.8875,
+      "step": 36890
+    },
+    {
+      "epoch": 136.67,
+      "grad_norm": 1.1719036102294922,
+      "learning_rate": 0.0006333333333333333,
+      "loss": 0.8559,
+      "step": 36900
+    },
+    {
+      "epoch": 136.7,
+      "grad_norm": 1.1026197671890259,
+      "learning_rate": 0.000632962962962963,
+      "loss": 0.8947,
+      "step": 36910
+    },
+    {
+      "epoch": 136.74,
+      "grad_norm": 1.146915316581726,
+      "learning_rate": 0.0006325925925925926,
+      "loss": 0.8931,
+      "step": 36920
+    },
+    {
+      "epoch": 136.78,
+      "grad_norm": 1.1203457117080688,
+      "learning_rate": 0.0006322222222222223,
+      "loss": 0.9019,
+      "step": 36930
+    },
+    {
+      "epoch": 136.81,
+      "grad_norm": 1.1791220903396606,
+      "learning_rate": 0.0006318518518518519,
+      "loss": 0.9112,
+      "step": 36940
+    },
+    {
+      "epoch": 136.85,
+      "grad_norm": 1.0644776821136475,
+      "learning_rate": 0.0006314814814814816,
+      "loss": 0.8946,
+      "step": 36950
+    },
+    {
+      "epoch": 136.89,
+      "grad_norm": 1.0956945419311523,
+      "learning_rate": 0.000631111111111111,
+      "loss": 0.9108,
+      "step": 36960
+    },
+    {
+      "epoch": 136.93,
+      "grad_norm": 1.064579725265503,
+      "learning_rate": 0.0006307407407407407,
+      "loss": 0.9043,
+      "step": 36970
+    },
+    {
+      "epoch": 136.96,
+      "grad_norm": 1.117305874824524,
+      "learning_rate": 0.0006303703703703703,
+      "loss": 0.8842,
+      "step": 36980
+    },
+    {
+      "epoch": 137.0,
+      "grad_norm": 2.0160574913024902,
+      "learning_rate": 0.00063,
+      "loss": 0.908,
+      "step": 36990
+    },
+    {
+      "epoch": 137.04,
+      "grad_norm": 1.0514990091323853,
+      "learning_rate": 0.0006296296296296296,
+      "loss": 0.7255,
+      "step": 37000
+    },
+    {
+      "epoch": 137.07,
+      "grad_norm": 1.0121877193450928,
+      "learning_rate": 0.0006292592592592593,
+      "loss": 0.7161,
+      "step": 37010
+    },
+    {
+      "epoch": 137.11,
+      "grad_norm": 1.0525656938552856,
+      "learning_rate": 0.000628888888888889,
+      "loss": 0.7277,
+      "step": 37020
+    },
+    {
+      "epoch": 137.15,
+      "grad_norm": 1.0623517036437988,
+      "learning_rate": 0.0006285185185185185,
+      "loss": 0.7355,
+      "step": 37030
+    },
+    {
+      "epoch": 137.19,
+      "grad_norm": 1.064374327659607,
+      "learning_rate": 0.0006281481481481482,
+      "loss": 0.7413,
+      "step": 37040
+    },
+    {
+      "epoch": 137.22,
+      "grad_norm": 1.085761547088623,
+      "learning_rate": 0.0006277777777777778,
+      "loss": 0.7539,
+      "step": 37050
+    },
+    {
+      "epoch": 137.26,
+      "grad_norm": 1.1157646179199219,
+      "learning_rate": 0.0006274074074074075,
+      "loss": 0.7912,
+      "step": 37060
+    },
+    {
+      "epoch": 137.3,
+      "grad_norm": 1.0635371208190918,
+      "learning_rate": 0.0006270370370370371,
+      "loss": 0.8024,
+      "step": 37070
+    },
+    {
+      "epoch": 137.33,
+      "grad_norm": 1.0670291185379028,
+      "learning_rate": 0.0006266666666666668,
+      "loss": 0.7945,
+      "step": 37080
+    },
+    {
+      "epoch": 137.37,
+      "grad_norm": 1.1312631368637085,
+      "learning_rate": 0.0006262962962962963,
+      "loss": 0.8225,
+      "step": 37090
+    },
+    {
+      "epoch": 137.41,
+      "grad_norm": 1.0298441648483276,
+      "learning_rate": 0.0006259259259259259,
+      "loss": 0.8254,
+      "step": 37100
+    },
+    {
+      "epoch": 137.44,
+      "grad_norm": 1.0292751789093018,
+      "learning_rate": 0.0006255555555555555,
+      "loss": 0.8184,
+      "step": 37110
+    },
+    {
+      "epoch": 137.48,
+      "grad_norm": 1.1706461906433105,
+      "learning_rate": 0.0006251851851851852,
+      "loss": 0.8487,
+      "step": 37120
+    },
+    {
+      "epoch": 137.52,
+      "grad_norm": 1.0919010639190674,
+      "learning_rate": 0.0006248148148148148,
+      "loss": 0.8532,
+      "step": 37130
+    },
+    {
+      "epoch": 137.56,
+      "grad_norm": 1.1142833232879639,
+      "learning_rate": 0.0006244444444444445,
+      "loss": 0.8286,
+      "step": 37140
+    },
+    {
+      "epoch": 137.59,
+      "grad_norm": 1.0280237197875977,
+      "learning_rate": 0.0006240740740740741,
+      "loss": 0.8573,
+      "step": 37150
+    },
+    {
+      "epoch": 137.63,
+      "grad_norm": 1.1540181636810303,
+      "learning_rate": 0.0006237037037037037,
+      "loss": 0.8634,
+      "step": 37160
+    },
+    {
+      "epoch": 137.67,
+      "grad_norm": 1.0904784202575684,
+      "learning_rate": 0.0006233333333333333,
+      "loss": 0.8658,
+      "step": 37170
+    },
+    {
+      "epoch": 137.7,
+      "grad_norm": 1.0962737798690796,
+      "learning_rate": 0.000622962962962963,
+      "loss": 0.872,
+      "step": 37180
+    },
+    {
+      "epoch": 137.74,
+      "grad_norm": 1.0626097917556763,
+      "learning_rate": 0.0006225925925925926,
+      "loss": 0.8977,
+      "step": 37190
+    },
+    {
+      "epoch": 137.78,
+      "grad_norm": 1.079049825668335,
+      "learning_rate": 0.0006222222222222223,
+      "loss": 0.8806,
+      "step": 37200
+    },
+    {
+      "epoch": 137.81,
+      "grad_norm": 1.0811468362808228,
+      "learning_rate": 0.0006218518518518518,
+      "loss": 0.8627,
+      "step": 37210
+    },
+    {
+      "epoch": 137.85,
+      "grad_norm": 1.1173999309539795,
+      "learning_rate": 0.0006214814814814815,
+      "loss": 0.909,
+      "step": 37220
+    },
+    {
+      "epoch": 137.89,
+      "grad_norm": 1.1013028621673584,
+      "learning_rate": 0.000621111111111111,
+      "loss": 0.9217,
+      "step": 37230
+    },
+    {
+      "epoch": 137.93,
+      "grad_norm": 1.2114920616149902,
+      "learning_rate": 0.0006207407407407407,
+      "loss": 0.9109,
+      "step": 37240
+    },
+    {
+      "epoch": 137.96,
+      "grad_norm": 1.1027235984802246,
+      "learning_rate": 0.0006203703703703704,
+      "loss": 0.9372,
+      "step": 37250
+    },
+    {
+      "epoch": 138.0,
+      "grad_norm": 1.8099175691604614,
+      "learning_rate": 0.00062,
+      "loss": 0.9212,
+      "step": 37260
+    },
+    {
+      "epoch": 138.04,
+      "grad_norm": 1.0433896780014038,
+      "learning_rate": 0.0006196296296296297,
+      "loss": 0.7166,
+      "step": 37270
+    },
+    {
+      "epoch": 138.07,
+      "grad_norm": 0.9847230315208435,
+      "learning_rate": 0.0006192592592592593,
+      "loss": 0.7069,
+      "step": 37280
+    },
+    {
+      "epoch": 138.11,
+      "grad_norm": 1.0111587047576904,
+      "learning_rate": 0.000618888888888889,
+      "loss": 0.7178,
+      "step": 37290
+    },
+    {
+      "epoch": 138.15,
+      "grad_norm": 1.0253404378890991,
+      "learning_rate": 0.0006185185185185185,
+      "loss": 0.7347,
+      "step": 37300
+    },
+    {
+      "epoch": 138.19,
+      "grad_norm": 1.0111534595489502,
+      "learning_rate": 0.0006181481481481482,
+      "loss": 0.7663,
+      "step": 37310
+    },
+    {
+      "epoch": 138.22,
+      "grad_norm": 1.0437885522842407,
+      "learning_rate": 0.0006177777777777777,
+      "loss": 0.7502,
+      "step": 37320
+    },
+    {
+      "epoch": 138.26,
+      "grad_norm": 1.1091742515563965,
+      "learning_rate": 0.0006174074074074074,
+      "loss": 0.7719,
+      "step": 37330
+    },
+    {
+      "epoch": 138.3,
+      "grad_norm": 1.1044790744781494,
+      "learning_rate": 0.000617037037037037,
+      "loss": 0.7994,
+      "step": 37340
+    },
+    {
+      "epoch": 138.33,
+      "grad_norm": 1.1266263723373413,
+      "learning_rate": 0.0006166666666666667,
+      "loss": 0.7864,
+      "step": 37350
+    },
+    {
+      "epoch": 138.37,
+      "grad_norm": 1.0152026414871216,
+      "learning_rate": 0.0006162962962962963,
+      "loss": 0.8299,
+      "step": 37360
+    },
+    {
+      "epoch": 138.41,
+      "grad_norm": 1.0453267097473145,
+      "learning_rate": 0.0006159259259259259,
+      "loss": 0.8212,
+      "step": 37370
+    },
+    {
+      "epoch": 138.44,
+      "grad_norm": 1.0323936939239502,
+      "learning_rate": 0.0006155555555555555,
+      "loss": 0.809,
+      "step": 37380
+    },
+    {
+      "epoch": 138.48,
+      "grad_norm": 1.1219218969345093,
+      "learning_rate": 0.0006151851851851852,
+      "loss": 0.8379,
+      "step": 37390
+    },
+    {
+      "epoch": 138.52,
+      "grad_norm": 1.0550750494003296,
+      "learning_rate": 0.0006148148148148148,
+      "loss": 0.8564,
+      "step": 37400
+    },
+    {
+      "epoch": 138.56,
+      "grad_norm": 1.142170786857605,
+      "learning_rate": 0.0006144444444444445,
+      "loss": 0.842,
+      "step": 37410
+    },
+    {
+      "epoch": 138.59,
+      "grad_norm": 1.0784144401550293,
+      "learning_rate": 0.0006140740740740741,
+      "loss": 0.8461,
+      "step": 37420
+    },
+    {
+      "epoch": 138.63,
+      "grad_norm": 0.992065966129303,
+      "learning_rate": 0.0006137037037037038,
+      "loss": 0.856,
+      "step": 37430
+    },
+    {
+      "epoch": 138.67,
+      "grad_norm": 1.1817371845245361,
+      "learning_rate": 0.0006133333333333334,
+      "loss": 0.8397,
+      "step": 37440
+    },
+    {
+      "epoch": 138.7,
+      "grad_norm": 1.1475744247436523,
+      "learning_rate": 0.0006129629629629629,
+      "loss": 0.8616,
+      "step": 37450
+    },
+    {
+      "epoch": 138.74,
+      "grad_norm": 1.1824603080749512,
+      "learning_rate": 0.0006125925925925926,
+      "loss": 0.8667,
+      "step": 37460
+    },
+    {
+      "epoch": 138.78,
+      "grad_norm": 1.0849061012268066,
+      "learning_rate": 0.0006122222222222222,
+      "loss": 0.8609,
+      "step": 37470
+    },
+    {
+      "epoch": 138.81,
+      "grad_norm": 1.088152527809143,
+      "learning_rate": 0.0006118518518518519,
+      "loss": 0.8914,
+      "step": 37480
+    },
+    {
+      "epoch": 138.85,
+      "grad_norm": 1.1588715314865112,
+      "learning_rate": 0.0006114814814814815,
+      "loss": 0.878,
+      "step": 37490
+    },
+    {
+      "epoch": 138.89,
+      "grad_norm": 1.1323331594467163,
+      "learning_rate": 0.0006111111111111112,
+      "loss": 0.8842,
+      "step": 37500
+    },
+    {
+      "epoch": 138.93,
+      "grad_norm": 1.0989532470703125,
+      "learning_rate": 0.0006107407407407407,
+      "loss": 0.9039,
+      "step": 37510
+    },
+    {
+      "epoch": 138.96,
+      "grad_norm": 1.1324613094329834,
+      "learning_rate": 0.0006103703703703704,
+      "loss": 0.9128,
+      "step": 37520
+    },
+    {
+      "epoch": 139.0,
+      "grad_norm": 2.4864327907562256,
+      "learning_rate": 0.00061,
+      "loss": 0.8831,
+      "step": 37530
+    },
+    {
+      "epoch": 139.04,
+      "grad_norm": 0.9940552711486816,
+      "learning_rate": 0.0006096296296296297,
+      "loss": 0.7002,
+      "step": 37540
+    },
+    {
+      "epoch": 139.07,
+      "grad_norm": 1.0477380752563477,
+      "learning_rate": 0.0006092592592592593,
+      "loss": 0.7302,
+      "step": 37550
+    },
+    {
+      "epoch": 139.11,
+      "grad_norm": 1.0235960483551025,
+      "learning_rate": 0.000608888888888889,
+      "loss": 0.7223,
+      "step": 37560
+    },
+    {
+      "epoch": 139.15,
+      "grad_norm": 0.9758222103118896,
+      "learning_rate": 0.0006085185185185185,
+      "loss": 0.7183,
+      "step": 37570
+    },
+    {
+      "epoch": 139.19,
+      "grad_norm": 1.018068552017212,
+      "learning_rate": 0.0006081481481481481,
+      "loss": 0.7327,
+      "step": 37580
+    },
+    {
+      "epoch": 139.22,
+      "grad_norm": 1.0142362117767334,
+      "learning_rate": 0.0006077777777777777,
+      "loss": 0.7574,
+      "step": 37590
+    },
+    {
+      "epoch": 139.26,
+      "grad_norm": 1.1294622421264648,
+      "learning_rate": 0.0006074074074074074,
+      "loss": 0.7817,
+      "step": 37600
+    },
+    {
+      "epoch": 139.3,
+      "grad_norm": 1.1239715814590454,
+      "learning_rate": 0.000607037037037037,
+      "loss": 0.7688,
+      "step": 37610
+    },
+    {
+      "epoch": 139.33,
+      "grad_norm": 1.0944328308105469,
+      "learning_rate": 0.0006066666666666667,
+      "loss": 0.79,
+      "step": 37620
+    },
+    {
+      "epoch": 139.37,
+      "grad_norm": 1.0795550346374512,
+      "learning_rate": 0.0006062962962962963,
+      "loss": 0.7937,
+      "step": 37630
+    },
+    {
+      "epoch": 139.41,
+      "grad_norm": 1.0517427921295166,
+      "learning_rate": 0.000605925925925926,
+      "loss": 0.8055,
+      "step": 37640
+    },
+    {
+      "epoch": 139.44,
+      "grad_norm": 1.1125982999801636,
+      "learning_rate": 0.0006055555555555556,
+      "loss": 0.8028,
+      "step": 37650
+    },
+    {
+      "epoch": 139.48,
+      "grad_norm": 1.0875502824783325,
+      "learning_rate": 0.0006051851851851852,
+      "loss": 0.8102,
+      "step": 37660
+    },
+    {
+      "epoch": 139.52,
+      "grad_norm": 1.1398983001708984,
+      "learning_rate": 0.0006048148148148149,
+      "loss": 0.8346,
+      "step": 37670
+    },
+    {
+      "epoch": 139.56,
+      "grad_norm": 1.16152024269104,
+      "learning_rate": 0.0006044444444444445,
+      "loss": 0.8348,
+      "step": 37680
+    },
+    {
+      "epoch": 139.59,
+      "grad_norm": 1.0522875785827637,
+      "learning_rate": 0.0006040740740740741,
+      "loss": 0.83,
+      "step": 37690
+    },
+    {
+      "epoch": 139.63,
+      "grad_norm": 1.1236652135849,
+      "learning_rate": 0.0006037037037037037,
+      "loss": 0.8631,
+      "step": 37700
+    },
+    {
+      "epoch": 139.67,
+      "grad_norm": 1.0937477350234985,
+      "learning_rate": 0.0006033333333333334,
+      "loss": 0.843,
+      "step": 37710
+    },
+    {
+      "epoch": 139.7,
+      "grad_norm": 1.1400094032287598,
+      "learning_rate": 0.0006029629629629629,
+      "loss": 0.857,
+      "step": 37720
+    },
+    {
+      "epoch": 139.74,
+      "grad_norm": 1.053101658821106,
+      "learning_rate": 0.0006025925925925926,
+      "loss": 0.8821,
+      "step": 37730
+    },
+    {
+      "epoch": 139.78,
+      "grad_norm": 1.1917682886123657,
+      "learning_rate": 0.0006022222222222222,
+      "loss": 0.8669,
+      "step": 37740
+    },
+    {
+      "epoch": 139.81,
+      "grad_norm": 1.0519508123397827,
+      "learning_rate": 0.0006018518518518519,
+      "loss": 0.8726,
+      "step": 37750
+    },
+    {
+      "epoch": 139.85,
+      "grad_norm": 1.091275691986084,
+      "learning_rate": 0.0006014814814814815,
+      "loss": 0.9034,
+      "step": 37760
+    },
+    {
+      "epoch": 139.89,
+      "grad_norm": 1.154052734375,
+      "learning_rate": 0.0006011111111111112,
+      "loss": 0.874,
+      "step": 37770
+    },
+    {
+      "epoch": 139.93,
+      "grad_norm": 1.1093419790267944,
+      "learning_rate": 0.0006007407407407407,
+      "loss": 0.87,
+      "step": 37780
+    },
+    {
+      "epoch": 139.96,
+      "grad_norm": 1.100222110748291,
+      "learning_rate": 0.0006003703703703704,
+      "loss": 0.8778,
+      "step": 37790
+    },
+    {
+      "epoch": 140.0,
+      "grad_norm": 1.992299199104309,
+      "learning_rate": 0.0006,
+      "loss": 0.8879,
+      "step": 37800
+    },
+    {
+      "epoch": 140.04,
+      "grad_norm": 0.997854471206665,
+      "learning_rate": 0.0005996296296296296,
+      "loss": 0.6986,
+      "step": 37810
+    },
+    {
+      "epoch": 140.07,
+      "grad_norm": 1.0388472080230713,
+      "learning_rate": 0.0005992592592592592,
+      "loss": 0.7165,
+      "step": 37820
+    },
+    {
+      "epoch": 140.11,
+      "grad_norm": 1.0725175142288208,
+      "learning_rate": 0.0005988888888888889,
+      "loss": 0.7285,
+      "step": 37830
+    },
+    {
+      "epoch": 140.15,
+      "grad_norm": 1.0202758312225342,
+      "learning_rate": 0.0005985185185185186,
+      "loss": 0.74,
+      "step": 37840
+    },
+    {
+      "epoch": 140.19,
+      "grad_norm": 1.115369439125061,
+      "learning_rate": 0.0005981481481481481,
+      "loss": 0.7243,
+      "step": 37850
+    },
+    {
+      "epoch": 140.22,
+      "grad_norm": 1.033743143081665,
+      "learning_rate": 0.0005977777777777778,
+      "loss": 0.7606,
+      "step": 37860
+    },
+    {
+      "epoch": 140.26,
+      "grad_norm": 1.1330405473709106,
+      "learning_rate": 0.0005974074074074074,
+      "loss": 0.746,
+      "step": 37870
+    },
+    {
+      "epoch": 140.3,
+      "grad_norm": 1.0477662086486816,
+      "learning_rate": 0.0005970370370370371,
+      "loss": 0.766,
+      "step": 37880
+    },
+    {
+      "epoch": 140.33,
+      "grad_norm": 1.0449988842010498,
+      "learning_rate": 0.0005966666666666667,
+      "loss": 0.7704,
+      "step": 37890
+    },
+    {
+      "epoch": 140.37,
+      "grad_norm": 1.0974394083023071,
+      "learning_rate": 0.0005962962962962964,
+      "loss": 0.7675,
+      "step": 37900
+    },
+    {
+      "epoch": 140.41,
+      "grad_norm": 1.111845850944519,
+      "learning_rate": 0.000595925925925926,
+      "loss": 0.7752,
+      "step": 37910
+    },
+    {
+      "epoch": 140.44,
+      "grad_norm": 1.2133774757385254,
+      "learning_rate": 0.0005955555555555556,
+      "loss": 0.8017,
+      "step": 37920
+    },
+    {
+      "epoch": 140.48,
+      "grad_norm": 1.1683170795440674,
+      "learning_rate": 0.0005951851851851851,
+      "loss": 0.8009,
+      "step": 37930
+    },
+    {
+      "epoch": 140.52,
+      "grad_norm": 1.181700348854065,
+      "learning_rate": 0.0005948148148148148,
+      "loss": 0.8308,
+      "step": 37940
+    },
+    {
+      "epoch": 140.56,
+      "grad_norm": 1.0495575666427612,
+      "learning_rate": 0.0005944444444444444,
+      "loss": 0.8417,
+      "step": 37950
+    },
+    {
+      "epoch": 140.59,
+      "grad_norm": 1.143425703048706,
+      "learning_rate": 0.0005940740740740741,
+      "loss": 0.8356,
+      "step": 37960
+    },
+    {
+      "epoch": 140.63,
+      "grad_norm": 1.1288803815841675,
+      "learning_rate": 0.0005937037037037037,
+      "loss": 0.8432,
+      "step": 37970
+    },
+    {
+      "epoch": 140.67,
+      "grad_norm": 1.1121859550476074,
+      "learning_rate": 0.0005933333333333334,
+      "loss": 0.8469,
+      "step": 37980
+    },
+    {
+      "epoch": 140.7,
+      "grad_norm": 1.1236166954040527,
+      "learning_rate": 0.0005929629629629629,
+      "loss": 0.8398,
+      "step": 37990
+    },
+    {
+      "epoch": 140.74,
+      "grad_norm": 1.1188994646072388,
+      "learning_rate": 0.0005925925925925926,
+      "loss": 0.8561,
+      "step": 38000
+    },
+    {
+      "epoch": 140.78,
+      "grad_norm": 1.1710904836654663,
+      "learning_rate": 0.0005922222222222222,
+      "loss": 0.8594,
+      "step": 38010
+    },
+    {
+      "epoch": 140.81,
+      "grad_norm": 1.1164005994796753,
+      "learning_rate": 0.0005918518518518519,
+      "loss": 0.8645,
+      "step": 38020
+    },
+    {
+      "epoch": 140.85,
+      "grad_norm": 1.1373645067214966,
+      "learning_rate": 0.0005914814814814815,
+      "loss": 0.8587,
+      "step": 38030
+    },
+    {
+      "epoch": 140.89,
+      "grad_norm": 1.1462770700454712,
+      "learning_rate": 0.0005911111111111112,
+      "loss": 0.8689,
+      "step": 38040
+    },
+    {
+      "epoch": 140.93,
+      "grad_norm": 1.100512981414795,
+      "learning_rate": 0.0005907407407407409,
+      "loss": 0.8783,
+      "step": 38050
+    },
+    {
+      "epoch": 140.96,
+      "grad_norm": 1.2024351358413696,
+      "learning_rate": 0.0005903703703703703,
+      "loss": 0.8815,
+      "step": 38060
+    },
+    {
+      "epoch": 141.0,
+      "grad_norm": 1.949777364730835,
+      "learning_rate": 0.00059,
+      "loss": 0.8838,
+      "step": 38070
+    },
+    {
+      "epoch": 141.04,
+      "grad_norm": 1.0184309482574463,
+      "learning_rate": 0.0005896296296296296,
+      "loss": 0.7022,
+      "step": 38080
+    },
+    {
+      "epoch": 141.07,
+      "grad_norm": 1.0396180152893066,
+      "learning_rate": 0.0005892592592592593,
+      "loss": 0.6793,
+      "step": 38090
+    },
+    {
+      "epoch": 141.11,
+      "grad_norm": 1.0512199401855469,
+      "learning_rate": 0.0005888888888888889,
+      "loss": 0.7005,
+      "step": 38100
+    },
+    {
+      "epoch": 141.15,
+      "grad_norm": 1.046024203300476,
+      "learning_rate": 0.0005885185185185186,
+      "loss": 0.7032,
+      "step": 38110
+    },
+    {
+      "epoch": 141.19,
+      "grad_norm": 1.0370922088623047,
+      "learning_rate": 0.0005881481481481482,
+      "loss": 0.721,
+      "step": 38120
+    },
+    {
+      "epoch": 141.22,
+      "grad_norm": 1.1544946432113647,
+      "learning_rate": 0.0005877777777777778,
+      "loss": 0.7438,
+      "step": 38130
+    },
+    {
+      "epoch": 141.26,
+      "grad_norm": 1.1410951614379883,
+      "learning_rate": 0.0005874074074074074,
+      "loss": 0.7771,
+      "step": 38140
+    },
+    {
+      "epoch": 141.3,
+      "grad_norm": 1.0561929941177368,
+      "learning_rate": 0.0005870370370370371,
+      "loss": 0.7694,
+      "step": 38150
+    },
+    {
+      "epoch": 141.33,
+      "grad_norm": 1.096622347831726,
+      "learning_rate": 0.0005866666666666667,
+      "loss": 0.7672,
+      "step": 38160
+    },
+    {
+      "epoch": 141.37,
+      "grad_norm": 1.0748016834259033,
+      "learning_rate": 0.0005862962962962963,
+      "loss": 0.7778,
+      "step": 38170
+    },
+    {
+      "epoch": 141.41,
+      "grad_norm": 1.121734857559204,
+      "learning_rate": 0.0005859259259259259,
+      "loss": 0.8151,
+      "step": 38180
+    },
+    {
+      "epoch": 141.44,
+      "grad_norm": 1.2701319456100464,
+      "learning_rate": 0.0005855555555555556,
+      "loss": 0.7857,
+      "step": 38190
+    },
+    {
+      "epoch": 141.48,
+      "grad_norm": 1.0674980878829956,
+      "learning_rate": 0.0005851851851851851,
+      "loss": 0.8126,
+      "step": 38200
+    },
+    {
+      "epoch": 141.52,
+      "grad_norm": 1.125149130821228,
+      "learning_rate": 0.0005848148148148148,
+      "loss": 0.8048,
+      "step": 38210
+    },
+    {
+      "epoch": 141.56,
+      "grad_norm": 1.0857785940170288,
+      "learning_rate": 0.0005844444444444444,
+      "loss": 0.8335,
+      "step": 38220
+    },
+    {
+      "epoch": 141.59,
+      "grad_norm": 1.1086872816085815,
+      "learning_rate": 0.0005840740740740741,
+      "loss": 0.8306,
+      "step": 38230
+    },
+    {
+      "epoch": 141.63,
+      "grad_norm": 1.1450704336166382,
+      "learning_rate": 0.0005837037037037037,
+      "loss": 0.8205,
+      "step": 38240
+    },
+    {
+      "epoch": 141.67,
+      "grad_norm": 1.1717641353607178,
+      "learning_rate": 0.0005833333333333334,
+      "loss": 0.837,
+      "step": 38250
+    },
+    {
+      "epoch": 141.7,
+      "grad_norm": 1.1084407567977905,
+      "learning_rate": 0.0005829629629629631,
+      "loss": 0.816,
+      "step": 38260
+    },
+    {
+      "epoch": 141.74,
+      "grad_norm": 1.1263971328735352,
+      "learning_rate": 0.0005825925925925926,
+      "loss": 0.8279,
+      "step": 38270
+    },
+    {
+      "epoch": 141.78,
+      "grad_norm": 1.140552043914795,
+      "learning_rate": 0.0005822222222222223,
+      "loss": 0.8537,
+      "step": 38280
+    },
+    {
+      "epoch": 141.81,
+      "grad_norm": 1.1099085807800293,
+      "learning_rate": 0.0005818518518518518,
+      "loss": 0.8805,
+      "step": 38290
+    },
+    {
+      "epoch": 141.85,
+      "grad_norm": 1.0791407823562622,
+      "learning_rate": 0.0005814814814814815,
+      "loss": 0.8542,
+      "step": 38300
+    },
+    {
+      "epoch": 141.89,
+      "grad_norm": 1.1818629503250122,
+      "learning_rate": 0.0005811111111111111,
+      "loss": 0.868,
+      "step": 38310
+    },
+    {
+      "epoch": 141.93,
+      "grad_norm": 1.0695195198059082,
+      "learning_rate": 0.0005807407407407408,
+      "loss": 0.8643,
+      "step": 38320
+    },
+    {
+      "epoch": 141.96,
+      "grad_norm": 1.1952699422836304,
+      "learning_rate": 0.0005803703703703704,
+      "loss": 0.8908,
+      "step": 38330
+    },
+    {
+      "epoch": 142.0,
+      "grad_norm": 2.1429734230041504,
+      "learning_rate": 0.00058,
+      "loss": 0.8539,
+      "step": 38340
+    },
+    {
+      "epoch": 142.04,
+      "grad_norm": 0.9811264872550964,
+      "learning_rate": 0.0005796296296296296,
+      "loss": 0.6761,
+      "step": 38350
+    },
+    {
+      "epoch": 142.07,
+      "grad_norm": 0.9972795844078064,
+      "learning_rate": 0.0005792592592592593,
+      "loss": 0.707,
+      "step": 38360
+    },
+    {
+      "epoch": 142.11,
+      "grad_norm": 1.0141464471817017,
+      "learning_rate": 0.0005788888888888889,
+      "loss": 0.6954,
+      "step": 38370
+    },
+    {
+      "epoch": 142.15,
+      "grad_norm": 1.0808486938476562,
+      "learning_rate": 0.0005785185185185186,
+      "loss": 0.7068,
+      "step": 38380
+    },
+    {
+      "epoch": 142.19,
+      "grad_norm": 1.119739055633545,
+      "learning_rate": 0.0005781481481481482,
+      "loss": 0.7176,
+      "step": 38390
+    },
+    {
+      "epoch": 142.22,
+      "grad_norm": 1.1214107275009155,
+      "learning_rate": 0.0005777777777777778,
+      "loss": 0.7453,
+      "step": 38400
+    },
+    {
+      "epoch": 142.26,
+      "grad_norm": 1.093930721282959,
+      "learning_rate": 0.0005774074074074073,
+      "loss": 0.7452,
+      "step": 38410
+    },
+    {
+      "epoch": 142.3,
+      "grad_norm": 1.102303147315979,
+      "learning_rate": 0.000577037037037037,
+      "loss": 0.7524,
+      "step": 38420
+    },
+    {
+      "epoch": 142.33,
+      "grad_norm": 1.0816642045974731,
+      "learning_rate": 0.0005766666666666666,
+      "loss": 0.7636,
+      "step": 38430
+    },
+    {
+      "epoch": 142.37,
+      "grad_norm": 1.0792324542999268,
+      "learning_rate": 0.0005762962962962963,
+      "loss": 0.77,
+      "step": 38440
+    },
+    {
+      "epoch": 142.41,
+      "grad_norm": 1.1112143993377686,
+      "learning_rate": 0.0005759259259259259,
+      "loss": 0.7736,
+      "step": 38450
+    },
+    {
+      "epoch": 142.44,
+      "grad_norm": 1.1527026891708374,
+      "learning_rate": 0.0005755555555555556,
+      "loss": 0.8019,
+      "step": 38460
+    },
+    {
+      "epoch": 142.48,
+      "grad_norm": 1.1157525777816772,
+      "learning_rate": 0.0005751851851851852,
+      "loss": 0.8126,
+      "step": 38470
+    },
+    {
+      "epoch": 142.52,
+      "grad_norm": 1.1247179508209229,
+      "learning_rate": 0.0005748148148148148,
+      "loss": 0.7943,
+      "step": 38480
+    },
+    {
+      "epoch": 142.56,
+      "grad_norm": 1.0445313453674316,
+      "learning_rate": 0.0005744444444444445,
+      "loss": 0.8086,
+      "step": 38490
+    },
+    {
+      "epoch": 142.59,
+      "grad_norm": 1.1115723848342896,
+      "learning_rate": 0.0005740740740740741,
+      "loss": 0.8125,
+      "step": 38500
+    },
+    {
+      "epoch": 142.63,
+      "grad_norm": 1.0390018224716187,
+      "learning_rate": 0.0005737037037037038,
+      "loss": 0.8189,
+      "step": 38510
+    },
+    {
+      "epoch": 142.67,
+      "grad_norm": 1.129560112953186,
+      "learning_rate": 0.0005733333333333334,
+      "loss": 0.8171,
+      "step": 38520
+    },
+    {
+      "epoch": 142.7,
+      "grad_norm": 1.1379806995391846,
+      "learning_rate": 0.000572962962962963,
+      "loss": 0.8264,
+      "step": 38530
+    },
+    {
+      "epoch": 142.74,
+      "grad_norm": 1.113497257232666,
+      "learning_rate": 0.0005725925925925925,
+      "loss": 0.8442,
+      "step": 38540
+    },
+    {
+      "epoch": 142.78,
+      "grad_norm": 1.118584394454956,
+      "learning_rate": 0.0005722222222222222,
+      "loss": 0.8348,
+      "step": 38550
+    },
+    {
+      "epoch": 142.81,
+      "grad_norm": 1.0676319599151611,
+      "learning_rate": 0.0005718518518518518,
+      "loss": 0.8437,
+      "step": 38560
+    },
+    {
+      "epoch": 142.85,
+      "grad_norm": 1.1327413320541382,
+      "learning_rate": 0.0005714814814814815,
+      "loss": 0.8419,
+      "step": 38570
+    },
+    {
+      "epoch": 142.89,
+      "grad_norm": 1.1048108339309692,
+      "learning_rate": 0.0005711111111111111,
+      "loss": 0.8532,
+      "step": 38580
+    },
+    {
+      "epoch": 142.93,
+      "grad_norm": 1.1380068063735962,
+      "learning_rate": 0.0005707407407407408,
+      "loss": 0.8534,
+      "step": 38590
+    },
+    {
+      "epoch": 142.96,
+      "grad_norm": 1.1003347635269165,
+      "learning_rate": 0.0005703703703703704,
+      "loss": 0.8627,
+      "step": 38600
+    },
+    {
+      "epoch": 143.0,
+      "grad_norm": 2.4374372959136963,
+      "learning_rate": 0.00057,
+      "loss": 0.8586,
+      "step": 38610
+    },
+    {
+      "epoch": 143.04,
+      "grad_norm": 0.9777768850326538,
+      "learning_rate": 0.0005696296296296296,
+      "loss": 0.6775,
+      "step": 38620
+    },
+    {
+      "epoch": 143.07,
+      "grad_norm": 1.0442510843276978,
+      "learning_rate": 0.0005692592592592593,
+      "loss": 0.6821,
+      "step": 38630
+    },
+    {
+      "epoch": 143.11,
+      "grad_norm": 1.043049693107605,
+      "learning_rate": 0.0005688888888888889,
+      "loss": 0.6977,
+      "step": 38640
+    },
+    {
+      "epoch": 143.15,
+      "grad_norm": 1.0662933588027954,
+      "learning_rate": 0.0005685185185185185,
+      "loss": 0.7017,
+      "step": 38650
+    },
+    {
+      "epoch": 143.19,
+      "grad_norm": 1.0610828399658203,
+      "learning_rate": 0.0005681481481481482,
+      "loss": 0.7077,
+      "step": 38660
+    },
+    {
+      "epoch": 143.22,
+      "grad_norm": 1.0228286981582642,
+      "learning_rate": 0.0005677777777777778,
+      "loss": 0.7324,
+      "step": 38670
+    },
+    {
+      "epoch": 143.26,
+      "grad_norm": 1.0251611471176147,
+      "learning_rate": 0.0005674074074074074,
+      "loss": 0.7398,
+      "step": 38680
+    },
+    {
+      "epoch": 143.3,
+      "grad_norm": 1.1510093212127686,
+      "learning_rate": 0.000567037037037037,
+      "loss": 0.7603,
+      "step": 38690
+    },
+    {
+      "epoch": 143.33,
+      "grad_norm": 1.1165571212768555,
+      "learning_rate": 0.0005666666666666667,
+      "loss": 0.7761,
+      "step": 38700
+    },
+    {
+      "epoch": 143.37,
+      "grad_norm": 1.1153897047042847,
+      "learning_rate": 0.0005662962962962963,
+      "loss": 0.7582,
+      "step": 38710
+    },
+    {
+      "epoch": 143.41,
+      "grad_norm": 1.1018493175506592,
+      "learning_rate": 0.000565925925925926,
+      "loss": 0.7589,
+      "step": 38720
+    },
+    {
+      "epoch": 143.44,
+      "grad_norm": 1.0931953191757202,
+      "learning_rate": 0.0005655555555555556,
+      "loss": 0.7809,
+      "step": 38730
+    },
+    {
+      "epoch": 143.48,
+      "grad_norm": 1.1556371450424194,
+      "learning_rate": 0.0005651851851851853,
+      "loss": 0.8009,
+      "step": 38740
+    },
+    {
+      "epoch": 143.52,
+      "grad_norm": 1.1340473890304565,
+      "learning_rate": 0.0005648148148148148,
+      "loss": 0.769,
+      "step": 38750
+    },
+    {
+      "epoch": 143.56,
+      "grad_norm": 1.0956592559814453,
+      "learning_rate": 0.0005644444444444445,
+      "loss": 0.806,
+      "step": 38760
+    },
+    {
+      "epoch": 143.59,
+      "grad_norm": 1.2477532625198364,
+      "learning_rate": 0.000564074074074074,
+      "loss": 0.8008,
+      "step": 38770
+    },
+    {
+      "epoch": 143.63,
+      "grad_norm": 1.0662354230880737,
+      "learning_rate": 0.0005637037037037037,
+      "loss": 0.8115,
+      "step": 38780
+    },
+    {
+      "epoch": 143.67,
+      "grad_norm": 1.212112307548523,
+      "learning_rate": 0.0005633333333333333,
+      "loss": 0.806,
+      "step": 38790
+    },
+    {
+      "epoch": 143.7,
+      "grad_norm": 1.0893350839614868,
+      "learning_rate": 0.000562962962962963,
+      "loss": 0.8145,
+      "step": 38800
+    },
+    {
+      "epoch": 143.74,
+      "grad_norm": 1.0864957571029663,
+      "learning_rate": 0.0005625925925925926,
+      "loss": 0.8426,
+      "step": 38810
+    },
+    {
+      "epoch": 143.78,
+      "grad_norm": 1.1166471242904663,
+      "learning_rate": 0.0005622222222222222,
+      "loss": 0.8316,
+      "step": 38820
+    },
+    {
+      "epoch": 143.81,
+      "grad_norm": 1.1510049104690552,
+      "learning_rate": 0.0005618518518518518,
+      "loss": 0.8229,
+      "step": 38830
+    },
+    {
+      "epoch": 143.85,
+      "grad_norm": 1.0466991662979126,
+      "learning_rate": 0.0005614814814814815,
+      "loss": 0.8368,
+      "step": 38840
+    },
+    {
+      "epoch": 143.89,
+      "grad_norm": 1.1131941080093384,
+      "learning_rate": 0.0005611111111111111,
+      "loss": 0.8588,
+      "step": 38850
+    },
+    {
+      "epoch": 143.93,
+      "grad_norm": 1.1063518524169922,
+      "learning_rate": 0.0005607407407407408,
+      "loss": 0.8594,
+      "step": 38860
+    },
+    {
+      "epoch": 143.96,
+      "grad_norm": 1.011124849319458,
+      "learning_rate": 0.0005603703703703705,
+      "loss": 0.8556,
+      "step": 38870
+    },
+    {
+      "epoch": 144.0,
+      "grad_norm": 1.844652533531189,
+      "learning_rate": 0.0005600000000000001,
+      "loss": 0.8584,
+      "step": 38880
+    },
+    {
+      "epoch": 144.04,
+      "grad_norm": 1.0818065404891968,
+      "learning_rate": 0.0005596296296296296,
+      "loss": 0.6472,
+      "step": 38890
+    },
+    {
+      "epoch": 144.07,
+      "grad_norm": 1.0280637741088867,
+      "learning_rate": 0.0005592592592592592,
+      "loss": 0.6668,
+      "step": 38900
+    },
+    {
+      "epoch": 144.11,
+      "grad_norm": 1.105465292930603,
+      "learning_rate": 0.0005588888888888889,
+      "loss": 0.7023,
+      "step": 38910
+    },
+    {
+      "epoch": 144.15,
+      "grad_norm": 1.1759517192840576,
+      "learning_rate": 0.0005585185185185185,
+      "loss": 0.6914,
+      "step": 38920
+    },
+    {
+      "epoch": 144.19,
+      "grad_norm": 1.0486412048339844,
+      "learning_rate": 0.0005581481481481482,
+      "loss": 0.7048,
+      "step": 38930
+    },
+    {
+      "epoch": 144.22,
+      "grad_norm": 1.0130939483642578,
+      "learning_rate": 0.0005577777777777778,
+      "loss": 0.7313,
+      "step": 38940
+    },
+    {
+      "epoch": 144.26,
+      "grad_norm": 1.0542813539505005,
+      "learning_rate": 0.0005574074074074075,
+      "loss": 0.7384,
+      "step": 38950
+    },
+    {
+      "epoch": 144.3,
+      "grad_norm": 1.1504030227661133,
+      "learning_rate": 0.000557037037037037,
+      "loss": 0.749,
+      "step": 38960
+    },
+    {
+      "epoch": 144.33,
+      "grad_norm": 1.2246326208114624,
+      "learning_rate": 0.0005566666666666667,
+      "loss": 0.753,
+      "step": 38970
+    },
+    {
+      "epoch": 144.37,
+      "grad_norm": 1.1878856420516968,
+      "learning_rate": 0.0005562962962962963,
+      "loss": 0.7525,
+      "step": 38980
+    },
+    {
+      "epoch": 144.41,
+      "grad_norm": 1.0348081588745117,
+      "learning_rate": 0.000555925925925926,
+      "loss": 0.7492,
+      "step": 38990
+    },
+    {
+      "epoch": 144.44,
+      "grad_norm": 1.1095950603485107,
+      "learning_rate": 0.0005555555555555556,
+      "loss": 0.7751,
+      "step": 39000
+    },
+    {
+      "epoch": 144.48,
+      "grad_norm": 1.143757700920105,
+      "learning_rate": 0.0005551851851851853,
+      "loss": 0.788,
+      "step": 39010
+    },
+    {
+      "epoch": 144.52,
+      "grad_norm": 1.1462836265563965,
+      "learning_rate": 0.0005548148148148147,
+      "loss": 0.7998,
+      "step": 39020
+    },
+    {
+      "epoch": 144.56,
+      "grad_norm": 1.0627027750015259,
+      "learning_rate": 0.0005544444444444444,
+      "loss": 0.7932,
+      "step": 39030
+    },
+    {
+      "epoch": 144.59,
