diff --git "a/Luminia-8B-RP/trainer_state.json" "b/Luminia-8B-RP/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/Luminia-8B-RP/trainer_state.json"
@@ -0,0 +1,7027 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 8739,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011442956860052637,
+      "grad_norm": 0.6716632843017578,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.5075,
+      "num_input_tokens_seen": 81920,
+      "step": 10
+    },
+    {
+      "epoch": 0.0022885913720105274,
+      "grad_norm": 0.8028005957603455,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.379,
+      "num_input_tokens_seen": 163840,
+      "step": 20
+    },
+    {
+      "epoch": 0.0034328870580157913,
+      "grad_norm": 0.6309388279914856,
+      "learning_rate": 1.5e-06,
+      "loss": 1.6386,
+      "num_input_tokens_seen": 245760,
+      "step": 30
+    },
+    {
+      "epoch": 0.004577182744021055,
+      "grad_norm": 2.590484380722046,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.5525,
+      "num_input_tokens_seen": 327680,
+      "step": 40
+    },
+    {
+      "epoch": 0.005721478430026319,
+      "grad_norm": 0.8970260620117188,
+      "learning_rate": 2.5e-06,
+      "loss": 1.651,
+      "num_input_tokens_seen": 409600,
+      "step": 50
+    },
+    {
+      "epoch": 0.006865774116031583,
+      "grad_norm": 0.7371882796287537,
+      "learning_rate": 3e-06,
+      "loss": 1.5067,
+      "num_input_tokens_seen": 491520,
+      "step": 60
+    },
+    {
+      "epoch": 0.008010069802036847,
+      "grad_norm": 2.1576790809631348,
+      "learning_rate": 3.5000000000000004e-06,
+      "loss": 1.3367,
+      "num_input_tokens_seen": 573440,
+      "step": 70
+    },
+    {
+      "epoch": 0.00915436548804211,
+      "grad_norm": 0.8213880658149719,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.6071,
+      "num_input_tokens_seen": 655360,
+      "step": 80
+    },
+    {
+      "epoch": 0.010298661174047374,
+      "grad_norm": 1.3453713655471802,
+      "learning_rate": 4.5e-06,
+      "loss": 1.522,
+      "num_input_tokens_seen": 737280,
+      "step": 90
+    },
+    {
+      "epoch": 0.011442956860052637,
+      "grad_norm": 1.1660873889923096,
+      "learning_rate": 5e-06,
+      "loss": 1.5853,
+      "num_input_tokens_seen": 819200,
+      "step": 100
+    },
+    {
+      "epoch": 0.012587252546057902,
+      "grad_norm": 1.4795037508010864,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 1.4081,
+      "num_input_tokens_seen": 901120,
+      "step": 110
+    },
+    {
+      "epoch": 0.013731548232063165,
+      "grad_norm": 2.582639455795288,
+      "learning_rate": 6e-06,
+      "loss": 1.5405,
+      "num_input_tokens_seen": 983040,
+      "step": 120
+    },
+    {
+      "epoch": 0.014875843918068428,
+      "grad_norm": 1.1922495365142822,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 1.3632,
+      "num_input_tokens_seen": 1064960,
+      "step": 130
+    },
+    {
+      "epoch": 0.016020139604073693,
+      "grad_norm": 1.249202847480774,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 1.128,
+      "num_input_tokens_seen": 1146880,
+      "step": 140
+    },
+    {
+      "epoch": 0.017164435290078956,
+      "grad_norm": 2.9147748947143555,
+      "learning_rate": 7.5e-06,
+      "loss": 1.3543,
+      "num_input_tokens_seen": 1228800,
+      "step": 150
+    },
+    {
+      "epoch": 0.01830873097608422,
+      "grad_norm": 1.3228706121444702,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.3541,
+      "num_input_tokens_seen": 1310720,
+      "step": 160
+    },
+    {
+      "epoch": 0.019453026662089482,
+      "grad_norm": 1.2447155714035034,
+      "learning_rate": 8.500000000000002e-06,
+      "loss": 1.4476,
+      "num_input_tokens_seen": 1392640,
+      "step": 170
+    },
+    {
+      "epoch": 0.02059732234809475,
+      "grad_norm": 1.2924906015396118,
+      "learning_rate": 9e-06,
+      "loss": 1.212,
+      "num_input_tokens_seen": 1474560,
+      "step": 180
+    },
+    {
+      "epoch": 0.021741618034100012,
+      "grad_norm": 1.352099061012268,
+      "learning_rate": 9.5e-06,
+      "loss": 1.3312,
+      "num_input_tokens_seen": 1556480,
+      "step": 190
+    },
+    {
+      "epoch": 0.022885913720105275,
+      "grad_norm": 2.618453025817871,
+      "learning_rate": 1e-05,
+      "loss": 1.3266,
+      "num_input_tokens_seen": 1638400,
+      "step": 200
+    },
+    {
+      "epoch": 0.024030209406110538,
+      "grad_norm": 1.2627308368682861,
+      "learning_rate": 1.05e-05,
+      "loss": 1.3106,
+      "num_input_tokens_seen": 1720320,
+      "step": 210
+    },
+    {
+      "epoch": 0.025174505092115804,
+      "grad_norm": 0.9320095181465149,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 1.4398,
+      "num_input_tokens_seen": 1802240,
+      "step": 220
+    },
+    {
+      "epoch": 0.026318800778121067,
+      "grad_norm": 1.7634745836257935,
+      "learning_rate": 1.1500000000000002e-05,
+      "loss": 1.1946,
+      "num_input_tokens_seen": 1884160,
+      "step": 230
+    },
+    {
+      "epoch": 0.02746309646412633,
+      "grad_norm": 1.1597224473953247,
+      "learning_rate": 1.2e-05,
+      "loss": 1.2912,
+      "num_input_tokens_seen": 1966080,
+      "step": 240
+    },
+    {
+      "epoch": 0.028607392150131594,
+      "grad_norm": 1.1823091506958008,
+      "learning_rate": 1.25e-05,
+      "loss": 1.5374,
+      "num_input_tokens_seen": 2048000,
+      "step": 250
+    },
+    {
+      "epoch": 0.029751687836136857,
+      "grad_norm": 0.88273024559021,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 1.4559,
+      "num_input_tokens_seen": 2129920,
+      "step": 260
+    },
+    {
+      "epoch": 0.030895983522142123,
+      "grad_norm": 1.2335752248764038,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 1.4361,
+      "num_input_tokens_seen": 2211840,
+      "step": 270
+    },
+    {
+      "epoch": 0.032040279208147386,
+      "grad_norm": 0.8680986166000366,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 1.7506,
+      "num_input_tokens_seen": 2293760,
+      "step": 280
+    },
+    {
+      "epoch": 0.03318457489415265,
+      "grad_norm": 0.8861328363418579,
+      "learning_rate": 1.45e-05,
+      "loss": 1.5946,
+      "num_input_tokens_seen": 2375680,
+      "step": 290
+    },
+    {
+      "epoch": 0.03432887058015791,
+      "grad_norm": 0.8510925769805908,
+      "learning_rate": 1.5e-05,
+      "loss": 1.386,
+      "num_input_tokens_seen": 2457600,
+      "step": 300
+    },
+    {
+      "epoch": 0.035473166266163175,
+      "grad_norm": 0.961478054523468,
+      "learning_rate": 1.55e-05,
+      "loss": 1.1091,
+      "num_input_tokens_seen": 2539520,
+      "step": 310
+    },
+    {
+      "epoch": 0.03661746195216844,
+      "grad_norm": 0.8009458780288696,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.3287,
+      "num_input_tokens_seen": 2621440,
+      "step": 320
+    },
+    {
+      "epoch": 0.0377617576381737,
+      "grad_norm": 1.1145437955856323,
+      "learning_rate": 1.65e-05,
+      "loss": 1.2537,
+      "num_input_tokens_seen": 2703360,
+      "step": 330
+    },
+    {
+      "epoch": 0.038906053324178964,
+      "grad_norm": 0.7063050270080566,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 1.4166,
+      "num_input_tokens_seen": 2785280,
+      "step": 340
+    },
+    {
+      "epoch": 0.040050349010184234,
+      "grad_norm": 0.8494574427604675,
+      "learning_rate": 1.75e-05,
+      "loss": 1.298,
+      "num_input_tokens_seen": 2867200,
+      "step": 350
+    },
+    {
+      "epoch": 0.0411946446961895,
+      "grad_norm": 0.8032189011573792,
+      "learning_rate": 1.8e-05,
+      "loss": 1.4127,
+      "num_input_tokens_seen": 2949120,
+      "step": 360
+    },
+    {
+      "epoch": 0.04233894038219476,
+      "grad_norm": 0.6479071974754333,
+      "learning_rate": 1.85e-05,
+      "loss": 1.1462,
+      "num_input_tokens_seen": 3031040,
+      "step": 370
+    },
+    {
+      "epoch": 0.043483236068200024,
+      "grad_norm": 1.165413737297058,
+      "learning_rate": 1.9e-05,
+      "loss": 1.5957,
+      "num_input_tokens_seen": 3112960,
+      "step": 380
+    },
+    {
+      "epoch": 0.04462753175420529,
+      "grad_norm": 0.8543304800987244,
+      "learning_rate": 1.9500000000000003e-05,
+      "loss": 1.3672,
+      "num_input_tokens_seen": 3194880,
+      "step": 390
+    },
+    {
+      "epoch": 0.04577182744021055,
+      "grad_norm": 0.8080208897590637,
+      "learning_rate": 2e-05,
+      "loss": 1.2616,
+      "num_input_tokens_seen": 3276800,
+      "step": 400
+    },
+    {
+      "epoch": 0.04691612312621581,
+      "grad_norm": 0.810095489025116,
+      "learning_rate": 2.05e-05,
+      "loss": 1.1038,
+      "num_input_tokens_seen": 3358720,
+      "step": 410
+    },
+    {
+      "epoch": 0.048060418812221076,
+      "grad_norm": 0.6553860902786255,
+      "learning_rate": 2.1e-05,
+      "loss": 1.3898,
+      "num_input_tokens_seen": 3440640,
+      "step": 420
+    },
+    {
+      "epoch": 0.04920471449822634,
+      "grad_norm": 3.40867018699646,
+      "learning_rate": 2.15e-05,
+      "loss": 1.2665,
+      "num_input_tokens_seen": 3522560,
+      "step": 430
+    },
+    {
+      "epoch": 0.05034901018423161,
+      "grad_norm": 0.6455990076065063,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.188,
+      "num_input_tokens_seen": 3604480,
+      "step": 440
+    },
+    {
+      "epoch": 0.05149330587023687,
+      "grad_norm": 0.5759983658790588,
+      "learning_rate": 2.25e-05,
+      "loss": 1.4607,
+      "num_input_tokens_seen": 3686400,
+      "step": 450
+    },
+    {
+      "epoch": 0.052637601556242135,
+      "grad_norm": 1.409822940826416,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 1.1622,
+      "num_input_tokens_seen": 3768320,
+      "step": 460
+    },
+    {
+      "epoch": 0.0537818972422474,
+      "grad_norm": 0.6812942028045654,
+      "learning_rate": 2.35e-05,
+      "loss": 1.3681,
+      "num_input_tokens_seen": 3850240,
+      "step": 470
+    },
+    {
+      "epoch": 0.05492619292825266,
+      "grad_norm": 1.1562724113464355,
+      "learning_rate": 2.4e-05,
+      "loss": 1.341,
+      "num_input_tokens_seen": 3932160,
+      "step": 480
+    },
+    {
+      "epoch": 0.056070488614257924,
+      "grad_norm": 1.1834927797317505,
+      "learning_rate": 2.45e-05,
+      "loss": 1.0104,
+      "num_input_tokens_seen": 4014080,
+      "step": 490
+    },
+    {
+      "epoch": 0.05721478430026319,
+      "grad_norm": 0.6955065727233887,
+      "learning_rate": 2.5e-05,
+      "loss": 1.4473,
+      "num_input_tokens_seen": 4096000,
+      "step": 500
+    },
+    {
+      "epoch": 0.05835907998626845,
+      "grad_norm": 5.183798313140869,
+      "learning_rate": 2.5500000000000003e-05,
+      "loss": 1.5255,
+      "num_input_tokens_seen": 4177920,
+      "step": 510
+    },
+    {
+      "epoch": 0.05950337567227371,
+      "grad_norm": 0.6789785027503967,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 1.3778,
+      "num_input_tokens_seen": 4259840,
+      "step": 520
+    },
+    {
+      "epoch": 0.060647671358278976,
+      "grad_norm": 0.5913729667663574,
+      "learning_rate": 2.6500000000000004e-05,
+      "loss": 1.1477,
+      "num_input_tokens_seen": 4341760,
+      "step": 530
+    },
+    {
+      "epoch": 0.061791967044284246,
+      "grad_norm": 1.2051905393600464,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 1.4827,
+      "num_input_tokens_seen": 4423680,
+      "step": 540
+    },
+    {
+      "epoch": 0.0629362627302895,
+      "grad_norm": 0.5383505821228027,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 1.635,
+      "num_input_tokens_seen": 4505600,
+      "step": 550
+    },
+    {
+      "epoch": 0.06408055841629477,
+      "grad_norm": 0.6724018454551697,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.2354,
+      "num_input_tokens_seen": 4587520,
+      "step": 560
+    },
+    {
+      "epoch": 0.06522485410230003,
+      "grad_norm": 0.8346174359321594,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 1.3294,
+      "num_input_tokens_seen": 4669440,
+      "step": 570
+    },
+    {
+      "epoch": 0.0663691497883053,
+      "grad_norm": 0.624640703201294,
+      "learning_rate": 2.9e-05,
+      "loss": 1.3841,
+      "num_input_tokens_seen": 4751360,
+      "step": 580
+    },
+    {
+      "epoch": 0.06751344547431057,
+      "grad_norm": 0.5437624454498291,
+      "learning_rate": 2.95e-05,
+      "loss": 1.3379,
+      "num_input_tokens_seen": 4833280,
+      "step": 590
+    },
+    {
+      "epoch": 0.06865774116031582,
+      "grad_norm": 0.5260199308395386,
+      "learning_rate": 3e-05,
+      "loss": 1.187,
+      "num_input_tokens_seen": 4915200,
+      "step": 600
+    },
+    {
+      "epoch": 0.0698020368463211,
+      "grad_norm": 0.6440827250480652,
+      "learning_rate": 3.05e-05,
+      "loss": 1.5078,
+      "num_input_tokens_seen": 4997120,
+      "step": 610
+    },
+    {
+      "epoch": 0.07094633253232635,
+      "grad_norm": 0.8612675666809082,
+      "learning_rate": 3.1e-05,
+      "loss": 1.4624,
+      "num_input_tokens_seen": 5079040,
+      "step": 620
+    },
+    {
+      "epoch": 0.07209062821833162,
+      "grad_norm": 0.5070405602455139,
+      "learning_rate": 3.15e-05,
+      "loss": 1.2244,
+      "num_input_tokens_seen": 5160960,
+      "step": 630
+    },
+    {
+      "epoch": 0.07323492390433688,
+      "grad_norm": 0.6587413549423218,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.4688,
+      "num_input_tokens_seen": 5242880,
+      "step": 640
+    },
+    {
+      "epoch": 0.07437921959034215,
+      "grad_norm": 0.6212195754051208,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 1.1677,
+      "num_input_tokens_seen": 5324800,
+      "step": 650
+    },
+    {
+      "epoch": 0.0755235152763474,
+      "grad_norm": 1.031207799911499,
+      "learning_rate": 3.3e-05,
+      "loss": 1.2562,
+      "num_input_tokens_seen": 5406720,
+      "step": 660
+    },
+    {
+      "epoch": 0.07666781096235267,
+      "grad_norm": 0.5637765526771545,
+      "learning_rate": 3.35e-05,
+      "loss": 1.3111,
+      "num_input_tokens_seen": 5488640,
+      "step": 670
+    },
+    {
+      "epoch": 0.07781210664835793,
+      "grad_norm": 1.6246665716171265,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 1.2953,
+      "num_input_tokens_seen": 5570560,
+      "step": 680
+    },
+    {
+      "epoch": 0.0789564023343632,
+      "grad_norm": 0.5538053512573242,
+      "learning_rate": 3.45e-05,
+      "loss": 1.2493,
+      "num_input_tokens_seen": 5652480,
+      "step": 690
+    },
+    {
+      "epoch": 0.08010069802036847,
+      "grad_norm": 0.566648542881012,
+      "learning_rate": 3.5e-05,
+      "loss": 1.3966,
+      "num_input_tokens_seen": 5734400,
+      "step": 700
+    },
+    {
+      "epoch": 0.08124499370637372,
+      "grad_norm": 0.5915348529815674,
+      "learning_rate": 3.55e-05,
+      "loss": 1.1898,
+      "num_input_tokens_seen": 5816320,
+      "step": 710
+    },
+    {
+      "epoch": 0.082389289392379,
+      "grad_norm": 0.5097474455833435,
+      "learning_rate": 3.6e-05,
+      "loss": 1.3558,
+      "num_input_tokens_seen": 5898240,
+      "step": 720
+    },
+    {
+      "epoch": 0.08353358507838425,
+      "grad_norm": 0.6351694464683533,
+      "learning_rate": 3.65e-05,
+      "loss": 1.6577,
+      "num_input_tokens_seen": 5980160,
+      "step": 730
+    },
+    {
+      "epoch": 0.08467788076438952,
+      "grad_norm": 0.5961860418319702,
+      "learning_rate": 3.7e-05,
+      "loss": 1.2571,
+      "num_input_tokens_seen": 6062080,
+      "step": 740
+    },
+    {
+      "epoch": 0.08582217645039478,
+      "grad_norm": 0.7016868591308594,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 1.3735,
+      "num_input_tokens_seen": 6144000,
+      "step": 750
+    },
+    {
+      "epoch": 0.08696647213640005,
+      "grad_norm": 0.5575875043869019,
+      "learning_rate": 3.8e-05,
+      "loss": 1.2731,
+      "num_input_tokens_seen": 6225920,
+      "step": 760
+    },
+    {
+      "epoch": 0.0881107678224053,
+      "grad_norm": 1.047059416770935,
+      "learning_rate": 3.85e-05,
+      "loss": 1.543,
+      "num_input_tokens_seen": 6307840,
+      "step": 770
+    },
+    {
+      "epoch": 0.08925506350841057,
+      "grad_norm": 0.8251460790634155,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 1.1869,
+      "num_input_tokens_seen": 6389760,
+      "step": 780
+    },
+    {
+      "epoch": 0.09039935919441584,
+      "grad_norm": 0.6270453333854675,
+      "learning_rate": 3.9500000000000005e-05,
+      "loss": 1.6793,
+      "num_input_tokens_seen": 6471680,
+      "step": 790
+    },
+    {
+      "epoch": 0.0915436548804211,
+      "grad_norm": 0.6033393740653992,
+      "learning_rate": 4e-05,
+      "loss": 1.7236,
+      "num_input_tokens_seen": 6553600,
+      "step": 800
+    },
+    {
+      "epoch": 0.09268795056642637,
+      "grad_norm": 1.2078917026519775,
+      "learning_rate": 4.05e-05,
+      "loss": 1.1466,
+      "num_input_tokens_seen": 6635520,
+      "step": 810
+    },
+    {
+      "epoch": 0.09383224625243163,
+      "grad_norm": 0.6712398529052734,
+      "learning_rate": 4.1e-05,
+      "loss": 1.308,
+      "num_input_tokens_seen": 6717440,
+      "step": 820
+    },
+    {
+      "epoch": 0.0949765419384369,
+      "grad_norm": 0.5450591444969177,
+      "learning_rate": 4.15e-05,
+      "loss": 1.1448,
+      "num_input_tokens_seen": 6799360,
+      "step": 830
+    },
+    {
+      "epoch": 0.09612083762444215,
+      "grad_norm": 0.6224645376205444,
+      "learning_rate": 4.2e-05,
+      "loss": 1.1618,
+      "num_input_tokens_seen": 6881280,
+      "step": 840
+    },
+    {
+      "epoch": 0.09726513331044742,
+      "grad_norm": 0.8741360306739807,
+      "learning_rate": 4.25e-05,
+      "loss": 1.3667,
+      "num_input_tokens_seen": 6963200,
+      "step": 850
+    },
+    {
+      "epoch": 0.09840942899645268,
+      "grad_norm": 0.6075563430786133,
+      "learning_rate": 4.3e-05,
+      "loss": 1.3752,
+      "num_input_tokens_seen": 7045120,
+      "step": 860
+    },
+    {
+      "epoch": 0.09955372468245795,
+      "grad_norm": 0.6345754265785217,
+      "learning_rate": 4.35e-05,
+      "loss": 1.3973,
+      "num_input_tokens_seen": 7127040,
+      "step": 870
+    },
+    {
+      "epoch": 0.10069802036846322,
+      "grad_norm": 0.7039983868598938,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.3936,
+      "num_input_tokens_seen": 7208960,
+      "step": 880
+    },
+    {
+      "epoch": 0.10184231605446847,
+      "grad_norm": 0.6462425589561462,
+      "learning_rate": 4.4500000000000004e-05,
+      "loss": 1.2224,
+      "num_input_tokens_seen": 7290880,
+      "step": 890
+    },
+    {
+      "epoch": 0.10298661174047374,
+      "grad_norm": 0.6138895153999329,
+      "learning_rate": 4.5e-05,
+      "loss": 1.0594,
+      "num_input_tokens_seen": 7372800,
+      "step": 900
+    },
+    {
+      "epoch": 0.104130907426479,
+      "grad_norm": 0.5026350021362305,
+      "learning_rate": 4.55e-05,
+      "loss": 1.122,
+      "num_input_tokens_seen": 7454720,
+      "step": 910
+    },
+    {
+      "epoch": 0.10527520311248427,
+      "grad_norm": 0.6174338459968567,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.2711,
+      "num_input_tokens_seen": 7536640,
+      "step": 920
+    },
+    {
+      "epoch": 0.10641949879848953,
+      "grad_norm": 0.736929178237915,
+      "learning_rate": 4.6500000000000005e-05,
+      "loss": 1.3302,
+      "num_input_tokens_seen": 7618560,
+      "step": 930
+    },
+    {
+      "epoch": 0.1075637944844948,
+      "grad_norm": 0.6089776158332825,
+      "learning_rate": 4.7e-05,
+      "loss": 1.2869,
+      "num_input_tokens_seen": 7700480,
+      "step": 940
+    },
+    {
+      "epoch": 0.10870809017050005,
+      "grad_norm": 0.7438560724258423,
+      "learning_rate": 4.75e-05,
+      "loss": 1.2435,
+      "num_input_tokens_seen": 7782400,
+      "step": 950
+    },
+    {
+      "epoch": 0.10985238585650532,
+      "grad_norm": 0.5268089175224304,
+      "learning_rate": 4.8e-05,
+      "loss": 1.0464,
+      "num_input_tokens_seen": 7864320,
+      "step": 960
+    },
+    {
+      "epoch": 0.11099668154251058,
+      "grad_norm": 0.6583835482597351,
+      "learning_rate": 4.85e-05,
+      "loss": 1.2715,
+      "num_input_tokens_seen": 7946240,
+      "step": 970
+    },
+    {
+      "epoch": 0.11214097722851585,
+      "grad_norm": 1.4557218551635742,
+      "learning_rate": 4.9e-05,
+      "loss": 1.4746,
+      "num_input_tokens_seen": 8028160,
+      "step": 980
+    },
+    {
+      "epoch": 0.11328527291452112,
+      "grad_norm": 1.0798985958099365,
+      "learning_rate": 4.9500000000000004e-05,
+      "loss": 1.1464,
+      "num_input_tokens_seen": 8110080,
+      "step": 990
+    },
+    {
+      "epoch": 0.11442956860052637,
+      "grad_norm": 0.5683252811431885,
+      "learning_rate": 5e-05,
+      "loss": 1.3901,
+      "num_input_tokens_seen": 8192000,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11557386428653164,
+      "grad_norm": 0.6159707903862,
+      "learning_rate": 4.999979401316311e-05,
+      "loss": 1.3986,
+      "num_input_tokens_seen": 8273920,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1167181599725369,
+      "grad_norm": 0.6029934883117676,
+      "learning_rate": 4.999917605604688e-05,
+      "loss": 1.3841,
+      "num_input_tokens_seen": 8355840,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11786245565854217,
+      "grad_norm": 0.6061372756958008,
+      "learning_rate": 4.999814613883459e-05,
+      "loss": 1.6034,
+      "num_input_tokens_seen": 8437760,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11900675134454743,
+      "grad_norm": 0.5895953178405762,
+      "learning_rate": 4.9996704278498185e-05,
+      "loss": 1.5685,
+      "num_input_tokens_seen": 8519680,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1201510470305527,
+      "grad_norm": 0.6715599894523621,
+      "learning_rate": 4.9994850498798026e-05,
+      "loss": 1.4407,
+      "num_input_tokens_seen": 8601600,
+      "step": 1050
+    },
+    {
+      "epoch": 0.12129534271655795,
+      "grad_norm": 0.6397563815116882,
+      "learning_rate": 4.999258483028243e-05,
+      "loss": 1.2713,
+      "num_input_tokens_seen": 8683520,
+      "step": 1060
+    },
+    {
+      "epoch": 0.12243963840256322,
+      "grad_norm": 0.5121895670890808,
+      "learning_rate": 4.9989907310287243e-05,
+      "loss": 1.2696,
+      "num_input_tokens_seen": 8765440,
+      "step": 1070
+    },
+    {
+      "epoch": 0.12358393408856849,
+      "grad_norm": 0.533311665058136,
+      "learning_rate": 4.998681798293516e-05,
+      "loss": 1.1279,
+      "num_input_tokens_seen": 8847360,
+      "step": 1080
+    },
+    {
+      "epoch": 0.12472822977457375,
+      "grad_norm": 0.4866422414779663,
+      "learning_rate": 4.998331689913506e-05,
+      "loss": 1.3104,
+      "num_input_tokens_seen": 8929280,
+      "step": 1090
+    },
+    {
+      "epoch": 0.125872525460579,
+      "grad_norm": 0.6534491181373596,
+      "learning_rate": 4.9979404116581104e-05,
+      "loss": 1.2877,
+      "num_input_tokens_seen": 9011200,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12701682114658427,
+      "grad_norm": 0.5954148173332214,
+      "learning_rate": 4.9975079699751825e-05,
+      "loss": 1.2768,
+      "num_input_tokens_seen": 9093120,
+      "step": 1110
+    },
+    {
+      "epoch": 0.12816111683258954,
+      "grad_norm": 0.49169713258743286,
+      "learning_rate": 4.997034371990907e-05,
+      "loss": 1.5502,
+      "num_input_tokens_seen": 9175040,
+      "step": 1120
+    },
+    {
+      "epoch": 0.12930541251859481,
+      "grad_norm": 0.6432734131813049,
+      "learning_rate": 4.99651962550968e-05,
+      "loss": 1.3732,
+      "num_input_tokens_seen": 9256960,
+      "step": 1130
+    },
+    {
+      "epoch": 0.13044970820460006,
+      "grad_norm": 0.6305902600288391,
+      "learning_rate": 4.9959637390139814e-05,
