diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7406 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 1052,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0019011406844106464,
+      "grad_norm": 3.417204165198468,
+      "learning_rate": 1.886792452830189e-07,
+      "loss": 0.7011,
+      "step": 1
+    },
+    {
+      "epoch": 0.0038022813688212928,
+      "grad_norm": 3.166157807561183,
+      "learning_rate": 3.773584905660378e-07,
+      "loss": 0.7101,
+      "step": 2
+    },
+    {
+      "epoch": 0.005703422053231939,
+      "grad_norm": 3.279205229040061,
+      "learning_rate": 5.660377358490567e-07,
+      "loss": 0.6891,
+      "step": 3
+    },
+    {
+      "epoch": 0.0076045627376425855,
+      "grad_norm": 3.2492416956945984,
+      "learning_rate": 7.547169811320755e-07,
+      "loss": 0.7193,
+      "step": 4
+    },
+    {
+      "epoch": 0.009505703422053232,
+      "grad_norm": 3.4427153509078754,
+      "learning_rate": 9.433962264150944e-07,
+      "loss": 0.7494,
+      "step": 5
+    },
+    {
+      "epoch": 0.011406844106463879,
+      "grad_norm": 3.0252626744532463,
+      "learning_rate": 1.1320754716981133e-06,
+      "loss": 0.7102,
+      "step": 6
+    },
+    {
+      "epoch": 0.013307984790874524,
+      "grad_norm": 3.158549478478432,
+      "learning_rate": 1.3207547169811322e-06,
+      "loss": 0.6877,
+      "step": 7
+    },
+    {
+      "epoch": 0.015209125475285171,
+      "grad_norm": 2.9058547765478,
+      "learning_rate": 1.509433962264151e-06,
+      "loss": 0.7019,
+      "step": 8
+    },
+    {
+      "epoch": 0.017110266159695818,
+      "grad_norm": 2.502993884200636,
+      "learning_rate": 1.6981132075471698e-06,
+      "loss": 0.7034,
+      "step": 9
+    },
+    {
+      "epoch": 0.019011406844106463,
+      "grad_norm": 2.2508996187569963,
+      "learning_rate": 1.8867924528301889e-06,
+      "loss": 0.6858,
+      "step": 10
+    },
+    {
+      "epoch": 0.02091254752851711,
+      "grad_norm": 1.58887064134842,
+      "learning_rate": 2.075471698113208e-06,
+      "loss": 0.661,
+      "step": 11
+    },
+    {
+      "epoch": 0.022813688212927757,
+      "grad_norm": 1.725768498676427,
+      "learning_rate": 2.2641509433962266e-06,
+      "loss": 0.6144,
+      "step": 12
+    },
+    {
+      "epoch": 0.024714828897338403,
+      "grad_norm": 1.605959053090016,
+      "learning_rate": 2.4528301886792453e-06,
+      "loss": 0.6011,
+      "step": 13
+    },
+    {
+      "epoch": 0.026615969581749048,
+      "grad_norm": 1.436543424379693,
+      "learning_rate": 2.6415094339622644e-06,
+      "loss": 0.5913,
+      "step": 14
+    },
+    {
+      "epoch": 0.028517110266159697,
+      "grad_norm": 1.9220056832166528,
+      "learning_rate": 2.830188679245283e-06,
+      "loss": 0.597,
+      "step": 15
+    },
+    {
+      "epoch": 0.030418250950570342,
+      "grad_norm": 2.3105218967746466,
+      "learning_rate": 3.018867924528302e-06,
+      "loss": 0.5766,
+      "step": 16
+    },
+    {
+      "epoch": 0.03231939163498099,
+      "grad_norm": 2.3955260373458347,
+      "learning_rate": 3.207547169811321e-06,
+      "loss": 0.6069,
+      "step": 17
+    },
+    {
+      "epoch": 0.034220532319391636,
+      "grad_norm": 1.728865236659783,
+      "learning_rate": 3.3962264150943395e-06,
+      "loss": 0.5875,
+      "step": 18
+    },
+    {
+      "epoch": 0.03612167300380228,
+      "grad_norm": 1.480017300597635,
+      "learning_rate": 3.5849056603773586e-06,
+      "loss": 0.5584,
+      "step": 19
+    },
+    {
+      "epoch": 0.03802281368821293,
+      "grad_norm": 1.1399821289031777,
+      "learning_rate": 3.7735849056603777e-06,
+      "loss": 0.563,
+      "step": 20
+    },
+    {
+      "epoch": 0.039923954372623575,
+      "grad_norm": 1.2810830825057045,
+      "learning_rate": 3.962264150943396e-06,
+      "loss": 0.5482,
+      "step": 21
+    },
+    {
+      "epoch": 0.04182509505703422,
+      "grad_norm": 1.347980096258214,
+      "learning_rate": 4.150943396226416e-06,
+      "loss": 0.5987,
+      "step": 22
+    },
+    {
+      "epoch": 0.043726235741444866,
+      "grad_norm": 1.4238951251311742,
+      "learning_rate": 4.339622641509435e-06,
+      "loss": 0.576,
+      "step": 23
+    },
+    {
+      "epoch": 0.045627376425855515,
+      "grad_norm": 1.1105240650881238,
+      "learning_rate": 4.528301886792453e-06,
+      "loss": 0.5565,
+      "step": 24
+    },
+    {
+      "epoch": 0.04752851711026616,
+      "grad_norm": 1.0450785437152297,
+      "learning_rate": 4.716981132075472e-06,
+      "loss": 0.5575,
+      "step": 25
+    },
+    {
+      "epoch": 0.049429657794676805,
+      "grad_norm": 0.970713607725702,
+      "learning_rate": 4.905660377358491e-06,
+      "loss": 0.561,
+      "step": 26
+    },
+    {
+      "epoch": 0.051330798479087454,
+      "grad_norm": 1.0540546940089415,
+      "learning_rate": 5.09433962264151e-06,
+      "loss": 0.5313,
+      "step": 27
+    },
+    {
+      "epoch": 0.053231939163498096,
+      "grad_norm": 0.969549015497069,
+      "learning_rate": 5.283018867924529e-06,
+      "loss": 0.5288,
+      "step": 28
+    },
+    {
+      "epoch": 0.055133079847908745,
+      "grad_norm": 1.008437748126313,
+      "learning_rate": 5.4716981132075475e-06,
+      "loss": 0.5283,
+      "step": 29
+    },
+    {
+      "epoch": 0.057034220532319393,
+      "grad_norm": 0.9198708469611205,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 0.5213,
+      "step": 30
+    },
+    {
+      "epoch": 0.058935361216730035,
+      "grad_norm": 0.8911406250599365,
+      "learning_rate": 5.849056603773585e-06,
+      "loss": 0.5216,
+      "step": 31
+    },
+    {
+      "epoch": 0.060836501901140684,
+      "grad_norm": 0.8602053176588361,
+      "learning_rate": 6.037735849056604e-06,
+      "loss": 0.501,
+      "step": 32
+    },
+    {
+      "epoch": 0.06273764258555133,
+      "grad_norm": 0.8945005572295428,
+      "learning_rate": 6.226415094339623e-06,
+      "loss": 0.512,
+      "step": 33
+    },
+    {
+      "epoch": 0.06463878326996197,
+      "grad_norm": 0.9281572289783664,
+      "learning_rate": 6.415094339622642e-06,
+      "loss": 0.5244,
+      "step": 34
+    },
+    {
+      "epoch": 0.06653992395437262,
+      "grad_norm": 0.936344200015766,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.5355,
+      "step": 35
+    },
+    {
+      "epoch": 0.06844106463878327,
+      "grad_norm": 0.8599876977308395,
+      "learning_rate": 6.792452830188679e-06,
+      "loss": 0.5172,
+      "step": 36
+    },
+    {
+      "epoch": 0.07034220532319392,
+      "grad_norm": 0.8207006388260467,
+      "learning_rate": 6.981132075471699e-06,
+      "loss": 0.4642,
+      "step": 37
+    },
+    {
+      "epoch": 0.07224334600760456,
+      "grad_norm": 0.9288223033334503,
+      "learning_rate": 7.169811320754717e-06,
+      "loss": 0.5452,
+      "step": 38
+    },
+    {
+      "epoch": 0.0741444866920152,
+      "grad_norm": 0.9221944941672239,
+      "learning_rate": 7.358490566037736e-06,
+      "loss": 0.5333,
+      "step": 39
+    },
+    {
+      "epoch": 0.07604562737642585,
+      "grad_norm": 0.8315334334128202,
+      "learning_rate": 7.5471698113207555e-06,
+      "loss": 0.502,
+      "step": 40
+    },
+    {
+      "epoch": 0.0779467680608365,
+      "grad_norm": 0.8182263145542923,
+      "learning_rate": 7.735849056603775e-06,
+      "loss": 0.5005,
+      "step": 41
+    },
+    {
+      "epoch": 0.07984790874524715,
+      "grad_norm": 0.7519422805517089,
+      "learning_rate": 7.924528301886793e-06,
+      "loss": 0.4949,
+      "step": 42
+    },
+    {
+      "epoch": 0.0817490494296578,
+      "grad_norm": 0.799114433836668,
+      "learning_rate": 8.113207547169812e-06,
+      "loss": 0.4722,
+      "step": 43
+    },
+    {
+      "epoch": 0.08365019011406843,
+      "grad_norm": 0.8098614779123615,
+      "learning_rate": 8.301886792452832e-06,
+      "loss": 0.4549,
+      "step": 44
+    },
+    {
+      "epoch": 0.08555133079847908,
+      "grad_norm": 0.819190583285894,
+      "learning_rate": 8.49056603773585e-06,
+      "loss": 0.5122,
+      "step": 45
+    },
+    {
+      "epoch": 0.08745247148288973,
+      "grad_norm": 0.7771545946614856,
+      "learning_rate": 8.67924528301887e-06,
+      "loss": 0.4775,
+      "step": 46
+    },
+    {
+      "epoch": 0.08935361216730038,
+      "grad_norm": 0.7826377222179967,
+      "learning_rate": 8.867924528301887e-06,
+      "loss": 0.4917,
+      "step": 47
+    },
+    {
+      "epoch": 0.09125475285171103,
+      "grad_norm": 0.7874338914775674,
+      "learning_rate": 9.056603773584907e-06,
+      "loss": 0.4722,
+      "step": 48
+    },
+    {
+      "epoch": 0.09315589353612168,
+      "grad_norm": 0.829116774708427,
+      "learning_rate": 9.245283018867926e-06,
+      "loss": 0.4856,
+      "step": 49
+    },
+    {
+      "epoch": 0.09505703422053231,
+      "grad_norm": 0.7921034625500607,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 0.4906,
+      "step": 50
+    },
+    {
+      "epoch": 0.09695817490494296,
+      "grad_norm": 0.8225701767879259,
+      "learning_rate": 9.622641509433963e-06,
+      "loss": 0.4769,
+      "step": 51
+    },
+    {
+      "epoch": 0.09885931558935361,
+      "grad_norm": 0.8531947375668958,
+      "learning_rate": 9.811320754716981e-06,
+      "loss": 0.4968,
+      "step": 52
+    },
+    {
+      "epoch": 0.10076045627376426,
+      "grad_norm": 0.9010720053634059,
+      "learning_rate": 1e-05,
+      "loss": 0.4829,
+      "step": 53
+    },
+    {
+      "epoch": 0.10266159695817491,
+      "grad_norm": 0.7714328393188834,
+      "learning_rate": 1.018867924528302e-05,
+      "loss": 0.4914,
+      "step": 54
+    },
+    {
+      "epoch": 0.10456273764258556,
+      "grad_norm": 0.8784000098522912,
+      "learning_rate": 1.0377358490566038e-05,
+      "loss": 0.4966,
+      "step": 55
+    },
+    {
+      "epoch": 0.10646387832699619,
+      "grad_norm": 0.8909819896545116,
+      "learning_rate": 1.0566037735849058e-05,
+      "loss": 0.4672,
+      "step": 56
+    },
+    {
+      "epoch": 0.10836501901140684,
+      "grad_norm": 0.8548288617033107,
+      "learning_rate": 1.0754716981132076e-05,
+      "loss": 0.4759,
+      "step": 57
+    },
+    {
+      "epoch": 0.11026615969581749,
+      "grad_norm": 0.9803966489687058,
+      "learning_rate": 1.0943396226415095e-05,
+      "loss": 0.4934,
+      "step": 58
+    },
+    {
+      "epoch": 0.11216730038022814,
+      "grad_norm": 0.8137052382540544,
+      "learning_rate": 1.1132075471698115e-05,
+      "loss": 0.4862,
+      "step": 59
+    },
+    {
+      "epoch": 0.11406844106463879,
+      "grad_norm": 0.9753818680791325,
+      "learning_rate": 1.1320754716981132e-05,
+      "loss": 0.493,
+      "step": 60
+    },
+    {
+      "epoch": 0.11596958174904944,
+      "grad_norm": 0.8784599765135663,
+      "learning_rate": 1.1509433962264152e-05,
+      "loss": 0.4648,
+      "step": 61
+    },
+    {
+      "epoch": 0.11787072243346007,
+      "grad_norm": 0.9396076832901826,
+      "learning_rate": 1.169811320754717e-05,
+      "loss": 0.496,
+      "step": 62
+    },
+    {
+      "epoch": 0.11977186311787072,
+      "grad_norm": 0.8449366736998304,
+      "learning_rate": 1.188679245283019e-05,
+      "loss": 0.4606,
+      "step": 63
+    },
+    {
+      "epoch": 0.12167300380228137,
+      "grad_norm": 0.8603686207854464,
+      "learning_rate": 1.2075471698113209e-05,
+      "loss": 0.5053,
+      "step": 64
+    },
+    {
+      "epoch": 0.12357414448669202,
+      "grad_norm": 0.8189655861346682,
+      "learning_rate": 1.2264150943396227e-05,
+      "loss": 0.4736,
+      "step": 65
+    },
+    {
+      "epoch": 0.12547528517110265,
+      "grad_norm": 0.8495159372035639,
+      "learning_rate": 1.2452830188679246e-05,
+      "loss": 0.48,
+      "step": 66
+    },
+    {
+      "epoch": 0.12737642585551331,
+      "grad_norm": 0.7916107361839776,
+      "learning_rate": 1.2641509433962264e-05,
+      "loss": 0.4488,
+      "step": 67
+    },
+    {
+      "epoch": 0.12927756653992395,
+      "grad_norm": 0.8824936977140291,
+      "learning_rate": 1.2830188679245283e-05,
+      "loss": 0.4786,
+      "step": 68
+    },
+    {
+      "epoch": 0.1311787072243346,
+      "grad_norm": 0.8512531620323498,
+      "learning_rate": 1.3018867924528303e-05,
+      "loss": 0.5031,
+      "step": 69
+    },
+    {
+      "epoch": 0.13307984790874525,
+      "grad_norm": 0.8004547900674819,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 0.4803,
+      "step": 70
+    },
+    {
+      "epoch": 0.13498098859315588,
+      "grad_norm": 0.8259470916113447,
+      "learning_rate": 1.339622641509434e-05,
+      "loss": 0.472,
+      "step": 71
+    },
+    {
+      "epoch": 0.13688212927756654,
+      "grad_norm": 0.8089583409764832,
+      "learning_rate": 1.3584905660377358e-05,
+      "loss": 0.4897,
+      "step": 72
+    },
+    {
+      "epoch": 0.13878326996197718,
+      "grad_norm": 0.8659015719148437,
+      "learning_rate": 1.3773584905660378e-05,
+      "loss": 0.4742,
+      "step": 73
+    },
+    {
+      "epoch": 0.14068441064638784,
+      "grad_norm": 0.922324155931076,
+      "learning_rate": 1.3962264150943397e-05,
+      "loss": 0.4512,
+      "step": 74
+    },
+    {
+      "epoch": 0.14258555133079848,
+      "grad_norm": 0.8486837219662609,
+      "learning_rate": 1.4150943396226415e-05,
+      "loss": 0.4782,
+      "step": 75
+    },
+    {
+      "epoch": 0.1444866920152091,
+      "grad_norm": 0.9680552281346009,
+      "learning_rate": 1.4339622641509435e-05,
+      "loss": 0.4874,
+      "step": 76
+    },
+    {
+      "epoch": 0.14638783269961977,
+      "grad_norm": 0.8784178593983059,
+      "learning_rate": 1.4528301886792452e-05,
+      "loss": 0.493,
+      "step": 77
+    },
+    {
+      "epoch": 0.1482889733840304,
+      "grad_norm": 0.8343578654956217,
+      "learning_rate": 1.4716981132075472e-05,
+      "loss": 0.4695,
+      "step": 78
+    },
+    {
+      "epoch": 0.15019011406844107,
+      "grad_norm": 0.8511083266044375,
+      "learning_rate": 1.4905660377358491e-05,
+      "loss": 0.4937,
+      "step": 79
+    },
+    {
+      "epoch": 0.1520912547528517,
+      "grad_norm": 0.7980069897041593,
+      "learning_rate": 1.5094339622641511e-05,
+      "loss": 0.4847,
+      "step": 80
+    },
+    {
+      "epoch": 0.15399239543726237,
+      "grad_norm": 0.8364801012146286,
+      "learning_rate": 1.5283018867924532e-05,
+      "loss": 0.4916,
+      "step": 81
+    },
+    {
+      "epoch": 0.155893536121673,
+      "grad_norm": 0.7957012167134059,
+      "learning_rate": 1.547169811320755e-05,
+      "loss": 0.4624,
+      "step": 82
+    },
+    {
+      "epoch": 0.15779467680608364,
+      "grad_norm": 0.7926240201924292,
+      "learning_rate": 1.5660377358490568e-05,
+      "loss": 0.4606,
+      "step": 83
+    },
+    {
+      "epoch": 0.1596958174904943,
+      "grad_norm": 0.84176871585359,
+      "learning_rate": 1.5849056603773586e-05,
+      "loss": 0.4777,
+      "step": 84
+    },
+    {
+      "epoch": 0.16159695817490494,
+      "grad_norm": 0.8138527990345641,
+      "learning_rate": 1.6037735849056607e-05,
+      "loss": 0.485,
+      "step": 85
+    },
+    {
+      "epoch": 0.1634980988593156,
+      "grad_norm": 0.8550228503601863,
+      "learning_rate": 1.6226415094339625e-05,
+      "loss": 0.4552,
+      "step": 86
+    },
+    {
+      "epoch": 0.16539923954372623,
+      "grad_norm": 0.8225355634094624,
+      "learning_rate": 1.6415094339622643e-05,
+      "loss": 0.4913,
+      "step": 87
+    },
+    {
+      "epoch": 0.16730038022813687,
+      "grad_norm": 0.8797710867058435,
+      "learning_rate": 1.6603773584905664e-05,
+      "loss": 0.4878,
+      "step": 88
+    },
+    {
+      "epoch": 0.16920152091254753,
+      "grad_norm": 0.8032669330407621,
+      "learning_rate": 1.679245283018868e-05,
+      "loss": 0.4688,
+      "step": 89
+    },
+    {
+      "epoch": 0.17110266159695817,
+      "grad_norm": 0.838091177207545,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.4894,
+      "step": 90
+    },
+    {
+      "epoch": 0.17300380228136883,
+      "grad_norm": 0.8441501653091029,
+      "learning_rate": 1.716981132075472e-05,
+      "loss": 0.4588,
+      "step": 91
+    },
+    {
+      "epoch": 0.17490494296577946,
+      "grad_norm": 0.8271641510189525,