+      "grad_norm": 1.0926839113235474,
+      "learning_rate": 0.000554074074074074,
+      "loss": 0.8,
+      "step": 39040
+    },
+    {
+      "epoch": 144.63,
+      "grad_norm": 1.0934412479400635,
+      "learning_rate": 0.0005537037037037037,
+      "loss": 0.7927,
+      "step": 39050
+    },
+    {
+      "epoch": 144.67,
+      "grad_norm": 1.0997884273529053,
+      "learning_rate": 0.0005533333333333333,
+      "loss": 0.8074,
+      "step": 39060
+    },
+    {
+      "epoch": 144.7,
+      "grad_norm": 1.0809195041656494,
+      "learning_rate": 0.000552962962962963,
+      "loss": 0.816,
+      "step": 39070
+    },
+    {
+      "epoch": 144.74,
+      "grad_norm": 1.1381590366363525,
+      "learning_rate": 0.0005525925925925927,
+      "loss": 0.8291,
+      "step": 39080
+    },
+    {
+      "epoch": 144.78,
+      "grad_norm": 1.1699448823928833,
+      "learning_rate": 0.0005522222222222222,
+      "loss": 0.8444,
+      "step": 39090
+    },
+    {
+      "epoch": 144.81,
+      "grad_norm": 1.0845420360565186,
+      "learning_rate": 0.0005518518518518519,
+      "loss": 0.8203,
+      "step": 39100
+    },
+    {
+      "epoch": 144.85,
+      "grad_norm": 1.1182432174682617,
+      "learning_rate": 0.0005514814814814815,
+      "loss": 0.8269,
+      "step": 39110
+    },
+    {
+      "epoch": 144.89,
+      "grad_norm": 1.1073358058929443,
+      "learning_rate": 0.0005511111111111112,
+      "loss": 0.8318,
+      "step": 39120
+    },
+    {
+      "epoch": 144.93,
+      "grad_norm": 1.1134344339370728,
+      "learning_rate": 0.0005507407407407407,
+      "loss": 0.8389,
+      "step": 39130
+    },
+    {
+      "epoch": 144.96,
+      "grad_norm": 1.1211203336715698,
+      "learning_rate": 0.0005503703703703704,
+      "loss": 0.8578,
+      "step": 39140
+    },
+    {
+      "epoch": 145.0,
+      "grad_norm": 1.924920678138733,
+      "learning_rate": 0.00055,
+      "loss": 0.8209,
+      "step": 39150
+    },
+    {
+      "epoch": 145.04,
+      "grad_norm": 1.0097651481628418,
+      "learning_rate": 0.0005496296296296296,
+      "loss": 0.6572,
+      "step": 39160
+    },
+    {
+      "epoch": 145.07,
+      "grad_norm": 1.04509699344635,
+      "learning_rate": 0.0005492592592592592,
+      "loss": 0.6626,
+      "step": 39170
+    },
+    {
+      "epoch": 145.11,
+      "grad_norm": 0.9996280670166016,
+      "learning_rate": 0.0005488888888888889,
+      "loss": 0.7079,
+      "step": 39180
+    },
+    {
+      "epoch": 145.15,
+      "grad_norm": 1.038508415222168,
+      "learning_rate": 0.0005485185185185185,
+      "loss": 0.6676,
+      "step": 39190
+    },
+    {
+      "epoch": 145.19,
+      "grad_norm": 1.0765340328216553,
+      "learning_rate": 0.0005481481481481482,
+      "loss": 0.6973,
+      "step": 39200
+    },
+    {
+      "epoch": 145.22,
+      "grad_norm": 1.0047717094421387,
+      "learning_rate": 0.0005477777777777778,
+      "loss": 0.71,
+      "step": 39210
+    },
+    {
+      "epoch": 145.26,
+      "grad_norm": 1.1309590339660645,
+      "learning_rate": 0.0005474074074074075,
+      "loss": 0.7334,
+      "step": 39220
+    },
+    {
+      "epoch": 145.3,
+      "grad_norm": 1.156137466430664,
+      "learning_rate": 0.000547037037037037,
+      "loss": 0.7345,
+      "step": 39230
+    },
+    {
+      "epoch": 145.33,
+      "grad_norm": 1.1593093872070312,
+      "learning_rate": 0.0005466666666666667,
+      "loss": 0.7398,
+      "step": 39240
+    },
+    {
+      "epoch": 145.37,
+      "grad_norm": 1.149204969406128,
+      "learning_rate": 0.0005462962962962962,
+      "loss": 0.7511,
+      "step": 39250
+    },
+    {
+      "epoch": 145.41,
+      "grad_norm": 1.0335086584091187,
+      "learning_rate": 0.0005459259259259259,
+      "loss": 0.7702,
+      "step": 39260
+    },
+    {
+      "epoch": 145.44,
+      "grad_norm": 1.116210699081421,
+      "learning_rate": 0.0005455555555555555,
+      "loss": 0.7714,
+      "step": 39270
+    },
+    {
+      "epoch": 145.48,
+      "grad_norm": 1.105791687965393,
+      "learning_rate": 0.0005451851851851852,
+      "loss": 0.7735,
+      "step": 39280
+    },
+    {
+      "epoch": 145.52,
+      "grad_norm": 1.0724138021469116,
+      "learning_rate": 0.0005448148148148149,
+      "loss": 0.7731,
+      "step": 39290
+    },
+    {
+      "epoch": 145.56,
+      "grad_norm": 1.1065338850021362,
+      "learning_rate": 0.0005444444444444444,
+      "loss": 0.7743,
+      "step": 39300
+    },
+    {
+      "epoch": 145.59,
+      "grad_norm": 1.1840715408325195,
+      "learning_rate": 0.0005440740740740741,
+      "loss": 0.798,
+      "step": 39310
+    },
+    {
+      "epoch": 145.63,
+      "grad_norm": 1.1680824756622314,
+      "learning_rate": 0.0005437037037037037,
+      "loss": 0.7973,
+      "step": 39320
+    },
+    {
+      "epoch": 145.67,
+      "grad_norm": 1.1493988037109375,
+      "learning_rate": 0.0005433333333333334,
+      "loss": 0.7932,
+      "step": 39330
+    },
+    {
+      "epoch": 145.7,
+      "grad_norm": 1.14024817943573,
+      "learning_rate": 0.000542962962962963,
+      "loss": 0.8029,
+      "step": 39340
+    },
+    {
+      "epoch": 145.74,
+      "grad_norm": 1.1363987922668457,
+      "learning_rate": 0.0005425925925925927,
+      "loss": 0.803,
+      "step": 39350
+    },
+    {
+      "epoch": 145.78,
+      "grad_norm": 1.156929612159729,
+      "learning_rate": 0.0005422222222222223,
+      "loss": 0.8108,
+      "step": 39360
+    },
+    {
+      "epoch": 145.81,
+      "grad_norm": 1.1919382810592651,
+      "learning_rate": 0.0005418518518518518,
+      "loss": 0.8119,
+      "step": 39370
+    },
+    {
+      "epoch": 145.85,
+      "grad_norm": 1.0990703105926514,
+      "learning_rate": 0.0005414814814814814,
+      "loss": 0.834,
+      "step": 39380
+    },
+    {
+      "epoch": 145.89,
+      "grad_norm": 1.1778970956802368,
+      "learning_rate": 0.0005411111111111111,
+      "loss": 0.8265,
+      "step": 39390
+    },
+    {
+      "epoch": 145.93,
+      "grad_norm": 1.1993062496185303,
+      "learning_rate": 0.0005407407407407407,
+      "loss": 0.8397,
+      "step": 39400
+    },
+    {
+      "epoch": 145.96,
+      "grad_norm": 1.1283531188964844,
+      "learning_rate": 0.0005403703703703704,
+      "loss": 0.8346,
+      "step": 39410
+    },
+    {
+      "epoch": 146.0,
+      "grad_norm": 2.261765241622925,
+      "learning_rate": 0.00054,
+      "loss": 0.8082,
+      "step": 39420
+    },
+    {
+      "epoch": 146.04,
+      "grad_norm": 1.0939857959747314,
+      "learning_rate": 0.0005396296296296297,
+      "loss": 0.6595,
+      "step": 39430
+    },
+    {
+      "epoch": 146.07,
+      "grad_norm": 1.005850911140442,
+      "learning_rate": 0.0005392592592592592,
+      "loss": 0.6712,
+      "step": 39440
+    },
+    {
+      "epoch": 146.11,
+      "grad_norm": 1.0855950117111206,
+      "learning_rate": 0.0005388888888888889,
+      "loss": 0.6865,
+      "step": 39450
+    },
+    {
+      "epoch": 146.15,
+      "grad_norm": 1.111831784248352,
+      "learning_rate": 0.0005385185185185185,
+      "loss": 0.6897,
+      "step": 39460
+    },
+    {
+      "epoch": 146.19,
+      "grad_norm": 1.1208680868148804,
+      "learning_rate": 0.0005381481481481482,
+      "loss": 0.6802,
+      "step": 39470
+    },
+    {
+      "epoch": 146.22,
+      "grad_norm": 1.0708884000778198,
+      "learning_rate": 0.0005377777777777779,
+      "loss": 0.701,
+      "step": 39480
+    },
+    {
+      "epoch": 146.26,
+      "grad_norm": 1.1542518138885498,
+      "learning_rate": 0.0005374074074074075,
+      "loss": 0.7043,
+      "step": 39490
+    },
+    {
+      "epoch": 146.3,
+      "grad_norm": 1.1241813898086548,
+      "learning_rate": 0.0005370370370370371,
+      "loss": 0.7316,
+      "step": 39500
+    },
+    {
+      "epoch": 146.33,
+      "grad_norm": 1.0299608707427979,
+      "learning_rate": 0.0005366666666666666,
+      "loss": 0.7348,
+      "step": 39510
+    },
+    {
+      "epoch": 146.37,
+      "grad_norm": 1.130296230316162,
+      "learning_rate": 0.0005362962962962963,
+      "loss": 0.7505,
+      "step": 39520
+    },
+    {
+      "epoch": 146.41,
+      "grad_norm": 1.0225753784179688,
+      "learning_rate": 0.0005359259259259259,
+      "loss": 0.736,
+      "step": 39530
+    },
+    {
+      "epoch": 146.44,
+      "grad_norm": 1.0999373197555542,
+      "learning_rate": 0.0005355555555555556,
+      "loss": 0.7673,
+      "step": 39540
+    },
+    {
+      "epoch": 146.48,
+      "grad_norm": 1.0895211696624756,
+      "learning_rate": 0.0005351851851851852,
+      "loss": 0.7544,
+      "step": 39550
+    },
+    {
+      "epoch": 146.52,
+      "grad_norm": 1.161813735961914,
+      "learning_rate": 0.0005348148148148149,
+      "loss": 0.7711,
+      "step": 39560
+    },
+    {
+      "epoch": 146.56,
+      "grad_norm": 1.2281898260116577,
+      "learning_rate": 0.0005344444444444445,
+      "loss": 0.769,
+      "step": 39570
+    },
+    {
+      "epoch": 146.59,
+      "grad_norm": 1.1053509712219238,
+      "learning_rate": 0.0005340740740740741,
+      "loss": 0.7886,
+      "step": 39580
+    },
+    {
+      "epoch": 146.63,
+      "grad_norm": 1.1369836330413818,
+      "learning_rate": 0.0005337037037037037,
+      "loss": 0.7959,
+      "step": 39590
+    },
+    {
+      "epoch": 146.67,
+      "grad_norm": 1.1838639974594116,
+      "learning_rate": 0.0005333333333333334,
+      "loss": 0.7876,
+      "step": 39600
+    },
+    {
+      "epoch": 146.7,
+      "grad_norm": 1.1545466184616089,
+      "learning_rate": 0.000532962962962963,
+      "loss": 0.788,
+      "step": 39610
+    },
+    {
+      "epoch": 146.74,
+      "grad_norm": 1.1716816425323486,
+      "learning_rate": 0.0005325925925925926,
+      "loss": 0.798,
+      "step": 39620
+    },
+    {
+      "epoch": 146.78,
+      "grad_norm": 1.1613768339157104,
+      "learning_rate": 0.0005322222222222222,
+      "loss": 0.8067,
+      "step": 39630
+    },
+    {
+      "epoch": 146.81,
+      "grad_norm": 1.1515871286392212,
+      "learning_rate": 0.0005318518518518518,
+      "loss": 0.8183,
+      "step": 39640
+    },
+    {
+      "epoch": 146.85,
+      "grad_norm": 1.1251693964004517,
+      "learning_rate": 0.0005314814814814814,
+      "loss": 0.8063,
+      "step": 39650
+    },
+    {
+      "epoch": 146.89,
+      "grad_norm": 1.1438850164413452,
+      "learning_rate": 0.0005311111111111111,
+      "loss": 0.7968,
+      "step": 39660
+    },
+    {
+      "epoch": 146.93,
+      "grad_norm": 1.132735252380371,
+      "learning_rate": 0.0005307407407407407,
+      "loss": 0.8234,
+      "step": 39670
+    },
+    {
+      "epoch": 146.96,
+      "grad_norm": 1.1190030574798584,
+      "learning_rate": 0.0005303703703703704,
+      "loss": 0.8268,
+      "step": 39680
+    },
+    {
+      "epoch": 147.0,
+      "grad_norm": 2.033181667327881,
+      "learning_rate": 0.0005300000000000001,
+      "loss": 0.8498,
+      "step": 39690
+    },
+    {
+      "epoch": 147.04,
+      "grad_norm": 1.0466933250427246,
+      "learning_rate": 0.0005296296296296297,
+      "loss": 0.6463,
+      "step": 39700
+    },
+    {
+      "epoch": 147.07,
+      "grad_norm": 1.0324013233184814,
+      "learning_rate": 0.0005292592592592593,
+      "loss": 0.6445,
+      "step": 39710
+    },
+    {
+      "epoch": 147.11,
+      "grad_norm": 1.004692554473877,
+      "learning_rate": 0.0005288888888888889,
+      "loss": 0.6506,
+      "step": 39720
+    },
+    {
+      "epoch": 147.15,
+      "grad_norm": 1.0634324550628662,
+      "learning_rate": 0.0005285185185185186,
+      "loss": 0.6859,
+      "step": 39730
+    },
+    {
+      "epoch": 147.19,
+      "grad_norm": 1.0275641679763794,
+      "learning_rate": 0.0005281481481481481,
+      "loss": 0.6817,
+      "step": 39740
+    },
+    {
+      "epoch": 147.22,
+      "grad_norm": 1.118804931640625,
+      "learning_rate": 0.0005277777777777778,
+      "loss": 0.7118,
+      "step": 39750
+    },
+    {
+      "epoch": 147.26,
+      "grad_norm": 1.160846471786499,
+      "learning_rate": 0.0005274074074074074,
+      "loss": 0.7119,
+      "step": 39760
+    },
+    {
+      "epoch": 147.3,
+      "grad_norm": 1.1287107467651367,
+      "learning_rate": 0.0005270370370370371,
+      "loss": 0.7383,
+      "step": 39770
+    },
+    {
+      "epoch": 147.33,
+      "grad_norm": 1.1108704805374146,
+      "learning_rate": 0.0005266666666666666,
+      "loss": 0.7156,
+      "step": 39780
+    },
+    {
+      "epoch": 147.37,
+      "grad_norm": 1.1652884483337402,
+      "learning_rate": 0.0005262962962962963,
+      "loss": 0.7476,
+      "step": 39790
+    },
+    {
+      "epoch": 147.41,
+      "grad_norm": 1.122714638710022,
+      "learning_rate": 0.0005259259259259259,
+      "loss": 0.7423,
+      "step": 39800
+    },
+    {
+      "epoch": 147.44,
+      "grad_norm": 1.1680740118026733,
+      "learning_rate": 0.0005255555555555556,
+      "loss": 0.743,
+      "step": 39810
+    },
+    {
+      "epoch": 147.48,
+      "grad_norm": 1.1342582702636719,
+      "learning_rate": 0.0005251851851851852,
+      "loss": 0.7522,
+      "step": 39820
+    },
+    {
+      "epoch": 147.52,
+      "grad_norm": 1.113469123840332,
+      "learning_rate": 0.0005248148148148149,
+      "loss": 0.7405,
+      "step": 39830
+    },
+    {
+      "epoch": 147.56,
+      "grad_norm": 1.109640121459961,
+      "learning_rate": 0.0005244444444444445,
+      "loss": 0.7548,
+      "step": 39840
+    },
+    {
+      "epoch": 147.59,
+      "grad_norm": 1.059773564338684,
+      "learning_rate": 0.000524074074074074,
+      "loss": 0.7609,
+      "step": 39850
+    },
+    {
+      "epoch": 147.63,
+      "grad_norm": 1.2097136974334717,
+      "learning_rate": 0.0005237037037037036,
+      "loss": 0.778,
+      "step": 39860
+    },
+    {
+      "epoch": 147.67,
+      "grad_norm": 1.1901289224624634,
+      "learning_rate": 0.0005233333333333333,
+      "loss": 0.774,
+      "step": 39870
+    },
+    {
+      "epoch": 147.7,
+      "grad_norm": 1.1514116525650024,
+      "learning_rate": 0.0005229629629629629,
+      "loss": 0.7879,
+      "step": 39880
+    },
+    {
+      "epoch": 147.74,
+      "grad_norm": 1.1901254653930664,
+      "learning_rate": 0.0005225925925925926,
+      "loss": 0.7876,
+      "step": 39890
+    },
+    {
+      "epoch": 147.78,
+      "grad_norm": 1.1982858180999756,
+      "learning_rate": 0.0005222222222222223,
+      "loss": 0.7948,
+      "step": 39900
+    },
+    {
+      "epoch": 147.81,
+      "grad_norm": 1.1372984647750854,
+      "learning_rate": 0.0005218518518518519,
+      "loss": 0.8026,
+      "step": 39910
+    },
+    {
+      "epoch": 147.85,
+      "grad_norm": 1.1477608680725098,
+      "learning_rate": 0.0005214814814814815,
+      "loss": 0.7825,
+      "step": 39920
+    },
+    {
+      "epoch": 147.89,
+      "grad_norm": 1.1834083795547485,
+      "learning_rate": 0.0005211111111111111,
+      "loss": 0.8171,
+      "step": 39930
+    },
+    {
+      "epoch": 147.93,
+      "grad_norm": 1.2041763067245483,
+      "learning_rate": 0.0005207407407407408,
+      "loss": 0.8157,
+      "step": 39940
+    },
+    {
+      "epoch": 147.96,
+      "grad_norm": 1.1421923637390137,
+      "learning_rate": 0.0005203703703703704,
+      "loss": 0.8245,
+      "step": 39950
+    },
+    {
+      "epoch": 148.0,
+      "grad_norm": 2.4720282554626465,
+      "learning_rate": 0.0005200000000000001,
+      "loss": 0.8358,
+      "step": 39960
+    },
+    {
+      "epoch": 148.04,
+      "grad_norm": 1.0653715133666992,
+      "learning_rate": 0.0005196296296296297,
+      "loss": 0.6536,
+      "step": 39970
+    },
+    {
+      "epoch": 148.07,
+      "grad_norm": 1.0128833055496216,
+      "learning_rate": 0.0005192592592592593,
+      "loss": 0.6595,
+      "step": 39980
+    },
+    {
+      "epoch": 148.11,
+      "grad_norm": 1.0346759557724,
+      "learning_rate": 0.0005188888888888888,
+      "loss": 0.6267,
+      "step": 39990
+    },
+    {
+      "epoch": 148.15,
+      "grad_norm": 0.9999263286590576,
+      "learning_rate": 0.0005185185185185185,
+      "loss": 0.6466,
+      "step": 40000
+    },
+    {
+      "epoch": 148.19,
+      "grad_norm": 1.1047943830490112,
+      "learning_rate": 0.0005181481481481481,
+      "loss": 0.6875,
+      "step": 40010
+    },
+    {
+      "epoch": 148.22,
+      "grad_norm": 1.1172022819519043,
+      "learning_rate": 0.0005177777777777778,
+      "loss": 0.6935,
+      "step": 40020
+    },
+    {
+      "epoch": 148.26,
+      "grad_norm": 1.1120051145553589,
+      "learning_rate": 0.0005174074074074074,
+      "loss": 0.7019,
+      "step": 40030
+    },
+    {
+      "epoch": 148.3,
+      "grad_norm": 1.0814529657363892,
+      "learning_rate": 0.0005170370370370371,
+      "loss": 0.7039,
+      "step": 40040
+    },
+    {
+      "epoch": 148.33,
+      "grad_norm": 1.0708006620407104,
+      "learning_rate": 0.0005166666666666667,
+      "loss": 0.7087,
+      "step": 40050
+    },
+    {
+      "epoch": 148.37,
+      "grad_norm": 1.0833553075790405,
+      "learning_rate": 0.0005162962962962963,
+      "loss": 0.7337,
+      "step": 40060
+    },
+    {
+      "epoch": 148.41,
+      "grad_norm": 1.0425792932510376,
+      "learning_rate": 0.0005159259259259259,
+      "loss": 0.7404,
+      "step": 40070
+    },
+    {
+      "epoch": 148.44,
+      "grad_norm": 1.1267426013946533,
+      "learning_rate": 0.0005155555555555556,
+      "loss": 0.7303,
+      "step": 40080
+    },
+    {
+      "epoch": 148.48,
+      "grad_norm": 1.1181142330169678,
+      "learning_rate": 0.0005151851851851851,
+      "loss": 0.7364,
+      "step": 40090
+    },
+    {
+      "epoch": 148.52,
+      "grad_norm": 1.0920939445495605,
+      "learning_rate": 0.0005148148148148148,
+      "loss": 0.7446,
+      "step": 40100
+    },
+    {
+      "epoch": 148.56,
+      "grad_norm": 1.160544991493225,
+      "learning_rate": 0.0005144444444444445,
+      "loss": 0.7639,
+      "step": 40110
+    },
+    {
+      "epoch": 148.59,
+      "grad_norm": 1.1735516786575317,
+      "learning_rate": 0.0005140740740740741,
+      "loss": 0.761,
+      "step": 40120
+    },
+    {
+      "epoch": 148.63,
+      "grad_norm": 1.1860623359680176,
+      "learning_rate": 0.0005137037037037037,
+      "loss": 0.7591,
+      "step": 40130
+    },
+    {
+      "epoch": 148.67,
+      "grad_norm": 1.2525490522384644,
+      "learning_rate": 0.0005133333333333333,
+      "loss": 0.7968,
+      "step": 40140
+    },
+    {
+      "epoch": 148.7,
+      "grad_norm": 1.1567634344100952,
+      "learning_rate": 0.000512962962962963,
+      "loss": 0.773,
+      "step": 40150
+    },
+    {
+      "epoch": 148.74,
+      "grad_norm": 1.1326799392700195,
+      "learning_rate": 0.0005125925925925926,
+      "loss": 0.789,
+      "step": 40160
+    },
+    {
+      "epoch": 148.78,
+      "grad_norm": 1.1794488430023193,
+      "learning_rate": 0.0005122222222222223,
+      "loss": 0.7952,
+      "step": 40170
+    },
+    {
+      "epoch": 148.81,
+      "grad_norm": 1.138416051864624,
+      "learning_rate": 0.0005118518518518519,
+      "loss": 0.7985,
+      "step": 40180
+    },
+    {
+      "epoch": 148.85,
+      "grad_norm": 1.2064673900604248,
+      "learning_rate": 0.0005114814814814816,
+      "loss": 0.8087,
+      "step": 40190
+    },
+    {
+      "epoch": 148.89,
+      "grad_norm": 1.1867910623550415,
+      "learning_rate": 0.0005111111111111111,
+      "loss": 0.8216,
+      "step": 40200
+    },
+    {
+      "epoch": 148.93,
+      "grad_norm": 1.1607937812805176,
+      "learning_rate": 0.0005107407407407408,
+      "loss": 0.7919,
+      "step": 40210
+    },
+    {
+      "epoch": 148.96,
+      "grad_norm": 1.1278940439224243,
+      "learning_rate": 0.0005103703703703703,
+      "loss": 0.7983,
+      "step": 40220
+    },
+    {
+      "epoch": 149.0,
+      "grad_norm": 1.926332712173462,
+      "learning_rate": 0.00051,
+      "loss": 0.8079,
+      "step": 40230
+    },
+    {
+      "epoch": 149.04,
+      "grad_norm": 1.08697509765625,
+      "learning_rate": 0.0005096296296296296,
+      "loss": 0.6352,
+      "step": 40240
+    },
+    {
+      "epoch": 149.07,
+      "grad_norm": 0.9892911911010742,
+      "learning_rate": 0.0005092592592592593,
+      "loss": 0.6308,
+      "step": 40250
+    },
+    {
+      "epoch": 149.11,
+      "grad_norm": 1.0473777055740356,
+      "learning_rate": 0.0005088888888888888,
+      "loss": 0.6406,
+      "step": 40260
+    },
+    {
+      "epoch": 149.15,
+      "grad_norm": 1.0929820537567139,
+      "learning_rate": 0.0005085185185185185,
+      "loss": 0.6733,
+      "step": 40270
+    },
+    {
+      "epoch": 149.19,
+      "grad_norm": 1.1148898601531982,
+      "learning_rate": 0.0005081481481481481,
+      "loss": 0.6696,
+      "step": 40280
+    },
+    {
+      "epoch": 149.22,
+      "grad_norm": 1.172523856163025,
+      "learning_rate": 0.0005077777777777778,
+      "loss": 0.6854,
+      "step": 40290
+    },
+    {
+      "epoch": 149.26,
+      "grad_norm": 1.094595193862915,
+      "learning_rate": 0.0005074074074074075,
+      "loss": 0.6874,
+      "step": 40300
+    },
+    {
+      "epoch": 149.3,
+      "grad_norm": 1.0399515628814697,
+      "learning_rate": 0.0005070370370370371,
+      "loss": 0.7074,
+      "step": 40310
+    },
+    {
+      "epoch": 149.33,
+      "grad_norm": 1.0524202585220337,
+      "learning_rate": 0.0005066666666666668,
+      "loss": 0.7228,
+      "step": 40320
+    },
+    {
+      "epoch": 149.37,
+      "grad_norm": 1.1507049798965454,
+      "learning_rate": 0.0005062962962962962,
+      "loss": 0.7194,
+      "step": 40330
+    },
+    {
+      "epoch": 149.41,
+      "grad_norm": 1.099669098854065,
+      "learning_rate": 0.0005059259259259259,
+      "loss": 0.7212,
+      "step": 40340
+    },
+    {
+      "epoch": 149.44,
+      "grad_norm": 1.093471646308899,
+      "learning_rate": 0.0005055555555555555,
+      "loss": 0.7277,
+      "step": 40350
+    },
+    {
+      "epoch": 149.48,
+      "grad_norm": 1.1976149082183838,
+      "learning_rate": 0.0005051851851851852,
+      "loss": 0.7222,
+      "step": 40360
+    },
+    {
+      "epoch": 149.52,
+      "grad_norm": 1.1135153770446777,
+      "learning_rate": 0.0005048148148148148,
+      "loss": 0.7621,
+      "step": 40370
+    },
+    {
+      "epoch": 149.56,
+      "grad_norm": 1.1930509805679321,
+      "learning_rate": 0.0005044444444444445,
+      "loss": 0.7326,
+      "step": 40380
+    },
+    {
+      "epoch": 149.59,
+      "grad_norm": 1.2204022407531738,
+      "learning_rate": 0.0005040740740740741,
+      "loss": 0.7684,
+      "step": 40390
+    },
+    {
+      "epoch": 149.63,
+      "grad_norm": 1.1036962270736694,
+      "learning_rate": 0.0005037037037037037,
+      "loss": 0.7623,
+      "step": 40400
+    },
+    {
+      "epoch": 149.67,
+      "grad_norm": 1.1521191596984863,
+      "learning_rate": 0.0005033333333333333,
+      "loss": 0.7744,
+      "step": 40410
+    },
+    {
+      "epoch": 149.7,
+      "grad_norm": 1.172692894935608,
+      "learning_rate": 0.000502962962962963,
+      "loss": 0.7797,
+      "step": 40420
+    },
+    {
+      "epoch": 149.74,
+      "grad_norm": 1.145432710647583,
+      "learning_rate": 0.0005025925925925926,
+      "loss": 0.7674,
+      "step": 40430
+    },
+    {
+      "epoch": 149.78,
+      "grad_norm": 1.1602561473846436,
+      "learning_rate": 0.0005022222222222223,
+      "loss": 0.7686,
+      "step": 40440
+    },
+    {
+      "epoch": 149.81,
+      "grad_norm": 1.1843475103378296,
+      "learning_rate": 0.0005018518518518519,
+      "loss": 0.7736,
+      "step": 40450
+    },
+    {
+      "epoch": 149.85,
+      "grad_norm": 1.2497076988220215,
+      "learning_rate": 0.0005014814814814815,
+      "loss": 0.7763,
+      "step": 40460
+    },
+    {
+      "epoch": 149.89,
+      "grad_norm": 1.1091344356536865,
+      "learning_rate": 0.000501111111111111,
+      "loss": 0.7915,
+      "step": 40470
+    },
+    {
+      "epoch": 149.93,
+      "grad_norm": 1.1577190160751343,
+      "learning_rate": 0.0005007407407407407,
+      "loss": 0.799,
+      "step": 40480
+    },
+    {
+      "epoch": 149.96,
+      "grad_norm": 1.1448825597763062,
+      "learning_rate": 0.0005003703703703703,
+      "loss": 0.791,
+      "step": 40490
+    },
+    {
+      "epoch": 150.0,
+      "grad_norm": 1.9420194625854492,
+      "learning_rate": 0.0005,
+      "loss": 0.8055,
+      "step": 40500
+    },
+    {
+      "epoch": 150.04,
+      "grad_norm": 1.0880106687545776,
+      "learning_rate": 0.0004996296296296297,
+      "loss": 0.6164,
+      "step": 40510
+    },
+    {
+      "epoch": 150.07,
+      "grad_norm": 1.135968804359436,
+      "learning_rate": 0.0004992592592592593,
+      "loss": 0.6499,
+      "step": 40520
+    },
+    {
+      "epoch": 150.11,
+      "grad_norm": 1.111045479774475,
+      "learning_rate": 0.0004988888888888889,
+      "loss": 0.6442,
+      "step": 40530
+    },
+    {
+      "epoch": 150.15,
+      "grad_norm": 0.9961147308349609,
+      "learning_rate": 0.0004985185185185186,
+      "loss": 0.6586,
+      "step": 40540
+    },
+    {
+      "epoch": 150.19,
+      "grad_norm": 1.0140842199325562,
+      "learning_rate": 0.0004981481481481482,
+      "loss": 0.6714,
+      "step": 40550
+    },
+    {
+      "epoch": 150.22,
+      "grad_norm": 1.0996960401535034,
+      "learning_rate": 0.0004977777777777778,
+      "loss": 0.6872,
+      "step": 40560
+    },
+    {
+      "epoch": 150.26,
+      "grad_norm": 1.110884428024292,
+      "learning_rate": 0.0004974074074074075,
+      "loss": 0.6758,
+      "step": 40570
+    },
+    {
+      "epoch": 150.3,
+      "grad_norm": 1.0551011562347412,
+      "learning_rate": 0.000497037037037037,
+      "loss": 0.695,
+      "step": 40580
+    },
+    {
+      "epoch": 150.33,
+      "grad_norm": 1.2043288946151733,
+      "learning_rate": 0.0004966666666666666,
+      "loss": 0.7031,
+      "step": 40590
+    },
+    {
+      "epoch": 150.37,
+      "grad_norm": 1.1636004447937012,
+      "learning_rate": 0.0004962962962962963,
+      "loss": 0.7112,
+      "step": 40600
+    },
+    {
+      "epoch": 150.41,
+      "grad_norm": 1.0336576700210571,
+      "learning_rate": 0.0004959259259259259,
+      "loss": 0.709,
+      "step": 40610
+    },
+    {
+      "epoch": 150.44,
+      "grad_norm": 1.1624946594238281,
+      "learning_rate": 0.0004955555555555556,
+      "loss": 0.7204,
+      "step": 40620
+    },
+    {
+      "epoch": 150.48,
+      "grad_norm": 1.1802222728729248,
+      "learning_rate": 0.0004951851851851852,
+      "loss": 0.7153,
+      "step": 40630
+    },
+    {
+      "epoch": 150.52,
+      "grad_norm": 1.1276952028274536,
+      "learning_rate": 0.0004948148148148148,
+      "loss": 0.7163,
+      "step": 40640
+    },
+    {
+      "epoch": 150.56,
+      "grad_norm": 1.0561846494674683,
+      "learning_rate": 0.0004944444444444445,
+      "loss": 0.7412,
+      "step": 40650
+    },
+    {
+      "epoch": 150.59,
+      "grad_norm": 1.2993978261947632,
+      "learning_rate": 0.0004940740740740741,
+      "loss": 0.7463,
+      "step": 40660
+    },
+    {
+      "epoch": 150.63,
+      "grad_norm": 1.2309373617172241,
+      "learning_rate": 0.0004937037037037037,
+      "loss": 0.7464,
+      "step": 40670
+    },
+    {
+      "epoch": 150.67,
+      "grad_norm": 1.1278002262115479,
+      "learning_rate": 0.0004933333333333334,
+      "loss": 0.7687,
+      "step": 40680
+    },
+    {
+      "epoch": 150.7,
+      "grad_norm": 1.2106173038482666,
+      "learning_rate": 0.000492962962962963,
+      "loss": 0.766,
+      "step": 40690
+    },
+    {
+      "epoch": 150.74,
+      "grad_norm": 1.2029941082000732,
+      "learning_rate": 0.0004925925925925925,
+      "loss": 0.7765,
+      "step": 40700
+    },
+    {
+      "epoch": 150.78,
+      "grad_norm": 1.258554458618164,
+      "learning_rate": 0.0004922222222222222,
+      "loss": 0.7825,
+      "step": 40710
+    },
+    {
+      "epoch": 150.81,
+      "grad_norm": 1.215246319770813,
+      "learning_rate": 0.0004918518518518519,
+      "loss": 0.7552,
+      "step": 40720
+    },
+    {
+      "epoch": 150.85,
+      "grad_norm": 1.0888087749481201,
+      "learning_rate": 0.0004914814814814815,
+      "loss": 0.7828,
+      "step": 40730
+    },
+    {
+      "epoch": 150.89,
+      "grad_norm": 1.1715198755264282,
+      "learning_rate": 0.0004911111111111111,
+      "loss": 0.7868,
+      "step": 40740
+    },
+    {
+      "epoch": 150.93,
+      "grad_norm": 1.1972428560256958,
+      "learning_rate": 0.0004907407407407408,
+      "loss": 0.79,
+      "step": 40750
+    },
+    {
+      "epoch": 150.96,
+      "grad_norm": 1.139693260192871,
+      "learning_rate": 0.0004903703703703704,
+      "loss": 0.7698,
+      "step": 40760
+    },
+    {
+      "epoch": 151.0,
+      "grad_norm": 1.8386411666870117,
+      "learning_rate": 0.00049,
+      "loss": 0.7949,
+      "step": 40770
+    },
+    {
+      "epoch": 151.04,
+      "grad_norm": 0.9836550354957581,
+      "learning_rate": 0.0004896296296296297,
+      "loss": 0.6448,
+      "step": 40780
+    },
+    {
+      "epoch": 151.07,
+      "grad_norm": 0.9819071292877197,
+      "learning_rate": 0.0004892592592592593,
+      "loss": 0.6159,
+      "step": 40790
+    },
+    {
+      "epoch": 151.11,
+      "grad_norm": 1.1223642826080322,
+      "learning_rate": 0.0004888888888888889,
+      "loss": 0.6392,
+      "step": 40800
+    },
+    {
+      "epoch": 151.15,
+      "grad_norm": 1.1107615232467651,
+      "learning_rate": 0.0004885185185185186,
+      "loss": 0.6437,
+      "step": 40810
+    },
+    {
+      "epoch": 151.19,
+      "grad_norm": 1.05820894241333,
+      "learning_rate": 0.0004881481481481482,
+      "loss": 0.6592,
+      "step": 40820
+    },
+    {
+      "epoch": 151.22,
+      "grad_norm": 1.0786594152450562,
+      "learning_rate": 0.0004877777777777778,
+      "loss": 0.6723,
+      "step": 40830
+    },
+    {
+      "epoch": 151.26,
+      "grad_norm": 1.0383567810058594,
+      "learning_rate": 0.00048740740740740743,
+      "loss": 0.6724,
+      "step": 40840
+    },
+    {
+      "epoch": 151.3,
+      "grad_norm": 1.1029598712921143,
+      "learning_rate": 0.00048703703703703707,
+      "loss": 0.6822,
+      "step": 40850
+    },
+    {
+      "epoch": 151.33,
+      "grad_norm": 1.18717360496521,
+      "learning_rate": 0.0004866666666666667,
+      "loss": 0.6843,
+      "step": 40860
+    },
+    {
+      "epoch": 151.37,
+      "grad_norm": 1.2326600551605225,
+      "learning_rate": 0.0004862962962962963,
+      "loss": 0.6921,
+      "step": 40870
+    },
+    {
+      "epoch": 151.41,
+      "grad_norm": 1.0844796895980835,
+      "learning_rate": 0.00048592592592592595,
+      "loss": 0.696,
+      "step": 40880
+    },
+    {
+      "epoch": 151.44,
+      "grad_norm": 1.0959242582321167,
+      "learning_rate": 0.0004855555555555556,
+      "loss": 0.7053,
+      "step": 40890
+    },
+    {
+      "epoch": 151.48,
+      "grad_norm": 1.081992268562317,
+      "learning_rate": 0.0004851851851851852,
+      "loss": 0.7171,
+      "step": 40900
+    },
+    {
+      "epoch": 151.52,
+      "grad_norm": 1.120227336883545,
+      "learning_rate": 0.0004848148148148148,
+      "loss": 0.742,
+      "step": 40910
+    },
+    {
+      "epoch": 151.56,
+      "grad_norm": 1.112587332725525,
+      "learning_rate": 0.00048444444444444446,
+      "loss": 0.7368,
+      "step": 40920
+    },
+    {
+      "epoch": 151.59,
+      "grad_norm": 1.1411511898040771,
+      "learning_rate": 0.0004840740740740741,
+      "loss": 0.7487,
+      "step": 40930
+    },
+    {
+      "epoch": 151.63,
+      "grad_norm": 1.1513372659683228,
+      "learning_rate": 0.0004837037037037037,
+      "loss": 0.7416,
+      "step": 40940
+    },
+    {
+      "epoch": 151.67,
+      "grad_norm": 1.2439723014831543,
+      "learning_rate": 0.00048333333333333334,
+      "loss": 0.7646,
+      "step": 40950
+    },
+    {
+      "epoch": 151.7,
+      "grad_norm": 1.1092541217803955,
+      "learning_rate": 0.000482962962962963,
+      "loss": 0.7572,
+      "step": 40960
+    },
+    {
+      "epoch": 151.74,
+      "grad_norm": 1.1118839979171753,
+      "learning_rate": 0.00048259259259259257,
+      "loss": 0.7438,
+      "step": 40970
+    },
+    {
+      "epoch": 151.78,
+      "grad_norm": 1.1118911504745483,
+      "learning_rate": 0.0004822222222222222,
+      "loss": 0.754,
+      "step": 40980
+    },
+    {
+      "epoch": 151.81,
+      "grad_norm": 1.1592152118682861,
+      "learning_rate": 0.00048185185185185185,
+      "loss": 0.7615,
+      "step": 40990
+    },
+    {
+      "epoch": 151.85,
+      "grad_norm": 1.1864370107650757,
+      "learning_rate": 0.00048148148148148144,
+      "loss": 0.7654,
+      "step": 41000
+    },
+    {
+      "epoch": 151.89,
+      "grad_norm": 1.1990009546279907,
+      "learning_rate": 0.0004811111111111111,
+      "loss": 0.7725,
+      "step": 41010
+    },
+    {
+      "epoch": 151.93,
+      "grad_norm": 1.149665117263794,
+      "learning_rate": 0.0004807407407407408,
+      "loss": 0.7701,
+      "step": 41020
+    },
+    {
+      "epoch": 151.96,
+      "grad_norm": 1.1219122409820557,
+      "learning_rate": 0.0004803703703703704,
+      "loss": 0.7903,
+      "step": 41030
+    },
+    {
+      "epoch": 152.0,
+      "grad_norm": 2.4777867794036865,
+      "learning_rate": 0.00048,
+      "loss": 0.7812,
+      "step": 41040
+    },
+    {
+      "epoch": 152.04,
+      "grad_norm": 1.1176631450653076,
+      "learning_rate": 0.00047962962962962965,
+      "loss": 0.601,
+      "step": 41050
+    },
+    {
+      "epoch": 152.07,
+      "grad_norm": 1.086923599243164,
+      "learning_rate": 0.0004792592592592593,
+      "loss": 0.6317,
+      "step": 41060
+    },
+    {
+      "epoch": 152.11,
+      "grad_norm": 0.9840608835220337,
+      "learning_rate": 0.0004788888888888889,
+      "loss": 0.6132,
+      "step": 41070
+    },
+    {
+      "epoch": 152.15,
+      "grad_norm": 0.9909200668334961,
+      "learning_rate": 0.00047851851851851853,
+      "loss": 0.6519,
+      "step": 41080
+    },
+    {
+      "epoch": 152.19,
+      "grad_norm": 1.117368459701538,
+      "learning_rate": 0.00047814814814814817,
+      "loss": 0.6605,
+      "step": 41090
+    },
+    {
+      "epoch": 152.22,
+      "grad_norm": 1.1555815935134888,
+      "learning_rate": 0.0004777777777777778,
+      "loss": 0.6538,
+      "step": 41100
+    },
+    {
+      "epoch": 152.26,
+      "grad_norm": 1.1172096729278564,
+      "learning_rate": 0.0004774074074074074,
+      "loss": 0.6665,
+      "step": 41110
+    },
+    {
+      "epoch": 152.3,
+      "grad_norm": 1.1070817708969116,
+      "learning_rate": 0.00047703703703703705,
+      "loss": 0.6885,
+      "step": 41120
+    },
+    {
+      "epoch": 152.33,
+      "grad_norm": 1.1454331874847412,
+      "learning_rate": 0.0004766666666666667,
+      "loss": 0.6653,
+      "step": 41130
+    },
+    {
+      "epoch": 152.37,
+      "grad_norm": 1.1130858659744263,
+      "learning_rate": 0.0004762962962962963,
+      "loss": 0.6984,
+      "step": 41140
+    },
+    {
+      "epoch": 152.41,
+      "grad_norm": 1.106524109840393,
+      "learning_rate": 0.0004759259259259259,
+      "loss": 0.6933,
+      "step": 41150
+    },
+    {
+      "epoch": 152.44,
+      "grad_norm": 1.1382125616073608,
+      "learning_rate": 0.00047555555555555556,
+      "loss": 0.6922,
+      "step": 41160
+    },
+    {
+      "epoch": 152.48,
+      "grad_norm": 1.1217646598815918,
+      "learning_rate": 0.0004751851851851852,
+      "loss": 0.7216,
+      "step": 41170
+    },
+    {
+      "epoch": 152.52,