+      "loss": 1.1296,
+      "num_input_tokens_seen": 9338880,
+      "step": 1140
+    },
+    {
+      "epoch": 0.13159400389060533,
+      "grad_norm": 0.5763765573501587,
+      "learning_rate": 4.995366721664234e-05,
+      "loss": 1.1064,
+      "num_input_tokens_seen": 9420800,
+      "step": 1150
+    },
+    {
+      "epoch": 0.1327382995766106,
+      "grad_norm": 0.5058463215827942,
+      "learning_rate": 4.9947285832986553e-05,
+      "loss": 1.3228,
+      "num_input_tokens_seen": 9502720,
+      "step": 1160
+    },
+    {
+      "epoch": 0.13388259526261587,
+      "grad_norm": 0.6652387380599976,
+      "learning_rate": 4.994049334433095e-05,
+      "loss": 1.3161,
+      "num_input_tokens_seen": 9584640,
+      "step": 1170
+    },
+    {
+      "epoch": 0.13502689094862114,
+      "grad_norm": 0.5453552007675171,
+      "learning_rate": 4.9933289862608584e-05,
+      "loss": 1.4789,
+      "num_input_tokens_seen": 9666560,
+      "step": 1180
+    },
+    {
+      "epoch": 0.13617118663462638,
+      "grad_norm": 0.5469970107078552,
+      "learning_rate": 4.992567550652525e-05,
+      "loss": 1.3033,
+      "num_input_tokens_seen": 9748480,
+      "step": 1190
+    },
+    {
+      "epoch": 0.13731548232063165,
+      "grad_norm": 0.7196422219276428,
+      "learning_rate": 4.9917650401557505e-05,
+      "loss": 1.1661,
+      "num_input_tokens_seen": 9830400,
+      "step": 1200
+    },
+    {
+      "epoch": 0.13845977800663692,
+      "grad_norm": 0.7473416328430176,
+      "learning_rate": 4.990921467995064e-05,
+      "loss": 1.3148,
+      "num_input_tokens_seen": 9912320,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1396040736926422,
+      "grad_norm": 0.6073607206344604,
+      "learning_rate": 4.9900368480716466e-05,
+      "loss": 1.186,
+      "num_input_tokens_seen": 9994240,
+      "step": 1220
+    },
+    {
+      "epoch": 0.14074836937864743,
+      "grad_norm": 0.49970775842666626,
+      "learning_rate": 4.9891111949631023e-05,
+      "loss": 1.0854,
+      "num_input_tokens_seen": 10076160,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1418926650646527,
+      "grad_norm": 0.5494070053100586,
+      "learning_rate": 4.988144523923221e-05,
+      "loss": 1.2775,
+      "num_input_tokens_seen": 10158080,
+      "step": 1240
+    },
+    {
+      "epoch": 0.14303696075065797,
+      "grad_norm": 0.49916988611221313,
+      "learning_rate": 4.987136850881721e-05,
+      "loss": 1.3006,
+      "num_input_tokens_seen": 10240000,
+      "step": 1250
+    },
+    {
+      "epoch": 0.14418125643666324,
+      "grad_norm": 0.46677911281585693,
+      "learning_rate": 4.986088192443995e-05,
+      "loss": 1.0796,
+      "num_input_tokens_seen": 10321920,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1453255521226685,
+      "grad_norm": 0.520656943321228,
+      "learning_rate": 4.9849985658908296e-05,
+      "loss": 1.2881,
+      "num_input_tokens_seen": 10403840,
+      "step": 1270
+    },
+    {
+      "epoch": 0.14646984780867375,
+      "grad_norm": 0.5875335335731506,
+      "learning_rate": 4.9838679891781214e-05,
+      "loss": 1.2889,
+      "num_input_tokens_seen": 10485760,
+      "step": 1280
+    },
+    {
+      "epoch": 0.14761414349467902,
+      "grad_norm": 0.431538850069046,
+      "learning_rate": 4.982696480936586e-05,
+      "loss": 1.0391,
+      "num_input_tokens_seen": 10567680,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1487584391806843,
+      "grad_norm": 0.7769544124603271,
+      "learning_rate": 4.981484060471444e-05,
+      "loss": 1.0809,
+      "num_input_tokens_seen": 10649600,
+      "step": 1300
+    },
+    {
+      "epoch": 0.14990273486668956,
+      "grad_norm": 0.5716103315353394,
+      "learning_rate": 4.9802307477621084e-05,
+      "loss": 1.1128,
+      "num_input_tokens_seen": 10731520,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1510470305526948,
+      "grad_norm": 0.546258807182312,
+      "learning_rate": 4.978936563461854e-05,
+      "loss": 1.2888,
+      "num_input_tokens_seen": 10813440,
+      "step": 1320
+    },
+    {
+      "epoch": 0.15219132623870008,
+      "grad_norm": 0.9458239674568176,
+      "learning_rate": 4.9776015288974736e-05,
+      "loss": 1.3032,
+      "num_input_tokens_seen": 10895360,
+      "step": 1330
+    },
+    {
+      "epoch": 0.15333562192470535,
+      "grad_norm": 0.43678519129753113,
+      "learning_rate": 4.976225666068932e-05,
+      "loss": 0.9843,
+      "num_input_tokens_seen": 10977280,
+      "step": 1340
+    },
+    {
+      "epoch": 0.15447991761071062,
+      "grad_norm": 0.5537676811218262,
+      "learning_rate": 4.9748089976489996e-05,
+      "loss": 1.3003,
+      "num_input_tokens_seen": 11059200,
+      "step": 1350
+    },
+    {
+      "epoch": 0.15562421329671586,
+      "grad_norm": 0.5468762516975403,
+      "learning_rate": 4.9733515469828795e-05,
+      "loss": 1.4718,
+      "num_input_tokens_seen": 11141120,
+      "step": 1360
+    },
+    {
+      "epoch": 0.15676850898272113,
+      "grad_norm": 0.5907567143440247,
+      "learning_rate": 4.971853338087825e-05,
+      "loss": 1.3703,
+      "num_input_tokens_seen": 11223040,
+      "step": 1370
+    },
+    {
+      "epoch": 0.1579128046687264,
+      "grad_norm": 0.6227702498435974,
+      "learning_rate": 4.97031439565274e-05,
+      "loss": 1.0876,
+      "num_input_tokens_seen": 11304960,
+      "step": 1380
+    },
+    {
+      "epoch": 0.15905710035473167,
+      "grad_norm": 0.5486903786659241,
+      "learning_rate": 4.9687347450377755e-05,
+      "loss": 1.2873,
+      "num_input_tokens_seen": 11386880,
+      "step": 1390
+    },
+    {
+      "epoch": 0.16020139604073694,
+      "grad_norm": 1.0276037454605103,
+      "learning_rate": 4.9671144122739106e-05,
+      "loss": 1.4622,
+      "num_input_tokens_seen": 11468800,
+      "step": 1400
+    },
+    {
+      "epoch": 0.16134569172674218,
+      "grad_norm": 0.5728926658630371,
+      "learning_rate": 4.9654534240625225e-05,
+      "loss": 1.0993,
+      "num_input_tokens_seen": 11550720,
+      "step": 1410
+    },
+    {
+      "epoch": 0.16248998741274745,
+      "grad_norm": 0.5300357341766357,
+      "learning_rate": 4.9637518077749476e-05,
+      "loss": 1.156,
+      "num_input_tokens_seen": 11632640,
+      "step": 1420
+    },
+    {
+      "epoch": 0.16363428309875272,
+      "grad_norm": 0.9682191610336304,
+      "learning_rate": 4.962009591452032e-05,
+      "loss": 1.266,
+      "num_input_tokens_seen": 11714560,
+      "step": 1430
+    },
+    {
+      "epoch": 0.164778578784758,
+      "grad_norm": 0.5290905237197876,
+      "learning_rate": 4.960226803803664e-05,
+      "loss": 1.2794,
+      "num_input_tokens_seen": 11796480,
+      "step": 1440
+    },
+    {
+      "epoch": 0.16592287447076323,
+      "grad_norm": 0.8597230315208435,
+      "learning_rate": 4.958403474208308e-05,
+      "loss": 1.2099,
+      "num_input_tokens_seen": 11878400,
+      "step": 1450
+    },
+    {
+      "epoch": 0.1670671701567685,
+      "grad_norm": 0.6521583795547485,
+      "learning_rate": 4.9565396327125155e-05,
+      "loss": 1.3631,
+      "num_input_tokens_seen": 11960320,
+      "step": 1460
+    },
+    {
+      "epoch": 0.16821146584277377,
+      "grad_norm": 0.6258875727653503,
+      "learning_rate": 4.95463531003043e-05,
+      "loss": 1.1778,
+      "num_input_tokens_seen": 12042240,
+      "step": 1470
+    },
+    {
+      "epoch": 0.16935576152877904,
+      "grad_norm": 3.744122266769409,
+      "learning_rate": 4.952690537543287e-05,
+      "loss": 1.3796,
+      "num_input_tokens_seen": 12124160,
+      "step": 1480
+    },
+    {
+      "epoch": 0.1705000572147843,
+      "grad_norm": 1.4813406467437744,
+      "learning_rate": 4.9507053472988867e-05,
+      "loss": 1.2806,
+      "num_input_tokens_seen": 12206080,
+      "step": 1490
+    },
+    {
+      "epoch": 0.17164435290078955,
+      "grad_norm": 0.5667291283607483,
+      "learning_rate": 4.9486797720110746e-05,
+      "loss": 1.3344,
+      "num_input_tokens_seen": 12288000,
+      "step": 1500
+    },
+    {
+      "epoch": 0.17278864858679482,
+      "grad_norm": 1.1978493928909302,
+      "learning_rate": 4.946613845059199e-05,
+      "loss": 1.2448,
+      "num_input_tokens_seen": 12369920,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1739329442728001,
+      "grad_norm": 1.018848180770874,
+      "learning_rate": 4.9445076004875596e-05,
+      "loss": 1.2916,
+      "num_input_tokens_seen": 12451840,
+      "step": 1520
+    },
+    {
+      "epoch": 0.17507723995880536,
+      "grad_norm": 0.5812198519706726,
+      "learning_rate": 4.9423610730048495e-05,
+      "loss": 1.3731,
+      "num_input_tokens_seen": 12533760,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1762215356448106,
+      "grad_norm": 0.5890350341796875,
+      "learning_rate": 4.940174297983581e-05,
+      "loss": 1.3654,
+      "num_input_tokens_seen": 12615680,
+      "step": 1540
+    },
+    {
+      "epoch": 0.17736583133081588,
+      "grad_norm": 0.5180538892745972,
+      "learning_rate": 4.937947311459503e-05,
+      "loss": 1.2584,
+      "num_input_tokens_seen": 12697600,
+      "step": 1550
+    },
+    {
+      "epoch": 0.17851012701682115,
+      "grad_norm": 0.6072027087211609,
+      "learning_rate": 4.9356801501310105e-05,
+      "loss": 1.372,
+      "num_input_tokens_seen": 12779520,
+      "step": 1560
+    },
+    {
+      "epoch": 0.17965442270282642,
+      "grad_norm": 0.5989131331443787,
+      "learning_rate": 4.933372851358532e-05,
+      "loss": 1.562,
+      "num_input_tokens_seen": 12861440,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1807987183888317,
+      "grad_norm": 0.6137218475341797,
+      "learning_rate": 4.9310254531639235e-05,
+      "loss": 1.2423,
+      "num_input_tokens_seen": 12943360,
+      "step": 1580
+    },
+    {
+      "epoch": 0.18194301407483693,
+      "grad_norm": 0.5463627576828003,
+      "learning_rate": 4.928637994229834e-05,
+      "loss": 1.3979,
+      "num_input_tokens_seen": 13025280,
+      "step": 1590
+    },
+    {
+      "epoch": 0.1830873097608422,
+      "grad_norm": 0.5558052062988281,
+      "learning_rate": 4.9262105138990745e-05,
+      "loss": 1.3811,
+      "num_input_tokens_seen": 13107200,
+      "step": 1600
+    },
+    {
+      "epoch": 0.18423160544684747,
+      "grad_norm": 0.6902844905853271,
+      "learning_rate": 4.9237430521739626e-05,
+      "loss": 1.4006,
+      "num_input_tokens_seen": 13189120,
+      "step": 1610
+    },
+    {
+      "epoch": 0.18537590113285274,
+      "grad_norm": 1.180159568786621,
+      "learning_rate": 4.92123564971567e-05,
+      "loss": 1.2214,
+      "num_input_tokens_seen": 13271040,
+      "step": 1620
+    },
+    {
+      "epoch": 0.18652019681885798,
+      "grad_norm": 0.6135653853416443,
+      "learning_rate": 4.918688347843549e-05,
+      "loss": 1.2192,
+      "num_input_tokens_seen": 13352960,
+      "step": 1630
+    },
+    {
+      "epoch": 0.18766449250486325,
+      "grad_norm": 0.5236485004425049,
+      "learning_rate": 4.916101188534452e-05,
+      "loss": 1.4384,
+      "num_input_tokens_seen": 13434880,
+      "step": 1640
+    },
+    {
+      "epoch": 0.18880878819086852,
+      "grad_norm": 0.6157830357551575,
+      "learning_rate": 4.9134742144220394e-05,
+      "loss": 1.2473,
+      "num_input_tokens_seen": 13516800,
+      "step": 1650
+    },
+    {
+      "epoch": 0.1899530838768738,
+      "grad_norm": 1.5510542392730713,
+      "learning_rate": 4.910807468796079e-05,
+      "loss": 1.277,
+      "num_input_tokens_seen": 13598720,
+      "step": 1660
+    },
+    {
+      "epoch": 0.19109737956287906,
+      "grad_norm": 0.5472043752670288,
+      "learning_rate": 4.90810099560173e-05,
+      "loss": 1.31,
+      "num_input_tokens_seen": 13680640,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1922416752488843,
+      "grad_norm": 0.5276287794113159,
+      "learning_rate": 4.90535483943882e-05,
+      "loss": 1.2949,
+      "num_input_tokens_seen": 13762560,
+      "step": 1680
+    },
+    {
+      "epoch": 0.19338597093488957,
+      "grad_norm": 0.5264617800712585,
+      "learning_rate": 4.902569045561113e-05,
+      "loss": 1.3735,
+      "num_input_tokens_seen": 13844480,
+      "step": 1690
+    },
+    {
+      "epoch": 0.19453026662089484,
+      "grad_norm": 0.5804072022438049,
+      "learning_rate": 4.899743659875556e-05,
+      "loss": 1.2904,
+      "num_input_tokens_seen": 13926400,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1956745623069001,
+      "grad_norm": 0.5250833630561829,
+      "learning_rate": 4.896878728941531e-05,
+      "loss": 1.3387,
+      "num_input_tokens_seen": 14008320,
+      "step": 1710
+    },
+    {
+      "epoch": 0.19681885799290535,
+      "grad_norm": 0.6384326815605164,
+      "learning_rate": 4.893974299970082e-05,
+      "loss": 1.1596,
+      "num_input_tokens_seen": 14090240,
+      "step": 1720
+    },
+    {
+      "epoch": 0.19796315367891062,
+      "grad_norm": 0.5223713517189026,
+      "learning_rate": 4.891030420823142e-05,
+      "loss": 1.296,
+      "num_input_tokens_seen": 14172160,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1991074493649159,
+      "grad_norm": 0.5310291051864624,
+      "learning_rate": 4.888047140012737e-05,
+      "loss": 1.3196,
+      "num_input_tokens_seen": 14254080,
+      "step": 1740
+    },
+    {
+      "epoch": 0.20025174505092116,
+      "grad_norm": 0.6313158869743347,
+      "learning_rate": 4.885024506700195e-05,
+      "loss": 1.1085,
+      "num_input_tokens_seen": 14336000,
+      "step": 1750
+    },
+    {
+      "epoch": 0.20139604073692643,
+      "grad_norm": 0.9189643263816833,
+      "learning_rate": 4.8819625706953286e-05,
+      "loss": 1.3757,
+      "num_input_tokens_seen": 14417920,
+      "step": 1760
+    },
+    {
+      "epoch": 0.20254033642293168,
+      "grad_norm": 0.5741713047027588,
+      "learning_rate": 4.8788613824556194e-05,
+      "loss": 1.1697,
+      "num_input_tokens_seen": 14499840,
+      "step": 1770
+    },
+    {
+      "epoch": 0.20368463210893695,
+      "grad_norm": 0.5382892489433289,
+      "learning_rate": 4.875720993085384e-05,
+      "loss": 1.2916,
+      "num_input_tokens_seen": 14581760,
+      "step": 1780
+    },
+    {
+      "epoch": 0.20482892779494222,
+      "grad_norm": 0.5520409941673279,
+      "learning_rate": 4.8725414543349326e-05,
+      "loss": 1.1496,
+      "num_input_tokens_seen": 14663680,
+      "step": 1790
+    },
+    {
+      "epoch": 0.2059732234809475,
+      "grad_norm": 0.5150988101959229,
+      "learning_rate": 4.869322818599714e-05,
+      "loss": 1.265,
+      "num_input_tokens_seen": 14745600,
+      "step": 1800
+    },
+    {
+      "epoch": 0.20711751916695273,
+      "grad_norm": 0.5061900019645691,
+      "learning_rate": 4.8660651389194576e-05,
+      "loss": 1.293,
+      "num_input_tokens_seen": 14827520,
+      "step": 1810
+    },
+    {
+      "epoch": 0.208261814852958,
+      "grad_norm": 0.5306046009063721,
+      "learning_rate": 4.862768468977293e-05,
+      "loss": 1.3073,
+      "num_input_tokens_seen": 14909440,
+      "step": 1820
+    },
+    {
+      "epoch": 0.20940611053896327,
+      "grad_norm": 0.5865935683250427,
+      "learning_rate": 4.8594328630988696e-05,
+      "loss": 1.171,
+      "num_input_tokens_seen": 14991360,
+      "step": 1830
+    },
+    {
+      "epoch": 0.21055040622496854,
+      "grad_norm": 0.5565701723098755,
+      "learning_rate": 4.8560583762514594e-05,
+      "loss": 1.2818,
+      "num_input_tokens_seen": 15073280,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2116947019109738,
+      "grad_norm": 0.504626452922821,
+      "learning_rate": 4.852645064043053e-05,
+      "loss": 1.0901,
+      "num_input_tokens_seen": 15155200,
+      "step": 1850
+    },
+    {
+      "epoch": 0.21283899759697905,
+      "grad_norm": 0.49429330229759216,
+      "learning_rate": 4.84919298272144e-05,
+      "loss": 1.2341,
+      "num_input_tokens_seen": 15237120,
+      "step": 1860
+    },
+    {
+      "epoch": 0.21398329328298432,
+      "grad_norm": 0.6143152117729187,
+      "learning_rate": 4.8457021891732866e-05,
+      "loss": 1.415,
+      "num_input_tokens_seen": 15319040,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2151275889689896,
+      "grad_norm": 0.5530739426612854,
+      "learning_rate": 4.842172740923194e-05,
+      "loss": 1.5628,
+      "num_input_tokens_seen": 15400960,
+      "step": 1880
+    },
+    {
+      "epoch": 0.21627188465499486,
+      "grad_norm": 0.5200616717338562,
+      "learning_rate": 4.838604696132753e-05,
+      "loss": 1.4214,
+      "num_input_tokens_seen": 15482880,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2174161803410001,
+      "grad_norm": 0.9391474723815918,
+      "learning_rate": 4.8349981135995826e-05,
+      "loss": 1.2436,
+      "num_input_tokens_seen": 15564800,
+      "step": 1900
+    },
+    {
+      "epoch": 0.21856047602700537,
+      "grad_norm": 0.49321290850639343,
+      "learning_rate": 4.831353052756367e-05,
+      "loss": 1.3363,
+      "num_input_tokens_seen": 15646720,
+      "step": 1910
+    },
+    {
+      "epoch": 0.21970477171301064,
+      "grad_norm": 0.5570663213729858,
+      "learning_rate": 4.8276695736698704e-05,
+      "loss": 1.3984,
+      "num_input_tokens_seen": 15728640,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2208490673990159,
+      "grad_norm": 0.5171666145324707,
+      "learning_rate": 4.823947737039948e-05,
+      "loss": 1.2282,
+      "num_input_tokens_seen": 15810560,
+      "step": 1930
+    },
+    {
+      "epoch": 0.22199336308502116,
+      "grad_norm": 0.6550266742706299,
+      "learning_rate": 4.8201876041985496e-05,
+      "loss": 1.32,
+      "num_input_tokens_seen": 15892480,
+      "step": 1940
+    },
+    {
+      "epoch": 0.22313765877102643,
+      "grad_norm": 0.6058536171913147,
+      "learning_rate": 4.8163892371087045e-05,
+      "loss": 1.2447,
+      "num_input_tokens_seen": 15974400,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2242819544570317,
+      "grad_norm": 0.7001304626464844,
+      "learning_rate": 4.812552698363502e-05,
+      "loss": 1.1607,
+      "num_input_tokens_seen": 16056320,
+      "step": 1960
+    },
+    {
+      "epoch": 0.22542625014303697,
+      "grad_norm": 0.8715665340423584,
+      "learning_rate": 4.8086780511850606e-05,
+      "loss": 1.3617,
+      "num_input_tokens_seen": 16138240,
+      "step": 1970
+    },
+    {
+      "epoch": 0.22657054582904224,
+      "grad_norm": 0.5489494800567627,
+      "learning_rate": 4.8047653594234855e-05,
+      "loss": 1.1251,
+      "num_input_tokens_seen": 16220160,
+      "step": 1980
+    },
+    {
+      "epoch": 0.22771484151504748,
+      "grad_norm": 0.5537709593772888,
+      "learning_rate": 4.800814687555817e-05,
+      "loss": 1.193,
+      "num_input_tokens_seen": 16302080,
+      "step": 1990
+    },
+    {
+      "epoch": 0.22885913720105275,
+      "grad_norm": 0.9744300842285156,
+      "learning_rate": 4.796826100684967e-05,
+      "loss": 1.1411,
+      "num_input_tokens_seen": 16384000,
+      "step": 2000
+    },
+    {
+      "epoch": 0.23000343288705802,
+      "grad_norm": 0.6724236607551575,
+      "learning_rate": 4.7927996645386476e-05,
+      "loss": 1.2578,
+      "num_input_tokens_seen": 16465920,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2311477285730633,
+      "grad_norm": 0.6202824711799622,
+      "learning_rate": 4.7887354454682854e-05,
+      "loss": 1.1249,
+      "num_input_tokens_seen": 16547840,
+      "step": 2020
+    },
+    {
+      "epoch": 0.23229202425906853,
+      "grad_norm": 0.5438826680183411,
+      "learning_rate": 4.784633510447932e-05,
+      "loss": 1.1754,
+      "num_input_tokens_seen": 16629760,
+      "step": 2030
+    },
+    {
+      "epoch": 0.2334363199450738,
+      "grad_norm": 0.5820137858390808,
+      "learning_rate": 4.7804939270731564e-05,
+      "loss": 1.2965,
+      "num_input_tokens_seen": 16711680,
+      "step": 2040
+    },
+    {
+      "epoch": 0.23458061563107907,
+      "grad_norm": 0.5191706418991089,
+      "learning_rate": 4.776316763559933e-05,
+      "loss": 1.2211,
+      "num_input_tokens_seen": 16793600,
+      "step": 2050
+    },
+    {
+      "epoch": 0.23572491131708434,
+      "grad_norm": 0.9790087342262268,
+      "learning_rate": 4.7721020887435186e-05,
+      "loss": 1.0741,
+      "num_input_tokens_seen": 16875520,
+      "step": 2060
+    },
+    {
+      "epoch": 0.2368692070030896,
+      "grad_norm": 0.48102623224258423,
+      "learning_rate": 4.767849972077315e-05,
+      "loss": 1.1755,
+      "num_input_tokens_seen": 16957440,
+      "step": 2070
+    },
+    {
+      "epoch": 0.23801350268909485,
+      "grad_norm": 0.5206550359725952,
+      "learning_rate": 4.763560483631728e-05,
+      "loss": 1.2918,
+      "num_input_tokens_seen": 17039360,
+      "step": 2080
+    },
+    {
+      "epoch": 0.23915779837510012,
+      "grad_norm": 0.620953381061554,
+      "learning_rate": 4.75923369409301e-05,
+      "loss": 1.2861,
+      "num_input_tokens_seen": 17121280,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2403020940611054,
+      "grad_norm": 0.5213440656661987,
+      "learning_rate": 4.7548696747620956e-05,
+      "loss": 1.2797,
+      "num_input_tokens_seen": 17203200,
+      "step": 2100
+    },
+    {
+      "epoch": 0.24144638974711066,
+      "grad_norm": 0.5252777934074402,
+      "learning_rate": 4.750468497553429e-05,
+      "loss": 1.2802,
+      "num_input_tokens_seen": 17285120,
+      "step": 2110
+    },
+    {
+      "epoch": 0.2425906854331159,
+      "grad_norm": 0.553130030632019,
+      "learning_rate": 4.746030234993775e-05,
+      "loss": 1.2026,
+      "num_input_tokens_seen": 17367040,
+      "step": 2120
+    },
+    {
+      "epoch": 0.24373498111912117,
+      "grad_norm": 0.5378229022026062,
+      "learning_rate": 4.741554960221027e-05,
+      "loss": 1.4968,
+      "num_input_tokens_seen": 17448960,
+      "step": 2130
+    },
+    {
+      "epoch": 0.24487927680512644,
+      "grad_norm": 2.1330316066741943,
+      "learning_rate": 4.7370427469830016e-05,
+      "loss": 1.4157,
+      "num_input_tokens_seen": 17530880,
+      "step": 2140
+    },
+    {
+      "epoch": 0.24602357249113171,
+      "grad_norm": 0.555316150188446,
+      "learning_rate": 4.73249366963622e-05,
+      "loss": 1.2478,
+      "num_input_tokens_seen": 17612800,
+      "step": 2150
+    },
+    {
+      "epoch": 0.24716786817713698,
+      "grad_norm": 0.5266692638397217,
+      "learning_rate": 4.727907803144686e-05,
+      "loss": 1.4627,
+      "num_input_tokens_seen": 17694720,
+      "step": 2160
+    },
+    {
+      "epoch": 0.24831216386314223,
+      "grad_norm": 0.5252066254615784,
+      "learning_rate": 4.723285223078653e-05,
+      "loss": 1.2206,
+      "num_input_tokens_seen": 17776640,
+      "step": 2170
+    },
+    {
+      "epoch": 0.2494564595491475,
+      "grad_norm": 0.6599643230438232,
+      "learning_rate": 4.71862600561337e-05,
+      "loss": 1.1268,
+      "num_input_tokens_seen": 17858560,
+      "step": 2180
+    },
+    {
+      "epoch": 0.25060075523515274,
+      "grad_norm": 1.4253127574920654,
+      "learning_rate": 4.713930227527836e-05,
+      "loss": 1.0187,
+      "num_input_tokens_seen": 17940480,
+      "step": 2190
+    },
+    {
+      "epoch": 0.251745050921158,
+      "grad_norm": 0.9951733946800232,
+      "learning_rate": 4.709197966203528e-05,
+      "loss": 0.875,
+      "num_input_tokens_seen": 18022400,
+      "step": 2200
+    },
+    {
+      "epoch": 0.2528893466071633,
+      "grad_norm": 0.552880585193634,
+      "learning_rate": 4.704429299623129e-05,
+      "loss": 1.2277,
+      "num_input_tokens_seen": 18104320,
+      "step": 2210
+    },
+    {
+      "epoch": 0.25403364229316855,
+      "grad_norm": 0.47009891271591187,
+      "learning_rate": 4.6996243063692446e-05,
+      "loss": 1.4153,
+      "num_input_tokens_seen": 18186240,
+      "step": 2220
+    },
+    {
+      "epoch": 0.2551779379791738,