+      "learning_rate": 1.735849056603774e-05,
+      "loss": 0.4756,
+      "step": 92
+    },
+    {
+      "epoch": 0.17680608365019013,
+      "grad_norm": 0.7720396092230314,
+      "learning_rate": 1.7547169811320756e-05,
+      "loss": 0.5013,
+      "step": 93
+    },
+    {
+      "epoch": 0.17870722433460076,
+      "grad_norm": 0.8834357064240668,
+      "learning_rate": 1.7735849056603774e-05,
+      "loss": 0.4769,
+      "step": 94
+    },
+    {
+      "epoch": 0.1806083650190114,
+      "grad_norm": 0.8848091343505556,
+      "learning_rate": 1.7924528301886795e-05,
+      "loss": 0.4851,
+      "step": 95
+    },
+    {
+      "epoch": 0.18250950570342206,
+      "grad_norm": 0.8182572225541355,
+      "learning_rate": 1.8113207547169813e-05,
+      "loss": 0.4705,
+      "step": 96
+    },
+    {
+      "epoch": 0.1844106463878327,
+      "grad_norm": 0.8949393194188863,
+      "learning_rate": 1.830188679245283e-05,
+      "loss": 0.4828,
+      "step": 97
+    },
+    {
+      "epoch": 0.18631178707224336,
+      "grad_norm": 0.9612926938538086,
+      "learning_rate": 1.8490566037735852e-05,
+      "loss": 0.4387,
+      "step": 98
+    },
+    {
+      "epoch": 0.188212927756654,
+      "grad_norm": 0.8239502840418622,
+      "learning_rate": 1.867924528301887e-05,
+      "loss": 0.4633,
+      "step": 99
+    },
+    {
+      "epoch": 0.19011406844106463,
+      "grad_norm": 0.8331507428905235,
+      "learning_rate": 1.8867924528301888e-05,
+      "loss": 0.4769,
+      "step": 100
+    },
+    {
+      "epoch": 0.1920152091254753,
+      "grad_norm": 0.8320653522758553,
+      "learning_rate": 1.905660377358491e-05,
+      "loss": 0.5175,
+      "step": 101
+    },
+    {
+      "epoch": 0.19391634980988592,
+      "grad_norm": 0.8460318556093025,
+      "learning_rate": 1.9245283018867927e-05,
+      "loss": 0.4576,
+      "step": 102
+    },
+    {
+      "epoch": 0.1958174904942966,
+      "grad_norm": 0.8097510189784369,
+      "learning_rate": 1.9433962264150945e-05,
+      "loss": 0.4802,
+      "step": 103
+    },
+    {
+      "epoch": 0.19771863117870722,
+      "grad_norm": 0.845078635056242,
+      "learning_rate": 1.9622641509433963e-05,
+      "loss": 0.4545,
+      "step": 104
+    },
+    {
+      "epoch": 0.19961977186311788,
+      "grad_norm": 0.7968963097247687,
+      "learning_rate": 1.9811320754716984e-05,
+      "loss": 0.4777,
+      "step": 105
+    },
+    {
+      "epoch": 0.20152091254752852,
+      "grad_norm": 0.8839286855229416,
+      "learning_rate": 2e-05,
+      "loss": 0.4722,
+      "step": 106
+    },
+    {
+      "epoch": 0.20342205323193915,
+      "grad_norm": 0.9174076678992683,
+      "learning_rate": 1.9999944857420527e-05,
+      "loss": 0.464,
+      "step": 107
+    },
+    {
+      "epoch": 0.20532319391634982,
+      "grad_norm": 0.9182852232007711,
+      "learning_rate": 1.9999779430290247e-05,
+      "loss": 0.4673,
+      "step": 108
+    },
+    {
+      "epoch": 0.20722433460076045,
+      "grad_norm": 0.8268746379749562,
+      "learning_rate": 1.9999503720433575e-05,
+      "loss": 0.5012,
+      "step": 109
+    },
+    {
+      "epoch": 0.20912547528517111,
+      "grad_norm": 0.9626844373812254,
+      "learning_rate": 1.999911773089118e-05,
+      "loss": 0.484,
+      "step": 110
+    },
+    {
+      "epoch": 0.21102661596958175,
+      "grad_norm": 0.8377332679229166,
+      "learning_rate": 1.999862146591996e-05,
+      "loss": 0.4464,
+      "step": 111
+    },
+    {
+      "epoch": 0.21292775665399238,
+      "grad_norm": 0.8501848725676991,
+      "learning_rate": 1.9998014930992976e-05,
+      "loss": 0.4512,
+      "step": 112
+    },
+    {
+      "epoch": 0.21482889733840305,
+      "grad_norm": 0.9390676008390613,
+      "learning_rate": 1.9997298132799408e-05,
+      "loss": 0.4916,
+      "step": 113
+    },
+    {
+      "epoch": 0.21673003802281368,
+      "grad_norm": 0.8414912820682342,
+      "learning_rate": 1.9996471079244477e-05,
+      "loss": 0.4872,
+      "step": 114
+    },
+    {
+      "epoch": 0.21863117870722434,
+      "grad_norm": 0.8946529309257297,
+      "learning_rate": 1.999553377944936e-05,
+      "loss": 0.4732,
+      "step": 115
+    },
+    {
+      "epoch": 0.22053231939163498,
+      "grad_norm": 0.8888484869996336,
+      "learning_rate": 1.9994486243751076e-05,
+      "loss": 0.4687,
+      "step": 116
+    },
+    {
+      "epoch": 0.2224334600760456,
+      "grad_norm": 0.8300369879800873,
+      "learning_rate": 1.9993328483702393e-05,
+      "loss": 0.4656,
+      "step": 117
+    },
+    {
+      "epoch": 0.22433460076045628,
+      "grad_norm": 0.9488681886095498,
+      "learning_rate": 1.999206051207169e-05,
+      "loss": 0.476,
+      "step": 118
+    },
+    {
+      "epoch": 0.2262357414448669,
+      "grad_norm": 0.821149717884847,
+      "learning_rate": 1.9990682342842805e-05,
+      "loss": 0.4843,
+      "step": 119
+    },
+    {
+      "epoch": 0.22813688212927757,
+      "grad_norm": 0.8116301203521895,
+      "learning_rate": 1.99891939912149e-05,
+      "loss": 0.4467,
+      "step": 120
+    },
+    {
+      "epoch": 0.2300380228136882,
+      "grad_norm": 0.8140848349038625,
+      "learning_rate": 1.9987595473602292e-05,
+      "loss": 0.4788,
+      "step": 121
+    },
+    {
+      "epoch": 0.23193916349809887,
+      "grad_norm": 1.0191471284237132,
+      "learning_rate": 1.9985886807634246e-05,
+      "loss": 0.5047,
+      "step": 122
+    },
+    {
+      "epoch": 0.2338403041825095,
+      "grad_norm": 0.7878887639872737,
+      "learning_rate": 1.9984068012154824e-05,
+      "loss": 0.4604,
+      "step": 123
+    },
+    {
+      "epoch": 0.23574144486692014,
+      "grad_norm": 0.9568456263799635,
+      "learning_rate": 1.9982139107222634e-05,
+      "loss": 0.4527,
+      "step": 124
+    },
+    {
+      "epoch": 0.2376425855513308,
+      "grad_norm": 0.9057385954561654,
+      "learning_rate": 1.9980100114110637e-05,
+      "loss": 0.4659,
+      "step": 125
+    },
+    {
+      "epoch": 0.23954372623574144,
+      "grad_norm": 0.8702652230034865,
+      "learning_rate": 1.99779510553059e-05,
+      "loss": 0.4502,
+      "step": 126
+    },
+    {
+      "epoch": 0.2414448669201521,
+      "grad_norm": 0.7982450971596367,
+      "learning_rate": 1.9975691954509347e-05,
+      "loss": 0.4775,
+      "step": 127
+    },
+    {
+      "epoch": 0.24334600760456274,
+      "grad_norm": 0.9339907254515178,
+      "learning_rate": 1.9973322836635517e-05,
+      "loss": 0.4873,
+      "step": 128
+    },
+    {
+      "epoch": 0.24524714828897337,
+      "grad_norm": 0.8755624036963652,
+      "learning_rate": 1.997084372781226e-05,
+      "loss": 0.4962,
+      "step": 129
+    },
+    {
+      "epoch": 0.24714828897338403,
+      "grad_norm": 0.8604658996662152,
+      "learning_rate": 1.9968254655380465e-05,
+      "loss": 0.483,
+      "step": 130
+    },
+    {
+      "epoch": 0.24904942965779467,
+      "grad_norm": 0.8838567627329856,
+      "learning_rate": 1.996555564789376e-05,
+      "loss": 0.4686,
+      "step": 131
+    },
+    {
+      "epoch": 0.2509505703422053,
+      "grad_norm": 0.9218028563406552,
+      "learning_rate": 1.996274673511819e-05,
+      "loss": 0.4399,
+      "step": 132
+    },
+    {
+      "epoch": 0.25285171102661597,
+      "grad_norm": 0.8322543404530368,
+      "learning_rate": 1.99598279480319e-05,
+      "loss": 0.447,
+      "step": 133
+    },
+    {
+      "epoch": 0.25475285171102663,
+      "grad_norm": 0.9068221460485459,
+      "learning_rate": 1.9956799318824776e-05,
+      "loss": 0.4633,
+      "step": 134
+    },
+    {
+      "epoch": 0.25665399239543724,
+      "grad_norm": 0.7890803297343486,
+      "learning_rate": 1.99536608808981e-05,
+      "loss": 0.4681,
+      "step": 135
+    },
+    {
+      "epoch": 0.2585551330798479,
+      "grad_norm": 0.785502190799272,
+      "learning_rate": 1.995041266886419e-05,
+      "loss": 0.4824,
+      "step": 136
+    },
+    {
+      "epoch": 0.26045627376425856,
+      "grad_norm": 0.8658384046258384,
+      "learning_rate": 1.9947054718545996e-05,
+      "loss": 0.4842,
+      "step": 137
+    },
+    {
+      "epoch": 0.2623574144486692,
+      "grad_norm": 0.8855912239510314,
+      "learning_rate": 1.994358706697674e-05,
+      "loss": 0.4504,
+      "step": 138
+    },
+    {
+      "epoch": 0.26425855513307983,
+      "grad_norm": 0.8142899019049212,
+      "learning_rate": 1.9940009752399462e-05,
+      "loss": 0.4769,
+      "step": 139
+    },
+    {
+      "epoch": 0.2661596958174905,
+      "grad_norm": 0.872523678791427,
+      "learning_rate": 1.9936322814266634e-05,
+      "loss": 0.4785,
+      "step": 140
+    },
+    {
+      "epoch": 0.26806083650190116,
+      "grad_norm": 0.7193570954156706,
+      "learning_rate": 1.9932526293239713e-05,
+      "loss": 0.4745,
+      "step": 141
+    },
+    {
+      "epoch": 0.26996197718631176,
+      "grad_norm": 0.8706462519900843,
+      "learning_rate": 1.9928620231188694e-05,
+      "loss": 0.4413,
+      "step": 142
+    },
+    {
+      "epoch": 0.2718631178707224,
+      "grad_norm": 0.8405071677677448,
+      "learning_rate": 1.992460467119164e-05,
+      "loss": 0.4719,
+      "step": 143
+    },
+    {
+      "epoch": 0.2737642585551331,
+      "grad_norm": 0.7568977431854605,
+      "learning_rate": 1.992047965753422e-05,
+      "loss": 0.4961,
+      "step": 144
+    },
+    {
+      "epoch": 0.27566539923954375,
+      "grad_norm": 0.7964113108994708,
+      "learning_rate": 1.991624523570922e-05,
+      "loss": 0.5009,
+      "step": 145
+    },
+    {
+      "epoch": 0.27756653992395436,
+      "grad_norm": 0.7615624563103405,
+      "learning_rate": 1.9911901452416012e-05,
+      "loss": 0.4551,
+      "step": 146
+    },
+    {
+      "epoch": 0.279467680608365,
+      "grad_norm": 0.7957202827835359,
+      "learning_rate": 1.9907448355560094e-05,
+      "loss": 0.4412,
+      "step": 147
+    },
+    {
+      "epoch": 0.2813688212927757,
+      "grad_norm": 0.8315397396423518,
+      "learning_rate": 1.9902885994252506e-05,
+      "loss": 0.4592,
+      "step": 148
+    },
+    {
+      "epoch": 0.2832699619771863,
+      "grad_norm": 0.7893609096844134,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.459,
+      "step": 149
+    },
+    {
+      "epoch": 0.28517110266159695,
+      "grad_norm": 0.8240275177677148,
+      "learning_rate": 1.9893433680751105e-05,
+      "loss": 0.4693,
+      "step": 150
+    },
+    {
+      "epoch": 0.2870722433460076,
+      "grad_norm": 0.7373204835516783,
+      "learning_rate": 1.9888543832802277e-05,
+      "loss": 0.4236,
+      "step": 151
+    },
+    {
+      "epoch": 0.2889733840304182,
+      "grad_norm": 0.7181726636423414,
+      "learning_rate": 1.9883544928890612e-05,
+      "loss": 0.4757,
+      "step": 152
+    },
+    {
+      "epoch": 0.2908745247148289,
+      "grad_norm": 0.8463786969451002,
+      "learning_rate": 1.9878437024146603e-05,
+      "loss": 0.5037,
+      "step": 153
+    },
+    {
+      "epoch": 0.29277566539923955,
+      "grad_norm": 0.800101771945551,
+      "learning_rate": 1.9873220174902857e-05,
+      "loss": 0.4566,
+      "step": 154
+    },
+    {
+      "epoch": 0.2946768060836502,
+      "grad_norm": 0.7060818872434408,
+      "learning_rate": 1.986789443869348e-05,
+      "loss": 0.4576,
+      "step": 155
+    },
+    {
+      "epoch": 0.2965779467680608,
+      "grad_norm": 0.7495873510493309,
+      "learning_rate": 1.9862459874253438e-05,
+      "loss": 0.4473,
+      "step": 156
+    },
+    {
+      "epoch": 0.2984790874524715,
+      "grad_norm": 0.8261211600590206,
+      "learning_rate": 1.985691654151791e-05,
+      "loss": 0.4587,
+      "step": 157
+    },
+    {
+      "epoch": 0.30038022813688214,
+      "grad_norm": 0.780658089784845,
+      "learning_rate": 1.9851264501621635e-05,
+      "loss": 0.4714,
+      "step": 158
+    },
+    {
+      "epoch": 0.30228136882129275,
+      "grad_norm": 0.8079889039798914,
+      "learning_rate": 1.984550381689822e-05,
+      "loss": 0.4574,
+      "step": 159
+    },
+    {
+      "epoch": 0.3041825095057034,
+      "grad_norm": 0.7099996945982795,
+      "learning_rate": 1.983963455087946e-05,
+      "loss": 0.4704,
+      "step": 160
+    },
+    {
+      "epoch": 0.3060836501901141,
+      "grad_norm": 0.7745334401909533,
+      "learning_rate": 1.983365676829466e-05,
+      "loss": 0.4726,
+      "step": 161
+    },
+    {
+      "epoch": 0.30798479087452474,
+      "grad_norm": 0.7314356714459969,
+      "learning_rate": 1.982757053506989e-05,
+      "loss": 0.4431,
+      "step": 162
+    },
+    {
+      "epoch": 0.30988593155893535,
+      "grad_norm": 0.7997080094531027,
+      "learning_rate": 1.9821375918327268e-05,
+      "loss": 0.4671,
+      "step": 163
+    },
+    {
+      "epoch": 0.311787072243346,
+      "grad_norm": 0.7010964415968214,
+      "learning_rate": 1.981507298638422e-05,
+      "loss": 0.4452,
+      "step": 164
+    },
+    {
+      "epoch": 0.31368821292775667,
+      "grad_norm": 0.7599313051588067,
+      "learning_rate": 1.9808661808752735e-05,
+      "loss": 0.4848,
+      "step": 165
+    },
+    {
+      "epoch": 0.3155893536121673,
+      "grad_norm": 0.7984654594215558,
+      "learning_rate": 1.980214245613858e-05,
+      "loss": 0.4558,
+      "step": 166
+    },
+    {
+      "epoch": 0.31749049429657794,
+      "grad_norm": 0.7277912525281827,
+      "learning_rate": 1.979551500044055e-05,
+      "loss": 0.4522,
+      "step": 167
+    },
+    {
+      "epoch": 0.3193916349809886,
+      "grad_norm": 0.6849277494650874,
+      "learning_rate": 1.9788779514749635e-05,
+      "loss": 0.4438,
+      "step": 168
+    },
+    {
+      "epoch": 0.32129277566539927,
+      "grad_norm": 0.7249174837495981,
+      "learning_rate": 1.978193607334826e-05,
+      "loss": 0.4578,
+      "step": 169
+    },
+    {
+      "epoch": 0.3231939163498099,
+      "grad_norm": 0.7019360128229704,
+      "learning_rate": 1.977498475170941e-05,
+      "loss": 0.4734,
+      "step": 170
+    },
+    {
+      "epoch": 0.32509505703422054,
+      "grad_norm": 0.7374004982983245,
+      "learning_rate": 1.9767925626495857e-05,
+      "loss": 0.454,
+      "step": 171
+    },
+    {
+      "epoch": 0.3269961977186312,
+      "grad_norm": 0.7181449888677062,
+      "learning_rate": 1.9760758775559275e-05,
+      "loss": 0.457,
+      "step": 172
+    },
+    {
+      "epoch": 0.3288973384030418,
+      "grad_norm": 0.8198957548721154,
+      "learning_rate": 1.975348427793939e-05,
+      "loss": 0.4667,
+      "step": 173
+    },
+    {
+      "epoch": 0.33079847908745247,
+      "grad_norm": 0.7892520627052213,
+      "learning_rate": 1.9746102213863113e-05,
+      "loss": 0.4438,
+      "step": 174
+    },
+    {
+      "epoch": 0.33269961977186313,
+      "grad_norm": 0.7594858705676638,
+      "learning_rate": 1.973861266474366e-05,
+      "loss": 0.4669,
+      "step": 175
+    },
+    {
+      "epoch": 0.33460076045627374,
+      "grad_norm": 0.787048537318064,
+      "learning_rate": 1.9731015713179643e-05,
+      "loss": 0.4701,
+      "step": 176
+    },
+    {
+      "epoch": 0.3365019011406844,
+      "grad_norm": 0.7789085269396826,
+      "learning_rate": 1.9723311442954163e-05,
+      "loss": 0.4776,
+      "step": 177
+    },
+    {
+      "epoch": 0.33840304182509506,
+      "grad_norm": 0.8600714411605456,
+      "learning_rate": 1.9715499939033883e-05,
+      "loss": 0.4592,
+      "step": 178
+    },
+    {
+      "epoch": 0.3403041825095057,
+      "grad_norm": 0.8118360689267554,
+      "learning_rate": 1.9707581287568094e-05,
+      "loss": 0.4283,
+      "step": 179
+    },
+    {
+      "epoch": 0.34220532319391633,
+      "grad_norm": 0.7596627762802998,
+      "learning_rate": 1.969955557588778e-05,
+      "loss": 0.4277,
+      "step": 180
+    },
+    {
+      "epoch": 0.344106463878327,
+      "grad_norm": 0.8456655840791617,
+      "learning_rate": 1.9691422892504626e-05,
+      "loss": 0.4544,
+      "step": 181
+    },
+    {
+      "epoch": 0.34600760456273766,
+      "grad_norm": 0.7910405929677755,
+      "learning_rate": 1.968318332711006e-05,
+      "loss": 0.4728,
+      "step": 182
+    },
+    {
+      "epoch": 0.34790874524714827,
+      "grad_norm": 0.7971060783746655,
+      "learning_rate": 1.9674836970574253e-05,
+      "loss": 0.4667,
+      "step": 183
+    },
+    {
+      "epoch": 0.34980988593155893,