+      "grad_norm": 1.1261752843856812,
+      "learning_rate": 0.0004748148148148148,
+      "loss": 0.7053,
+      "step": 41180
+    },
+    {
+      "epoch": 152.56,
+      "grad_norm": 1.1546183824539185,
+      "learning_rate": 0.00047444444444444444,
+      "loss": 0.7199,
+      "step": 41190
+    },
+    {
+      "epoch": 152.59,
+      "grad_norm": 1.0895978212356567,
+      "learning_rate": 0.0004740740740740741,
+      "loss": 0.7476,
+      "step": 41200
+    },
+    {
+      "epoch": 152.63,
+      "grad_norm": 1.170520305633545,
+      "learning_rate": 0.00047370370370370367,
+      "loss": 0.7436,
+      "step": 41210
+    },
+    {
+      "epoch": 152.67,
+      "grad_norm": 1.1396688222885132,
+      "learning_rate": 0.00047333333333333336,
+      "loss": 0.738,
+      "step": 41220
+    },
+    {
+      "epoch": 152.7,
+      "grad_norm": 1.1664234399795532,
+      "learning_rate": 0.000472962962962963,
+      "loss": 0.7358,
+      "step": 41230
+    },
+    {
+      "epoch": 152.74,
+      "grad_norm": 1.2222336530685425,
+      "learning_rate": 0.00047259259259259265,
+      "loss": 0.7317,
+      "step": 41240
+    },
+    {
+      "epoch": 152.78,
+      "grad_norm": 1.122406244277954,
+      "learning_rate": 0.00047222222222222224,
+      "loss": 0.7431,
+      "step": 41250
+    },
+    {
+      "epoch": 152.81,
+      "grad_norm": 1.2545862197875977,
+      "learning_rate": 0.0004718518518518519,
+      "loss": 0.7522,
+      "step": 41260
+    },
+    {
+      "epoch": 152.85,
+      "grad_norm": 1.1529812812805176,
+      "learning_rate": 0.0004714814814814815,
+      "loss": 0.7779,
+      "step": 41270
+    },
+    {
+      "epoch": 152.89,
+      "grad_norm": 1.2208104133605957,
+      "learning_rate": 0.0004711111111111111,
+      "loss": 0.7697,
+      "step": 41280
+    },
+    {
+      "epoch": 152.93,
+      "grad_norm": 1.1053566932678223,
+      "learning_rate": 0.00047074074074074075,
+      "loss": 0.7773,
+      "step": 41290
+    },
+    {
+      "epoch": 152.96,
+      "grad_norm": 1.2086186408996582,
+      "learning_rate": 0.0004703703703703704,
+      "loss": 0.7723,
+      "step": 41300
+    },
+    {
+      "epoch": 153.0,
+      "grad_norm": 2.352893114089966,
+      "learning_rate": 0.00047,
+      "loss": 0.7823,
+      "step": 41310
+    },
+    {
+      "epoch": 153.04,
+      "grad_norm": 1.0613468885421753,
+      "learning_rate": 0.00046962962962962963,
+      "loss": 0.6175,
+      "step": 41320
+    },
+    {
+      "epoch": 153.07,
+      "grad_norm": 1.0901178121566772,
+      "learning_rate": 0.00046925925925925927,
+      "loss": 0.6191,
+      "step": 41330
+    },
+    {
+      "epoch": 153.11,
+      "grad_norm": 1.1153274774551392,
+      "learning_rate": 0.0004688888888888889,
+      "loss": 0.6235,
+      "step": 41340
+    },
+    {
+      "epoch": 153.15,
+      "grad_norm": 1.109421730041504,
+      "learning_rate": 0.0004685185185185185,
+      "loss": 0.6197,
+      "step": 41350
+    },
+    {
+      "epoch": 153.19,
+      "grad_norm": 1.1132631301879883,
+      "learning_rate": 0.00046814814814814815,
+      "loss": 0.6406,
+      "step": 41360
+    },
+    {
+      "epoch": 153.22,
+      "grad_norm": 1.1249178647994995,
+      "learning_rate": 0.0004677777777777778,
+      "loss": 0.6494,
+      "step": 41370
+    },
+    {
+      "epoch": 153.26,
+      "grad_norm": 1.061094045639038,
+      "learning_rate": 0.0004674074074074074,
+      "loss": 0.6581,
+      "step": 41380
+    },
+    {
+      "epoch": 153.3,
+      "grad_norm": 1.105556607246399,
+      "learning_rate": 0.000467037037037037,
+      "loss": 0.6691,
+      "step": 41390
+    },
+    {
+      "epoch": 153.33,
+      "grad_norm": 1.1356477737426758,
+      "learning_rate": 0.00046666666666666666,
+      "loss": 0.6856,
+      "step": 41400
+    },
+    {
+      "epoch": 153.37,
+      "grad_norm": 1.0905135869979858,
+      "learning_rate": 0.0004662962962962963,
+      "loss": 0.6867,
+      "step": 41410
+    },
+    {
+      "epoch": 153.41,
+      "grad_norm": 1.1868871450424194,
+      "learning_rate": 0.0004659259259259259,
+      "loss": 0.689,
+      "step": 41420
+    },
+    {
+      "epoch": 153.44,
+      "grad_norm": 1.1457031965255737,
+      "learning_rate": 0.0004655555555555556,
+      "loss": 0.6948,
+      "step": 41430
+    },
+    {
+      "epoch": 153.48,
+      "grad_norm": 1.1043384075164795,
+      "learning_rate": 0.00046518518518518523,
+      "loss": 0.699,
+      "step": 41440
+    },
+    {
+      "epoch": 153.52,
+      "grad_norm": 1.0756326913833618,
+      "learning_rate": 0.0004648148148148148,
+      "loss": 0.7106,
+      "step": 41450
+    },
+    {
+      "epoch": 153.56,
+      "grad_norm": 1.1827819347381592,
+      "learning_rate": 0.00046444444444444446,
+      "loss": 0.7403,
+      "step": 41460
+    },
+    {
+      "epoch": 153.59,
+      "grad_norm": 1.1916935443878174,
+      "learning_rate": 0.0004640740740740741,
+      "loss": 0.7038,
+      "step": 41470
+    },
+    {
+      "epoch": 153.63,
+      "grad_norm": 1.1236070394515991,
+      "learning_rate": 0.00046370370370370375,
+      "loss": 0.7164,
+      "step": 41480
+    },
+    {
+      "epoch": 153.67,
+      "grad_norm": 1.1120916604995728,
+      "learning_rate": 0.00046333333333333334,
+      "loss": 0.7496,
+      "step": 41490
+    },
+    {
+      "epoch": 153.7,
+      "grad_norm": 1.1568827629089355,
+      "learning_rate": 0.000462962962962963,
+      "loss": 0.7386,
+      "step": 41500
+    },
+    {
+      "epoch": 153.74,
+      "grad_norm": 1.1520304679870605,
+      "learning_rate": 0.0004625925925925926,
+      "loss": 0.7354,
+      "step": 41510
+    },
+    {
+      "epoch": 153.78,
+      "grad_norm": 1.194348931312561,
+      "learning_rate": 0.0004622222222222222,
+      "loss": 0.7527,
+      "step": 41520
+    },
+    {
+      "epoch": 153.81,
+      "grad_norm": 1.0823569297790527,
+      "learning_rate": 0.00046185185185185185,
+      "loss": 0.7386,
+      "step": 41530
+    },
+    {
+      "epoch": 153.85,
+      "grad_norm": 1.2366108894348145,
+      "learning_rate": 0.0004614814814814815,
+      "loss": 0.7405,
+      "step": 41540
+    },
+    {
+      "epoch": 153.89,
+      "grad_norm": 1.2345499992370605,
+      "learning_rate": 0.00046111111111111114,
+      "loss": 0.7587,
+      "step": 41550
+    },
+    {
+      "epoch": 153.93,
+      "grad_norm": 1.1638497114181519,
+      "learning_rate": 0.00046074074074074073,
+      "loss": 0.7585,
+      "step": 41560
+    },
+    {
+      "epoch": 153.96,
+      "grad_norm": 1.1882902383804321,
+      "learning_rate": 0.00046037037037037037,
+      "loss": 0.7315,
+      "step": 41570
+    },
+    {
+      "epoch": 154.0,
+      "grad_norm": 2.5309550762176514,
+      "learning_rate": 0.00046,
+      "loss": 0.7686,
+      "step": 41580
+    },
+    {
+      "epoch": 154.04,
+      "grad_norm": 1.1524628400802612,
+      "learning_rate": 0.0004596296296296296,
+      "loss": 0.5923,
+      "step": 41590
+    },
+    {
+      "epoch": 154.07,
+      "grad_norm": 1.1245719194412231,
+      "learning_rate": 0.00045925925925925925,
+      "loss": 0.5912,
+      "step": 41600
+    },
+    {
+      "epoch": 154.11,
+      "grad_norm": 1.08280611038208,
+      "learning_rate": 0.0004588888888888889,
+      "loss": 0.6281,
+      "step": 41610
+    },
+    {
+      "epoch": 154.15,
+      "grad_norm": 1.1774368286132812,
+      "learning_rate": 0.0004585185185185185,
+      "loss": 0.6129,
+      "step": 41620
+    },
+    {
+      "epoch": 154.19,
+      "grad_norm": 1.1452480554580688,
+      "learning_rate": 0.0004581481481481482,
+      "loss": 0.6267,
+      "step": 41630
+    },
+    {
+      "epoch": 154.22,
+      "grad_norm": 1.070086121559143,
+      "learning_rate": 0.0004577777777777778,
+      "loss": 0.6509,
+      "step": 41640
+    },
+    {
+      "epoch": 154.26,
+      "grad_norm": 1.218158483505249,
+      "learning_rate": 0.00045740740740740746,
+      "loss": 0.6465,
+      "step": 41650
+    },
+    {
+      "epoch": 154.3,
+      "grad_norm": 1.0811407566070557,
+      "learning_rate": 0.00045703703703703705,
+      "loss": 0.6516,
+      "step": 41660
+    },
+    {
+      "epoch": 154.33,
+      "grad_norm": 1.122278094291687,
+      "learning_rate": 0.0004566666666666667,
+      "loss": 0.6576,
+      "step": 41670
+    },
+    {
+      "epoch": 154.37,
+      "grad_norm": 1.0787327289581299,
+      "learning_rate": 0.00045629629629629633,
+      "loss": 0.6562,
+      "step": 41680
+    },
+    {
+      "epoch": 154.41,
+      "grad_norm": 1.099676489830017,
+      "learning_rate": 0.0004559259259259259,
+      "loss": 0.6897,
+      "step": 41690
+    },
+    {
+      "epoch": 154.44,
+      "grad_norm": 1.1147680282592773,
+      "learning_rate": 0.00045555555555555556,
+      "loss": 0.6945,
+      "step": 41700
+    },
+    {
+      "epoch": 154.48,
+      "grad_norm": 1.1603626012802124,
+      "learning_rate": 0.0004551851851851852,
+      "loss": 0.6991,
+      "step": 41710
+    },
+    {
+      "epoch": 154.52,
+      "grad_norm": 1.1135655641555786,
+      "learning_rate": 0.00045481481481481485,
+      "loss": 0.7082,
+      "step": 41720
+    },
+    {
+      "epoch": 154.56,
+      "grad_norm": 1.1685587167739868,
+      "learning_rate": 0.00045444444444444444,
+      "loss": 0.6951,
+      "step": 41730
+    },
+    {
+      "epoch": 154.59,
+      "grad_norm": 1.0956676006317139,
+      "learning_rate": 0.0004540740740740741,
+      "loss": 0.7139,
+      "step": 41740
+    },
+    {
+      "epoch": 154.63,
+      "grad_norm": 1.1997812986373901,
+      "learning_rate": 0.0004537037037037037,
+      "loss": 0.7053,
+      "step": 41750
+    },
+    {
+      "epoch": 154.67,
+      "grad_norm": 1.2050188779830933,
+      "learning_rate": 0.0004533333333333333,
+      "loss": 0.7286,
+      "step": 41760
+    },
+    {
+      "epoch": 154.7,
+      "grad_norm": 1.1398041248321533,
+      "learning_rate": 0.00045296296296296295,
+      "loss": 0.7291,
+      "step": 41770
+    },
+    {
+      "epoch": 154.74,
+      "grad_norm": 1.146764874458313,
+      "learning_rate": 0.0004525925925925926,
+      "loss": 0.7308,
+      "step": 41780
+    },
+    {
+      "epoch": 154.78,
+      "grad_norm": 1.1888115406036377,
+      "learning_rate": 0.00045222222222222224,
+      "loss": 0.7576,
+      "step": 41790
+    },
+    {
+      "epoch": 154.81,
+      "grad_norm": 1.1720606088638306,
+      "learning_rate": 0.00045185185185185183,
+      "loss": 0.7278,
+      "step": 41800
+    },
+    {
+      "epoch": 154.85,
+      "grad_norm": 1.167518138885498,
+      "learning_rate": 0.00045148148148148147,
+      "loss": 0.7516,
+      "step": 41810
+    },
+    {
+      "epoch": 154.89,
+      "grad_norm": 1.218790054321289,
+      "learning_rate": 0.0004511111111111111,
+      "loss": 0.7533,
+      "step": 41820
+    },
+    {
+      "epoch": 154.93,
+      "grad_norm": 1.1420875787734985,
+      "learning_rate": 0.0004507407407407407,
+      "loss": 0.7621,
+      "step": 41830
+    },
+    {
+      "epoch": 154.96,
+      "grad_norm": 1.1655449867248535,
+      "learning_rate": 0.0004503703703703704,
+      "loss": 0.7515,
+      "step": 41840
+    },
+    {
+      "epoch": 155.0,
+      "grad_norm": 2.171473503112793,
+      "learning_rate": 0.00045000000000000004,
+      "loss": 0.7627,
+      "step": 41850
+    },
+    {
+      "epoch": 155.04,
+      "grad_norm": 1.02388334274292,
+      "learning_rate": 0.0004496296296296297,
+      "loss": 0.5863,
+      "step": 41860
+    },
+    {
+      "epoch": 155.07,
+      "grad_norm": 1.0087072849273682,
+      "learning_rate": 0.0004492592592592593,
+      "loss": 0.6048,
+      "step": 41870
+    },
+    {
+      "epoch": 155.11,
+      "grad_norm": 1.1261789798736572,
+      "learning_rate": 0.0004488888888888889,
+      "loss": 0.624,
+      "step": 41880
+    },
+    {
+      "epoch": 155.15,
+      "grad_norm": 1.0319106578826904,
+      "learning_rate": 0.00044851851851851856,
+      "loss": 0.6099,
+      "step": 41890
+    },
+    {
+      "epoch": 155.19,
+      "grad_norm": 1.107433557510376,
+      "learning_rate": 0.00044814814814814815,
+      "loss": 0.6352,
+      "step": 41900
+    },
+    {
+      "epoch": 155.22,
+      "grad_norm": 1.175877332687378,
+      "learning_rate": 0.0004477777777777778,
+      "loss": 0.6475,
+      "step": 41910
+    },
+    {
+      "epoch": 155.26,
+      "grad_norm": 1.1328219175338745,
+      "learning_rate": 0.00044740740740740743,
+      "loss": 0.6498,
+      "step": 41920
+    },
+    {
+      "epoch": 155.3,
+      "grad_norm": 1.1170285940170288,
+      "learning_rate": 0.000447037037037037,
+      "loss": 0.6436,
+      "step": 41930
+    },
+    {
+      "epoch": 155.33,
+      "grad_norm": 1.050402045249939,
+      "learning_rate": 0.00044666666666666666,
+      "loss": 0.6543,
+      "step": 41940
+    },
+    {
+      "epoch": 155.37,
+      "grad_norm": 1.1217918395996094,
+      "learning_rate": 0.0004462962962962963,
+      "loss": 0.6589,
+      "step": 41950
+    },
+    {
+      "epoch": 155.41,
+      "grad_norm": 1.080417513847351,
+      "learning_rate": 0.00044592592592592595,
+      "loss": 0.6785,
+      "step": 41960
+    },
+    {
+      "epoch": 155.44,
+      "grad_norm": 1.1411579847335815,
+      "learning_rate": 0.00044555555555555554,
+      "loss": 0.6696,
+      "step": 41970
+    },
+    {
+      "epoch": 155.48,
+      "grad_norm": 1.161450982093811,
+      "learning_rate": 0.0004451851851851852,
+      "loss": 0.6909,
+      "step": 41980
+    },
+    {
+      "epoch": 155.52,
+      "grad_norm": 1.2231916189193726,
+      "learning_rate": 0.0004448148148148148,
+      "loss": 0.6869,
+      "step": 41990
+    },
+    {
+      "epoch": 155.56,
+      "grad_norm": 1.163811445236206,
+      "learning_rate": 0.0004444444444444444,
+      "loss": 0.6922,
+      "step": 42000
+    },
+    {
+      "epoch": 155.59,
+      "grad_norm": 1.1644994020462036,
+      "learning_rate": 0.00044407407407407405,
+      "loss": 0.696,
+      "step": 42010
+    },
+    {
+      "epoch": 155.63,
+      "grad_norm": 1.1336703300476074,
+      "learning_rate": 0.0004437037037037037,
+      "loss": 0.7217,
+      "step": 42020
+    },
+    {
+      "epoch": 155.67,
+      "grad_norm": 1.075753927230835,
+      "learning_rate": 0.00044333333333333334,
+      "loss": 0.7201,
+      "step": 42030
+    },
+    {
+      "epoch": 155.7,
+      "grad_norm": 1.1686182022094727,
+      "learning_rate": 0.000442962962962963,
+      "loss": 0.722,
+      "step": 42040
+    },
+    {
+      "epoch": 155.74,
+      "grad_norm": 1.1514602899551392,
+      "learning_rate": 0.0004425925925925926,
+      "loss": 0.7139,
+      "step": 42050
+    },
+    {
+      "epoch": 155.78,
+      "grad_norm": 1.1290898323059082,
+      "learning_rate": 0.00044222222222222227,
+      "loss": 0.7301,
+      "step": 42060
+    },
+    {
+      "epoch": 155.81,
+      "grad_norm": 1.2161083221435547,
+      "learning_rate": 0.00044185185185185186,
+      "loss": 0.7197,
+      "step": 42070
+    },
+    {
+      "epoch": 155.85,
+      "grad_norm": 1.1686118841171265,
+      "learning_rate": 0.0004414814814814815,
+      "loss": 0.7286,
+      "step": 42080
+    },
+    {
+      "epoch": 155.89,
+      "grad_norm": 1.2055273056030273,
+      "learning_rate": 0.00044111111111111114,
+      "loss": 0.7256,
+      "step": 42090
+    },
+    {
+      "epoch": 155.93,
+      "grad_norm": 1.1590404510498047,
+      "learning_rate": 0.0004407407407407408,
+      "loss": 0.7513,
+      "step": 42100
+    },
+    {
+      "epoch": 155.96,
+      "grad_norm": 1.1125705242156982,
+      "learning_rate": 0.0004403703703703704,
+      "loss": 0.7336,
+      "step": 42110
+    },
+    {
+      "epoch": 156.0,
+      "grad_norm": 2.0711772441864014,
+      "learning_rate": 0.00044,
+      "loss": 0.754,
+      "step": 42120
+    },
+    {
+      "epoch": 156.04,
+      "grad_norm": 1.1298962831497192,
+      "learning_rate": 0.00043962962962962966,
+      "loss": 0.6014,
+      "step": 42130
+    },
+    {
+      "epoch": 156.07,
+      "grad_norm": 1.0285028219223022,
+      "learning_rate": 0.00043925925925925925,
+      "loss": 0.5644,
+      "step": 42140
+    },
+    {
+      "epoch": 156.11,
+      "grad_norm": 1.0838044881820679,
+      "learning_rate": 0.0004388888888888889,
+      "loss": 0.6023,
+      "step": 42150
+    },
+    {
+      "epoch": 156.15,
+      "grad_norm": 1.0490028858184814,
+      "learning_rate": 0.00043851851851851853,
+      "loss": 0.6113,
+      "step": 42160
+    },
+    {
+      "epoch": 156.19,
+      "grad_norm": 1.0957796573638916,
+      "learning_rate": 0.0004381481481481482,
+      "loss": 0.6303,
+      "step": 42170
+    },
+    {
+      "epoch": 156.22,
+      "grad_norm": 1.0779811143875122,
+      "learning_rate": 0.00043777777777777776,
+      "loss": 0.6318,
+      "step": 42180
+    },
+    {
+      "epoch": 156.26,
+      "grad_norm": 1.145910382270813,
+      "learning_rate": 0.0004374074074074074,
+      "loss": 0.6353,
+      "step": 42190
+    },
+    {
+      "epoch": 156.3,
+      "grad_norm": 1.0913562774658203,
+      "learning_rate": 0.00043703703703703705,
+      "loss": 0.6348,
+      "step": 42200
+    },
+    {
+      "epoch": 156.33,
+      "grad_norm": 1.3088796138763428,
+      "learning_rate": 0.00043666666666666664,
+      "loss": 0.648,
+      "step": 42210
+    },
+    {
+      "epoch": 156.37,
+      "grad_norm": 1.1091128587722778,
+      "learning_rate": 0.0004362962962962963,
+      "loss": 0.6479,
+      "step": 42220
+    },
+    {
+      "epoch": 156.41,
+      "grad_norm": 1.184120535850525,
+      "learning_rate": 0.0004359259259259259,
+      "loss": 0.6536,
+      "step": 42230
+    },
+    {
+      "epoch": 156.44,
+      "grad_norm": 1.130748987197876,
+      "learning_rate": 0.0004355555555555555,
+      "loss": 0.6752,
+      "step": 42240
+    },
+    {
+      "epoch": 156.48,
+      "grad_norm": 1.153894305229187,
+      "learning_rate": 0.0004351851851851852,
+      "loss": 0.669,
+      "step": 42250
+    },
+    {
+      "epoch": 156.52,
+      "grad_norm": 1.2015818357467651,
+      "learning_rate": 0.00043481481481481485,
+      "loss": 0.6832,
+      "step": 42260
+    },
+    {
+      "epoch": 156.56,
+      "grad_norm": 1.1696724891662598,
+      "learning_rate": 0.0004344444444444445,
+      "loss": 0.6921,
+      "step": 42270
+    },
+    {
+      "epoch": 156.59,
+      "grad_norm": 1.0543904304504395,
+      "learning_rate": 0.0004340740740740741,
+      "loss": 0.6832,
+      "step": 42280
+    },
+    {
+      "epoch": 156.63,
+      "grad_norm": 1.197202205657959,
+      "learning_rate": 0.0004337037037037037,
+      "loss": 0.7122,
+      "step": 42290
+    },
+    {
+      "epoch": 156.67,
+      "grad_norm": 1.171535849571228,
+      "learning_rate": 0.00043333333333333337,
+      "loss": 0.7122,
+      "step": 42300
+    },
+    {
+      "epoch": 156.7,
+      "grad_norm": 1.1360234022140503,
+      "learning_rate": 0.00043296296296296296,
+      "loss": 0.725,
+      "step": 42310
+    },
+    {
+      "epoch": 156.74,
+      "grad_norm": 1.1630491018295288,
+      "learning_rate": 0.0004325925925925926,
+      "loss": 0.7112,
+      "step": 42320
+    },
+    {
+      "epoch": 156.78,
+      "grad_norm": 1.121999979019165,
+      "learning_rate": 0.00043222222222222224,
+      "loss": 0.7346,
+      "step": 42330
+    },
+    {
+      "epoch": 156.81,
+      "grad_norm": 1.2397000789642334,
+      "learning_rate": 0.0004318518518518519,
+      "loss": 0.7055,
+      "step": 42340
+    },
+    {
+      "epoch": 156.85,
+      "grad_norm": 1.1417386531829834,
+      "learning_rate": 0.00043148148148148147,
+      "loss": 0.7317,
+      "step": 42350
+    },
+    {
+      "epoch": 156.89,
+      "grad_norm": 1.1937369108200073,
+      "learning_rate": 0.0004311111111111111,
+      "loss": 0.7176,
+      "step": 42360
+    },
+    {
+      "epoch": 156.93,
+      "grad_norm": 1.2735649347305298,
+      "learning_rate": 0.00043074074074074076,
+      "loss": 0.7317,
+      "step": 42370
+    },
+    {
+      "epoch": 156.96,
+      "grad_norm": 1.2109440565109253,
+      "learning_rate": 0.00043037037037037035,
+      "loss": 0.7359,
+      "step": 42380
+    },
+    {
+      "epoch": 157.0,
+      "grad_norm": 2.1727218627929688,
+      "learning_rate": 0.00043,
+      "loss": 0.7676,
+      "step": 42390
+    },
+    {
+      "epoch": 157.04,
+      "grad_norm": 1.0113236904144287,
+      "learning_rate": 0.00042962962962962963,
+      "loss": 0.5726,
+      "step": 42400
+    },
+    {
+      "epoch": 157.07,
+      "grad_norm": 1.0725810527801514,
+      "learning_rate": 0.0004292592592592593,
+      "loss": 0.5732,
+      "step": 42410
+    },
+    {
+      "epoch": 157.11,
+      "grad_norm": 1.0465131998062134,
+      "learning_rate": 0.00042888888888888886,
+      "loss": 0.5963,
+      "step": 42420
+    },
+    {
+      "epoch": 157.15,
+      "grad_norm": 1.0574426651000977,
+      "learning_rate": 0.0004285185185185185,
+      "loss": 0.604,
+      "step": 42430
+    },
+    {
+      "epoch": 157.19,
+      "grad_norm": 1.084412932395935,
+      "learning_rate": 0.00042814814814814815,
+      "loss": 0.6123,
+      "step": 42440
+    },
+    {
+      "epoch": 157.22,
+      "grad_norm": 1.1177759170532227,
+      "learning_rate": 0.0004277777777777778,
+      "loss": 0.6294,
+      "step": 42450
+    },
+    {
+      "epoch": 157.26,
+      "grad_norm": 1.0315555334091187,
+      "learning_rate": 0.00042740740740740743,
+      "loss": 0.6354,
+      "step": 42460
+    },
+    {
+      "epoch": 157.3,
+      "grad_norm": 1.1477491855621338,
+      "learning_rate": 0.0004270370370370371,
+      "loss": 0.6207,
+      "step": 42470
+    },
+    {
+      "epoch": 157.33,
+      "grad_norm": 1.078210473060608,
+      "learning_rate": 0.0004266666666666667,
+      "loss": 0.6378,
+      "step": 42480
+    },
+    {
+      "epoch": 157.37,
+      "grad_norm": 1.1546977758407593,
+      "learning_rate": 0.0004262962962962963,
+      "loss": 0.6455,
+      "step": 42490
+    },
+    {
+      "epoch": 157.41,
+      "grad_norm": 1.1062132120132446,
+      "learning_rate": 0.00042592592592592595,
+      "loss": 0.6488,
+      "step": 42500
+    },
+    {
+      "epoch": 157.44,
+      "grad_norm": 1.1070458889007568,
+      "learning_rate": 0.0004255555555555556,
+      "loss": 0.6639,
+      "step": 42510
+    },
+    {
+      "epoch": 157.48,
+      "grad_norm": 1.1309759616851807,
+      "learning_rate": 0.0004251851851851852,
+      "loss": 0.6768,
+      "step": 42520
+    },
+    {
+      "epoch": 157.52,
+      "grad_norm": 1.1529152393341064,
+      "learning_rate": 0.0004248148148148148,
+      "loss": 0.6733,
+      "step": 42530
+    },
+    {
+      "epoch": 157.56,
+      "grad_norm": 1.1645874977111816,
+      "learning_rate": 0.00042444444444444447,
+      "loss": 0.6805,
+      "step": 42540
+    },
+    {
+      "epoch": 157.59,
+      "grad_norm": 1.1483224630355835,
+      "learning_rate": 0.00042407407407407406,
+      "loss": 0.6802,
+      "step": 42550
+    },
+    {
+      "epoch": 157.63,
+      "grad_norm": 1.1695054769515991,
+      "learning_rate": 0.0004237037037037037,
+      "loss": 0.6865,
+      "step": 42560
+    },
+    {
+      "epoch": 157.67,
+      "grad_norm": 1.21616792678833,
+      "learning_rate": 0.00042333333333333334,
+      "loss": 0.6992,
+      "step": 42570
+    },
+    {
+      "epoch": 157.7,
+      "grad_norm": 1.126473069190979,
+      "learning_rate": 0.000422962962962963,
+      "loss": 0.702,
+      "step": 42580
+    },
+    {
+      "epoch": 157.74,
+      "grad_norm": 1.2085374593734741,
+      "learning_rate": 0.00042259259259259257,
+      "loss": 0.7191,
+      "step": 42590
+    },
+    {
+      "epoch": 157.78,
+      "grad_norm": 1.2415745258331299,
+      "learning_rate": 0.0004222222222222222,
+      "loss": 0.715,
+      "step": 42600
+    },
+    {
+      "epoch": 157.81,
+      "grad_norm": 1.2480597496032715,
+      "learning_rate": 0.00042185185185185186,
+      "loss": 0.7173,
+      "step": 42610
+    },
+    {
+      "epoch": 157.85,
+      "grad_norm": 1.1208328008651733,
+      "learning_rate": 0.00042148148148148145,
+      "loss": 0.7093,
+      "step": 42620
+    },
+    {
+      "epoch": 157.89,
+      "grad_norm": 1.2241952419281006,
+      "learning_rate": 0.0004211111111111111,
+      "loss": 0.7226,
+      "step": 42630
+    },
+    {
+      "epoch": 157.93,
+      "grad_norm": 1.106781005859375,
+      "learning_rate": 0.00042074074074074073,
+      "loss": 0.7443,
+      "step": 42640
+    },
+    {
+      "epoch": 157.96,
+      "grad_norm": 1.174251675605774,
+      "learning_rate": 0.00042037037037037043,
+      "loss": 0.722,
+      "step": 42650
+    },
+    {
+      "epoch": 158.0,
+      "grad_norm": 2.178400993347168,
+      "learning_rate": 0.00042,
+      "loss": 0.724,
+      "step": 42660
+    },
+    {
+      "epoch": 158.04,
+      "grad_norm": 1.0325825214385986,
+      "learning_rate": 0.00041962962962962966,
+      "loss": 0.5514,
+      "step": 42670
+    },
+    {
+      "epoch": 158.07,
+      "grad_norm": 1.107414960861206,
+      "learning_rate": 0.0004192592592592593,
+      "loss": 0.599,
+      "step": 42680
+    },
+    {
+      "epoch": 158.11,
+      "grad_norm": 1.071731686592102,
+      "learning_rate": 0.0004188888888888889,
+      "loss": 0.5881,
+      "step": 42690
+    },
+    {
+      "epoch": 158.15,
+      "grad_norm": 1.1810855865478516,
+      "learning_rate": 0.00041851851851851853,
+      "loss": 0.5949,
+      "step": 42700
+    },
+    {
+      "epoch": 158.19,
+      "grad_norm": 1.0901010036468506,
+      "learning_rate": 0.0004181481481481482,
+      "loss": 0.5967,
+      "step": 42710
+    },
+    {
+      "epoch": 158.22,
+      "grad_norm": 1.074057936668396,
+      "learning_rate": 0.0004177777777777778,
+      "loss": 0.6035,
+      "step": 42720
+    },
+    {
+      "epoch": 158.26,
+      "grad_norm": 1.017867088317871,
+      "learning_rate": 0.0004174074074074074,
+      "loss": 0.6148,
+      "step": 42730
+    },
+    {
+      "epoch": 158.3,
+      "grad_norm": 1.0730246305465698,
+      "learning_rate": 0.00041703703703703705,
+      "loss": 0.628,
+      "step": 42740
+    },
+    {
+      "epoch": 158.33,
+      "grad_norm": 1.1320347785949707,
+      "learning_rate": 0.0004166666666666667,
+      "loss": 0.6399,
+      "step": 42750
+    },
+    {
+      "epoch": 158.37,
+      "grad_norm": 1.1257990598678589,
+      "learning_rate": 0.0004162962962962963,
+      "loss": 0.6465,
+      "step": 42760
+    },
+    {
+      "epoch": 158.41,
+      "grad_norm": 1.1283479928970337,
+      "learning_rate": 0.0004159259259259259,
+      "loss": 0.6439,
+      "step": 42770
+    },
+    {
+      "epoch": 158.44,
+      "grad_norm": 1.1584910154342651,
+      "learning_rate": 0.00041555555555555557,
+      "loss": 0.6622,
+      "step": 42780
+    },
+    {
+      "epoch": 158.48,
+      "grad_norm": 1.1500332355499268,
+      "learning_rate": 0.0004151851851851852,
+      "loss": 0.6598,
+      "step": 42790
+    },
+    {
+      "epoch": 158.52,
+      "grad_norm": 1.0922963619232178,
+      "learning_rate": 0.0004148148148148148,
+      "loss": 0.6723,
+      "step": 42800
+    },
+    {
+      "epoch": 158.56,
+      "grad_norm": 1.2014588117599487,
+      "learning_rate": 0.00041444444444444444,
+      "loss": 0.6881,
+      "step": 42810
+    },
+    {
+      "epoch": 158.59,
+      "grad_norm": 1.1734037399291992,
+      "learning_rate": 0.0004140740740740741,
+      "loss": 0.6841,
+      "step": 42820
+    },
+    {
+      "epoch": 158.63,
+      "grad_norm": 1.1617283821105957,
+      "learning_rate": 0.00041370370370370367,
+      "loss": 0.6866,
+      "step": 42830
+    },
+    {
+      "epoch": 158.67,
+      "grad_norm": 1.1515743732452393,
+      "learning_rate": 0.0004133333333333333,
+      "loss": 0.6856,
+      "step": 42840
+    },
+    {
+      "epoch": 158.7,
+      "grad_norm": 1.1375868320465088,
+      "learning_rate": 0.00041296296296296296,
+      "loss": 0.6761,
+      "step": 42850
+    },
+    {
+      "epoch": 158.74,
+      "grad_norm": 1.1889747381210327,
+      "learning_rate": 0.0004125925925925926,
+      "loss": 0.7045,
+      "step": 42860
+    },
+    {
+      "epoch": 158.78,
+      "grad_norm": 1.1877621412277222,
+      "learning_rate": 0.00041222222222222224,
+      "loss": 0.6864,
+      "step": 42870
+    },
+    {
+      "epoch": 158.81,
+      "grad_norm": 1.2101047039031982,
+      "learning_rate": 0.0004118518518518519,
+      "loss": 0.7121,
+      "step": 42880
+    },
+    {
+      "epoch": 158.85,
+      "grad_norm": 1.1501836776733398,
+      "learning_rate": 0.00041148148148148153,
+      "loss": 0.6959,
+      "step": 42890
+    },
+    {
+      "epoch": 158.89,
+      "grad_norm": 1.1214256286621094,
+      "learning_rate": 0.0004111111111111111,
+      "loss": 0.7205,
+      "step": 42900
+    },
+    {
+      "epoch": 158.93,
+      "grad_norm": 1.2997136116027832,
+      "learning_rate": 0.00041074074074074076,
+      "loss": 0.7005,
+      "step": 42910
+    },
+    {
+      "epoch": 158.96,
+      "grad_norm": 1.1824320554733276,
+      "learning_rate": 0.0004103703703703704,
+      "loss": 0.722,
+      "step": 42920
+    },
+    {
+      "epoch": 159.0,
+      "grad_norm": 2.101818084716797,
+      "learning_rate": 0.00041,
+      "loss": 0.7108,
+      "step": 42930
+    },
+    {
+      "epoch": 159.04,
+      "grad_norm": 1.0299731492996216,
+      "learning_rate": 0.00040962962962962963,
+      "loss": 0.5651,
+      "step": 42940
+    },
+    {
+      "epoch": 159.07,
+      "grad_norm": 1.1588903665542603,
+      "learning_rate": 0.0004092592592592593,
+      "loss": 0.5606,
+      "step": 42950
+    },
+    {
+      "epoch": 159.11,
+      "grad_norm": 1.090228796005249,
+      "learning_rate": 0.0004088888888888889,
+      "loss": 0.58,
+      "step": 42960
+    },
+    {
+      "epoch": 159.15,
+      "grad_norm": 1.138484239578247,
+      "learning_rate": 0.0004085185185185185,
+      "loss": 0.596,
+      "step": 42970
+    },
+    {
+      "epoch": 159.19,
+      "grad_norm": 1.0343921184539795,
+      "learning_rate": 0.00040814814814814815,
+      "loss": 0.6039,
+      "step": 42980
+    },
+    {
+      "epoch": 159.22,
+      "grad_norm": 1.137735366821289,
+      "learning_rate": 0.0004077777777777778,
+      "loss": 0.6124,
+      "step": 42990
+    },
+    {
+      "epoch": 159.26,
+      "grad_norm": 1.0816062688827515,
+      "learning_rate": 0.0004074074074074074,
+      "loss": 0.6041,
+      "step": 43000
+    },
+    {
+      "epoch": 159.3,
+      "grad_norm": 1.1814439296722412,
+      "learning_rate": 0.000407037037037037,
+      "loss": 0.6102,
+      "step": 43010
+    },
+    {
+      "epoch": 159.33,
+      "grad_norm": 1.1498757600784302,
+      "learning_rate": 0.00040666666666666667,
+      "loss": 0.6114,
+      "step": 43020
+    },
+    {
+      "epoch": 159.37,
+      "grad_norm": 1.0893162488937378,
+      "learning_rate": 0.0004062962962962963,
+      "loss": 0.6276,
+      "step": 43030
+    },
+    {
+      "epoch": 159.41,
+      "grad_norm": 1.1037179231643677,
+      "learning_rate": 0.0004059259259259259,
+      "loss": 0.6303,
+      "step": 43040
+    },
+    {
+      "epoch": 159.44,
+      "grad_norm": 1.0938615798950195,
+      "learning_rate": 0.00040555555555555554,
+      "loss": 0.6602,
+      "step": 43050
+    },
+    {
+      "epoch": 159.48,
+      "grad_norm": 1.2383320331573486,
+      "learning_rate": 0.00040518518518518524,
+      "loss": 0.672,
+      "step": 43060
+    },
+    {
+      "epoch": 159.52,
+      "grad_norm": 1.1433382034301758,
+      "learning_rate": 0.0004048148148148148,
+      "loss": 0.6534,
+      "step": 43070
+    },
+    {
+      "epoch": 159.56,
+      "grad_norm": 1.1631580591201782,
+      "learning_rate": 0.00040444444444444447,
+      "loss": 0.6525,
+      "step": 43080
+    },
+    {
+      "epoch": 159.59,
+      "grad_norm": 1.1940419673919678,
+      "learning_rate": 0.0004040740740740741,
+      "loss": 0.6784,
+      "step": 43090
+    },
+    {
+      "epoch": 159.63,
+      "grad_norm": 1.2554081678390503,
+      "learning_rate": 0.00040370370370370375,
+      "loss": 0.6611,
+      "step": 43100
+    },
+    {
+      "epoch": 159.67,
+      "grad_norm": 1.2060763835906982,
+      "learning_rate": 0.00040333333333333334,
+      "loss": 0.6781,
+      "step": 43110
+    },
+    {
+      "epoch": 159.7,
+      "grad_norm": 1.21128511428833,
+      "learning_rate": 0.000402962962962963,
+      "loss": 0.6779,
+      "step": 43120
+    },
+    {
+      "epoch": 159.74,
+      "grad_norm": 1.1475038528442383,
+      "learning_rate": 0.00040259259259259263,
+      "loss": 0.7052,
+      "step": 43130
+    },
+    {
+      "epoch": 159.78,
+      "grad_norm": 1.1346590518951416,
+      "learning_rate": 0.0004022222222222222,
+      "loss": 0.6993,
+      "step": 43140
+    },
+    {
+      "epoch": 159.81,
+      "grad_norm": 1.2075724601745605,
+      "learning_rate": 0.00040185185185185186,
+      "loss": 0.6974,
+      "step": 43150
+    },
+    {
+      "epoch": 159.85,
+      "grad_norm": 1.1834874153137207,
+      "learning_rate": 0.0004014814814814815,
+      "loss": 0.6846,
+      "step": 43160
+    },
+    {
+      "epoch": 159.89,
+      "grad_norm": 1.2245084047317505,
+      "learning_rate": 0.0004011111111111111,
+      "loss": 0.7141,
+      "step": 43170
+    },
+    {
+      "epoch": 159.93,
+      "grad_norm": 1.1702778339385986,
+      "learning_rate": 0.00040074074074074073,
+      "loss": 0.7074,
+      "step": 43180
+    },
+    {
+      "epoch": 159.96,
+      "grad_norm": 1.2088655233383179,
+      "learning_rate": 0.0004003703703703704,
+      "loss": 0.7047,
+      "step": 43190
+    },
+    {
+      "epoch": 160.0,
+      "grad_norm": 2.2771294116973877,
+      "learning_rate": 0.0004,
+      "loss": 0.7056,
+      "step": 43200
+    },
+    {
+      "epoch": 160.04,
+      "grad_norm": 1.0756103992462158,
+      "learning_rate": 0.0003996296296296296,
+      "loss": 0.5502,
+      "step": 43210
+    },
+    {
+      "epoch": 160.07,
+      "grad_norm": 1.0512828826904297,
+      "learning_rate": 0.00039925925925925925,
+      "loss": 0.5546,
+      "step": 43220
+    },
+    {
+      "epoch": 160.11,
+      "grad_norm": 1.0669021606445312,
+      "learning_rate": 0.0003988888888888889,
+      "loss": 0.5743,
+      "step": 43230
+    },
+    {
+      "epoch": 160.15,
+      "grad_norm": 1.011168360710144,
+      "learning_rate": 0.0003985185185185185,
+      "loss": 0.581,
+      "step": 43240
+    },
+    {
+      "epoch": 160.19,
+      "grad_norm": 1.0751960277557373,
+      "learning_rate": 0.0003981481481481481,
+      "loss": 0.6031,
+      "step": 43250
+    },
+    {
+      "epoch": 160.22,
+      "grad_norm": 1.123462438583374,
+      "learning_rate": 0.00039777777777777777,
+      "loss": 0.5781,
+      "step": 43260
+    },
+    {
+      "epoch": 160.26,
+      "grad_norm": 1.0884305238723755,
+      "learning_rate": 0.00039740740740740746,
+      "loss": 0.6239,
+      "step": 43270
+    },
+    {
+      "epoch": 160.3,
+      "grad_norm": 1.1230250597000122,
+      "learning_rate": 0.00039703703703703705,
+      "loss": 0.6233,
+      "step": 43280
+    },
+    {
+      "epoch": 160.33,
+      "grad_norm": 1.1342320442199707,
+      "learning_rate": 0.0003966666666666667,
+      "loss": 0.6271,
+      "step": 43290
+    },
+    {
+      "epoch": 160.37,
+      "grad_norm": 1.0703272819519043,
+      "learning_rate": 0.00039629629629629634,
+      "loss": 0.621,
+      "step": 43300
+    },
+    {
+      "epoch": 160.41,
+      "grad_norm": 1.132581353187561,
+      "learning_rate": 0.0003959259259259259,
+      "loss": 0.6373,
+      "step": 43310
+    },