+      "grad_norm": 0.5296884179115295,
+      "learning_rate": 4.694783065623102e-05,
+      "loss": 1.1905,
+      "num_input_tokens_seen": 18268160,
+      "step": 2230
+    },
+    {
+      "epoch": 0.2563222336651791,
+      "grad_norm": 0.561241865158081,
+      "learning_rate": 4.68990565716325e-05,
+      "loss": 1.363,
+      "num_input_tokens_seen": 18350080,
+      "step": 2240
+    },
+    {
+      "epoch": 0.25746652935118436,
+      "grad_norm": 0.4963686466217041,
+      "learning_rate": 4.6849921613642456e-05,
+      "loss": 1.3028,
+      "num_input_tokens_seen": 18432000,
+      "step": 2250
+    },
+    {
+      "epoch": 0.25861082503718963,
+      "grad_norm": 0.5019493103027344,
+      "learning_rate": 4.680042659195325e-05,
+      "loss": 1.0479,
+      "num_input_tokens_seen": 18513920,
+      "step": 2260
+    },
+    {
+      "epoch": 0.2597551207231949,
+      "grad_norm": 0.5507912039756775,
+      "learning_rate": 4.6750572322190716e-05,
+      "loss": 1.4381,
+      "num_input_tokens_seen": 18595840,
+      "step": 2270
+    },
+    {
+      "epoch": 0.2608994164092001,
+      "grad_norm": 0.45838436484336853,
+      "learning_rate": 4.6700359625900724e-05,
+      "loss": 1.1211,
+      "num_input_tokens_seen": 18677760,
+      "step": 2280
+    },
+    {
+      "epoch": 0.2620437120952054,
+      "grad_norm": 0.49079829454421997,
+      "learning_rate": 4.664978933053562e-05,
+      "loss": 1.1206,
+      "num_input_tokens_seen": 18759680,
+      "step": 2290
+    },
+    {
+      "epoch": 0.26318800778121065,
+      "grad_norm": 1.0304409265518188,
+      "learning_rate": 4.659886226944063e-05,
+      "loss": 1.2446,
+      "num_input_tokens_seen": 18841600,
+      "step": 2300
+    },
+    {
+      "epoch": 0.2643323034672159,
+      "grad_norm": 0.6165328025817871,
+      "learning_rate": 4.65475792818401e-05,
+      "loss": 1.1047,
+      "num_input_tokens_seen": 18923520,
+      "step": 2310
+    },
+    {
+      "epoch": 0.2654765991532212,
+      "grad_norm": 1.5652453899383545,
+      "learning_rate": 4.6495941212823644e-05,
+      "loss": 1.175,
+      "num_input_tokens_seen": 19005440,
+      "step": 2320
+    },
+    {
+      "epoch": 0.26662089483922646,
+      "grad_norm": 0.5280311107635498,
+      "learning_rate": 4.644394891333227e-05,
+      "loss": 1.1473,
+      "num_input_tokens_seen": 19087360,
+      "step": 2330
+    },
+    {
+      "epoch": 0.26776519052523173,
+      "grad_norm": 1.034646987915039,
+      "learning_rate": 4.639160324014433e-05,
+      "loss": 1.1003,
+      "num_input_tokens_seen": 19169280,
+      "step": 2340
+    },
+    {
+      "epoch": 0.268909486211237,
+      "grad_norm": 0.4900410771369934,
+      "learning_rate": 4.633890505586139e-05,
+      "loss": 1.5043,
+      "num_input_tokens_seen": 19251200,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2700537818972423,
+      "grad_norm": 1.9830961227416992,
+      "learning_rate": 4.6285855228894025e-05,
+      "loss": 1.3875,
+      "num_input_tokens_seen": 19333120,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2711980775832475,
+      "grad_norm": 0.5432376265525818,
+      "learning_rate": 4.623245463344753e-05,
+      "loss": 1.2485,
+      "num_input_tokens_seen": 19415040,
+      "step": 2370
+    },
+    {
+      "epoch": 0.27234237326925276,
+      "grad_norm": 0.6268254518508911,
+      "learning_rate": 4.617870414950748e-05,
+      "loss": 1.6017,
+      "num_input_tokens_seen": 19496960,
+      "step": 2380
+    },
+    {
+      "epoch": 0.273486668955258,
+      "grad_norm": 0.6118723750114441,
+      "learning_rate": 4.612460466282525e-05,
+      "loss": 1.1282,
+      "num_input_tokens_seen": 19578880,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2746309646412633,
+      "grad_norm": 0.6583623290061951,
+      "learning_rate": 4.607015706490341e-05,
+      "loss": 1.1795,
+      "num_input_tokens_seen": 19660800,
+      "step": 2400
+    },
+    {
+      "epoch": 0.27577526032726857,
+      "grad_norm": 0.6352823376655579,
+      "learning_rate": 4.601536225298104e-05,
+      "loss": 1.111,
+      "num_input_tokens_seen": 19742720,
+      "step": 2410
+    },
+    {
+      "epoch": 0.27691955601327384,
+      "grad_norm": 1.524483561515808,
+      "learning_rate": 4.5960221130018946e-05,
+      "loss": 1.1325,
+      "num_input_tokens_seen": 19824640,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2780638516992791,
+      "grad_norm": 0.41133037209510803,
+      "learning_rate": 4.590473460468475e-05,
+      "loss": 1.2953,
+      "num_input_tokens_seen": 19906560,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2792081473852844,
+      "grad_norm": 0.8059386610984802,
+      "learning_rate": 4.584890359133797e-05,
+      "loss": 1.168,
+      "num_input_tokens_seen": 19988480,
+      "step": 2440
+    },
+    {
+      "epoch": 0.28035244307128965,
+      "grad_norm": 1.0323024988174438,
+      "learning_rate": 4.579272901001491e-05,
+      "loss": 1.3607,
+      "num_input_tokens_seen": 20070400,
+      "step": 2450
+    },
+    {
+      "epoch": 0.28149673875729486,
+      "grad_norm": 0.5809211134910583,
+      "learning_rate": 4.5736211786413524e-05,
+      "loss": 1.293,
+      "num_input_tokens_seen": 20152320,
+      "step": 2460
+    },
+    {
+      "epoch": 0.28264103444330013,
+      "grad_norm": 0.7068034410476685,
+      "learning_rate": 4.5679352851878135e-05,
+      "loss": 1.3167,
+      "num_input_tokens_seen": 20234240,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2837853301293054,
+      "grad_norm": 0.5605278611183167,
+      "learning_rate": 4.562215314338411e-05,
+      "loss": 1.3806,
+      "num_input_tokens_seen": 20316160,
+      "step": 2480
+    },
+    {
+      "epoch": 0.28492962581531067,
+      "grad_norm": 0.5452982187271118,
+      "learning_rate": 4.556461360352241e-05,
+      "loss": 1.0428,
+      "num_input_tokens_seen": 20398080,
+      "step": 2490
+    },
+    {
+      "epoch": 0.28607392150131594,
+      "grad_norm": 0.5816521048545837,
+      "learning_rate": 4.550673518048405e-05,
+      "loss": 1.0979,
+      "num_input_tokens_seen": 20480000,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2872182171873212,
+      "grad_norm": 0.6600253582000732,
+      "learning_rate": 4.5448518828044515e-05,
+      "loss": 1.2309,
+      "num_input_tokens_seen": 20561920,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2883625128733265,
+      "grad_norm": 0.5788952708244324,
+      "learning_rate": 4.538996550554798e-05,
+      "loss": 1.1428,
+      "num_input_tokens_seen": 20643840,
+      "step": 2520
+    },
+    {
+      "epoch": 0.28950680855933175,
+      "grad_norm": 0.5803161263465881,
+      "learning_rate": 4.5331076177891527e-05,
+      "loss": 1.5423,
+      "num_input_tokens_seen": 20725760,
+      "step": 2530
+    },
+    {
+      "epoch": 0.290651104245337,
+      "grad_norm": 0.5454373955726624,
+      "learning_rate": 4.527185181550928e-05,
+      "loss": 1.31,
+      "num_input_tokens_seen": 20807680,
+      "step": 2540
+    },
+    {
+      "epoch": 0.29179539993134224,
+      "grad_norm": 1.4661908149719238,
+      "learning_rate": 4.5212293394356356e-05,
+      "loss": 1.0873,
+      "num_input_tokens_seen": 20889600,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2929396956173475,
+      "grad_norm": 0.6312928199768066,
+      "learning_rate": 4.515240189589282e-05,
+      "loss": 1.1553,
+      "num_input_tokens_seen": 20971520,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2940839913033528,
+      "grad_norm": 0.5782475471496582,
+      "learning_rate": 4.509217830706749e-05,
+      "loss": 1.267,
+      "num_input_tokens_seen": 21053440,
+      "step": 2570
+    },
+    {
+      "epoch": 0.29522828698935805,
+      "grad_norm": 0.5917658805847168,
+      "learning_rate": 4.50316236203017e-05,
+      "loss": 1.074,
+      "num_input_tokens_seen": 21135360,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2963725826753633,
+      "grad_norm": 0.5403671264648438,
+      "learning_rate": 4.497073883347293e-05,
+      "loss": 1.238,
+      "num_input_tokens_seen": 21217280,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2975168783613686,
+      "grad_norm": 0.5732063055038452,
+      "learning_rate": 4.490952494989834e-05,
+      "loss": 1.2223,
+      "num_input_tokens_seen": 21299200,
+      "step": 2600
+    },
+    {
+      "epoch": 0.29866117404737386,
+      "grad_norm": 0.5230866074562073,
+      "learning_rate": 4.484798297831826e-05,
+      "loss": 1.2797,
+      "num_input_tokens_seen": 21381120,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2998054697333791,
+      "grad_norm": 0.5615763664245605,
+      "learning_rate": 4.4786113932879605e-05,
+      "loss": 1.1556,
+      "num_input_tokens_seen": 21463040,
+      "step": 2620
+    },
+    {
+      "epoch": 0.3009497654193844,
+      "grad_norm": 0.8755237460136414,
+      "learning_rate": 4.472391883311906e-05,
+      "loss": 1.156,
+      "num_input_tokens_seen": 21544960,
+      "step": 2630
+    },
+    {
+      "epoch": 0.3020940611053896,
+      "grad_norm": 0.5666770935058594,
+      "learning_rate": 4.4661398703946396e-05,
+      "loss": 1.1363,
+      "num_input_tokens_seen": 21626880,
+      "step": 2640
+    },
+    {
+      "epoch": 0.3032383567913949,
+      "grad_norm": 0.5907579064369202,
+      "learning_rate": 4.4598554575627495e-05,
+      "loss": 1.4523,
+      "num_input_tokens_seen": 21708800,
+      "step": 2650
+    },
+    {
+      "epoch": 0.30438265247740015,
+      "grad_norm": 0.5281969904899597,
+      "learning_rate": 4.453538748376742e-05,
+      "loss": 0.9537,
+      "num_input_tokens_seen": 21790720,
+      "step": 2660
+    },
+    {
+      "epoch": 0.3055269481634054,
+      "grad_norm": 0.6245942115783691,
+      "learning_rate": 4.4471898469293324e-05,
+      "loss": 1.1622,
+      "num_input_tokens_seen": 21872640,
+      "step": 2670
+    },
+    {
+      "epoch": 0.3066712438494107,
+      "grad_norm": 0.5124683976173401,
+      "learning_rate": 4.44080885784373e-05,
+      "loss": 1.3041,
+      "num_input_tokens_seen": 21954560,
+      "step": 2680
+    },
+    {
+      "epoch": 0.30781553953541596,
+      "grad_norm": 0.536454439163208,
+      "learning_rate": 4.434395886271917e-05,
+      "loss": 1.3722,
+      "num_input_tokens_seen": 22036480,
+      "step": 2690
+    },
+    {
+      "epoch": 0.30895983522142123,
+      "grad_norm": 0.5136358737945557,
+      "learning_rate": 4.427951037892911e-05,
+      "loss": 1.4421,
+      "num_input_tokens_seen": 22118400,
+      "step": 2700
+    },
+    {
+      "epoch": 0.3101041309074265,
+      "grad_norm": 0.4972691237926483,
+      "learning_rate": 4.4214744189110266e-05,
+      "loss": 1.5259,
+      "num_input_tokens_seen": 22200320,
+      "step": 2710
+    },
+    {
+      "epoch": 0.3112484265934317,
+      "grad_norm": 0.5017683506011963,
+      "learning_rate": 4.414966136054125e-05,
+      "loss": 1.1897,
+      "num_input_tokens_seen": 22282240,
+      "step": 2720
+    },
+    {
+      "epoch": 0.312392722279437,
+      "grad_norm": 0.468337744474411,
+      "learning_rate": 4.408426296571852e-05,
+      "loss": 1.0431,
+      "num_input_tokens_seen": 22364160,
+      "step": 2730
+    },
+    {
+      "epoch": 0.31353701796544226,
+      "grad_norm": 0.6088995337486267,
+      "learning_rate": 4.401855008233879e-05,
+      "loss": 1.3437,
+      "num_input_tokens_seen": 22446080,
+      "step": 2740
+    },
+    {
+      "epoch": 0.3146813136514475,
+      "grad_norm": 0.6361683011054993,
+      "learning_rate": 4.395252379328115e-05,
+      "loss": 1.4596,
+      "num_input_tokens_seen": 22528000,
+      "step": 2750
+    },
+    {
+      "epoch": 0.3158256093374528,
+      "grad_norm": 0.4672735631465912,
+      "learning_rate": 4.388618518658932e-05,
+      "loss": 1.1596,
+      "num_input_tokens_seen": 22609920,
+      "step": 2760
+    },
+    {
+      "epoch": 0.31696990502345807,
+      "grad_norm": 1.0056240558624268,
+      "learning_rate": 4.381953535545369e-05,
+      "loss": 1.2776,
+      "num_input_tokens_seen": 22691840,
+      "step": 2770
+    },
+    {
+      "epoch": 0.31811420070946334,
+      "grad_norm": 0.8526020646095276,
+      "learning_rate": 4.375257539819328e-05,
+      "loss": 1.1147,
+      "num_input_tokens_seen": 22773760,
+      "step": 2780
+    },
+    {
+      "epoch": 0.3192584963954686,
+      "grad_norm": 0.527472972869873,
+      "learning_rate": 4.368530641823769e-05,
+      "loss": 1.1802,
+      "num_input_tokens_seen": 22855680,
+      "step": 2790
+    },
+    {
+      "epoch": 0.3204027920814739,
+      "grad_norm": 0.5480664372444153,
+      "learning_rate": 4.361772952410886e-05,
+      "loss": 1.0881,
+      "num_input_tokens_seen": 22937600,
+      "step": 2800
+    },
+    {
+      "epoch": 0.3215470877674791,
+      "grad_norm": 0.6834565997123718,
+      "learning_rate": 4.354984582940285e-05,
+      "loss": 1.1945,
+      "num_input_tokens_seen": 23019520,
+      "step": 2810
+    },
+    {
+      "epoch": 0.32269138345348436,
+      "grad_norm": 0.5412240028381348,
+      "learning_rate": 4.348165645277145e-05,
+      "loss": 1.3629,
+      "num_input_tokens_seen": 23101440,
+      "step": 2820
+    },
+    {
+      "epoch": 0.32383567913948963,
+      "grad_norm": 0.5615849494934082,
+      "learning_rate": 4.34131625179038e-05,
+      "loss": 1.3362,
+      "num_input_tokens_seen": 23183360,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3249799748254949,
+      "grad_norm": 0.5076019763946533,
+      "learning_rate": 4.334436515350779e-05,
+      "loss": 1.2316,
+      "num_input_tokens_seen": 23265280,
+      "step": 2840
+    },
+    {
+      "epoch": 0.32612427051150017,
+      "grad_norm": 0.568336546421051,
+      "learning_rate": 4.327526549329157e-05,
+      "loss": 1.7573,
+      "num_input_tokens_seen": 23347200,
+      "step": 2850
+    },
+    {
+      "epoch": 0.32726856619750544,
+      "grad_norm": 0.6184702515602112,
+      "learning_rate": 4.320586467594476e-05,
+      "loss": 1.4037,
+      "num_input_tokens_seen": 23429120,
+      "step": 2860
+    },
+    {
+      "epoch": 0.3284128618835107,
+      "grad_norm": 0.8441096544265747,
+      "learning_rate": 4.313616384511976e-05,
+      "loss": 1.4623,
+      "num_input_tokens_seen": 23511040,
+      "step": 2870
+    },
+    {
+      "epoch": 0.329557157569516,
+      "grad_norm": 0.41614028811454773,
+      "learning_rate": 4.3066164149412844e-05,
+      "loss": 1.1027,
+      "num_input_tokens_seen": 23592960,
+      "step": 2880
+    },
+    {
+      "epoch": 0.33070145325552125,
+      "grad_norm": 0.6045880317687988,
+      "learning_rate": 4.299586674234529e-05,
+      "loss": 1.2341,
+      "num_input_tokens_seen": 23674880,
+      "step": 2890
+    },
+    {
+      "epoch": 0.33184574894152646,
+      "grad_norm": 0.5018451809883118,
+      "learning_rate": 4.292527278234435e-05,
+      "loss": 1.2661,
+      "num_input_tokens_seen": 23756800,
+      "step": 2900
+    },
+    {
+      "epoch": 0.33299004462753173,
+      "grad_norm": 0.6029537320137024,
+      "learning_rate": 4.285438343272414e-05,
+      "loss": 1.375,
+      "num_input_tokens_seen": 23838720,
+      "step": 2910
+    },
+    {
+      "epoch": 0.334134340313537,
+      "grad_norm": 0.7714540958404541,
+      "learning_rate": 4.278319986166649e-05,
+      "loss": 1.2119,
+      "num_input_tokens_seen": 23920640,
+      "step": 2920
+    },
+    {
+      "epoch": 0.3352786359995423,
+      "grad_norm": 0.5025473237037659,
+      "learning_rate": 4.2711723242201695e-05,
+      "loss": 1.322,
+      "num_input_tokens_seen": 24002560,
+      "step": 2930
+    },
+    {
+      "epoch": 0.33642293168554754,
+      "grad_norm": 0.5495464205741882,
+      "learning_rate": 4.263995475218917e-05,
+      "loss": 1.4806,
+      "num_input_tokens_seen": 24084480,
+      "step": 2940
+    },
+    {
+      "epoch": 0.3375672273715528,
+      "grad_norm": 0.5496026277542114,
+      "learning_rate": 4.256789557429806e-05,
+      "loss": 1.6319,
+      "num_input_tokens_seen": 24166400,
+      "step": 2950
+    },
+    {
+      "epoch": 0.3387115230575581,
+      "grad_norm": 2.653693199157715,
+      "learning_rate": 4.2495546895987724e-05,
+      "loss": 1.2252,
+      "num_input_tokens_seen": 24248320,
+      "step": 2960
+    },
+    {
+      "epoch": 0.33985581874356335,
+      "grad_norm": 0.7369899153709412,
+      "learning_rate": 4.242290990948821e-05,
+      "loss": 1.1941,
+      "num_input_tokens_seen": 24330240,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3410001144295686,
+      "grad_norm": 0.6210314631462097,
+      "learning_rate": 4.234998581178056e-05,
+      "loss": 1.2853,
+      "num_input_tokens_seen": 24412160,
+      "step": 2980
+    },
+    {
+      "epoch": 0.34214441011557384,
+      "grad_norm": 0.4932264983654022,
+      "learning_rate": 4.227677580457711e-05,
+      "loss": 1.157,
+      "num_input_tokens_seen": 24494080,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3432887058015791,
+      "grad_norm": 0.7470729947090149,
+      "learning_rate": 4.220328109430167e-05,
+      "loss": 1.1064,
+      "num_input_tokens_seen": 24576000,
+      "step": 3000
+    },
+    {
+      "epoch": 0.3444330014875844,
+      "grad_norm": 0.5054774880409241,
+      "learning_rate": 4.21295028920697e-05,
+      "loss": 1.3768,
+      "num_input_tokens_seen": 24657920,
+      "step": 3010
+    },
+    {
+      "epoch": 0.34557729717358965,
+      "grad_norm": 0.5471848845481873,
+      "learning_rate": 4.2055442413668264e-05,
+      "loss": 1.1968,
+      "num_input_tokens_seen": 24739840,
+      "step": 3020
+    },
+    {
+      "epoch": 0.3467215928595949,
+      "grad_norm": 0.48451676964759827,
+      "learning_rate": 4.198110087953606e-05,
+      "loss": 1.3736,
+      "num_input_tokens_seen": 24821760,
+      "step": 3030
+    },
+    {
+      "epoch": 0.3478658885456002,
+      "grad_norm": 0.4831438958644867,
+      "learning_rate": 4.190647951474328e-05,
+      "loss": 1.2485,
+      "num_input_tokens_seen": 24903680,
+      "step": 3040
+    },
+    {
+      "epoch": 0.34901018423160546,
+      "grad_norm": 0.5406004190444946,
+      "learning_rate": 4.183157954897144e-05,
+      "loss": 1.1501,
+      "num_input_tokens_seen": 24985600,
+      "step": 3050
+    },
+    {
+      "epoch": 0.35015447991761073,
+      "grad_norm": 0.5474951863288879,
+      "learning_rate": 4.1756402216493115e-05,
+      "loss": 1.1709,
+      "num_input_tokens_seen": 25067520,
+      "step": 3060
+    },
+    {
+      "epoch": 0.351298775603616,
+      "grad_norm": 0.6155632138252258,
+      "learning_rate": 4.1680948756151564e-05,
+      "loss": 1.2692,
+      "num_input_tokens_seen": 25149440,
+      "step": 3070
+    },
+    {
+      "epoch": 0.3524430712896212,
+      "grad_norm": 0.47120699286460876,
+      "learning_rate": 4.160522041134035e-05,
+      "loss": 1.2124,
+      "num_input_tokens_seen": 25231360,
+      "step": 3080
+    },
+    {
+      "epoch": 0.3535873669756265,
+      "grad_norm": 0.5330422520637512,
+      "learning_rate": 4.152921842998287e-05,
+      "loss": 1.1728,
+      "num_input_tokens_seen": 25313280,
+      "step": 3090
+    },
+    {
+      "epoch": 0.35473166266163175,
+      "grad_norm": 0.48196467757225037,
+      "learning_rate": 4.145294406451173e-05,
+      "loss": 1.309,
+      "num_input_tokens_seen": 25395200,
+      "step": 3100
+    },
+    {
+      "epoch": 0.355875958347637,
+      "grad_norm": 0.5000414252281189,
+      "learning_rate": 4.137639857184815e-05,
+      "loss": 1.3163,
+      "num_input_tokens_seen": 25477120,
+      "step": 3110
+    },
+    {
+      "epoch": 0.3570202540336423,
+      "grad_norm": 0.5418947339057922,
+      "learning_rate": 4.129958321338127e-05,
+      "loss": 1.1725,
+      "num_input_tokens_seen": 25559040,
+      "step": 3120
+    },
+    {
+      "epoch": 0.35816454971964756,
+      "grad_norm": 0.5317056775093079,
+      "learning_rate": 4.122249925494726e-05,
+      "loss": 1.1192,
+      "num_input_tokens_seen": 25640960,
+      "step": 3130
+    },
+    {
+      "epoch": 0.35930884540565283,
+      "grad_norm": 0.6244747638702393,
+      "learning_rate": 4.114514796680862e-05,
+      "loss": 1.2959,
+      "num_input_tokens_seen": 25722880,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3604531410916581,
+      "grad_norm": 0.5483636856079102,
+      "learning_rate": 4.106753062363311e-05,
+      "loss": 1.2905,
+      "num_input_tokens_seen": 25804800,
+      "step": 3150
+    },
+    {
+      "epoch": 0.3615974367776634,
+      "grad_norm": 0.5433777570724487,
+      "learning_rate": 4.098964850447281e-05,
+      "loss": 1.315,
+      "num_input_tokens_seen": 25886720,
+      "step": 3160
+    },
+    {
+      "epoch": 0.3627417324636686,
+      "grad_norm": 0.90435391664505,
+      "learning_rate": 4.0911502892743035e-05,
+      "loss": 1.28,
+      "num_input_tokens_seen": 25968640,
+      "step": 3170
+    },
+    {
+      "epoch": 0.36388602814967386,
+      "grad_norm": 0.5195088386535645,
+      "learning_rate": 4.083309507620118e-05,
+      "loss": 1.2515,
+      "num_input_tokens_seen": 26050560,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3650303238356791,
+      "grad_norm": 0.5890297293663025,
+      "learning_rate": 4.075442634692548e-05,
+      "loss": 1.1768,
+      "num_input_tokens_seen": 26132480,
+      "step": 3190
+    },
+    {
+      "epoch": 0.3661746195216844,
+      "grad_norm": 0.6505255103111267,
+      "learning_rate": 4.067549800129375e-05,
+      "loss": 1.2891,
+      "num_input_tokens_seen": 26214400,
+      "step": 3200
+    },
+    {
+      "epoch": 0.36731891520768967,
+      "grad_norm": 0.6695213317871094,
+      "learning_rate": 4.059631133996203e-05,
+      "loss": 1.3888,
+      "num_input_tokens_seen": 26296320,
+      "step": 3210
+    },
+    {
+      "epoch": 0.36846321089369494,
+      "grad_norm": 0.4889095723628998,
+      "learning_rate": 4.05168676678431e-05,
+      "loss": 0.9946,
+      "num_input_tokens_seen": 26378240,
+      "step": 3220
+    },
+    {
+      "epoch": 0.3696075065797002,
+      "grad_norm": 2.1291022300720215,
+      "learning_rate": 4.0437168294085013e-05,
+      "loss": 1.4232,
+      "num_input_tokens_seen": 26460160,
+      "step": 3230
+    },
+    {
+      "epoch": 0.3707518022657055,
+      "grad_norm": 0.4844609797000885,
+      "learning_rate": 4.0357214532049535e-05,
+      "loss": 1.1119,
+      "num_input_tokens_seen": 26542080,
+      "step": 3240
+    },
+    {
+      "epoch": 0.37189609795171075,
+      "grad_norm": 0.509472668170929,
+      "learning_rate": 4.027700769929046e-05,
+      "loss": 1.1302,
+      "num_input_tokens_seen": 26624000,
+      "step": 3250
+    },
+    {
+      "epoch": 0.37304039363771596,
+      "grad_norm": 0.5184939503669739,
+      "learning_rate": 4.019654911753193e-05,
+      "loss": 1.2648,
+      "num_input_tokens_seen": 26705920,
+      "step": 3260
+    },
+    {
+      "epoch": 0.37418468932372123,
+      "grad_norm": 0.5771918892860413,
+      "learning_rate": 4.011584011264665e-05,
+      "loss": 1.204,
+      "num_input_tokens_seen": 26787840,
+      "step": 3270
+    },
+    {
+      "epoch": 0.3753289850097265,
+      "grad_norm": 0.5364697575569153,
+      "learning_rate": 4.0034882014634015e-05,
+      "loss": 1.3242,
+      "num_input_tokens_seen": 26869760,
+      "step": 3280
+    },
+    {
+      "epoch": 0.37647328069573177,
+      "grad_norm": 0.5383545160293579,
+      "learning_rate": 3.995367615759825e-05,
+      "loss": 1.235,
+      "num_input_tokens_seen": 26951680,
+      "step": 3290
+    },
+    {
+      "epoch": 0.37761757638173704,
+      "grad_norm": 0.5549062490463257,
+      "learning_rate": 3.9872223879726356e-05,
+      "loss": 1.1119,
+      "num_input_tokens_seen": 27033600,
+      "step": 3300
+    },
+    {
+      "epoch": 0.3787618720677423,
+      "grad_norm": 0.8855923414230347,
+      "learning_rate": 3.979052652326609e-05,
+      "loss": 1.2295,
+      "num_input_tokens_seen": 27115520,
+      "step": 3310
+    },
+    {