+      "grad_norm": 0.8702703139124975,
+      "learning_rate": 1.966638391494514e-05,
+      "loss": 0.4549,
+      "step": 184
+    },
+    {
+      "epoch": 0.3517110266159696,
+      "grad_norm": 0.7449851203440129,
+      "learning_rate": 1.9657824253447378e-05,
+      "loss": 0.4398,
+      "step": 185
+    },
+    {
+      "epoch": 0.35361216730038025,
+      "grad_norm": 0.7725911172155271,
+      "learning_rate": 1.9649158080481327e-05,
+      "loss": 0.4639,
+      "step": 186
+    },
+    {
+      "epoch": 0.35551330798479086,
+      "grad_norm": 0.9398412015886246,
+      "learning_rate": 1.964038549162201e-05,
+      "loss": 0.4467,
+      "step": 187
+    },
+    {
+      "epoch": 0.3574144486692015,
+      "grad_norm": 0.7696939193576091,
+      "learning_rate": 1.963150658361807e-05,
+      "loss": 0.4546,
+      "step": 188
+    },
+    {
+      "epoch": 0.3593155893536122,
+      "grad_norm": 0.9387338941905596,
+      "learning_rate": 1.962252145439068e-05,
+      "loss": 0.4298,
+      "step": 189
+    },
+    {
+      "epoch": 0.3612167300380228,
+      "grad_norm": 0.8360787601391084,
+      "learning_rate": 1.9613430203032486e-05,
+      "loss": 0.4496,
+      "step": 190
+    },
+    {
+      "epoch": 0.36311787072243346,
+      "grad_norm": 0.861024512047741,
+      "learning_rate": 1.9604232929806493e-05,
+      "loss": 0.481,
+      "step": 191
+    },
+    {
+      "epoch": 0.3650190114068441,
+      "grad_norm": 0.9226517326038568,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.4439,
+      "step": 192
+    },
+    {
+      "epoch": 0.3669201520912547,
+      "grad_norm": 0.727443441007983,
+      "learning_rate": 1.9585520724648354e-05,
+      "loss": 0.4462,
+      "step": 193
+    },
+    {
+      "epoch": 0.3688212927756654,
+      "grad_norm": 0.8877165995142432,
+      "learning_rate": 1.957600599908406e-05,
+      "loss": 0.4484,
+      "step": 194
+    },
+    {
+      "epoch": 0.37072243346007605,
+      "grad_norm": 0.8359516770322809,
+      "learning_rate": 1.95663856643854e-05,
+      "loss": 0.4293,
+      "step": 195
+    },
+    {
+      "epoch": 0.3726235741444867,
+      "grad_norm": 0.8462915279780584,
+      "learning_rate": 1.955665982665038e-05,
+      "loss": 0.4505,
+      "step": 196
+    },
+    {
+      "epoch": 0.3745247148288973,
+      "grad_norm": 0.7363213541030138,
+      "learning_rate": 1.9546828593140565e-05,
+      "loss": 0.4577,
+      "step": 197
+    },
+    {
+      "epoch": 0.376425855513308,
+      "grad_norm": 0.7711815266523486,
+      "learning_rate": 1.9536892072279863e-05,
+      "loss": 0.475,
+      "step": 198
+    },
+    {
+      "epoch": 0.37832699619771865,
+      "grad_norm": 0.8138440374635927,
+      "learning_rate": 1.9526850373653356e-05,
+      "loss": 0.4596,
+      "step": 199
+    },
+    {
+      "epoch": 0.38022813688212925,
+      "grad_norm": 0.7868148318246388,
+      "learning_rate": 1.9516703608006074e-05,
+      "loss": 0.4255,
+      "step": 200
+    },
+    {
+      "epoch": 0.3821292775665399,
+      "grad_norm": 0.8071745709077621,
+      "learning_rate": 1.9506451887241787e-05,
+      "loss": 0.4638,
+      "step": 201
+    },
+    {
+      "epoch": 0.3840304182509506,
+      "grad_norm": 0.733323214581715,
+      "learning_rate": 1.949609532442176e-05,
+      "loss": 0.4684,
+      "step": 202
+    },
+    {
+      "epoch": 0.38593155893536124,
+      "grad_norm": 0.7786528423535014,
+      "learning_rate": 1.9485634033763507e-05,
+      "loss": 0.4325,
+      "step": 203
+    },
+    {
+      "epoch": 0.38783269961977185,
+      "grad_norm": 0.7487048557688153,
+      "learning_rate": 1.9475068130639543e-05,
+      "loss": 0.4437,
+      "step": 204
+    },
+    {
+      "epoch": 0.3897338403041825,
+      "grad_norm": 0.7474781253372679,
+      "learning_rate": 1.9464397731576093e-05,
+      "loss": 0.4397,
+      "step": 205
+    },
+    {
+      "epoch": 0.3916349809885932,
+      "grad_norm": 0.7572187050446576,
+      "learning_rate": 1.945362295425183e-05,
+      "loss": 0.4615,
+      "step": 206
+    },
+    {
+      "epoch": 0.3935361216730038,
+      "grad_norm": 0.6701870136654475,
+      "learning_rate": 1.944274391749655e-05,
+      "loss": 0.4297,
+      "step": 207
+    },
+    {
+      "epoch": 0.39543726235741444,
+      "grad_norm": 0.7631553304389546,
+      "learning_rate": 1.9431760741289886e-05,
+      "loss": 0.4415,
+      "step": 208
+    },
+    {
+      "epoch": 0.3973384030418251,
+      "grad_norm": 0.7220133503724472,
+      "learning_rate": 1.942067354675997e-05,
+      "loss": 0.4642,
+      "step": 209
+    },
+    {
+      "epoch": 0.39923954372623577,
+      "grad_norm": 0.7703817294531721,
+      "learning_rate": 1.9409482456182105e-05,
+      "loss": 0.4452,
+      "step": 210
+    },
+    {
+      "epoch": 0.4011406844106464,
+      "grad_norm": 0.684633996371507,
+      "learning_rate": 1.939818759297741e-05,
+      "loss": 0.4444,
+      "step": 211
+    },
+    {
+      "epoch": 0.40304182509505704,
+      "grad_norm": 0.7398254352812675,
+      "learning_rate": 1.9386789081711465e-05,
+      "loss": 0.4472,
+      "step": 212
+    },
+    {
+      "epoch": 0.4049429657794677,
+      "grad_norm": 0.7133712268624046,
+      "learning_rate": 1.9375287048092927e-05,
+      "loss": 0.4476,
+      "step": 213
+    },
+    {
+      "epoch": 0.4068441064638783,
+      "grad_norm": 0.7809381978844976,
+      "learning_rate": 1.9363681618972166e-05,
+      "loss": 0.4555,
+      "step": 214
+    },
+    {
+      "epoch": 0.40874524714828897,
+      "grad_norm": 0.7146620205108506,
+      "learning_rate": 1.9351972922339835e-05,
+      "loss": 0.4732,
+      "step": 215
+    },
+    {
+      "epoch": 0.41064638783269963,
+      "grad_norm": 0.8434266092013188,
+      "learning_rate": 1.9340161087325483e-05,
+      "loss": 0.4562,
+      "step": 216
+    },
+    {
+      "epoch": 0.41254752851711024,
+      "grad_norm": 0.7180241901297593,
+      "learning_rate": 1.9328246244196117e-05,
+      "loss": 0.4414,
+      "step": 217
+    },
+    {
+      "epoch": 0.4144486692015209,
+      "grad_norm": 0.7355924088180167,
+      "learning_rate": 1.931622852435478e-05,
+      "loss": 0.4447,
+      "step": 218
+    },
+    {
+      "epoch": 0.41634980988593157,
+      "grad_norm": 0.7213007439802591,
+      "learning_rate": 1.930410806033908e-05,
+      "loss": 0.4206,
+      "step": 219
+    },
+    {
+      "epoch": 0.41825095057034223,
+      "grad_norm": 0.7523263560987146,
+      "learning_rate": 1.929188498581975e-05,
+      "loss": 0.4382,
+      "step": 220
+    },
+    {
+      "epoch": 0.42015209125475284,
+      "grad_norm": 0.7324257943530699,
+      "learning_rate": 1.9279559435599164e-05,
+      "loss": 0.4405,
+      "step": 221
+    },
+    {
+      "epoch": 0.4220532319391635,
+      "grad_norm": 0.8226508520626221,
+      "learning_rate": 1.926713154560984e-05,
+      "loss": 0.4603,
+      "step": 222
+    },
+    {
+      "epoch": 0.42395437262357416,
+      "grad_norm": 0.7133376496281909,
+      "learning_rate": 1.9254601452912972e-05,
+      "loss": 0.4709,
+      "step": 223
+    },
+    {
+      "epoch": 0.42585551330798477,
+      "grad_norm": 0.7896916068017726,
+      "learning_rate": 1.924196929569688e-05,
+      "loss": 0.4248,
+      "step": 224
+    },
+    {
+      "epoch": 0.42775665399239543,
+      "grad_norm": 0.7597621689902878,
+      "learning_rate": 1.922923521327551e-05,
+      "loss": 0.503,
+      "step": 225
+    },
+    {
+      "epoch": 0.4296577946768061,
+      "grad_norm": 0.7181294810165241,
+      "learning_rate": 1.9216399346086893e-05,
+      "loss": 0.4492,
+      "step": 226
+    },
+    {
+      "epoch": 0.43155893536121676,
+      "grad_norm": 0.7676930134829011,
+      "learning_rate": 1.9203461835691596e-05,
+      "loss": 0.4638,
+      "step": 227
+    },
+    {
+      "epoch": 0.43346007604562736,
+      "grad_norm": 0.693470477294498,
+      "learning_rate": 1.9190422824771158e-05,
+      "loss": 0.4348,
+      "step": 228
+    },
+    {
+      "epoch": 0.435361216730038,
+      "grad_norm": 0.7475855625509044,
+      "learning_rate": 1.9177282457126515e-05,
+      "loss": 0.4379,
+      "step": 229
+    },
+    {
+      "epoch": 0.4372623574144487,
+      "grad_norm": 0.7500300730978319,
+      "learning_rate": 1.9164040877676425e-05,
+      "loss": 0.4577,
+      "step": 230
+    },
+    {
+      "epoch": 0.4391634980988593,
+      "grad_norm": 0.7512016120542404,
+      "learning_rate": 1.9150698232455853e-05,
+      "loss": 0.4166,
+      "step": 231
+    },
+    {
+      "epoch": 0.44106463878326996,
+      "grad_norm": 0.7395554426424591,
+      "learning_rate": 1.913725466861438e-05,
+      "loss": 0.4544,
+      "step": 232
+    },
+    {
+      "epoch": 0.4429657794676806,
+      "grad_norm": 0.856479598330211,
+      "learning_rate": 1.9123710334414552e-05,
+      "loss": 0.4477,
+      "step": 233
+    },
+    {
+      "epoch": 0.4448669201520912,
+      "grad_norm": 0.8026421294944246,
+      "learning_rate": 1.911006537923029e-05,
+      "loss": 0.4304,
+      "step": 234
+    },
+    {
+      "epoch": 0.4467680608365019,
+      "grad_norm": 0.7083360114353457,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.4228,
+      "step": 235
+    },
+    {
+      "epoch": 0.44866920152091255,
+      "grad_norm": 0.753083611197222,
+      "learning_rate": 1.908247420895089e-05,
+      "loss": 0.4367,
+      "step": 236
+    },
+    {
+      "epoch": 0.4505703422053232,
+      "grad_norm": 0.8124857729958662,
+      "learning_rate": 1.9068528298145418e-05,
+      "loss": 0.4626,
+      "step": 237
+    },
+    {
+      "epoch": 0.4524714828897338,
+      "grad_norm": 0.8343797942728003,
+      "learning_rate": 1.905448237493147e-05,
+      "loss": 0.4446,
+      "step": 238
+    },
+    {
+      "epoch": 0.4543726235741445,
+      "grad_norm": 0.7726589591821661,
+      "learning_rate": 1.9040336594214727e-05,
+      "loss": 0.4438,
+      "step": 239
+    },
+    {
+      "epoch": 0.45627376425855515,
+      "grad_norm": 0.7888635359670993,
+      "learning_rate": 1.9026091112002163e-05,
+      "loss": 0.4347,
+      "step": 240
+    },
+    {
+      "epoch": 0.45817490494296575,
+      "grad_norm": 0.7366950129327747,
+      "learning_rate": 1.90117460854003e-05,
+      "loss": 0.4337,
+      "step": 241
+    },
+    {
+      "epoch": 0.4600760456273764,
+      "grad_norm": 0.739710648607917,
+      "learning_rate": 1.8997301672613496e-05,
+      "loss": 0.4741,
+      "step": 242
+    },
+    {
+      "epoch": 0.4619771863117871,
+      "grad_norm": 0.7466782628008918,
+      "learning_rate": 1.8982758032942184e-05,
+      "loss": 0.4647,
+      "step": 243
+    },
+    {
+      "epoch": 0.46387832699619774,
+      "grad_norm": 0.7498984313900852,
+      "learning_rate": 1.896811532678113e-05,
+      "loss": 0.4394,
+      "step": 244
+    },
+    {
+      "epoch": 0.46577946768060835,
+      "grad_norm": 0.7680399841586866,
+      "learning_rate": 1.8953373715617646e-05,
+      "loss": 0.453,
+      "step": 245
+    },
+    {
+      "epoch": 0.467680608365019,
+      "grad_norm": 0.6821247213706364,
+      "learning_rate": 1.893853336202983e-05,
+      "loss": 0.4269,
+      "step": 246
+    },
+    {
+      "epoch": 0.4695817490494297,
+      "grad_norm": 0.803838173431942,
+      "learning_rate": 1.892359442968475e-05,
+      "loss": 0.444,
+      "step": 247
+    },
+    {
+      "epoch": 0.4714828897338403,
+      "grad_norm": 0.7321264678676883,
+      "learning_rate": 1.8908557083336668e-05,
+      "loss": 0.4438,
+      "step": 248
+    },
+    {
+      "epoch": 0.47338403041825095,
+      "grad_norm": 0.8011887217840076,
+      "learning_rate": 1.889342148882519e-05,
+      "loss": 0.4433,
+      "step": 249
+    },
+    {
+      "epoch": 0.4752851711026616,
+      "grad_norm": 0.8140018018401939,
+      "learning_rate": 1.8878187813073465e-05,
+      "loss": 0.4612,
+      "step": 250
+    },
+    {
+      "epoch": 0.47718631178707227,
+      "grad_norm": 0.7217120363665857,
+      "learning_rate": 1.886285622408633e-05,
+      "loss": 0.4212,
+      "step": 251
+    },
+    {
+      "epoch": 0.4790874524714829,
+      "grad_norm": 0.8379492337778576,
+      "learning_rate": 1.8847426890948447e-05,
+      "loss": 0.4458,
+      "step": 252
+    },
+    {
+      "epoch": 0.48098859315589354,
+      "grad_norm": 0.7522963277942867,
+      "learning_rate": 1.8831899983822475e-05,
+      "loss": 0.4677,
+      "step": 253
+    },
+    {
+      "epoch": 0.4828897338403042,
+      "grad_norm": 0.745937249894783,
+      "learning_rate": 1.8816275673947148e-05,
+      "loss": 0.4364,
+      "step": 254
+    },
+    {
+      "epoch": 0.4847908745247148,
+      "grad_norm": 0.7454520991441149,
+      "learning_rate": 1.8800554133635417e-05,
+      "loss": 0.437,
+      "step": 255
+    },
+    {
+      "epoch": 0.4866920152091255,
+      "grad_norm": 0.6918530756120918,
+      "learning_rate": 1.8784735536272543e-05,
+      "loss": 0.4285,
+      "step": 256
+    },
+    {
+      "epoch": 0.48859315589353614,
+      "grad_norm": 0.8648890375130991,
+      "learning_rate": 1.8768820056314173e-05,
+      "loss": 0.4302,
+      "step": 257
+    },
+    {
+      "epoch": 0.49049429657794674,
+      "grad_norm": 1.3723744377777394,
+      "learning_rate": 1.875280786928444e-05,
+      "loss": 0.4556,
+      "step": 258
+    },
+    {
+      "epoch": 0.4923954372623574,
+      "grad_norm": 0.7597614247709502,
+      "learning_rate": 1.873669915177399e-05,
+      "loss": 0.4604,
+      "step": 259
+    },
+    {
+      "epoch": 0.49429657794676807,
+      "grad_norm": 0.7947445856200992,
+      "learning_rate": 1.872049408143808e-05,
+      "loss": 0.4209,
+      "step": 260
+    },
+    {
+      "epoch": 0.49619771863117873,
+      "grad_norm": 0.6361303424420126,
+      "learning_rate": 1.8704192836994578e-05,
+      "loss": 0.4175,
+      "step": 261
+    },
+    {
+      "epoch": 0.49809885931558934,
+      "grad_norm": 0.7272127082042135,
+      "learning_rate": 1.8687795598222024e-05,
+      "loss": 0.4318,
+      "step": 262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7805323590263087,
+      "learning_rate": 1.8671302545957628e-05,
+      "loss": 0.4491,
+      "step": 263
+    },
+    {
+      "epoch": 0.5019011406844106,
+      "grad_norm": 0.7097642566629339,
+      "learning_rate": 1.8654713862095272e-05,
+      "loss": 0.4338,
+      "step": 264
+    },
+    {
+      "epoch": 0.5038022813688213,
+      "grad_norm": 0.7115360578554114,
+      "learning_rate": 1.8638029729583524e-05,
+      "loss": 0.4426,
+      "step": 265
+    },
+    {
+      "epoch": 0.5057034220532319,
+      "grad_norm": 0.692053846517522,
+      "learning_rate": 1.8621250332423603e-05,
+      "loss": 0.4374,
+      "step": 266
+    },
+    {
+      "epoch": 0.5076045627376425,
+      "grad_norm": 0.7619493905095778,
+      "learning_rate": 1.860437585566736e-05,
+      "loss": 0.428,
+      "step": 267
+    },
+    {
+      "epoch": 0.5095057034220533,
+      "grad_norm": 0.6843641202448926,
+      "learning_rate": 1.8587406485415226e-05,
+      "loss": 0.4485,
+      "step": 268
+    },
+    {
+      "epoch": 0.5114068441064639,
+      "grad_norm": 0.8030580736022731,
+      "learning_rate": 1.8570342408814173e-05,
+      "loss": 0.4331,
+      "step": 269
+    },
+    {
+      "epoch": 0.5133079847908745,
+      "grad_norm": 0.7704258526878163,
+      "learning_rate": 1.855318381405564e-05,
+      "loss": 0.4511,
+      "step": 270
+    },
+    {
+      "epoch": 0.5152091254752852,
+      "grad_norm": 0.8384791299677623,
+      "learning_rate": 1.8535930890373467e-05,
+      "loss": 0.4445,
+      "step": 271
+    },
+    {
+      "epoch": 0.5171102661596958,
+      "grad_norm": 0.7991906372160948,
+      "learning_rate": 1.8518583828041787e-05,
+      "loss": 0.4436,
+      "step": 272
+    },
+    {
+      "epoch": 0.5190114068441065,
+      "grad_norm": 0.7763199252672867,
+      "learning_rate": 1.8501142818372964e-05,
+      "loss": 0.4436,
+      "step": 273
+    },
+    {
+      "epoch": 0.5209125475285171,
+      "grad_norm": 0.8227042699053724,
+      "learning_rate": 1.848360805371544e-05,
+      "loss": 0.4584,
+      "step": 274
+    },
+    {
+      "epoch": 0.5228136882129277,
+      "grad_norm": 0.674275377032333,
+      "learning_rate": 1.8465979727451653e-05,
+      "loss": 0.4227,
+      "step": 275
+    },
+    {