+    {
+      "epoch": 160.44,
+      "grad_norm": 1.085752248764038,
+      "learning_rate": 0.00039555555555555557,
+      "loss": 0.6455,
+      "step": 43320
+    },
+    {
+      "epoch": 160.48,
+      "grad_norm": 1.1563119888305664,
+      "learning_rate": 0.0003951851851851852,
+      "loss": 0.6423,
+      "step": 43330
+    },
+    {
+      "epoch": 160.52,
+      "grad_norm": 1.2023621797561646,
+      "learning_rate": 0.00039481481481481485,
+      "loss": 0.657,
+      "step": 43340
+    },
+    {
+      "epoch": 160.56,
+      "grad_norm": 1.1549417972564697,
+      "learning_rate": 0.00039444444444444444,
+      "loss": 0.6601,
+      "step": 43350
+    },
+    {
+      "epoch": 160.59,
+      "grad_norm": 1.18620765209198,
+      "learning_rate": 0.0003940740740740741,
+      "loss": 0.6578,
+      "step": 43360
+    },
+    {
+      "epoch": 160.63,
+      "grad_norm": 1.1379402875900269,
+      "learning_rate": 0.00039370370370370373,
+      "loss": 0.6663,
+      "step": 43370
+    },
+    {
+      "epoch": 160.67,
+      "grad_norm": 1.1374726295471191,
+      "learning_rate": 0.0003933333333333333,
+      "loss": 0.6819,
+      "step": 43380
+    },
+    {
+      "epoch": 160.7,
+      "grad_norm": 1.245043158531189,
+      "learning_rate": 0.00039296296296296296,
+      "loss": 0.6728,
+      "step": 43390
+    },
+    {
+      "epoch": 160.74,
+      "grad_norm": 1.167038917541504,
+      "learning_rate": 0.0003925925925925926,
+      "loss": 0.668,
+      "step": 43400
+    },
+    {
+      "epoch": 160.78,
+      "grad_norm": 1.1261484622955322,
+      "learning_rate": 0.00039222222222222225,
+      "loss": 0.675,
+      "step": 43410
+    },
+    {
+      "epoch": 160.81,
+      "grad_norm": 1.1429600715637207,
+      "learning_rate": 0.00039185185185185183,
+      "loss": 0.6842,
+      "step": 43420
+    },
+    {
+      "epoch": 160.85,
+      "grad_norm": 1.22481369972229,
+      "learning_rate": 0.0003914814814814815,
+      "loss": 0.6691,
+      "step": 43430
+    },
+    {
+      "epoch": 160.89,
+      "grad_norm": 1.1873610019683838,
+      "learning_rate": 0.0003911111111111111,
+      "loss": 0.6811,
+      "step": 43440
+    },
+    {
+      "epoch": 160.93,
+      "grad_norm": 1.1495413780212402,
+      "learning_rate": 0.0003907407407407407,
+      "loss": 0.6972,
+      "step": 43450
+    },
+    {
+      "epoch": 160.96,
+      "grad_norm": 1.1746470928192139,
+      "learning_rate": 0.00039037037037037035,
+      "loss": 0.6851,
+      "step": 43460
+    },
+    {
+      "epoch": 161.0,
+      "grad_norm": 2.1555185317993164,
+      "learning_rate": 0.00039000000000000005,
+      "loss": 0.6896,
+      "step": 43470
+    },
+    {
+      "epoch": 161.04,
+      "grad_norm": 1.07627534866333,
+      "learning_rate": 0.00038962962962962964,
+      "loss": 0.5404,
+      "step": 43480
+    },
+    {
+      "epoch": 161.07,
+      "grad_norm": 1.0882971286773682,
+      "learning_rate": 0.0003892592592592593,
+      "loss": 0.5519,
+      "step": 43490
+    },
+    {
+      "epoch": 161.11,
+      "grad_norm": 1.0781952142715454,
+      "learning_rate": 0.0003888888888888889,
+      "loss": 0.5598,
+      "step": 43500
+    },
+    {
+      "epoch": 161.15,
+      "grad_norm": 1.0606975555419922,
+      "learning_rate": 0.00038851851851851856,
+      "loss": 0.5889,
+      "step": 43510
+    },
+    {
+      "epoch": 161.19,
+      "grad_norm": 1.0489012002944946,
+      "learning_rate": 0.00038814814814814815,
+      "loss": 0.5871,
+      "step": 43520
+    },
+    {
+      "epoch": 161.22,
+      "grad_norm": 1.1675165891647339,
+      "learning_rate": 0.0003877777777777778,
+      "loss": 0.6078,
+      "step": 43530
+    },
+    {
+      "epoch": 161.26,
+      "grad_norm": 1.1082695722579956,
+      "learning_rate": 0.00038740740740740744,
+      "loss": 0.5854,
+      "step": 43540
+    },
+    {
+      "epoch": 161.3,
+      "grad_norm": 1.103467583656311,
+      "learning_rate": 0.000387037037037037,
+      "loss": 0.5976,
+      "step": 43550
+    },
+    {
+      "epoch": 161.33,
+      "grad_norm": 1.0924053192138672,
+      "learning_rate": 0.00038666666666666667,
+      "loss": 0.6114,
+      "step": 43560
+    },
+    {
+      "epoch": 161.37,
+      "grad_norm": 1.1562738418579102,
+      "learning_rate": 0.0003862962962962963,
+      "loss": 0.6082,
+      "step": 43570
+    },
+    {
+      "epoch": 161.41,
+      "grad_norm": 1.0727171897888184,
+      "learning_rate": 0.00038592592592592595,
+      "loss": 0.6207,
+      "step": 43580
+    },
+    {
+      "epoch": 161.44,
+      "grad_norm": 1.0738524198532104,
+      "learning_rate": 0.00038555555555555554,
+      "loss": 0.6413,
+      "step": 43590
+    },
+    {
+      "epoch": 161.48,
+      "grad_norm": 1.1282587051391602,
+      "learning_rate": 0.0003851851851851852,
+      "loss": 0.6179,
+      "step": 43600
+    },
+    {
+      "epoch": 161.52,
+      "grad_norm": 1.1933233737945557,
+      "learning_rate": 0.00038481481481481483,
+      "loss": 0.6282,
+      "step": 43610
+    },
+    {
+      "epoch": 161.56,
+      "grad_norm": 1.1274137496948242,
+      "learning_rate": 0.0003844444444444444,
+      "loss": 0.6456,
+      "step": 43620
+    },
+    {
+      "epoch": 161.59,
+      "grad_norm": 1.0575157403945923,
+      "learning_rate": 0.00038407407407407406,
+      "loss": 0.6469,
+      "step": 43630
+    },
+    {
+      "epoch": 161.63,
+      "grad_norm": 1.1219662427902222,
+      "learning_rate": 0.0003837037037037037,
+      "loss": 0.6514,
+      "step": 43640
+    },
+    {
+      "epoch": 161.67,
+      "grad_norm": 1.160602331161499,
+      "learning_rate": 0.00038333333333333334,
+      "loss": 0.6681,
+      "step": 43650
+    },
+    {
+      "epoch": 161.7,
+      "grad_norm": 1.1490904092788696,
+      "learning_rate": 0.00038296296296296293,
+      "loss": 0.6755,
+      "step": 43660
+    },
+    {
+      "epoch": 161.74,
+      "grad_norm": 1.1233144998550415,
+      "learning_rate": 0.0003825925925925926,
+      "loss": 0.6718,
+      "step": 43670
+    },
+    {
+      "epoch": 161.78,
+      "grad_norm": 1.1319135427474976,
+      "learning_rate": 0.0003822222222222223,
+      "loss": 0.68,
+      "step": 43680
+    },
+    {
+      "epoch": 161.81,
+      "grad_norm": 1.164638638496399,
+      "learning_rate": 0.00038185185185185186,
+      "loss": 0.6809,
+      "step": 43690
+    },
+    {
+      "epoch": 161.85,
+      "grad_norm": 1.2083297967910767,
+      "learning_rate": 0.0003814814814814815,
+      "loss": 0.6953,
+      "step": 43700
+    },
+    {
+      "epoch": 161.89,
+      "grad_norm": 1.1989994049072266,
+      "learning_rate": 0.00038111111111111115,
+      "loss": 0.691,
+      "step": 43710
+    },
+    {
+      "epoch": 161.93,
+      "grad_norm": 1.2003412246704102,
+      "learning_rate": 0.0003807407407407408,
+      "loss": 0.6784,
+      "step": 43720
+    },
+    {
+      "epoch": 161.96,
+      "grad_norm": 1.177047610282898,
+      "learning_rate": 0.0003803703703703704,
+      "loss": 0.6812,
+      "step": 43730
+    },
+    {
+      "epoch": 162.0,
+      "grad_norm": 2.104875326156616,
+      "learning_rate": 0.00038,
+      "loss": 0.6884,
+      "step": 43740
+    },
+    {
+      "epoch": 162.04,
+      "grad_norm": 1.0653473138809204,
+      "learning_rate": 0.00037962962962962966,
+      "loss": 0.5542,
+      "step": 43750
+    },
+    {
+      "epoch": 162.07,
+      "grad_norm": 1.1091032028198242,
+      "learning_rate": 0.00037925925925925925,
+      "loss": 0.5432,
+      "step": 43760
+    },
+    {
+      "epoch": 162.11,
+      "grad_norm": 0.9920121431350708,
+      "learning_rate": 0.0003788888888888889,
+      "loss": 0.5498,
+      "step": 43770
+    },
+    {
+      "epoch": 162.15,
+      "grad_norm": 1.0519205331802368,
+      "learning_rate": 0.00037851851851851854,
+      "loss": 0.555,
+      "step": 43780
+    },
+    {
+      "epoch": 162.19,
+      "grad_norm": 1.044488787651062,
+      "learning_rate": 0.0003781481481481481,
+      "loss": 0.5752,
+      "step": 43790
+    },
+    {
+      "epoch": 162.22,
+      "grad_norm": 1.165174126625061,
+      "learning_rate": 0.00037777777777777777,
+      "loss": 0.5811,
+      "step": 43800
+    },
+    {
+      "epoch": 162.26,
+      "grad_norm": 1.1170642375946045,
+      "learning_rate": 0.0003774074074074074,
+      "loss": 0.5996,
+      "step": 43810
+    },
+    {
+      "epoch": 162.3,
+      "grad_norm": 1.0534666776657104,
+      "learning_rate": 0.00037703703703703705,
+      "loss": 0.6039,
+      "step": 43820
+    },
+    {
+      "epoch": 162.33,
+      "grad_norm": 1.1040012836456299,
+      "learning_rate": 0.00037666666666666664,
+      "loss": 0.6114,
+      "step": 43830
+    },
+    {
+      "epoch": 162.37,
+      "grad_norm": 1.1486854553222656,
+      "learning_rate": 0.0003762962962962963,
+      "loss": 0.6191,
+      "step": 43840
+    },
+    {
+      "epoch": 162.41,
+      "grad_norm": 1.1414391994476318,
+      "learning_rate": 0.00037592592592592593,
+      "loss": 0.6203,
+      "step": 43850
+    },
+    {
+      "epoch": 162.44,
+      "grad_norm": 1.213144063949585,
+      "learning_rate": 0.0003755555555555555,
+      "loss": 0.6278,
+      "step": 43860
+    },
+    {
+      "epoch": 162.48,
+      "grad_norm": 1.127974033355713,
+      "learning_rate": 0.00037518518518518516,
+      "loss": 0.6341,
+      "step": 43870
+    },
+    {
+      "epoch": 162.52,
+      "grad_norm": 1.1302952766418457,
+      "learning_rate": 0.00037481481481481486,
+      "loss": 0.623,
+      "step": 43880
+    },
+    {
+      "epoch": 162.56,
+      "grad_norm": 1.1535022258758545,
+      "learning_rate": 0.0003744444444444445,
+      "loss": 0.66,
+      "step": 43890
+    },
+    {
+      "epoch": 162.59,
+      "grad_norm": 1.2633349895477295,
+      "learning_rate": 0.0003740740740740741,
+      "loss": 0.6539,
+      "step": 43900
+    },
+    {
+      "epoch": 162.63,
+      "grad_norm": 1.1691466569900513,
+      "learning_rate": 0.00037370370370370373,
+      "loss": 0.6448,
+      "step": 43910
+    },
+    {
+      "epoch": 162.67,
+      "grad_norm": 1.20991051197052,
+      "learning_rate": 0.0003733333333333334,
+      "loss": 0.6407,
+      "step": 43920
+    },
+    {
+      "epoch": 162.7,
+      "grad_norm": 1.1535216569900513,
+      "learning_rate": 0.00037296296296296296,
+      "loss": 0.6649,
+      "step": 43930
+    },
+    {
+      "epoch": 162.74,
+      "grad_norm": 1.1679757833480835,
+      "learning_rate": 0.0003725925925925926,
+      "loss": 0.6485,
+      "step": 43940
+    },
+    {
+      "epoch": 162.78,
+      "grad_norm": 1.1296665668487549,
+      "learning_rate": 0.00037222222222222225,
+      "loss": 0.6507,
+      "step": 43950
+    },
+    {
+      "epoch": 162.81,
+      "grad_norm": 1.1518938541412354,
+      "learning_rate": 0.0003718518518518519,
+      "loss": 0.66,
+      "step": 43960
+    },
+    {
+      "epoch": 162.85,
+      "grad_norm": 1.1524932384490967,
+      "learning_rate": 0.0003714814814814815,
+      "loss": 0.67,
+      "step": 43970
+    },
+    {
+      "epoch": 162.89,
+      "grad_norm": 1.1855889558792114,
+      "learning_rate": 0.0003711111111111111,
+      "loss": 0.6697,
+      "step": 43980
+    },
+    {
+      "epoch": 162.93,
+      "grad_norm": 1.200710415840149,
+      "learning_rate": 0.00037074074074074076,
+      "loss": 0.6605,
+      "step": 43990
+    },
+    {
+      "epoch": 162.96,
+      "grad_norm": 1.20967435836792,
+      "learning_rate": 0.00037037037037037035,
+      "loss": 0.6831,
+      "step": 44000
+    },
+    {
+      "epoch": 163.0,
+      "grad_norm": 3.300697088241577,
+      "learning_rate": 0.00037,
+      "loss": 0.663,
+      "step": 44010
+    },
+    {
+      "epoch": 163.04,
+      "grad_norm": 1.1124473810195923,
+      "learning_rate": 0.00036962962962962964,
+      "loss": 0.5306,
+      "step": 44020
+    },
+    {
+      "epoch": 163.07,
+      "grad_norm": 1.1132569313049316,
+      "learning_rate": 0.0003692592592592592,
+      "loss": 0.5461,
+      "step": 44030
+    },
+    {
+      "epoch": 163.11,
+      "grad_norm": 1.1740922927856445,
+      "learning_rate": 0.00036888888888888887,
+      "loss": 0.5393,
+      "step": 44040
+    },
+    {
+      "epoch": 163.15,
+      "grad_norm": 1.1028344631195068,
+      "learning_rate": 0.0003685185185185185,
+      "loss": 0.5571,
+      "step": 44050
+    },
+    {
+      "epoch": 163.19,
+      "grad_norm": 1.029313564300537,
+      "learning_rate": 0.00036814814814814815,
+      "loss": 0.5373,
+      "step": 44060
+    },
+    {
+      "epoch": 163.22,
+      "grad_norm": 1.1128367185592651,
+      "learning_rate": 0.00036777777777777774,
+      "loss": 0.5853,
+      "step": 44070
+    },
+    {
+      "epoch": 163.26,
+      "grad_norm": 1.0616388320922852,
+      "learning_rate": 0.00036740740740740744,
+      "loss": 0.5839,
+      "step": 44080
+    },
+    {
+      "epoch": 163.3,
+      "grad_norm": 1.1045690774917603,
+      "learning_rate": 0.0003670370370370371,
+      "loss": 0.5866,
+      "step": 44090
+    },
+    {
+      "epoch": 163.33,
+      "grad_norm": 1.0463433265686035,
+      "learning_rate": 0.00036666666666666667,
+      "loss": 0.5984,
+      "step": 44100
+    },
+    {
+      "epoch": 163.37,
+      "grad_norm": 1.213541865348816,
+      "learning_rate": 0.0003662962962962963,
+      "loss": 0.6033,
+      "step": 44110
+    },
+    {
+      "epoch": 163.41,
+      "grad_norm": 1.070902943611145,
+      "learning_rate": 0.00036592592592592596,
+      "loss": 0.6147,
+      "step": 44120
+    },
+    {
+      "epoch": 163.44,
+      "grad_norm": 1.0918152332305908,
+      "learning_rate": 0.0003655555555555556,
+      "loss": 0.6183,
+      "step": 44130
+    },
+    {
+      "epoch": 163.48,
+      "grad_norm": 1.1264549493789673,
+      "learning_rate": 0.0003651851851851852,
+      "loss": 0.6266,
+      "step": 44140
+    },
+    {
+      "epoch": 163.52,
+      "grad_norm": 1.174247145652771,
+      "learning_rate": 0.00036481481481481483,
+      "loss": 0.6421,
+      "step": 44150
+    },
+    {
+      "epoch": 163.56,
+      "grad_norm": 1.1570000648498535,
+      "learning_rate": 0.00036444444444444447,
+      "loss": 0.6308,
+      "step": 44160
+    },
+    {
+      "epoch": 163.59,
+      "grad_norm": 1.2352962493896484,
+      "learning_rate": 0.00036407407407407406,
+      "loss": 0.6307,
+      "step": 44170
+    },
+    {
+      "epoch": 163.63,
+      "grad_norm": 1.1054643392562866,
+      "learning_rate": 0.0003637037037037037,
+      "loss": 0.6464,
+      "step": 44180
+    },
+    {
+      "epoch": 163.67,
+      "grad_norm": 1.065085530281067,
+      "learning_rate": 0.00036333333333333335,
+      "loss": 0.6382,
+      "step": 44190
+    },
+    {
+      "epoch": 163.7,
+      "grad_norm": 1.1962122917175293,
+      "learning_rate": 0.000362962962962963,
+      "loss": 0.6451,
+      "step": 44200
+    },
+    {
+      "epoch": 163.74,
+      "grad_norm": 1.2009409666061401,
+      "learning_rate": 0.0003625925925925926,
+      "loss": 0.6479,
+      "step": 44210
+    },
+    {
+      "epoch": 163.78,
+      "grad_norm": 1.1283183097839355,
+      "learning_rate": 0.0003622222222222222,
+      "loss": 0.6485,
+      "step": 44220
+    },
+    {
+      "epoch": 163.81,
+      "grad_norm": 1.1767650842666626,
+      "learning_rate": 0.00036185185185185186,
+      "loss": 0.6593,
+      "step": 44230
+    },
+    {
+      "epoch": 163.85,
+      "grad_norm": 1.154915690422058,
+      "learning_rate": 0.00036148148148148145,
+      "loss": 0.675,
+      "step": 44240
+    },
+    {
+      "epoch": 163.89,
+      "grad_norm": 1.1761237382888794,
+      "learning_rate": 0.0003611111111111111,
+      "loss": 0.662,
+      "step": 44250
+    },
+    {
+      "epoch": 163.93,
+      "grad_norm": 1.2740514278411865,
+      "learning_rate": 0.00036074074074074074,
+      "loss": 0.6671,
+      "step": 44260
+    },
+    {
+      "epoch": 163.96,
+      "grad_norm": 1.1661285161972046,
+      "learning_rate": 0.0003603703703703704,
+      "loss": 0.6729,
+      "step": 44270
+    },
+    {
+      "epoch": 164.0,
+      "grad_norm": 2.0054233074188232,
+      "learning_rate": 0.00035999999999999997,
+      "loss": 0.6686,
+      "step": 44280
+    },
+    {
+      "epoch": 164.04,
+      "grad_norm": 0.9928212761878967,
+      "learning_rate": 0.00035962962962962967,
+      "loss": 0.5332,
+      "step": 44290
+    },
+    {
+      "epoch": 164.07,
+      "grad_norm": 1.057337760925293,
+      "learning_rate": 0.0003592592592592593,
+      "loss": 0.5347,
+      "step": 44300
+    },
+    {
+      "epoch": 164.11,
+      "grad_norm": 1.0809550285339355,
+      "learning_rate": 0.0003588888888888889,
+      "loss": 0.5378,
+      "step": 44310
+    },
+    {
+      "epoch": 164.15,
+      "grad_norm": 0.9812051653862,
+      "learning_rate": 0.00035851851851851854,
+      "loss": 0.5619,
+      "step": 44320
+    },
+    {
+      "epoch": 164.19,
+      "grad_norm": 1.0844264030456543,
+      "learning_rate": 0.0003581481481481482,
+      "loss": 0.5586,
+      "step": 44330
+    },
+    {
+      "epoch": 164.22,
+      "grad_norm": 1.1218096017837524,
+      "learning_rate": 0.00035777777777777777,
+      "loss": 0.5732,
+      "step": 44340
+    },
+    {
+      "epoch": 164.26,
+      "grad_norm": 1.1573761701583862,
+      "learning_rate": 0.0003574074074074074,
+      "loss": 0.5704,
+      "step": 44350
+    },
+    {
+      "epoch": 164.3,
+      "grad_norm": 1.2225539684295654,
+      "learning_rate": 0.00035703703703703706,
+      "loss": 0.5686,
+      "step": 44360
+    },
+    {
+      "epoch": 164.33,
+      "grad_norm": 1.0911756753921509,
+      "learning_rate": 0.0003566666666666667,
+      "loss": 0.5971,
+      "step": 44370
+    },
+    {
+      "epoch": 164.37,
+      "grad_norm": 1.0996484756469727,
+      "learning_rate": 0.0003562962962962963,
+      "loss": 0.5912,
+      "step": 44380
+    },
+    {
+      "epoch": 164.41,
+      "grad_norm": 1.1536465883255005,
+      "learning_rate": 0.00035592592592592593,
+      "loss": 0.6102,
+      "step": 44390
+    },
+    {
+      "epoch": 164.44,
+      "grad_norm": 1.212332844734192,
+      "learning_rate": 0.00035555555555555557,
+      "loss": 0.6031,
+      "step": 44400
+    },
+    {
+      "epoch": 164.48,
+      "grad_norm": 1.1705257892608643,
+      "learning_rate": 0.00035518518518518516,
+      "loss": 0.5988,
+      "step": 44410
+    },
+    {
+      "epoch": 164.52,
+      "grad_norm": 1.1567353010177612,
+      "learning_rate": 0.0003548148148148148,
+      "loss": 0.6318,
+      "step": 44420
+    },
+    {
+      "epoch": 164.56,
+      "grad_norm": 1.1380826234817505,
+      "learning_rate": 0.00035444444444444445,
+      "loss": 0.6225,
+      "step": 44430
+    },
+    {
+      "epoch": 164.59,
+      "grad_norm": 1.2010952234268188,
+      "learning_rate": 0.0003540740740740741,
+      "loss": 0.6333,
+      "step": 44440
+    },
+    {
+      "epoch": 164.63,
+      "grad_norm": 1.1813087463378906,
+      "learning_rate": 0.0003537037037037037,
+      "loss": 0.6223,
+      "step": 44450
+    },
+    {
+      "epoch": 164.67,
+      "grad_norm": 1.2010884284973145,
+      "learning_rate": 0.0003533333333333333,
+      "loss": 0.6246,
+      "step": 44460
+    },
+    {
+      "epoch": 164.7,
+      "grad_norm": 1.1103819608688354,
+      "learning_rate": 0.00035296296296296296,
+      "loss": 0.6417,
+      "step": 44470
+    },
+    {
+      "epoch": 164.74,
+      "grad_norm": 1.2009899616241455,
+      "learning_rate": 0.00035259259259259255,
+      "loss": 0.6318,
+      "step": 44480
+    },
+    {
+      "epoch": 164.78,
+      "grad_norm": 1.2279729843139648,
+      "learning_rate": 0.00035222222222222225,
+      "loss": 0.6437,
+      "step": 44490
+    },
+    {
+      "epoch": 164.81,
+      "grad_norm": 1.1632214784622192,
+      "learning_rate": 0.0003518518518518519,
+      "loss": 0.6419,
+      "step": 44500
+    },
+    {
+      "epoch": 164.85,
+      "grad_norm": 1.2233213186264038,
+      "learning_rate": 0.00035148148148148153,
+      "loss": 0.6453,
+      "step": 44510
+    },
+    {
+      "epoch": 164.89,
+      "grad_norm": 1.2718476057052612,
+      "learning_rate": 0.0003511111111111111,
+      "loss": 0.6696,
+      "step": 44520
+    },
+    {
+      "epoch": 164.93,
+      "grad_norm": 1.1639232635498047,
+      "learning_rate": 0.00035074074074074077,
+      "loss": 0.6662,
+      "step": 44530
+    },
+    {
+      "epoch": 164.96,
+      "grad_norm": 1.1556235551834106,
+      "learning_rate": 0.0003503703703703704,
+      "loss": 0.6543,
+      "step": 44540
+    },
+    {
+      "epoch": 165.0,
+      "grad_norm": 2.22497296333313,
+      "learning_rate": 0.00035,
+      "loss": 0.6562,
+      "step": 44550
+    },
+    {
+      "epoch": 165.04,
+      "grad_norm": 0.9978173971176147,
+      "learning_rate": 0.00034962962962962964,
+      "loss": 0.5205,
+      "step": 44560
+    },
+    {
+      "epoch": 165.07,
+      "grad_norm": 0.9967636466026306,
+      "learning_rate": 0.0003492592592592593,
+      "loss": 0.5098,
+      "step": 44570
+    },
+    {
+      "epoch": 165.11,
+      "grad_norm": 1.1091960668563843,
+      "learning_rate": 0.0003488888888888889,
+      "loss": 0.5431,
+      "step": 44580
+    },
+    {
+      "epoch": 165.15,
+      "grad_norm": 1.0127582550048828,
+      "learning_rate": 0.0003485185185185185,
+      "loss": 0.5463,
+      "step": 44590
+    },
+    {
+      "epoch": 165.19,
+      "grad_norm": 1.1253612041473389,
+      "learning_rate": 0.00034814814814814816,
+      "loss": 0.5565,
+      "step": 44600
+    },
+    {
+      "epoch": 165.22,
+      "grad_norm": 1.1062602996826172,
+      "learning_rate": 0.0003477777777777778,
+      "loss": 0.5653,
+      "step": 44610
+    },
+    {
+      "epoch": 165.26,
+      "grad_norm": 1.0730866193771362,
+      "learning_rate": 0.0003474074074074074,
+      "loss": 0.5714,
+      "step": 44620
+    },
+    {
+      "epoch": 165.3,
+      "grad_norm": 1.1267579793930054,
+      "learning_rate": 0.00034703703703703703,
+      "loss": 0.5642,
+      "step": 44630
+    },
+    {
+      "epoch": 165.33,
+      "grad_norm": 1.1193773746490479,
+      "learning_rate": 0.00034666666666666667,
+      "loss": 0.5742,
+      "step": 44640
+    },
+    {
+      "epoch": 165.37,
+      "grad_norm": 1.1450809240341187,
+      "learning_rate": 0.00034629629629629626,
+      "loss": 0.5998,
+      "step": 44650
+    },
+    {
+      "epoch": 165.41,
+      "grad_norm": 1.165490746498108,
+      "learning_rate": 0.0003459259259259259,
+      "loss": 0.593,
+      "step": 44660
+    },
+    {
+      "epoch": 165.44,
+      "grad_norm": 1.1239349842071533,
+      "learning_rate": 0.00034555555555555555,
+      "loss": 0.5855,
+      "step": 44670
+    },
+    {
+      "epoch": 165.48,
+      "grad_norm": 1.146379828453064,
+      "learning_rate": 0.0003451851851851852,
+      "loss": 0.6084,
+      "step": 44680
+    },
+    {
+      "epoch": 165.52,
+      "grad_norm": 1.2076387405395508,
+      "learning_rate": 0.0003448148148148148,
+      "loss": 0.6022,
+      "step": 44690
+    },
+    {
+      "epoch": 165.56,
+      "grad_norm": 1.1927461624145508,
+      "learning_rate": 0.0003444444444444445,
+      "loss": 0.6044,
+      "step": 44700
+    },
+    {
+      "epoch": 165.59,
+      "grad_norm": 1.2381534576416016,
+      "learning_rate": 0.0003440740740740741,
+      "loss": 0.6095,
+      "step": 44710
+    },
+    {
+      "epoch": 165.63,
+      "grad_norm": 1.1408498287200928,
+      "learning_rate": 0.0003437037037037037,
+      "loss": 0.6088,
+      "step": 44720
+    },
+    {
+      "epoch": 165.67,
+      "grad_norm": 1.1445554494857788,
+      "learning_rate": 0.00034333333333333335,
+      "loss": 0.6393,
+      "step": 44730
+    },
+    {
+      "epoch": 165.7,
+      "grad_norm": 1.2025806903839111,
+      "learning_rate": 0.000342962962962963,
+      "loss": 0.6335,
+      "step": 44740
+    },
+    {
+      "epoch": 165.74,
+      "grad_norm": 1.2227649688720703,
+      "learning_rate": 0.00034259259259259263,
+      "loss": 0.6401,
+      "step": 44750
+    },
+    {
+      "epoch": 165.78,
+      "grad_norm": 1.217417597770691,
+      "learning_rate": 0.0003422222222222222,
+      "loss": 0.6424,
+      "step": 44760
+    },
+    {
+      "epoch": 165.81,
+      "grad_norm": 1.2169790267944336,
+      "learning_rate": 0.00034185185185185187,
+      "loss": 0.6449,
+      "step": 44770
+    },
+    {
+      "epoch": 165.85,
+      "grad_norm": 1.1809437274932861,
+      "learning_rate": 0.0003414814814814815,
+      "loss": 0.6452,
+      "step": 44780
+    },
+    {
+      "epoch": 165.89,
+      "grad_norm": 1.1234064102172852,
+      "learning_rate": 0.0003411111111111111,
+      "loss": 0.6436,
+      "step": 44790
+    },
+    {
+      "epoch": 165.93,
+      "grad_norm": 1.191071629524231,
+      "learning_rate": 0.00034074074074074074,
+      "loss": 0.6425,
+      "step": 44800
+    },
+    {
+      "epoch": 165.96,
+      "grad_norm": 1.2422044277191162,
+      "learning_rate": 0.0003403703703703704,
+      "loss": 0.6441,
+      "step": 44810
+    },
+    {
+      "epoch": 166.0,
+      "grad_norm": 1.9944589138031006,
+      "learning_rate": 0.00034,
+      "loss": 0.6568,
+      "step": 44820
+    },
+    {
+      "epoch": 166.04,
+      "grad_norm": 1.0018469095230103,
+      "learning_rate": 0.0003396296296296296,
+      "loss": 0.5225,
+      "step": 44830
+    },
+    {
+      "epoch": 166.07,
+      "grad_norm": 1.1185327768325806,
+      "learning_rate": 0.00033925925925925926,
+      "loss": 0.5186,
+      "step": 44840
+    },
+    {
+      "epoch": 166.11,
+      "grad_norm": 1.0297755002975464,
+      "learning_rate": 0.0003388888888888889,
+      "loss": 0.5285,
+      "step": 44850
+    },
+    {
+      "epoch": 166.15,
+      "grad_norm": 1.054671049118042,
+      "learning_rate": 0.0003385185185185185,
+      "loss": 0.5527,
+      "step": 44860
+    },
+    {
+      "epoch": 166.19,
+      "grad_norm": 1.0687144994735718,
+      "learning_rate": 0.00033814814814814813,
+      "loss": 0.5511,
+      "step": 44870
+    },
+    {
+      "epoch": 166.22,
+      "grad_norm": 1.1111799478530884,
+      "learning_rate": 0.00033777777777777777,
+      "loss": 0.5365,
+      "step": 44880
+    },
+    {
+      "epoch": 166.26,
+      "grad_norm": 1.1225335597991943,
+      "learning_rate": 0.0003374074074074074,
+      "loss": 0.5641,
+      "step": 44890
+    },
+    {
+      "epoch": 166.3,
+      "grad_norm": 1.1786049604415894,
+      "learning_rate": 0.00033703703703703706,
+      "loss": 0.5798,
+      "step": 44900
+    },
+    {
+      "epoch": 166.33,
+      "grad_norm": 1.1004527807235718,
+      "learning_rate": 0.0003366666666666667,
+      "loss": 0.5602,
+      "step": 44910
+    },
+    {
+      "epoch": 166.37,
+      "grad_norm": 1.1575289964675903,
+      "learning_rate": 0.00033629629629629634,
+      "loss": 0.5808,
+      "step": 44920
+    },
+    {
+      "epoch": 166.41,
+      "grad_norm": 1.1519818305969238,
+      "learning_rate": 0.00033592592592592593,
+      "loss": 0.581,
+      "step": 44930
+    },
+    {
+      "epoch": 166.44,
+      "grad_norm": 1.2028568983078003,
+      "learning_rate": 0.0003355555555555556,
+      "loss": 0.5835,
+      "step": 44940
+    },
+    {
+      "epoch": 166.48,
+      "grad_norm": 1.12227201461792,
+      "learning_rate": 0.0003351851851851852,
+      "loss": 0.5909,
+      "step": 44950
+    },
+    {
+      "epoch": 166.52,
+      "grad_norm": 1.1458628177642822,
+      "learning_rate": 0.0003348148148148148,
+      "loss": 0.5777,
+      "step": 44960
+    },
+    {
+      "epoch": 166.56,
+      "grad_norm": 1.2187118530273438,
+      "learning_rate": 0.00033444444444444445,
+      "loss": 0.6024,
+      "step": 44970
+    },
+    {
+      "epoch": 166.59,
+      "grad_norm": 1.142576813697815,
+      "learning_rate": 0.0003340740740740741,
+      "loss": 0.6061,
+      "step": 44980
+    },
+    {
+      "epoch": 166.63,
+      "grad_norm": 1.136089563369751,
+      "learning_rate": 0.00033370370370370373,
+      "loss": 0.6222,
+      "step": 44990
+    },
+    {
+      "epoch": 166.67,
+      "grad_norm": 1.1738704442977905,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 0.6156,
+      "step": 45000
+    },
+    {
+      "epoch": 166.7,
+      "grad_norm": 1.1535799503326416,
+      "learning_rate": 0.00033296296296296296,
+      "loss": 0.6352,
+      "step": 45010
+    },
+    {
+      "epoch": 166.74,
+      "grad_norm": 1.1678482294082642,
+      "learning_rate": 0.0003325925925925926,
+      "loss": 0.612,
+      "step": 45020
+    },
+    {
+      "epoch": 166.78,
+      "grad_norm": 1.1075305938720703,
+      "learning_rate": 0.0003322222222222222,
+      "loss": 0.6254,
+      "step": 45030
+    },
+    {
+      "epoch": 166.81,
+      "grad_norm": 1.2057133913040161,
+      "learning_rate": 0.00033185185185185184,
+      "loss": 0.6143,
+      "step": 45040
+    },
+    {
+      "epoch": 166.85,
+      "grad_norm": 1.150559425354004,
+      "learning_rate": 0.0003314814814814815,
+      "loss": 0.6465,
+      "step": 45050
+    },
+    {
+      "epoch": 166.89,
+      "grad_norm": 1.2744975090026855,
+      "learning_rate": 0.0003311111111111111,
+      "loss": 0.6407,
+      "step": 45060
+    },
+    {
+      "epoch": 166.93,
+      "grad_norm": 1.1772456169128418,
+      "learning_rate": 0.0003307407407407407,
+      "loss": 0.6456,
+      "step": 45070
+    },
+    {
+      "epoch": 166.96,
+      "grad_norm": 1.225772500038147,
+      "learning_rate": 0.00033037037037037036,
+      "loss": 0.6467,
+      "step": 45080
+    },
+    {
+      "epoch": 167.0,
+      "grad_norm": 2.134972095489502,
+      "learning_rate": 0.00033,
+      "loss": 0.6476,
+      "step": 45090
+    },
+    {
+      "epoch": 167.04,
+      "grad_norm": 1.0380076169967651,
+      "learning_rate": 0.0003296296296296296,
+      "loss": 0.503,
+      "step": 45100
+    },
+    {
+      "epoch": 167.07,
+      "grad_norm": 0.9915314316749573,
+      "learning_rate": 0.0003292592592592593,
+      "loss": 0.5193,
+      "step": 45110
+    },
+    {
+      "epoch": 167.11,
+      "grad_norm": 0.9994888305664062,
+      "learning_rate": 0.0003288888888888889,
+      "loss": 0.515,
+      "step": 45120
+    },
+    {
+      "epoch": 167.15,
+      "grad_norm": 1.0514576435089111,
+      "learning_rate": 0.00032851851851851857,
+      "loss": 0.5342,
+      "step": 45130
+    },
+    {
+      "epoch": 167.19,
+      "grad_norm": 0.9884275197982788,
+      "learning_rate": 0.00032814814814814816,
+      "loss": 0.5299,
+      "step": 45140
+    },
+    {
+      "epoch": 167.22,
+      "grad_norm": 1.1106001138687134,
+      "learning_rate": 0.0003277777777777778,
+      "loss": 0.5489,
+      "step": 45150
+    },
+    {
+      "epoch": 167.26,
+      "grad_norm": 1.092725157737732,
+      "learning_rate": 0.00032740740740740744,
+      "loss": 0.5464,
+      "step": 45160
+    },
+    {
+      "epoch": 167.3,
+      "grad_norm": 1.180444359779358,
+      "learning_rate": 0.00032703703703703703,
+      "loss": 0.5625,
+      "step": 45170
+    },
+    {
+      "epoch": 167.33,
+      "grad_norm": 1.0736066102981567,
+      "learning_rate": 0.0003266666666666667,
+      "loss": 0.5595,
+      "step": 45180
+    },
+    {
+      "epoch": 167.37,
+      "grad_norm": 1.1144076585769653,
+      "learning_rate": 0.0003262962962962963,
+      "loss": 0.5551,
+      "step": 45190
+    },
+    {
+      "epoch": 167.41,
+      "grad_norm": 1.1394331455230713,
+      "learning_rate": 0.00032592592592592596,
+      "loss": 0.5559,
+      "step": 45200
+    },
+    {
+      "epoch": 167.44,
+      "grad_norm": 1.0996838808059692,
+      "learning_rate": 0.00032555555555555555,
+      "loss": 0.5846,
+      "step": 45210
+    },
+    {
+      "epoch": 167.48,
+      "grad_norm": 1.1391139030456543,
+      "learning_rate": 0.0003251851851851852,
+      "loss": 0.5875,
+      "step": 45220
+    },
+    {
+      "epoch": 167.52,
+      "grad_norm": 1.0955880880355835,
+      "learning_rate": 0.00032481481481481483,
+      "loss": 0.5841,
+      "step": 45230
+    },
+    {
+      "epoch": 167.56,
+      "grad_norm": 1.1299819946289062,
+      "learning_rate": 0.0003244444444444444,
+      "loss": 0.6052,
+      "step": 45240
+    },
+    {
+      "epoch": 167.59,
+      "grad_norm": 1.1784850358963013,
+      "learning_rate": 0.00032407407407407406,
+      "loss": 0.6128,
+      "step": 45250
+    },
+    {
+      "epoch": 167.63,
+      "grad_norm": 1.2381561994552612,
+      "learning_rate": 0.0003237037037037037,
+      "loss": 0.6095,
+      "step": 45260
+    },
+    {
+      "epoch": 167.67,
+      "grad_norm": 1.1277177333831787,
+      "learning_rate": 0.0003233333333333333,
+      "loss": 0.6038,
+      "step": 45270
+    },
+    {
+      "epoch": 167.7,
+      "grad_norm": 1.1459439992904663,
+      "learning_rate": 0.00032296296296296294,
+      "loss": 0.6173,
+      "step": 45280
+    },
+    {
+      "epoch": 167.74,
+      "grad_norm": 1.2963416576385498,
+      "learning_rate": 0.0003225925925925926,
+      "loss": 0.6303,
+      "step": 45290
+    },
+    {
+      "epoch": 167.78,
+      "grad_norm": 1.1779850721359253,
+      "learning_rate": 0.0003222222222222222,
+      "loss": 0.6227,
+      "step": 45300
+    },
+    {
+      "epoch": 167.81,
+      "grad_norm": 1.140274167060852,
+      "learning_rate": 0.00032185185185185187,
+      "loss": 0.6168,
+      "step": 45310
+    },
+    {
+      "epoch": 167.85,
+      "grad_norm": 1.3273096084594727,
+      "learning_rate": 0.0003214814814814815,
+      "loss": 0.6326,
+      "step": 45320
+    },
+    {
+      "epoch": 167.89,
+      "grad_norm": 1.1361197233200073,
+      "learning_rate": 0.00032111111111111115,
+      "loss": 0.6225,
+      "step": 45330
+    },
+    {
+      "epoch": 167.93,
+      "grad_norm": 1.1830846071243286,
+      "learning_rate": 0.00032074074074074074,
+      "loss": 0.6335,
+      "step": 45340
+    },
+    {
+      "epoch": 167.96,
+      "grad_norm": 1.2070780992507935,
+      "learning_rate": 0.0003203703703703704,
+      "loss": 0.6328,
+      "step": 45350
+    },
+    {
+      "epoch": 168.0,
+      "grad_norm": 2.6271791458129883,
+      "learning_rate": 0.00032,
+      "loss": 0.6206,
+      "step": 45360
+    },
+    {
+      "epoch": 168.04,
+      "grad_norm": 1.031340479850769,
+      "learning_rate": 0.00031962962962962967,
+      "loss": 0.5002,
+      "step": 45370
+    },
+    {
+      "epoch": 168.07,
+      "grad_norm": 1.0149120092391968,
+      "learning_rate": 0.00031925925925925926,
+      "loss": 0.5076,
+      "step": 45380
+    },
+    {
+      "epoch": 168.11,
+      "grad_norm": 1.0120203495025635,
+      "learning_rate": 0.0003188888888888889,
+      "loss": 0.5156,
+      "step": 45390
+    },
+    {
+      "epoch": 168.15,
+      "grad_norm": 1.0357682704925537,
+      "learning_rate": 0.00031851851851851854,
+      "loss": 0.5137,
+      "step": 45400
+    },
+    {
+      "epoch": 168.19,
+      "grad_norm": 1.0558855533599854,
+      "learning_rate": 0.00031814814814814813,
+      "loss": 0.523,
+      "step": 45410
+    },
+    {
+      "epoch": 168.22,
+      "grad_norm": 1.186493992805481,
+      "learning_rate": 0.0003177777777777778,
+      "loss": 0.5383,
+      "step": 45420
+    },
+    {
+      "epoch": 168.26,
+      "grad_norm": 1.0746095180511475,
+      "learning_rate": 0.0003174074074074074,
+      "loss": 0.5434,
+      "step": 45430
+    },
+    {
+      "epoch": 168.3,
+      "grad_norm": 1.0704513788223267,
+      "learning_rate": 0.00031703703703703706,
+      "loss": 0.5586,
+      "step": 45440
+    },
+    {
+      "epoch": 168.33,
+      "grad_norm": 1.1498819589614868,