+      "epoch": 0.3799061677537476,
+      "grad_norm": 0.5798035264015198,
+      "learning_rate": 3.970858543450387e-05,
+      "loss": 1.4858,
+      "num_input_tokens_seen": 27197440,
+      "step": 3320
+    },
+    {
+      "epoch": 0.38105046343975285,
+      "grad_norm": 0.9078068137168884,
+      "learning_rate": 3.962640196374254e-05,
+      "loss": 1.1919,
+      "num_input_tokens_seen": 27279360,
+      "step": 3330
+    },
+    {
+      "epoch": 0.3821947591257581,
+      "grad_norm": 0.5055866241455078,
+      "learning_rate": 3.954397746527916e-05,
+      "loss": 1.1531,
+      "num_input_tokens_seen": 27361280,
+      "step": 3340
+    },
+    {
+      "epoch": 0.38333905481176334,
+      "grad_norm": 0.5842207670211792,
+      "learning_rate": 3.9461313297382666e-05,
+      "loss": 1.1539,
+      "num_input_tokens_seen": 27443200,
+      "step": 3350
+    },
+    {
+      "epoch": 0.3844833504977686,
+      "grad_norm": 0.5523999929428101,
+      "learning_rate": 3.93784108222715e-05,
+      "loss": 1.6298,
+      "num_input_tokens_seen": 27525120,
+      "step": 3360
+    },
+    {
+      "epoch": 0.3856276461837739,
+      "grad_norm": 0.5067601799964905,
+      "learning_rate": 3.929527140609115e-05,
+      "loss": 1.2188,
+      "num_input_tokens_seen": 27607040,
+      "step": 3370
+    },
+    {
+      "epoch": 0.38677194186977915,
+      "grad_norm": 0.5155417919158936,
+      "learning_rate": 3.921189641889163e-05,
+      "loss": 1.2925,
+      "num_input_tokens_seen": 27688960,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3879162375557844,
+      "grad_norm": 0.5797560811042786,
+      "learning_rate": 3.912828723460495e-05,
+      "loss": 1.3113,
+      "num_input_tokens_seen": 27770880,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3890605332417897,
+      "grad_norm": 0.5342340469360352,
+      "learning_rate": 3.904444523102242e-05,
+      "loss": 1.2155,
+      "num_input_tokens_seen": 27852800,
+      "step": 3400
+    },
+    {
+      "epoch": 0.39020482892779496,
+      "grad_norm": 0.867396354675293,
+      "learning_rate": 3.896037178977196e-05,
+      "loss": 1.4058,
+      "num_input_tokens_seen": 27934720,
+      "step": 3410
+    },
+    {
+      "epoch": 0.3913491246138002,
+      "grad_norm": 0.904750645160675,
+      "learning_rate": 3.887606829629536e-05,
+      "loss": 1.2582,
+      "num_input_tokens_seen": 28016640,
+      "step": 3420
+    },
+    {
+      "epoch": 0.3924934202998055,
+      "grad_norm": 0.49150246381759644,
+      "learning_rate": 3.87915361398254e-05,
+      "loss": 1.2471,
+      "num_input_tokens_seen": 28098560,
+      "step": 3430
+    },
+    {
+      "epoch": 0.3936377159858107,
+      "grad_norm": 1.6605616807937622,
+      "learning_rate": 3.8706776713363025e-05,
+      "loss": 1.2815,
+      "num_input_tokens_seen": 28180480,
+      "step": 3440
+    },
+    {
+      "epoch": 0.394782011671816,
+      "grad_norm": 0.5466508865356445,
+      "learning_rate": 3.862179141365431e-05,
+      "loss": 1.4604,
+      "num_input_tokens_seen": 28262400,
+      "step": 3450
+    },
+    {
+      "epoch": 0.39592630735782125,
+      "grad_norm": 0.8647962808609009,
+      "learning_rate": 3.8536581641167506e-05,
+      "loss": 1.2577,
+      "num_input_tokens_seen": 28344320,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3970706030438265,
+      "grad_norm": 2.0921196937561035,
+      "learning_rate": 3.845114880006994e-05,
+      "loss": 1.2211,
+      "num_input_tokens_seen": 28426240,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3982148987298318,
+      "grad_norm": 1.0453864336013794,
+      "learning_rate": 3.836549429820485e-05,
+      "loss": 1.2543,
+      "num_input_tokens_seen": 28508160,
+      "step": 3480
+    },
+    {
+      "epoch": 0.39935919441583706,
+      "grad_norm": 1.259315013885498,
+      "learning_rate": 3.827961954706825e-05,
+      "loss": 1.2716,
+      "num_input_tokens_seen": 28590080,
+      "step": 3490
+    },
+    {
+      "epoch": 0.40050349010184233,
+      "grad_norm": 0.49976375699043274,
+      "learning_rate": 3.8193525961785584e-05,
+      "loss": 0.9854,
+      "num_input_tokens_seen": 28672000,
+      "step": 3500
+    },
+    {
+      "epoch": 0.4016477857878476,
+      "grad_norm": 0.4869581162929535,
+      "learning_rate": 3.81072149610885e-05,
+      "loss": 1.1416,
+      "num_input_tokens_seen": 28753920,
+      "step": 3510
+    },
+    {
+      "epoch": 0.40279208147385287,
+      "grad_norm": 0.5591604113578796,
+      "learning_rate": 3.802068796729139e-05,
+      "loss": 1.3587,
+      "num_input_tokens_seen": 28835840,
+      "step": 3520
+    },
+    {
+      "epoch": 0.4039363771598581,
+      "grad_norm": 0.5056213736534119,
+      "learning_rate": 3.7933946406268e-05,
+      "loss": 1.2926,
+      "num_input_tokens_seen": 28917760,
+      "step": 3530
+    },
+    {
+      "epoch": 0.40508067284586335,
+      "grad_norm": 0.4954633414745331,
+      "learning_rate": 3.7846991707427905e-05,
+      "loss": 1.2945,
+      "num_input_tokens_seen": 28999680,
+      "step": 3540
+    },
+    {
+      "epoch": 0.4062249685318686,
+      "grad_norm": 0.823973536491394,
+      "learning_rate": 3.775982530369298e-05,
+      "loss": 1.2348,
+      "num_input_tokens_seen": 29081600,
+      "step": 3550
+    },
+    {
+      "epoch": 0.4073692642178739,
+      "grad_norm": 0.5769624710083008,
+      "learning_rate": 3.767244863147377e-05,
+      "loss": 1.2022,
+      "num_input_tokens_seen": 29163520,
+      "step": 3560
+    },
+    {
+      "epoch": 0.40851355990387916,
+      "grad_norm": 0.5166818499565125,
+      "learning_rate": 3.75848631306458e-05,
+      "loss": 1.5009,
+      "num_input_tokens_seen": 29245440,
+      "step": 3570
+    },
+    {
+      "epoch": 0.40965785558988443,
+      "grad_norm": 0.8839643001556396,
+      "learning_rate": 3.7497070244525925e-05,
+      "loss": 1.2186,
+      "num_input_tokens_seen": 29327360,
+      "step": 3580
+    },
+    {
+      "epoch": 0.4108021512758897,
+      "grad_norm": 0.5029737949371338,
+      "learning_rate": 3.7409071419848436e-05,
+      "loss": 1.3161,
+      "num_input_tokens_seen": 29409280,
+      "step": 3590
+    },
+    {
+      "epoch": 0.411946446961895,
+      "grad_norm": 0.46834880113601685,
+      "learning_rate": 3.73208681067413e-05,
+      "loss": 1.0534,
+      "num_input_tokens_seen": 29491200,
+      "step": 3600
+    },
+    {
+      "epoch": 0.41309074264790024,
+      "grad_norm": 0.6240310668945312,
+      "learning_rate": 3.7232461758702244e-05,
+      "loss": 1.3398,
+      "num_input_tokens_seen": 29573120,
+      "step": 3610
+    },
+    {
+      "epoch": 0.41423503833390546,
+      "grad_norm": 0.5298212766647339,
+      "learning_rate": 3.714385383257477e-05,
+      "loss": 1.1538,
+      "num_input_tokens_seen": 29655040,
+      "step": 3620
+    },
+    {
+      "epoch": 0.41537933401991073,
+      "grad_norm": 0.46233418583869934,
+      "learning_rate": 3.7055045788524214e-05,
+      "loss": 1.0701,
+      "num_input_tokens_seen": 29736960,
+      "step": 3630
+    },
+    {
+      "epoch": 0.416523629705916,
+      "grad_norm": 0.5719970464706421,
+      "learning_rate": 3.696603909001361e-05,
+      "loss": 1.4487,
+      "num_input_tokens_seen": 29818880,
+      "step": 3640
+    },
+    {
+      "epoch": 0.41766792539192127,
+      "grad_norm": 0.5453386902809143,
+      "learning_rate": 3.6876835203779615e-05,
+      "loss": 1.3053,
+      "num_input_tokens_seen": 29900800,
+      "step": 3650
+    },
+    {
+      "epoch": 0.41881222107792654,
+      "grad_norm": 0.4228511452674866,
+      "learning_rate": 3.678743559980835e-05,
+      "loss": 1.4515,
+      "num_input_tokens_seen": 29982720,
+      "step": 3660
+    },
+    {
+      "epoch": 0.4199565167639318,
+      "grad_norm": 0.5006650686264038,
+      "learning_rate": 3.669784175131115e-05,
+      "loss": 1.4384,
+      "num_input_tokens_seen": 30064640,
+      "step": 3670
+    },
+    {
+      "epoch": 0.4211008124499371,
+      "grad_norm": 0.49229928851127625,
+      "learning_rate": 3.660805513470027e-05,
+      "loss": 1.4419,
+      "num_input_tokens_seen": 30146560,
+      "step": 3680
+    },
+    {
+      "epoch": 0.42224510813594235,
+      "grad_norm": 0.5263097286224365,
+      "learning_rate": 3.651807722956462e-05,
+      "loss": 1.2246,
+      "num_input_tokens_seen": 30228480,
+      "step": 3690
+    },
+    {
+      "epoch": 0.4233894038219476,
+      "grad_norm": 0.48281192779541016,
+      "learning_rate": 3.642790951864532e-05,
+      "loss": 1.0901,
+      "num_input_tokens_seen": 30310400,
+      "step": 3700
+    },
+    {
+      "epoch": 0.42453369950795283,
+      "grad_norm": 0.6877309083938599,
+      "learning_rate": 3.63375534878113e-05,
+      "loss": 1.2148,
+      "num_input_tokens_seen": 30392320,
+      "step": 3710
+    },
+    {
+      "epoch": 0.4256779951939581,
+      "grad_norm": 2.1429624557495117,
+      "learning_rate": 3.6247010626034795e-05,
+      "loss": 1.3764,
+      "num_input_tokens_seen": 30474240,
+      "step": 3720
+    },
+    {
+      "epoch": 0.4268222908799634,
+      "grad_norm": 0.6136846542358398,
+      "learning_rate": 3.615628242536682e-05,
+      "loss": 1.216,
+      "num_input_tokens_seen": 30556160,
+      "step": 3730
+    },
+    {
+      "epoch": 0.42796658656596864,
+      "grad_norm": 0.5359023809432983,
+      "learning_rate": 3.6065370380912587e-05,
+      "loss": 1.2571,
+      "num_input_tokens_seen": 30638080,
+      "step": 3740
+    },
+    {
+      "epoch": 0.4291108822519739,
+      "grad_norm": 1.0357657670974731,
+      "learning_rate": 3.5974275990806846e-05,
+      "loss": 1.1939,
+      "num_input_tokens_seen": 30720000,
+      "step": 3750
+    },
+    {
+      "epoch": 0.4302551779379792,
+      "grad_norm": 0.5274159908294678,
+      "learning_rate": 3.588300075618922e-05,
+      "loss": 1.1014,
+      "num_input_tokens_seen": 30801920,
+      "step": 3760
+    },
+    {
+      "epoch": 0.43139947362398445,
+      "grad_norm": 0.48430922627449036,
+      "learning_rate": 3.579154618117946e-05,
+      "loss": 1.1029,
+      "num_input_tokens_seen": 30883840,
+      "step": 3770
+    },
+    {
+      "epoch": 0.4325437693099897,
+      "grad_norm": 0.5668869018554688,
+      "learning_rate": 3.5699913772852664e-05,
+      "loss": 1.3753,
+      "num_input_tokens_seen": 30965760,
+      "step": 3780
+    },
+    {
+      "epoch": 0.43368806499599494,
+      "grad_norm": 1.2524999380111694,
+      "learning_rate": 3.560810504121441e-05,
+      "loss": 1.405,
+      "num_input_tokens_seen": 31047680,
+      "step": 3790
+    },
+    {
+      "epoch": 0.4348323606820002,
+      "grad_norm": 0.49741896986961365,
+      "learning_rate": 3.551612149917593e-05,
+      "loss": 1.1393,
+      "num_input_tokens_seen": 31129600,
+      "step": 3800
+    },
+    {
+      "epoch": 0.4359766563680055,
+      "grad_norm": 0.5412270426750183,
+      "learning_rate": 3.542396466252913e-05,
+      "loss": 1.3569,
+      "num_input_tokens_seen": 31211520,
+      "step": 3810
+    },
+    {
+      "epoch": 0.43712095205401075,
+      "grad_norm": 2.5141255855560303,
+      "learning_rate": 3.533163604992163e-05,
+      "loss": 1.4194,
+      "num_input_tokens_seen": 31293440,
+      "step": 3820
+    },
+    {
+      "epoch": 0.438265247740016,
+      "grad_norm": 0.5513648390769958,
+      "learning_rate": 3.523913718283175e-05,
+      "loss": 1.2899,
+      "num_input_tokens_seen": 31375360,
+      "step": 3830
+    },
+    {
+      "epoch": 0.4394095434260213,
+      "grad_norm": 0.583692729473114,
+      "learning_rate": 3.514646958554339e-05,
+      "loss": 1.3624,
+      "num_input_tokens_seen": 31457280,
+      "step": 3840
+    },
+    {
+      "epoch": 0.44055383911202656,
+      "grad_norm": 0.5681483149528503,
+      "learning_rate": 3.5053634785121e-05,
+      "loss": 1.1517,
+      "num_input_tokens_seen": 31539200,
+      "step": 3850
+    },
+    {
+      "epoch": 0.4416981347980318,
+      "grad_norm": 0.4962402880191803,
+      "learning_rate": 3.496063431138431e-05,
+      "loss": 1.2125,
+      "num_input_tokens_seen": 31621120,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4428424304840371,
+      "grad_norm": 0.4531221389770508,
+      "learning_rate": 3.4867469696883204e-05,
+      "loss": 1.287,
+      "num_input_tokens_seen": 31703040,
+      "step": 3870
+    },
+    {
+      "epoch": 0.4439867261700423,
+      "grad_norm": 0.4687143564224243,
+      "learning_rate": 3.477414247687241e-05,
+      "loss": 1.0745,
+      "num_input_tokens_seen": 31784960,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4451310218560476,
+      "grad_norm": 0.5784918665885925,
+      "learning_rate": 3.468065418928625e-05,
+      "loss": 1.146,
+      "num_input_tokens_seen": 31866880,
+      "step": 3890
+    },
+    {
+      "epoch": 0.44627531754205285,
+      "grad_norm": 0.4997071623802185,
+      "learning_rate": 3.458700637471325e-05,
+      "loss": 1.0793,
+      "num_input_tokens_seen": 31948800,
+      "step": 3900
+    },
+    {
+      "epoch": 0.4474196132280581,
+      "grad_norm": 0.5425757765769958,
+      "learning_rate": 3.4493200576370776e-05,
+      "loss": 1.1415,
+      "num_input_tokens_seen": 32030720,
+      "step": 3910
+    },
+    {
+      "epoch": 0.4485639089140634,
+      "grad_norm": 0.5466029644012451,
+      "learning_rate": 3.4399238340079607e-05,
+      "loss": 1.2075,
+      "num_input_tokens_seen": 32112640,
+      "step": 3920
+    },
+    {
+      "epoch": 0.44970820460006866,
+      "grad_norm": 0.5251049995422363,
+      "learning_rate": 3.4305121214238446e-05,
+      "loss": 1.1582,
+      "num_input_tokens_seen": 32194560,
+      "step": 3930
+    },
+    {
+      "epoch": 0.45085250028607393,
+      "grad_norm": 0.5477258563041687,
+      "learning_rate": 3.4210850749798415e-05,
+      "loss": 1.3309,
+      "num_input_tokens_seen": 32276480,
+      "step": 3940
+    },
+    {
+      "epoch": 0.4519967959720792,
+      "grad_norm": 0.621071457862854,
+      "learning_rate": 3.411642850023751e-05,
+      "loss": 1.2072,
+      "num_input_tokens_seen": 32358400,
+      "step": 3950
+    },
+    {
+      "epoch": 0.45314109165808447,
+      "grad_norm": 0.5287960767745972,
+      "learning_rate": 3.402185602153495e-05,
+      "loss": 1.1731,
+      "num_input_tokens_seen": 32440320,
+      "step": 3960
+    },
+    {
+      "epoch": 0.4542853873440897,
+      "grad_norm": 0.5123234987258911,
+      "learning_rate": 3.392713487214561e-05,
+      "loss": 1.1238,
+      "num_input_tokens_seen": 32522240,
+      "step": 3970
+    },
+    {
+      "epoch": 0.45542968303009496,
+      "grad_norm": 0.5226893424987793,
+      "learning_rate": 3.38322666129743e-05,
+      "loss": 1.1262,
+      "num_input_tokens_seen": 32604160,
+      "step": 3980
+    },
+    {
+      "epoch": 0.4565739787161002,
+      "grad_norm": 0.6006979942321777,
+      "learning_rate": 3.373725280735e-05,
+      "loss": 1.2084,
+      "num_input_tokens_seen": 32686080,
+      "step": 3990
+    },
+    {
+      "epoch": 0.4577182744021055,
+      "grad_norm": 0.475813627243042,
+      "learning_rate": 3.3642095021000184e-05,
+      "loss": 1.2836,
+      "num_input_tokens_seen": 32768000,
+      "step": 4000
+    },
+    {
+      "epoch": 0.45886257008811077,
+      "grad_norm": 0.5431310534477234,
+      "learning_rate": 3.3546794822024976e-05,
+      "loss": 1.295,
+      "num_input_tokens_seen": 32849920,
+      "step": 4010
+    },
+    {
+      "epoch": 0.46000686577411604,
+      "grad_norm": 0.48410120606422424,
+      "learning_rate": 3.3451353780871286e-05,
+      "loss": 1.3341,
+      "num_input_tokens_seen": 32931840,
+      "step": 4020
+    },
+    {
+      "epoch": 0.4611511614601213,
+      "grad_norm": 0.525595486164093,
+      "learning_rate": 3.335577347030697e-05,
+      "loss": 1.2892,
+      "num_input_tokens_seen": 33013760,
+      "step": 4030
+    },
+    {
+      "epoch": 0.4622954571461266,
+      "grad_norm": 0.5761798620223999,
+      "learning_rate": 3.32600554653949e-05,
+      "loss": 1.5631,
+      "num_input_tokens_seen": 33095680,
+      "step": 4040
+    },
+    {
+      "epoch": 0.46343975283213185,
+      "grad_norm": 0.48856690526008606,
+      "learning_rate": 3.316420134346701e-05,
+      "loss": 0.9705,
+      "num_input_tokens_seen": 33177600,
+      "step": 4050
+    },
+    {
+      "epoch": 0.46458404851813706,
+      "grad_norm": 0.7112519145011902,
+      "learning_rate": 3.306821268409827e-05,
+      "loss": 1.3763,
+      "num_input_tokens_seen": 33259520,
+      "step": 4060
+    },
+    {
+      "epoch": 0.46572834420414233,
+      "grad_norm": 0.9292272925376892,
+      "learning_rate": 3.297209106908072e-05,
+      "loss": 1.2818,
+      "num_input_tokens_seen": 33341440,
+      "step": 4070
+    },
+    {
+      "epoch": 0.4668726398901476,
+      "grad_norm": 0.7249779105186462,
+      "learning_rate": 3.287583808239735e-05,
+      "loss": 1.1623,
+      "num_input_tokens_seen": 33423360,
+      "step": 4080
+    },
+    {
+      "epoch": 0.46801693557615287,
+      "grad_norm": 0.5191545486450195,
+      "learning_rate": 3.277945531019601e-05,
+      "loss": 1.5152,
+      "num_input_tokens_seen": 33505280,
+      "step": 4090
+    },
+    {
+      "epoch": 0.46916123126215814,
+      "grad_norm": 0.5034803748130798,
+      "learning_rate": 3.268294434076332e-05,
+      "loss": 1.5291,
+      "num_input_tokens_seen": 33587200,
+      "step": 4100
+    },
+    {
+      "epoch": 0.4703055269481634,
+      "grad_norm": 0.5027021765708923,
+      "learning_rate": 3.2586306764498395e-05,
+      "loss": 1.1453,
+      "num_input_tokens_seen": 33669120,
+      "step": 4110
+    },
+    {
+      "epoch": 0.4714498226341687,
+      "grad_norm": 0.6401455998420715,
+      "learning_rate": 3.2489544173886745e-05,
+      "loss": 1.398,
+      "num_input_tokens_seen": 33751040,
+      "step": 4120
+    },
+    {
+      "epoch": 0.47259411832017395,
+      "grad_norm": 0.34582868218421936,
+      "learning_rate": 3.239265816347397e-05,
+      "loss": 1.0819,
+      "num_input_tokens_seen": 33832960,
+      "step": 4130
+    },
+    {
+      "epoch": 0.4737384140061792,
+      "grad_norm": 0.5344799757003784,
+      "learning_rate": 3.2295650329839474e-05,
+      "loss": 1.3502,
+      "num_input_tokens_seen": 33914880,
+      "step": 4140
+    },
+    {
+      "epoch": 0.47488270969218443,
+      "grad_norm": 0.48560017347335815,
+      "learning_rate": 3.219852227157022e-05,
+      "loss": 1.1419,
+      "num_input_tokens_seen": 33996800,
+      "step": 4150
+    },
+    {
+      "epoch": 0.4760270053781897,
+      "grad_norm": 0.6051056981086731,
+      "learning_rate": 3.210127558923434e-05,
+      "loss": 1.3278,
+      "num_input_tokens_seen": 34078720,
+      "step": 4160
+    },
+    {
+      "epoch": 0.477171301064195,
+      "grad_norm": 0.47563403844833374,
+      "learning_rate": 3.200391188535472e-05,
+      "loss": 1.2711,
+      "num_input_tokens_seen": 34160640,
+      "step": 4170
+    },
+    {
+      "epoch": 0.47831559675020024,
+      "grad_norm": 0.5107710361480713,
+      "learning_rate": 3.1906432764382695e-05,
+      "loss": 1.4901,
+      "num_input_tokens_seen": 34242560,
+      "step": 4180
+    },
+    {
+      "epoch": 0.4794598924362055,
+      "grad_norm": 0.5463585257530212,
+      "learning_rate": 3.1808839832671523e-05,
+      "loss": 1.4351,
+      "num_input_tokens_seen": 34324480,
+      "step": 4190
+    },
+    {
+      "epoch": 0.4806041881222108,
+      "grad_norm": 0.4871821403503418,
+      "learning_rate": 3.1711134698449946e-05,
+      "loss": 1.1299,
+      "num_input_tokens_seen": 34406400,
+      "step": 4200
+    },
+    {
+      "epoch": 0.48174848380821605,
+      "grad_norm": 0.5575620532035828,
+      "learning_rate": 3.161331897179568e-05,
+      "loss": 1.3597,
+      "num_input_tokens_seen": 34488320,
+      "step": 4210
+    },
+    {
+      "epoch": 0.4828927794942213,
+      "grad_norm": 0.5072765350341797,
+      "learning_rate": 3.151539426460892e-05,
+      "loss": 1.4977,
+      "num_input_tokens_seen": 34570240,
+      "step": 4220
+    },
+    {
+      "epoch": 0.4840370751802266,
+      "grad_norm": 0.517631471157074,
+      "learning_rate": 3.14173621905857e-05,
+      "loss": 1.1672,
+      "num_input_tokens_seen": 34652160,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4851813708662318,
+      "grad_norm": 0.5261096954345703,
+      "learning_rate": 3.1319224365191366e-05,
+      "loss": 1.495,
+      "num_input_tokens_seen": 34734080,
+      "step": 4240
+    },
+    {
+      "epoch": 0.4863256665522371,
+      "grad_norm": 0.5559542179107666,
+      "learning_rate": 3.122098240563396e-05,
+      "loss": 1.6386,
+      "num_input_tokens_seen": 34816000,
+      "step": 4250
+    },
+    {
+      "epoch": 0.48746996223824235,
+      "grad_norm": 0.8691073060035706,
+      "learning_rate": 3.1122637930837486e-05,
+      "loss": 1.2742,
+      "num_input_tokens_seen": 34897920,
+      "step": 4260
+    },
+    {
+      "epoch": 0.4886142579242476,
+      "grad_norm": 0.5249228477478027,
+      "learning_rate": 3.102419256141536e-05,
+      "loss": 1.2173,
+      "num_input_tokens_seen": 34979840,
+      "step": 4270
+    },
+    {
+      "epoch": 0.4897585536102529,
+      "grad_norm": 0.5527949333190918,
+      "learning_rate": 3.092564791964358e-05,
+      "loss": 1.187,
+      "num_input_tokens_seen": 35061760,
+      "step": 4280
+    },
+    {
+      "epoch": 0.49090284929625816,
+      "grad_norm": 0.48992666602134705,
+      "learning_rate": 3.082700562943409e-05,
+      "loss": 1.5849,
+      "num_input_tokens_seen": 35143680,
+      "step": 4290
+    },
+    {
+      "epoch": 0.49204714498226343,
+      "grad_norm": 0.4628704786300659,
+      "learning_rate": 3.0728267316307945e-05,
+      "loss": 1.2716,
+      "num_input_tokens_seen": 35225600,
+      "step": 4300
+    },
+    {
+      "epoch": 0.4931914406682687,
+      "grad_norm": 1.0824905633926392,
+      "learning_rate": 3.062943460736857e-05,
+      "loss": 1.2538,
+      "num_input_tokens_seen": 35307520,
+      "step": 4310
+    },
+    {
+      "epoch": 0.49433573635427397,
+      "grad_norm": 0.5195997953414917,
+      "learning_rate": 3.0530509131274935e-05,
+      "loss": 1.1509,
+      "num_input_tokens_seen": 35389440,
+      "step": 4320
+    },
+    {
+      "epoch": 0.4954800320402792,
+      "grad_norm": 0.5249128937721252,
+      "learning_rate": 3.04314925182147e-05,
+      "loss": 1.1693,
+      "num_input_tokens_seen": 35471360,
+      "step": 4330
+    },
+    {
+      "epoch": 0.49662432772628445,
+      "grad_norm": 0.5838843584060669,
+      "learning_rate": 3.03323863998774e-05,
+      "loss": 1.2192,
+      "num_input_tokens_seen": 35553280,
+      "step": 4340
+    },
+    {
+      "epoch": 0.4977686234122897,
+      "grad_norm": 0.5815695524215698,
+      "learning_rate": 3.0233192409427492e-05,
+      "loss": 1.2383,
+      "num_input_tokens_seen": 35635200,
+      "step": 4350
+    },
+    {
+      "epoch": 0.498912919098295,
+      "grad_norm": 0.47992345690727234,
+      "learning_rate": 3.0133912181477475e-05,
+      "loss": 1.1951,
+      "num_input_tokens_seen": 35717120,
+      "step": 4360
+    },
+    {
+      "epoch": 0.5000572147843002,
+      "grad_norm": 0.5813825726509094,
+      "learning_rate": 3.003454735206097e-05,
+      "loss": 1.4106,
+      "num_input_tokens_seen": 35799040,
+      "step": 4370
+    },
+    {
+      "epoch": 0.5012015104703055,
+      "grad_norm": 0.5316118597984314,
+      "learning_rate": 2.9935099558605728e-05,
+      "loss": 1.2978,
+      "num_input_tokens_seen": 35880960,
+      "step": 4380
+    },
+    {
+      "epoch": 0.5023458061563107,
+      "grad_norm": 0.5098001956939697,
+      "learning_rate": 2.9835570439906657e-05,
+      "loss": 1.1646,
+      "num_input_tokens_seen": 35962880,
+      "step": 4390
+    },
+    {
+      "epoch": 0.503490101842316,
+      "grad_norm": 0.482377290725708,
+      "learning_rate": 2.973596163609883e-05,
+      "loss": 1.2874,
+      "num_input_tokens_seen": 36044800,