+      "epoch": 0.5247148288973384,
+      "grad_norm": 0.8392772491073504,
+      "learning_rate": 1.8448258033995877e-05,
+      "loss": 0.4332,
+      "step": 276
+    },
+    {
+      "epoch": 0.526615969581749,
+      "grad_norm": 0.7101347164788872,
+      "learning_rate": 1.8430443168792087e-05,
+      "loss": 0.4334,
+      "step": 277
+    },
+    {
+      "epoch": 0.5285171102661597,
+      "grad_norm": 0.7226360482181258,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.4472,
+      "step": 278
+    },
+    {
+      "epoch": 0.5304182509505704,
+      "grad_norm": 0.7709804234238199,
+      "learning_rate": 1.8394534710051956e-05,
+      "loss": 0.45,
+      "step": 279
+    },
+    {
+      "epoch": 0.532319391634981,
+      "grad_norm": 0.7232938907917893,
+      "learning_rate": 1.8376441512532617e-05,
+      "loss": 0.4641,
+      "step": 280
+    },
+    {
+      "epoch": 0.5342205323193916,
+      "grad_norm": 0.7458716197522715,
+      "learning_rate": 1.835825593529492e-05,
+      "loss": 0.4196,
+      "step": 281
+    },
+    {
+      "epoch": 0.5361216730038023,
+      "grad_norm": 0.7074343056521704,
+      "learning_rate": 1.833997817889878e-05,
+      "loss": 0.4268,
+      "step": 282
+    },
+    {
+      "epoch": 0.5380228136882129,
+      "grad_norm": 0.7629966945686958,
+      "learning_rate": 1.8321608444920738e-05,
+      "loss": 0.4186,
+      "step": 283
+    },
+    {
+      "epoch": 0.5399239543726235,
+      "grad_norm": 0.7658329258385395,
+      "learning_rate": 1.830314693595169e-05,
+      "loss": 0.4356,
+      "step": 284
+    },
+    {
+      "epoch": 0.5418250950570342,
+      "grad_norm": 0.6880289771009196,
+      "learning_rate": 1.828459385559468e-05,
+      "loss": 0.4187,
+      "step": 285
+    },
+    {
+      "epoch": 0.5437262357414449,
+      "grad_norm": 0.6308636275687333,
+      "learning_rate": 1.8265949408462657e-05,
+      "loss": 0.4219,
+      "step": 286
+    },
+    {
+      "epoch": 0.5456273764258555,
+      "grad_norm": 0.7070642866052863,
+      "learning_rate": 1.8247213800176192e-05,
+      "loss": 0.4455,
+      "step": 287
+    },
+    {
+      "epoch": 0.5475285171102662,
+      "grad_norm": 0.6731517796903704,
+      "learning_rate": 1.8228387237361245e-05,
+      "loss": 0.4244,
+      "step": 288
+    },
+    {
+      "epoch": 0.5494296577946768,
+      "grad_norm": 0.6859070598433792,
+      "learning_rate": 1.8209469927646863e-05,
+      "loss": 0.4281,
+      "step": 289
+    },
+    {
+      "epoch": 0.5513307984790875,
+      "grad_norm": 0.6838169207127082,
+      "learning_rate": 1.8190462079662897e-05,
+      "loss": 0.4532,
+      "step": 290
+    },
+    {
+      "epoch": 0.5532319391634981,
+      "grad_norm": 0.7024952636065559,
+      "learning_rate": 1.81713639030377e-05,
+      "loss": 0.4423,
+      "step": 291
+    },
+    {
+      "epoch": 0.5551330798479087,
+      "grad_norm": 0.7113808340823516,
+      "learning_rate": 1.8152175608395814e-05,
+      "loss": 0.43,
+      "step": 292
+    },
+    {
+      "epoch": 0.5570342205323194,
+      "grad_norm": 0.7386002729857145,
+      "learning_rate": 1.8132897407355657e-05,
+      "loss": 0.4287,
+      "step": 293
+    },
+    {
+      "epoch": 0.55893536121673,
+      "grad_norm": 0.6563212206236789,
+      "learning_rate": 1.811352951252717e-05,
+      "loss": 0.4097,
+      "step": 294
+    },
+    {
+      "epoch": 0.5608365019011406,
+      "grad_norm": 0.6952946147762175,
+      "learning_rate": 1.809407213750949e-05,
+      "loss": 0.4255,
+      "step": 295
+    },
+    {
+      "epoch": 0.5627376425855514,
+      "grad_norm": 0.6627403249107185,
+      "learning_rate": 1.807452549688859e-05,
+      "loss": 0.45,
+      "step": 296
+    },
+    {
+      "epoch": 0.564638783269962,
+      "grad_norm": 0.6926188095638577,
+      "learning_rate": 1.8054889806234906e-05,
+      "loss": 0.4653,
+      "step": 297
+    },
+    {
+      "epoch": 0.5665399239543726,
+      "grad_norm": 0.7030066647161937,
+      "learning_rate": 1.8035165282100963e-05,
+      "loss": 0.454,
+      "step": 298
+    },
+    {
+      "epoch": 0.5684410646387833,
+      "grad_norm": 0.7052972587921341,
+      "learning_rate": 1.8015352142018984e-05,
+      "loss": 0.426,
+      "step": 299
+    },
+    {
+      "epoch": 0.5703422053231939,
+      "grad_norm": 0.6621603013168155,
+      "learning_rate": 1.799545060449851e-05,
+      "loss": 0.4226,
+      "step": 300
+    },
+    {
+      "epoch": 0.5722433460076045,
+      "grad_norm": 0.6714577247062239,
+      "learning_rate": 1.797546088902396e-05,
+      "loss": 0.4346,
+      "step": 301
+    },
+    {
+      "epoch": 0.5741444866920152,
+      "grad_norm": 0.6890671511738551,
+      "learning_rate": 1.7955383216052224e-05,
+      "loss": 0.442,
+      "step": 302
+    },
+    {
+      "epoch": 0.5760456273764258,
+      "grad_norm": 0.6748590389682685,
+      "learning_rate": 1.7935217807010238e-05,
+      "loss": 0.4237,
+      "step": 303
+    },
+    {
+      "epoch": 0.5779467680608364,
+      "grad_norm": 0.7175518446822167,
+      "learning_rate": 1.7914964884292543e-05,
+      "loss": 0.4265,
+      "step": 304
+    },
+    {
+      "epoch": 0.5798479087452472,
+      "grad_norm": 0.634268965146618,
+      "learning_rate": 1.7894624671258813e-05,
+      "loss": 0.4367,
+      "step": 305
+    },
+    {
+      "epoch": 0.5817490494296578,
+      "grad_norm": 0.6733745275246174,
+      "learning_rate": 1.7874197392231414e-05,
+      "loss": 0.4213,
+      "step": 306
+    },
+    {
+      "epoch": 0.5836501901140685,
+      "grad_norm": 0.6817369025650951,
+      "learning_rate": 1.7853683272492913e-05,
+      "loss": 0.4525,
+      "step": 307
+    },
+    {
+      "epoch": 0.5855513307984791,
+      "grad_norm": 0.611540828238282,
+      "learning_rate": 1.7833082538283615e-05,
+      "loss": 0.4252,
+      "step": 308
+    },
+    {
+      "epoch": 0.5874524714828897,
+      "grad_norm": 0.6545901239405563,
+      "learning_rate": 1.7812395416799034e-05,
+      "loss": 0.4246,
+      "step": 309
+    },
+    {
+      "epoch": 0.5893536121673004,
+      "grad_norm": 0.7201340046913922,
+      "learning_rate": 1.7791622136187422e-05,
+      "loss": 0.4653,
+      "step": 310
+    },
+    {
+      "epoch": 0.591254752851711,
+      "grad_norm": 0.7525512724063577,
+      "learning_rate": 1.7770762925547235e-05,
+      "loss": 0.4649,
+      "step": 311
+    },
+    {
+      "epoch": 0.5931558935361216,
+      "grad_norm": 0.6979731234407439,
+      "learning_rate": 1.7749818014924612e-05,
+      "loss": 0.4329,
+      "step": 312
+    },
+    {
+      "epoch": 0.5950570342205324,
+      "grad_norm": 0.6473578801126666,
+      "learning_rate": 1.7728787635310828e-05,
+      "loss": 0.4231,
+      "step": 313
+    },
+    {
+      "epoch": 0.596958174904943,
+      "grad_norm": 0.736305770883599,
+      "learning_rate": 1.770767201863976e-05,
+      "loss": 0.4473,
+      "step": 314
+    },
+    {
+      "epoch": 0.5988593155893536,
+      "grad_norm": 0.716977872915759,
+      "learning_rate": 1.7686471397785322e-05,
+      "loss": 0.4284,
+      "step": 315
+    },
+    {
+      "epoch": 0.6007604562737643,
+      "grad_norm": 0.7243251374715308,
+      "learning_rate": 1.76651860065589e-05,
+      "loss": 0.4072,
+      "step": 316
+    },
+    {
+      "epoch": 0.6026615969581749,
+      "grad_norm": 0.8067283007141923,
+      "learning_rate": 1.764381607970677e-05,
+      "loss": 0.4323,
+      "step": 317
+    },
+    {
+      "epoch": 0.6045627376425855,
+      "grad_norm": 0.6460023954212142,
+      "learning_rate": 1.7622361852907506e-05,
+      "loss": 0.4196,
+      "step": 318
+    },
+    {
+      "epoch": 0.6064638783269962,
+      "grad_norm": 0.8739609399195127,
+      "learning_rate": 1.760082356276939e-05,
+      "loss": 0.429,
+      "step": 319
+    },
+    {
+      "epoch": 0.6083650190114068,
+      "grad_norm": 0.6716689865379148,
+      "learning_rate": 1.75792014468278e-05,
+      "loss": 0.4296,
+      "step": 320
+    },
+    {
+      "epoch": 0.6102661596958175,
+      "grad_norm": 0.7426959557128658,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.4555,
+      "step": 321
+    },
+    {
+      "epoch": 0.6121673003802282,
+      "grad_norm": 0.8274596525387122,
+      "learning_rate": 1.7535706692295436e-05,
+      "loss": 0.4495,
+      "step": 322
+    },
+    {
+      "epoch": 0.6140684410646388,
+      "grad_norm": 0.6400206675359528,
+      "learning_rate": 1.7513834533387256e-05,
+      "loss": 0.4222,
+      "step": 323
+    },
+    {
+      "epoch": 0.6159695817490495,
+      "grad_norm": 0.7559105536973265,
+      "learning_rate": 1.7491879508035488e-05,
+      "loss": 0.4297,
+      "step": 324
+    },
+    {
+      "epoch": 0.6178707224334601,
+      "grad_norm": 0.680328440479777,
+      "learning_rate": 1.746984185837149e-05,
+      "loss": 0.4183,
+      "step": 325
+    },
+    {
+      "epoch": 0.6197718631178707,
+      "grad_norm": 0.6986796593379792,
+      "learning_rate": 1.744772182743782e-05,
+      "loss": 0.4229,
+      "step": 326
+    },
+    {
+      "epoch": 0.6216730038022814,
+      "grad_norm": 0.6788258037958755,
+      "learning_rate": 1.7425519659185596e-05,
+      "loss": 0.4309,
+      "step": 327
+    },
+    {
+      "epoch": 0.623574144486692,
+      "grad_norm": 0.6597090620019976,
+      "learning_rate": 1.740323559847179e-05,
+      "loss": 0.4226,
+      "step": 328
+    },
+    {
+      "epoch": 0.6254752851711026,
+      "grad_norm": 0.6631302841702288,
+      "learning_rate": 1.738086989105651e-05,
+      "loss": 0.401,
+      "step": 329
+    },
+    {
+      "epoch": 0.6273764258555133,
+      "grad_norm": 0.7110117223466982,
+      "learning_rate": 1.735842278360032e-05,
+      "loss": 0.4678,
+      "step": 330
+    },
+    {
+      "epoch": 0.629277566539924,
+      "grad_norm": 0.7292508851270849,
+      "learning_rate": 1.73358945236615e-05,
+      "loss": 0.4262,
+      "step": 331
+    },
+    {
+      "epoch": 0.6311787072243346,
+      "grad_norm": 0.6403054374850369,
+      "learning_rate": 1.7313285359693322e-05,
+      "loss": 0.4478,
+      "step": 332
+    },
+    {
+      "epoch": 0.6330798479087453,
+      "grad_norm": 0.6210175646945111,
+      "learning_rate": 1.7290595541041312e-05,
+      "loss": 0.414,
+      "step": 333
+    },
+    {
+      "epoch": 0.6349809885931559,
+      "grad_norm": 0.6458275050940676,
+      "learning_rate": 1.7267825317940494e-05,
+      "loss": 0.4422,
+      "step": 334
+    },
+    {
+      "epoch": 0.6368821292775665,
+      "grad_norm": 0.6869869574640594,
+      "learning_rate": 1.724497494151264e-05,
+      "loss": 0.4055,
+      "step": 335
+    },
+    {
+      "epoch": 0.6387832699619772,
+      "grad_norm": 0.7118806502819282,
+      "learning_rate": 1.7222044663763484e-05,
+      "loss": 0.43,
+      "step": 336
+    },
+    {
+      "epoch": 0.6406844106463878,
+      "grad_norm": 0.640150328798902,
+      "learning_rate": 1.7199034737579962e-05,
+      "loss": 0.4369,
+      "step": 337
+    },
+    {
+      "epoch": 0.6425855513307985,
+      "grad_norm": 0.7089732928015482,
+      "learning_rate": 1.7175945416727405e-05,
+      "loss": 0.4397,
+      "step": 338
+    },
+    {
+      "epoch": 0.6444866920152091,
+      "grad_norm": 0.7103991952820954,
+      "learning_rate": 1.7152776955846768e-05,
+      "loss": 0.4193,
+      "step": 339
+    },
+    {
+      "epoch": 0.6463878326996197,
+      "grad_norm": 0.6202808889289216,
+      "learning_rate": 1.7129529610451775e-05,
+      "loss": 0.3981,
+      "step": 340
+    },
+    {
+      "epoch": 0.6482889733840305,
+      "grad_norm": 0.6617723006748341,
+      "learning_rate": 1.7106203636926154e-05,
+      "loss": 0.4366,
+      "step": 341
+    },
+    {
+      "epoch": 0.6501901140684411,
+      "grad_norm": 0.6818093793143916,
+      "learning_rate": 1.7082799292520767e-05,
+      "loss": 0.4249,
+      "step": 342
+    },
+    {
+      "epoch": 0.6520912547528517,
+      "grad_norm": 0.7165538155146259,
+      "learning_rate": 1.7059316835350806e-05,
+      "loss": 0.4371,
+      "step": 343
+    },
+    {
+      "epoch": 0.6539923954372624,
+      "grad_norm": 0.6727569609593109,
+      "learning_rate": 1.7035756524392924e-05,
+      "loss": 0.4223,
+      "step": 344
+    },
+    {
+      "epoch": 0.655893536121673,
+      "grad_norm": 0.7126060741007787,
+      "learning_rate": 1.7012118619482376e-05,
+      "loss": 0.4365,
+      "step": 345
+    },
+    {
+      "epoch": 0.6577946768060836,
+      "grad_norm": 0.698562128720747,
+      "learning_rate": 1.6988403381310177e-05,
+      "loss": 0.451,
+      "step": 346
+    },
+    {
+      "epoch": 0.6596958174904943,
+      "grad_norm": 0.6200536410263091,
+      "learning_rate": 1.696461107142021e-05,
+      "loss": 0.4425,
+      "step": 347
+    },
+    {
+      "epoch": 0.6615969581749049,
+      "grad_norm": 0.6756990713122462,
+      "learning_rate": 1.6940741952206342e-05,
+      "loss": 0.423,
+      "step": 348
+    },
+    {
+      "epoch": 0.6634980988593155,
+      "grad_norm": 0.7399644821055874,
+      "learning_rate": 1.691679628690953e-05,
+      "loss": 0.4347,
+      "step": 349
+    },
+    {
+      "epoch": 0.6653992395437263,
+      "grad_norm": 0.6310254774285764,
+      "learning_rate": 1.6892774339614927e-05,
+      "loss": 0.4098,
+      "step": 350
+    },
+    {
+      "epoch": 0.6673003802281369,
+      "grad_norm": 0.66669752039808,
+      "learning_rate": 1.686867637524896e-05,
+      "loss": 0.432,
+      "step": 351
+    },
+    {
+      "epoch": 0.6692015209125475,
+      "grad_norm": 0.6738514583274932,
+      "learning_rate": 1.6844502659576414e-05,
+      "loss": 0.4227,
+      "step": 352
+    },
+    {
+      "epoch": 0.6711026615969582,
+      "grad_norm": 0.7028774383826191,
+      "learning_rate": 1.6820253459197493e-05,
+      "loss": 0.4595,
+      "step": 353
+    },
+    {
+      "epoch": 0.6730038022813688,
+      "grad_norm": 0.7194031634018965,
+      "learning_rate": 1.679592904154489e-05,
+      "loss": 0.4261,
+      "step": 354
+    },
+    {
+      "epoch": 0.6749049429657795,
+      "grad_norm": 0.6639128000195935,
+      "learning_rate": 1.677152967488084e-05,
+      "loss": 0.4455,
+      "step": 355
+    },
+    {
+      "epoch": 0.6768060836501901,
+      "grad_norm": 0.7013139248665963,
+      "learning_rate": 1.6747055628294134e-05,
+      "loss": 0.4393,
+      "step": 356
+    },
+    {
+      "epoch": 0.6787072243346007,
+      "grad_norm": 0.7286047299541876,
+      "learning_rate": 1.6722507171697184e-05,
+      "loss": 0.4399,
+      "step": 357
+    },
+    {
+      "epoch": 0.6806083650190115,
+      "grad_norm": 0.6653840354319058,
+      "learning_rate": 1.669788457582304e-05,
+      "loss": 0.4543,
+      "step": 358
+    },
+    {
+      "epoch": 0.6825095057034221,
+      "grad_norm": 0.6964943672651862,
+      "learning_rate": 1.6673188112222394e-05,
+      "loss": 0.4157,
+      "step": 359
+    },
+    {
+      "epoch": 0.6844106463878327,
+      "grad_norm": 0.7074424760667147,
+      "learning_rate": 1.6648418053260585e-05,
+      "loss": 0.4297,
+      "step": 360
+    },
+    {
+      "epoch": 0.6863117870722434,
+      "grad_norm": 0.6215401441557756,
+      "learning_rate": 1.6623574672114596e-05,
+      "loss": 0.4319,
+      "step": 361
+    },
+    {
+      "epoch": 0.688212927756654,
+      "grad_norm": 0.7342134016704186,
+      "learning_rate": 1.6598658242770054e-05,
+      "loss": 0.4334,
+      "step": 362
+    },
+    {
+      "epoch": 0.6901140684410646,
+      "grad_norm": 0.7965253378724604,
+      "learning_rate": 1.6573669040018202e-05,
+      "loss": 0.4233,
+      "step": 363
+    },
+    {
+      "epoch": 0.6920152091254753,
+      "grad_norm": 0.6398428412224858,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.416,
+      "step": 364
+    },
+    {
+      "epoch": 0.6939163498098859,
+      "grad_norm": 0.6604114625988922,
+      "learning_rate": 1.652347341746737e-05,
+      "loss": 0.4303,
+      "step": 365
+    },
+    {
+      "epoch": 0.6958174904942965,
+      "grad_norm": 0.7014715733807383,
+      "learning_rate": 1.6498267551251618e-05,
+      "loss": 0.4217,
+      "step": 366
+    },
+    {
+      "epoch": 0.6977186311787072,
+      "grad_norm": 0.6113482323588421,
+      "learning_rate": 1.6472990018788884e-05,
+      "loss": 0.4207,
+      "step": 367
+    },
+    {