+      "learning_rate": 0.00031666666666666665,
+      "loss": 0.5414,
+      "step": 45450
+    },
+    {
+      "epoch": 168.37,
+      "grad_norm": 1.046481728553772,
+      "learning_rate": 0.0003162962962962963,
+      "loss": 0.5719,
+      "step": 45460
+    },
+    {
+      "epoch": 168.41,
+      "grad_norm": 1.098426342010498,
+      "learning_rate": 0.00031592592592592593,
+      "loss": 0.559,
+      "step": 45470
+    },
+    {
+      "epoch": 168.44,
+      "grad_norm": 1.1073126792907715,
+      "learning_rate": 0.0003155555555555555,
+      "loss": 0.5704,
+      "step": 45480
+    },
+    {
+      "epoch": 168.48,
+      "grad_norm": 1.1616339683532715,
+      "learning_rate": 0.00031518518518518516,
+      "loss": 0.577,
+      "step": 45490
+    },
+    {
+      "epoch": 168.52,
+      "grad_norm": 1.1858264207839966,
+      "learning_rate": 0.0003148148148148148,
+      "loss": 0.5877,
+      "step": 45500
+    },
+    {
+      "epoch": 168.56,
+      "grad_norm": 1.1451951265335083,
+      "learning_rate": 0.0003144444444444445,
+      "loss": 0.5846,
+      "step": 45510
+    },
+    {
+      "epoch": 168.59,
+      "grad_norm": 1.1581388711929321,
+      "learning_rate": 0.0003140740740740741,
+      "loss": 0.603,
+      "step": 45520
+    },
+    {
+      "epoch": 168.63,
+      "grad_norm": 1.1635332107543945,
+      "learning_rate": 0.00031370370370370374,
+      "loss": 0.5953,
+      "step": 45530
+    },
+    {
+      "epoch": 168.67,
+      "grad_norm": 1.1322771310806274,
+      "learning_rate": 0.0003133333333333334,
+      "loss": 0.5944,
+      "step": 45540
+    },
+    {
+      "epoch": 168.7,
+      "grad_norm": 1.2044880390167236,
+      "learning_rate": 0.00031296296296296297,
+      "loss": 0.6159,
+      "step": 45550
+    },
+    {
+      "epoch": 168.74,
+      "grad_norm": 1.1870005130767822,
+      "learning_rate": 0.0003125925925925926,
+      "loss": 0.5962,
+      "step": 45560
+    },
+    {
+      "epoch": 168.78,
+      "grad_norm": 1.1659802198410034,
+      "learning_rate": 0.00031222222222222225,
+      "loss": 0.5992,
+      "step": 45570
+    },
+    {
+      "epoch": 168.81,
+      "grad_norm": 1.113723635673523,
+      "learning_rate": 0.00031185185185185184,
+      "loss": 0.6133,
+      "step": 45580
+    },
+    {
+      "epoch": 168.85,
+      "grad_norm": 1.2763746976852417,
+      "learning_rate": 0.0003114814814814815,
+      "loss": 0.627,
+      "step": 45590
+    },
+    {
+      "epoch": 168.89,
+      "grad_norm": 1.1561163663864136,
+      "learning_rate": 0.0003111111111111111,
+      "loss": 0.6268,
+      "step": 45600
+    },
+    {
+      "epoch": 168.93,
+      "grad_norm": 1.2296404838562012,
+      "learning_rate": 0.00031074074074074077,
+      "loss": 0.613,
+      "step": 45610
+    },
+    {
+      "epoch": 168.96,
+      "grad_norm": 1.1916122436523438,
+      "learning_rate": 0.00031037037037037036,
+      "loss": 0.6283,
+      "step": 45620
+    },
+    {
+      "epoch": 169.0,
+      "grad_norm": 2.852231740951538,
+      "learning_rate": 0.00031,
+      "loss": 0.6324,
+      "step": 45630
+    },
+    {
+      "epoch": 169.04,
+      "grad_norm": 1.0171003341674805,
+      "learning_rate": 0.00030962962962962964,
+      "loss": 0.4882,
+      "step": 45640
+    },
+    {
+      "epoch": 169.07,
+      "grad_norm": 1.0816006660461426,
+      "learning_rate": 0.00030925925925925923,
+      "loss": 0.4913,
+      "step": 45650
+    },
+    {
+      "epoch": 169.11,
+      "grad_norm": 1.1123528480529785,
+      "learning_rate": 0.0003088888888888889,
+      "loss": 0.5073,
+      "step": 45660
+    },
+    {
+      "epoch": 169.15,
+      "grad_norm": 1.0543406009674072,
+      "learning_rate": 0.0003085185185185185,
+      "loss": 0.5064,
+      "step": 45670
+    },
+    {
+      "epoch": 169.19,
+      "grad_norm": 1.0657761096954346,
+      "learning_rate": 0.00030814814814814816,
+      "loss": 0.5045,
+      "step": 45680
+    },
+    {
+      "epoch": 169.22,
+      "grad_norm": 1.077353596687317,
+      "learning_rate": 0.00030777777777777775,
+      "loss": 0.534,
+      "step": 45690
+    },
+    {
+      "epoch": 169.26,
+      "grad_norm": 1.0854156017303467,
+      "learning_rate": 0.0003074074074074074,
+      "loss": 0.5381,
+      "step": 45700
+    },
+    {
+      "epoch": 169.3,
+      "grad_norm": 1.0773741006851196,
+      "learning_rate": 0.00030703703703703703,
+      "loss": 0.5381,
+      "step": 45710
+    },
+    {
+      "epoch": 169.33,
+      "grad_norm": 1.239553451538086,
+      "learning_rate": 0.0003066666666666667,
+      "loss": 0.5517,
+      "step": 45720
+    },
+    {
+      "epoch": 169.37,
+      "grad_norm": 1.1670728921890259,
+      "learning_rate": 0.0003062962962962963,
+      "loss": 0.5636,
+      "step": 45730
+    },
+    {
+      "epoch": 169.41,
+      "grad_norm": 1.0546120405197144,
+      "learning_rate": 0.00030592592592592596,
+      "loss": 0.5617,
+      "step": 45740
+    },
+    {
+      "epoch": 169.44,
+      "grad_norm": 1.1616939306259155,
+      "learning_rate": 0.0003055555555555556,
+      "loss": 0.5634,
+      "step": 45750
+    },
+    {
+      "epoch": 169.48,
+      "grad_norm": 1.1171976327896118,
+      "learning_rate": 0.0003051851851851852,
+      "loss": 0.5781,
+      "step": 45760
+    },
+    {
+      "epoch": 169.52,
+      "grad_norm": 1.0559581518173218,
+      "learning_rate": 0.00030481481481481484,
+      "loss": 0.5842,
+      "step": 45770
+    },
+    {
+      "epoch": 169.56,
+      "grad_norm": 1.1623177528381348,
+      "learning_rate": 0.0003044444444444445,
+      "loss": 0.5636,
+      "step": 45780
+    },
+    {
+      "epoch": 169.59,
+      "grad_norm": 1.1423313617706299,
+      "learning_rate": 0.00030407407407407407,
+      "loss": 0.5774,
+      "step": 45790
+    },
+    {
+      "epoch": 169.63,
+      "grad_norm": 1.2008979320526123,
+      "learning_rate": 0.0003037037037037037,
+      "loss": 0.5891,
+      "step": 45800
+    },
+    {
+      "epoch": 169.67,
+      "grad_norm": 1.146252155303955,
+      "learning_rate": 0.00030333333333333335,
+      "loss": 0.6024,
+      "step": 45810
+    },
+    {
+      "epoch": 169.7,
+      "grad_norm": 1.2153797149658203,
+      "learning_rate": 0.000302962962962963,
+      "loss": 0.5919,
+      "step": 45820
+    },
+    {
+      "epoch": 169.74,
+      "grad_norm": 1.2314988374710083,
+      "learning_rate": 0.0003025925925925926,
+      "loss": 0.5962,
+      "step": 45830
+    },
+    {
+      "epoch": 169.78,
+      "grad_norm": 1.1446729898452759,
+      "learning_rate": 0.0003022222222222222,
+      "loss": 0.5979,
+      "step": 45840
+    },
+    {
+      "epoch": 169.81,
+      "grad_norm": 1.1876449584960938,
+      "learning_rate": 0.00030185185185185187,
+      "loss": 0.5944,
+      "step": 45850
+    },
+    {
+      "epoch": 169.85,
+      "grad_norm": 1.222914695739746,
+      "learning_rate": 0.00030148148148148146,
+      "loss": 0.6196,
+      "step": 45860
+    },
+    {
+      "epoch": 169.89,
+      "grad_norm": 1.2583873271942139,
+      "learning_rate": 0.0003011111111111111,
+      "loss": 0.6079,
+      "step": 45870
+    },
+    {
+      "epoch": 169.93,
+      "grad_norm": 1.2030463218688965,
+      "learning_rate": 0.00030074074074074074,
+      "loss": 0.6167,
+      "step": 45880
+    },
+    {
+      "epoch": 169.96,
+      "grad_norm": 1.1214995384216309,
+      "learning_rate": 0.00030037037037037033,
+      "loss": 0.6141,
+      "step": 45890
+    },
+    {
+      "epoch": 170.0,
+      "grad_norm": 2.046583652496338,
+      "learning_rate": 0.0003,
+      "loss": 0.6237,
+      "step": 45900
+    },
+    {
+      "epoch": 170.04,
+      "grad_norm": 0.9938816428184509,
+      "learning_rate": 0.0002996296296296296,
+      "loss": 0.4945,
+      "step": 45910
+    },
+    {
+      "epoch": 170.07,
+      "grad_norm": 1.0744401216506958,
+      "learning_rate": 0.0002992592592592593,
+      "loss": 0.4909,
+      "step": 45920
+    },
+    {
+      "epoch": 170.11,
+      "grad_norm": 1.006912350654602,
+      "learning_rate": 0.0002988888888888889,
+      "loss": 0.5027,
+      "step": 45930
+    },
+    {
+      "epoch": 170.15,
+      "grad_norm": 1.0353368520736694,
+      "learning_rate": 0.00029851851851851854,
+      "loss": 0.5027,
+      "step": 45940
+    },
+    {
+      "epoch": 170.19,
+      "grad_norm": 1.0443346500396729,
+      "learning_rate": 0.0002981481481481482,
+      "loss": 0.5333,
+      "step": 45950
+    },
+    {
+      "epoch": 170.22,
+      "grad_norm": 1.082785964012146,
+      "learning_rate": 0.0002977777777777778,
+      "loss": 0.5197,
+      "step": 45960
+    },
+    {
+      "epoch": 170.26,
+      "grad_norm": 1.0719095468521118,
+      "learning_rate": 0.0002974074074074074,
+      "loss": 0.5193,
+      "step": 45970
+    },
+    {
+      "epoch": 170.3,
+      "grad_norm": 1.0522955656051636,
+      "learning_rate": 0.00029703703703703706,
+      "loss": 0.5319,
+      "step": 45980
+    },
+    {
+      "epoch": 170.33,
+      "grad_norm": 1.0776923894882202,
+      "learning_rate": 0.0002966666666666667,
+      "loss": 0.5204,
+      "step": 45990
+    },
+    {
+      "epoch": 170.37,
+      "grad_norm": 1.1122677326202393,
+      "learning_rate": 0.0002962962962962963,
+      "loss": 0.5291,
+      "step": 46000
+    },
+    {
+      "epoch": 170.41,
+      "grad_norm": 1.1635634899139404,
+      "learning_rate": 0.00029592592592592594,
+      "loss": 0.5479,
+      "step": 46010
+    },
+    {
+      "epoch": 170.44,
+      "grad_norm": 1.2051600217819214,
+      "learning_rate": 0.0002955555555555556,
+      "loss": 0.5419,
+      "step": 46020
+    },
+    {
+      "epoch": 170.48,
+      "grad_norm": 1.1886661052703857,
+      "learning_rate": 0.00029518518518518517,
+      "loss": 0.571,
+      "step": 46030
+    },
+    {
+      "epoch": 170.52,
+      "grad_norm": 1.1528897285461426,
+      "learning_rate": 0.0002948148148148148,
+      "loss": 0.5575,
+      "step": 46040
+    },
+    {
+      "epoch": 170.56,
+      "grad_norm": 1.2057989835739136,
+      "learning_rate": 0.00029444444444444445,
+      "loss": 0.5681,
+      "step": 46050
+    },
+    {
+      "epoch": 170.59,
+      "grad_norm": 1.1691752672195435,
+      "learning_rate": 0.0002940740740740741,
+      "loss": 0.573,
+      "step": 46060
+    },
+    {
+      "epoch": 170.63,
+      "grad_norm": 1.1771482229232788,
+      "learning_rate": 0.0002937037037037037,
+      "loss": 0.5758,
+      "step": 46070
+    },
+    {
+      "epoch": 170.67,
+      "grad_norm": 1.1611257791519165,
+      "learning_rate": 0.0002933333333333333,
+      "loss": 0.5841,
+      "step": 46080
+    },
+    {
+      "epoch": 170.7,
+      "grad_norm": 1.1115514039993286,
+      "learning_rate": 0.00029296296296296297,
+      "loss": 0.5776,
+      "step": 46090
+    },
+    {
+      "epoch": 170.74,
+      "grad_norm": 1.1738566160202026,
+      "learning_rate": 0.00029259259259259256,
+      "loss": 0.594,
+      "step": 46100
+    },
+    {
+      "epoch": 170.78,
+      "grad_norm": 1.2470712661743164,
+      "learning_rate": 0.0002922222222222222,
+      "loss": 0.6051,
+      "step": 46110
+    },
+    {
+      "epoch": 170.81,
+      "grad_norm": 1.1863471269607544,
+      "learning_rate": 0.00029185185185185184,
+      "loss": 0.6104,
+      "step": 46120
+    },
+    {
+      "epoch": 170.85,
+      "grad_norm": 1.0882984399795532,
+      "learning_rate": 0.00029148148148148154,
+      "loss": 0.605,
+      "step": 46130
+    },
+    {
+      "epoch": 170.89,
+      "grad_norm": 1.20223069190979,
+      "learning_rate": 0.00029111111111111113,
+      "loss": 0.5981,
+      "step": 46140
+    },
+    {
+      "epoch": 170.93,
+      "grad_norm": 1.1262892484664917,
+      "learning_rate": 0.00029074074074074077,
+      "loss": 0.6117,
+      "step": 46150
+    },
+    {
+      "epoch": 170.96,
+      "grad_norm": 1.2084956169128418,
+      "learning_rate": 0.0002903703703703704,
+      "loss": 0.6077,
+      "step": 46160
+    },
+    {
+      "epoch": 171.0,
+      "grad_norm": 2.073090076446533,
+      "learning_rate": 0.00029,
+      "loss": 0.6047,
+      "step": 46170
+    },
+    {
+      "epoch": 171.04,
+      "grad_norm": 0.9190633296966553,
+      "learning_rate": 0.00028962962962962964,
+      "loss": 0.4772,
+      "step": 46180
+    },
+    {
+      "epoch": 171.07,
+      "grad_norm": 1.0562139749526978,
+      "learning_rate": 0.0002892592592592593,
+      "loss": 0.4771,
+      "step": 46190
+    },
+    {
+      "epoch": 171.11,
+      "grad_norm": 1.0270583629608154,
+      "learning_rate": 0.0002888888888888889,
+      "loss": 0.4823,
+      "step": 46200
+    },
+    {
+      "epoch": 171.15,
+      "grad_norm": 0.9361873865127563,
+      "learning_rate": 0.0002885185185185185,
+      "loss": 0.4957,
+      "step": 46210
+    },
+    {
+      "epoch": 171.19,
+      "grad_norm": 1.0008320808410645,
+      "learning_rate": 0.00028814814814814816,
+      "loss": 0.5127,
+      "step": 46220
+    },
+    {
+      "epoch": 171.22,
+      "grad_norm": 0.9977531433105469,
+      "learning_rate": 0.0002877777777777778,
+      "loss": 0.497,
+      "step": 46230
+    },
+    {
+      "epoch": 171.26,
+      "grad_norm": 1.1229784488677979,
+      "learning_rate": 0.0002874074074074074,
+      "loss": 0.5163,
+      "step": 46240
+    },
+    {
+      "epoch": 171.3,
+      "grad_norm": 1.0604517459869385,
+      "learning_rate": 0.00028703703703703703,
+      "loss": 0.5257,
+      "step": 46250
+    },
+    {
+      "epoch": 171.33,
+      "grad_norm": 1.1533794403076172,
+      "learning_rate": 0.0002866666666666667,
+      "loss": 0.5284,
+      "step": 46260
+    },
+    {
+      "epoch": 171.37,
+      "grad_norm": 1.1037262678146362,
+      "learning_rate": 0.00028629629629629627,
+      "loss": 0.5391,
+      "step": 46270
+    },
+    {
+      "epoch": 171.41,
+      "grad_norm": 1.1206879615783691,
+      "learning_rate": 0.0002859259259259259,
+      "loss": 0.5462,
+      "step": 46280
+    },
+    {
+      "epoch": 171.44,
+      "grad_norm": 1.1554394960403442,
+      "learning_rate": 0.00028555555555555555,
+      "loss": 0.5422,
+      "step": 46290
+    },
+    {
+      "epoch": 171.48,
+      "grad_norm": 1.1456857919692993,
+      "learning_rate": 0.0002851851851851852,
+      "loss": 0.5466,
+      "step": 46300
+    },
+    {
+      "epoch": 171.52,
+      "grad_norm": 1.1105420589447021,
+      "learning_rate": 0.0002848148148148148,
+      "loss": 0.5529,
+      "step": 46310
+    },
+    {
+      "epoch": 171.56,
+      "grad_norm": 1.1906827688217163,
+      "learning_rate": 0.0002844444444444444,
+      "loss": 0.5663,
+      "step": 46320
+    },
+    {
+      "epoch": 171.59,
+      "grad_norm": 1.1303131580352783,
+      "learning_rate": 0.0002840740740740741,
+      "loss": 0.5582,
+      "step": 46330
+    },
+    {
+      "epoch": 171.63,
+      "grad_norm": 1.1390284299850464,
+      "learning_rate": 0.0002837037037037037,
+      "loss": 0.5775,
+      "step": 46340
+    },
+    {
+      "epoch": 171.67,
+      "grad_norm": 1.1789822578430176,
+      "learning_rate": 0.00028333333333333335,
+      "loss": 0.5834,
+      "step": 46350
+    },
+    {
+      "epoch": 171.7,
+      "grad_norm": 1.152759075164795,
+      "learning_rate": 0.000282962962962963,
+      "loss": 0.5897,
+      "step": 46360
+    },
+    {
+      "epoch": 171.74,
+      "grad_norm": 1.1558884382247925,
+      "learning_rate": 0.00028259259259259264,
+      "loss": 0.59,
+      "step": 46370
+    },
+    {
+      "epoch": 171.78,
+      "grad_norm": 1.1207094192504883,
+      "learning_rate": 0.00028222222222222223,
+      "loss": 0.5746,
+      "step": 46380
+    },
+    {
+      "epoch": 171.81,
+      "grad_norm": 1.1491044759750366,
+      "learning_rate": 0.00028185185185185187,
+      "loss": 0.588,
+      "step": 46390
+    },
+    {
+      "epoch": 171.85,
+      "grad_norm": 1.1739957332611084,
+      "learning_rate": 0.0002814814814814815,
+      "loss": 0.6055,
+      "step": 46400
+    },
+    {
+      "epoch": 171.89,
+      "grad_norm": 1.1879185438156128,
+      "learning_rate": 0.0002811111111111111,
+      "loss": 0.5832,
+      "step": 46410
+    },
+    {
+      "epoch": 171.93,
+      "grad_norm": 1.1847784519195557,
+      "learning_rate": 0.00028074074074074074,
+      "loss": 0.6117,
+      "step": 46420
+    },
+    {
+      "epoch": 171.96,
+      "grad_norm": 1.1747502088546753,
+      "learning_rate": 0.0002803703703703704,
+      "loss": 0.5978,
+      "step": 46430
+    },
+    {
+      "epoch": 172.0,
+      "grad_norm": 2.1897666454315186,
+      "learning_rate": 0.00028000000000000003,
+      "loss": 0.6104,
+      "step": 46440
+    },
+    {
+      "epoch": 172.04,
+      "grad_norm": 0.9945102334022522,
+      "learning_rate": 0.0002796296296296296,
+      "loss": 0.479,
+      "step": 46450
+    },
+    {
+      "epoch": 172.07,
+      "grad_norm": 0.9526515007019043,
+      "learning_rate": 0.00027925925925925926,
+      "loss": 0.4701,
+      "step": 46460
+    },
+    {
+      "epoch": 172.11,
+      "grad_norm": 1.0737017393112183,
+      "learning_rate": 0.0002788888888888889,
+      "loss": 0.4879,
+      "step": 46470
+    },
+    {
+      "epoch": 172.15,
+      "grad_norm": 1.036736249923706,
+      "learning_rate": 0.0002785185185185185,
+      "loss": 0.4923,
+      "step": 46480
+    },
+    {
+      "epoch": 172.19,
+      "grad_norm": 1.0772678852081299,
+      "learning_rate": 0.00027814814814814813,
+      "loss": 0.5052,
+      "step": 46490
+    },
+    {
+      "epoch": 172.22,
+      "grad_norm": 1.1343975067138672,
+      "learning_rate": 0.0002777777777777778,
+      "loss": 0.4975,
+      "step": 46500
+    },
+    {
+      "epoch": 172.26,
+      "grad_norm": 1.077694058418274,
+      "learning_rate": 0.00027740740740740737,
+      "loss": 0.507,
+      "step": 46510
+    },
+    {
+      "epoch": 172.3,
+      "grad_norm": 1.0724430084228516,
+      "learning_rate": 0.000277037037037037,
+      "loss": 0.5183,
+      "step": 46520
+    },
+    {
+      "epoch": 172.33,
+      "grad_norm": 1.1396112442016602,
+      "learning_rate": 0.00027666666666666665,
+      "loss": 0.5216,
+      "step": 46530
+    },
+    {
+      "epoch": 172.37,
+      "grad_norm": 1.0889900922775269,
+      "learning_rate": 0.00027629629629629635,
+      "loss": 0.5374,
+      "step": 46540
+    },
+    {
+      "epoch": 172.41,
+      "grad_norm": 1.0954948663711548,
+      "learning_rate": 0.00027592592592592594,
+      "loss": 0.5285,
+      "step": 46550
+    },
+    {
+      "epoch": 172.44,
+      "grad_norm": 1.1125346422195435,
+      "learning_rate": 0.0002755555555555556,
+      "loss": 0.5407,
+      "step": 46560
+    },
+    {
+      "epoch": 172.48,
+      "grad_norm": 1.0506683588027954,
+      "learning_rate": 0.0002751851851851852,
+      "loss": 0.5337,
+      "step": 46570
+    },
+    {
+      "epoch": 172.52,
+      "grad_norm": 1.1666901111602783,
+      "learning_rate": 0.0002748148148148148,
+      "loss": 0.5568,
+      "step": 46580
+    },
+    {
+      "epoch": 172.56,
+      "grad_norm": 1.1451871395111084,
+      "learning_rate": 0.00027444444444444445,
+      "loss": 0.5502,
+      "step": 46590
+    },
+    {
+      "epoch": 172.59,
+      "grad_norm": 1.1013054847717285,
+      "learning_rate": 0.0002740740740740741,
+      "loss": 0.5603,
+      "step": 46600
+    },
+    {
+      "epoch": 172.63,
+      "grad_norm": 1.1822503805160522,
+      "learning_rate": 0.00027370370370370374,
+      "loss": 0.5711,
+      "step": 46610
+    },
+    {
+      "epoch": 172.67,
+      "grad_norm": 1.215627908706665,
+      "learning_rate": 0.00027333333333333333,
+      "loss": 0.5529,
+      "step": 46620
+    },
+    {
+      "epoch": 172.7,
+      "grad_norm": 1.1580448150634766,
+      "learning_rate": 0.00027296296296296297,
+      "loss": 0.5801,
+      "step": 46630
+    },
+    {
+      "epoch": 172.74,
+      "grad_norm": 1.1888889074325562,
+      "learning_rate": 0.0002725925925925926,
+      "loss": 0.5762,
+      "step": 46640
+    },
+    {
+      "epoch": 172.78,
+      "grad_norm": 1.1299220323562622,
+      "learning_rate": 0.0002722222222222222,
+      "loss": 0.5743,
+      "step": 46650
+    },
+    {
+      "epoch": 172.81,
+      "grad_norm": 1.142557978630066,
+      "learning_rate": 0.00027185185185185184,
+      "loss": 0.5833,
+      "step": 46660
+    },
+    {
+      "epoch": 172.85,
+      "grad_norm": 1.1916149854660034,
+      "learning_rate": 0.0002714814814814815,
+      "loss": 0.5906,
+      "step": 46670
+    },
+    {
+      "epoch": 172.89,
+      "grad_norm": 1.1581250429153442,
+      "learning_rate": 0.00027111111111111113,
+      "loss": 0.5878,
+      "step": 46680
+    },
+    {
+      "epoch": 172.93,
+      "grad_norm": 1.2044334411621094,
+      "learning_rate": 0.0002707407407407407,
+      "loss": 0.5708,
+      "step": 46690
+    },
+    {
+      "epoch": 172.96,
+      "grad_norm": 1.279520869255066,
+      "learning_rate": 0.00027037037037037036,
+      "loss": 0.597,
+      "step": 46700
+    },
+    {
+      "epoch": 173.0,
+      "grad_norm": 2.107897996902466,
+      "learning_rate": 0.00027,
+      "loss": 0.5836,
+      "step": 46710
+    },
+    {
+      "epoch": 173.04,
+      "grad_norm": 1.0807936191558838,
+      "learning_rate": 0.0002696296296296296,
+      "loss": 0.4548,
+      "step": 46720
+    },
+    {
+      "epoch": 173.07,
+      "grad_norm": 0.9994587302207947,
+      "learning_rate": 0.00026925925925925923,
+      "loss": 0.4707,
+      "step": 46730
+    },
+    {
+      "epoch": 173.11,
+      "grad_norm": 1.0478770732879639,
+      "learning_rate": 0.00026888888888888893,
+      "loss": 0.4722,
+      "step": 46740
+    },
+    {
+      "epoch": 173.15,
+      "grad_norm": 1.013415813446045,
+      "learning_rate": 0.0002685185185185186,
+      "loss": 0.5006,
+      "step": 46750
+    },
+    {
+      "epoch": 173.19,
+      "grad_norm": 1.012677550315857,
+      "learning_rate": 0.00026814814814814816,
+      "loss": 0.4959,
+      "step": 46760
+    },
+    {
+      "epoch": 173.22,
+      "grad_norm": 1.0860716104507446,
+      "learning_rate": 0.0002677777777777778,
+      "loss": 0.4915,
+      "step": 46770
+    },
+    {
+      "epoch": 173.26,
+      "grad_norm": 1.157912254333496,
+      "learning_rate": 0.00026740740740740745,
+      "loss": 0.5134,
+      "step": 46780
+    },
+    {
+      "epoch": 173.3,
+      "grad_norm": 1.1048921346664429,
+      "learning_rate": 0.00026703703703703704,
+      "loss": 0.5078,
+      "step": 46790
+    },
+    {
+      "epoch": 173.33,
+      "grad_norm": 1.1350382566452026,
+      "learning_rate": 0.0002666666666666667,
+      "loss": 0.5103,
+      "step": 46800
+    },
+    {
+      "epoch": 173.37,
+      "grad_norm": 1.0764796733856201,
+      "learning_rate": 0.0002662962962962963,
+      "loss": 0.5152,
+      "step": 46810
+    },
+    {
+      "epoch": 173.41,
+      "grad_norm": 1.0592381954193115,
+      "learning_rate": 0.0002659259259259259,
+      "loss": 0.5038,
+      "step": 46820
+    },
+    {
+      "epoch": 173.44,
+      "grad_norm": 1.0957568883895874,
+      "learning_rate": 0.00026555555555555555,
+      "loss": 0.5321,
+      "step": 46830
+    },
+    {
+      "epoch": 173.48,
+      "grad_norm": 1.0465227365493774,
+      "learning_rate": 0.0002651851851851852,
+      "loss": 0.5263,
+      "step": 46840
+    },
+    {
+      "epoch": 173.52,
+      "grad_norm": 1.106717824935913,
+      "learning_rate": 0.00026481481481481484,
+      "loss": 0.5371,
+      "step": 46850
+    },
+    {
+      "epoch": 173.56,
+      "grad_norm": 1.0666769742965698,
+      "learning_rate": 0.00026444444444444443,
+      "loss": 0.5489,
+      "step": 46860
+    },
+    {
+      "epoch": 173.59,
+      "grad_norm": 1.1631163358688354,
+      "learning_rate": 0.00026407407407407407,
+      "loss": 0.5543,
+      "step": 46870
+    },
+    {
+      "epoch": 173.63,
+      "grad_norm": 1.2045526504516602,
+      "learning_rate": 0.0002637037037037037,
+      "loss": 0.5638,
+      "step": 46880
+    },
+    {
+      "epoch": 173.67,
+      "grad_norm": 1.1705162525177002,
+      "learning_rate": 0.0002633333333333333,
+      "loss": 0.5475,
+      "step": 46890
+    },
+    {
+      "epoch": 173.7,
+      "grad_norm": 1.1357591152191162,
+      "learning_rate": 0.00026296296296296294,
+      "loss": 0.5737,
+      "step": 46900
+    },
+    {
+      "epoch": 173.74,
+      "grad_norm": 1.2002065181732178,
+      "learning_rate": 0.0002625925925925926,
+      "loss": 0.5689,
+      "step": 46910
+    },
+    {
+      "epoch": 173.78,
+      "grad_norm": 1.1841059923171997,
+      "learning_rate": 0.00026222222222222223,
+      "loss": 0.5563,
+      "step": 46920
+    },
+    {
+      "epoch": 173.81,
+      "grad_norm": 1.2427780628204346,
+      "learning_rate": 0.0002618518518518518,
+      "loss": 0.5768,
+      "step": 46930
+    },
+    {
+      "epoch": 173.85,
+      "grad_norm": 1.1587187051773071,
+      "learning_rate": 0.00026148148148148146,
+      "loss": 0.5747,
+      "step": 46940
+    },
+    {
+      "epoch": 173.89,
+      "grad_norm": 1.2276320457458496,
+      "learning_rate": 0.00026111111111111116,
+      "loss": 0.5696,
+      "step": 46950
+    },
+    {
+      "epoch": 173.93,
+      "grad_norm": 1.2079700231552124,
+      "learning_rate": 0.00026074074074074075,
+      "loss": 0.5909,
+      "step": 46960
+    },
+    {
+      "epoch": 173.96,
+      "grad_norm": 1.270209789276123,
+      "learning_rate": 0.0002603703703703704,
+      "loss": 0.5803,
+      "step": 46970
+    },
+    {
+      "epoch": 174.0,
+      "grad_norm": 2.246609687805176,
+      "learning_rate": 0.00026000000000000003,
+      "loss": 0.5801,
+      "step": 46980
+    },
+    {
+      "epoch": 174.04,
+      "grad_norm": 0.9848663806915283,
+      "learning_rate": 0.0002596296296296297,
+      "loss": 0.4537,
+      "step": 46990
+    },
+    {
+      "epoch": 174.07,
+      "grad_norm": 1.026583194732666,
+      "learning_rate": 0.00025925925925925926,
+      "loss": 0.4583,
+      "step": 47000
+    },
+    {
+      "epoch": 174.11,
+      "grad_norm": 1.0775405168533325,
+      "learning_rate": 0.0002588888888888889,
+      "loss": 0.4698,
+      "step": 47010
+    },
+    {
+      "epoch": 174.15,
+      "grad_norm": 0.9840408563613892,
+      "learning_rate": 0.00025851851851851855,
+      "loss": 0.4818,
+      "step": 47020
+    },
+    {
+      "epoch": 174.19,
+      "grad_norm": 0.9654091000556946,
+      "learning_rate": 0.00025814814814814814,
+      "loss": 0.4712,
+      "step": 47030
+    },
+    {
+      "epoch": 174.22,
+      "grad_norm": 1.057830810546875,
+      "learning_rate": 0.0002577777777777778,
+      "loss": 0.4991,
+      "step": 47040
+    },
+    {
+      "epoch": 174.26,
+      "grad_norm": 1.0927250385284424,
+      "learning_rate": 0.0002574074074074074,
+      "loss": 0.4966,
+      "step": 47050
+    },
+    {
+      "epoch": 174.3,
+      "grad_norm": 1.0779294967651367,
+      "learning_rate": 0.00025703703703703706,
+      "loss": 0.5209,
+      "step": 47060
+    },
+    {
+      "epoch": 174.33,
+      "grad_norm": 1.0855543613433838,
+      "learning_rate": 0.00025666666666666665,
+      "loss": 0.5019,
+      "step": 47070
+    },
+    {
+      "epoch": 174.37,
+      "grad_norm": 1.1064202785491943,
+      "learning_rate": 0.0002562962962962963,
+      "loss": 0.5069,
+      "step": 47080
+    },
+    {
+      "epoch": 174.41,
+      "grad_norm": 1.0853049755096436,
+      "learning_rate": 0.00025592592592592594,
+      "loss": 0.5084,
+      "step": 47090
+    },
+    {
+      "epoch": 174.44,
+      "grad_norm": 1.0846391916275024,
+      "learning_rate": 0.00025555555555555553,
+      "loss": 0.5129,
+      "step": 47100
+    },
+    {
+      "epoch": 174.48,
+      "grad_norm": 1.1598272323608398,
+      "learning_rate": 0.00025518518518518517,
+      "loss": 0.5432,
+      "step": 47110
+    },
+    {
+      "epoch": 174.52,
+      "grad_norm": 1.1339960098266602,
+      "learning_rate": 0.0002548148148148148,
+      "loss": 0.5323,
+      "step": 47120
+    },
+    {
+      "epoch": 174.56,
+      "grad_norm": 1.1392383575439453,
+      "learning_rate": 0.0002544444444444444,
+      "loss": 0.5429,
+      "step": 47130
+    },
+    {
+      "epoch": 174.59,
+      "grad_norm": 1.126197338104248,
+      "learning_rate": 0.00025407407407407404,
+      "loss": 0.5453,
+      "step": 47140
+    },
+    {
+      "epoch": 174.63,
+      "grad_norm": 1.1894769668579102,
+      "learning_rate": 0.00025370370370370374,
+      "loss": 0.5416,
+      "step": 47150
+    },
+    {
+      "epoch": 174.67,
+      "grad_norm": 1.1315686702728271,
+      "learning_rate": 0.0002533333333333334,
+      "loss": 0.5651,
+      "step": 47160
+    },
+    {
+      "epoch": 174.7,
+      "grad_norm": 1.0671674013137817,
+      "learning_rate": 0.00025296296296296297,
+      "loss": 0.5458,
+      "step": 47170
+    },
+    {
+      "epoch": 174.74,
+      "grad_norm": 1.1603813171386719,
+      "learning_rate": 0.0002525925925925926,
+      "loss": 0.5408,
+      "step": 47180
+    },
+    {
+      "epoch": 174.78,
+      "grad_norm": 1.2179709672927856,
+      "learning_rate": 0.00025222222222222226,
+      "loss": 0.5569,
+      "step": 47190
+    },
+    {
+      "epoch": 174.81,
+      "grad_norm": 1.1304550170898438,
+      "learning_rate": 0.00025185185185185185,
+      "loss": 0.5602,
+      "step": 47200
+    },
+    {
+      "epoch": 174.85,
+      "grad_norm": 1.1317181587219238,
+      "learning_rate": 0.0002514814814814815,
+      "loss": 0.5668,
+      "step": 47210
+    },
+    {
+      "epoch": 174.89,
+      "grad_norm": 1.1473468542099,
+      "learning_rate": 0.00025111111111111113,
+      "loss": 0.5592,
+      "step": 47220
+    },
+    {
+      "epoch": 174.93,
+      "grad_norm": 1.1926883459091187,
+      "learning_rate": 0.0002507407407407408,
+      "loss": 0.563,
+      "step": 47230
+    },
+    {
+      "epoch": 174.96,
+      "grad_norm": 1.1721458435058594,
+      "learning_rate": 0.00025037037037037036,
+      "loss": 0.5819,
+      "step": 47240
+    },
+    {
+      "epoch": 175.0,
+      "grad_norm": 2.136143207550049,
+      "learning_rate": 0.00025,
+      "loss": 0.5768,
+      "step": 47250
+    },
+    {
+      "epoch": 175.04,
+      "grad_norm": 0.958111584186554,
+      "learning_rate": 0.00024962962962962965,
+      "loss": 0.4661,
+      "step": 47260
+    },
+    {
+      "epoch": 175.07,
+      "grad_norm": 0.9953908324241638,
+      "learning_rate": 0.0002492592592592593,
+      "loss": 0.4594,
+      "step": 47270
+    },
+    {
+      "epoch": 175.11,
+      "grad_norm": 0.9846945405006409,
+      "learning_rate": 0.0002488888888888889,
+      "loss": 0.4578,
+      "step": 47280
+    },
+    {
+      "epoch": 175.15,
+      "grad_norm": 1.0534905195236206,
+      "learning_rate": 0.0002485185185185185,
+      "loss": 0.4543,
+      "step": 47290
+    },
+    {
+      "epoch": 175.19,
+      "grad_norm": 1.0469465255737305,
+      "learning_rate": 0.00024814814814814816,
+      "loss": 0.4752,
+      "step": 47300
+    },
+    {
+      "epoch": 175.22,
+      "grad_norm": 1.0580891370773315,
+      "learning_rate": 0.0002477777777777778,
+      "loss": 0.483,
+      "step": 47310
+    },
+    {
+      "epoch": 175.26,
+      "grad_norm": 1.050082802772522,
+      "learning_rate": 0.0002474074074074074,
+      "loss": 0.485,
+      "step": 47320
+    },
+    {
+      "epoch": 175.3,
+      "grad_norm": 1.0860666036605835,
+      "learning_rate": 0.00024703703703703704,
+      "loss": 0.5029,
+      "step": 47330
+    },
+    {
+      "epoch": 175.33,
+      "grad_norm": 1.01170015335083,
+      "learning_rate": 0.0002466666666666667,
+      "loss": 0.5031,
+      "step": 47340
+    },
+    {
+      "epoch": 175.37,
+      "grad_norm": 1.034176230430603,
+      "learning_rate": 0.00024629629629629627,
+      "loss": 0.5019,
+      "step": 47350
+    },
+    {
+      "epoch": 175.41,
+      "grad_norm": 1.0043485164642334,
+      "learning_rate": 0.00024592592592592597,
+      "loss": 0.5153,
+      "step": 47360
+    },
+    {
+      "epoch": 175.44,
+      "grad_norm": 1.0900912284851074,
+      "learning_rate": 0.00024555555555555556,
+      "loss": 0.5226,
+      "step": 47370
+    },
+    {
+      "epoch": 175.48,
+      "grad_norm": 1.1493656635284424,
+      "learning_rate": 0.0002451851851851852,
+      "loss": 0.5326,
+      "step": 47380
+    },
+    {
+      "epoch": 175.52,
+      "grad_norm": 1.1280782222747803,
+      "learning_rate": 0.00024481481481481484,
+      "loss": 0.52,
+      "step": 47390
+    },
+    {
+      "epoch": 175.56,
+      "grad_norm": 1.1246672868728638,
+      "learning_rate": 0.00024444444444444443,
+      "loss": 0.5281,
+      "step": 47400
+    },
+    {
+      "epoch": 175.59,
+      "grad_norm": 1.1507177352905273,
+      "learning_rate": 0.0002440740740740741,
+      "loss": 0.5317,
+      "step": 47410
+    },
+    {
+      "epoch": 175.63,
+      "grad_norm": 1.1389291286468506,
+      "learning_rate": 0.00024370370370370371,
+      "loss": 0.5342,
+      "step": 47420
+    },
+    {
+      "epoch": 175.67,
+      "grad_norm": 1.155777931213379,
+      "learning_rate": 0.00024333333333333336,
+      "loss": 0.5374,
+      "step": 47430
+    },
+    {
+      "epoch": 175.7,
+      "grad_norm": 1.1306384801864624,
+      "learning_rate": 0.00024296296296296297,
+      "loss": 0.5269,
+      "step": 47440
+    },
+    {
+      "epoch": 175.74,
+      "grad_norm": 1.1292226314544678,
+      "learning_rate": 0.0002425925925925926,
+      "loss": 0.5608,
+      "step": 47450
+    },
+    {
+      "epoch": 175.78,
+      "grad_norm": 1.0860081911087036,
+      "learning_rate": 0.00024222222222222223,
+      "loss": 0.5423,
+      "step": 47460
+    },
+    {
+      "epoch": 175.81,
+      "grad_norm": 1.207112431526184,
+      "learning_rate": 0.00024185185185185185,
+      "loss": 0.544,
+      "step": 47470
+    },
+    {
+      "epoch": 175.85,
+      "grad_norm": 1.0947352647781372,
+      "learning_rate": 0.0002414814814814815,
+      "loss": 0.563,
+      "step": 47480
+    },
+    {
+      "epoch": 175.89,
+      "grad_norm": 1.2116717100143433,
+      "learning_rate": 0.0002411111111111111,
+      "loss": 0.562,
+      "step": 47490
+    },
+    {
+      "epoch": 175.93,
+      "grad_norm": 1.1679658889770508,
+      "learning_rate": 0.00024074074074074072,
+      "loss": 0.5582,
+      "step": 47500
+    },
+    {
+      "epoch": 175.96,
+      "grad_norm": 1.1572608947753906,
+      "learning_rate": 0.0002403703703703704,
+      "loss": 0.5604,
+      "step": 47510
+    },
+    {
+      "epoch": 176.0,
+      "grad_norm": 1.975321888923645,
+      "learning_rate": 0.00024,
+      "loss": 0.5608,
+      "step": 47520
+    },
+    {
+      "epoch": 176.04,
+      "grad_norm": 0.9298834800720215,
+      "learning_rate": 0.00023962962962962965,
+      "loss": 0.4581,
+      "step": 47530
+    },
+    {
+      "epoch": 176.07,
+      "grad_norm": 1.068063497543335,
+      "learning_rate": 0.00023925925925925926,
+      "loss": 0.4384,
+      "step": 47540
+    },
+    {
+      "epoch": 176.11,
+      "grad_norm": 1.0253227949142456,
+      "learning_rate": 0.0002388888888888889,
+      "loss": 0.4508,
+      "step": 47550
+    },
+    {
+      "epoch": 176.15,
+      "grad_norm": 1.007856011390686,
+      "learning_rate": 0.00023851851851851852,
+      "loss": 0.4536,
+      "step": 47560
+    },
+    {
+      "epoch": 176.19,
+      "grad_norm": 1.0197805166244507,
+      "learning_rate": 0.00023814814814814814,
+      "loss": 0.4724,
+      "step": 47570
+    },
+    {
+      "epoch": 176.22,
+      "grad_norm": 1.1127703189849854,
+      "learning_rate": 0.00023777777777777778,
+      "loss": 0.4804,
+      "step": 47580
+    },