+      "step": 4400
+    },
+    {
+      "epoch": 0.5046343975283213,
+      "grad_norm": 0.5183489322662354,
+      "learning_rate": 2.9636274788630437e-05,
+      "loss": 1.2198,
+      "num_input_tokens_seen": 36126720,
+      "step": 4410
+    },
+    {
+      "epoch": 0.5057786932143266,
+      "grad_norm": 0.4759490191936493,
+      "learning_rate": 2.9536511540235744e-05,
+      "loss": 1.2435,
+      "num_input_tokens_seen": 36208640,
+      "step": 4420
+    },
+    {
+      "epoch": 0.5069229889003318,
+      "grad_norm": 0.5005761384963989,
+      "learning_rate": 2.9436673534908044e-05,
+      "loss": 1.367,
+      "num_input_tokens_seen": 36290560,
+      "step": 4430
+    },
+    {
+      "epoch": 0.5080672845863371,
+      "grad_norm": 0.48932281136512756,
+      "learning_rate": 2.9336762417872516e-05,
+      "loss": 1.1908,
+      "num_input_tokens_seen": 36372480,
+      "step": 4440
+    },
+    {
+      "epoch": 0.5092115802723424,
+      "grad_norm": 0.49353188276290894,
+      "learning_rate": 2.9236779835559165e-05,
+      "loss": 1.2709,
+      "num_input_tokens_seen": 36454400,
+      "step": 4450
+    },
+    {
+      "epoch": 0.5103558759583476,
+      "grad_norm": 0.5292254686355591,
+      "learning_rate": 2.913672743557565e-05,
+      "loss": 1.1411,
+      "num_input_tokens_seen": 36536320,
+      "step": 4460
+    },
+    {
+      "epoch": 0.5115001716443529,
+      "grad_norm": 0.4894375801086426,
+      "learning_rate": 2.9036606866680187e-05,
+      "loss": 1.1795,
+      "num_input_tokens_seen": 36618240,
+      "step": 4470
+    },
+    {
+      "epoch": 0.5126444673303582,
+      "grad_norm": 0.7328277826309204,
+      "learning_rate": 2.8936419778754294e-05,
+      "loss": 1.4518,
+      "num_input_tokens_seen": 36700160,
+      "step": 4480
+    },
+    {
+      "epoch": 0.5137887630163634,
+      "grad_norm": 0.5245153903961182,
+      "learning_rate": 2.883616782277569e-05,
+      "loss": 1.1565,
+      "num_input_tokens_seen": 36782080,
+      "step": 4490
+    },
+    {
+      "epoch": 0.5149330587023687,
+      "grad_norm": 0.5825649499893188,
+      "learning_rate": 2.8735852650791035e-05,
+      "loss": 1.1245,
+      "num_input_tokens_seen": 36864000,
+      "step": 4500
+    },
+    {
+      "epoch": 0.516077354388374,
+      "grad_norm": 0.4824535548686981,
+      "learning_rate": 2.8635475915888732e-05,
+      "loss": 1.2498,
+      "num_input_tokens_seen": 36945920,
+      "step": 4510
+    },
+    {
+      "epoch": 0.5172216500743793,
+      "grad_norm": 0.51859050989151,
+      "learning_rate": 2.853503927217167e-05,
+      "loss": 1.222,
+      "num_input_tokens_seen": 37027840,
+      "step": 4520
+    },
+    {
+      "epoch": 0.5183659457603845,
+      "grad_norm": 0.4794926345348358,
+      "learning_rate": 2.8434544374729965e-05,
+      "loss": 1.3003,
+      "num_input_tokens_seen": 37109760,
+      "step": 4530
+    },
+    {
+      "epoch": 0.5195102414463898,
+      "grad_norm": 0.8347005248069763,
+      "learning_rate": 2.8333992879613712e-05,
+      "loss": 1.1049,
+      "num_input_tokens_seen": 37191680,
+      "step": 4540
+    },
+    {
+      "epoch": 0.520654537132395,
+      "grad_norm": 0.5049195289611816,
+      "learning_rate": 2.823338644380566e-05,
+      "loss": 1.2885,
+      "num_input_tokens_seen": 37273600,
+      "step": 4550
+    },
+    {
+      "epoch": 0.5217988328184002,
+      "grad_norm": 0.6773023009300232,
+      "learning_rate": 2.8132726725193926e-05,
+      "loss": 1.3466,
+      "num_input_tokens_seen": 37355520,
+      "step": 4560
+    },
+    {
+      "epoch": 0.5229431285044055,
+      "grad_norm": 0.4892396628856659,
+      "learning_rate": 2.803201538254467e-05,
+      "loss": 1.3192,
+      "num_input_tokens_seen": 37437440,
+      "step": 4570
+    },
+    {
+      "epoch": 0.5240874241904108,
+      "grad_norm": 0.5198492407798767,
+      "learning_rate": 2.7931254075474768e-05,
+      "loss": 1.2215,
+      "num_input_tokens_seen": 37519360,
+      "step": 4580
+    },
+    {
+      "epoch": 0.525231719876416,
+      "grad_norm": 0.4889214336872101,
+      "learning_rate": 2.7830444464424466e-05,
+      "loss": 1.1609,
+      "num_input_tokens_seen": 37601280,
+      "step": 4590
+    },
+    {
+      "epoch": 0.5263760155624213,
+      "grad_norm": 0.512380838394165,
+      "learning_rate": 2.772958821062997e-05,
+      "loss": 1.0642,
+      "num_input_tokens_seen": 37683200,
+      "step": 4600
+    },
+    {
+      "epoch": 0.5275203112484266,
+      "grad_norm": 0.792269766330719,
+      "learning_rate": 2.7628686976096164e-05,
+      "loss": 1.1454,
+      "num_input_tokens_seen": 37765120,
+      "step": 4610
+    },
+    {
+      "epoch": 0.5286646069344318,
+      "grad_norm": 0.5080438256263733,
+      "learning_rate": 2.7527742423569124e-05,
+      "loss": 1.1747,
+      "num_input_tokens_seen": 37847040,
+      "step": 4620
+    },
+    {
+      "epoch": 0.5298089026204371,
+      "grad_norm": 0.494386225938797,
+      "learning_rate": 2.7426756216508776e-05,
+      "loss": 1.1861,
+      "num_input_tokens_seen": 37928960,
+      "step": 4630
+    },
+    {
+      "epoch": 0.5309531983064424,
+      "grad_norm": 0.7930368185043335,
+      "learning_rate": 2.7325730019061474e-05,
+      "loss": 1.3794,
+      "num_input_tokens_seen": 38010880,
+      "step": 4640
+    },
+    {
+      "epoch": 0.5320974939924477,
+      "grad_norm": 0.5523006319999695,
+      "learning_rate": 2.7224665496032565e-05,
+      "loss": 1.1395,
+      "num_input_tokens_seen": 38092800,
+      "step": 4650
+    },
+    {
+      "epoch": 0.5332417896784529,
+      "grad_norm": 0.6003334522247314,
+      "learning_rate": 2.712356431285896e-05,
+      "loss": 1.1517,
+      "num_input_tokens_seen": 38174720,
+      "step": 4660
+    },
+    {
+      "epoch": 0.5343860853644582,
+      "grad_norm": 0.5045771598815918,
+      "learning_rate": 2.70224281355817e-05,
+      "loss": 1.2753,
+      "num_input_tokens_seen": 38256640,
+      "step": 4670
+    },
+    {
+      "epoch": 0.5355303810504635,
+      "grad_norm": 0.5754336714744568,
+      "learning_rate": 2.6921258630818475e-05,
+      "loss": 1.1533,
+      "num_input_tokens_seen": 38338560,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5366746767364687,
+      "grad_norm": 0.5257157683372498,
+      "learning_rate": 2.6820057465736197e-05,
+      "loss": 1.1615,
+      "num_input_tokens_seen": 38420480,
+      "step": 4690
+    },
+    {
+      "epoch": 0.537818972422474,
+      "grad_norm": 0.4963821768760681,
+      "learning_rate": 2.6718826308023487e-05,
+      "loss": 1.1366,
+      "num_input_tokens_seen": 38502400,
+      "step": 4700
+    },
+    {
+      "epoch": 0.5389632681084793,
+      "grad_norm": 0.5854734778404236,
+      "learning_rate": 2.6617566825863237e-05,
+      "loss": 1.129,
+      "num_input_tokens_seen": 38584320,
+      "step": 4710
+    },
+    {
+      "epoch": 0.5401075637944845,
+      "grad_norm": 0.5400436520576477,
+      "learning_rate": 2.651628068790507e-05,
+      "loss": 1.0386,
+      "num_input_tokens_seen": 38666240,
+      "step": 4720
+    },
+    {
+      "epoch": 0.5412518594804897,
+      "grad_norm": 1.165490984916687,
+      "learning_rate": 2.6414969563237874e-05,
+      "loss": 1.0903,
+      "num_input_tokens_seen": 38748160,
+      "step": 4730
+    },
+    {
+      "epoch": 0.542396155166495,
+      "grad_norm": 1.1871403455734253,
+      "learning_rate": 2.6313635121362322e-05,
+      "loss": 1.2361,
+      "num_input_tokens_seen": 38830080,
+      "step": 4740
+    },
+    {
+      "epoch": 0.5435404508525002,
+      "grad_norm": 0.5504677295684814,
+      "learning_rate": 2.6212279032163283e-05,
+      "loss": 1.5279,
+      "num_input_tokens_seen": 38912000,
+      "step": 4750
+    },
+    {
+      "epoch": 0.5446847465385055,
+      "grad_norm": 0.9438173174858093,
+      "learning_rate": 2.6110902965882383e-05,
+      "loss": 1.3035,
+      "num_input_tokens_seen": 38993920,
+      "step": 4760
+    },
+    {
+      "epoch": 0.5458290422245108,
+      "grad_norm": 0.5121588110923767,
+      "learning_rate": 2.6009508593090448e-05,
+      "loss": 1.2656,
+      "num_input_tokens_seen": 39075840,
+      "step": 4770
+    },
+    {
+      "epoch": 0.546973337910516,
+      "grad_norm": 0.5496011972427368,
+      "learning_rate": 2.590809758465995e-05,
+      "loss": 1.3347,
+      "num_input_tokens_seen": 39157760,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5481176335965213,
+      "grad_norm": 0.5163994431495667,
+      "learning_rate": 2.580667161173753e-05,
+      "loss": 1.0816,
+      "num_input_tokens_seen": 39239680,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5492619292825266,
+      "grad_norm": 0.547443687915802,
+      "learning_rate": 2.570523234571642e-05,
+      "loss": 1.072,
+      "num_input_tokens_seen": 39321600,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5504062249685319,
+      "grad_norm": 0.49614182114601135,
+      "learning_rate": 2.5603781458208885e-05,
+      "loss": 1.105,
+      "num_input_tokens_seen": 39403520,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5515505206545371,
+      "grad_norm": 0.6012304425239563,
+      "learning_rate": 2.5502320621018732e-05,
+      "loss": 1.1652,
+      "num_input_tokens_seen": 39485440,
+      "step": 4820
+    },
+    {
+      "epoch": 0.5526948163405424,
+      "grad_norm": 1.310507893562317,
+      "learning_rate": 2.5400851506113728e-05,
+      "loss": 1.2073,
+      "num_input_tokens_seen": 39567360,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5538391120265477,
+      "grad_norm": 0.5216242671012878,
+      "learning_rate": 2.5299375785598005e-05,
+      "loss": 1.2457,
+      "num_input_tokens_seen": 39649280,
+      "step": 4840
+    },
+    {
+      "epoch": 0.554983407712553,
+      "grad_norm": 0.4636791944503784,
+      "learning_rate": 2.519789513168459e-05,
+      "loss": 1.3313,
+      "num_input_tokens_seen": 39731200,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5561277033985582,
+      "grad_norm": 0.4918244779109955,
+      "learning_rate": 2.509641121666781e-05,
+      "loss": 1.1312,
+      "num_input_tokens_seen": 39813120,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5572719990845635,
+      "grad_norm": 0.5367238521575928,
+      "learning_rate": 2.4994925712895697e-05,
+      "loss": 1.1891,
+      "num_input_tokens_seen": 39895040,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5584162947705688,
+      "grad_norm": 0.45426082611083984,
+      "learning_rate": 2.489344029274249e-05,
+      "loss": 1.4017,
+      "num_input_tokens_seen": 39976960,
+      "step": 4880
+    },
+    {
+      "epoch": 0.559560590456574,
+      "grad_norm": 1.55329167842865,
+      "learning_rate": 2.479195662858105e-05,
+      "loss": 1.1554,
+      "num_input_tokens_seen": 40058880,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5607048861425793,
+      "grad_norm": 0.5056139230728149,
+      "learning_rate": 2.4690476392755298e-05,
+      "loss": 1.3402,
+      "num_input_tokens_seen": 40140800,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5618491818285845,
+      "grad_norm": 0.45648014545440674,
+      "learning_rate": 2.4589001257552637e-05,
+      "loss": 1.288,
+      "num_input_tokens_seen": 40222720,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5629934775145897,
+      "grad_norm": 0.5516757369041443,
+      "learning_rate": 2.4487532895176457e-05,
+      "loss": 1.2969,
+      "num_input_tokens_seen": 40304640,
+      "step": 4920
+    },
+    {
+      "epoch": 0.564137773200595,
+      "grad_norm": 0.5544970631599426,
+      "learning_rate": 2.4386072977718503e-05,
+      "loss": 1.346,
+      "num_input_tokens_seen": 40386560,
+      "step": 4930
+    },
+    {
+      "epoch": 0.5652820688866003,
+      "grad_norm": 0.48032233119010925,
+      "learning_rate": 2.4284623177131395e-05,
+      "loss": 1.2043,
+      "num_input_tokens_seen": 40468480,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5664263645726055,
+      "grad_norm": 1.1765650510787964,
+      "learning_rate": 2.4183185165200998e-05,
+      "loss": 1.1418,
+      "num_input_tokens_seen": 40550400,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5675706602586108,
+      "grad_norm": 0.5061176419258118,
+      "learning_rate": 2.4081760613518924e-05,
+      "loss": 1.0886,
+      "num_input_tokens_seen": 40632320,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5687149559446161,
+      "grad_norm": 1.4397417306900024,
+      "learning_rate": 2.3980351193455e-05,
+      "loss": 1.3159,
+      "num_input_tokens_seen": 40714240,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5698592516306213,
+      "grad_norm": 0.46491971611976624,
+      "learning_rate": 2.3878958576129664e-05,
+      "loss": 1.2942,
+      "num_input_tokens_seen": 40796160,
+      "step": 4980
+    },
+    {
+      "epoch": 0.5710035473166266,
+      "grad_norm": 0.5516128540039062,
+      "learning_rate": 2.3777584432386474e-05,
+      "loss": 1.3472,
+      "num_input_tokens_seen": 40878080,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5721478430026319,
+      "grad_norm": 0.4662507176399231,
+      "learning_rate": 2.367623043276459e-05,
+      "loss": 1.1461,
+      "num_input_tokens_seen": 40960000,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5732921386886372,
+      "grad_norm": 0.4610395133495331,
+      "learning_rate": 2.3574898247471167e-05,
+      "loss": 0.997,
+      "num_input_tokens_seen": 41041920,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5744364343746424,
+      "grad_norm": 0.6839539408683777,
+      "learning_rate": 2.347358954635393e-05,
+      "loss": 1.2608,
+      "num_input_tokens_seen": 41123840,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5755807300606477,
+      "grad_norm": 0.5418996810913086,
+      "learning_rate": 2.337230599887358e-05,
+      "loss": 1.1886,
+      "num_input_tokens_seen": 41205760,
+      "step": 5030
+    },
+    {
+      "epoch": 0.576725025746653,
+      "grad_norm": 0.5929667949676514,
+      "learning_rate": 2.327104927407634e-05,
+      "loss": 1.2545,
+      "num_input_tokens_seen": 41287680,
+      "step": 5040
+    },
+    {
+      "epoch": 0.5778693214326582,
+      "grad_norm": 0.49478694796562195,
+      "learning_rate": 2.3169821040566387e-05,
+      "loss": 1.0687,
+      "num_input_tokens_seen": 41369600,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5790136171186635,
+      "grad_norm": 0.48581549525260925,
+      "learning_rate": 2.306862296647841e-05,
+      "loss": 1.1071,
+      "num_input_tokens_seen": 41451520,
+      "step": 5060
+    },
+    {
+      "epoch": 0.5801579128046688,
+      "grad_norm": 1.5436527729034424,
+      "learning_rate": 2.2967456719450127e-05,
+      "loss": 1.0824,
+      "num_input_tokens_seen": 41533440,
+      "step": 5070
+    },
+    {
+      "epoch": 0.581302208490674,
+      "grad_norm": 0.49023711681365967,
+      "learning_rate": 2.2866323966594736e-05,
+      "loss": 1.451,
+      "num_input_tokens_seen": 41615360,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5824465041766792,
+      "grad_norm": 0.48039644956588745,
+      "learning_rate": 2.2765226374473504e-05,
+      "loss": 1.2207,
+      "num_input_tokens_seen": 41697280,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5835907998626845,
+      "grad_norm": 0.46668538451194763,
+      "learning_rate": 2.2664165609068304e-05,
+      "loss": 1.1504,
+      "num_input_tokens_seen": 41779200,
+      "step": 5100
+    },
+    {
+      "epoch": 0.5847350955486897,
+      "grad_norm": 0.418855220079422,
+      "learning_rate": 2.2563143335754118e-05,
+      "loss": 1.1436,
+      "num_input_tokens_seen": 41861120,
+      "step": 5110
+    },
+    {
+      "epoch": 0.585879391234695,
+      "grad_norm": 0.6656162142753601,
+      "learning_rate": 2.2462161219271622e-05,
+      "loss": 1.5087,
+      "num_input_tokens_seen": 41943040,
+      "step": 5120
+    },
+    {
+      "epoch": 0.5870236869207003,
+      "grad_norm": 0.5571346282958984,
+      "learning_rate": 2.236122092369977e-05,
+      "loss": 1.2635,
+      "num_input_tokens_seen": 42024960,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5881679826067056,
+      "grad_norm": 0.5325060486793518,
+      "learning_rate": 2.2260324112428336e-05,
+      "loss": 1.4867,
+      "num_input_tokens_seen": 42106880,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5893122782927108,
+      "grad_norm": 0.48673689365386963,
+      "learning_rate": 2.2159472448130513e-05,
+      "loss": 1.2188,
+      "num_input_tokens_seen": 42188800,
+      "step": 5150
+    },
+    {
+      "epoch": 0.5904565739787161,
+      "grad_norm": 0.5229732394218445,
+      "learning_rate": 2.2058667592735532e-05,
+      "loss": 1.3696,
+      "num_input_tokens_seen": 42270720,
+      "step": 5160
+    },
+    {
+      "epoch": 0.5916008696647214,
+      "grad_norm": 0.6097173094749451,
+      "learning_rate": 2.1957911207401267e-05,
+      "loss": 1.6279,
+      "num_input_tokens_seen": 42352640,
+      "step": 5170
+    },
+    {
+      "epoch": 0.5927451653507266,
+      "grad_norm": 0.5959046483039856,
+      "learning_rate": 2.1857204952486824e-05,
+      "loss": 1.521,
+      "num_input_tokens_seen": 42434560,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5938894610367319,
+      "grad_norm": 0.7557221055030823,
+      "learning_rate": 2.1756550487525247e-05,
+      "loss": 1.3385,
+      "num_input_tokens_seen": 42516480,
+      "step": 5190
+    },
+    {
+      "epoch": 0.5950337567227372,
+      "grad_norm": 0.48483288288116455,
+      "learning_rate": 2.165594947119613e-05,
+      "loss": 1.1232,
+      "num_input_tokens_seen": 42598400,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5961780524087424,
+      "grad_norm": 0.48301464319229126,
+      "learning_rate": 2.1555403561298287e-05,
+      "loss": 1.3692,
+      "num_input_tokens_seen": 42680320,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5973223480947477,
+      "grad_norm": 0.4715411961078644,
+      "learning_rate": 2.1454914414722417e-05,
+      "loss": 1.2656,
+      "num_input_tokens_seen": 42762240,
+      "step": 5220
+    },
+    {
+      "epoch": 0.598466643780753,
+      "grad_norm": 0.4841952323913574,
+      "learning_rate": 2.135448368742385e-05,
+      "loss": 1.2838,
+      "num_input_tokens_seen": 42844160,
+      "step": 5230
+    },
+    {
+      "epoch": 0.5996109394667583,
+      "grad_norm": 0.4453716278076172,
+      "learning_rate": 2.1254113034395212e-05,
+      "loss": 1.2797,
+      "num_input_tokens_seen": 42926080,
+      "step": 5240
+    },
+    {
+      "epoch": 0.6007552351527635,
+      "grad_norm": 1.0583953857421875,
+      "learning_rate": 2.1153804109639157e-05,
+      "loss": 1.2836,
+      "num_input_tokens_seen": 43008000,
+      "step": 5250
+    },
+    {
+      "epoch": 0.6018995308387688,
+      "grad_norm": 0.5487982630729675,
+      "learning_rate": 2.105355856614115e-05,
+      "loss": 1.1154,
+      "num_input_tokens_seen": 43089920,
+      "step": 5260
+    },
+    {
+      "epoch": 0.603043826524774,
+      "grad_norm": 0.5370051264762878,
+      "learning_rate": 2.0953378055842183e-05,
+      "loss": 1.3152,
+      "num_input_tokens_seen": 43171840,
+      "step": 5270
+    },
+    {
+      "epoch": 0.6041881222107792,
+      "grad_norm": 0.574317455291748,
+      "learning_rate": 2.0853264229611557e-05,
+      "loss": 0.9663,
+      "num_input_tokens_seen": 43253760,
+      "step": 5280
+    },
+    {
+      "epoch": 0.6053324178967845,
+      "grad_norm": 0.48023343086242676,
+      "learning_rate": 2.075321873721972e-05,
+      "loss": 1.0387,
+      "num_input_tokens_seen": 43335680,
+      "step": 5290
+    },
+    {
+      "epoch": 0.6064767135827898,
+      "grad_norm": 0.5434787273406982,
+      "learning_rate": 2.0653243227311014e-05,
+      "loss": 1.3762,
+      "num_input_tokens_seen": 43417600,
+      "step": 5300
+    },
+    {
+      "epoch": 0.607621009268795,
+      "grad_norm": 0.5317404866218567,
+      "learning_rate": 2.0553339347376592e-05,
+      "loss": 1.187,
+      "num_input_tokens_seen": 43499520,
+      "step": 5310
+    },
+    {
+      "epoch": 0.6087653049548003,
+      "grad_norm": 0.4761092960834503,
+      "learning_rate": 2.045350874372717e-05,
+      "loss": 1.0539,
+      "num_input_tokens_seen": 43581440,
+      "step": 5320
+    },
+    {
+      "epoch": 0.6099096006408056,
+      "grad_norm": 0.45552459359169006,
+      "learning_rate": 2.0353753061465972e-05,
+      "loss": 1.3189,
+      "num_input_tokens_seen": 43663360,
+      "step": 5330
+    },
+    {
+      "epoch": 0.6110538963268108,
+      "grad_norm": 0.6579371690750122,
+      "learning_rate": 2.0254073944461603e-05,
+      "loss": 1.2721,
+      "num_input_tokens_seen": 43745280,
+      "step": 5340
+    },
+    {
+      "epoch": 0.6121981920128161,
+      "grad_norm": 0.5771154761314392,
+      "learning_rate": 2.0154473035320936e-05,
+      "loss": 1.1475,
+      "num_input_tokens_seen": 43827200,
+      "step": 5350
+    },
+    {
+      "epoch": 0.6133424876988214,
+      "grad_norm": 0.500095009803772,
+      "learning_rate": 2.0054951975362067e-05,
+      "loss": 1.2035,
+      "num_input_tokens_seen": 43909120,
+      "step": 5360
+    },
+    {
+      "epoch": 0.6144867833848267,
+      "grad_norm": 0.46520867943763733,
+      "learning_rate": 1.995551240458728e-05,
+      "loss": 1.1699,
+      "num_input_tokens_seen": 43991040,
+      "step": 5370
+    },
+    {
+      "epoch": 0.6156310790708319,
+      "grad_norm": 0.46139946579933167,
+      "learning_rate": 1.985615596165597e-05,
+      "loss": 1.2418,
+      "num_input_tokens_seen": 44072960,
+      "step": 5380
+    },
+    {
+      "epoch": 0.6167753747568372,
+      "grad_norm": 0.49758806824684143,
+      "learning_rate": 1.9756884283857685e-05,
+      "loss": 1.4331,
+      "num_input_tokens_seen": 44154880,
+      "step": 5390
+    },
+    {
+      "epoch": 0.6179196704428425,
+      "grad_norm": 0.48119744658470154,
+      "learning_rate": 1.965769900708515e-05,
+      "loss": 1.3412,
+      "num_input_tokens_seen": 44236800,
+      "step": 5400
+    },
+    {
+      "epoch": 0.6190639661288477,
+      "grad_norm": 0.5523199439048767,
+      "learning_rate": 1.955860176580729e-05,
+      "loss": 1.1733,
+      "num_input_tokens_seen": 44318720,
+      "step": 5410
+    },
+    {
+      "epoch": 0.620208261814853,
+      "grad_norm": 0.5241261124610901,
+      "learning_rate": 1.945959419304226e-05,
+      "loss": 1.2493,
+      "num_input_tokens_seen": 44400640,
+      "step": 5420
+    },
+    {
+      "epoch": 0.6213525575008583,
+      "grad_norm": 0.5371418595314026,
+      "learning_rate": 1.936067792033061e-05,
+      "loss": 1.2361,
+      "num_input_tokens_seen": 44482560,
+      "step": 5430
+    },
+    {
+      "epoch": 0.6224968531868634,
+      "grad_norm": 0.5102439522743225,
+      "learning_rate": 1.9261854577708366e-05,
+      "loss": 1.2999,
+      "num_input_tokens_seen": 44564480,
+      "step": 5440
+    },
+    {
+      "epoch": 0.6236411488728687,
+      "grad_norm": 0.9786542654037476,
+      "learning_rate": 1.9163125793680125e-05,
+      "loss": 1.4174,
+      "num_input_tokens_seen": 44646400,
+      "step": 5450
+    },
+    {
+      "epoch": 0.624785444558874,
+      "grad_norm": 0.5264445543289185,
+      "learning_rate": 1.9064493195192293e-05,
+      "loss": 1.0868,
+      "num_input_tokens_seen": 44728320,
+      "step": 5460
+    },
+    {
+      "epoch": 0.6259297402448792,
+      "grad_norm": 0.5219578742980957,
+      "learning_rate": 1.8965958407606236e-05,
+      "loss": 1.1703,
+      "num_input_tokens_seen": 44810240,
+      "step": 5470
+    },
+    {
+      "epoch": 0.6270740359308845,
+      "grad_norm": 0.4690404534339905,
+      "learning_rate": 1.8867523054671475e-05,
+      "loss": 1.2942,
+      "num_input_tokens_seen": 44892160,
+      "step": 5480
+    },
+    {
+      "epoch": 0.6282183316168898,
+      "grad_norm": 0.5699681043624878,
+      "learning_rate": 1.8769188758498973e-05,