+      "epoch": 0.6996197718631179,
+      "grad_norm": 0.642687875343853,
+      "learning_rate": 1.644764109885284e-05,
+      "loss": 0.441,
+      "step": 368
+    },
+    {
+      "epoch": 0.7015209125475285,
+      "grad_norm": 0.7595072400925998,
+      "learning_rate": 1.642222107100446e-05,
+      "loss": 0.4372,
+      "step": 369
+    },
+    {
+      "epoch": 0.7034220532319392,
+      "grad_norm": 0.6850130290445166,
+      "learning_rate": 1.6396730215588913e-05,
+      "loss": 0.4335,
+      "step": 370
+    },
+    {
+      "epoch": 0.7053231939163498,
+      "grad_norm": 0.7533775105677387,
+      "learning_rate": 1.6371168813732514e-05,
+      "loss": 0.4215,
+      "step": 371
+    },
+    {
+      "epoch": 0.7072243346007605,
+      "grad_norm": 0.6597760164801185,
+      "learning_rate": 1.6345537147339578e-05,
+      "loss": 0.417,
+      "step": 372
+    },
+    {
+      "epoch": 0.7091254752851711,
+      "grad_norm": 0.6701191108201907,
+      "learning_rate": 1.6319835499089358e-05,
+      "loss": 0.3997,
+      "step": 373
+    },
+    {
+      "epoch": 0.7110266159695817,
+      "grad_norm": 0.677228004682316,
+      "learning_rate": 1.6294064152432878e-05,
+      "loss": 0.4182,
+      "step": 374
+    },
+    {
+      "epoch": 0.7129277566539924,
+      "grad_norm": 0.6558490282354752,
+      "learning_rate": 1.626822339158985e-05,
+      "loss": 0.4061,
+      "step": 375
+    },
+    {
+      "epoch": 0.714828897338403,
+      "grad_norm": 0.6574849005432378,
+      "learning_rate": 1.6242313501545522e-05,
+      "loss": 0.4266,
+      "step": 376
+    },
+    {
+      "epoch": 0.7167300380228137,
+      "grad_norm": 0.6686970300935623,
+      "learning_rate": 1.621633476804752e-05,
+      "loss": 0.4469,
+      "step": 377
+    },
+    {
+      "epoch": 0.7186311787072244,
+      "grad_norm": 0.6585843577537227,
+      "learning_rate": 1.6190287477602716e-05,
+      "loss": 0.4223,
+      "step": 378
+    },
+    {
+      "epoch": 0.720532319391635,
+      "grad_norm": 0.6874653499532027,
+      "learning_rate": 1.6164171917474078e-05,
+      "loss": 0.4186,
+      "step": 379
+    },
+    {
+      "epoch": 0.7224334600760456,
+      "grad_norm": 0.6953364571051063,
+      "learning_rate": 1.6137988375677466e-05,
+      "loss": 0.472,
+      "step": 380
+    },
+    {
+      "epoch": 0.7243346007604563,
+      "grad_norm": 0.6643376464866113,
+      "learning_rate": 1.6111737140978495e-05,
+      "loss": 0.4182,
+      "step": 381
+    },
+    {
+      "epoch": 0.7262357414448669,
+      "grad_norm": 0.6440559396830852,
+      "learning_rate": 1.6085418502889315e-05,
+      "loss": 0.426,
+      "step": 382
+    },
+    {
+      "epoch": 0.7281368821292775,
+      "grad_norm": 0.7443861679510732,
+      "learning_rate": 1.6059032751665454e-05,
+      "loss": 0.4525,
+      "step": 383
+    },
+    {
+      "epoch": 0.7300380228136882,
+      "grad_norm": 0.7544747826221699,
+      "learning_rate": 1.6032580178302585e-05,
+      "loss": 0.4548,
+      "step": 384
+    },
+    {
+      "epoch": 0.7319391634980988,
+      "grad_norm": 0.7101133516046464,
+      "learning_rate": 1.600606107453333e-05,
+      "loss": 0.4284,
+      "step": 385
+    },
+    {
+      "epoch": 0.7338403041825095,
+      "grad_norm": 0.6467712497112342,
+      "learning_rate": 1.597947573282405e-05,
+      "loss": 0.4288,
+      "step": 386
+    },
+    {
+      "epoch": 0.7357414448669202,
+      "grad_norm": 0.8285602540140268,
+      "learning_rate": 1.5952824446371608e-05,
+      "loss": 0.4625,
+      "step": 387
+    },
+    {
+      "epoch": 0.7376425855513308,
+      "grad_norm": 0.6274516214837191,
+      "learning_rate": 1.592610750910014e-05,
+      "loss": 0.414,
+      "step": 388
+    },
+    {
+      "epoch": 0.7395437262357415,
+      "grad_norm": 0.6616928035334787,
+      "learning_rate": 1.589932521565781e-05,
+      "loss": 0.4184,
+      "step": 389
+    },
+    {
+      "epoch": 0.7414448669201521,
+      "grad_norm": 0.7238516405222247,
+      "learning_rate": 1.587247786141358e-05,
+      "loss": 0.425,
+      "step": 390
+    },
+    {
+      "epoch": 0.7433460076045627,
+      "grad_norm": 0.6801422344967514,
+      "learning_rate": 1.5845565742453906e-05,
+      "loss": 0.4243,
+      "step": 391
+    },
+    {
+      "epoch": 0.7452471482889734,
+      "grad_norm": 0.6825219208634737,
+      "learning_rate": 1.581858915557953e-05,
+      "loss": 0.4359,
+      "step": 392
+    },
+    {
+      "epoch": 0.747148288973384,
+      "grad_norm": 0.7642489423557044,
+      "learning_rate": 1.5791548398302167e-05,
+      "loss": 0.4237,
+      "step": 393
+    },
+    {
+      "epoch": 0.7490494296577946,
+      "grad_norm": 0.6252312550975114,
+      "learning_rate": 1.5764443768841234e-05,
+      "loss": 0.3976,
+      "step": 394
+    },
+    {
+      "epoch": 0.7509505703422054,
+      "grad_norm": 0.7642977075462019,
+      "learning_rate": 1.5737275566120577e-05,
+      "loss": 0.4331,
+      "step": 395
+    },
+    {
+      "epoch": 0.752851711026616,
+      "grad_norm": 0.63348697398912,
+      "learning_rate": 1.5710044089765144e-05,
+      "loss": 0.4335,
+      "step": 396
+    },
+    {
+      "epoch": 0.7547528517110266,
+      "grad_norm": 0.6548373981216499,
+      "learning_rate": 1.5682749640097708e-05,
+      "loss": 0.4238,
+      "step": 397
+    },
+    {
+      "epoch": 0.7566539923954373,
+      "grad_norm": 0.6695247148969895,
+      "learning_rate": 1.565539251813554e-05,
+      "loss": 0.3983,
+      "step": 398
+    },
+    {
+      "epoch": 0.7585551330798479,
+      "grad_norm": 0.6936441735854132,
+      "learning_rate": 1.5627973025587093e-05,
+      "loss": 0.447,
+      "step": 399
+    },
+    {
+      "epoch": 0.7604562737642585,
+      "grad_norm": 0.6829450328460898,
+      "learning_rate": 1.560049146484868e-05,
+      "loss": 0.4339,
+      "step": 400
+    },
+    {
+      "epoch": 0.7623574144486692,
+      "grad_norm": 0.6831295920285226,
+      "learning_rate": 1.5572948139001128e-05,
+      "loss": 0.433,
+      "step": 401
+    },
+    {
+      "epoch": 0.7642585551330798,
+      "grad_norm": 0.5875633318170729,
+      "learning_rate": 1.5545343351806443e-05,
+      "loss": 0.3983,
+      "step": 402
+    },
+    {
+      "epoch": 0.7661596958174905,
+      "grad_norm": 0.6365055624072052,
+      "learning_rate": 1.551767740770446e-05,
+      "loss": 0.411,
+      "step": 403
+    },
+    {
+      "epoch": 0.7680608365019012,
+      "grad_norm": 0.6796189162611886,
+      "learning_rate": 1.5489950611809484e-05,
+      "loss": 0.415,
+      "step": 404
+    },
+    {
+      "epoch": 0.7699619771863118,
+      "grad_norm": 0.5974993486194421,
+      "learning_rate": 1.5462163269906928e-05,
+      "loss": 0.4297,
+      "step": 405
+    },
+    {
+      "epoch": 0.7718631178707225,
+      "grad_norm": 0.6644522363450603,
+      "learning_rate": 1.5434315688449924e-05,
+      "loss": 0.4402,
+      "step": 406
+    },
+    {
+      "epoch": 0.7737642585551331,
+      "grad_norm": 0.6646893143068616,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.4061,
+      "step": 407
+    },
+    {
+      "epoch": 0.7756653992395437,
+      "grad_norm": 0.6440666355794912,
+      "learning_rate": 1.5378441036003543e-05,
+      "loss": 0.4381,
+      "step": 408
+    },
+    {
+      "epoch": 0.7775665399239544,
+      "grad_norm": 0.7859236359603424,
+      "learning_rate": 1.535041458122865e-05,
+      "loss": 0.4109,
+      "step": 409
+    },
+    {
+      "epoch": 0.779467680608365,
+      "grad_norm": 0.6429359494979248,
+      "learning_rate": 1.5322329119321508e-05,
+      "loss": 0.4152,
+      "step": 410
+    },
+    {
+      "epoch": 0.7813688212927756,
+      "grad_norm": 0.706404918789796,
+      "learning_rate": 1.529418496002308e-05,
+      "loss": 0.4248,
+      "step": 411
+    },
+    {
+      "epoch": 0.7832699619771863,
+      "grad_norm": 0.6436928059769736,
+      "learning_rate": 1.5265982413721662e-05,
+      "loss": 0.4005,
+      "step": 412
+    },
+    {
+      "epoch": 0.785171102661597,
+      "grad_norm": 0.6391533149458034,
+      "learning_rate": 1.5237721791449497e-05,
+      "loss": 0.4111,
+      "step": 413
+    },
+    {
+      "epoch": 0.7870722433460076,
+      "grad_norm": 0.6515561967119322,
+      "learning_rate": 1.5209403404879305e-05,
+      "loss": 0.418,
+      "step": 414
+    },
+    {
+      "epoch": 0.7889733840304183,
+      "grad_norm": 0.6368167004607705,
+      "learning_rate": 1.5181027566320858e-05,
+      "loss": 0.4159,
+      "step": 415
+    },
+    {
+      "epoch": 0.7908745247148289,
+      "grad_norm": 0.6351695005051988,
+      "learning_rate": 1.5152594588717544e-05,
+      "loss": 0.4407,
+      "step": 416
+    },
+    {
+      "epoch": 0.7927756653992395,
+      "grad_norm": 0.7686405559820786,
+      "learning_rate": 1.5124104785642909e-05,
+      "loss": 0.433,
+      "step": 417
+    },
+    {
+      "epoch": 0.7946768060836502,
+      "grad_norm": 0.646702747191919,
+      "learning_rate": 1.5095558471297196e-05,
+      "loss": 0.4292,
+      "step": 418
+    },
+    {
+      "epoch": 0.7965779467680608,
+      "grad_norm": 0.6319810875105967,
+      "learning_rate": 1.5066955960503893e-05,
+      "loss": 0.435,
+      "step": 419
+    },
+    {
+      "epoch": 0.7984790874524715,
+      "grad_norm": 0.6755618924957245,
+      "learning_rate": 1.5038297568706244e-05,
+      "loss": 0.4301,
+      "step": 420
+    },
+    {
+      "epoch": 0.8003802281368821,
+      "grad_norm": 0.6399496735222758,
+      "learning_rate": 1.5009583611963772e-05,
+      "loss": 0.4419,
+      "step": 421
+    },
+    {
+      "epoch": 0.8022813688212928,
+      "grad_norm": 0.6104585101347578,
+      "learning_rate": 1.4980814406948806e-05,
+      "loss": 0.4232,
+      "step": 422
+    },
+    {
+      "epoch": 0.8041825095057035,
+      "grad_norm": 0.6464691255511562,
+      "learning_rate": 1.4951990270942991e-05,
+      "loss": 0.4219,
+      "step": 423
+    },
+    {
+      "epoch": 0.8060836501901141,
+      "grad_norm": 0.5882841991668659,
+      "learning_rate": 1.492311152183376e-05,
+      "loss": 0.3986,
+      "step": 424
+    },
+    {
+      "epoch": 0.8079847908745247,
+      "grad_norm": 0.6537065391248243,
+      "learning_rate": 1.4894178478110856e-05,
+      "loss": 0.4212,
+      "step": 425
+    },
+    {
+      "epoch": 0.8098859315589354,
+      "grad_norm": 0.5701443605402853,
+      "learning_rate": 1.4865191458862816e-05,
+      "loss": 0.4086,
+      "step": 426
+    },
+    {
+      "epoch": 0.811787072243346,
+      "grad_norm": 0.6578355816780781,
+      "learning_rate": 1.4836150783773442e-05,
+      "loss": 0.4043,
+      "step": 427
+    },
+    {
+      "epoch": 0.8136882129277566,
+      "grad_norm": 0.6516830518252411,
+      "learning_rate": 1.4807056773118276e-05,
+      "loss": 0.4393,
+      "step": 428
+    },
+    {
+      "epoch": 0.8155893536121673,
+      "grad_norm": 0.5693114254684896,
+      "learning_rate": 1.4777909747761085e-05,
+      "loss": 0.4028,
+      "step": 429
+    },
+    {
+      "epoch": 0.8174904942965779,
+      "grad_norm": 0.6806921217447237,
+      "learning_rate": 1.4748710029150296e-05,
+      "loss": 0.4221,
+      "step": 430
+    },
+    {
+      "epoch": 0.8193916349809885,
+      "grad_norm": 0.6521245597036365,
+      "learning_rate": 1.4719457939315468e-05,
+      "loss": 0.425,
+      "step": 431
+    },
+    {
+      "epoch": 0.8212927756653993,
+      "grad_norm": 0.6470885438077629,
+      "learning_rate": 1.4690153800863743e-05,
+      "loss": 0.4003,
+      "step": 432
+    },
+    {
+      "epoch": 0.8231939163498099,
+      "grad_norm": 0.620809219133859,
+      "learning_rate": 1.4660797936976278e-05,
+      "loss": 0.4088,
+      "step": 433
+    },
+    {
+      "epoch": 0.8250950570342205,
+      "grad_norm": 0.5797646417810149,
+      "learning_rate": 1.4631390671404682e-05,
+      "loss": 0.4015,
+      "step": 434
+    },
+    {
+      "epoch": 0.8269961977186312,
+      "grad_norm": 0.6586587263921959,
+      "learning_rate": 1.460193232846745e-05,
+      "loss": 0.4439,
+      "step": 435
+    },
+    {
+      "epoch": 0.8288973384030418,
+      "grad_norm": 0.6889970521645293,
+      "learning_rate": 1.4572423233046386e-05,
+      "loss": 0.4318,
+      "step": 436
+    },
+    {
+      "epoch": 0.8307984790874525,
+      "grad_norm": 0.6326201069773926,
+      "learning_rate": 1.4542863710583022e-05,
+      "loss": 0.4322,
+      "step": 437
+    },
+    {
+      "epoch": 0.8326996197718631,
+      "grad_norm": 0.6080188080052302,
+      "learning_rate": 1.4513254087075015e-05,
+      "loss": 0.424,
+      "step": 438
+    },
+    {
+      "epoch": 0.8346007604562737,
+      "grad_norm": 0.6163198199318526,
+      "learning_rate": 1.4483594689072571e-05,
+      "loss": 0.4112,
+      "step": 439
+    },
+    {
+      "epoch": 0.8365019011406845,
+      "grad_norm": 0.6362744727754932,
+      "learning_rate": 1.4453885843674837e-05,
+      "loss": 0.4217,
+      "step": 440
+    },
+    {
+      "epoch": 0.8384030418250951,
+      "grad_norm": 0.6820784181713129,
+      "learning_rate": 1.4424127878526278e-05,
+      "loss": 0.4263,
+      "step": 441
+    },
+    {
+      "epoch": 0.8403041825095057,
+      "grad_norm": 0.643612502055302,
+      "learning_rate": 1.4394321121813093e-05,
+      "loss": 0.4075,
+      "step": 442
+    },
+    {
+      "epoch": 0.8422053231939164,
+      "grad_norm": 0.6564550619308712,
+      "learning_rate": 1.436446590225957e-05,
+      "loss": 0.4291,
+      "step": 443
+    },
+    {
+      "epoch": 0.844106463878327,
+      "grad_norm": 0.6138393284551312,
+      "learning_rate": 1.433456254912447e-05,
+      "loss": 0.3934,
+      "step": 444
+    },
+    {
+      "epoch": 0.8460076045627376,
+      "grad_norm": 0.766957990765335,
+      "learning_rate": 1.4304611392197399e-05,
+      "loss": 0.4186,
+      "step": 445
+    },
+    {
+      "epoch": 0.8479087452471483,
+      "grad_norm": 0.7298652465877792,
+      "learning_rate": 1.427461276179517e-05,
+      "loss": 0.4369,
+      "step": 446
+    },
+    {
+      "epoch": 0.8498098859315589,
+      "grad_norm": 0.6284698144096496,
+      "learning_rate": 1.4244566988758152e-05,
+      "loss": 0.415,
+      "step": 447
+    },
+    {
+      "epoch": 0.8517110266159695,
+      "grad_norm": 0.6332420376619956,
+      "learning_rate": 1.4214474404446633e-05,
+      "loss": 0.4089,
+      "step": 448
+    },
+    {
+      "epoch": 0.8536121673003803,
+      "grad_norm": 0.6804506480077341,
+      "learning_rate": 1.4184335340737158e-05,
+      "loss": 0.4156,
+      "step": 449
+    },
+    {
+      "epoch": 0.8555133079847909,
+      "grad_norm": 0.7327321799801652,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.422,
+      "step": 450
+    },
+    {
+      "epoch": 0.8574144486692015,
+      "grad_norm": 0.6371639351762105,
+      "learning_rate": 1.4123919105189836e-05,
+      "loss": 0.4333,
+      "step": 451
+    },
+    {
+      "epoch": 0.8593155893536122,
+      "grad_norm": 0.7269433665153127,
+      "learning_rate": 1.4093642599653406e-05,
+      "loss": 0.4325,
+      "step": 452
+    },
+    {
+      "epoch": 0.8612167300380228,
+      "grad_norm": 0.6101990972658262,
+      "learning_rate": 1.40633209473145e-05,
+      "loss": 0.4237,
+      "step": 453
+    },
+    {
+      "epoch": 0.8631178707224335,
+      "grad_norm": 0.7103400151507081,
+      "learning_rate": 1.4032954482575938e-05,
+      "loss": 0.4222,
+      "step": 454
+    },
+    {
+      "epoch": 0.8650190114068441,
+      "grad_norm": 0.6571365049330196,
+      "learning_rate": 1.4002543540334766e-05,
+      "loss": 0.4313,
+      "step": 455
+    },
+    {
+      "epoch": 0.8669201520912547,
+      "grad_norm": 0.6180547188795785,
+      "learning_rate": 1.3972088455978537e-05,
+      "loss": 0.4183,
+      "step": 456
+    },
+    {
+      "epoch": 0.8688212927756654,
+      "grad_norm": 0.6370942015731359,
+      "learning_rate": 1.3941589565381635e-05,
+      "loss": 0.4068,
+      "step": 457
+    },
+    {
+      "epoch": 0.870722433460076,
+      "grad_norm": 0.7012219529358206,
+      "learning_rate": 1.391104720490156e-05,
+      "loss": 0.4509,
+      "step": 458
+    },
+    {
+      "epoch": 0.8726235741444867,
+      "grad_norm": 0.6373101149215029,
+      "learning_rate": 1.3880461711375224e-05,
+      "loss": 0.4009,
+      "step": 459
+    },
+    {