+    {
+      "epoch": 176.26,
+      "grad_norm": 1.1433093547821045,
+      "learning_rate": 0.0002374074074074074,
+      "loss": 0.4765,
+      "step": 47590
+    },
+    {
+      "epoch": 176.3,
+      "grad_norm": 1.0717138051986694,
+      "learning_rate": 0.00023703703703703704,
+      "loss": 0.5006,
+      "step": 47600
+    },
+    {
+      "epoch": 176.33,
+      "grad_norm": 1.1593323945999146,
+      "learning_rate": 0.00023666666666666668,
+      "loss": 0.4878,
+      "step": 47610
+    },
+    {
+      "epoch": 176.37,
+      "grad_norm": 1.0496388673782349,
+      "learning_rate": 0.00023629629629629632,
+      "loss": 0.5008,
+      "step": 47620
+    },
+    {
+      "epoch": 176.41,
+      "grad_norm": 1.0645349025726318,
+      "learning_rate": 0.00023592592592592594,
+      "loss": 0.5019,
+      "step": 47630
+    },
+    {
+      "epoch": 176.44,
+      "grad_norm": 1.1091699600219727,
+      "learning_rate": 0.00023555555555555556,
+      "loss": 0.5004,
+      "step": 47640
+    },
+    {
+      "epoch": 176.48,
+      "grad_norm": 1.156760573387146,
+      "learning_rate": 0.0002351851851851852,
+      "loss": 0.523,
+      "step": 47650
+    },
+    {
+      "epoch": 176.52,
+      "grad_norm": 1.110307216644287,
+      "learning_rate": 0.00023481481481481481,
+      "loss": 0.5202,
+      "step": 47660
+    },
+    {
+      "epoch": 176.56,
+      "grad_norm": 1.1058788299560547,
+      "learning_rate": 0.00023444444444444446,
+      "loss": 0.5114,
+      "step": 47670
+    },
+    {
+      "epoch": 176.59,
+      "grad_norm": 1.1229724884033203,
+      "learning_rate": 0.00023407407407407407,
+      "loss": 0.533,
+      "step": 47680
+    },
+    {
+      "epoch": 176.63,
+      "grad_norm": 1.1795262098312378,
+      "learning_rate": 0.0002337037037037037,
+      "loss": 0.5311,
+      "step": 47690
+    },
+    {
+      "epoch": 176.67,
+      "grad_norm": 1.178025722503662,
+      "learning_rate": 0.00023333333333333333,
+      "loss": 0.5181,
+      "step": 47700
+    },
+    {
+      "epoch": 176.7,
+      "grad_norm": 1.1427292823791504,
+      "learning_rate": 0.00023296296296296295,
+      "loss": 0.5399,
+      "step": 47710
+    },
+    {
+      "epoch": 176.74,
+      "grad_norm": 1.0837512016296387,
+      "learning_rate": 0.00023259259259259262,
+      "loss": 0.5336,
+      "step": 47720
+    },
+    {
+      "epoch": 176.78,
+      "grad_norm": 1.1414895057678223,
+      "learning_rate": 0.00023222222222222223,
+      "loss": 0.5309,
+      "step": 47730
+    },
+    {
+      "epoch": 176.81,
+      "grad_norm": 1.1707152128219604,
+      "learning_rate": 0.00023185185185185187,
+      "loss": 0.5419,
+      "step": 47740
+    },
+    {
+      "epoch": 176.85,
+      "grad_norm": 1.2246688604354858,
+      "learning_rate": 0.0002314814814814815,
+      "loss": 0.5507,
+      "step": 47750
+    },
+    {
+      "epoch": 176.89,
+      "grad_norm": 1.0815807580947876,
+      "learning_rate": 0.0002311111111111111,
+      "loss": 0.5502,
+      "step": 47760
+    },
+    {
+      "epoch": 176.93,
+      "grad_norm": 1.1408597230911255,
+      "learning_rate": 0.00023074074074074075,
+      "loss": 0.5663,
+      "step": 47770
+    },
+    {
+      "epoch": 176.96,
+      "grad_norm": 1.1771601438522339,
+      "learning_rate": 0.00023037037037037036,
+      "loss": 0.5381,
+      "step": 47780
+    },
+    {
+      "epoch": 177.0,
+      "grad_norm": 2.0750699043273926,
+      "learning_rate": 0.00023,
+      "loss": 0.5629,
+      "step": 47790
+    },
+    {
+      "epoch": 177.04,
+      "grad_norm": 0.947895884513855,
+      "learning_rate": 0.00022962962962962962,
+      "loss": 0.4314,
+      "step": 47800
+    },
+    {
+      "epoch": 177.07,
+      "grad_norm": 0.9859268069267273,
+      "learning_rate": 0.00022925925925925924,
+      "loss": 0.4374,
+      "step": 47810
+    },
+    {
+      "epoch": 177.11,
+      "grad_norm": 0.9737774133682251,
+      "learning_rate": 0.0002288888888888889,
+      "loss": 0.4595,
+      "step": 47820
+    },
+    {
+      "epoch": 177.15,
+      "grad_norm": 0.9961501955986023,
+      "learning_rate": 0.00022851851851851852,
+      "loss": 0.451,
+      "step": 47830
+    },
+    {
+      "epoch": 177.19,
+      "grad_norm": 0.9911838173866272,
+      "learning_rate": 0.00022814814814814817,
+      "loss": 0.4618,
+      "step": 47840
+    },
+    {
+      "epoch": 177.22,
+      "grad_norm": 1.0495216846466064,
+      "learning_rate": 0.00022777777777777778,
+      "loss": 0.4797,
+      "step": 47850
+    },
+    {
+      "epoch": 177.26,
+      "grad_norm": 1.0654104948043823,
+      "learning_rate": 0.00022740740740740742,
+      "loss": 0.4764,
+      "step": 47860
+    },
+    {
+      "epoch": 177.3,
+      "grad_norm": 1.047002911567688,
+      "learning_rate": 0.00022703703703703704,
+      "loss": 0.4677,
+      "step": 47870
+    },
+    {
+      "epoch": 177.33,
+      "grad_norm": 1.1380468606948853,
+      "learning_rate": 0.00022666666666666666,
+      "loss": 0.4924,
+      "step": 47880
+    },
+    {
+      "epoch": 177.37,
+      "grad_norm": 1.1353617906570435,
+      "learning_rate": 0.0002262962962962963,
+      "loss": 0.4937,
+      "step": 47890
+    },
+    {
+      "epoch": 177.41,
+      "grad_norm": 1.0635641813278198,
+      "learning_rate": 0.00022592592592592591,
+      "loss": 0.4951,
+      "step": 47900
+    },
+    {
+      "epoch": 177.44,
+      "grad_norm": 1.0273109674453735,
+      "learning_rate": 0.00022555555555555556,
+      "loss": 0.5008,
+      "step": 47910
+    },
+    {
+      "epoch": 177.48,
+      "grad_norm": 1.1229729652404785,
+      "learning_rate": 0.0002251851851851852,
+      "loss": 0.5104,
+      "step": 47920
+    },
+    {
+      "epoch": 177.52,
+      "grad_norm": 1.1669470071792603,
+      "learning_rate": 0.00022481481481481484,
+      "loss": 0.4993,
+      "step": 47930
+    },
+    {
+      "epoch": 177.56,
+      "grad_norm": 1.099764108657837,
+      "learning_rate": 0.00022444444444444446,
+      "loss": 0.5089,
+      "step": 47940
+    },
+    {
+      "epoch": 177.59,
+      "grad_norm": 1.1565130949020386,
+      "learning_rate": 0.00022407407407407407,
+      "loss": 0.5289,
+      "step": 47950
+    },
+    {
+      "epoch": 177.63,
+      "grad_norm": 1.0863348245620728,
+      "learning_rate": 0.00022370370370370372,
+      "loss": 0.5218,
+      "step": 47960
+    },
+    {
+      "epoch": 177.67,
+      "grad_norm": 1.1537193059921265,
+      "learning_rate": 0.00022333333333333333,
+      "loss": 0.5335,
+      "step": 47970
+    },
+    {
+      "epoch": 177.7,
+      "grad_norm": 1.1691408157348633,
+      "learning_rate": 0.00022296296296296297,
+      "loss": 0.5228,
+      "step": 47980
+    },
+    {
+      "epoch": 177.74,
+      "grad_norm": 1.1775586605072021,
+      "learning_rate": 0.0002225925925925926,
+      "loss": 0.5109,
+      "step": 47990
+    },
+    {
+      "epoch": 177.78,
+      "grad_norm": 1.1857826709747314,
+      "learning_rate": 0.0002222222222222222,
+      "loss": 0.5208,
+      "step": 48000
+    },
+    {
+      "epoch": 177.81,
+      "grad_norm": 1.2024893760681152,
+      "learning_rate": 0.00022185185185185185,
+      "loss": 0.5377,
+      "step": 48010
+    },
+    {
+      "epoch": 177.85,
+      "grad_norm": 1.0835148096084595,
+      "learning_rate": 0.0002214814814814815,
+      "loss": 0.5361,
+      "step": 48020
+    },
+    {
+      "epoch": 177.89,
+      "grad_norm": 1.0845205783843994,
+      "learning_rate": 0.00022111111111111113,
+      "loss": 0.5415,
+      "step": 48030
+    },
+    {
+      "epoch": 177.93,
+      "grad_norm": 1.2087277173995972,
+      "learning_rate": 0.00022074074074074075,
+      "loss": 0.5304,
+      "step": 48040
+    },
+    {
+      "epoch": 177.96,
+      "grad_norm": 1.1453560590744019,
+      "learning_rate": 0.0002203703703703704,
+      "loss": 0.5516,
+      "step": 48050
+    },
+    {
+      "epoch": 178.0,
+      "grad_norm": 2.6465954780578613,
+      "learning_rate": 0.00022,
+      "loss": 0.5592,
+      "step": 48060
+    },
+    {
+      "epoch": 178.04,
+      "grad_norm": 1.0190789699554443,
+      "learning_rate": 0.00021962962962962962,
+      "loss": 0.439,
+      "step": 48070
+    },
+    {
+      "epoch": 178.07,
+      "grad_norm": 0.9200291633605957,
+      "learning_rate": 0.00021925925925925927,
+      "loss": 0.4346,
+      "step": 48080
+    },
+    {
+      "epoch": 178.11,
+      "grad_norm": 0.9328666925430298,
+      "learning_rate": 0.00021888888888888888,
+      "loss": 0.4468,
+      "step": 48090
+    },
+    {
+      "epoch": 178.15,
+      "grad_norm": 1.0163941383361816,
+      "learning_rate": 0.00021851851851851852,
+      "loss": 0.463,
+      "step": 48100
+    },
+    {
+      "epoch": 178.19,
+      "grad_norm": 0.9540250301361084,
+      "learning_rate": 0.00021814814814814814,
+      "loss": 0.4362,
+      "step": 48110
+    },
+    {
+      "epoch": 178.22,
+      "grad_norm": 1.0213724374771118,
+      "learning_rate": 0.00021777777777777776,
+      "loss": 0.4546,
+      "step": 48120
+    },
+    {
+      "epoch": 178.26,
+      "grad_norm": 0.9999773502349854,
+      "learning_rate": 0.00021740740740740743,
+      "loss": 0.4761,
+      "step": 48130
+    },
+    {
+      "epoch": 178.3,
+      "grad_norm": 0.992883026599884,
+      "learning_rate": 0.00021703703703703704,
+      "loss": 0.4708,
+      "step": 48140
+    },
+    {
+      "epoch": 178.33,
+      "grad_norm": 1.0905901193618774,
+      "learning_rate": 0.00021666666666666668,
+      "loss": 0.476,
+      "step": 48150
+    },
+    {
+      "epoch": 178.37,
+      "grad_norm": 1.0493611097335815,
+      "learning_rate": 0.0002162962962962963,
+      "loss": 0.4743,
+      "step": 48160
+    },
+    {
+      "epoch": 178.41,
+      "grad_norm": 1.0430622100830078,
+      "learning_rate": 0.00021592592592592594,
+      "loss": 0.4763,
+      "step": 48170
+    },
+    {
+      "epoch": 178.44,
+      "grad_norm": 1.0460885763168335,
+      "learning_rate": 0.00021555555555555556,
+      "loss": 0.48,
+      "step": 48180
+    },
+    {
+      "epoch": 178.48,
+      "grad_norm": 1.0845812559127808,
+      "learning_rate": 0.00021518518518518517,
+      "loss": 0.4926,
+      "step": 48190
+    },
+    {
+      "epoch": 178.52,
+      "grad_norm": 1.0286200046539307,
+      "learning_rate": 0.00021481481481481482,
+      "loss": 0.4815,
+      "step": 48200
+    },
+    {
+      "epoch": 178.56,
+      "grad_norm": 1.0686702728271484,
+      "learning_rate": 0.00021444444444444443,
+      "loss": 0.5101,
+      "step": 48210
+    },
+    {
+      "epoch": 178.59,
+      "grad_norm": 1.112003207206726,
+      "learning_rate": 0.00021407407407407407,
+      "loss": 0.5155,
+      "step": 48220
+    },
+    {
+      "epoch": 178.63,
+      "grad_norm": 1.1415358781814575,
+      "learning_rate": 0.00021370370370370372,
+      "loss": 0.5,
+      "step": 48230
+    },
+    {
+      "epoch": 178.67,
+      "grad_norm": 1.135293960571289,
+      "learning_rate": 0.00021333333333333336,
+      "loss": 0.5151,
+      "step": 48240
+    },
+    {
+      "epoch": 178.7,
+      "grad_norm": 1.0956019163131714,
+      "learning_rate": 0.00021296296296296298,
+      "loss": 0.5123,
+      "step": 48250
+    },
+    {
+      "epoch": 178.74,
+      "grad_norm": 1.150425910949707,
+      "learning_rate": 0.0002125925925925926,
+      "loss": 0.5311,
+      "step": 48260
+    },
+    {
+      "epoch": 178.78,
+      "grad_norm": 1.1303415298461914,
+      "learning_rate": 0.00021222222222222223,
+      "loss": 0.5246,
+      "step": 48270
+    },
+    {
+      "epoch": 178.81,
+      "grad_norm": 1.262482762336731,
+      "learning_rate": 0.00021185185185185185,
+      "loss": 0.5445,
+      "step": 48280
+    },
+    {
+      "epoch": 178.85,
+      "grad_norm": 1.236796259880066,
+      "learning_rate": 0.0002114814814814815,
+      "loss": 0.5318,
+      "step": 48290
+    },
+    {
+      "epoch": 178.89,
+      "grad_norm": 1.13998281955719,
+      "learning_rate": 0.0002111111111111111,
+      "loss": 0.5236,
+      "step": 48300
+    },
+    {
+      "epoch": 178.93,
+      "grad_norm": 1.1601678133010864,
+      "learning_rate": 0.00021074074074074072,
+      "loss": 0.5397,
+      "step": 48310
+    },
+    {
+      "epoch": 178.96,
+      "grad_norm": 1.1455624103546143,
+      "learning_rate": 0.00021037037037037037,
+      "loss": 0.5432,
+      "step": 48320
+    },
+    {
+      "epoch": 179.0,
+      "grad_norm": 2.160454034805298,
+      "learning_rate": 0.00021,
+      "loss": 0.5366,
+      "step": 48330
+    },
+    {
+      "epoch": 179.04,
+      "grad_norm": 1.0072879791259766,
+      "learning_rate": 0.00020962962962962965,
+      "loss": 0.4454,
+      "step": 48340
+    },
+    {
+      "epoch": 179.07,
+      "grad_norm": 0.9614301919937134,
+      "learning_rate": 0.00020925925925925927,
+      "loss": 0.4311,
+      "step": 48350
+    },
+    {
+      "epoch": 179.11,
+      "grad_norm": 0.9831352233886719,
+      "learning_rate": 0.0002088888888888889,
+      "loss": 0.4359,
+      "step": 48360
+    },
+    {
+      "epoch": 179.15,
+      "grad_norm": 0.9975229501724243,
+      "learning_rate": 0.00020851851851851853,
+      "loss": 0.4511,
+      "step": 48370
+    },
+    {
+      "epoch": 179.19,
+      "grad_norm": 1.075417399406433,
+      "learning_rate": 0.00020814814814814814,
+      "loss": 0.4466,
+      "step": 48380
+    },
+    {
+      "epoch": 179.22,
+      "grad_norm": 0.9628404378890991,
+      "learning_rate": 0.00020777777777777778,
+      "loss": 0.4474,
+      "step": 48390
+    },
+    {
+      "epoch": 179.26,
+      "grad_norm": 1.052886724472046,
+      "learning_rate": 0.0002074074074074074,
+      "loss": 0.446,
+      "step": 48400
+    },
+    {
+      "epoch": 179.3,
+      "grad_norm": 1.0707443952560425,
+      "learning_rate": 0.00020703703703703704,
+      "loss": 0.4662,
+      "step": 48410
+    },
+    {
+      "epoch": 179.33,
+      "grad_norm": 1.0438921451568604,
+      "learning_rate": 0.00020666666666666666,
+      "loss": 0.4695,
+      "step": 48420
+    },
+    {
+      "epoch": 179.37,
+      "grad_norm": 1.076917052268982,
+      "learning_rate": 0.0002062962962962963,
+      "loss": 0.4781,
+      "step": 48430
+    },
+    {
+      "epoch": 179.41,
+      "grad_norm": 1.067229986190796,
+      "learning_rate": 0.00020592592592592594,
+      "loss": 0.4713,
+      "step": 48440
+    },
+    {
+      "epoch": 179.44,
+      "grad_norm": 1.0892492532730103,
+      "learning_rate": 0.00020555555555555556,
+      "loss": 0.4814,
+      "step": 48450
+    },
+    {
+      "epoch": 179.48,
+      "grad_norm": 1.1713968515396118,
+      "learning_rate": 0.0002051851851851852,
+      "loss": 0.4865,
+      "step": 48460
+    },
+    {
+      "epoch": 179.52,
+      "grad_norm": 1.089644193649292,
+      "learning_rate": 0.00020481481481481482,
+      "loss": 0.4923,
+      "step": 48470
+    },
+    {
+      "epoch": 179.56,
+      "grad_norm": 1.1268770694732666,
+      "learning_rate": 0.00020444444444444446,
+      "loss": 0.52,
+      "step": 48480
+    },
+    {
+      "epoch": 179.59,
+      "grad_norm": 1.0705839395523071,
+      "learning_rate": 0.00020407407407407408,
+      "loss": 0.5031,
+      "step": 48490
+    },
+    {
+      "epoch": 179.63,
+      "grad_norm": 1.0443114042282104,
+      "learning_rate": 0.0002037037037037037,
+      "loss": 0.4888,
+      "step": 48500
+    },
+    {
+      "epoch": 179.67,
+      "grad_norm": 1.0626062154769897,
+      "learning_rate": 0.00020333333333333333,
+      "loss": 0.509,
+      "step": 48510
+    },
+    {
+      "epoch": 179.7,
+      "grad_norm": 1.1279469728469849,
+      "learning_rate": 0.00020296296296296295,
+      "loss": 0.5049,
+      "step": 48520
+    },
+    {
+      "epoch": 179.74,
+      "grad_norm": 1.089354157447815,
+      "learning_rate": 0.00020259259259259262,
+      "loss": 0.4979,
+      "step": 48530
+    },
+    {
+      "epoch": 179.78,
+      "grad_norm": 1.1244808435440063,
+      "learning_rate": 0.00020222222222222223,
+      "loss": 0.5134,
+      "step": 48540
+    },
+    {
+      "epoch": 179.81,
+      "grad_norm": 1.0787869691848755,
+      "learning_rate": 0.00020185185185185188,
+      "loss": 0.5184,
+      "step": 48550
+    },
+    {
+      "epoch": 179.85,
+      "grad_norm": 1.1457653045654297,
+      "learning_rate": 0.0002014814814814815,
+      "loss": 0.5201,
+      "step": 48560
+    },
+    {
+      "epoch": 179.89,
+      "grad_norm": 1.1280841827392578,
+      "learning_rate": 0.0002011111111111111,
+      "loss": 0.5205,
+      "step": 48570
+    },
+    {
+      "epoch": 179.93,
+      "grad_norm": 1.1580936908721924,
+      "learning_rate": 0.00020074074074074075,
+      "loss": 0.5329,
+      "step": 48580
+    },
+    {
+      "epoch": 179.96,
+      "grad_norm": 1.0993684530258179,
+      "learning_rate": 0.00020037037037037037,
+      "loss": 0.5289,
+      "step": 48590
+    },
+    {
+      "epoch": 180.0,
+      "grad_norm": 2.2351152896881104,
+      "learning_rate": 0.0002,
+      "loss": 0.5151,
+      "step": 48600
+    },
+    {
+      "epoch": 180.04,
+      "grad_norm": 0.9843377470970154,
+      "learning_rate": 0.00019962962962962963,
+      "loss": 0.4304,
+      "step": 48610
+    },
+    {
+      "epoch": 180.07,
+      "grad_norm": 0.8959612250328064,
+      "learning_rate": 0.00019925925925925924,
+      "loss": 0.4105,
+      "step": 48620
+    },
+    {
+      "epoch": 180.11,
+      "grad_norm": 0.9292177557945251,
+      "learning_rate": 0.00019888888888888888,
+      "loss": 0.4352,
+      "step": 48630
+    },
+    {
+      "epoch": 180.15,
+      "grad_norm": 0.9821608662605286,
+      "learning_rate": 0.00019851851851851853,
+      "loss": 0.4408,
+      "step": 48640
+    },
+    {
+      "epoch": 180.19,
+      "grad_norm": 1.0155918598175049,
+      "learning_rate": 0.00019814814814814817,
+      "loss": 0.4278,
+      "step": 48650
+    },
+    {
+      "epoch": 180.22,
+      "grad_norm": 1.047547698020935,
+      "learning_rate": 0.00019777777777777778,
+      "loss": 0.4486,
+      "step": 48660
+    },
+    {
+      "epoch": 180.26,
+      "grad_norm": 1.0473016500473022,
+      "learning_rate": 0.00019740740740740743,
+      "loss": 0.4448,
+      "step": 48670
+    },
+    {
+      "epoch": 180.3,
+      "grad_norm": 1.0475410223007202,
+      "learning_rate": 0.00019703703703703704,
+      "loss": 0.4602,
+      "step": 48680
+    },
+    {
+      "epoch": 180.33,
+      "grad_norm": 1.0084902048110962,
+      "learning_rate": 0.00019666666666666666,
+      "loss": 0.4469,
+      "step": 48690
+    },
+    {
+      "epoch": 180.37,
+      "grad_norm": 1.1199017763137817,
+      "learning_rate": 0.0001962962962962963,
+      "loss": 0.4632,
+      "step": 48700
+    },
+    {
+      "epoch": 180.41,
+      "grad_norm": 1.0491690635681152,
+      "learning_rate": 0.00019592592592592592,
+      "loss": 0.4626,
+      "step": 48710
+    },
+    {
+      "epoch": 180.44,
+      "grad_norm": 1.0795011520385742,
+      "learning_rate": 0.00019555555555555556,
+      "loss": 0.4826,
+      "step": 48720
+    },
+    {
+      "epoch": 180.48,
+      "grad_norm": 1.0450068712234497,
+      "learning_rate": 0.00019518518518518518,
+      "loss": 0.4836,
+      "step": 48730
+    },
+    {
+      "epoch": 180.52,
+      "grad_norm": 1.0705045461654663,
+      "learning_rate": 0.00019481481481481482,
+      "loss": 0.479,
+      "step": 48740
+    },
+    {
+      "epoch": 180.56,
+      "grad_norm": 1.0864088535308838,
+      "learning_rate": 0.00019444444444444446,
+      "loss": 0.4816,
+      "step": 48750
+    },
+    {
+      "epoch": 180.59,
+      "grad_norm": 1.0538465976715088,
+      "learning_rate": 0.00019407407407407408,
+      "loss": 0.4927,
+      "step": 48760
+    },
+    {
+      "epoch": 180.63,
+      "grad_norm": 1.0714560747146606,
+      "learning_rate": 0.00019370370370370372,
+      "loss": 0.5075,
+      "step": 48770
+    },
+    {
+      "epoch": 180.67,
+      "grad_norm": 1.0492222309112549,
+      "learning_rate": 0.00019333333333333333,
+      "loss": 0.4991,
+      "step": 48780
+    },
+    {
+      "epoch": 180.7,
+      "grad_norm": 1.095389485359192,
+      "learning_rate": 0.00019296296296296298,
+      "loss": 0.5102,
+      "step": 48790
+    },
+    {
+      "epoch": 180.74,
+      "grad_norm": 1.0268661975860596,
+      "learning_rate": 0.0001925925925925926,
+      "loss": 0.5101,
+      "step": 48800
+    },
+    {
+      "epoch": 180.78,
+      "grad_norm": 1.0888630151748657,
+      "learning_rate": 0.0001922222222222222,
+      "loss": 0.5086,
+      "step": 48810
+    },
+    {
+      "epoch": 180.81,
+      "grad_norm": 1.1224223375320435,
+      "learning_rate": 0.00019185185185185185,
+      "loss": 0.506,
+      "step": 48820
+    },
+    {
+      "epoch": 180.85,
+      "grad_norm": 1.0878612995147705,
+      "learning_rate": 0.00019148148148148147,
+      "loss": 0.5174,
+      "step": 48830
+    },
+    {
+      "epoch": 180.89,
+      "grad_norm": 1.1202062368392944,
+      "learning_rate": 0.00019111111111111114,
+      "loss": 0.5192,
+      "step": 48840
+    },
+    {
+      "epoch": 180.93,
+      "grad_norm": 1.115593671798706,
+      "learning_rate": 0.00019074074074074075,
+      "loss": 0.5133,
+      "step": 48850
+    },
+    {
+      "epoch": 180.96,
+      "grad_norm": 1.1733633279800415,
+      "learning_rate": 0.0001903703703703704,
+      "loss": 0.5132,
+      "step": 48860
+    },
+    {
+      "epoch": 181.0,
+      "grad_norm": 2.1127378940582275,
+      "learning_rate": 0.00019,
+      "loss": 0.5097,
+      "step": 48870
+    },
+    {
+      "epoch": 181.04,
+      "grad_norm": 0.9501312375068665,
+      "learning_rate": 0.00018962962962962963,
+      "loss": 0.4238,
+      "step": 48880
+    },
+    {
+      "epoch": 181.07,
+      "grad_norm": 0.9944826364517212,
+      "learning_rate": 0.00018925925925925927,
+      "loss": 0.421,
+      "step": 48890
+    },
+    {
+      "epoch": 181.11,
+      "grad_norm": 0.958714485168457,
+      "learning_rate": 0.00018888888888888888,
+      "loss": 0.4229,
+      "step": 48900
+    },
+    {
+      "epoch": 181.15,
+      "grad_norm": 1.00678288936615,
+      "learning_rate": 0.00018851851851851853,
+      "loss": 0.4263,
+      "step": 48910
+    },
+    {
+      "epoch": 181.19,
+      "grad_norm": 0.9769282937049866,
+      "learning_rate": 0.00018814814814814814,
+      "loss": 0.443,
+      "step": 48920
+    },
+    {
+      "epoch": 181.22,
+      "grad_norm": 1.0257822275161743,
+      "learning_rate": 0.00018777777777777776,
+      "loss": 0.439,
+      "step": 48930
+    },
+    {
+      "epoch": 181.26,
+      "grad_norm": 0.9551071524620056,
+      "learning_rate": 0.00018740740740740743,
+      "loss": 0.4545,
+      "step": 48940
+    },
+    {
+      "epoch": 181.3,
+      "grad_norm": 0.9713193774223328,
+      "learning_rate": 0.00018703703703703704,
+      "loss": 0.4456,
+      "step": 48950
+    },
+    {
+      "epoch": 181.33,
+      "grad_norm": 1.0076040029525757,
+      "learning_rate": 0.0001866666666666667,
+      "loss": 0.4457,
+      "step": 48960
+    },
+    {
+      "epoch": 181.37,
+      "grad_norm": 1.1038206815719604,
+      "learning_rate": 0.0001862962962962963,
+      "loss": 0.4587,
+      "step": 48970
+    },
+    {
+      "epoch": 181.41,
+      "grad_norm": 1.092207908630371,
+      "learning_rate": 0.00018592592592592594,
+      "loss": 0.4683,
+      "step": 48980
+    },
+    {
+      "epoch": 181.44,
+      "grad_norm": 1.0330462455749512,
+      "learning_rate": 0.00018555555555555556,
+      "loss": 0.4657,
+      "step": 48990
+    },
+    {
+      "epoch": 181.48,
+      "grad_norm": 1.0785881280899048,
+      "learning_rate": 0.00018518518518518518,
+      "loss": 0.4782,
+      "step": 49000
+    },
+    {
+      "epoch": 181.52,
+      "grad_norm": 1.0754797458648682,
+      "learning_rate": 0.00018481481481481482,
+      "loss": 0.4602,
+      "step": 49010
+    },
+    {
+      "epoch": 181.56,
+      "grad_norm": 1.1690996885299683,
+      "learning_rate": 0.00018444444444444443,
+      "loss": 0.4749,
+      "step": 49020
+    },
+    {
+      "epoch": 181.59,
+      "grad_norm": 1.0903944969177246,
+      "learning_rate": 0.00018407407407407408,
+      "loss": 0.4791,
+      "step": 49030
+    },
+    {
+      "epoch": 181.63,
+      "grad_norm": 1.09013032913208,
+      "learning_rate": 0.00018370370370370372,
+      "loss": 0.4915,
+      "step": 49040
+    },
+    {
+      "epoch": 181.67,
+      "grad_norm": 1.0747276544570923,
+      "learning_rate": 0.00018333333333333334,
+      "loss": 0.491,
+      "step": 49050
+    },
+    {
+      "epoch": 181.7,
+      "grad_norm": 1.0240190029144287,
+      "learning_rate": 0.00018296296296296298,
+      "loss": 0.4837,
+      "step": 49060
+    },
+    {
+      "epoch": 181.74,
+      "grad_norm": 1.0852088928222656,
+      "learning_rate": 0.0001825925925925926,
+      "loss": 0.4984,
+      "step": 49070
+    },
+    {
+      "epoch": 181.78,
+      "grad_norm": 1.1203248500823975,
+      "learning_rate": 0.00018222222222222224,
+      "loss": 0.5084,
+      "step": 49080
+    },
+    {
+      "epoch": 181.81,
+      "grad_norm": 1.1268478631973267,
+      "learning_rate": 0.00018185185185185185,
+      "loss": 0.4992,
+      "step": 49090
+    },
+    {
+      "epoch": 181.85,
+      "grad_norm": 1.137208342552185,
+      "learning_rate": 0.0001814814814814815,
+      "loss": 0.4983,
+      "step": 49100
+    },
+    {
+      "epoch": 181.89,
+      "grad_norm": 1.1762620210647583,
+      "learning_rate": 0.0001811111111111111,
+      "loss": 0.5113,
+      "step": 49110
+    },
+    {
+      "epoch": 181.93,
+      "grad_norm": 1.0956618785858154,
+      "learning_rate": 0.00018074074074074073,
+      "loss": 0.4969,
+      "step": 49120
+    },
+    {
+      "epoch": 181.96,
+      "grad_norm": 1.088881015777588,
+      "learning_rate": 0.00018037037037037037,
+      "loss": 0.5037,
+      "step": 49130
+    },
+    {
+      "epoch": 182.0,
+      "grad_norm": 2.1335837841033936,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 0.5282,
+      "step": 49140
+    },
+    {
+      "epoch": 182.04,
+      "grad_norm": 0.9405384659767151,
+      "learning_rate": 0.00017962962962962965,
+      "loss": 0.4223,
+      "step": 49150
+    },
+    {
+      "epoch": 182.07,
+      "grad_norm": 0.9342048168182373,
+      "learning_rate": 0.00017925925925925927,
+      "loss": 0.4154,
+      "step": 49160
+    },
+    {
+      "epoch": 182.11,
+      "grad_norm": 0.8926421999931335,
+      "learning_rate": 0.00017888888888888889,
+      "loss": 0.419,
+      "step": 49170
+    },
+    {
+      "epoch": 182.15,
+      "grad_norm": 0.9453332424163818,
+      "learning_rate": 0.00017851851851851853,
+      "loss": 0.4238,
+      "step": 49180
+    },
+    {
+      "epoch": 182.19,
+      "grad_norm": 0.9516483545303345,
+      "learning_rate": 0.00017814814814814814,
+      "loss": 0.4312,
+      "step": 49190
+    },
+    {
+      "epoch": 182.22,
+      "grad_norm": 0.9720765948295593,
+      "learning_rate": 0.00017777777777777779,
+      "loss": 0.4319,
+      "step": 49200
+    },
+    {
+      "epoch": 182.26,
+      "grad_norm": 0.9715983867645264,
+      "learning_rate": 0.0001774074074074074,
+      "loss": 0.4378,
+      "step": 49210
+    },
+    {
+      "epoch": 182.3,
+      "grad_norm": 1.0771870613098145,
+      "learning_rate": 0.00017703703703703704,
+      "loss": 0.4448,
+      "step": 49220
+    },
+    {
+      "epoch": 182.33,
+      "grad_norm": 1.0263774394989014,
+      "learning_rate": 0.00017666666666666666,
+      "loss": 0.4491,
+      "step": 49230
+    },
+    {
+      "epoch": 182.37,
+      "grad_norm": 1.0612064599990845,
+      "learning_rate": 0.00017629629629629628,
+      "loss": 0.4352,
+      "step": 49240
+    },
+    {
+      "epoch": 182.41,
+      "grad_norm": 1.0565451383590698,
+      "learning_rate": 0.00017592592592592595,
+      "loss": 0.4549,
+      "step": 49250
+    },
+    {
+      "epoch": 182.44,
+      "grad_norm": 1.0483824014663696,
+      "learning_rate": 0.00017555555555555556,
+      "loss": 0.4589,
+      "step": 49260
+    },
+    {
+      "epoch": 182.48,
+      "grad_norm": 1.0516337156295776,
+      "learning_rate": 0.0001751851851851852,
+      "loss": 0.4741,
+      "step": 49270
+    },
+    {
+      "epoch": 182.52,
+      "grad_norm": 1.0335724353790283,
+      "learning_rate": 0.00017481481481481482,
+      "loss": 0.4833,
+      "step": 49280
+    },
+    {
+      "epoch": 182.56,
+      "grad_norm": 1.072335124015808,
+      "learning_rate": 0.00017444444444444446,
+      "loss": 0.4679,
+      "step": 49290
+    },
+    {
+      "epoch": 182.59,
+      "grad_norm": 1.1325234174728394,
+      "learning_rate": 0.00017407407407407408,
+      "loss": 0.4653,
+      "step": 49300
+    },
+    {
+      "epoch": 182.63,
+      "grad_norm": 1.0939109325408936,
+      "learning_rate": 0.0001737037037037037,
+      "loss": 0.4782,
+      "step": 49310
+    },
+    {
+      "epoch": 182.67,
+      "grad_norm": 1.039179801940918,
+      "learning_rate": 0.00017333333333333334,
+      "loss": 0.4767,
+      "step": 49320
+    },
+    {
+      "epoch": 182.7,
+      "grad_norm": 1.054984450340271,
+      "learning_rate": 0.00017296296296296295,
+      "loss": 0.4917,
+      "step": 49330
+    },
+    {
+      "epoch": 182.74,
+      "grad_norm": 1.0941307544708252,
+      "learning_rate": 0.0001725925925925926,
+      "loss": 0.4915,
+      "step": 49340
+    },
+    {
+      "epoch": 182.78,
+      "grad_norm": 1.044124960899353,
+      "learning_rate": 0.00017222222222222224,
+      "loss": 0.4686,
+      "step": 49350
+    },
+    {
+      "epoch": 182.81,
+      "grad_norm": 1.0401660203933716,
+      "learning_rate": 0.00017185185185185185,
+      "loss": 0.4846,
+      "step": 49360
+    },
+    {
+      "epoch": 182.85,
+      "grad_norm": 1.0751421451568604,
+      "learning_rate": 0.0001714814814814815,
+      "loss": 0.4927,
+      "step": 49370
+    },
+    {
+      "epoch": 182.89,
+      "grad_norm": 1.1070232391357422,
+      "learning_rate": 0.0001711111111111111,
+      "loss": 0.4959,
+      "step": 49380
+    },
+    {
+      "epoch": 182.93,
+      "grad_norm": 1.1466519832611084,
+      "learning_rate": 0.00017074074074074075,
+      "loss": 0.4969,
+      "step": 49390
+    },
+    {
+      "epoch": 182.96,
+      "grad_norm": 1.1511880159378052,
+      "learning_rate": 0.00017037037037037037,
+      "loss": 0.5007,
+      "step": 49400
+    },
+    {
+      "epoch": 183.0,
+      "grad_norm": 1.9645870923995972,
+      "learning_rate": 0.00017,
+      "loss": 0.509,
+      "step": 49410
+    },
+    {
+      "epoch": 183.04,
+      "grad_norm": 0.9212923049926758,
+      "learning_rate": 0.00016962962962962963,
+      "loss": 0.4048,
+      "step": 49420
+    },
+    {
+      "epoch": 183.07,
+      "grad_norm": 0.9352565407752991,
+      "learning_rate": 0.00016925925925925924,
+      "loss": 0.4027,
+      "step": 49430
+    },
+    {
+      "epoch": 183.11,
+      "grad_norm": 0.9354068040847778,
+      "learning_rate": 0.00016888888888888889,
+      "loss": 0.4185,
+      "step": 49440
+    },
+    {
+      "epoch": 183.15,
+      "grad_norm": 1.0052289962768555,
+      "learning_rate": 0.00016851851851851853,
+      "loss": 0.4161,
+      "step": 49450
+    },
+    {
+      "epoch": 183.19,
+      "grad_norm": 0.936885416507721,
+      "learning_rate": 0.00016814814814814817,
+      "loss": 0.4236,
+      "step": 49460
+    },
+    {
+      "epoch": 183.22,
+      "grad_norm": 1.0176805257797241,
+      "learning_rate": 0.0001677777777777778,
+      "loss": 0.426,
+      "step": 49470
+    },
+    {
+      "epoch": 183.26,
+      "grad_norm": 1.0213100910186768,
+      "learning_rate": 0.0001674074074074074,
+      "loss": 0.4317,
+      "step": 49480
+    },
+    {
+      "epoch": 183.3,
+      "grad_norm": 0.9696800708770752,
+      "learning_rate": 0.00016703703703703705,
+      "loss": 0.4173,
+      "step": 49490
+    },
+    {
+      "epoch": 183.33,
+      "grad_norm": 1.0048062801361084,
+      "learning_rate": 0.00016666666666666666,
+      "loss": 0.433,
+      "step": 49500
+    },
+    {
+      "epoch": 183.37,
+      "grad_norm": 1.0107778310775757,
+      "learning_rate": 0.0001662962962962963,
+      "loss": 0.4351,
+      "step": 49510
+    },
+    {
+      "epoch": 183.41,
+      "grad_norm": 0.9895990490913391,
+      "learning_rate": 0.00016592592592592592,
+      "loss": 0.4458,
+      "step": 49520
+    },
+    {
+      "epoch": 183.44,
+      "grad_norm": 0.9655704498291016,
+      "learning_rate": 0.00016555555555555556,
+      "loss": 0.4602,
+      "step": 49530
+    },
+    {
+      "epoch": 183.48,
+      "grad_norm": 1.0920255184173584,
+      "learning_rate": 0.00016518518518518518,
+      "loss": 0.4657,
+      "step": 49540
+    },
+    {
+      "epoch": 183.52,
+      "grad_norm": 1.0697599649429321,
+      "learning_rate": 0.0001648148148148148,
+      "loss": 0.4542,
+      "step": 49550
+    },
+    {
+      "epoch": 183.56,
+      "grad_norm": 1.09999680519104,
+      "learning_rate": 0.00016444444444444446,
+      "loss": 0.4724,
+      "step": 49560
+    },
+    {
+      "epoch": 183.59,
+      "grad_norm": 1.034386157989502,
+      "learning_rate": 0.00016407407407407408,
+      "loss": 0.4664,
+      "step": 49570
+    },
+    {
+      "epoch": 183.63,
+      "grad_norm": 1.041436791419983,
+      "learning_rate": 0.00016370370370370372,
+      "loss": 0.4695,
+      "step": 49580
+    },
+    {
+      "epoch": 183.67,
+      "grad_norm": 1.0949628353118896,
+      "learning_rate": 0.00016333333333333334,
+      "loss": 0.4737,
+      "step": 49590
+    },
+    {
+      "epoch": 183.7,
+      "grad_norm": 1.0208719968795776,
+      "learning_rate": 0.00016296296296296298,
+      "loss": 0.4803,
+      "step": 49600
+    },
+    {
+      "epoch": 183.74,
+      "grad_norm": 1.042176604270935,
+      "learning_rate": 0.0001625925925925926,
+      "loss": 0.4867,
+      "step": 49610
+    },
+    {
+      "epoch": 183.78,
+      "grad_norm": 1.0463050603866577,
+      "learning_rate": 0.0001622222222222222,
+      "loss": 0.4704,
+      "step": 49620
+    },
+    {
+      "epoch": 183.81,
+      "grad_norm": 1.1194572448730469,
+      "learning_rate": 0.00016185185185185185,
+      "loss": 0.4778,
+      "step": 49630
+    },
+    {
+      "epoch": 183.85,
+      "grad_norm": 1.06777024269104,
+      "learning_rate": 0.00016148148148148147,
+      "loss": 0.4745,
+      "step": 49640
+    },
+    {
+      "epoch": 183.89,
+      "grad_norm": 1.1410287618637085,
+      "learning_rate": 0.0001611111111111111,
+      "loss": 0.496,
+      "step": 49650
+    },
+    {
+      "epoch": 183.93,
+      "grad_norm": 1.0613508224487305,
+      "learning_rate": 0.00016074074074074075,
+      "loss": 0.4862,
+      "step": 49660
+    },
+    {
+      "epoch": 183.96,
+      "grad_norm": 1.103725552558899,
+      "learning_rate": 0.00016037037037037037,
+      "loss": 0.5019,
+      "step": 49670
+    },
+    {
+      "epoch": 184.0,
+      "grad_norm": 2.150892734527588,
+      "learning_rate": 0.00016,
+      "loss": 0.4927,
+      "step": 49680
+    },
+    {
+      "epoch": 184.04,
+      "grad_norm": 0.8993950486183167,
+      "learning_rate": 0.00015962962962962963,
+      "loss": 0.3915,
+      "step": 49690
+    },
+    {
+      "epoch": 184.07,
+      "grad_norm": 0.9276669025421143,
+      "learning_rate": 0.00015925925925925927,
+      "loss": 0.4048,
+      "step": 49700
+    },
+    {
+      "epoch": 184.11,
+      "grad_norm": 0.9294317960739136,
+      "learning_rate": 0.0001588888888888889,
+      "loss": 0.402,
+      "step": 49710
+    },
+    {
+      "epoch": 184.15,
+      "grad_norm": 1.004883050918579,