+      "loss": 1.1778,
+      "num_input_tokens_seen": 44974080,
+      "step": 5490
+    },
+    {
+      "epoch": 0.629362627302895,
+      "grad_norm": 0.4668191075325012,
+      "learning_rate": 1.867095713953439e-05,
+      "loss": 1.2899,
+      "num_input_tokens_seen": 45056000,
+      "step": 5500
+    },
+    {
+      "epoch": 0.6305069229889003,
+      "grad_norm": 1.0437618494033813,
+      "learning_rate": 1.8572829816531364e-05,
+      "loss": 1.2486,
+      "num_input_tokens_seen": 45137920,
+      "step": 5510
+    },
+    {
+      "epoch": 0.6316512186749056,
+      "grad_norm": 0.4803614318370819,
+      "learning_rate": 1.847480840652483e-05,
+      "loss": 1.2433,
+      "num_input_tokens_seen": 45219840,
+      "step": 5520
+    },
+    {
+      "epoch": 0.6327955143609109,
+      "grad_norm": 0.5359634160995483,
+      "learning_rate": 1.8376894524804416e-05,
+      "loss": 1.2503,
+      "num_input_tokens_seen": 45301760,
+      "step": 5530
+    },
+    {
+      "epoch": 0.6339398100469161,
+      "grad_norm": 0.4710131883621216,
+      "learning_rate": 1.827908978488779e-05,
+      "loss": 1.3593,
+      "num_input_tokens_seen": 45383680,
+      "step": 5540
+    },
+    {
+      "epoch": 0.6350841057329214,
+      "grad_norm": 0.4845348298549652,
+      "learning_rate": 1.8181395798494048e-05,
+      "loss": 1.3471,
+      "num_input_tokens_seen": 45465600,
+      "step": 5550
+    },
+    {
+      "epoch": 0.6362284014189267,
+      "grad_norm": 0.6170366406440735,
+      "learning_rate": 1.8083814175517234e-05,
+      "loss": 1.3265,
+      "num_input_tokens_seen": 45547520,
+      "step": 5560
+    },
+    {
+      "epoch": 0.6373726971049319,
+      "grad_norm": 0.5501654148101807,
+      "learning_rate": 1.798634652399972e-05,
+      "loss": 1.6024,
+      "num_input_tokens_seen": 45629440,
+      "step": 5570
+    },
+    {
+      "epoch": 0.6385169927909372,
+      "grad_norm": 1.0594714879989624,
+      "learning_rate": 1.7888994450105788e-05,
+      "loss": 1.2416,
+      "num_input_tokens_seen": 45711360,
+      "step": 5580
+    },
+    {
+      "epoch": 0.6396612884769425,
+      "grad_norm": 0.4413740634918213,
+      "learning_rate": 1.7791759558095077e-05,
+      "loss": 1.3582,
+      "num_input_tokens_seen": 45793280,
+      "step": 5590
+    },
+    {
+      "epoch": 0.6408055841629478,
+      "grad_norm": 0.6288554668426514,
+      "learning_rate": 1.7694643450296216e-05,
+      "loss": 1.4037,
+      "num_input_tokens_seen": 45875200,
+      "step": 5600
+    },
+    {
+      "epoch": 0.641949879848953,
+      "grad_norm": 0.5537555813789368,
+      "learning_rate": 1.7597647727080408e-05,
+      "loss": 1.3859,
+      "num_input_tokens_seen": 45957120,
+      "step": 5610
+    },
+    {
+      "epoch": 0.6430941755349582,
+      "grad_norm": 0.5998051762580872,
+      "learning_rate": 1.7500773986835013e-05,
+      "loss": 1.0984,
+      "num_input_tokens_seen": 46039040,
+      "step": 5620
+    },
+    {
+      "epoch": 0.6442384712209634,
+      "grad_norm": 0.8084359765052795,
+      "learning_rate": 1.740402382593727e-05,
+      "loss": 1.3061,
+      "num_input_tokens_seen": 46120960,
+      "step": 5630
+    },
+    {
+      "epoch": 0.6453827669069687,
+      "grad_norm": 0.8041592240333557,
+      "learning_rate": 1.730739883872795e-05,
+      "loss": 1.0178,
+      "num_input_tokens_seen": 46202880,
+      "step": 5640
+    },
+    {
+      "epoch": 0.646527062592974,
+      "grad_norm": 0.5870161652565002,
+      "learning_rate": 1.7210900617485075e-05,
+      "loss": 1.2927,
+      "num_input_tokens_seen": 46284800,
+      "step": 5650
+    },
+    {
+      "epoch": 0.6476713582789793,
+      "grad_norm": 0.4723074734210968,
+      "learning_rate": 1.711453075239773e-05,
+      "loss": 1.205,
+      "num_input_tokens_seen": 46366720,
+      "step": 5660
+    },
+    {
+      "epoch": 0.6488156539649845,
+      "grad_norm": 0.9023962020874023,
+      "learning_rate": 1.7018290831539795e-05,
+      "loss": 1.378,
+      "num_input_tokens_seen": 46448640,
+      "step": 5670
+    },
+    {
+      "epoch": 0.6499599496509898,
+      "grad_norm": 0.7521951794624329,
+      "learning_rate": 1.6922182440843843e-05,
+      "loss": 1.0102,
+      "num_input_tokens_seen": 46530560,
+      "step": 5680
+    },
+    {
+      "epoch": 0.6511042453369951,
+      "grad_norm": 0.4786534309387207,
+      "learning_rate": 1.6826207164074924e-05,
+      "loss": 1.2101,
+      "num_input_tokens_seen": 46612480,
+      "step": 5690
+    },
+    {
+      "epoch": 0.6522485410230003,
+      "grad_norm": 1.202210783958435,
+      "learning_rate": 1.6730366582804535e-05,
+      "loss": 1.1996,
+      "num_input_tokens_seen": 46694400,
+      "step": 5700
+    },
+    {
+      "epoch": 0.6533928367090056,
+      "grad_norm": 0.4971697926521301,
+      "learning_rate": 1.6634662276384548e-05,
+      "loss": 1.2356,
+      "num_input_tokens_seen": 46776320,
+      "step": 5710
+    },
+    {
+      "epoch": 0.6545371323950109,
+      "grad_norm": 0.5253583788871765,
+      "learning_rate": 1.6539095821921136e-05,
+      "loss": 1.2006,
+      "num_input_tokens_seen": 46858240,
+      "step": 5720
+    },
+    {
+      "epoch": 0.6556814280810161,
+      "grad_norm": 0.558186411857605,
+      "learning_rate": 1.6443668794248828e-05,
+      "loss": 1.2621,
+      "num_input_tokens_seen": 46940160,
+      "step": 5730
+    },
+    {
+      "epoch": 0.6568257237670214,
+      "grad_norm": 1.3204737901687622,
+      "learning_rate": 1.6348382765904567e-05,
+      "loss": 1.0727,
+      "num_input_tokens_seen": 47022080,
+      "step": 5740
+    },
+    {
+      "epoch": 0.6579700194530267,
+      "grad_norm": 0.5247931480407715,
+      "learning_rate": 1.6253239307101748e-05,
+      "loss": 1.212,
+      "num_input_tokens_seen": 47104000,
+      "step": 5750
+    },
+    {
+      "epoch": 0.659114315139032,
+      "grad_norm": 0.4864051043987274,
+      "learning_rate": 1.6158239985704378e-05,
+      "loss": 1.3306,
+      "num_input_tokens_seen": 47185920,
+      "step": 5760
+    },
+    {
+      "epoch": 0.6602586108250372,
+      "grad_norm": 0.8225966691970825,
+      "learning_rate": 1.606338636720125e-05,
+      "loss": 1.0573,
+      "num_input_tokens_seen": 47267840,
+      "step": 5770
+    },
+    {
+      "epoch": 0.6614029065110425,
+      "grad_norm": 1.1698306798934937,
+      "learning_rate": 1.5968680014680105e-05,
+      "loss": 1.4235,
+      "num_input_tokens_seen": 47349760,
+      "step": 5780
+    },
+    {
+      "epoch": 0.6625472021970478,
+      "grad_norm": 0.4029783010482788,
+      "learning_rate": 1.5874122488801888e-05,
+      "loss": 1.1993,
+      "num_input_tokens_seen": 47431680,
+      "step": 5790
+    },
+    {
+      "epoch": 0.6636914978830529,
+      "grad_norm": 0.4945356249809265,
+      "learning_rate": 1.577971534777507e-05,
+      "loss": 1.1437,
+      "num_input_tokens_seen": 47513600,
+      "step": 5800
+    },
+    {
+      "epoch": 0.6648357935690582,
+      "grad_norm": 0.5037471652030945,
+      "learning_rate": 1.5685460147329917e-05,
+      "loss": 1.2579,
+      "num_input_tokens_seen": 47595520,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6659800892550635,
+      "grad_norm": 0.5378947257995605,
+      "learning_rate": 1.5591358440692865e-05,
+      "loss": 1.2693,
+      "num_input_tokens_seen": 47677440,
+      "step": 5820
+    },
+    {
+      "epoch": 0.6671243849410687,
+      "grad_norm": 0.49818259477615356,
+      "learning_rate": 1.5497411778560954e-05,
+      "loss": 1.1159,
+      "num_input_tokens_seen": 47759360,
+      "step": 5830
+    },
+    {
+      "epoch": 0.668268680627074,
+      "grad_norm": 0.4517885744571686,
+      "learning_rate": 1.5403621709076247e-05,
+      "loss": 1.1901,
+      "num_input_tokens_seen": 47841280,
+      "step": 5840
+    },
+    {
+      "epoch": 0.6694129763130793,
+      "grad_norm": 0.8571302890777588,
+      "learning_rate": 1.530998977780033e-05,
+      "loss": 1.4804,
+      "num_input_tokens_seen": 47923200,
+      "step": 5850
+    },
+    {
+      "epoch": 0.6705572719990845,
+      "grad_norm": 0.8022615313529968,
+      "learning_rate": 1.5216517527688818e-05,
+      "loss": 1.1893,
+      "num_input_tokens_seen": 48005120,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6717015676850898,
+      "grad_norm": 0.630002498626709,
+      "learning_rate": 1.5123206499065967e-05,
+      "loss": 1.085,
+      "num_input_tokens_seen": 48087040,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6728458633710951,
+      "grad_norm": 0.49356332421302795,
+      "learning_rate": 1.5030058229599275e-05,
+      "loss": 1.0382,
+      "num_input_tokens_seen": 48168960,
+      "step": 5880
+    },
+    {
+      "epoch": 0.6739901590571004,
+      "grad_norm": 0.6504026055335999,
+      "learning_rate": 1.4937074254274117e-05,
+      "loss": 1.3036,
+      "num_input_tokens_seen": 48250880,
+      "step": 5890
+    },
+    {
+      "epoch": 0.6751344547431056,
+      "grad_norm": 0.4629366993904114,
+      "learning_rate": 1.4844256105368504e-05,
+      "loss": 1.3465,
+      "num_input_tokens_seen": 48332800,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6762787504291109,
+      "grad_norm": 0.4517896771430969,
+      "learning_rate": 1.4751605312427786e-05,
+      "loss": 1.4707,
+      "num_input_tokens_seen": 48414720,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6774230461151162,
+      "grad_norm": 1.4166773557662964,
+      "learning_rate": 1.4659123402239454e-05,
+      "loss": 1.1061,
+      "num_input_tokens_seen": 48496640,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6785673418011214,
+      "grad_norm": 0.7780046463012695,
+      "learning_rate": 1.4566811898808013e-05,
+      "loss": 1.2819,
+      "num_input_tokens_seen": 48578560,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6797116374871267,
+      "grad_norm": 0.55788654088974,
+      "learning_rate": 1.4474672323329819e-05,
+      "loss": 1.4707,
+      "num_input_tokens_seen": 48660480,
+      "step": 5940
+    },
+    {
+      "epoch": 0.680855933173132,
+      "grad_norm": 0.5268608331680298,
+      "learning_rate": 1.4382706194168066e-05,
+      "loss": 1.4235,
+      "num_input_tokens_seen": 48742400,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6820002288591372,
+      "grad_norm": 0.48586344718933105,
+      "learning_rate": 1.42909150268277e-05,
+      "loss": 1.2991,
+      "num_input_tokens_seen": 48824320,
+      "step": 5960
+    },
+    {
+      "epoch": 0.6831445245451425,
+      "grad_norm": 0.5041683912277222,
+      "learning_rate": 1.4199300333930515e-05,
+      "loss": 1.0258,
+      "num_input_tokens_seen": 48906240,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6842888202311477,
+      "grad_norm": 0.5228765606880188,
+      "learning_rate": 1.4107863625190163e-05,
+      "loss": 1.2446,
+      "num_input_tokens_seen": 48988160,
+      "step": 5980
+    },
+    {
+      "epoch": 0.685433115917153,
+      "grad_norm": 0.4611828029155731,
+      "learning_rate": 1.4016606407387312e-05,
+      "loss": 1.2735,
+      "num_input_tokens_seen": 49070080,
+      "step": 5990
+    },
+    {
+      "epoch": 0.6865774116031582,
+      "grad_norm": 0.543980598449707,
+      "learning_rate": 1.3925530184344818e-05,
+      "loss": 1.3767,
+      "num_input_tokens_seen": 49152000,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6877217072891635,
+      "grad_norm": 0.5816802978515625,
+      "learning_rate": 1.3834636456902944e-05,
+      "loss": 1.4241,
+      "num_input_tokens_seen": 49233920,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6888660029751688,
+      "grad_norm": 0.4795803725719452,
+      "learning_rate": 1.3743926722894579e-05,
+      "loss": 1.1986,
+      "num_input_tokens_seen": 49315840,
+      "step": 6020
+    },
+    {
+      "epoch": 0.690010298661174,
+      "grad_norm": 0.4873317778110504,
+      "learning_rate": 1.365340247712064e-05,
+      "loss": 0.9731,
+      "num_input_tokens_seen": 49397760,
+      "step": 6030
+    },
+    {
+      "epoch": 0.6911545943471793,
+      "grad_norm": 0.5345495939254761,
+      "learning_rate": 1.3563065211325349e-05,
+      "loss": 1.2101,
+      "num_input_tokens_seen": 49479680,
+      "step": 6040
+    },
+    {
+      "epoch": 0.6922988900331846,
+      "grad_norm": 0.6827751398086548,
+      "learning_rate": 1.3472916414171738e-05,
+      "loss": 1.0661,
+      "num_input_tokens_seen": 49561600,
+      "step": 6050
+    },
+    {
+      "epoch": 0.6934431857191898,
+      "grad_norm": 0.533750057220459,
+      "learning_rate": 1.338295757121703e-05,
+      "loss": 1.2239,
+      "num_input_tokens_seen": 49643520,
+      "step": 6060
+    },
+    {
+      "epoch": 0.6945874814051951,
+      "grad_norm": 0.47301891446113586,
+      "learning_rate": 1.3293190164888242e-05,
+      "loss": 1.3588,
+      "num_input_tokens_seen": 49725440,
+      "step": 6070
+    },
+    {
+      "epoch": 0.6957317770912004,
+      "grad_norm": 0.4108991324901581,
+      "learning_rate": 1.3203615674457709e-05,
+      "loss": 1.1447,
+      "num_input_tokens_seen": 49807360,
+      "step": 6080
+    },
+    {
+      "epoch": 0.6968760727772056,
+      "grad_norm": 0.5564691424369812,
+      "learning_rate": 1.3114235576018686e-05,
+      "loss": 1.3927,
+      "num_input_tokens_seen": 49889280,
+      "step": 6090
+    },
+    {
+      "epoch": 0.6980203684632109,
+      "grad_norm": 0.55669766664505,
+      "learning_rate": 1.3025051342461087e-05,
+      "loss": 1.2068,
+      "num_input_tokens_seen": 49971200,
+      "step": 6100
+    },
+    {
+      "epoch": 0.6991646641492162,
+      "grad_norm": 0.47335195541381836,
+      "learning_rate": 1.2936064443447157e-05,
+      "loss": 1.5996,
+      "num_input_tokens_seen": 50053120,
+      "step": 6110
+    },
+    {
+      "epoch": 0.7003089598352215,
+      "grad_norm": 0.5122705698013306,
+      "learning_rate": 1.2847276345387299e-05,
+      "loss": 1.0368,
+      "num_input_tokens_seen": 50135040,
+      "step": 6120
+    },
+    {
+      "epoch": 0.7014532555212267,
+      "grad_norm": 0.47498294711112976,
+      "learning_rate": 1.2758688511415848e-05,
+      "loss": 1.3925,
+      "num_input_tokens_seen": 50216960,
+      "step": 6130
+    },
+    {
+      "epoch": 0.702597551207232,
+      "grad_norm": 0.4896914064884186,
+      "learning_rate": 1.2670302401367035e-05,
+      "loss": 1.1923,
+      "num_input_tokens_seen": 50298880,
+      "step": 6140
+    },
+    {
+      "epoch": 0.7037418468932373,
+      "grad_norm": 0.7128376960754395,
+      "learning_rate": 1.2582119471750888e-05,
+      "loss": 1.3158,
+      "num_input_tokens_seen": 50380800,
+      "step": 6150
+    },
+    {
+      "epoch": 0.7048861425792424,
+      "grad_norm": 0.4625462293624878,
+      "learning_rate": 1.2494141175729216e-05,
+      "loss": 1.1663,
+      "num_input_tokens_seen": 50462720,
+      "step": 6160
+    },
+    {
+      "epoch": 0.7060304382652477,
+      "grad_norm": 0.5610952377319336,
+      "learning_rate": 1.240636896309168e-05,
+      "loss": 1.1836,
+      "num_input_tokens_seen": 50544640,
+      "step": 6170
+    },
+    {
+      "epoch": 0.707174733951253,
+      "grad_norm": 0.8366256952285767,
+      "learning_rate": 1.2318804280231939e-05,
+      "loss": 1.3311,
+      "num_input_tokens_seen": 50626560,
+      "step": 6180
+    },
+    {
+      "epoch": 0.7083190296372582,
+      "grad_norm": 0.6703715324401855,
+      "learning_rate": 1.2231448570123732e-05,
+      "loss": 1.1546,
+      "num_input_tokens_seen": 50708480,
+      "step": 6190
+    },
+    {
+      "epoch": 0.7094633253232635,
+      "grad_norm": 0.41081807017326355,
+      "learning_rate": 1.2144303272297186e-05,
+      "loss": 1.1482,
+      "num_input_tokens_seen": 50790400,
+      "step": 6200
+    },
+    {
+      "epoch": 0.7106076210092688,
+      "grad_norm": 0.6010390520095825,
+      "learning_rate": 1.2057369822815051e-05,
+      "loss": 1.2103,
+      "num_input_tokens_seen": 50872320,
+      "step": 6210
+    },
+    {
+      "epoch": 0.711751916695274,
+      "grad_norm": 0.5093455910682678,
+      "learning_rate": 1.1970649654249017e-05,
+      "loss": 1.5064,
+      "num_input_tokens_seen": 50954240,
+      "step": 6220
+    },
+    {
+      "epoch": 0.7128962123812793,
+      "grad_norm": 0.45238783955574036,
+      "learning_rate": 1.1884144195656133e-05,
+      "loss": 1.3582,
+      "num_input_tokens_seen": 51036160,
+      "step": 6230
+    },
+    {
+      "epoch": 0.7140405080672846,
+      "grad_norm": 0.48569175601005554,
+      "learning_rate": 1.1797854872555272e-05,
+      "loss": 1.3876,
+      "num_input_tokens_seen": 51118080,
+      "step": 6240
+    },
+    {
+      "epoch": 0.7151848037532899,
+      "grad_norm": 0.5373899340629578,
+      "learning_rate": 1.171178310690362e-05,
+      "loss": 1.07,
+      "num_input_tokens_seen": 51200000,
+      "step": 6250
+    },
+    {
+      "epoch": 0.7163290994392951,
+      "grad_norm": 0.568634569644928,
+      "learning_rate": 1.1625930317073221e-05,
+      "loss": 1.2047,
+      "num_input_tokens_seen": 51281920,
+      "step": 6260
+    },
+    {
+      "epoch": 0.7174733951253004,
+      "grad_norm": 0.49342796206474304,
+      "learning_rate": 1.154029791782765e-05,
+      "loss": 1.2819,
+      "num_input_tokens_seen": 51363840,
+      "step": 6270
+    },
+    {
+      "epoch": 0.7186176908113057,
+      "grad_norm": 0.8063413500785828,
+      "learning_rate": 1.1454887320298686e-05,
+      "loss": 1.2803,
+      "num_input_tokens_seen": 51445760,
+      "step": 6280
+    },
+    {
+      "epoch": 0.7197619864973109,
+      "grad_norm": 0.5260705947875977,
+      "learning_rate": 1.1369699931963018e-05,
+      "loss": 1.1276,
+      "num_input_tokens_seen": 51527680,
+      "step": 6290
+    },
+    {
+      "epoch": 0.7209062821833162,
+      "grad_norm": 0.7325502038002014,
+      "learning_rate": 1.1284737156619096e-05,
+      "loss": 1.1392,
+      "num_input_tokens_seen": 51609600,
+      "step": 6300
+    },
+    {
+      "epoch": 0.7220505778693215,
+      "grad_norm": 0.3689436614513397,
+      "learning_rate": 1.1200000394363996e-05,
+      "loss": 1.1951,
+      "num_input_tokens_seen": 51691520,
+      "step": 6310
+    },
+    {
+      "epoch": 0.7231948735553267,
+      "grad_norm": 0.5486378073692322,
+      "learning_rate": 1.1115491041570337e-05,
+      "loss": 1.0797,
+      "num_input_tokens_seen": 51773440,
+      "step": 6320
+    },
+    {
+      "epoch": 0.724339169241332,
+      "grad_norm": 0.4265560209751129,
+      "learning_rate": 1.103121049086324e-05,
+      "loss": 1.3091,
+      "num_input_tokens_seen": 51855360,
+      "step": 6330
+    },
+    {
+      "epoch": 0.7254834649273372,
+      "grad_norm": 0.5804295539855957,
+      "learning_rate": 1.094716013109745e-05,
+      "loss": 1.1386,
+      "num_input_tokens_seen": 51937280,
+      "step": 6340
+    },
+    {
+      "epoch": 0.7266277606133424,
+      "grad_norm": 0.5425553321838379,
+      "learning_rate": 1.0863341347334376e-05,
+      "loss": 1.2296,
+      "num_input_tokens_seen": 52019200,
+      "step": 6350
+    },
+    {
+      "epoch": 0.7277720562993477,
+      "grad_norm": 0.584921658039093,
+      "learning_rate": 1.0779755520819302e-05,
+      "loss": 1.2027,
+      "num_input_tokens_seen": 52101120,
+      "step": 6360
+    },
+    {
+      "epoch": 0.728916351985353,
+      "grad_norm": 0.6812318563461304,
+      "learning_rate": 1.0696404028958634e-05,
+      "loss": 1.1153,
+      "num_input_tokens_seen": 52183040,
+      "step": 6370
+    },
+    {
+      "epoch": 0.7300606476713583,
+      "grad_norm": 0.4892879128456116,
+      "learning_rate": 1.0613288245297193e-05,
+      "loss": 1.097,
+      "num_input_tokens_seen": 52264960,
+      "step": 6380
+    },
+    {
+      "epoch": 0.7312049433573635,
+      "grad_norm": 0.4659218490123749,
+      "learning_rate": 1.053040953949557e-05,
+      "loss": 1.2608,
+      "num_input_tokens_seen": 52346880,
+      "step": 6390
+    },
+    {
+      "epoch": 0.7323492390433688,
+      "grad_norm": 0.5382397770881653,
+      "learning_rate": 1.0447769277307554e-05,
+      "loss": 1.0397,
+      "num_input_tokens_seen": 52428800,
+      "step": 6400
+    },
+    {
+      "epoch": 0.7334935347293741,
+      "grad_norm": 0.5254513621330261,
+      "learning_rate": 1.0365368820557633e-05,
+      "loss": 1.0589,
+      "num_input_tokens_seen": 52510720,
+      "step": 6410
+    },
+    {
+      "epoch": 0.7346378304153793,
+      "grad_norm": 1.027398943901062,
+      "learning_rate": 1.0283209527118584e-05,
+      "loss": 1.0307,
+      "num_input_tokens_seen": 52592640,
+      "step": 6420
+    },
+    {
+      "epoch": 0.7357821261013846,
+      "grad_norm": 0.5271130800247192,
+      "learning_rate": 1.0201292750889022e-05,
+      "loss": 1.3241,
+      "num_input_tokens_seen": 52674560,
+      "step": 6430
+    },
+    {
+      "epoch": 0.7369264217873899,
+      "grad_norm": 0.45055344700813293,
+      "learning_rate": 1.011961984177117e-05,
+      "loss": 1.1761,
+      "num_input_tokens_seen": 52756480,
+      "step": 6440
+    },
+    {
+      "epoch": 0.7380707174733951,
+      "grad_norm": 0.5612509846687317,
+      "learning_rate": 1.0038192145648567e-05,
+      "loss": 1.0847,
+      "num_input_tokens_seen": 52838400,
+      "step": 6450
+    },
+    {
+      "epoch": 0.7392150131594004,
+      "grad_norm": 0.5806179642677307,
+      "learning_rate": 9.95701100436389e-06,
+      "loss": 1.1278,
+      "num_input_tokens_seen": 52920320,
+      "step": 6460
+    },
+    {
+      "epoch": 0.7403593088454057,
+      "grad_norm": 0.5459727048873901,
+      "learning_rate": 9.876077755696868e-06,
+      "loss": 1.2579,
+      "num_input_tokens_seen": 53002240,
+      "step": 6470
+    },
+    {
+      "epoch": 0.741503604531411,
+      "grad_norm": 0.4346022307872772,
+      "learning_rate": 9.795393733342203e-06,
+      "loss": 1.3319,
+      "num_input_tokens_seen": 53084160,
+      "step": 6480
+    },
+    {
+      "epoch": 0.7426479002174162,
+      "grad_norm": 1.9029626846313477,
+      "learning_rate": 9.71496026688763e-06,
+      "loss": 1.4099,
+      "num_input_tokens_seen": 53166080,
+      "step": 6490
+    },
+    {
+      "epoch": 0.7437921959034215,
+      "grad_norm": 0.5004605054855347,
+      "learning_rate": 9.634778681791962e-06,
+      "loss": 1.2894,
+      "num_input_tokens_seen": 53248000,
+      "step": 6500
+    },
+    {
+      "epoch": 0.7449364915894267,
+      "grad_norm": 0.7425548434257507,
+      "learning_rate": 9.554850299363294e-06,
+      "loss": 1.4331,
+      "num_input_tokens_seen": 53329920,
+      "step": 6510
+    },
+    {
+      "epoch": 0.7460807872754319,
+      "grad_norm": 0.507979691028595,
+      "learning_rate": 9.47517643673721e-06,
+      "loss": 1.1623,
+      "num_input_tokens_seen": 53411840,
+      "step": 6520
+    },
+    {
+      "epoch": 0.7472250829614372,
+      "grad_norm": 0.47209110856056213,
+      "learning_rate": 9.395758406855053e-06,
+      "loss": 1.195,
+      "num_input_tokens_seen": 53493760,
+      "step": 6530
+    },
+    {
+      "epoch": 0.7483693786474425,
+      "grad_norm": 0.4794630706310272,
+      "learning_rate": 9.31659751844232e-06,
+      "loss": 1.2136,
+      "num_input_tokens_seen": 53575680,
+      "step": 6540
+    },
+    {
+      "epoch": 0.7495136743334477,
+      "grad_norm": 0.48791974782943726,
+      "learning_rate": 9.237695075987106e-06,
+      "loss": 1.476,
+      "num_input_tokens_seen": 53657600,
+      "step": 6550
+    },
+    {
+      "epoch": 0.750657970019453,
+      "grad_norm": 0.4857141971588135,
+      "learning_rate": 9.15905237971856e-06,
+      "loss": 1.3678,
+      "num_input_tokens_seen": 53739520,
+      "step": 6560
+    },
+    {
+      "epoch": 0.7518022657054583,
+      "grad_norm": 0.5270280241966248,
+      "learning_rate": 9.080670725585511e-06,
+      "loss": 1.4367,
+      "num_input_tokens_seen": 53821440,
+      "step": 6570
+    },
+    {
+      "epoch": 0.7529465613914635,
+      "grad_norm": 0.48381784558296204,