+      "epoch": 0.8745247148288974,
+      "grad_norm": 0.6858223509340476,
+      "learning_rate": 1.3849833422115221e-05,
+      "loss": 0.4285,
+      "step": 460
+    },
+    {
+      "epoch": 0.876425855513308,
+      "grad_norm": 0.6128745991660574,
+      "learning_rate": 1.3819162674906134e-05,
+      "loss": 0.422,
+      "step": 461
+    },
+    {
+      "epoch": 0.8783269961977186,
+      "grad_norm": 0.648129348909501,
+      "learning_rate": 1.378844980800078e-05,
+      "loss": 0.4061,
+      "step": 462
+    },
+    {
+      "epoch": 0.8802281368821293,
+      "grad_norm": 0.6700872992135477,
+      "learning_rate": 1.3757695160116502e-05,
+      "loss": 0.4488,
+      "step": 463
+    },
+    {
+      "epoch": 0.8821292775665399,
+      "grad_norm": 0.5933754973878622,
+      "learning_rate": 1.3726899070431423e-05,
+      "loss": 0.4176,
+      "step": 464
+    },
+    {
+      "epoch": 0.8840304182509505,
+      "grad_norm": 0.6376208348664184,
+      "learning_rate": 1.3696061878580707e-05,
+      "loss": 0.4358,
+      "step": 465
+    },
+    {
+      "epoch": 0.8859315589353612,
+      "grad_norm": 0.6144318643702079,
+      "learning_rate": 1.3665183924652817e-05,
+      "loss": 0.4264,
+      "step": 466
+    },
+    {
+      "epoch": 0.8878326996197718,
+      "grad_norm": 0.5968680174347751,
+      "learning_rate": 1.3634265549185755e-05,
+      "loss": 0.4061,
+      "step": 467
+    },
+    {
+      "epoch": 0.8897338403041825,
+      "grad_norm": 0.6300952807942402,
+      "learning_rate": 1.3603307093163319e-05,
+      "loss": 0.4454,
+      "step": 468
+    },
+    {
+      "epoch": 0.8916349809885932,
+      "grad_norm": 0.6503995723770262,
+      "learning_rate": 1.3572308898011328e-05,
+      "loss": 0.4198,
+      "step": 469
+    },
+    {
+      "epoch": 0.8935361216730038,
+      "grad_norm": 0.6016579977211428,
+      "learning_rate": 1.3541271305593878e-05,
+      "loss": 0.4267,
+      "step": 470
+    },
+    {
+      "epoch": 0.8954372623574145,
+      "grad_norm": 0.6155931894049604,
+      "learning_rate": 1.3510194658209547e-05,
+      "loss": 0.4219,
+      "step": 471
+    },
+    {
+      "epoch": 0.8973384030418251,
+      "grad_norm": 0.6251824588453468,
+      "learning_rate": 1.3479079298587634e-05,
+      "loss": 0.4087,
+      "step": 472
+    },
+    {
+      "epoch": 0.8992395437262357,
+      "grad_norm": 0.6246206861378312,
+      "learning_rate": 1.3447925569884374e-05,
+      "loss": 0.4169,
+      "step": 473
+    },
+    {
+      "epoch": 0.9011406844106464,
+      "grad_norm": 0.6437020276029598,
+      "learning_rate": 1.3416733815679166e-05,
+      "loss": 0.4181,
+      "step": 474
+    },
+    {
+      "epoch": 0.903041825095057,
+      "grad_norm": 0.6570051501669705,
+      "learning_rate": 1.3385504379970764e-05,
+      "loss": 0.4343,
+      "step": 475
+    },
+    {
+      "epoch": 0.9049429657794676,
+      "grad_norm": 0.6209378450559351,
+      "learning_rate": 1.3354237607173494e-05,
+      "loss": 0.4236,
+      "step": 476
+    },
+    {
+      "epoch": 0.9068441064638784,
+      "grad_norm": 0.6282984336963544,
+      "learning_rate": 1.3322933842113457e-05,
+      "loss": 0.4049,
+      "step": 477
+    },
+    {
+      "epoch": 0.908745247148289,
+      "grad_norm": 0.6730942448082525,
+      "learning_rate": 1.3291593430024727e-05,
+      "loss": 0.4205,
+      "step": 478
+    },
+    {
+      "epoch": 0.9106463878326996,
+      "grad_norm": 0.6358059934332029,
+      "learning_rate": 1.3260216716545534e-05,
+      "loss": 0.4173,
+      "step": 479
+    },
+    {
+      "epoch": 0.9125475285171103,
+      "grad_norm": 0.5966364822772298,
+      "learning_rate": 1.3228804047714462e-05,
+      "loss": 0.4032,
+      "step": 480
+    },
+    {
+      "epoch": 0.9144486692015209,
+      "grad_norm": 0.6338310009812316,
+      "learning_rate": 1.319735576996663e-05,
+      "loss": 0.4453,
+      "step": 481
+    },
+    {
+      "epoch": 0.9163498098859315,
+      "grad_norm": 0.625677440236051,
+      "learning_rate": 1.3165872230129869e-05,
+      "loss": 0.4173,
+      "step": 482
+    },
+    {
+      "epoch": 0.9182509505703422,
+      "grad_norm": 0.665352590056403,
+      "learning_rate": 1.3134353775420895e-05,
+      "loss": 0.4171,
+      "step": 483
+    },
+    {
+      "epoch": 0.9201520912547528,
+      "grad_norm": 0.6001235200645851,
+      "learning_rate": 1.3102800753441488e-05,
+      "loss": 0.4317,
+      "step": 484
+    },
+    {
+      "epoch": 0.9220532319391636,
+      "grad_norm": 0.6604942294876789,
+      "learning_rate": 1.3071213512174655e-05,
+      "loss": 0.3996,
+      "step": 485
+    },
+    {
+      "epoch": 0.9239543726235742,
+      "grad_norm": 0.6255711483205718,
+      "learning_rate": 1.3039592399980785e-05,
+      "loss": 0.4001,
+      "step": 486
+    },
+    {
+      "epoch": 0.9258555133079848,
+      "grad_norm": 0.6699243426942582,
+      "learning_rate": 1.3007937765593818e-05,
+      "loss": 0.4446,
+      "step": 487
+    },
+    {
+      "epoch": 0.9277566539923955,
+      "grad_norm": 0.6154153967597945,
+      "learning_rate": 1.2976249958117395e-05,
+      "loss": 0.4092,
+      "step": 488
+    },
+    {
+      "epoch": 0.9296577946768061,
+      "grad_norm": 0.6674175018085043,
+      "learning_rate": 1.2944529327021002e-05,
+      "loss": 0.4084,
+      "step": 489
+    },
+    {
+      "epoch": 0.9315589353612167,
+      "grad_norm": 0.6671703602261613,
+      "learning_rate": 1.291277622213612e-05,
+      "loss": 0.4072,
+      "step": 490
+    },
+    {
+      "epoch": 0.9334600760456274,
+      "grad_norm": 0.6273893293237608,
+      "learning_rate": 1.2880990993652379e-05,
+      "loss": 0.4209,
+      "step": 491
+    },
+    {
+      "epoch": 0.935361216730038,
+      "grad_norm": 0.6117796812456406,
+      "learning_rate": 1.2849173992113669e-05,
+      "loss": 0.4212,
+      "step": 492
+    },
+    {
+      "epoch": 0.9372623574144486,
+      "grad_norm": 0.6181494101476228,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.428,
+      "step": 493
+    },
+    {
+      "epoch": 0.9391634980988594,
+      "grad_norm": 0.7273393309788759,
+      "learning_rate": 1.2785446073795118e-05,
+      "loss": 0.4116,
+      "step": 494
+    },
+    {
+      "epoch": 0.94106463878327,
+      "grad_norm": 0.6767040540513543,
+      "learning_rate": 1.2753535859839638e-05,
+      "loss": 0.4212,
+      "step": 495
+    },
+    {
+      "epoch": 0.9429657794676806,
+      "grad_norm": 0.6217787850737705,
+      "learning_rate": 1.272159527847016e-05,
+      "loss": 0.4092,
+      "step": 496
+    },
+    {
+      "epoch": 0.9448669201520913,
+      "grad_norm": 0.6325058664070493,
+      "learning_rate": 1.2689624681943897e-05,
+      "loss": 0.4404,
+      "step": 497
+    },
+    {
+      "epoch": 0.9467680608365019,
+      "grad_norm": 0.6810237842054134,
+      "learning_rate": 1.2657624422849077e-05,
+      "loss": 0.436,
+      "step": 498
+    },
+    {
+      "epoch": 0.9486692015209125,
+      "grad_norm": 0.6490530802959269,
+      "learning_rate": 1.2625594854101066e-05,
+      "loss": 0.4341,
+      "step": 499
+    },
+    {
+      "epoch": 0.9505703422053232,
+      "grad_norm": 0.6350846489273293,
+      "learning_rate": 1.2593536328938471e-05,
+      "loss": 0.4265,
+      "step": 500
+    },
+    {
+      "epoch": 0.9524714828897338,
+      "grad_norm": 0.6748581679410075,
+      "learning_rate": 1.2561449200919253e-05,
+      "loss": 0.3987,
+      "step": 501
+    },
+    {
+      "epoch": 0.9543726235741445,
+      "grad_norm": 0.5656159130929157,
+      "learning_rate": 1.2529333823916807e-05,
+      "loss": 0.406,
+      "step": 502
+    },
+    {
+      "epoch": 0.9562737642585551,
+      "grad_norm": 0.6303710871278102,
+      "learning_rate": 1.2497190552116082e-05,
+      "loss": 0.4264,
+      "step": 503
+    },
+    {
+      "epoch": 0.9581749049429658,
+      "grad_norm": 0.6398016146690956,
+      "learning_rate": 1.2465019740009662e-05,
+      "loss": 0.4113,
+      "step": 504
+    },
+    {
+      "epoch": 0.9600760456273765,
+      "grad_norm": 0.5965106140502064,
+      "learning_rate": 1.2432821742393854e-05,
+      "loss": 0.415,
+      "step": 505
+    },
+    {
+      "epoch": 0.9619771863117871,
+      "grad_norm": 0.6695604111444653,
+      "learning_rate": 1.2400596914364792e-05,
+      "loss": 0.4104,
+      "step": 506
+    },
+    {
+      "epoch": 0.9638783269961977,
+      "grad_norm": 0.6581528364436859,
+      "learning_rate": 1.2368345611314508e-05,
+      "loss": 0.4168,
+      "step": 507
+    },
+    {
+      "epoch": 0.9657794676806084,
+      "grad_norm": 0.6167627899157956,
+      "learning_rate": 1.2336068188927002e-05,
+      "loss": 0.4092,
+      "step": 508
+    },
+    {
+      "epoch": 0.967680608365019,
+      "grad_norm": 0.6047519925905133,
+      "learning_rate": 1.2303765003174342e-05,
+      "loss": 0.4201,
+      "step": 509
+    },
+    {
+      "epoch": 0.9695817490494296,
+      "grad_norm": 0.6184925217192032,
+      "learning_rate": 1.2271436410312727e-05,
+      "loss": 0.4384,
+      "step": 510
+    },
+    {
+      "epoch": 0.9714828897338403,
+      "grad_norm": 0.630861620394623,
+      "learning_rate": 1.2239082766878557e-05,
+      "loss": 0.4308,
+      "step": 511
+    },
+    {
+      "epoch": 0.973384030418251,
+      "grad_norm": 0.6749925922751687,
+      "learning_rate": 1.2206704429684504e-05,
+      "loss": 0.4366,
+      "step": 512
+    },
+    {
+      "epoch": 0.9752851711026616,
+      "grad_norm": 0.5826742426817829,
+      "learning_rate": 1.2174301755815572e-05,
+      "loss": 0.4221,
+      "step": 513
+    },
+    {
+      "epoch": 0.9771863117870723,
+      "grad_norm": 0.630771658911766,
+      "learning_rate": 1.2141875102625166e-05,
+      "loss": 0.4067,
+      "step": 514
+    },
+    {
+      "epoch": 0.9790874524714829,
+      "grad_norm": 0.6541500497911698,
+      "learning_rate": 1.2109424827731144e-05,
+      "loss": 0.421,
+      "step": 515
+    },
+    {
+      "epoch": 0.9809885931558935,
+      "grad_norm": 0.6199947283738567,
+      "learning_rate": 1.2076951289011884e-05,
+      "loss": 0.4345,
+      "step": 516
+    },
+    {
+      "epoch": 0.9828897338403042,
+      "grad_norm": 0.6957933168559665,
+      "learning_rate": 1.204445484460232e-05,
+      "loss": 0.4284,
+      "step": 517
+    },
+    {
+      "epoch": 0.9847908745247148,
+      "grad_norm": 0.6187195977259033,
+      "learning_rate": 1.2011935852890004e-05,
+      "loss": 0.4159,
+      "step": 518
+    },
+    {
+      "epoch": 0.9866920152091255,
+      "grad_norm": 0.6846454880459624,
+      "learning_rate": 1.1979394672511156e-05,
+      "loss": 0.3931,
+      "step": 519
+    },
+    {
+      "epoch": 0.9885931558935361,
+      "grad_norm": 0.6983866879763947,
+      "learning_rate": 1.19468316623467e-05,
+      "loss": 0.4425,
+      "step": 520
+    },
+    {
+      "epoch": 0.9904942965779467,
+      "grad_norm": 0.6409799621748642,
+      "learning_rate": 1.1914247181518312e-05,
+      "loss": 0.4045,
+      "step": 521
+    },
+    {
+      "epoch": 0.9923954372623575,
+      "grad_norm": 0.6613348843196384,
+      "learning_rate": 1.1881641589384456e-05,
+      "loss": 0.4282,
+      "step": 522
+    },
+    {
+      "epoch": 0.9942965779467681,
+      "grad_norm": 0.6245921104406804,
+      "learning_rate": 1.1849015245536424e-05,
+      "loss": 0.4321,
+      "step": 523
+    },
+    {
+      "epoch": 0.9961977186311787,
+      "grad_norm": 0.6059598558539324,
+      "learning_rate": 1.1816368509794365e-05,
+      "loss": 0.4231,
+      "step": 524
+    },
+    {
+      "epoch": 0.9980988593155894,
+      "grad_norm": 0.6525328754595001,
+      "learning_rate": 1.1783701742203326e-05,
+      "loss": 0.3872,
+      "step": 525
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6132253139977111,
+      "learning_rate": 1.1751015303029272e-05,
+      "loss": 0.4178,
+      "step": 526
+    },
+    {
+      "epoch": 1.0019011406844107,
+      "grad_norm": 0.8006565194435021,
+      "learning_rate": 1.1718309552755118e-05,
+      "loss": 0.3304,
+      "step": 527
+    },
+    {
+      "epoch": 1.0038022813688212,
+      "grad_norm": 0.7657577845765433,
+      "learning_rate": 1.1685584852076746e-05,
+      "loss": 0.335,
+      "step": 528
+    },
+    {
+      "epoch": 1.005703422053232,
+      "grad_norm": 0.6142222913024673,
+      "learning_rate": 1.1652841561899042e-05,
+      "loss": 0.3246,
+      "step": 529
+    },
+    {
+      "epoch": 1.0076045627376427,
+      "grad_norm": 0.608928243732463,
+      "learning_rate": 1.1620080043331901e-05,
+      "loss": 0.3122,
+      "step": 530
+    },
+    {
+      "epoch": 1.0095057034220531,
+      "grad_norm": 0.7697422948905064,
+      "learning_rate": 1.1587300657686254e-05,
+      "loss": 0.3334,
+      "step": 531
+    },
+    {
+      "epoch": 1.0114068441064639,
+      "grad_norm": 0.7656692565553174,
+      "learning_rate": 1.1554503766470069e-05,
+      "loss": 0.3244,
+      "step": 532
+    },
+    {
+      "epoch": 1.0133079847908746,
+      "grad_norm": 0.780562025140702,
+      "learning_rate": 1.1521689731384391e-05,
+      "loss": 0.3356,
+      "step": 533
+    },
+    {
+      "epoch": 1.015209125475285,
+      "grad_norm": 0.752933278659135,
+      "learning_rate": 1.1488858914319321e-05,
+      "loss": 0.3264,
+      "step": 534
+    },
+    {
+      "epoch": 1.0171102661596958,
+      "grad_norm": 0.6748302791115485,
+      "learning_rate": 1.1456011677350052e-05,
+      "loss": 0.3074,
+      "step": 535
+    },
+    {
+      "epoch": 1.0190114068441065,
+      "grad_norm": 0.6736181146135486,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.3302,
+      "step": 536
+    },
+    {
+      "epoch": 1.020912547528517,
+      "grad_norm": 0.7213457946548599,
+      "learning_rate": 1.1390269392901096e-05,
+      "loss": 0.3144,
+      "step": 537
+    },
+    {
+      "epoch": 1.0228136882129277,
+      "grad_norm": 0.6537213510152631,
+      "learning_rate": 1.1357375070461241e-05,
+      "loss": 0.3107,
+      "step": 538
+    },
+    {
+      "epoch": 1.0247148288973384,
+      "grad_norm": 0.6495715054864254,
+      "learning_rate": 1.1324465778188846e-05,
+      "loss": 0.3253,
+      "step": 539
+    },
+    {
+      "epoch": 1.026615969581749,
+      "grad_norm": 0.6907962860713626,
+      "learning_rate": 1.1291541879024568e-05,
+      "loss": 0.3318,
+      "step": 540
+    },
+    {
+      "epoch": 1.0285171102661597,
+      "grad_norm": 0.6342419142188326,
+      "learning_rate": 1.1258603736070145e-05,
+      "loss": 0.2977,
+      "step": 541
+    },
+    {
+      "epoch": 1.0304182509505704,
+      "grad_norm": 0.6573226817001695,
+      "learning_rate": 1.1225651712584413e-05,
+      "loss": 0.3214,
+      "step": 542
+    },
+    {
+      "epoch": 1.032319391634981,
+      "grad_norm": 0.6700148455556502,
+      "learning_rate": 1.1192686171979288e-05,
+      "loss": 0.3229,
+      "step": 543
+    },
+    {
+      "epoch": 1.0342205323193916,
+      "grad_norm": 0.6465298686791402,
+      "learning_rate": 1.1159707477815756e-05,
+      "loss": 0.3056,
+      "step": 544
+    },
+    {
+      "epoch": 1.0361216730038023,
+      "grad_norm": 0.6683181379039095,
+      "learning_rate": 1.1126715993799875e-05,
+      "loss": 0.3109,
+      "step": 545
+    },
+    {
+      "epoch": 1.038022813688213,
+      "grad_norm": 0.6269966144191473,
+      "learning_rate": 1.1093712083778748e-05,
+      "loss": 0.3107,
+      "step": 546
+    },
+    {
+      "epoch": 1.0399239543726235,
+      "grad_norm": 0.6620472814466998,
+      "learning_rate": 1.1060696111736515e-05,
+      "loss": 0.3162,
+      "step": 547
+    },
+    {
+      "epoch": 1.0418250950570342,
+      "grad_norm": 0.6642661451936002,
+      "learning_rate": 1.1027668441790358e-05,
+      "loss": 0.3083,
+      "step": 548
+    },
+    {
+      "epoch": 1.043726235741445,
+      "grad_norm": 0.7363895997405029,
+      "learning_rate": 1.099462943818646e-05,
+      "loss": 0.3225,
+      "step": 549
+    },
+    {
+      "epoch": 1.0456273764258555,
+      "grad_norm": 0.6436015034359204,
+      "learning_rate": 1.0961579465295987e-05,
+      "loss": 0.3108,
+      "step": 550
+    },
+    {
+      "epoch": 1.0475285171102662,
+      "grad_norm": 0.7344966534581019,
+      "learning_rate": 1.0928518887611099e-05,
+      "loss": 0.323,
+      "step": 551
+    },