+      "learning_rate": 0.00015851851851851853,
+      "loss": 0.4077,
+      "step": 49720
+    },
+    {
+      "epoch": 184.19,
+      "grad_norm": 0.9653359055519104,
+      "learning_rate": 0.00015814814814814815,
+      "loss": 0.4204,
+      "step": 49730
+    },
+    {
+      "epoch": 184.22,
+      "grad_norm": 0.9280202984809875,
+      "learning_rate": 0.00015777777777777776,
+      "loss": 0.4257,
+      "step": 49740
+    },
+    {
+      "epoch": 184.26,
+      "grad_norm": 0.9918959736824036,
+      "learning_rate": 0.0001574074074074074,
+      "loss": 0.4261,
+      "step": 49750
+    },
+    {
+      "epoch": 184.3,
+      "grad_norm": 1.0165759325027466,
+      "learning_rate": 0.00015703703703703705,
+      "loss": 0.4336,
+      "step": 49760
+    },
+    {
+      "epoch": 184.33,
+      "grad_norm": 1.0142242908477783,
+      "learning_rate": 0.0001566666666666667,
+      "loss": 0.4342,
+      "step": 49770
+    },
+    {
+      "epoch": 184.37,
+      "grad_norm": 1.0278903245925903,
+      "learning_rate": 0.0001562962962962963,
+      "loss": 0.4458,
+      "step": 49780
+    },
+    {
+      "epoch": 184.41,
+      "grad_norm": 0.9913373589515686,
+      "learning_rate": 0.00015592592592592592,
+      "loss": 0.4443,
+      "step": 49790
+    },
+    {
+      "epoch": 184.44,
+      "grad_norm": 1.059424877166748,
+      "learning_rate": 0.00015555555555555556,
+      "loss": 0.4446,
+      "step": 49800
+    },
+    {
+      "epoch": 184.48,
+      "grad_norm": 1.0725332498550415,
+      "learning_rate": 0.00015518518518518518,
+      "loss": 0.4465,
+      "step": 49810
+    },
+    {
+      "epoch": 184.52,
+      "grad_norm": 1.0381534099578857,
+      "learning_rate": 0.00015481481481481482,
+      "loss": 0.4481,
+      "step": 49820
+    },
+    {
+      "epoch": 184.56,
+      "grad_norm": 0.9897781014442444,
+      "learning_rate": 0.00015444444444444444,
+      "loss": 0.4626,
+      "step": 49830
+    },
+    {
+      "epoch": 184.59,
+      "grad_norm": 1.1018785238265991,
+      "learning_rate": 0.00015407407407407408,
+      "loss": 0.4642,
+      "step": 49840
+    },
+    {
+      "epoch": 184.63,
+      "grad_norm": 1.0702872276306152,
+      "learning_rate": 0.0001537037037037037,
+      "loss": 0.4719,
+      "step": 49850
+    },
+    {
+      "epoch": 184.67,
+      "grad_norm": 1.0390729904174805,
+      "learning_rate": 0.00015333333333333334,
+      "loss": 0.4525,
+      "step": 49860
+    },
+    {
+      "epoch": 184.7,
+      "grad_norm": 1.0610655546188354,
+      "learning_rate": 0.00015296296296296298,
+      "loss": 0.4687,
+      "step": 49870
+    },
+    {
+      "epoch": 184.74,
+      "grad_norm": 1.009779453277588,
+      "learning_rate": 0.0001525925925925926,
+      "loss": 0.4672,
+      "step": 49880
+    },
+    {
+      "epoch": 184.78,
+      "grad_norm": 1.141605019569397,
+      "learning_rate": 0.00015222222222222224,
+      "loss": 0.4612,
+      "step": 49890
+    },
+    {
+      "epoch": 184.81,
+      "grad_norm": 1.0932852029800415,
+      "learning_rate": 0.00015185185185185185,
+      "loss": 0.4866,
+      "step": 49900
+    },
+    {
+      "epoch": 184.85,
+      "grad_norm": 1.0668141841888428,
+      "learning_rate": 0.0001514814814814815,
+      "loss": 0.4736,
+      "step": 49910
+    },
+    {
+      "epoch": 184.89,
+      "grad_norm": 1.0443545579910278,
+      "learning_rate": 0.0001511111111111111,
+      "loss": 0.4682,
+      "step": 49920
+    },
+    {
+      "epoch": 184.93,
+      "grad_norm": 1.1250598430633545,
+      "learning_rate": 0.00015074074074074073,
+      "loss": 0.4702,
+      "step": 49930
+    },
+    {
+      "epoch": 184.96,
+      "grad_norm": 1.1155190467834473,
+      "learning_rate": 0.00015037037037037037,
+      "loss": 0.4846,
+      "step": 49940
+    },
+    {
+      "epoch": 185.0,
+      "grad_norm": 2.0481882095336914,
+      "learning_rate": 0.00015,
+      "loss": 0.4802,
+      "step": 49950
+    },
+    {
+      "epoch": 185.04,
+      "grad_norm": 0.8488351702690125,
+      "learning_rate": 0.00014962962962962966,
+      "loss": 0.3989,
+      "step": 49960
+    },
+    {
+      "epoch": 185.07,
+      "grad_norm": 0.9094061255455017,
+      "learning_rate": 0.00014925925925925927,
+      "loss": 0.3977,
+      "step": 49970
+    },
+    {
+      "epoch": 185.11,
+      "grad_norm": 0.9350351095199585,
+      "learning_rate": 0.0001488888888888889,
+      "loss": 0.4065,
+      "step": 49980
+    },
+    {
+      "epoch": 185.15,
+      "grad_norm": 0.938935399055481,
+      "learning_rate": 0.00014851851851851853,
+      "loss": 0.4046,
+      "step": 49990
+    },
+    {
+      "epoch": 185.19,
+      "grad_norm": 0.9022473096847534,
+      "learning_rate": 0.00014814814814814815,
+      "loss": 0.4004,
+      "step": 50000
+    },
+    {
+      "epoch": 185.22,
+      "grad_norm": 0.9724648594856262,
+      "learning_rate": 0.0001477777777777778,
+      "loss": 0.4207,
+      "step": 50010
+    },
+    {
+      "epoch": 185.26,
+      "grad_norm": 0.9791123270988464,
+      "learning_rate": 0.0001474074074074074,
+      "loss": 0.4104,
+      "step": 50020
+    },
+    {
+      "epoch": 185.3,
+      "grad_norm": 1.002153992652893,
+      "learning_rate": 0.00014703703703703705,
+      "loss": 0.4349,
+      "step": 50030
+    },
+    {
+      "epoch": 185.33,
+      "grad_norm": 0.9777942299842834,
+      "learning_rate": 0.00014666666666666666,
+      "loss": 0.4431,
+      "step": 50040
+    },
+    {
+      "epoch": 185.37,
+      "grad_norm": 0.9837242364883423,
+      "learning_rate": 0.00014629629629629628,
+      "loss": 0.4335,
+      "step": 50050
+    },
+    {
+      "epoch": 185.41,
+      "grad_norm": 1.0102285146713257,
+      "learning_rate": 0.00014592592592592592,
+      "loss": 0.4342,
+      "step": 50060
+    },
+    {
+      "epoch": 185.44,
+      "grad_norm": 0.9805873036384583,
+      "learning_rate": 0.00014555555555555556,
+      "loss": 0.4377,
+      "step": 50070
+    },
+    {
+      "epoch": 185.48,
+      "grad_norm": 0.9970133900642395,
+      "learning_rate": 0.0001451851851851852,
+      "loss": 0.4319,
+      "step": 50080
+    },
+    {
+      "epoch": 185.52,
+      "grad_norm": 1.07515549659729,
+      "learning_rate": 0.00014481481481481482,
+      "loss": 0.4466,
+      "step": 50090
+    },
+    {
+      "epoch": 185.56,
+      "grad_norm": 0.9581437110900879,
+      "learning_rate": 0.00014444444444444444,
+      "loss": 0.4423,
+      "step": 50100
+    },
+    {
+      "epoch": 185.59,
+      "grad_norm": 1.0771484375,
+      "learning_rate": 0.00014407407407407408,
+      "loss": 0.4528,
+      "step": 50110
+    },
+    {
+      "epoch": 185.63,
+      "grad_norm": 1.1003572940826416,
+      "learning_rate": 0.0001437037037037037,
+      "loss": 0.449,
+      "step": 50120
+    },
+    {
+      "epoch": 185.67,
+      "grad_norm": 1.0268014669418335,
+      "learning_rate": 0.00014333333333333334,
+      "loss": 0.4548,
+      "step": 50130
+    },
+    {
+      "epoch": 185.7,
+      "grad_norm": 1.0169177055358887,
+      "learning_rate": 0.00014296296296296295,
+      "loss": 0.4519,
+      "step": 50140
+    },
+    {
+      "epoch": 185.74,
+      "grad_norm": 1.0599011182785034,
+      "learning_rate": 0.0001425925925925926,
+      "loss": 0.4671,
+      "step": 50150
+    },
+    {
+      "epoch": 185.78,
+      "grad_norm": 0.9963357448577881,
+      "learning_rate": 0.0001422222222222222,
+      "loss": 0.4674,
+      "step": 50160
+    },
+    {
+      "epoch": 185.81,
+      "grad_norm": 1.004539966583252,
+      "learning_rate": 0.00014185185185185186,
+      "loss": 0.4611,
+      "step": 50170
+    },
+    {
+      "epoch": 185.85,
+      "grad_norm": 1.0894609689712524,
+      "learning_rate": 0.0001414814814814815,
+      "loss": 0.461,
+      "step": 50180
+    },
+    {
+      "epoch": 185.89,
+      "grad_norm": 1.0321896076202393,
+      "learning_rate": 0.00014111111111111111,
+      "loss": 0.4665,
+      "step": 50190
+    },
+    {
+      "epoch": 185.93,
+      "grad_norm": 1.1291207075119019,
+      "learning_rate": 0.00014074074074074076,
+      "loss": 0.4683,
+      "step": 50200
+    },
+    {
+      "epoch": 185.96,
+      "grad_norm": 1.114660382270813,
+      "learning_rate": 0.00014037037037037037,
+      "loss": 0.4552,
+      "step": 50210
+    },
+    {
+      "epoch": 186.0,
+      "grad_norm": 2.1208958625793457,
+      "learning_rate": 0.00014000000000000001,
+      "loss": 0.4891,
+      "step": 50220
+    },
+    {
+      "epoch": 186.04,
+      "grad_norm": 0.8631995916366577,
+      "learning_rate": 0.00013962962962962963,
+      "loss": 0.383,
+      "step": 50230
+    },
+    {
+      "epoch": 186.07,
+      "grad_norm": 0.8872973322868347,
+      "learning_rate": 0.00013925925925925925,
+      "loss": 0.3906,
+      "step": 50240
+    },
+    {
+      "epoch": 186.11,
+      "grad_norm": 0.8547492027282715,
+      "learning_rate": 0.0001388888888888889,
+      "loss": 0.3952,
+      "step": 50250
+    },
+    {
+      "epoch": 186.15,
+      "grad_norm": 0.9142735004425049,
+      "learning_rate": 0.0001385185185185185,
+      "loss": 0.4063,
+      "step": 50260
+    },
+    {
+      "epoch": 186.19,
+      "grad_norm": 0.9310815930366516,
+      "learning_rate": 0.00013814814814814817,
+      "loss": 0.4043,
+      "step": 50270
+    },
+    {
+      "epoch": 186.22,
+      "grad_norm": 0.9587202668190002,
+      "learning_rate": 0.0001377777777777778,
+      "loss": 0.4163,
+      "step": 50280
+    },
+    {
+      "epoch": 186.26,
+      "grad_norm": 0.8949682116508484,
+      "learning_rate": 0.0001374074074074074,
+      "loss": 0.409,
+      "step": 50290
+    },
+    {
+      "epoch": 186.3,
+      "grad_norm": 0.9883646368980408,
+      "learning_rate": 0.00013703703703703705,
+      "loss": 0.4019,
+      "step": 50300
+    },
+    {
+      "epoch": 186.33,
+      "grad_norm": 1.0231529474258423,
+      "learning_rate": 0.00013666666666666666,
+      "loss": 0.4209,
+      "step": 50310
+    },
+    {
+      "epoch": 186.37,
+      "grad_norm": 0.9546399712562561,
+      "learning_rate": 0.0001362962962962963,
+      "loss": 0.4255,
+      "step": 50320
+    },
+    {
+      "epoch": 186.41,
+      "grad_norm": 0.9848965406417847,
+      "learning_rate": 0.00013592592592592592,
+      "loss": 0.415,
+      "step": 50330
+    },
+    {
+      "epoch": 186.44,
+      "grad_norm": 0.9679991006851196,
+      "learning_rate": 0.00013555555555555556,
+      "loss": 0.428,
+      "step": 50340
+    },
+    {
+      "epoch": 186.48,
+      "grad_norm": 1.0232571363449097,
+      "learning_rate": 0.00013518518518518518,
+      "loss": 0.4354,
+      "step": 50350
+    },
+    {
+      "epoch": 186.52,
+      "grad_norm": 1.018001675605774,
+      "learning_rate": 0.0001348148148148148,
+      "loss": 0.4345,
+      "step": 50360
+    },
+    {
+      "epoch": 186.56,
+      "grad_norm": 1.016631841659546,
+      "learning_rate": 0.00013444444444444447,
+      "loss": 0.4381,
+      "step": 50370
+    },
+    {
+      "epoch": 186.59,
+      "grad_norm": 1.0199625492095947,
+      "learning_rate": 0.00013407407407407408,
+      "loss": 0.4364,
+      "step": 50380
+    },
+    {
+      "epoch": 186.63,
+      "grad_norm": 1.0431368350982666,
+      "learning_rate": 0.00013370370370370372,
+      "loss": 0.4425,
+      "step": 50390
+    },
+    {
+      "epoch": 186.67,
+      "grad_norm": 1.0619491338729858,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.4425,
+      "step": 50400
+    },
+    {
+      "epoch": 186.7,
+      "grad_norm": 0.9911066293716431,
+      "learning_rate": 0.00013296296296296296,
+      "loss": 0.4477,
+      "step": 50410
+    },
+    {
+      "epoch": 186.74,
+      "grad_norm": 1.1205402612686157,
+      "learning_rate": 0.0001325925925925926,
+      "loss": 0.4635,
+      "step": 50420
+    },
+    {
+      "epoch": 186.78,
+      "grad_norm": 1.0640531778335571,
+      "learning_rate": 0.00013222222222222221,
+      "loss": 0.4656,
+      "step": 50430
+    },
+    {
+      "epoch": 186.81,
+      "grad_norm": 1.092454195022583,
+      "learning_rate": 0.00013185185185185186,
+      "loss": 0.4605,
+      "step": 50440
+    },
+    {
+      "epoch": 186.85,
+      "grad_norm": 1.084949254989624,
+      "learning_rate": 0.00013148148148148147,
+      "loss": 0.4588,
+      "step": 50450
+    },
+    {
+      "epoch": 186.89,
+      "grad_norm": 1.0481915473937988,
+      "learning_rate": 0.00013111111111111111,
+      "loss": 0.4642,
+      "step": 50460
+    },
+    {
+      "epoch": 186.93,
+      "grad_norm": 1.0836292505264282,
+      "learning_rate": 0.00013074074074074073,
+      "loss": 0.4644,
+      "step": 50470
+    },
+    {
+      "epoch": 186.96,
+      "grad_norm": 1.1251612901687622,
+      "learning_rate": 0.00013037037037037037,
+      "loss": 0.4697,
+      "step": 50480
+    },
+    {
+      "epoch": 187.0,
+      "grad_norm": 1.925938606262207,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 0.4711,
+      "step": 50490
+    },
+    {
+      "epoch": 187.04,
+      "grad_norm": 0.8456882834434509,
+      "learning_rate": 0.00012962962962962963,
+      "loss": 0.3921,
+      "step": 50500
+    },
+    {
+      "epoch": 187.07,
+      "grad_norm": 0.9210270047187805,
+      "learning_rate": 0.00012925925925925927,
+      "loss": 0.3783,
+      "step": 50510
+    },
+    {
+      "epoch": 187.11,
+      "grad_norm": 0.8768045902252197,
+      "learning_rate": 0.0001288888888888889,
+      "loss": 0.3902,
+      "step": 50520
+    },
+    {
+      "epoch": 187.15,
+      "grad_norm": 0.8845224380493164,
+      "learning_rate": 0.00012851851851851853,
+      "loss": 0.3985,
+      "step": 50530
+    },
+    {
+      "epoch": 187.19,
+      "grad_norm": 0.9048132300376892,
+      "learning_rate": 0.00012814814814814815,
+      "loss": 0.3951,
+      "step": 50540
+    },
+    {
+      "epoch": 187.22,
+      "grad_norm": 0.8952040076255798,
+      "learning_rate": 0.00012777777777777776,
+      "loss": 0.4099,
+      "step": 50550
+    },
+    {
+      "epoch": 187.26,
+      "grad_norm": 0.962620198726654,
+      "learning_rate": 0.0001274074074074074,
+      "loss": 0.4141,
+      "step": 50560
+    },
+    {
+      "epoch": 187.3,
+      "grad_norm": 0.9428215622901917,
+      "learning_rate": 0.00012703703703703702,
+      "loss": 0.4095,
+      "step": 50570
+    },
+    {
+      "epoch": 187.33,
+      "grad_norm": 0.9892339110374451,
+      "learning_rate": 0.0001266666666666667,
+      "loss": 0.4171,
+      "step": 50580
+    },
+    {
+      "epoch": 187.37,
+      "grad_norm": 1.010561466217041,
+      "learning_rate": 0.0001262962962962963,
+      "loss": 0.4103,
+      "step": 50590
+    },
+    {
+      "epoch": 187.41,
+      "grad_norm": 0.9834034442901611,
+      "learning_rate": 0.00012592592592592592,
+      "loss": 0.4231,
+      "step": 50600
+    },
+    {
+      "epoch": 187.44,
+      "grad_norm": 0.9447629451751709,
+      "learning_rate": 0.00012555555555555557,
+      "loss": 0.4162,
+      "step": 50610
+    },
+    {
+      "epoch": 187.48,
+      "grad_norm": 1.0351061820983887,
+      "learning_rate": 0.00012518518518518518,
+      "loss": 0.4317,
+      "step": 50620
+    },
+    {
+      "epoch": 187.52,
+      "grad_norm": 0.973498523235321,
+      "learning_rate": 0.00012481481481481482,
+      "loss": 0.4243,
+      "step": 50630
+    },
+    {
+      "epoch": 187.56,
+      "grad_norm": 0.9797120094299316,
+      "learning_rate": 0.00012444444444444444,
+      "loss": 0.4352,
+      "step": 50640
+    },
+    {
+      "epoch": 187.59,
+      "grad_norm": 0.987112283706665,
+      "learning_rate": 0.00012407407407407408,
+      "loss": 0.4287,
+      "step": 50650
+    },
+    {
+      "epoch": 187.63,
+      "grad_norm": 1.0124151706695557,
+      "learning_rate": 0.0001237037037037037,
+      "loss": 0.4365,
+      "step": 50660
+    },
+    {
+      "epoch": 187.67,
+      "grad_norm": 0.9849717617034912,
+      "learning_rate": 0.00012333333333333334,
+      "loss": 0.4454,
+      "step": 50670
+    },
+    {
+      "epoch": 187.7,
+      "grad_norm": 1.003054141998291,
+      "learning_rate": 0.00012296296296296298,
+      "loss": 0.4325,
+      "step": 50680
+    },
+    {
+      "epoch": 187.74,
+      "grad_norm": 1.047780156135559,
+      "learning_rate": 0.0001225925925925926,
+      "loss": 0.4407,
+      "step": 50690
+    },
+    {
+      "epoch": 187.78,
+      "grad_norm": 1.075384259223938,
+      "learning_rate": 0.00012222222222222221,
+      "loss": 0.4403,
+      "step": 50700
+    },
+    {
+      "epoch": 187.81,
+      "grad_norm": 1.0416277647018433,
+      "learning_rate": 0.00012185185185185186,
+      "loss": 0.4485,
+      "step": 50710
+    },
+    {
+      "epoch": 187.85,
+      "grad_norm": 1.1392602920532227,
+      "learning_rate": 0.00012148148148148149,
+      "loss": 0.4507,
+      "step": 50720
+    },
+    {
+      "epoch": 187.89,
+      "grad_norm": 1.0072075128555298,
+      "learning_rate": 0.00012111111111111112,
+      "loss": 0.4673,
+      "step": 50730
+    },
+    {
+      "epoch": 187.93,
+      "grad_norm": 1.0627766847610474,
+      "learning_rate": 0.00012074074074074074,
+      "loss": 0.4486,
+      "step": 50740
+    },
+    {
+      "epoch": 187.96,
+      "grad_norm": 1.0712461471557617,
+      "learning_rate": 0.00012037037037037036,
+      "loss": 0.4472,
+      "step": 50750
+    },
+    {
+      "epoch": 188.0,
+      "grad_norm": 1.8942756652832031,
+      "learning_rate": 0.00012,
+      "loss": 0.4591,
+      "step": 50760
+    },
+    {
+      "epoch": 188.04,
+      "grad_norm": 0.862216055393219,
+      "learning_rate": 0.00011962962962962963,
+      "loss": 0.3789,
+      "step": 50770
+    },
+    {
+      "epoch": 188.07,
+      "grad_norm": 0.8385865688323975,
+      "learning_rate": 0.00011925925925925926,
+      "loss": 0.3917,
+      "step": 50780
+    },
+    {
+      "epoch": 188.11,
+      "grad_norm": 0.9033856391906738,
+      "learning_rate": 0.00011888888888888889,
+      "loss": 0.3856,
+      "step": 50790
+    },
+    {
+      "epoch": 188.15,
+      "grad_norm": 0.9251241683959961,
+      "learning_rate": 0.00011851851851851852,
+      "loss": 0.3865,
+      "step": 50800
+    },
+    {
+      "epoch": 188.19,
+      "grad_norm": 0.8424283266067505,
+      "learning_rate": 0.00011814814814814816,
+      "loss": 0.3984,
+      "step": 50810
+    },
+    {
+      "epoch": 188.22,
+      "grad_norm": 0.9130561351776123,
+      "learning_rate": 0.00011777777777777778,
+      "loss": 0.3929,
+      "step": 50820
+    },
+    {
+      "epoch": 188.26,
+      "grad_norm": 0.9251916408538818,
+      "learning_rate": 0.00011740740740740741,
+      "loss": 0.4009,
+      "step": 50830
+    },
+    {
+      "epoch": 188.3,
+      "grad_norm": 0.935077965259552,
+      "learning_rate": 0.00011703703703703704,
+      "loss": 0.4057,
+      "step": 50840
+    },
+    {
+      "epoch": 188.33,
+      "grad_norm": 0.9387704133987427,
+      "learning_rate": 0.00011666666666666667,
+      "loss": 0.4097,
+      "step": 50850
+    },
+    {
+      "epoch": 188.37,
+      "grad_norm": 0.970709502696991,
+      "learning_rate": 0.00011629629629629631,
+      "loss": 0.411,
+      "step": 50860
+    },
+    {
+      "epoch": 188.41,
+      "grad_norm": 0.9760639667510986,
+      "learning_rate": 0.00011592592592592594,
+      "loss": 0.4105,
+      "step": 50870
+    },
+    {
+      "epoch": 188.44,
+      "grad_norm": 0.9423500895500183,
+      "learning_rate": 0.00011555555555555555,
+      "loss": 0.4136,
+      "step": 50880
+    },
+    {
+      "epoch": 188.48,
+      "grad_norm": 0.9758352041244507,
+      "learning_rate": 0.00011518518518518518,
+      "loss": 0.4092,
+      "step": 50890
+    },
+    {
+      "epoch": 188.52,
+      "grad_norm": 1.012181043624878,
+      "learning_rate": 0.00011481481481481481,
+      "loss": 0.4238,
+      "step": 50900
+    },
+    {
+      "epoch": 188.56,
+      "grad_norm": 1.0230450630187988,
+      "learning_rate": 0.00011444444444444445,
+      "loss": 0.4244,
+      "step": 50910
+    },
+    {
+      "epoch": 188.59,
+      "grad_norm": 0.9596276879310608,
+      "learning_rate": 0.00011407407407407408,
+      "loss": 0.4173,
+      "step": 50920
+    },
+    {
+      "epoch": 188.63,
+      "grad_norm": 0.9961845278739929,
+      "learning_rate": 0.00011370370370370371,
+      "loss": 0.4259,
+      "step": 50930
+    },
+    {
+      "epoch": 188.67,
+      "grad_norm": 1.0274980068206787,
+      "learning_rate": 0.00011333333333333333,
+      "loss": 0.4284,
+      "step": 50940
+    },
+    {
+      "epoch": 188.7,
+      "grad_norm": 1.023209571838379,
+      "learning_rate": 0.00011296296296296296,
+      "loss": 0.4429,
+      "step": 50950
+    },
+    {
+      "epoch": 188.74,
+      "grad_norm": 1.0073031187057495,
+      "learning_rate": 0.0001125925925925926,
+      "loss": 0.4287,
+      "step": 50960
+    },
+    {
+      "epoch": 188.78,
+      "grad_norm": 1.0359915494918823,
+      "learning_rate": 0.00011222222222222223,
+      "loss": 0.4405,
+      "step": 50970
+    },
+    {
+      "epoch": 188.81,
+      "grad_norm": 0.9596119523048401,
+      "learning_rate": 0.00011185185185185186,
+      "loss": 0.4434,
+      "step": 50980
+    },
+    {
+      "epoch": 188.85,
+      "grad_norm": 0.9704906344413757,
+      "learning_rate": 0.00011148148148148149,
+      "loss": 0.4347,
+      "step": 50990
+    },
+    {
+      "epoch": 188.89,
+      "grad_norm": 1.03348970413208,
+      "learning_rate": 0.0001111111111111111,
+      "loss": 0.4527,
+      "step": 51000
+    },
+    {
+      "epoch": 188.93,
+      "grad_norm": 1.060300588607788,
+      "learning_rate": 0.00011074074074074075,
+      "loss": 0.4533,
+      "step": 51010
+    },
+    {
+      "epoch": 188.96,
+      "grad_norm": 1.0980509519577026,
+      "learning_rate": 0.00011037037037037037,
+      "loss": 0.4441,
+      "step": 51020
+    },
+    {
+      "epoch": 189.0,
+      "grad_norm": 2.1591084003448486,
+      "learning_rate": 0.00011,
+      "loss": 0.4329,
+      "step": 51030
+    },
+    {
+      "epoch": 189.04,
+      "grad_norm": 0.8770800828933716,
+      "learning_rate": 0.00010962962962962963,
+      "loss": 0.3757,
+      "step": 51040
+    },
+    {
+      "epoch": 189.07,
+      "grad_norm": 0.8365693688392639,
+      "learning_rate": 0.00010925925925925926,
+      "loss": 0.3843,
+      "step": 51050
+    },
+    {
+      "epoch": 189.11,
+      "grad_norm": 0.8623953461647034,
+      "learning_rate": 0.00010888888888888888,
+      "loss": 0.3778,
+      "step": 51060
+    },
+    {
+      "epoch": 189.15,
+      "grad_norm": 0.8259143233299255,
+      "learning_rate": 0.00010851851851851852,
+      "loss": 0.3798,
+      "step": 51070
+    },
+    {
+      "epoch": 189.19,
+      "grad_norm": 0.8634757995605469,
+      "learning_rate": 0.00010814814814814815,
+      "loss": 0.3778,
+      "step": 51080
+    },
+    {
+      "epoch": 189.22,
+      "grad_norm": 0.8648687601089478,
+      "learning_rate": 0.00010777777777777778,
+      "loss": 0.3899,
+      "step": 51090
+    },
+    {
+      "epoch": 189.26,
+      "grad_norm": 0.8897496461868286,
+      "learning_rate": 0.00010740740740740741,
+      "loss": 0.3879,
+      "step": 51100
+    },
+    {
+      "epoch": 189.3,
+      "grad_norm": 0.887324333190918,
+      "learning_rate": 0.00010703703703703704,
+      "loss": 0.3915,
+      "step": 51110
+    },
+    {
+      "epoch": 189.33,
+      "grad_norm": 0.9308483600616455,
+      "learning_rate": 0.00010666666666666668,
+      "loss": 0.4084,
+      "step": 51120
+    },
+    {
+      "epoch": 189.37,
+      "grad_norm": 0.9871852993965149,
+      "learning_rate": 0.0001062962962962963,
+      "loss": 0.4119,
+      "step": 51130
+    },
+    {
+      "epoch": 189.41,
+      "grad_norm": 0.9372105002403259,
+      "learning_rate": 0.00010592592592592592,
+      "loss": 0.4044,
+      "step": 51140
+    },
+    {
+      "epoch": 189.44,
+      "grad_norm": 0.8799377083778381,
+      "learning_rate": 0.00010555555555555555,
+      "loss": 0.4072,
+      "step": 51150
+    },
+    {
+      "epoch": 189.48,
+      "grad_norm": 0.9507384896278381,
+      "learning_rate": 0.00010518518518518518,
+      "loss": 0.4168,
+      "step": 51160
+    },
+    {
+      "epoch": 189.52,
+      "grad_norm": 0.9813037514686584,
+      "learning_rate": 0.00010481481481481483,
+      "loss": 0.4126,
+      "step": 51170
+    },
+    {
+      "epoch": 189.56,
+      "grad_norm": 0.9744457006454468,
+      "learning_rate": 0.00010444444444444445,
+      "loss": 0.4229,
+      "step": 51180
+    },
+    {
+      "epoch": 189.59,
+      "grad_norm": 1.0010546445846558,
+      "learning_rate": 0.00010407407407407407,
+      "loss": 0.4178,
+      "step": 51190
+    },
+    {
+      "epoch": 189.63,
+      "grad_norm": 1.0330417156219482,
+      "learning_rate": 0.0001037037037037037,
+      "loss": 0.4212,
+      "step": 51200
+    },
+    {
+      "epoch": 189.67,
+      "grad_norm": 0.99884033203125,
+      "learning_rate": 0.00010333333333333333,
+      "loss": 0.4267,
+      "step": 51210
+    },
+    {
+      "epoch": 189.7,
+      "grad_norm": 1.0357877016067505,
+      "learning_rate": 0.00010296296296296297,
+      "loss": 0.4199,
+      "step": 51220
+    },
+    {
+      "epoch": 189.74,
+      "grad_norm": 0.9598915576934814,
+      "learning_rate": 0.0001025925925925926,
+      "loss": 0.4331,
+      "step": 51230
+    },
+    {
+      "epoch": 189.78,
+      "grad_norm": 1.012041449546814,
+      "learning_rate": 0.00010222222222222223,
+      "loss": 0.4225,
+      "step": 51240
+    },
+    {
+      "epoch": 189.81,
+      "grad_norm": 0.9845399260520935,
+      "learning_rate": 0.00010185185185185185,
+      "loss": 0.4214,
+      "step": 51250
+    },
+    {
+      "epoch": 189.85,
+      "grad_norm": 0.9941114187240601,
+      "learning_rate": 0.00010148148148148147,
+      "loss": 0.4436,
+      "step": 51260
+    },
+    {
+      "epoch": 189.89,
+      "grad_norm": 1.0230532884597778,
+      "learning_rate": 0.00010111111111111112,
+      "loss": 0.4325,
+      "step": 51270
+    },
+    {
+      "epoch": 189.93,
+      "grad_norm": 1.022606372833252,
+      "learning_rate": 0.00010074074074074075,
+      "loss": 0.4266,
+      "step": 51280
+    },
+    {
+      "epoch": 189.96,
+      "grad_norm": 0.9943325519561768,
+      "learning_rate": 0.00010037037037037038,
+      "loss": 0.4408,
+      "step": 51290
+    },
+    {
+      "epoch": 190.0,
+      "grad_norm": 1.894622564315796,
+      "learning_rate": 0.0001,
+      "loss": 0.4405,
+      "step": 51300
+    },
+    {
+      "epoch": 190.04,
+      "grad_norm": 0.7975081205368042,
+      "learning_rate": 9.962962962962962e-05,
+      "loss": 0.3793,
+      "step": 51310
+    },
+    {
+      "epoch": 190.07,
+      "grad_norm": 0.828575611114502,
+      "learning_rate": 9.925925925925926e-05,
+      "loss": 0.3897,
+      "step": 51320
+    },
+    {
+      "epoch": 190.11,
+      "grad_norm": 0.8125698566436768,
+      "learning_rate": 9.888888888888889e-05,
+      "loss": 0.3756,
+      "step": 51330
+    },
+    {
+      "epoch": 190.15,
+      "grad_norm": 0.9077195525169373,
+      "learning_rate": 9.851851851851852e-05,
+      "loss": 0.3801,
+      "step": 51340
+    },
+    {
+      "epoch": 190.19,
+      "grad_norm": 0.8815000057220459,
+      "learning_rate": 9.814814814814815e-05,
+      "loss": 0.3766,
+      "step": 51350
+    },
+    {
+      "epoch": 190.22,
+      "grad_norm": 0.8989863395690918,
+      "learning_rate": 9.777777777777778e-05,
+      "loss": 0.3803,
+      "step": 51360
+    },
+    {
+      "epoch": 190.26,
+      "grad_norm": 0.8354569673538208,
+      "learning_rate": 9.740740740740741e-05,
+      "loss": 0.3859,
+      "step": 51370
+    },
+    {
+      "epoch": 190.3,
+      "grad_norm": 0.9326614141464233,
+      "learning_rate": 9.703703703703704e-05,
+      "loss": 0.3903,
+      "step": 51380
+    },
+    {
+      "epoch": 190.33,
+      "grad_norm": 0.9138501286506653,
+      "learning_rate": 9.666666666666667e-05,
+      "loss": 0.3835,
+      "step": 51390
+    },
+    {
+      "epoch": 190.37,
+      "grad_norm": 0.9268086552619934,
+      "learning_rate": 9.62962962962963e-05,
+      "loss": 0.3952,
+      "step": 51400
+    },
+    {
+      "epoch": 190.41,
+      "grad_norm": 0.8911784887313843,
+      "learning_rate": 9.592592592592593e-05,
+      "loss": 0.3999,
+      "step": 51410
+    },
+    {
+      "epoch": 190.44,
+      "grad_norm": 1.0090967416763306,
+      "learning_rate": 9.555555555555557e-05,
+      "loss": 0.4031,
+      "step": 51420
+    },
+    {
+      "epoch": 190.48,
+      "grad_norm": 0.9447932243347168,
+      "learning_rate": 9.51851851851852e-05,
+      "loss": 0.4105,
+      "step": 51430
+    },
+    {
+      "epoch": 190.52,
+      "grad_norm": 0.9692918658256531,
+      "learning_rate": 9.481481481481481e-05,
+      "loss": 0.3992,
+      "step": 51440
+    },
+    {
+      "epoch": 190.56,
+      "grad_norm": 0.9399166703224182,
+      "learning_rate": 9.444444444444444e-05,
+      "loss": 0.4175,
+      "step": 51450
+    },
+    {
+      "epoch": 190.59,
+      "grad_norm": 0.9361637830734253,
+      "learning_rate": 9.407407407407407e-05,
+      "loss": 0.4015,
+      "step": 51460
+    },
+    {
+      "epoch": 190.63,
+      "grad_norm": 0.9695074558258057,
+      "learning_rate": 9.370370370370371e-05,
+      "loss": 0.4076,
+      "step": 51470
+    },
+    {
+      "epoch": 190.67,
+      "grad_norm": 0.957758367061615,
+      "learning_rate": 9.333333333333334e-05,
+      "loss": 0.4195,
+      "step": 51480
+    },
+    {
+      "epoch": 190.7,
+      "grad_norm": 0.9265300631523132,
+      "learning_rate": 9.296296296296297e-05,
+      "loss": 0.4211,
+      "step": 51490
+    },
+    {
+      "epoch": 190.74,
+      "grad_norm": 0.9656269550323486,
+      "learning_rate": 9.259259259259259e-05,
+      "loss": 0.4063,
+      "step": 51500
+    },
+    {
+      "epoch": 190.78,
+      "grad_norm": 1.0088058710098267,
+      "learning_rate": 9.222222222222222e-05,
+      "loss": 0.4265,
+      "step": 51510
+    },
+    {
+      "epoch": 190.81,
+      "grad_norm": 0.989260733127594,
+      "learning_rate": 9.185185185185186e-05,
+      "loss": 0.4262,
+      "step": 51520
+    },
+    {
+      "epoch": 190.85,
+      "grad_norm": 1.060614824295044,
+      "learning_rate": 9.148148148148149e-05,
+      "loss": 0.4307,
+      "step": 51530
+    },
+    {
+      "epoch": 190.89,
+      "grad_norm": 1.0297032594680786,
+      "learning_rate": 9.111111111111112e-05,
+      "loss": 0.4281,
+      "step": 51540
+    },
+    {
+      "epoch": 190.93,
+      "grad_norm": 0.989966869354248,
+      "learning_rate": 9.074074074074075e-05,
+      "loss": 0.4226,
+      "step": 51550
+    },
+    {
+      "epoch": 190.96,
+      "grad_norm": 1.0322147607803345,
+      "learning_rate": 9.037037037037036e-05,
+      "loss": 0.4361,
+      "step": 51560
+    },
+    {
+      "epoch": 191.0,
+      "grad_norm": 1.8464624881744385,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 0.4243,
+      "step": 51570
+    },
+    {
+      "epoch": 191.04,
+      "grad_norm": 0.8357323408126831,
+      "learning_rate": 8.962962962962963e-05,
+      "loss": 0.3711,
+      "step": 51580
+    },
+    {
+      "epoch": 191.07,
+      "grad_norm": 0.8935667276382446,
+      "learning_rate": 8.925925925925926e-05,
+      "loss": 0.3692,
+      "step": 51590
+    },
+    {
+      "epoch": 191.11,
+      "grad_norm": 0.8498260378837585,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 0.385,
+      "step": 51600
+    },
+    {
+      "epoch": 191.15,
+      "grad_norm": 0.837672770023346,
+      "learning_rate": 8.851851851851852e-05,
+      "loss": 0.3816,
+      "step": 51610
+    },
+    {
+      "epoch": 191.19,
+      "grad_norm": 0.8953566551208496,
+      "learning_rate": 8.814814814814814e-05,
+      "loss": 0.3835,
+      "step": 51620
+    },
+    {
+      "epoch": 191.22,
+      "grad_norm": 0.8604071736335754,
+      "learning_rate": 8.777777777777778e-05,
+      "loss": 0.3691,
+      "step": 51630
+    },
+    {
+      "epoch": 191.26,
+      "grad_norm": 0.9045225381851196,
+      "learning_rate": 8.740740740740741e-05,
+      "loss": 0.3754,
+      "step": 51640
+    },
+    {
+      "epoch": 191.3,
+      "grad_norm": 0.9023613333702087,
+      "learning_rate": 8.703703703703704e-05,
+      "loss": 0.3867,
+      "step": 51650
+    },
+    {
+      "epoch": 191.33,
+      "grad_norm": 0.8950648307800293,
+      "learning_rate": 8.666666666666667e-05,
+      "loss": 0.3728,
+      "step": 51660
+    },
+    {
+      "epoch": 191.37,
+      "grad_norm": 0.9078792929649353,
+      "learning_rate": 8.62962962962963e-05,
+      "loss": 0.3841,
+      "step": 51670
+    },
+    {
+      "epoch": 191.41,
+      "grad_norm": 0.8773408532142639,
+      "learning_rate": 8.592592592592593e-05,
+      "loss": 0.4035,
+      "step": 51680
+    },
+    {
+      "epoch": 191.44,
+      "grad_norm": 0.9532427191734314,
+      "learning_rate": 8.555555555555556e-05,
+      "loss": 0.3946,
+      "step": 51690
+    },
+    {
+      "epoch": 191.48,
+      "grad_norm": 0.9748596549034119,
+      "learning_rate": 8.518518518518518e-05,
+      "loss": 0.3969,
+      "step": 51700
+    },
+    {
+      "epoch": 191.52,
+      "grad_norm": 0.913264811038971,
+      "learning_rate": 8.481481481481481e-05,
+      "loss": 0.4019,
+      "step": 51710
+    },
+    {
+      "epoch": 191.56,
+      "grad_norm": 0.8992056250572205,
+      "learning_rate": 8.444444444444444e-05,
+      "loss": 0.4035,
+      "step": 51720
+    },
+    {
+      "epoch": 191.59,
+      "grad_norm": 0.9542635679244995,
+      "learning_rate": 8.407407407407409e-05,
+      "loss": 0.4045,
+      "step": 51730
+    },
+    {
+      "epoch": 191.63,
+      "grad_norm": 1.0166466236114502,
+      "learning_rate": 8.37037037037037e-05,
+      "loss": 0.4131,
+      "step": 51740
+    },
+    {
+      "epoch": 191.67,
+      "grad_norm": 0.9533120393753052,
+      "learning_rate": 8.333333333333333e-05,
+      "loss": 0.4115,
+      "step": 51750
+    },
+    {
+      "epoch": 191.7,
+      "grad_norm": 1.03888738155365,
+      "learning_rate": 8.296296296296296e-05,
+      "loss": 0.4052,
+      "step": 51760
+    },
+    {
+      "epoch": 191.74,
+      "grad_norm": 0.9723010063171387,
+      "learning_rate": 8.259259259259259e-05,
+      "loss": 0.4189,
+      "step": 51770
+    },
+    {
+      "epoch": 191.78,
+      "grad_norm": 0.9556922912597656,
+      "learning_rate": 8.222222222222223e-05,
+      "loss": 0.4126,
+      "step": 51780
+    },
+    {
+      "epoch": 191.81,
+      "grad_norm": 0.9536474347114563,
+      "learning_rate": 8.185185185185186e-05,
+      "loss": 0.4117,
+      "step": 51790
+    },
+    {
+      "epoch": 191.85,
+      "grad_norm": 1.0156913995742798,
+      "learning_rate": 8.148148148148149e-05,
+      "loss": 0.4031,
+      "step": 51800
+    },
+    {
+      "epoch": 191.89,
+      "grad_norm": 0.9756938815116882,
+      "learning_rate": 8.11111111111111e-05,
+      "loss": 0.4143,
+      "step": 51810
+    },
+    {
+      "epoch": 191.93,
+      "grad_norm": 0.9222599864006042,
+      "learning_rate": 8.074074074074073e-05,
+      "loss": 0.4268,
+      "step": 51820
+    },
+    {
+      "epoch": 191.96,
+      "grad_norm": 1.0029851198196411,
+      "learning_rate": 8.037037037037038e-05,
+      "loss": 0.4188,
+      "step": 51830
+    },
+    {
+      "epoch": 192.0,
+      "grad_norm": 1.8502293825149536,
+      "learning_rate": 8e-05,
+      "loss": 0.422,
+      "step": 51840
+    },
+    {
+      "epoch": 192.04,
+      "grad_norm": 0.8421656489372253,
+      "learning_rate": 7.962962962962964e-05,
+      "loss": 0.3644,