+      "learning_rate": 9.002551405235082e-06,
+      "loss": 1.2801,
+      "num_input_tokens_seen": 53903360,
+      "step": 6580
+    },
+    {
+      "epoch": 0.7540908570774688,
+      "grad_norm": 0.7849867343902588,
+      "learning_rate": 8.924695705991407e-06,
+      "loss": 1.4574,
+      "num_input_tokens_seen": 53985280,
+      "step": 6590
+    },
+    {
+      "epoch": 0.7552351527634741,
+      "grad_norm": 0.45584842562675476,
+      "learning_rate": 8.847104910834414e-06,
+      "loss": 1.4978,
+      "num_input_tokens_seen": 54067200,
+      "step": 6600
+    },
+    {
+      "epoch": 0.7563794484494794,
+      "grad_norm": 0.5474436283111572,
+      "learning_rate": 8.769780298378705e-06,
+      "loss": 1.277,
+      "num_input_tokens_seen": 54149120,
+      "step": 6610
+    },
+    {
+      "epoch": 0.7575237441354846,
+      "grad_norm": 0.49009135365486145,
+      "learning_rate": 8.69272314285248e-06,
+      "loss": 1.0243,
+      "num_input_tokens_seen": 54231040,
+      "step": 6620
+    },
+    {
+      "epoch": 0.7586680398214899,
+      "grad_norm": 0.4982384443283081,
+      "learning_rate": 8.6159347140765e-06,
+      "loss": 1.2217,
+      "num_input_tokens_seen": 54312960,
+      "step": 6630
+    },
+    {
+      "epoch": 0.7598123355074952,
+      "grad_norm": 0.5559042096138,
+      "learning_rate": 8.539416277443218e-06,
+      "loss": 1.2161,
+      "num_input_tokens_seen": 54394880,
+      "step": 6640
+    },
+    {
+      "epoch": 0.7609566311935004,
+      "grad_norm": 0.4743431508541107,
+      "learning_rate": 8.463169093895887e-06,
+      "loss": 1.069,
+      "num_input_tokens_seen": 54476800,
+      "step": 6650
+    },
+    {
+      "epoch": 0.7621009268795057,
+      "grad_norm": 0.5154614448547363,
+      "learning_rate": 8.38719441990781e-06,
+      "loss": 1.5262,
+      "num_input_tokens_seen": 54558720,
+      "step": 6660
+    },
+    {
+      "epoch": 0.763245222565511,
+      "grad_norm": 0.5313172340393066,
+      "learning_rate": 8.311493507461593e-06,
+      "loss": 1.3362,
+      "num_input_tokens_seen": 54640640,
+      "step": 6670
+    },
+    {
+      "epoch": 0.7643895182515162,
+      "grad_norm": 0.4694889187812805,
+      "learning_rate": 8.236067604028563e-06,
+      "loss": 1.2781,
+      "num_input_tokens_seen": 54722560,
+      "step": 6680
+    },
+    {
+      "epoch": 0.7655338139375214,
+      "grad_norm": 0.5564500689506531,
+      "learning_rate": 8.160917952548197e-06,
+      "loss": 1.4282,
+      "num_input_tokens_seen": 54804480,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7666781096235267,
+      "grad_norm": 0.49310502409935,
+      "learning_rate": 8.08604579140759e-06,
+      "loss": 1.2766,
+      "num_input_tokens_seen": 54886400,
+      "step": 6700
+    },
+    {
+      "epoch": 0.7678224053095319,
+      "grad_norm": 0.4908527731895447,
+      "learning_rate": 8.011452354421136e-06,
+      "loss": 1.4154,
+      "num_input_tokens_seen": 54968320,
+      "step": 6710
+    },
+    {
+      "epoch": 0.7689667009955372,
+      "grad_norm": 0.47903621196746826,
+      "learning_rate": 7.937138870810115e-06,
+      "loss": 0.9568,
+      "num_input_tokens_seen": 55050240,
+      "step": 6720
+    },
+    {
+      "epoch": 0.7701109966815425,
+      "grad_norm": 0.820591926574707,
+      "learning_rate": 7.863106565182474e-06,
+      "loss": 1.4689,
+      "num_input_tokens_seen": 55132160,
+      "step": 6730
+    },
+    {
+      "epoch": 0.7712552923675478,
+      "grad_norm": 0.5931017994880676,
+      "learning_rate": 7.78935665751266e-06,
+      "loss": 1.3219,
+      "num_input_tokens_seen": 55214080,
+      "step": 6740
+    },
+    {
+      "epoch": 0.772399588053553,
+      "grad_norm": 0.5245538949966431,
+      "learning_rate": 7.715890363121484e-06,
+      "loss": 1.1528,
+      "num_input_tokens_seen": 55296000,
+      "step": 6750
+    },
+    {
+      "epoch": 0.7735438837395583,
+      "grad_norm": 0.44471973180770874,
+      "learning_rate": 7.642708892656125e-06,
+      "loss": 1.129,
+      "num_input_tokens_seen": 55377920,
+      "step": 6760
+    },
+    {
+      "epoch": 0.7746881794255636,
+      "grad_norm": 0.5023803114891052,
+      "learning_rate": 7.569813452070146e-06,
+      "loss": 1.0879,
+      "num_input_tokens_seen": 55459840,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7758324751115688,
+      "grad_norm": 0.4996906518936157,
+      "learning_rate": 7.497205242603636e-06,
+      "loss": 1.2824,
+      "num_input_tokens_seen": 55541760,
+      "step": 6780
+    },
+    {
+      "epoch": 0.7769767707975741,
+      "grad_norm": 0.4397946298122406,
+      "learning_rate": 7.424885460763442e-06,
+      "loss": 1.2056,
+      "num_input_tokens_seen": 55623680,
+      "step": 6790
+    },
+    {
+      "epoch": 0.7781210664835794,
+      "grad_norm": 0.5805292725563049,
+      "learning_rate": 7.3528552983033985e-06,
+      "loss": 1.2665,
+      "num_input_tokens_seen": 55705600,
+      "step": 6800
+    },
+    {
+      "epoch": 0.7792653621695846,
+      "grad_norm": 0.4549366533756256,
+      "learning_rate": 7.281115942204739e-06,
+      "loss": 1.0662,
+      "num_input_tokens_seen": 55787520,
+      "step": 6810
+    },
+    {
+      "epoch": 0.7804096578555899,
+      "grad_norm": 0.547260046005249,
+      "learning_rate": 7.209668574656514e-06,
+      "loss": 1.1699,
+      "num_input_tokens_seen": 55869440,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7815539535415952,
+      "grad_norm": 0.5504161715507507,
+      "learning_rate": 7.138514373036098e-06,
+      "loss": 1.3507,
+      "num_input_tokens_seen": 55951360,
+      "step": 6830
+    },
+    {
+      "epoch": 0.7826982492276005,
+      "grad_norm": 0.5131449699401855,
+      "learning_rate": 7.0676545098897956e-06,
+      "loss": 1.3175,
+      "num_input_tokens_seen": 56033280,
+      "step": 6840
+    },
+    {
+      "epoch": 0.7838425449136057,
+      "grad_norm": 0.4970407783985138,
+      "learning_rate": 6.997090152913535e-06,
+      "loss": 1.1614,
+      "num_input_tokens_seen": 56115200,
+      "step": 6850
+    },
+    {
+      "epoch": 0.784986840599611,
+      "grad_norm": 1.6976333856582642,
+      "learning_rate": 6.92682246493363e-06,
+      "loss": 1.0802,
+      "num_input_tokens_seen": 56197120,
+      "step": 6860
+    },
+    {
+      "epoch": 0.7861311362856161,
+      "grad_norm": 2.0605082511901855,
+      "learning_rate": 6.856852603887556e-06,
+      "loss": 1.2859,
+      "num_input_tokens_seen": 56279040,
+      "step": 6870
+    },
+    {
+      "epoch": 0.7872754319716214,
+      "grad_norm": 0.4618769586086273,
+      "learning_rate": 6.787181722804959e-06,
+      "loss": 1.0909,
+      "num_input_tokens_seen": 56360960,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7884197276576267,
+      "grad_norm": 0.5229918360710144,
+      "learning_rate": 6.717810969788596e-06,
+      "loss": 1.2816,
+      "num_input_tokens_seen": 56442880,
+      "step": 6890
+    },
+    {
+      "epoch": 0.789564023343632,
+      "grad_norm": 0.54283607006073,
+      "learning_rate": 6.648741487995416e-06,
+      "loss": 1.036,
+      "num_input_tokens_seen": 56524800,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7907083190296372,
+      "grad_norm": 0.6073585748672485,
+      "learning_rate": 6.57997441561774e-06,
+      "loss": 0.9993,
+      "num_input_tokens_seen": 56606720,
+      "step": 6910
+    },
+    {
+      "epoch": 0.7918526147156425,
+      "grad_norm": 0.4236718416213989,
+      "learning_rate": 6.511510885864516e-06,
+      "loss": 1.1192,
+      "num_input_tokens_seen": 56688640,
+      "step": 6920
+    },
+    {
+      "epoch": 0.7929969104016478,
+      "grad_norm": 0.5184228420257568,
+      "learning_rate": 6.44335202694262e-06,
+      "loss": 1.3028,
+      "num_input_tokens_seen": 56770560,
+      "step": 6930
+    },
+    {
+      "epoch": 0.794141206087653,
+      "grad_norm": 0.6947381496429443,
+      "learning_rate": 6.375498962038265e-06,
+      "loss": 1.4295,
+      "num_input_tokens_seen": 56852480,
+      "step": 6940
+    },
+    {
+      "epoch": 0.7952855017736583,
+      "grad_norm": 0.5223413705825806,
+      "learning_rate": 6.307952809298517e-06,
+      "loss": 1.503,
+      "num_input_tokens_seen": 56934400,
+      "step": 6950
+    },
+    {
+      "epoch": 0.7964297974596636,
+      "grad_norm": 0.5551154613494873,
+      "learning_rate": 6.240714681812837e-06,
+      "loss": 1.1341,
+      "num_input_tokens_seen": 57016320,
+      "step": 6960
+    },
+    {
+      "epoch": 0.7975740931456688,
+      "grad_norm": 0.5838407278060913,
+      "learning_rate": 6.173785687594761e-06,
+      "loss": 1.14,
+      "num_input_tokens_seen": 57098240,
+      "step": 6970
+    },
+    {
+      "epoch": 0.7987183888316741,
+      "grad_norm": 0.4963679611682892,
+      "learning_rate": 6.107166929563629e-06,
+      "loss": 1.2954,
+      "num_input_tokens_seen": 57180160,
+      "step": 6980
+    },
+    {
+      "epoch": 0.7998626845176794,
+      "grad_norm": 0.5126423835754395,
+      "learning_rate": 6.040859505526439e-06,
+      "loss": 1.3446,
+      "num_input_tokens_seen": 57262080,
+      "step": 6990
+    },
+    {
+      "epoch": 0.8010069802036847,
+      "grad_norm": 0.48465293645858765,
+      "learning_rate": 5.974864508159692e-06,
+      "loss": 1.0721,
+      "num_input_tokens_seen": 57344000,
+      "step": 7000
+    },
+    {
+      "epoch": 0.8021512758896899,
+      "grad_norm": 0.5144678950309753,
+      "learning_rate": 5.9091830249914685e-06,
+      "loss": 1.3073,
+      "num_input_tokens_seen": 57425920,
+      "step": 7010
+    },
+    {
+      "epoch": 0.8032955715756952,
+      "grad_norm": 0.4844573140144348,
+      "learning_rate": 5.843816138383429e-06,
+      "loss": 1.3409,
+      "num_input_tokens_seen": 57507840,
+      "step": 7020
+    },
+    {
+      "epoch": 0.8044398672617005,
+      "grad_norm": 0.538421094417572,
+      "learning_rate": 5.778764925513045e-06,
+      "loss": 1.3687,
+      "num_input_tokens_seen": 57589760,
+      "step": 7030
+    },
+    {
+      "epoch": 0.8055841629477057,
+      "grad_norm": 0.49097201228141785,
+      "learning_rate": 5.714030458355784e-06,
+      "loss": 1.099,
+      "num_input_tokens_seen": 57671680,
+      "step": 7040
+    },
+    {
+      "epoch": 0.8067284586337109,
+      "grad_norm": 0.5372655391693115,
+      "learning_rate": 5.649613803667511e-06,
+      "loss": 1.3109,
+      "num_input_tokens_seen": 57753600,
+      "step": 7050
+    },
+    {
+      "epoch": 0.8078727543197162,
+      "grad_norm": 0.5266016721725464,
+      "learning_rate": 5.5855160229668636e-06,
+      "loss": 1.2747,
+      "num_input_tokens_seen": 57835520,
+      "step": 7060
+    },
+    {
+      "epoch": 0.8090170500057214,
+      "grad_norm": 0.5975900888442993,
+      "learning_rate": 5.5217381725177624e-06,
+      "loss": 1.2696,
+      "num_input_tokens_seen": 57917440,
+      "step": 7070
+    },
+    {
+      "epoch": 0.8101613456917267,
+      "grad_norm": 0.5652705430984497,
+      "learning_rate": 5.458281303312016e-06,
+      "loss": 1.1509,
+      "num_input_tokens_seen": 57999360,
+      "step": 7080
+    },
+    {
+      "epoch": 0.811305641377732,
+      "grad_norm": 0.7816336154937744,
+      "learning_rate": 5.39514646105202e-06,
+      "loss": 1.2708,
+      "num_input_tokens_seen": 58081280,
+      "step": 7090
+    },
+    {
+      "epoch": 0.8124499370637372,
+      "grad_norm": 0.48780134320259094,
+      "learning_rate": 5.332334686133475e-06,
+      "loss": 1.4797,
+      "num_input_tokens_seen": 58163200,
+      "step": 7100
+    },
+    {
+      "epoch": 0.8135942327497425,
+      "grad_norm": 0.5095590949058533,
+      "learning_rate": 5.269847013628299e-06,
+      "loss": 1.192,
+      "num_input_tokens_seen": 58245120,
+      "step": 7110
+    },
+    {
+      "epoch": 0.8147385284357478,
+      "grad_norm": 0.4638192057609558,
+      "learning_rate": 5.207684473267527e-06,
+      "loss": 1.5395,
+      "num_input_tokens_seen": 58327040,
+      "step": 7120
+    },
+    {
+      "epoch": 0.8158828241217531,
+      "grad_norm": 0.538078784942627,
+      "learning_rate": 5.145848089424374e-06,
+      "loss": 1.2233,
+      "num_input_tokens_seen": 58408960,
+      "step": 7130
+    },
+    {
+      "epoch": 0.8170271198077583,
+      "grad_norm": 0.544783353805542,
+      "learning_rate": 5.0843388810973195e-06,
+      "loss": 1.4877,
+      "num_input_tokens_seen": 58490880,
+      "step": 7140
+    },
+    {
+      "epoch": 0.8181714154937636,
+      "grad_norm": 0.9503467679023743,
+      "learning_rate": 5.02315786189334e-06,
+      "loss": 1.1406,
+      "num_input_tokens_seen": 58572800,
+      "step": 7150
+    },
+    {
+      "epoch": 0.8193157111797689,
+      "grad_norm": 0.5432871580123901,
+      "learning_rate": 4.962306040011222e-06,
+      "loss": 1.1193,
+      "num_input_tokens_seen": 58654720,
+      "step": 7160
+    },
+    {
+      "epoch": 0.8204600068657741,
+      "grad_norm": 0.6386141777038574,
+      "learning_rate": 4.901784418224892e-06,
+      "loss": 1.0802,
+      "num_input_tokens_seen": 58736640,
+      "step": 7170
+    },
+    {
+      "epoch": 0.8216043025517794,
+      "grad_norm": 0.5047373175621033,
+      "learning_rate": 4.841593993866949e-06,
+      "loss": 1.0499,
+      "num_input_tokens_seen": 58818560,
+      "step": 7180
+    },
+    {
+      "epoch": 0.8227485982377847,
+      "grad_norm": 0.4890262484550476,
+      "learning_rate": 4.781735758812217e-06,
+      "loss": 1.2981,
+      "num_input_tokens_seen": 58900480,
+      "step": 7190
+    },
+    {
+      "epoch": 0.82389289392379,
+      "grad_norm": 0.744850754737854,
+      "learning_rate": 4.7222106994613655e-06,
+      "loss": 1.1535,
+      "num_input_tokens_seen": 58982400,
+      "step": 7200
+    },
+    {
+      "epoch": 0.8250371896097952,
+      "grad_norm": 0.4771808087825775,
+      "learning_rate": 4.663019796724685e-06,
+      "loss": 1.2276,
+      "num_input_tokens_seen": 59064320,
+      "step": 7210
+    },
+    {
+      "epoch": 0.8261814852958005,
+      "grad_norm": 0.463220477104187,
+      "learning_rate": 4.604164026005925e-06,
+      "loss": 1.3646,
+      "num_input_tokens_seen": 59146240,
+      "step": 7220
+    },
+    {
+      "epoch": 0.8273257809818056,
+      "grad_norm": 0.46109485626220703,
+      "learning_rate": 4.5456443571862185e-06,
+      "loss": 1.3941,
+      "num_input_tokens_seen": 59228160,
+      "step": 7230
+    },
+    {
+      "epoch": 0.8284700766678109,
+      "grad_norm": 0.4565638303756714,
+      "learning_rate": 4.487461754608066e-06,
+      "loss": 1.102,
+      "num_input_tokens_seen": 59310080,
+      "step": 7240
+    },
+    {
+      "epoch": 0.8296143723538162,
+      "grad_norm": 0.5542830228805542,
+      "learning_rate": 4.429617177059508e-06,
+      "loss": 1.167,
+      "num_input_tokens_seen": 59392000,
+      "step": 7250
+    },
+    {
+      "epoch": 0.8307586680398215,
+      "grad_norm": 0.44131019711494446,
+      "learning_rate": 4.372111577758261e-06,
+      "loss": 1.0619,
+      "num_input_tokens_seen": 59473920,
+      "step": 7260
+    },
+    {
+      "epoch": 0.8319029637258267,
+      "grad_norm": 0.523678183555603,
+      "learning_rate": 4.314945904336037e-06,
+      "loss": 1.2679,
+      "num_input_tokens_seen": 59555840,
+      "step": 7270
+    },
+    {
+      "epoch": 0.833047259411832,
+      "grad_norm": 0.485726535320282,
+      "learning_rate": 4.258121098822945e-06,
+      "loss": 1.2982,
+      "num_input_tokens_seen": 59637760,
+      "step": 7280
+    },
+    {
+      "epoch": 0.8341915550978373,
+      "grad_norm": 0.49582865834236145,
+      "learning_rate": 4.201638097631938e-06,
+      "loss": 1.2368,
+      "num_input_tokens_seen": 59719680,
+      "step": 7290
+    },
+    {
+      "epoch": 0.8353358507838425,
+      "grad_norm": 0.44791966676712036,
+      "learning_rate": 4.145497831543402e-06,
+      "loss": 0.9904,
+      "num_input_tokens_seen": 59801600,
+      "step": 7300
+    },
+    {
+      "epoch": 0.8364801464698478,
+      "grad_norm": 0.4786857068538666,
+      "learning_rate": 4.089701225689793e-06,
+      "loss": 1.186,
+      "num_input_tokens_seen": 59883520,
+      "step": 7310
+    },
+    {
+      "epoch": 0.8376244421558531,
+      "grad_norm": 0.48226872086524963,
+      "learning_rate": 4.034249199540432e-06,
+      "loss": 1.359,
+      "num_input_tokens_seen": 59965440,
+      "step": 7320
+    },
+    {
+      "epoch": 0.8387687378418583,
+      "grad_norm": 0.5494146347045898,
+      "learning_rate": 3.97914266688631e-06,
+      "loss": 1.1773,
+      "num_input_tokens_seen": 60047360,
+      "step": 7330
+    },
+    {
+      "epoch": 0.8399130335278636,
+      "grad_norm": 0.4615972638130188,
+      "learning_rate": 3.924382535825047e-06,
+      "loss": 1.2996,
+      "num_input_tokens_seen": 60129280,
+      "step": 7340
+    },
+    {
+      "epoch": 0.8410573292138689,
+      "grad_norm": 1.4144350290298462,
+      "learning_rate": 3.869969708745946e-06,
+      "loss": 1.1155,
+      "num_input_tokens_seen": 60211200,
+      "step": 7350
+    },
+    {
+      "epoch": 0.8422016248998742,
+      "grad_norm": 0.5016247034072876,
+      "learning_rate": 3.815905082315102e-06,
+      "loss": 1.058,
+      "num_input_tokens_seen": 60293120,
+      "step": 7360
+    },
+    {
+      "epoch": 0.8433459205858794,
+      "grad_norm": 0.4713045358657837,
+      "learning_rate": 3.762189547460615e-06,
+      "loss": 1.2502,
+      "num_input_tokens_seen": 60375040,
+      "step": 7370
+    },
+    {
+      "epoch": 0.8444902162718847,
+      "grad_norm": 0.8532644510269165,
+      "learning_rate": 3.7088239893579456e-06,
+      "loss": 1.2315,
+      "num_input_tokens_seen": 60456960,
+      "step": 7380
+    },
+    {
+      "epoch": 0.84563451195789,
+      "grad_norm": 0.5714083313941956,
+      "learning_rate": 3.655809287415285e-06,
+      "loss": 1.46,
+      "num_input_tokens_seen": 60538880,
+      "step": 7390
+    },
+    {
+      "epoch": 0.8467788076438952,
+      "grad_norm": 0.5227552652359009,
+      "learning_rate": 3.603146315259104e-06,
+      "loss": 1.2398,
+      "num_input_tokens_seen": 60620800,
+      "step": 7400
+    },
+    {
+      "epoch": 0.8479231033299004,
+      "grad_norm": 0.5611023306846619,
+      "learning_rate": 3.5508359407197157e-06,
+      "loss": 1.2431,
+      "num_input_tokens_seen": 60702720,
+      "step": 7410
+    },
+    {
+      "epoch": 0.8490673990159057,
+      "grad_norm": 0.8136146068572998,
+      "learning_rate": 3.4988790258170146e-06,
+      "loss": 0.9851,
+      "num_input_tokens_seen": 60784640,
+      "step": 7420
+    },
+    {
+      "epoch": 0.8502116947019109,
+      "grad_norm": 0.5456047058105469,
+      "learning_rate": 3.4472764267462486e-06,
+      "loss": 1.1814,
+      "num_input_tokens_seen": 60866560,
+      "step": 7430
+    },
+    {
+      "epoch": 0.8513559903879162,
+      "grad_norm": 0.4470462501049042,
+      "learning_rate": 3.396028993863906e-06,
+      "loss": 1.0398,
+      "num_input_tokens_seen": 60948480,
+      "step": 7440
+    },
+    {
+      "epoch": 0.8525002860739215,
+      "grad_norm": 0.4927099049091339,
+      "learning_rate": 3.3451375716737067e-06,
+      "loss": 1.1797,
+      "num_input_tokens_seen": 61030400,
+      "step": 7450
+    },
+    {
+      "epoch": 0.8536445817599267,
+      "grad_norm": 0.6371917724609375,
+      "learning_rate": 3.2946029988127068e-06,
+      "loss": 1.3581,
+      "num_input_tokens_seen": 61112320,
+      "step": 7460
+    },
+    {
+      "epoch": 0.854788877445932,
+      "grad_norm": 0.4422919452190399,
+      "learning_rate": 3.2444261080374546e-06,
+      "loss": 1.1539,
+      "num_input_tokens_seen": 61194240,
+      "step": 7470
+    },
+    {
+      "epoch": 0.8559331731319373,
+      "grad_norm": 0.4536419212818146,
+      "learning_rate": 3.194607726210261e-06,
+      "loss": 1.0741,
+      "num_input_tokens_seen": 61276160,
+      "step": 7480
+    },
+    {
+      "epoch": 0.8570774688179426,
+      "grad_norm": 0.5793977379798889,
+      "learning_rate": 3.1451486742856055e-06,
+      "loss": 1.2566,
+      "num_input_tokens_seen": 61358080,
+      "step": 7490
+    },
+    {
+      "epoch": 0.8582217645039478,
+      "grad_norm": 0.6592020392417908,
+      "learning_rate": 3.0960497672965825e-06,
+      "loss": 1.2257,
+      "num_input_tokens_seen": 61440000,
+      "step": 7500
+    },
+    {
+      "epoch": 0.8593660601899531,
+      "grad_norm": 0.49061110615730286,
+      "learning_rate": 3.0473118143414634e-06,
+      "loss": 1.4217,
+      "num_input_tokens_seen": 61521920,
+      "step": 7510
+    },
+    {
+      "epoch": 0.8605103558759584,
+      "grad_norm": 0.4329964518547058,
+      "learning_rate": 2.9989356185703975e-06,
+      "loss": 1.4657,
+      "num_input_tokens_seen": 61603840,
+      "step": 7520
+    },
+    {
+      "epoch": 0.8616546515619636,
+      "grad_norm": 0.5590507388114929,
+      "learning_rate": 2.950921977172155e-06,
+      "loss": 1.0984,
+      "num_input_tokens_seen": 61685760,
+      "step": 7530
+    },
+    {
+      "epoch": 0.8627989472479689,
+      "grad_norm": 0.7304471135139465,
+      "learning_rate": 2.9032716813609723e-06,
+      "loss": 1.1865,
+      "num_input_tokens_seen": 61767680,
+      "step": 7540
+    },
+    {
+      "epoch": 0.8639432429339742,
+      "grad_norm": 0.48121222853660583,
+      "learning_rate": 2.8559855163635544e-06,
+      "loss": 1.1777,
+      "num_input_tokens_seen": 61849600,
+      "step": 7550
+    },
+    {
+      "epoch": 0.8650875386199794,
+      "grad_norm": 0.469186007976532,
+      "learning_rate": 2.809064261406111e-06,
+      "loss": 1.5473,
+      "num_input_tokens_seen": 61931520,
+      "step": 7560
+    },
+    {
+      "epoch": 0.8662318343059847,
+      "grad_norm": 0.5229854583740234,
+      "learning_rate": 2.762508689701504e-06,
+      "loss": 1.1929,
+      "num_input_tokens_seen": 62013440,
+      "step": 7570
+    },
+    {
+      "epoch": 0.8673761299919899,
+      "grad_norm": 0.5912187099456787,
+      "learning_rate": 2.716319568436529e-06,
+      "loss": 1.0352,
+      "num_input_tokens_seen": 62095360,
+      "step": 7580
+    },
+    {
+      "epoch": 0.8685204256779951,
+      "grad_norm": 0.5109303593635559,
+      "learning_rate": 2.6704976587592688e-06,
+      "loss": 1.0181,
+      "num_input_tokens_seen": 62177280,
+      "step": 7590
+    },
+    {
+      "epoch": 0.8696647213640004,
+      "grad_norm": 0.7814948558807373,
+      "learning_rate": 2.6250437157665455e-06,
+      "loss": 1.2663,
+      "num_input_tokens_seen": 62259200,
+      "step": 7600
+    },
+    {
+      "epoch": 0.8708090170500057,
+      "grad_norm": 1.977430820465088,
+      "learning_rate": 2.5799584884914685e-06,
+      "loss": 1.0863,
+      "num_input_tokens_seen": 62341120,
+      "step": 7610
+    },
+    {
+      "epoch": 0.871953312736011,
+      "grad_norm": 0.4965369701385498,
+      "learning_rate": 2.535242719891112e-06,
+      "loss": 1.1277,
+      "num_input_tokens_seen": 62423040,
+      "step": 7620
+    },
+    {
+      "epoch": 0.8730976084220162,
+      "grad_norm": 1.2752573490142822,
+      "learning_rate": 2.4908971468342535e-06,
+      "loss": 1.067,
+      "num_input_tokens_seen": 62504960,
+      "step": 7630
+    },
+    {
+      "epoch": 0.8742419041080215,
+      "grad_norm": 0.5822865962982178,
+      "learning_rate": 2.44692250008923e-06,
+      "loss": 1.2818,
+      "num_input_tokens_seen": 62586880,
+      "step": 7640
+    },
+    {
+      "epoch": 0.8753861997940268,
+      "grad_norm": 0.5638536214828491,
+      "learning_rate": 2.403319504311921e-06,
+      "loss": 1.1506,
+      "num_input_tokens_seen": 62668800,
+      "step": 7650
+    },
+    {
+      "epoch": 0.876530495480032,
+      "grad_norm": 0.48504772782325745,
+      "learning_rate": 2.360088878033778e-06,
+      "loss": 1.5572,
+      "num_input_tokens_seen": 62750720,
+      "step": 7660
+    },
+    {
+      "epoch": 0.8776747911660373,
+      "grad_norm": 0.4426518976688385,
+      "learning_rate": 2.317231333650005e-06,