+    {
+      "epoch": 1.049429657794677,
+      "grad_norm": 0.6772518626597273,
+      "learning_rate": 1.0895448069740902e-05,
+      "loss": 0.3211,
+      "step": 552
+    },
+    {
+      "epoch": 1.0513307984790874,
+      "grad_norm": 0.6354982894550579,
+      "learning_rate": 1.0862367376407433e-05,
+      "loss": 0.3088,
+      "step": 553
+    },
+    {
+      "epoch": 1.053231939163498,
+      "grad_norm": 0.6894784300347995,
+      "learning_rate": 1.0829277172441648e-05,
+      "loss": 0.324,
+      "step": 554
+    },
+    {
+      "epoch": 1.0551330798479088,
+      "grad_norm": 0.6940763741696211,
+      "learning_rate": 1.0796177822779384e-05,
+      "loss": 0.3223,
+      "step": 555
+    },
+    {
+      "epoch": 1.0570342205323193,
+      "grad_norm": 0.7360005823156429,
+      "learning_rate": 1.0763069692457346e-05,
+      "loss": 0.3167,
+      "step": 556
+    },
+    {
+      "epoch": 1.05893536121673,
+      "grad_norm": 0.6876125931257574,
+      "learning_rate": 1.0729953146609076e-05,
+      "loss": 0.3334,
+      "step": 557
+    },
+    {
+      "epoch": 1.0608365019011408,
+      "grad_norm": 0.7991561950953083,
+      "learning_rate": 1.0696828550460928e-05,
+      "loss": 0.3129,
+      "step": 558
+    },
+    {
+      "epoch": 1.0627376425855513,
+      "grad_norm": 0.7890798952761741,
+      "learning_rate": 1.0663696269328034e-05,
+      "loss": 0.3379,
+      "step": 559
+    },
+    {
+      "epoch": 1.064638783269962,
+      "grad_norm": 0.614876417508881,
+      "learning_rate": 1.0630556668610286e-05,
+      "loss": 0.2895,
+      "step": 560
+    },
+    {
+      "epoch": 1.0665399239543727,
+      "grad_norm": 0.839869731753892,
+      "learning_rate": 1.059741011378829e-05,
+      "loss": 0.3084,
+      "step": 561
+    },
+    {
+      "epoch": 1.0684410646387832,
+      "grad_norm": 0.6716771988189453,
+      "learning_rate": 1.0564256970419367e-05,
+      "loss": 0.3169,
+      "step": 562
+    },
+    {
+      "epoch": 1.070342205323194,
+      "grad_norm": 0.6480892215578606,
+      "learning_rate": 1.0531097604133473e-05,
+      "loss": 0.3199,
+      "step": 563
+    },
+    {
+      "epoch": 1.0722433460076046,
+      "grad_norm": 0.726897910174651,
+      "learning_rate": 1.0497932380629207e-05,
+      "loss": 0.3257,
+      "step": 564
+    },
+    {
+      "epoch": 1.0741444866920151,
+      "grad_norm": 0.7420412843722907,
+      "learning_rate": 1.0464761665669771e-05,
+      "loss": 0.3181,
+      "step": 565
+    },
+    {
+      "epoch": 1.0760456273764258,
+      "grad_norm": 0.647386690986311,
+      "learning_rate": 1.0431585825078916e-05,
+      "loss": 0.3145,
+      "step": 566
+    },
+    {
+      "epoch": 1.0779467680608366,
+      "grad_norm": 0.6864342448364056,
+      "learning_rate": 1.0398405224736927e-05,
+      "loss": 0.3127,
+      "step": 567
+    },
+    {
+      "epoch": 1.079847908745247,
+      "grad_norm": 0.6700581210314871,
+      "learning_rate": 1.0365220230576592e-05,
+      "loss": 0.2992,
+      "step": 568
+    },
+    {
+      "epoch": 1.0817490494296578,
+      "grad_norm": 0.6945855597922971,
+      "learning_rate": 1.0332031208579133e-05,
+      "loss": 0.3092,
+      "step": 569
+    },
+    {
+      "epoch": 1.0836501901140685,
+      "grad_norm": 0.6647499003220013,
+      "learning_rate": 1.0298838524770212e-05,
+      "loss": 0.2999,
+      "step": 570
+    },
+    {
+      "epoch": 1.085551330798479,
+      "grad_norm": 0.6385465861056062,
+      "learning_rate": 1.0265642545215872e-05,
+      "loss": 0.3145,
+      "step": 571
+    },
+    {
+      "epoch": 1.0874524714828897,
+      "grad_norm": 0.7525825678222171,
+      "learning_rate": 1.0232443636018502e-05,
+      "loss": 0.3242,
+      "step": 572
+    },
+    {
+      "epoch": 1.0893536121673004,
+      "grad_norm": 0.7807508547070708,
+      "learning_rate": 1.0199242163312794e-05,
+      "loss": 0.2983,
+      "step": 573
+    },
+    {
+      "epoch": 1.091254752851711,
+      "grad_norm": 0.6363833892399261,
+      "learning_rate": 1.0166038493261723e-05,
+      "loss": 0.2914,
+      "step": 574
+    },
+    {
+      "epoch": 1.0931558935361216,
+      "grad_norm": 0.8881711147926716,
+      "learning_rate": 1.013283299205249e-05,
+      "loss": 0.3052,
+      "step": 575
+    },
+    {
+      "epoch": 1.0950570342205324,
+      "grad_norm": 0.8005244886940339,
+      "learning_rate": 1.0099626025892491e-05,
+      "loss": 0.3048,
+      "step": 576
+    },
+    {
+      "epoch": 1.0969581749049429,
+      "grad_norm": 0.6983302950744485,
+      "learning_rate": 1.0066417961005283e-05,
+      "loss": 0.3124,
+      "step": 577
+    },
+    {
+      "epoch": 1.0988593155893536,
+      "grad_norm": 0.7728914074284565,
+      "learning_rate": 1.0033209163626539e-05,
+      "loss": 0.3132,
+      "step": 578
+    },
+    {
+      "epoch": 1.1007604562737643,
+      "grad_norm": 0.7886501715429274,
+      "learning_rate": 1e-05,
+      "loss": 0.3208,
+      "step": 579
+    },
+    {
+      "epoch": 1.102661596958175,
+      "grad_norm": 0.6473053674998779,
+      "learning_rate": 9.966790836373465e-06,
+      "loss": 0.3037,
+      "step": 580
+    },
+    {
+      "epoch": 1.1045627376425855,
+      "grad_norm": 0.657175499636643,
+      "learning_rate": 9.933582038994719e-06,
+      "loss": 0.3027,
+      "step": 581
+    },
+    {
+      "epoch": 1.1064638783269962,
+      "grad_norm": 0.7794391035494178,
+      "learning_rate": 9.90037397410751e-06,
+      "loss": 0.3198,
+      "step": 582
+    },
+    {
+      "epoch": 1.108365019011407,
+      "grad_norm": 0.7628090167991661,
+      "learning_rate": 9.867167007947511e-06,
+      "loss": 0.3112,
+      "step": 583
+    },
+    {
+      "epoch": 1.1102661596958174,
+      "grad_norm": 0.6578383217183126,
+      "learning_rate": 9.833961506738282e-06,
+      "loss": 0.3114,
+      "step": 584
+    },
+    {
+      "epoch": 1.1121673003802282,
+      "grad_norm": 0.7506195755600762,
+      "learning_rate": 9.80075783668721e-06,
+      "loss": 0.3017,
+      "step": 585
+    },
+    {
+      "epoch": 1.1140684410646389,
+      "grad_norm": 0.7742655530860587,
+      "learning_rate": 9.767556363981503e-06,
+      "loss": 0.3338,
+      "step": 586
+    },
+    {
+      "epoch": 1.1159695817490494,
+      "grad_norm": 0.6807864153878859,
+      "learning_rate": 9.734357454784131e-06,
+      "loss": 0.3247,
+      "step": 587
+    },
+    {
+      "epoch": 1.11787072243346,
+      "grad_norm": 0.7380207497667768,
+      "learning_rate": 9.701161475229791e-06,
+      "loss": 0.3125,
+      "step": 588
+    },
+    {
+      "epoch": 1.1197718631178708,
+      "grad_norm": 0.7052663565337727,
+      "learning_rate": 9.66796879142087e-06,
+      "loss": 0.3007,
+      "step": 589
+    },
+    {
+      "epoch": 1.1216730038022813,
+      "grad_norm": 0.6616679865479671,
+      "learning_rate": 9.634779769423412e-06,
+      "loss": 0.3007,
+      "step": 590
+    },
+    {
+      "epoch": 1.123574144486692,
+      "grad_norm": 0.6452783276901434,
+      "learning_rate": 9.601594775263073e-06,
+      "loss": 0.3186,
+      "step": 591
+    },
+    {
+      "epoch": 1.1254752851711027,
+      "grad_norm": 0.6782290326372009,
+      "learning_rate": 9.568414174921085e-06,
+      "loss": 0.3178,
+      "step": 592
+    },
+    {
+      "epoch": 1.1273764258555132,
+      "grad_norm": 0.7252900976075151,
+      "learning_rate": 9.535238334330234e-06,
+      "loss": 0.3117,
+      "step": 593
+    },
+    {
+      "epoch": 1.129277566539924,
+      "grad_norm": 0.711413459213573,
+      "learning_rate": 9.502067619370794e-06,
+      "loss": 0.315,
+      "step": 594
+    },
+    {
+      "epoch": 1.1311787072243347,
+      "grad_norm": 0.672151400243723,
+      "learning_rate": 9.468902395866532e-06,
+      "loss": 0.3072,
+      "step": 595
+    },
+    {
+      "epoch": 1.1330798479087452,
+      "grad_norm": 0.6721618389686488,
+      "learning_rate": 9.435743029580638e-06,
+      "loss": 0.3148,
+      "step": 596
+    },
+    {
+      "epoch": 1.1349809885931559,
+      "grad_norm": 0.6543075461895181,
+      "learning_rate": 9.402589886211711e-06,
+      "loss": 0.2996,
+      "step": 597
+    },
+    {
+      "epoch": 1.1368821292775666,
+      "grad_norm": 0.6798179973484861,
+      "learning_rate": 9.369443331389718e-06,
+      "loss": 0.3187,
+      "step": 598
+    },
+    {
+      "epoch": 1.138783269961977,
+      "grad_norm": 0.7453382816367166,
+      "learning_rate": 9.336303730671968e-06,
+      "loss": 0.3231,
+      "step": 599
+    },
+    {
+      "epoch": 1.1406844106463878,
+      "grad_norm": 0.6868238577210452,
+      "learning_rate": 9.303171449539074e-06,
+      "loss": 0.3123,
+      "step": 600
+    },
+    {
+      "epoch": 1.1425855513307985,
+      "grad_norm": 0.6841617373171418,
+      "learning_rate": 9.270046853390924e-06,
+      "loss": 0.2912,
+      "step": 601
+    },
+    {
+      "epoch": 1.144486692015209,
+      "grad_norm": 0.6514780469266361,
+      "learning_rate": 9.236930307542654e-06,
+      "loss": 0.3099,
+      "step": 602
+    },
+    {
+      "epoch": 1.1463878326996197,
+      "grad_norm": 0.624250938754594,
+      "learning_rate": 9.203822177220621e-06,
+      "loss": 0.2999,
+      "step": 603
+    },
+    {
+      "epoch": 1.1482889733840305,
+      "grad_norm": 0.6988345580587945,
+      "learning_rate": 9.170722827558357e-06,
+      "loss": 0.3342,
+      "step": 604
+    },
+    {
+      "epoch": 1.1501901140684412,
+      "grad_norm": 0.6763049377710139,
+      "learning_rate": 9.13763262359257e-06,
+      "loss": 0.3242,
+      "step": 605
+    },
+    {
+      "epoch": 1.1520912547528517,
+      "grad_norm": 0.684379695797014,
+      "learning_rate": 9.104551930259101e-06,
+      "loss": 0.3166,
+      "step": 606
+    },
+    {
+      "epoch": 1.1539923954372624,
+      "grad_norm": 0.7527930820893866,
+      "learning_rate": 9.071481112388905e-06,
+      "loss": 0.3399,
+      "step": 607
+    },
+    {
+      "epoch": 1.1558935361216731,
+      "grad_norm": 0.6489935180973732,
+      "learning_rate": 9.038420534704015e-06,
+      "loss": 0.3114,
+      "step": 608
+    },
+    {
+      "epoch": 1.1577946768060836,
+      "grad_norm": 0.6285217266384527,
+      "learning_rate": 9.005370561813545e-06,
+      "loss": 0.3208,
+      "step": 609
+    },
+    {
+      "epoch": 1.1596958174904943,
+      "grad_norm": 0.6700926089542495,
+      "learning_rate": 8.972331558209644e-06,
+      "loss": 0.3316,
+      "step": 610
+    },
+    {
+      "epoch": 1.161596958174905,
+      "grad_norm": 0.6365367179122374,
+      "learning_rate": 8.939303888263485e-06,
+      "loss": 0.3089,
+      "step": 611
+    },
+    {
+      "epoch": 1.1634980988593155,
+      "grad_norm": 0.6616789239212519,
+      "learning_rate": 8.906287916221259e-06,
+      "loss": 0.3108,
+      "step": 612
+    },
+    {
+      "epoch": 1.1653992395437263,
+      "grad_norm": 0.6717428893123029,
+      "learning_rate": 8.873284006200129e-06,
+      "loss": 0.3231,
+      "step": 613
+    },
+    {
+      "epoch": 1.167300380228137,
+      "grad_norm": 0.6799875382381467,
+      "learning_rate": 8.840292522184247e-06,
+      "loss": 0.2954,
+      "step": 614
+    },
+    {
+      "epoch": 1.1692015209125475,
+      "grad_norm": 0.6364378935645879,
+      "learning_rate": 8.807313828020715e-06,
+      "loss": 0.3258,
+      "step": 615
+    },
+    {
+      "epoch": 1.1711026615969582,
+      "grad_norm": 0.6855167383729526,
+      "learning_rate": 8.774348287415589e-06,
+      "loss": 0.3198,
+      "step": 616
+    },
+    {
+      "epoch": 1.173003802281369,
+      "grad_norm": 0.6932944613081954,
+      "learning_rate": 8.74139626392986e-06,
+      "loss": 0.3353,
+      "step": 617
+    },
+    {
+      "epoch": 1.1749049429657794,
+      "grad_norm": 0.6624591138934531,
+      "learning_rate": 8.708458120975436e-06,
+      "loss": 0.323,
+      "step": 618
+    },
+    {
+      "epoch": 1.1768060836501901,
+      "grad_norm": 0.6253694801540307,
+      "learning_rate": 8.675534221811156e-06,
+      "loss": 0.3011,
+      "step": 619
+    },
+    {
+      "epoch": 1.1787072243346008,
+      "grad_norm": 0.722519030738269,
+      "learning_rate": 8.64262492953876e-06,
+      "loss": 0.3149,
+      "step": 620
+    },
+    {
+      "epoch": 1.1806083650190113,
+      "grad_norm": 0.6832002410610039,
+      "learning_rate": 8.60973060709891e-06,
+      "loss": 0.3412,
+      "step": 621
+    },
+    {
+      "epoch": 1.182509505703422,
+      "grad_norm": 0.6419777945134574,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.3186,
+      "step": 622
+    },
+    {
+      "epoch": 1.1844106463878328,
+      "grad_norm": 0.7187627041714545,
+      "learning_rate": 8.543988322649954e-06,
+      "loss": 0.3291,
+      "step": 623
+    },
+    {
+      "epoch": 1.1863117870722433,
+      "grad_norm": 0.7392888649823336,
+      "learning_rate": 8.511141085680684e-06,
+      "loss": 0.3268,
+      "step": 624
+    },
+    {
+      "epoch": 1.188212927756654,
+      "grad_norm": 0.6667086104970468,
+      "learning_rate": 8.478310268615612e-06,
+      "loss": 0.3115,
+      "step": 625
+    },
+    {
+      "epoch": 1.1901140684410647,
+      "grad_norm": 0.6876961051295815,
+      "learning_rate": 8.445496233529934e-06,
+      "loss": 0.314,
+      "step": 626
+    },
+    {
+      "epoch": 1.1920152091254752,
+      "grad_norm": 0.6871762260376846,
+      "learning_rate": 8.41269934231375e-06,
+      "loss": 0.3218,
+      "step": 627
+    },
+    {
+      "epoch": 1.193916349809886,
+      "grad_norm": 0.8400217726179474,
+      "learning_rate": 8.3799199566681e-06,
+      "loss": 0.2992,
+      "step": 628
+    },
+    {
+      "epoch": 1.1958174904942966,
+      "grad_norm": 0.6360889255308926,
+      "learning_rate": 8.34715843810096e-06,
+      "loss": 0.2903,
+      "step": 629
+    },
+    {
+      "epoch": 1.1977186311787071,
+      "grad_norm": 0.6612295437520829,
+      "learning_rate": 8.314415147923254e-06,
+      "loss": 0.3223,
+      "step": 630
+    },
+    {
+      "epoch": 1.1996197718631179,
+      "grad_norm": 0.6608796692172322,
+      "learning_rate": 8.281690447244887e-06,
+      "loss": 0.3069,
+      "step": 631
+    },
+    {
+      "epoch": 1.2015209125475286,
+      "grad_norm": 0.7305042503329781,
+      "learning_rate": 8.248984696970732e-06,
+      "loss": 0.3156,
+      "step": 632
+    },
+    {
+      "epoch": 1.203422053231939,
+      "grad_norm": 0.6988199036881295,
+      "learning_rate": 8.216298257796677e-06,
+      "loss": 0.353,
+      "step": 633
+    },
+    {
+      "epoch": 1.2053231939163498,
+      "grad_norm": 0.6296362793319676,
+      "learning_rate": 8.183631490205636e-06,
+      "loss": 0.299,
+      "step": 634
+    },
+    {
+      "epoch": 1.2072243346007605,
+      "grad_norm": 0.7056580077842795,
+      "learning_rate": 8.150984754463578e-06,
+      "loss": 0.3166,
+      "step": 635
+    },
+    {
+      "epoch": 1.209125475285171,
+      "grad_norm": 0.6612887193447069,
+      "learning_rate": 8.118358410615545e-06,
+      "loss": 0.3036,
+      "step": 636
+    },
+    {
+      "epoch": 1.2110266159695817,
+      "grad_norm": 0.6449677821210904,
+      "learning_rate": 8.08575281848169e-06,
+      "loss": 0.2994,
+      "step": 637
+    },
+    {
+      "epoch": 1.2129277566539924,
+      "grad_norm": 0.6363722001533265,
+      "learning_rate": 8.0531683376533e-06,
+      "loss": 0.2981,
+      "step": 638
+    },
+    {
+      "epoch": 1.214828897338403,
+      "grad_norm": 0.6885008571525498,
+      "learning_rate": 8.020605327488846e-06,
+      "loss": 0.3214,
+      "step": 639
+    },
+    {
+      "epoch": 1.2167300380228137,
+      "grad_norm": 0.7361757448214602,
+      "learning_rate": 7.988064147110001e-06,
+      "loss": 0.3196,
+      "step": 640
+    },
+    {
+      "epoch": 1.2186311787072244,
+      "grad_norm": 0.6670610092945182,
+      "learning_rate": 7.955545155397684e-06,
+      "loss": 0.3157,
+      "step": 641
+    },
+    {
+      "epoch": 1.2205323193916349,
+      "grad_norm": 0.6765129520484091,
+      "learning_rate": 7.923048710988119e-06,
+      "loss": 0.3206,
+      "step": 642
+    },
+    {
+      "epoch": 1.2224334600760456,
+      "grad_norm": 0.7018350317512431,
+      "learning_rate": 7.890575172268858e-06,
+      "loss": 0.3201,
+      "step": 643
+    },
+    {
+      "epoch": 1.2243346007604563,