+      "step": 51850
+    },
+    {
+      "epoch": 192.07,
+      "grad_norm": 0.7998185157775879,
+      "learning_rate": 7.925925925925926e-05,
+      "loss": 0.368,
+      "step": 51860
+    },
+    {
+      "epoch": 192.11,
+      "grad_norm": 0.8072201013565063,
+      "learning_rate": 7.888888888888888e-05,
+      "loss": 0.3671,
+      "step": 51870
+    },
+    {
+      "epoch": 192.15,
+      "grad_norm": 0.8365472555160522,
+      "learning_rate": 7.851851851851852e-05,
+      "loss": 0.3645,
+      "step": 51880
+    },
+    {
+      "epoch": 192.19,
+      "grad_norm": 0.8714311718940735,
+      "learning_rate": 7.814814814814815e-05,
+      "loss": 0.3776,
+      "step": 51890
+    },
+    {
+      "epoch": 192.22,
+      "grad_norm": 0.8516049385070801,
+      "learning_rate": 7.777777777777778e-05,
+      "loss": 0.3858,
+      "step": 51900
+    },
+    {
+      "epoch": 192.26,
+      "grad_norm": 0.8652268052101135,
+      "learning_rate": 7.740740740740741e-05,
+      "loss": 0.387,
+      "step": 51910
+    },
+    {
+      "epoch": 192.3,
+      "grad_norm": 0.8955600261688232,
+      "learning_rate": 7.703703703703704e-05,
+      "loss": 0.3827,
+      "step": 51920
+    },
+    {
+      "epoch": 192.33,
+      "grad_norm": 0.9093692302703857,
+      "learning_rate": 7.666666666666667e-05,
+      "loss": 0.3742,
+      "step": 51930
+    },
+    {
+      "epoch": 192.37,
+      "grad_norm": 0.8596659898757935,
+      "learning_rate": 7.62962962962963e-05,
+      "loss": 0.3819,
+      "step": 51940
+    },
+    {
+      "epoch": 192.41,
+      "grad_norm": 0.9333339333534241,
+      "learning_rate": 7.592592592592593e-05,
+      "loss": 0.3825,
+      "step": 51950
+    },
+    {
+      "epoch": 192.44,
+      "grad_norm": 0.960281252861023,
+      "learning_rate": 7.555555555555556e-05,
+      "loss": 0.3917,
+      "step": 51960
+    },
+    {
+      "epoch": 192.48,
+      "grad_norm": 0.905362069606781,
+      "learning_rate": 7.518518518518519e-05,
+      "loss": 0.3956,
+      "step": 51970
+    },
+    {
+      "epoch": 192.52,
+      "grad_norm": 0.9406252503395081,
+      "learning_rate": 7.481481481481483e-05,
+      "loss": 0.3814,
+      "step": 51980
+    },
+    {
+      "epoch": 192.56,
+      "grad_norm": 0.8892878890037537,
+      "learning_rate": 7.444444444444444e-05,
+      "loss": 0.3888,
+      "step": 51990
+    },
+    {
+      "epoch": 192.59,
+      "grad_norm": 0.9101101756095886,
+      "learning_rate": 7.407407407407407e-05,
+      "loss": 0.3983,
+      "step": 52000
+    },
+    {
+      "epoch": 192.63,
+      "grad_norm": 1.0517656803131104,
+      "learning_rate": 7.37037037037037e-05,
+      "loss": 0.3925,
+      "step": 52010
+    },
+    {
+      "epoch": 192.67,
+      "grad_norm": 0.9784597754478455,
+      "learning_rate": 7.333333333333333e-05,
+      "loss": 0.4013,
+      "step": 52020
+    },
+    {
+      "epoch": 192.7,
+      "grad_norm": 0.9348440766334534,
+      "learning_rate": 7.296296296296296e-05,
+      "loss": 0.3899,
+      "step": 52030
+    },
+    {
+      "epoch": 192.74,
+      "grad_norm": 0.9543972015380859,
+      "learning_rate": 7.25925925925926e-05,
+      "loss": 0.3974,
+      "step": 52040
+    },
+    {
+      "epoch": 192.78,
+      "grad_norm": 0.97087162733078,
+      "learning_rate": 7.222222222222222e-05,
+      "loss": 0.4147,
+      "step": 52050
+    },
+    {
+      "epoch": 192.81,
+      "grad_norm": 0.9236013889312744,
+      "learning_rate": 7.185185185185185e-05,
+      "loss": 0.4116,
+      "step": 52060
+    },
+    {
+      "epoch": 192.85,
+      "grad_norm": 0.9513102769851685,
+      "learning_rate": 7.148148148148148e-05,
+      "loss": 0.4038,
+      "step": 52070
+    },
+    {
+      "epoch": 192.89,
+      "grad_norm": 0.9712694883346558,
+      "learning_rate": 7.11111111111111e-05,
+      "loss": 0.4164,
+      "step": 52080
+    },
+    {
+      "epoch": 192.93,
+      "grad_norm": 0.9384704828262329,
+      "learning_rate": 7.074074074074075e-05,
+      "loss": 0.4071,
+      "step": 52090
+    },
+    {
+      "epoch": 192.96,
+      "grad_norm": 0.9361010789871216,
+      "learning_rate": 7.037037037037038e-05,
+      "loss": 0.4165,
+      "step": 52100
+    },
+    {
+      "epoch": 193.0,
+      "grad_norm": 1.679486632347107,
+      "learning_rate": 7.000000000000001e-05,
+      "loss": 0.4205,
+      "step": 52110
+    },
+    {
+      "epoch": 193.04,
+      "grad_norm": 0.812853991985321,
+      "learning_rate": 6.962962962962962e-05,
+      "loss": 0.3496,
+      "step": 52120
+    },
+    {
+      "epoch": 193.07,
+      "grad_norm": 0.7948511838912964,
+      "learning_rate": 6.925925925925925e-05,
+      "loss": 0.3662,
+      "step": 52130
+    },
+    {
+      "epoch": 193.11,
+      "grad_norm": 0.841355562210083,
+      "learning_rate": 6.88888888888889e-05,
+      "loss": 0.3572,
+      "step": 52140
+    },
+    {
+      "epoch": 193.15,
+      "grad_norm": 0.8733651041984558,
+      "learning_rate": 6.851851851851852e-05,
+      "loss": 0.3658,
+      "step": 52150
+    },
+    {
+      "epoch": 193.19,
+      "grad_norm": 0.8185422420501709,
+      "learning_rate": 6.814814814814815e-05,
+      "loss": 0.3748,
+      "step": 52160
+    },
+    {
+      "epoch": 193.22,
+      "grad_norm": 0.8644742369651794,
+      "learning_rate": 6.777777777777778e-05,
+      "loss": 0.3768,
+      "step": 52170
+    },
+    {
+      "epoch": 193.26,
+      "grad_norm": 0.8009930849075317,
+      "learning_rate": 6.74074074074074e-05,
+      "loss": 0.3775,
+      "step": 52180
+    },
+    {
+      "epoch": 193.3,
+      "grad_norm": 0.8725280165672302,
+      "learning_rate": 6.703703703703704e-05,
+      "loss": 0.3705,
+      "step": 52190
+    },
+    {
+      "epoch": 193.33,
+      "grad_norm": 0.8709507584571838,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.3712,
+      "step": 52200
+    },
+    {
+      "epoch": 193.37,
+      "grad_norm": 0.8681923747062683,
+      "learning_rate": 6.62962962962963e-05,
+      "loss": 0.372,
+      "step": 52210
+    },
+    {
+      "epoch": 193.41,
+      "grad_norm": 0.8824298977851868,
+      "learning_rate": 6.592592592592593e-05,
+      "loss": 0.3832,
+      "step": 52220
+    },
+    {
+      "epoch": 193.44,
+      "grad_norm": 0.8672235608100891,
+      "learning_rate": 6.555555555555556e-05,
+      "loss": 0.3698,
+      "step": 52230
+    },
+    {
+      "epoch": 193.48,
+      "grad_norm": 0.8637567162513733,
+      "learning_rate": 6.518518518518519e-05,
+      "loss": 0.3877,
+      "step": 52240
+    },
+    {
+      "epoch": 193.52,
+      "grad_norm": 0.8623505234718323,
+      "learning_rate": 6.481481481481482e-05,
+      "loss": 0.3889,
+      "step": 52250
+    },
+    {
+      "epoch": 193.56,
+      "grad_norm": 0.904331386089325,
+      "learning_rate": 6.444444444444444e-05,
+      "loss": 0.4042,
+      "step": 52260
+    },
+    {
+      "epoch": 193.59,
+      "grad_norm": 0.9281720519065857,
+      "learning_rate": 6.407407407407407e-05,
+      "loss": 0.3965,
+      "step": 52270
+    },
+    {
+      "epoch": 193.63,
+      "grad_norm": 0.9167261123657227,
+      "learning_rate": 6.37037037037037e-05,
+      "loss": 0.3921,
+      "step": 52280
+    },
+    {
+      "epoch": 193.67,
+      "grad_norm": 0.889362633228302,
+      "learning_rate": 6.333333333333335e-05,
+      "loss": 0.3895,
+      "step": 52290
+    },
+    {
+      "epoch": 193.7,
+      "grad_norm": 0.9352006912231445,
+      "learning_rate": 6.296296296296296e-05,
+      "loss": 0.3967,
+      "step": 52300
+    },
+    {
+      "epoch": 193.74,
+      "grad_norm": 0.9324561953544617,
+      "learning_rate": 6.259259259259259e-05,
+      "loss": 0.3975,
+      "step": 52310
+    },
+    {
+      "epoch": 193.78,
+      "grad_norm": 0.9033541083335876,
+      "learning_rate": 6.222222222222222e-05,
+      "loss": 0.3965,
+      "step": 52320
+    },
+    {
+      "epoch": 193.81,
+      "grad_norm": 0.9311366677284241,
+      "learning_rate": 6.185185185185185e-05,
+      "loss": 0.4015,
+      "step": 52330
+    },
+    {
+      "epoch": 193.85,
+      "grad_norm": 0.9424006938934326,
+      "learning_rate": 6.148148148148149e-05,
+      "loss": 0.3873,
+      "step": 52340
+    },
+    {
+      "epoch": 193.89,
+      "grad_norm": 0.9936563968658447,
+      "learning_rate": 6.111111111111111e-05,
+      "loss": 0.4089,
+      "step": 52350
+    },
+    {
+      "epoch": 193.93,
+      "grad_norm": 0.927615225315094,
+      "learning_rate": 6.074074074074074e-05,
+      "loss": 0.3986,
+      "step": 52360
+    },
+    {
+      "epoch": 193.96,
+      "grad_norm": 0.9637607932090759,
+      "learning_rate": 6.037037037037037e-05,
+      "loss": 0.4097,
+      "step": 52370
+    },
+    {
+      "epoch": 194.0,
+      "grad_norm": 1.747179388999939,
+      "learning_rate": 6e-05,
+      "loss": 0.4011,
+      "step": 52380
+    },
+    {
+      "epoch": 194.04,
+      "grad_norm": 0.8385775089263916,
+      "learning_rate": 5.962962962962963e-05,
+      "loss": 0.3581,
+      "step": 52390
+    },
+    {
+      "epoch": 194.07,
+      "grad_norm": 0.8298104405403137,
+      "learning_rate": 5.925925925925926e-05,
+      "loss": 0.3515,
+      "step": 52400
+    },
+    {
+      "epoch": 194.11,
+      "grad_norm": 0.8504098653793335,
+      "learning_rate": 5.888888888888889e-05,
+      "loss": 0.3593,
+      "step": 52410
+    },
+    {
+      "epoch": 194.15,
+      "grad_norm": 0.8432555794715881,
+      "learning_rate": 5.851851851851852e-05,
+      "loss": 0.3615,
+      "step": 52420
+    },
+    {
+      "epoch": 194.19,
+      "grad_norm": 0.8588807582855225,
+      "learning_rate": 5.8148148148148154e-05,
+      "loss": 0.3665,
+      "step": 52430
+    },
+    {
+      "epoch": 194.22,
+      "grad_norm": 0.8673165440559387,
+      "learning_rate": 5.7777777777777776e-05,
+      "loss": 0.3729,
+      "step": 52440
+    },
+    {
+      "epoch": 194.26,
+      "grad_norm": 0.8235089182853699,
+      "learning_rate": 5.7407407407407406e-05,
+      "loss": 0.3636,
+      "step": 52450
+    },
+    {
+      "epoch": 194.3,
+      "grad_norm": 0.8563822507858276,
+      "learning_rate": 5.703703703703704e-05,
+      "loss": 0.3736,
+      "step": 52460
+    },
+    {
+      "epoch": 194.33,
+      "grad_norm": 0.881270170211792,
+      "learning_rate": 5.6666666666666664e-05,
+      "loss": 0.365,
+      "step": 52470
+    },
+    {
+      "epoch": 194.37,
+      "grad_norm": 0.9025720357894897,
+      "learning_rate": 5.62962962962963e-05,
+      "loss": 0.3782,
+      "step": 52480
+    },
+    {
+      "epoch": 194.41,
+      "grad_norm": 0.858967661857605,
+      "learning_rate": 5.592592592592593e-05,
+      "loss": 0.3863,
+      "step": 52490
+    },
+    {
+      "epoch": 194.44,
+      "grad_norm": 0.8928357362747192,
+      "learning_rate": 5.555555555555555e-05,
+      "loss": 0.3872,
+      "step": 52500
+    },
+    {
+      "epoch": 194.48,
+      "grad_norm": 0.8482757806777954,
+      "learning_rate": 5.518518518518519e-05,
+      "loss": 0.3785,
+      "step": 52510
+    },
+    {
+      "epoch": 194.52,
+      "grad_norm": 0.930858850479126,
+      "learning_rate": 5.4814814814814817e-05,
+      "loss": 0.3812,
+      "step": 52520
+    },
+    {
+      "epoch": 194.56,
+      "grad_norm": 0.8745943307876587,
+      "learning_rate": 5.444444444444444e-05,
+      "loss": 0.3792,
+      "step": 52530
+    },
+    {
+      "epoch": 194.59,
+      "grad_norm": 0.9000018239021301,
+      "learning_rate": 5.4074074074074075e-05,
+      "loss": 0.3812,
+      "step": 52540
+    },
+    {
+      "epoch": 194.63,
+      "grad_norm": 0.8568631410598755,
+      "learning_rate": 5.3703703703703704e-05,
+      "loss": 0.3838,
+      "step": 52550
+    },
+    {
+      "epoch": 194.67,
+      "grad_norm": 0.8851279020309448,
+      "learning_rate": 5.333333333333334e-05,
+      "loss": 0.3793,
+      "step": 52560
+    },
+    {
+      "epoch": 194.7,
+      "grad_norm": 0.9476543068885803,
+      "learning_rate": 5.296296296296296e-05,
+      "loss": 0.3962,
+      "step": 52570
+    },
+    {
+      "epoch": 194.74,
+      "grad_norm": 0.8866179585456848,
+      "learning_rate": 5.259259259259259e-05,
+      "loss": 0.3878,
+      "step": 52580
+    },
+    {
+      "epoch": 194.78,
+      "grad_norm": 0.8752300143241882,
+      "learning_rate": 5.222222222222223e-05,
+      "loss": 0.3951,
+      "step": 52590
+    },
+    {
+      "epoch": 194.81,
+      "grad_norm": 0.9218076467514038,
+      "learning_rate": 5.185185185185185e-05,
+      "loss": 0.3894,
+      "step": 52600
+    },
+    {
+      "epoch": 194.85,
+      "grad_norm": 0.960334300994873,
+      "learning_rate": 5.1481481481481486e-05,
+      "loss": 0.3855,
+      "step": 52610
+    },
+    {
+      "epoch": 194.89,
+      "grad_norm": 0.9215881824493408,
+      "learning_rate": 5.1111111111111115e-05,
+      "loss": 0.3969,
+      "step": 52620
+    },
+    {
+      "epoch": 194.93,
+      "grad_norm": 0.9425987005233765,
+      "learning_rate": 5.074074074074074e-05,
+      "loss": 0.389,
+      "step": 52630
+    },
+    {
+      "epoch": 194.96,
+      "grad_norm": 0.969688355922699,
+      "learning_rate": 5.037037037037037e-05,
+      "loss": 0.3959,
+      "step": 52640
+    },
+    {
+      "epoch": 195.0,
+      "grad_norm": 1.8812493085861206,
+      "learning_rate": 5e-05,
+      "loss": 0.3978,
+      "step": 52650
+    },
+    {
+      "epoch": 195.04,
+      "grad_norm": 0.8659247159957886,
+      "learning_rate": 4.962962962962963e-05,
+      "loss": 0.3513,
+      "step": 52660
+    },
+    {
+      "epoch": 195.07,
+      "grad_norm": 0.7738781571388245,
+      "learning_rate": 4.925925925925926e-05,
+      "loss": 0.3622,
+      "step": 52670
+    },
+    {
+      "epoch": 195.11,
+      "grad_norm": 0.8059188723564148,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 0.363,
+      "step": 52680
+    },
+    {
+      "epoch": 195.15,
+      "grad_norm": 0.7817255258560181,
+      "learning_rate": 4.851851851851852e-05,
+      "loss": 0.359,
+      "step": 52690
+    },
+    {
+      "epoch": 195.19,
+      "grad_norm": 0.8181976079940796,
+      "learning_rate": 4.814814814814815e-05,
+      "loss": 0.3601,
+      "step": 52700
+    },
+    {
+      "epoch": 195.22,
+      "grad_norm": 0.8579437732696533,
+      "learning_rate": 4.7777777777777784e-05,
+      "loss": 0.3626,
+      "step": 52710
+    },
+    {
+      "epoch": 195.26,
+      "grad_norm": 0.8376036882400513,
+      "learning_rate": 4.7407407407407407e-05,
+      "loss": 0.3516,
+      "step": 52720
+    },
+    {
+      "epoch": 195.3,
+      "grad_norm": 0.8274129033088684,
+      "learning_rate": 4.7037037037037036e-05,
+      "loss": 0.3735,
+      "step": 52730
+    },
+    {
+      "epoch": 195.33,
+      "grad_norm": 0.8801832795143127,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.3638,
+      "step": 52740
+    },
+    {
+      "epoch": 195.37,
+      "grad_norm": 0.8347715139389038,
+      "learning_rate": 4.6296296296296294e-05,
+      "loss": 0.3796,
+      "step": 52750
+    },
+    {
+      "epoch": 195.41,
+      "grad_norm": 0.8824911117553711,
+      "learning_rate": 4.592592592592593e-05,
+      "loss": 0.3704,
+      "step": 52760
+    },
+    {
+      "epoch": 195.44,
+      "grad_norm": 0.8434776067733765,
+      "learning_rate": 4.555555555555556e-05,
+      "loss": 0.3651,
+      "step": 52770
+    },
+    {
+      "epoch": 195.48,
+      "grad_norm": 0.8641818761825562,
+      "learning_rate": 4.518518518518518e-05,
+      "loss": 0.3677,
+      "step": 52780
+    },
+    {
+      "epoch": 195.52,
+      "grad_norm": 0.909089982509613,
+      "learning_rate": 4.481481481481482e-05,
+      "loss": 0.3743,
+      "step": 52790
+    },
+    {
+      "epoch": 195.56,
+      "grad_norm": 0.8464562892913818,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.3863,
+      "step": 52800
+    },
+    {
+      "epoch": 195.59,
+      "grad_norm": 0.9092419147491455,
+      "learning_rate": 4.407407407407407e-05,
+      "loss": 0.3722,
+      "step": 52810
+    },
+    {
+      "epoch": 195.63,
+      "grad_norm": 0.8889215588569641,
+      "learning_rate": 4.3703703703703705e-05,
+      "loss": 0.3844,
+      "step": 52820
+    },
+    {
+      "epoch": 195.67,
+      "grad_norm": 0.8936458230018616,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.3887,
+      "step": 52830
+    },
+    {
+      "epoch": 195.7,
+      "grad_norm": 0.9093278646469116,
+      "learning_rate": 4.296296296296296e-05,
+      "loss": 0.3724,
+      "step": 52840
+    },
+    {
+      "epoch": 195.74,
+      "grad_norm": 0.8708871006965637,
+      "learning_rate": 4.259259259259259e-05,
+      "loss": 0.3799,
+      "step": 52850
+    },
+    {
+      "epoch": 195.78,
+      "grad_norm": 0.9522839784622192,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 0.3799,
+      "step": 52860
+    },
+    {
+      "epoch": 195.81,
+      "grad_norm": 0.8670366406440735,
+      "learning_rate": 4.185185185185185e-05,
+      "loss": 0.3832,
+      "step": 52870
+    },
+    {
+      "epoch": 195.85,
+      "grad_norm": 0.8927628993988037,
+      "learning_rate": 4.148148148148148e-05,
+      "loss": 0.3794,
+      "step": 52880
+    },
+    {
+      "epoch": 195.89,
+      "grad_norm": 0.8557700514793396,
+      "learning_rate": 4.1111111111111116e-05,
+      "loss": 0.3816,
+      "step": 52890
+    },
+    {
+      "epoch": 195.93,
+      "grad_norm": 0.8442705869674683,
+      "learning_rate": 4.0740740740740745e-05,
+      "loss": 0.3874,
+      "step": 52900
+    },
+    {
+      "epoch": 195.96,
+      "grad_norm": 0.9126988649368286,
+      "learning_rate": 4.037037037037037e-05,
+      "loss": 0.3942,
+      "step": 52910
+    },
+    {
+      "epoch": 196.0,
+      "grad_norm": 1.4788200855255127,
+      "learning_rate": 4e-05,
+      "loss": 0.3881,
+      "step": 52920
+    },
+    {
+      "epoch": 196.04,
+      "grad_norm": 0.8152386546134949,
+      "learning_rate": 3.962962962962963e-05,
+      "loss": 0.3601,
+      "step": 52930
+    },
+    {
+      "epoch": 196.07,
+      "grad_norm": 0.7856703996658325,
+      "learning_rate": 3.925925925925926e-05,
+      "loss": 0.3639,
+      "step": 52940
+    },
+    {
+      "epoch": 196.11,
+      "grad_norm": 0.8162696957588196,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.3462,
+      "step": 52950
+    },
+    {
+      "epoch": 196.15,
+      "grad_norm": 0.7858487963676453,
+      "learning_rate": 3.851851851851852e-05,
+      "loss": 0.3605,
+      "step": 52960
+    },
+    {
+      "epoch": 196.19,
+      "grad_norm": 0.8109729290008545,
+      "learning_rate": 3.814814814814815e-05,
+      "loss": 0.3607,
+      "step": 52970
+    },
+    {
+      "epoch": 196.22,
+      "grad_norm": 0.8202926516532898,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 0.368,
+      "step": 52980
+    },
+    {
+      "epoch": 196.26,
+      "grad_norm": 0.7958359122276306,
+      "learning_rate": 3.7407407407407414e-05,
+      "loss": 0.3602,
+      "step": 52990
+    },
+    {
+      "epoch": 196.3,
+      "grad_norm": 0.7976880669593811,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.3616,
+      "step": 53000
+    },
+    {
+      "epoch": 196.33,
+      "grad_norm": 0.8153302073478699,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 0.3633,
+      "step": 53010
+    },
+    {
+      "epoch": 196.37,
+      "grad_norm": 0.8204575777053833,
+      "learning_rate": 3.62962962962963e-05,
+      "loss": 0.3551,
+      "step": 53020
+    },
+    {
+      "epoch": 196.41,
+      "grad_norm": 0.8472855687141418,
+      "learning_rate": 3.5925925925925924e-05,
+      "loss": 0.3595,
+      "step": 53030
+    },
+    {
+      "epoch": 196.44,
+      "grad_norm": 0.7960186004638672,
+      "learning_rate": 3.555555555555555e-05,
+      "loss": 0.3706,
+      "step": 53040
+    },
+    {
+      "epoch": 196.48,
+      "grad_norm": 0.8351572751998901,
+      "learning_rate": 3.518518518518519e-05,
+      "loss": 0.3705,
+      "step": 53050
+    },
+    {
+      "epoch": 196.52,
+      "grad_norm": 0.8574457764625549,
+      "learning_rate": 3.481481481481481e-05,
+      "loss": 0.3612,
+      "step": 53060
+    },
+    {
+      "epoch": 196.56,
+      "grad_norm": 0.8318818211555481,
+      "learning_rate": 3.444444444444445e-05,
+      "loss": 0.3752,
+      "step": 53070
+    },
+    {
+      "epoch": 196.59,
+      "grad_norm": 0.8394012451171875,
+      "learning_rate": 3.4074074074074077e-05,
+      "loss": 0.3672,
+      "step": 53080
+    },
+    {
+      "epoch": 196.63,
+      "grad_norm": 0.8550794124603271,
+      "learning_rate": 3.37037037037037e-05,
+      "loss": 0.3623,
+      "step": 53090
+    },
+    {
+      "epoch": 196.67,
+      "grad_norm": 0.8827661275863647,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.371,
+      "step": 53100
+    },
+    {
+      "epoch": 196.7,
+      "grad_norm": 0.8500509262084961,
+      "learning_rate": 3.2962962962962964e-05,
+      "loss": 0.3822,
+      "step": 53110
+    },
+    {
+      "epoch": 196.74,
+      "grad_norm": 0.8798251748085022,
+      "learning_rate": 3.259259259259259e-05,
+      "loss": 0.3791,
+      "step": 53120
+    },
+    {
+      "epoch": 196.78,
+      "grad_norm": 0.8392333388328552,
+      "learning_rate": 3.222222222222222e-05,
+      "loss": 0.368,
+      "step": 53130
+    },
+    {
+      "epoch": 196.81,
+      "grad_norm": 0.8432697653770447,
+      "learning_rate": 3.185185185185185e-05,
+      "loss": 0.3705,
+      "step": 53140
+    },
+    {
+      "epoch": 196.85,
+      "grad_norm": 0.8685855269432068,
+      "learning_rate": 3.148148148148148e-05,
+      "loss": 0.3806,
+      "step": 53150
+    },
+    {
+      "epoch": 196.89,
+      "grad_norm": 0.8553511500358582,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 0.3804,
+      "step": 53160
+    },
+    {
+      "epoch": 196.93,
+      "grad_norm": 0.8809592127799988,
+      "learning_rate": 3.0740740740740746e-05,
+      "loss": 0.379,
+      "step": 53170
+    },
+    {
+      "epoch": 196.96,
+      "grad_norm": 0.899813711643219,
+      "learning_rate": 3.037037037037037e-05,
+      "loss": 0.3809,
+      "step": 53180
+    },
+    {
+      "epoch": 197.0,
+      "grad_norm": 1.6756294965744019,
+      "learning_rate": 3e-05,
+      "loss": 0.3679,
+      "step": 53190
+    },
+    {
+      "epoch": 197.04,
+      "grad_norm": 0.8270112872123718,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.3528,
+      "step": 53200
+    },
+    {
+      "epoch": 197.07,
+      "grad_norm": 0.7883965969085693,
+      "learning_rate": 2.925925925925926e-05,
+      "loss": 0.3525,
+      "step": 53210
+    },
+    {
+      "epoch": 197.11,
+      "grad_norm": 0.8234068155288696,
+      "learning_rate": 2.8888888888888888e-05,
+      "loss": 0.3484,
+      "step": 53220
+    },
+    {
+      "epoch": 197.15,
+      "grad_norm": 0.7610680460929871,
+      "learning_rate": 2.851851851851852e-05,
+      "loss": 0.358,
+      "step": 53230
+    },
+    {
+      "epoch": 197.19,
+      "grad_norm": 0.8109197616577148,
+      "learning_rate": 2.814814814814815e-05,
+      "loss": 0.3538,
+      "step": 53240
+    },
+    {
+      "epoch": 197.22,
+      "grad_norm": 0.8277642130851746,
+      "learning_rate": 2.7777777777777776e-05,
+      "loss": 0.3544,
+      "step": 53250
+    },
+    {
+      "epoch": 197.26,
+      "grad_norm": 0.8044574856758118,
+      "learning_rate": 2.7407407407407408e-05,
+      "loss": 0.3626,
+      "step": 53260
+    },
+    {
+      "epoch": 197.3,
+      "grad_norm": 0.7997308373451233,
+      "learning_rate": 2.7037037037037037e-05,
+      "loss": 0.3608,
+      "step": 53270
+    },
+    {
+      "epoch": 197.33,
+      "grad_norm": 0.8243195414543152,
+      "learning_rate": 2.666666666666667e-05,
+      "loss": 0.3591,
+      "step": 53280
+    },
+    {
+      "epoch": 197.37,
+      "grad_norm": 0.8348344564437866,
+      "learning_rate": 2.6296296296296296e-05,
+      "loss": 0.3582,
+      "step": 53290
+    },
+    {
+      "epoch": 197.41,
+      "grad_norm": 0.8257560729980469,
+      "learning_rate": 2.5925925925925925e-05,
+      "loss": 0.3658,
+      "step": 53300
+    },
+    {
+      "epoch": 197.44,
+      "grad_norm": 0.8056928515434265,
+      "learning_rate": 2.5555555555555557e-05,
+      "loss": 0.3671,
+      "step": 53310
+    },
+    {
+      "epoch": 197.48,
+      "grad_norm": 0.8453297019004822,
+      "learning_rate": 2.5185185185185187e-05,
+      "loss": 0.3618,
+      "step": 53320
+    },
+    {
+      "epoch": 197.52,
+      "grad_norm": 0.8153665065765381,
+      "learning_rate": 2.4814814814814816e-05,
+      "loss": 0.3614,
+      "step": 53330
+    },
+    {
+      "epoch": 197.56,
+      "grad_norm": 0.8234355449676514,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 0.3604,
+      "step": 53340
+    },
+    {
+      "epoch": 197.59,
+      "grad_norm": 0.8666888475418091,
+      "learning_rate": 2.4074074074074074e-05,
+      "loss": 0.3733,
+      "step": 53350
+    },
+    {
+      "epoch": 197.63,
+      "grad_norm": 0.843656599521637,
+      "learning_rate": 2.3703703703703703e-05,
+      "loss": 0.3698,
+      "step": 53360
+    },
+    {
+      "epoch": 197.67,
+      "grad_norm": 0.8797212243080139,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 0.3657,
+      "step": 53370
+    },
+    {
+      "epoch": 197.7,
+      "grad_norm": 0.8750057816505432,
+      "learning_rate": 2.2962962962962965e-05,
+      "loss": 0.3562,
+      "step": 53380
+    },
+    {
+      "epoch": 197.74,
+      "grad_norm": 0.8547198176383972,
+      "learning_rate": 2.259259259259259e-05,
+      "loss": 0.3728,
+      "step": 53390
+    },
+    {
+      "epoch": 197.78,
+      "grad_norm": 0.8509737849235535,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.3611,
+      "step": 53400
+    },
+    {
+      "epoch": 197.81,
+      "grad_norm": 0.8830751776695251,
+      "learning_rate": 2.1851851851851852e-05,
+      "loss": 0.3614,
+      "step": 53410
+    },
+    {
+      "epoch": 197.85,
+      "grad_norm": 0.8684916496276855,
+      "learning_rate": 2.148148148148148e-05,
+      "loss": 0.3631,
+      "step": 53420
+    },
+    {
+      "epoch": 197.89,
+      "grad_norm": 0.8446648120880127,
+      "learning_rate": 2.111111111111111e-05,
+      "loss": 0.3635,
+      "step": 53430
+    },
+    {
+      "epoch": 197.93,
+      "grad_norm": 0.8646011352539062,
+      "learning_rate": 2.074074074074074e-05,
+      "loss": 0.3771,
+      "step": 53440
+    },
+    {
+      "epoch": 197.96,
+      "grad_norm": 0.8776282072067261,
+      "learning_rate": 2.0370370370370372e-05,
+      "loss": 0.3721,
+      "step": 53450
+    },
+    {
+      "epoch": 198.0,
+      "grad_norm": 1.4938170909881592,
+      "learning_rate": 2e-05,
+      "loss": 0.3705,
+      "step": 53460
+    },
+    {
+      "epoch": 198.04,
+      "grad_norm": 0.7895182371139526,
+      "learning_rate": 1.962962962962963e-05,
+      "loss": 0.3582,
+      "step": 53470
+    },
+    {
+      "epoch": 198.07,
+      "grad_norm": 0.7492555379867554,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 0.359,
+      "step": 53480
+    },
+    {
+      "epoch": 198.11,
+      "grad_norm": 0.7928748726844788,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 0.3429,
+      "step": 53490
+    },
+    {
+      "epoch": 198.15,
+      "grad_norm": 0.7984386682510376,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.3529,
+      "step": 53500
+    },
+    {
+      "epoch": 198.19,
+      "grad_norm": 0.8060497641563416,
+      "learning_rate": 1.814814814814815e-05,
+      "loss": 0.3463,
+      "step": 53510
+    },
+    {
+      "epoch": 198.22,
+      "grad_norm": 0.82963627576828,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.3447,
+      "step": 53520
+    },
+    {
+      "epoch": 198.26,
+      "grad_norm": 0.8006116151809692,
+      "learning_rate": 1.7407407407407406e-05,
+      "loss": 0.3503,
+      "step": 53530
+    },
+    {
+      "epoch": 198.3,
+      "grad_norm": 0.8024885654449463,
+      "learning_rate": 1.7037037037037038e-05,
+      "loss": 0.3566,
+      "step": 53540
+    },
+    {
+      "epoch": 198.33,
+      "grad_norm": 0.8089790344238281,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.3642,
+      "step": 53550
+    },
+    {
+      "epoch": 198.37,
+      "grad_norm": 0.7876638174057007,
+      "learning_rate": 1.6296296296296297e-05,
+      "loss": 0.3563,
+      "step": 53560
+    },
+    {
+      "epoch": 198.41,
+      "grad_norm": 0.8106840252876282,
+      "learning_rate": 1.5925925925925926e-05,
+      "loss": 0.3584,
+      "step": 53570
+    },
+    {
+      "epoch": 198.44,
+      "grad_norm": 0.7648861408233643,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 0.3572,
+      "step": 53580
+    },
+    {
+      "epoch": 198.48,
+      "grad_norm": 0.8709505796432495,
+      "learning_rate": 1.5185185185185186e-05,
+      "loss": 0.3603,
+      "step": 53590
+    },
+    {
+      "epoch": 198.52,
+      "grad_norm": 0.8210094571113586,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.352,
+      "step": 53600
+    },
+    {
+      "epoch": 198.56,
+      "grad_norm": 0.8383370041847229,
+      "learning_rate": 1.4444444444444444e-05,
+      "loss": 0.3518,
+      "step": 53610
+    },
+    {
+      "epoch": 198.59,
+      "grad_norm": 0.8203353881835938,
+      "learning_rate": 1.4074074074074075e-05,
+      "loss": 0.3533,
+      "step": 53620
+    },
+    {
+      "epoch": 198.63,
+      "grad_norm": 0.8129621744155884,
+      "learning_rate": 1.3703703703703704e-05,
+      "loss": 0.3533,
+      "step": 53630
+    },
+    {
+      "epoch": 198.67,
+      "grad_norm": 0.8171769380569458,
+      "learning_rate": 1.3333333333333335e-05,
+      "loss": 0.3596,
+      "step": 53640
+    },
+    {
+      "epoch": 198.7,
+      "grad_norm": 0.8063119053840637,
+      "learning_rate": 1.2962962962962962e-05,
+      "loss": 0.3718,
+      "step": 53650
+    },
+    {
+      "epoch": 198.74,
+      "grad_norm": 0.8649743795394897,
+      "learning_rate": 1.2592592592592593e-05,
+      "loss": 0.3586,
+      "step": 53660
+    },
+    {
+      "epoch": 198.78,
+      "grad_norm": 0.8293347954750061,
+      "learning_rate": 1.2222222222222222e-05,
+      "loss": 0.3656,
+      "step": 53670
+    },
+    {
+      "epoch": 198.81,
+      "grad_norm": 0.8939828872680664,
+      "learning_rate": 1.1851851851851852e-05,
+      "loss": 0.3542,
+      "step": 53680
+    },
+    {
+      "epoch": 198.85,
+      "grad_norm": 0.8524936437606812,
+      "learning_rate": 1.1481481481481482e-05,
+      "loss": 0.3573,
+      "step": 53690
+    },
+    {
+      "epoch": 198.89,
+      "grad_norm": 0.8106054067611694,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.3635,
+      "step": 53700
+    },
+    {
+      "epoch": 198.93,
+      "grad_norm": 0.7986220717430115,
+      "learning_rate": 1.074074074074074e-05,
+      "loss": 0.3605,
+      "step": 53710
+    },
+    {
+      "epoch": 198.96,
+      "grad_norm": 0.8215298056602478,
+      "learning_rate": 1.037037037037037e-05,
+      "loss": 0.3688,
+      "step": 53720
+    },
+    {
+      "epoch": 199.0,
+      "grad_norm": 1.5070199966430664,
+      "learning_rate": 1e-05,
+      "loss": 0.3647,
+      "step": 53730
+    },
+    {
+      "epoch": 199.04,
+      "grad_norm": 0.8176571130752563,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.3479,
+      "step": 53740
+    },
+    {
+      "epoch": 199.07,
+      "grad_norm": 0.7820712327957153,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.3476,
+      "step": 53750
+    },
+    {
+      "epoch": 199.11,
+      "grad_norm": 0.7757356762886047,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.3408,
+      "step": 53760
+    },
+    {
+      "epoch": 199.15,
+      "grad_norm": 0.8038589954376221,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 0.3503,
+      "step": 53770
+    },
+    {
+      "epoch": 199.19,
+      "grad_norm": 0.8047194480895996,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 0.3551,
+      "step": 53780
+    },
+    {
+      "epoch": 199.22,
+      "grad_norm": 0.7661307454109192,
+      "learning_rate": 7.777777777777777e-06,
+      "loss": 0.3497,
+      "step": 53790
+    },
+    {
+      "epoch": 199.26,
+      "grad_norm": 0.7506800889968872,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.3555,
+      "step": 53800
+    },
+    {
+      "epoch": 199.3,
+      "grad_norm": 0.7798722386360168,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 0.3453,
+      "step": 53810
+    },
+    {
+      "epoch": 199.33,
+      "grad_norm": 0.7841687202453613,
+      "learning_rate": 6.6666666666666675e-06,
+      "loss": 0.3571,
+      "step": 53820
+    },
+    {
+      "epoch": 199.37,
+      "grad_norm": 0.817516565322876,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 0.3488,
+      "step": 53830
+    },
+    {
+      "epoch": 199.41,
+      "grad_norm": 0.8079009652137756,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.3561,
+      "step": 53840
+    },
+    {
+      "epoch": 199.44,
+      "grad_norm": 0.8136209845542908,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.3479,
+      "step": 53850
+    },
+    {
+      "epoch": 199.48,
+      "grad_norm": 0.8167023658752441,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.3494,
+      "step": 53860
+    },
+    {
+      "epoch": 199.52,
+      "grad_norm": 0.7994495630264282,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 0.3535,
+      "step": 53870
+    },
+    {
+      "epoch": 199.56,
+      "grad_norm": 0.822665810585022,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.3514,
+      "step": 53880
+    },
+    {
+      "epoch": 199.59,
+      "grad_norm": 0.808405339717865,
+      "learning_rate": 4.074074074074074e-06,
+      "loss": 0.3577,
+      "step": 53890
+    },
+    {
+      "epoch": 199.63,
+      "grad_norm": 0.7848332524299622,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.3524,
+      "step": 53900
+    },
+    {
+      "epoch": 199.67,
+      "grad_norm": 0.8101984262466431,
+      "learning_rate": 3.3333333333333337e-06,
+      "loss": 0.3549,
+      "step": 53910
+    },
+    {
+      "epoch": 199.7,
+      "grad_norm": 0.7706851959228516,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 0.3526,
+      "step": 53920
+    },
+    {
+      "epoch": 199.74,
+      "grad_norm": 0.8184105753898621,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 0.3487,
+      "step": 53930
+    },
+    {
+      "epoch": 199.78,
+      "grad_norm": 0.8454097509384155,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.3529,
+      "step": 53940
+    },
+    {
+      "epoch": 199.81,
+      "grad_norm": 0.7991722226142883,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.3572,
+      "step": 53950
+    },
+    {
+      "epoch": 199.85,
+      "grad_norm": 0.8001537919044495,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 0.3614,
+      "step": 53960
+    },
+    {
+      "epoch": 199.89,
+      "grad_norm": 0.7946948409080505,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.3427,
+      "step": 53970
+    },
+    {
+      "epoch": 199.93,
+      "grad_norm": 0.8384917974472046,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 0.3575,
+      "step": 53980
+    },
+    {
+      "epoch": 199.96,
+      "grad_norm": 0.7977772355079651,
+      "learning_rate": 3.7037037037037036e-07,
+      "loss": 0.3586,
+      "step": 53990
+    },
+    {
+      "epoch": 200.0,
+      "grad_norm": 1.766687035560608,
+      "learning_rate": 0.0,
+      "loss": 0.3543,
+      "step": 54000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 54000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 200,
+  "save_steps": 500,
+  "total_flos": 6.764336504975697e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}