+      "loss": 1.2879,
+      "num_input_tokens_seen": 62832640,
+      "step": 7670
+    },
+    {
+      "epoch": 0.8788190868520426,
+      "grad_norm": 0.5259247422218323,
+      "learning_rate": 2.2747475774077986e-06,
+      "loss": 1.1017,
+      "num_input_tokens_seen": 62914560,
+      "step": 7680
+    },
+    {
+      "epoch": 0.8799633825380478,
+      "grad_norm": 0.4538893401622772,
+      "learning_rate": 2.2326383093947135e-06,
+      "loss": 1.2244,
+      "num_input_tokens_seen": 62996480,
+      "step": 7690
+    },
+    {
+      "epoch": 0.8811076782240531,
+      "grad_norm": 0.6097177267074585,
+      "learning_rate": 2.1909042235271597e-06,
+      "loss": 1.3507,
+      "num_input_tokens_seen": 63078400,
+      "step": 7700
+    },
+    {
+      "epoch": 0.8822519739100584,
+      "grad_norm": 0.5675943493843079,
+      "learning_rate": 2.1495460075389133e-06,
+      "loss": 1.0351,
+      "num_input_tokens_seen": 63160320,
+      "step": 7710
+    },
+    {
+      "epoch": 0.8833962695960637,
+      "grad_norm": 0.4782324731349945,
+      "learning_rate": 2.1085643429698236e-06,
+      "loss": 1.1214,
+      "num_input_tokens_seen": 63242240,
+      "step": 7720
+    },
+    {
+      "epoch": 0.8845405652820689,
+      "grad_norm": 0.5161313414573669,
+      "learning_rate": 2.067959905154568e-06,
+      "loss": 1.1543,
+      "num_input_tokens_seen": 63324160,
+      "step": 7730
+    },
+    {
+      "epoch": 0.8856848609680742,
+      "grad_norm": 0.4714735150337219,
+      "learning_rate": 2.0277333632115288e-06,
+      "loss": 0.9881,
+      "num_input_tokens_seen": 63406080,
+      "step": 7740
+    },
+    {
+      "epoch": 0.8868291566540795,
+      "grad_norm": 0.5167281031608582,
+      "learning_rate": 1.9878853800317535e-06,
+      "loss": 1.4237,
+      "num_input_tokens_seen": 63488000,
+      "step": 7750
+    },
+    {
+      "epoch": 0.8879734523400846,
+      "grad_norm": 0.47065913677215576,
+      "learning_rate": 1.948416612268034e-06,
+      "loss": 1.1573,
+      "num_input_tokens_seen": 63569920,
+      "step": 7760
+    },
+    {
+      "epoch": 0.8891177480260899,
+      "grad_norm": 0.6462358236312866,
+      "learning_rate": 1.909327710324116e-06,
+      "loss": 1.303,
+      "num_input_tokens_seen": 63651840,
+      "step": 7770
+    },
+    {
+      "epoch": 0.8902620437120952,
+      "grad_norm": 0.4849631190299988,
+      "learning_rate": 1.8706193183439247e-06,
+      "loss": 1.2495,
+      "num_input_tokens_seen": 63733760,
+      "step": 7780
+    },
+    {
+      "epoch": 0.8914063393981004,
+      "grad_norm": 0.4865788519382477,
+      "learning_rate": 1.8322920742010086e-06,
+      "loss": 1.1877,
+      "num_input_tokens_seen": 63815680,
+      "step": 7790
+    },
+    {
+      "epoch": 0.8925506350841057,
+      "grad_norm": 0.48363032937049866,
+      "learning_rate": 1.7943466094879902e-06,
+      "loss": 1.1676,
+      "num_input_tokens_seen": 63897600,
+      "step": 7800
+    },
+    {
+      "epoch": 0.893694930770111,
+      "grad_norm": 0.6026335954666138,
+      "learning_rate": 1.7567835495061718e-06,
+      "loss": 1.1575,
+      "num_input_tokens_seen": 63979520,
+      "step": 7810
+    },
+    {
+      "epoch": 0.8948392264561162,
+      "grad_norm": 0.49204021692276,
+      "learning_rate": 1.7196035132552135e-06,
+      "loss": 1.0517,
+      "num_input_tokens_seen": 64061440,
+      "step": 7820
+    },
+    {
+      "epoch": 0.8959835221421215,
+      "grad_norm": 0.5136172771453857,
+      "learning_rate": 1.682807113422971e-06,
+      "loss": 1.4192,
+      "num_input_tokens_seen": 64143360,
+      "step": 7830
+    },
+    {
+      "epoch": 0.8971278178281268,
+      "grad_norm": 0.516801655292511,
+      "learning_rate": 1.646394956375369e-06,
+      "loss": 1.5286,
+      "num_input_tokens_seen": 64225280,
+      "step": 7840
+    },
+    {
+      "epoch": 0.898272113514132,
+      "grad_norm": 1.405717134475708,
+      "learning_rate": 1.6103676421463986e-06,
+      "loss": 0.9688,
+      "num_input_tokens_seen": 64307200,
+      "step": 7850
+    },
+    {
+      "epoch": 0.8994164092001373,
+      "grad_norm": 0.4660581350326538,
+      "learning_rate": 1.5747257644282726e-06,
+      "loss": 1.1801,
+      "num_input_tokens_seen": 64389120,
+      "step": 7860
+    },
+    {
+      "epoch": 0.9005607048861426,
+      "grad_norm": 0.5325869917869568,
+      "learning_rate": 1.5394699105616002e-06,
+      "loss": 1.0305,
+      "num_input_tokens_seen": 64471040,
+      "step": 7870
+    },
+    {
+      "epoch": 0.9017050005721479,
+      "grad_norm": 0.5679177641868591,
+      "learning_rate": 1.504600661525718e-06,
+      "loss": 1.1399,
+      "num_input_tokens_seen": 64552960,
+      "step": 7880
+    },
+    {
+      "epoch": 0.9028492962581531,
+      "grad_norm": 0.49606427550315857,
+      "learning_rate": 1.4701185919291372e-06,
+      "loss": 1.4673,
+      "num_input_tokens_seen": 64634880,
+      "step": 7890
+    },
+    {
+      "epoch": 0.9039935919441584,
+      "grad_norm": 0.49286141991615295,
+      "learning_rate": 1.436024270000058e-06,
+      "loss": 1.0752,
+      "num_input_tokens_seen": 64716800,
+      "step": 7900
+    },
+    {
+      "epoch": 0.9051378876301637,
+      "grad_norm": 0.5626493096351624,
+      "learning_rate": 1.4023182575769956e-06,
+      "loss": 1.1824,
+      "num_input_tokens_seen": 64798720,
+      "step": 7910
+    },
+    {
+      "epoch": 0.9062821833161689,
+      "grad_norm": 0.513041079044342,
+      "learning_rate": 1.3690011100995437e-06,
+      "loss": 1.0409,
+      "num_input_tokens_seen": 64880640,
+      "step": 7920
+    },
+    {
+      "epoch": 0.9074264790021742,
+      "grad_norm": 0.46281594038009644,
+      "learning_rate": 1.3360733765992116e-06,
+      "loss": 1.2628,
+      "num_input_tokens_seen": 64962560,
+      "step": 7930
+    },
+    {
+      "epoch": 0.9085707746881794,
+      "grad_norm": 0.626043975353241,
+      "learning_rate": 1.3035355996903697e-06,
+      "loss": 1.307,
+      "num_input_tokens_seen": 65044480,
+      "step": 7940
+    },
+    {
+      "epoch": 0.9097150703741846,
+      "grad_norm": 0.6638424396514893,
+      "learning_rate": 1.2713883155613144e-06,
+      "loss": 1.061,
+      "num_input_tokens_seen": 65126400,
+      "step": 7950
+    },
+    {
+      "epoch": 0.9108593660601899,
+      "grad_norm": 0.8179208040237427,
+      "learning_rate": 1.2396320539654366e-06,
+      "loss": 1.178,
+      "num_input_tokens_seen": 65208320,
+      "step": 7960
+    },
+    {
+      "epoch": 0.9120036617461952,
+      "grad_norm": 0.44587135314941406,
+      "learning_rate": 1.208267338212493e-06,
+      "loss": 1.3368,
+      "num_input_tokens_seen": 65290240,
+      "step": 7970
+    },
+    {
+      "epoch": 0.9131479574322005,
+      "grad_norm": 0.4682336449623108,
+      "learning_rate": 1.177294685159963e-06,
+      "loss": 1.3091,
+      "num_input_tokens_seen": 65372160,
+      "step": 7980
+    },
+    {
+      "epoch": 0.9142922531182057,
+      "grad_norm": 0.4648427367210388,
+      "learning_rate": 1.1467146052045603e-06,
+      "loss": 1.2932,
+      "num_input_tokens_seen": 65454080,
+      "step": 7990
+    },
+    {
+      "epoch": 0.915436548804211,
+      "grad_norm": 0.5007482171058655,
+      "learning_rate": 1.1165276022737926e-06,
+      "loss": 1.1777,
+      "num_input_tokens_seen": 65536000,
+      "step": 8000
+    },
+    {
+      "epoch": 0.9165808444902163,
+      "grad_norm": 0.5155405402183533,
+      "learning_rate": 1.0867341738176857e-06,
+      "loss": 1.3184,
+      "num_input_tokens_seen": 65617920,
+      "step": 8010
+    },
+    {
+      "epoch": 0.9177251401762215,
+      "grad_norm": 0.6284940242767334,
+      "learning_rate": 1.0573348108005614e-06,
+      "loss": 1.2283,
+      "num_input_tokens_seen": 65699840,
+      "step": 8020
+    },
+    {
+      "epoch": 0.9188694358622268,
+      "grad_norm": 4.340449333190918,
+      "learning_rate": 1.0283299976929672e-06,
+      "loss": 1.4513,
+      "num_input_tokens_seen": 65781760,
+      "step": 8030
+    },
+    {
+      "epoch": 0.9200137315482321,
+      "grad_norm": 0.5452487468719482,
+      "learning_rate": 9.997202124636785e-07,
+      "loss": 1.2401,
+      "num_input_tokens_seen": 65863680,
+      "step": 8040
+    },
+    {
+      "epoch": 0.9211580272342373,
+      "grad_norm": 0.5035956501960754,
+      "learning_rate": 9.715059265718335e-07,
+      "loss": 1.0646,
+      "num_input_tokens_seen": 65945600,
+      "step": 8050
+    },
+    {
+      "epoch": 0.9223023229202426,
+      "grad_norm": 0.4889717996120453,
+      "learning_rate": 9.436876049591398e-07,
+      "loss": 1.4221,
+      "num_input_tokens_seen": 66027520,
+      "step": 8060
+    },
+    {
+      "epoch": 0.9234466186062479,
+      "grad_norm": 0.48875346779823303,
+      "learning_rate": 9.162657060422574e-07,
+      "loss": 1.2108,
+      "num_input_tokens_seen": 66109440,
+      "step": 8070
+    },
+    {
+      "epoch": 0.9245909142922532,
+      "grad_norm": 0.5173611044883728,
+      "learning_rate": 8.892406817051946e-07,
+      "loss": 1.4522,
+      "num_input_tokens_seen": 66191360,
+      "step": 8080
+    },
+    {
+      "epoch": 0.9257352099782584,
+      "grad_norm": 0.598610520362854,
+      "learning_rate": 8.626129772918962e-07,
+      "loss": 1.2966,
+      "num_input_tokens_seen": 66273280,
+      "step": 8090
+    },
+    {
+      "epoch": 0.9268795056642637,
+      "grad_norm": 0.5002040863037109,
+      "learning_rate": 8.363830315988947e-07,
+      "loss": 1.2018,
+      "num_input_tokens_seen": 66355200,
+      "step": 8100
+    },
+    {
+      "epoch": 0.928023801350269,
+      "grad_norm": 0.42786675691604614,
+      "learning_rate": 8.105512768680712e-07,
+      "loss": 1.1427,
+      "num_input_tokens_seen": 66437120,
+      "step": 8110
+    },
+    {
+      "epoch": 0.9291680970362741,
+      "grad_norm": 0.5229976177215576,
+      "learning_rate": 7.851181387795392e-07,
+      "loss": 1.1591,
+      "num_input_tokens_seen": 66519040,
+      "step": 8120
+    },
+    {
+      "epoch": 0.9303123927222794,
+      "grad_norm": 0.469088613986969,
+      "learning_rate": 7.600840364446333e-07,
+      "loss": 1.4109,
+      "num_input_tokens_seen": 66600960,
+      "step": 8130
+    },
+    {
+      "epoch": 0.9314566884082847,
+      "grad_norm": 0.46560269594192505,
+      "learning_rate": 7.354493823990006e-07,
+      "loss": 1.4194,
+      "num_input_tokens_seen": 66682880,
+      "step": 8140
+    },
+    {
+      "epoch": 0.9326009840942899,
+      "grad_norm": 0.6745224595069885,
+      "learning_rate": 7.112145825957927e-07,
+      "loss": 1.1635,
+      "num_input_tokens_seen": 66764800,
+      "step": 8150
+    },
+    {
+      "epoch": 0.9337452797802952,
+      "grad_norm": 0.6367943286895752,
+      "learning_rate": 6.873800363989935e-07,
+      "loss": 1.2245,
+      "num_input_tokens_seen": 66846720,
+      "step": 8160
+    },
+    {
+      "epoch": 0.9348895754663005,
+      "grad_norm": 0.9506150484085083,
+      "learning_rate": 6.63946136576829e-07,
+      "loss": 1.4055,
+      "num_input_tokens_seen": 66928640,
+      "step": 8170
+    },
+    {
+      "epoch": 0.9360338711523057,
+      "grad_norm": 0.5516346096992493,
+      "learning_rate": 6.409132692952874e-07,
+      "loss": 1.134,
+      "num_input_tokens_seen": 67010560,
+      "step": 8180
+    },
+    {
+      "epoch": 0.937178166838311,
+      "grad_norm": 0.48264941573143005,
+      "learning_rate": 6.182818141117625e-07,
+      "loss": 1.1915,
+      "num_input_tokens_seen": 67092480,
+      "step": 8190
+    },
+    {
+      "epoch": 0.9383224625243163,
+      "grad_norm": 0.5029363632202148,
+      "learning_rate": 5.960521439688088e-07,
+      "loss": 1.0641,
+      "num_input_tokens_seen": 67174400,
+      "step": 8200
+    },
+    {
+      "epoch": 0.9394667582103216,
+      "grad_norm": 0.5256086587905884,
+      "learning_rate": 5.742246251879829e-07,
+      "loss": 1.1629,
+      "num_input_tokens_seen": 67256320,
+      "step": 8210
+    },
+    {
+      "epoch": 0.9406110538963268,
+      "grad_norm": 0.5969172120094299,
+      "learning_rate": 5.527996174638061e-07,
+      "loss": 1.1297,
+      "num_input_tokens_seen": 67338240,
+      "step": 8220
+    },
+    {
+      "epoch": 0.9417553495823321,
+      "grad_norm": 0.6032472252845764,
+      "learning_rate": 5.317774738578446e-07,
+      "loss": 1.269,
+      "num_input_tokens_seen": 67420160,
+      "step": 8230
+    },
+    {
+      "epoch": 0.9428996452683374,
+      "grad_norm": 0.4904973804950714,
+      "learning_rate": 5.111585407928887e-07,
+      "loss": 1.0976,
+      "num_input_tokens_seen": 67502080,
+      "step": 8240
+    },
+    {
+      "epoch": 0.9440439409543426,
+      "grad_norm": 0.49423882365226746,
+      "learning_rate": 4.909431580472385e-07,
+      "loss": 1.2052,
+      "num_input_tokens_seen": 67584000,
+      "step": 8250
+    },
+    {
+      "epoch": 0.9451882366403479,
+      "grad_norm": 0.5731542110443115,
+      "learning_rate": 4.711316587491188e-07,
+      "loss": 1.1078,
+      "num_input_tokens_seen": 67665920,
+      "step": 8260
+    },
+    {
+      "epoch": 0.9463325323263532,
+      "grad_norm": 0.6199080348014832,
+      "learning_rate": 4.5172436937117036e-07,
+      "loss": 1.2543,
+      "num_input_tokens_seen": 67747840,
+      "step": 8270
+    },
+    {
+      "epoch": 0.9474768280123584,
+      "grad_norm": 0.5399270057678223,
+      "learning_rate": 4.3272160972509524e-07,
+      "loss": 1.1164,
+      "num_input_tokens_seen": 67829760,
+      "step": 8280
+    },
+    {
+      "epoch": 0.9486211236983637,
+      "grad_norm": 0.49898287653923035,
+      "learning_rate": 4.1412369295635023e-07,
+      "loss": 1.1804,
+      "num_input_tokens_seen": 67911680,
+      "step": 8290
+    },
+    {
+      "epoch": 0.9497654193843689,
+      "grad_norm": 0.5566830039024353,
+      "learning_rate": 3.9593092553902587e-07,
+      "loss": 1.1059,
+      "num_input_tokens_seen": 67993600,
+      "step": 8300
+    },
+    {
+      "epoch": 0.9509097150703741,
+      "grad_norm": 0.5725921988487244,
+      "learning_rate": 3.7814360727076724e-07,
+      "loss": 1.1965,
+      "num_input_tokens_seen": 68075520,
+      "step": 8310
+    },
+    {
+      "epoch": 0.9520540107563794,
+      "grad_norm": 0.5144397616386414,
+      "learning_rate": 3.607620312678528e-07,
+      "loss": 1.3606,
+      "num_input_tokens_seen": 68157440,
+      "step": 8320
+    },
+    {
+      "epoch": 0.9531983064423847,
+      "grad_norm": 0.5176523327827454,
+      "learning_rate": 3.437864839603455e-07,
+      "loss": 1.0969,
+      "num_input_tokens_seen": 68239360,
+      "step": 8330
+    },
+    {
+      "epoch": 0.95434260212839,
+      "grad_norm": 0.8841016888618469,
+      "learning_rate": 3.272172450873967e-07,
+      "loss": 1.6294,
+      "num_input_tokens_seen": 68321280,
+      "step": 8340
+    },
+    {
+      "epoch": 0.9554868978143952,
+      "grad_norm": 0.5609331727027893,
+      "learning_rate": 3.11054587692608e-07,
+      "loss": 1.0976,
+      "num_input_tokens_seen": 68403200,
+      "step": 8350
+    },
+    {
+      "epoch": 0.9566311935004005,
+      "grad_norm": 0.5448135137557983,
+      "learning_rate": 2.952987781195599e-07,
+      "loss": 1.1881,
+      "num_input_tokens_seen": 68485120,
+      "step": 8360
+    },
+    {
+      "epoch": 0.9577754891864058,
+      "grad_norm": 0.5337486863136292,
+      "learning_rate": 2.799500760073931e-07,
+      "loss": 1.1826,
+      "num_input_tokens_seen": 68567040,
+      "step": 8370
+    },
+    {
+      "epoch": 0.958919784872411,
+      "grad_norm": 0.5066606402397156,
+      "learning_rate": 2.6500873428656483e-07,
+      "loss": 1.1567,
+      "num_input_tokens_seen": 68648960,
+      "step": 8380
+    },
+    {
+      "epoch": 0.9600640805584163,
+      "grad_norm": 0.5534819960594177,
+      "learning_rate": 2.5047499917464636e-07,
+      "loss": 0.9999,
+      "num_input_tokens_seen": 68730880,
+      "step": 8390
+    },
+    {
+      "epoch": 0.9612083762444216,
+      "grad_norm": 0.5017804503440857,
+      "learning_rate": 2.3634911017229034e-07,
+      "loss": 1.347,
+      "num_input_tokens_seen": 68812800,
+      "step": 8400
+    },
+    {
+      "epoch": 0.9623526719304268,
+      "grad_norm": 0.4598214328289032,
+      "learning_rate": 2.2263130005927558e-07,
+      "loss": 1.0357,
+      "num_input_tokens_seen": 68894720,
+      "step": 8410
+    },
+    {
+      "epoch": 0.9634969676164321,
+      "grad_norm": 0.5540017485618591,
+      "learning_rate": 2.0932179489066006e-07,
+      "loss": 1.3101,
+      "num_input_tokens_seen": 68976640,
+      "step": 8420
+    },
+    {
+      "epoch": 0.9646412633024374,
+      "grad_norm": 0.7891622185707092,
+      "learning_rate": 1.9642081399307844e-07,
+      "loss": 1.4347,
+      "num_input_tokens_seen": 69058560,
+      "step": 8430
+    },
+    {
+      "epoch": 0.9657855589884426,
+      "grad_norm": 0.8638253211975098,
+      "learning_rate": 1.8392856996110875e-07,
+      "loss": 0.9711,
+      "num_input_tokens_seen": 69140480,
+      "step": 8440
+    },
+    {
+      "epoch": 0.9669298546744479,
+      "grad_norm": 0.49595341086387634,
+      "learning_rate": 1.7184526865377805e-07,
+      "loss": 1.1246,
+      "num_input_tokens_seen": 69222400,
+      "step": 8450
+    },
+    {
+      "epoch": 0.9680741503604532,
+      "grad_norm": 0.5145378708839417,
+      "learning_rate": 1.6017110919116786e-07,
+      "loss": 1.182,
+      "num_input_tokens_seen": 69304320,
+      "step": 8460
+    },
+    {
+      "epoch": 0.9692184460464585,
+      "grad_norm": 0.5098930597305298,
+      "learning_rate": 1.4890628395113072e-07,
+      "loss": 1.2913,
+      "num_input_tokens_seen": 69386240,
+      "step": 8470
+    },
+    {
+      "epoch": 0.9703627417324636,
+      "grad_norm": 0.5276376008987427,
+      "learning_rate": 1.380509785661288e-07,
+      "loss": 1.1669,
+      "num_input_tokens_seen": 69468160,
+      "step": 8480
+    },
+    {
+      "epoch": 0.9715070374184689,
+      "grad_norm": 0.5354599952697754,
+      "learning_rate": 1.2760537192015866e-07,
+      "loss": 1.2136,
+      "num_input_tokens_seen": 69550080,
+      "step": 8490
+    },
+    {
+      "epoch": 0.9726513331044742,
+      "grad_norm": 0.43569329380989075,
+      "learning_rate": 1.1756963614582006e-07,
+      "loss": 1.4041,
+      "num_input_tokens_seen": 69632000,
+      "step": 8500
+    },
+    {
+      "epoch": 0.9737956287904794,
+      "grad_norm": 0.4199809432029724,
+      "learning_rate": 1.0794393662147129e-07,
+      "loss": 1.4145,
+      "num_input_tokens_seen": 69713920,
+      "step": 8510
+    },
+    {
+      "epoch": 0.9749399244764847,
+      "grad_norm": 0.45975956320762634,
+      "learning_rate": 9.872843196850057e-08,
+      "loss": 1.1534,
+      "num_input_tokens_seen": 69795840,
+      "step": 8520
+    },
+    {
+      "epoch": 0.97608422016249,
+      "grad_norm": 0.47319695353507996,
+      "learning_rate": 8.992327404872825e-08,
+      "loss": 1.1406,
+      "num_input_tokens_seen": 69877760,
+      "step": 8530
+    },
+    {
+      "epoch": 0.9772285158484952,
+      "grad_norm": 0.6402721405029297,
+      "learning_rate": 8.152860796187545e-08,
+      "loss": 1.4171,
+      "num_input_tokens_seen": 69959680,
+      "step": 8540
+    },
+    {
+      "epoch": 0.9783728115345005,
+      "grad_norm": 0.5459085702896118,
+      "learning_rate": 7.354457204320486e-08,
+      "loss": 1.2596,
+      "num_input_tokens_seen": 70041600,
+      "step": 8550
+    },
+    {
+      "epoch": 0.9795171072205058,
+      "grad_norm": 0.49747464060783386,
+      "learning_rate": 6.59712978612198e-08,
+      "loss": 1.4422,
+      "num_input_tokens_seen": 70123520,
+      "step": 8560
+    },
+    {
+      "epoch": 0.980661402906511,
+      "grad_norm": 0.5260741114616394,
+      "learning_rate": 5.880891021549928e-08,
+      "loss": 1.1456,
+      "num_input_tokens_seen": 70205440,
+      "step": 8570
+    },
+    {
+      "epoch": 0.9818056985925163,
+      "grad_norm": 0.5129789113998413,
+      "learning_rate": 5.205752713465794e-08,
+      "loss": 1.235,
+      "num_input_tokens_seen": 70287360,
+      "step": 8580
+    },
+    {
+      "epoch": 0.9829499942785216,
+      "grad_norm": 0.44936448335647583,
+      "learning_rate": 4.57172598743727e-08,
+      "loss": 1.2076,
+      "num_input_tokens_seen": 70369280,
+      "step": 8590
+    },
+    {
+      "epoch": 0.9840942899645269,
+      "grad_norm": 0.5832532644271851,
+      "learning_rate": 3.9788212915573e-08,
+      "loss": 1.1558,
+      "num_input_tokens_seen": 70451200,
+      "step": 8600
+    },
+    {
+      "epoch": 0.9852385856505321,
+      "grad_norm": 0.44010186195373535,
+      "learning_rate": 3.427048396271171e-08,
+      "loss": 1.2666,
+      "num_input_tokens_seen": 70533120,
+      "step": 8610
+    },
+    {
+      "epoch": 0.9863828813365374,
+      "grad_norm": 0.479427695274353,
+      "learning_rate": 2.9164163942146937e-08,
+      "loss": 1.4532,
+      "num_input_tokens_seen": 70615040,
+      "step": 8620
+    },
+    {
+      "epoch": 0.9875271770225427,
+      "grad_norm": 0.5471181273460388,
+      "learning_rate": 2.44693370006599e-08,
+      "loss": 1.4154,
+      "num_input_tokens_seen": 70696960,
+      "step": 8630
+    },
+    {
+      "epoch": 0.9886714727085479,
+      "grad_norm": 0.5122755765914917,
+      "learning_rate": 2.0186080504050466e-08,
+      "loss": 1.062,
+      "num_input_tokens_seen": 70778880,
+      "step": 8640
+    },
+    {
+      "epoch": 0.9898157683945531,
+      "grad_norm": 0.5013795495033264,
+      "learning_rate": 1.6314465035879855e-08,
+      "loss": 1.0299,
+      "num_input_tokens_seen": 70860800,
+      "step": 8650
+    },
+    {
+      "epoch": 0.9909600640805584,
+      "grad_norm": 0.4794003963470459,
+      "learning_rate": 1.2854554396291018e-08,
+      "loss": 1.1402,
+      "num_input_tokens_seen": 70942720,
+      "step": 8660
+    },
+    {
+      "epoch": 0.9921043597665636,
+      "grad_norm": 1.2419332265853882,
+      "learning_rate": 9.806405600967794e-09,
+      "loss": 1.3485,
+      "num_input_tokens_seen": 71024640,
+      "step": 8670
+    },
+    {
+      "epoch": 0.9932486554525689,
+      "grad_norm": 0.5120067596435547,
+      "learning_rate": 7.1700688801940034e-09,
+      "loss": 1.26,
+      "num_input_tokens_seen": 71106560,
+      "step": 8680
+    },
+    {
+      "epoch": 0.9943929511385742,
+      "grad_norm": 0.5355421900749207,
+      "learning_rate": 4.94558767802078e-09,
+      "loss": 1.1129,
+      "num_input_tokens_seen": 71188480,
+      "step": 8690
+    },
+    {
+      "epoch": 0.9955372468245794,
+      "grad_norm": 0.6079626083374023,
+      "learning_rate": 3.1329986515560295e-09,
+      "loss": 1.2407,
+      "num_input_tokens_seen": 71270400,
+      "step": 8700
+    },
+    {
+      "epoch": 0.9966815425105847,
+      "grad_norm": 0.48402678966522217,
+      "learning_rate": 1.7323316703621305e-09,
+      "loss": 1.2159,
+      "num_input_tokens_seen": 71352320,
+      "step": 8710
+    },
+    {
+      "epoch": 0.99782583819659,
+      "grad_norm": 0.45963913202285767,
+      "learning_rate": 7.436098159480099e-10,
+      "loss": 1.0369,
+      "num_input_tokens_seen": 71434240,
+      "step": 8720
+    },
+    {
+      "epoch": 0.9989701338825953,
+      "grad_norm": 0.5231521725654602,
+      "learning_rate": 1.6684938141664498e-10,
+      "loss": 1.1446,
+      "num_input_tokens_seen": 71516160,
+      "step": 8730
+    },
+    {
+      "epoch": 1.0,
+      "num_input_tokens_seen": 71589888,
+      "step": 8739,
+      "total_flos": 3.2596926707332547e+18,
+      "train_loss": 1.260579330722832,
+      "train_runtime": 44987.2731,
+      "train_samples_per_second": 0.194,
+      "train_steps_per_second": 0.194
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8739,
+  "num_input_tokens_seen": 71589888,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.2596926707332547e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}