+      "grad_norm": 0.7051466428050569,
+      "learning_rate": 7.858124897374837e-06,
+      "loss": 0.298,
+      "step": 644
+    },
+    {
+      "epoch": 1.2262357414448668,
+      "grad_norm": 0.6571005783684197,
+      "learning_rate": 7.825698244184432e-06,
+      "loss": 0.2938,
+      "step": 645
+    },
+    {
+      "epoch": 1.2281368821292775,
+      "grad_norm": 0.7224097996130295,
+      "learning_rate": 7.7932955703155e-06,
+      "loss": 0.3109,
+      "step": 646
+    },
+    {
+      "epoch": 1.2300380228136882,
+      "grad_norm": 0.6896490816288485,
+      "learning_rate": 7.760917233121443e-06,
+      "loss": 0.3124,
+      "step": 647
+    },
+    {
+      "epoch": 1.231939163498099,
+      "grad_norm": 0.6915381668799482,
+      "learning_rate": 7.728563589687275e-06,
+      "loss": 0.3109,
+      "step": 648
+    },
+    {
+      "epoch": 1.2338403041825095,
+      "grad_norm": 0.7082172869434794,
+      "learning_rate": 7.696234996825663e-06,
+      "loss": 0.3266,
+      "step": 649
+    },
+    {
+      "epoch": 1.2357414448669202,
+      "grad_norm": 0.6363462099911186,
+      "learning_rate": 7.663931811073003e-06,
+      "loss": 0.3028,
+      "step": 650
+    },
+    {
+      "epoch": 1.2376425855513309,
+      "grad_norm": 0.6771335464107202,
+      "learning_rate": 7.631654388685496e-06,
+      "loss": 0.3148,
+      "step": 651
+    },
+    {
+      "epoch": 1.2395437262357414,
+      "grad_norm": 0.6660873577742755,
+      "learning_rate": 7.599403085635208e-06,
+      "loss": 0.3215,
+      "step": 652
+    },
+    {
+      "epoch": 1.241444866920152,
+      "grad_norm": 0.6968778941139329,
+      "learning_rate": 7.567178257606147e-06,
+      "loss": 0.3282,
+      "step": 653
+    },
+    {
+      "epoch": 1.2433460076045628,
+      "grad_norm": 0.6396797551782774,
+      "learning_rate": 7.534980259990341e-06,
+      "loss": 0.3129,
+      "step": 654
+    },
+    {
+      "epoch": 1.2452471482889733,
+      "grad_norm": 0.6557078854815676,
+      "learning_rate": 7.50280944788392e-06,
+      "loss": 0.2975,
+      "step": 655
+    },
+    {
+      "epoch": 1.247148288973384,
+      "grad_norm": 0.6659623091544976,
+      "learning_rate": 7.470666176083193e-06,
+      "loss": 0.3259,
+      "step": 656
+    },
+    {
+      "epoch": 1.2490494296577948,
+      "grad_norm": 0.6727568656455613,
+      "learning_rate": 7.438550799080746e-06,
+      "loss": 0.324,
+      "step": 657
+    },
+    {
+      "epoch": 1.2509505703422052,
+      "grad_norm": 0.6305321930339485,
+      "learning_rate": 7.40646367106153e-06,
+      "loss": 0.3075,
+      "step": 658
+    },
+    {
+      "epoch": 1.252851711026616,
+      "grad_norm": 0.6544733113550633,
+      "learning_rate": 7.3744051458989395e-06,
+      "loss": 0.3096,
+      "step": 659
+    },
+    {
+      "epoch": 1.2547528517110267,
+      "grad_norm": 0.672100209311043,
+      "learning_rate": 7.342375577150928e-06,
+      "loss": 0.3223,
+      "step": 660
+    },
+    {
+      "epoch": 1.2566539923954372,
+      "grad_norm": 0.6439040695228999,
+      "learning_rate": 7.310375318056107e-06,
+      "loss": 0.3122,
+      "step": 661
+    },
+    {
+      "epoch": 1.258555133079848,
+      "grad_norm": 0.6792687497106056,
+      "learning_rate": 7.278404721529843e-06,
+      "loss": 0.3268,
+      "step": 662
+    },
+    {
+      "epoch": 1.2604562737642586,
+      "grad_norm": 0.6378118432200549,
+      "learning_rate": 7.246464140160365e-06,
+      "loss": 0.3168,
+      "step": 663
+    },
+    {
+      "epoch": 1.2623574144486693,
+      "grad_norm": 0.6538418374771207,
+      "learning_rate": 7.214553926204884e-06,
+      "loss": 0.3096,
+      "step": 664
+    },
+    {
+      "epoch": 1.2642585551330798,
+      "grad_norm": 0.6760210885621455,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.3083,
+      "step": 665
+    },
+    {
+      "epoch": 1.2661596958174905,
+      "grad_norm": 0.6678587307530544,
+      "learning_rate": 7.150826007886334e-06,
+      "loss": 0.3027,
+      "step": 666
+    },
+    {
+      "epoch": 1.2680608365019013,
+      "grad_norm": 0.6422862161312628,
+      "learning_rate": 7.119009006347625e-06,
+      "loss": 0.2957,
+      "step": 667
+    },
+    {
+      "epoch": 1.2699619771863118,
+      "grad_norm": 0.6896312300716236,
+      "learning_rate": 7.087223777863883e-06,
+      "loss": 0.3037,
+      "step": 668
+    },
+    {
+      "epoch": 1.2718631178707225,
+      "grad_norm": 0.7109438931788244,
+      "learning_rate": 7.055470672979003e-06,
+      "loss": 0.3372,
+      "step": 669
+    },
+    {
+      "epoch": 1.2737642585551332,
+      "grad_norm": 0.6444023137626911,
+      "learning_rate": 7.023750041882609e-06,
+      "loss": 0.3129,
+      "step": 670
+    },
+    {
+      "epoch": 1.2756653992395437,
+      "grad_norm": 0.6246780836436571,
+      "learning_rate": 6.992062234406185e-06,
+      "loss": 0.3139,
+      "step": 671
+    },
+    {
+      "epoch": 1.2775665399239544,
+      "grad_norm": 0.6678579717480345,
+      "learning_rate": 6.960407600019217e-06,
+      "loss": 0.3029,
+      "step": 672
+    },
+    {
+      "epoch": 1.2794676806083651,
+      "grad_norm": 0.6481174375582444,
+      "learning_rate": 6.9287864878253475e-06,
+      "loss": 0.3101,
+      "step": 673
+    },
+    {
+      "epoch": 1.2813688212927756,
+      "grad_norm": 0.6656343549935969,
+      "learning_rate": 6.897199246558515e-06,
+      "loss": 0.319,
+      "step": 674
+    },
+    {
+      "epoch": 1.2832699619771863,
+      "grad_norm": 0.6679247636852839,
+      "learning_rate": 6.865646224579108e-06,
+      "loss": 0.3018,
+      "step": 675
+    },
+    {
+      "epoch": 1.285171102661597,
+      "grad_norm": 0.674469356966251,
+      "learning_rate": 6.834127769870134e-06,
+      "loss": 0.3032,
+      "step": 676
+    },
+    {
+      "epoch": 1.2870722433460076,
+      "grad_norm": 0.6496806296522415,
+      "learning_rate": 6.802644230033373e-06,
+      "loss": 0.3182,
+      "step": 677
+    },
+    {
+      "epoch": 1.2889733840304183,
+      "grad_norm": 0.6549615072963446,
+      "learning_rate": 6.771195952285541e-06,
+      "loss": 0.2922,
+      "step": 678
+    },
+    {
+      "epoch": 1.290874524714829,
+      "grad_norm": 0.6795885818582195,
+      "learning_rate": 6.739783283454469e-06,
+      "loss": 0.3178,
+      "step": 679
+    },
+    {
+      "epoch": 1.2927756653992395,
+      "grad_norm": 0.6734496926319783,
+      "learning_rate": 6.708406569975274e-06,
+      "loss": 0.2885,
+      "step": 680
+    },
+    {
+      "epoch": 1.2946768060836502,
+      "grad_norm": 0.6450335445471536,
+      "learning_rate": 6.6770661578865444e-06,
+      "loss": 0.3033,
+      "step": 681
+    },
+    {
+      "epoch": 1.296577946768061,
+      "grad_norm": 0.6363337188021119,
+      "learning_rate": 6.645762392826509e-06,
+      "loss": 0.3069,
+      "step": 682
+    },
+    {
+      "epoch": 1.2984790874524714,
+      "grad_norm": 0.7021565044352918,
+      "learning_rate": 6.614495620029238e-06,
+      "loss": 0.3262,
+      "step": 683
+    },
+    {
+      "epoch": 1.3003802281368821,
+      "grad_norm": 0.6880116306185828,
+      "learning_rate": 6.583266184320836e-06,
+      "loss": 0.3055,
+      "step": 684
+    },
+    {
+      "epoch": 1.3022813688212929,
+      "grad_norm": 0.6922660095928748,
+      "learning_rate": 6.552074430115624e-06,
+      "loss": 0.3298,
+      "step": 685
+    },
+    {
+      "epoch": 1.3041825095057034,
+      "grad_norm": 0.6613208612275078,
+      "learning_rate": 6.520920701412371e-06,
+      "loss": 0.2927,
+      "step": 686
+    },
+    {
+      "epoch": 1.306083650190114,
+      "grad_norm": 0.7127351298157687,
+      "learning_rate": 6.489805341790456e-06,
+      "loss": 0.3127,
+      "step": 687
+    },
+    {
+      "epoch": 1.3079847908745248,
+      "grad_norm": 0.6701809464526972,
+      "learning_rate": 6.458728694406124e-06,
+      "loss": 0.3109,
+      "step": 688
+    },
+    {
+      "epoch": 1.3098859315589353,
+      "grad_norm": 0.6346020268748015,
+      "learning_rate": 6.427691101988673e-06,
+      "loss": 0.3074,
+      "step": 689
+    },
+    {
+      "epoch": 1.311787072243346,
+      "grad_norm": 0.6950277468526841,
+      "learning_rate": 6.396692906836686e-06,
+      "loss": 0.3278,
+      "step": 690
+    },
+    {
+      "epoch": 1.3136882129277567,
+      "grad_norm": 0.6955208821483757,
+      "learning_rate": 6.3657344508142495e-06,
+      "loss": 0.2921,
+      "step": 691
+    },
+    {
+      "epoch": 1.3155893536121672,
+      "grad_norm": 0.6382243153101852,
+      "learning_rate": 6.334816075347185e-06,
+      "loss": 0.2945,
+      "step": 692
+    },
+    {
+      "epoch": 1.317490494296578,
+      "grad_norm": 0.6687590976661751,
+      "learning_rate": 6.303938121419295e-06,
+      "loss": 0.2978,
+      "step": 693
+    },
+    {
+      "epoch": 1.3193916349809887,
+      "grad_norm": 0.6803713169337828,
+      "learning_rate": 6.273100929568579e-06,
+      "loss": 0.3107,
+      "step": 694
+    },
+    {
+      "epoch": 1.3212927756653992,
+      "grad_norm": 0.6834322733945087,
+      "learning_rate": 6.242304839883502e-06,
+      "loss": 0.307,
+      "step": 695
+    },
+    {
+      "epoch": 1.3231939163498099,
+      "grad_norm": 0.6657659090293411,
+      "learning_rate": 6.211550191999223e-06,
+      "loss": 0.3239,
+      "step": 696
+    },
+    {
+      "epoch": 1.3250950570342206,
+      "grad_norm": 0.703797287794531,
+      "learning_rate": 6.18083732509387e-06,
+      "loss": 0.3362,
+      "step": 697
+    },
+    {
+      "epoch": 1.326996197718631,
+      "grad_norm": 0.6380256397022588,
+      "learning_rate": 6.150166577884781e-06,
+      "loss": 0.3038,
+      "step": 698
+    },
+    {
+      "epoch": 1.3288973384030418,
+      "grad_norm": 0.7196606333114327,
+      "learning_rate": 6.119538288624778e-06,
+      "loss": 0.2978,
+      "step": 699
+    },
+    {
+      "epoch": 1.3307984790874525,
+      "grad_norm": 0.662014340169013,
+      "learning_rate": 6.088952795098442e-06,
+      "loss": 0.3012,
+      "step": 700
+    },
+    {
+      "epoch": 1.332699619771863,
+      "grad_norm": 0.6854618045958362,
+      "learning_rate": 6.058410434618367e-06,
+      "loss": 0.3181,
+      "step": 701
+    },
+    {
+      "epoch": 1.3346007604562737,
+      "grad_norm": 0.6681438346121338,
+      "learning_rate": 6.027911544021465e-06,
+      "loss": 0.3084,
+      "step": 702
+    },
+    {
+      "epoch": 1.3365019011406845,
+      "grad_norm": 0.6346348094473896,
+      "learning_rate": 5.997456459665237e-06,
+      "loss": 0.3106,
+      "step": 703
+    },
+    {
+      "epoch": 1.338403041825095,
+      "grad_norm": 0.6401936190152808,
+      "learning_rate": 5.967045517424062e-06,
+      "loss": 0.3112,
+      "step": 704
+    },
+    {
+      "epoch": 1.3403041825095057,
+      "grad_norm": 0.7748108248822071,
+      "learning_rate": 5.936679052685505e-06,
+      "loss": 0.3299,
+      "step": 705
+    },
+    {
+      "epoch": 1.3422053231939164,
+      "grad_norm": 0.7381067950191753,
+      "learning_rate": 5.906357400346596e-06,
+      "loss": 0.315,
+      "step": 706
+    },
+    {
+      "epoch": 1.3441064638783269,
+      "grad_norm": 0.6528483457259935,
+      "learning_rate": 5.876080894810167e-06,
+      "loss": 0.3081,
+      "step": 707
+    },
+    {
+      "epoch": 1.3460076045627376,
+      "grad_norm": 0.6705863066546487,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.304,
+      "step": 708
+    },
+    {
+      "epoch": 1.3479087452471483,
+      "grad_norm": 0.6813816649941541,
+      "learning_rate": 5.815664659262845e-06,
+      "loss": 0.3245,
+      "step": 709
+    },
+    {
+      "epoch": 1.3498098859315588,
+      "grad_norm": 0.6875874980288026,
+      "learning_rate": 5.78552559555337e-06,
+      "loss": 0.3139,
+      "step": 710
+    },
+    {
+      "epoch": 1.3517110266159695,
+      "grad_norm": 0.6461631644329731,
+      "learning_rate": 5.755433011241851e-06,
+      "loss": 0.3078,
+      "step": 711
+    },
+    {
+      "epoch": 1.3536121673003803,
+      "grad_norm": 0.6443287096213095,
+      "learning_rate": 5.725387238204831e-06,
+      "loss": 0.3209,
+      "step": 712
+    },
+    {
+      "epoch": 1.3555133079847907,
+      "grad_norm": 0.6839337699747792,
+      "learning_rate": 5.695388607802603e-06,
+      "loss": 0.3192,
+      "step": 713
+    },
+    {
+      "epoch": 1.3574144486692015,
+      "grad_norm": 0.638925973544638,
+      "learning_rate": 5.665437450875534e-06,
+      "loss": 0.3176,
+      "step": 714
+    },
+    {
+      "epoch": 1.3593155893536122,
+      "grad_norm": 0.6639190902228325,
+      "learning_rate": 5.635534097740435e-06,
+      "loss": 0.3107,
+      "step": 715
+    },
+    {
+      "epoch": 1.3612167300380227,
+      "grad_norm": 0.6393560958010144,
+      "learning_rate": 5.605678878186911e-06,
+      "loss": 0.2958,
+      "step": 716
+    },
+    {
+      "epoch": 1.3631178707224334,
+      "grad_norm": 0.6285708773882102,
+      "learning_rate": 5.575872121473722e-06,
+      "loss": 0.2888,
+      "step": 717
+    },
+    {
+      "epoch": 1.3650190114068441,
+      "grad_norm": 0.6846241408983079,
+      "learning_rate": 5.546114156325166e-06,
+      "loss": 0.31,
+      "step": 718
+    },
+    {
+      "epoch": 1.3669201520912546,
+      "grad_norm": 0.6871887033935491,
+      "learning_rate": 5.516405310927431e-06,
+      "loss": 0.307,
+      "step": 719
+    },
+    {
+      "epoch": 1.3688212927756653,
+      "grad_norm": 0.6627953656618311,
+      "learning_rate": 5.4867459129249846e-06,
+      "loss": 0.3036,
+      "step": 720
+    },
+    {
+      "epoch": 1.370722433460076,
+      "grad_norm": 0.6826781750051257,
+      "learning_rate": 5.4571362894169795e-06,
+      "loss": 0.3203,
+      "step": 721
+    },
+    {
+      "epoch": 1.3726235741444868,
+      "grad_norm": 0.6749204384541736,
+      "learning_rate": 5.427576766953615e-06,
+      "loss": 0.3092,
+      "step": 722
+    },
+    {
+      "epoch": 1.3745247148288973,
+      "grad_norm": 0.645977062094342,
+      "learning_rate": 5.398067671532554e-06,
+      "loss": 0.3059,
+      "step": 723
+    },
+    {
+      "epoch": 1.376425855513308,
+      "grad_norm": 0.6441840584986329,
+      "learning_rate": 5.368609328595323e-06,
+      "loss": 0.3006,
+      "step": 724
+    },
+    {
+      "epoch": 1.3783269961977187,
+      "grad_norm": 0.6553415796476495,
+      "learning_rate": 5.339202063023727e-06,
+      "loss": 0.3004,
+      "step": 725
+    },
+    {
+      "epoch": 1.3802281368821292,
+      "grad_norm": 0.6679615430077358,
+      "learning_rate": 5.309846199136258e-06,
+      "loss": 0.3021,
+      "step": 726
+    },
+    {
+      "epoch": 1.38212927756654,
+      "grad_norm": 0.6326190581538851,
+      "learning_rate": 5.280542060684535e-06,
+      "loss": 0.3125,
+      "step": 727
+    },
+    {
+      "epoch": 1.3840304182509506,
+      "grad_norm": 0.6185028307269258,
+      "learning_rate": 5.2512899708497086e-06,
+      "loss": 0.2983,
+      "step": 728
+    },
+    {
+      "epoch": 1.3859315589353614,
+      "grad_norm": 0.6325976864827284,
+      "learning_rate": 5.222090252238916e-06,
+      "loss": 0.3041,
+      "step": 729
+    },
+    {
+      "epoch": 1.3878326996197718,
+      "grad_norm": 0.654860537012981,
+      "learning_rate": 5.192943226881724e-06,
+      "loss": 0.2799,
+      "step": 730
+    },
+    {
+      "epoch": 1.3897338403041826,
+      "grad_norm": 0.714215672954267,
+      "learning_rate": 5.163849216226562e-06,
+      "loss": 0.3084,
+      "step": 731
+    },
+    {
+      "epoch": 1.3916349809885933,
+      "grad_norm": 0.6843725415850207,
+      "learning_rate": 5.134808541137183e-06,
+      "loss": 0.3246,
+      "step": 732
+    },
+    {
+      "epoch": 1.3935361216730038,
+      "grad_norm": 0.6440204443738418,
+      "learning_rate": 5.105821521889147e-06,
+      "loss": 0.3068,
+      "step": 733
+    },
+    {
+      "epoch": 1.3954372623574145,
+      "grad_norm": 0.6611952654303866,
+      "learning_rate": 5.076888478166247e-06,
+      "loss": 0.3099,
+      "step": 734
+    },
+    {
+      "epoch": 1.3973384030418252,
+      "grad_norm": 0.6980582246375392,
+      "learning_rate": 5.048009729057012e-06,
+      "loss": 0.315,
+      "step": 735
+    },
+    {
+      "epoch": 1.3992395437262357,
+      "grad_norm": 0.6758047357097503,
+      "learning_rate": 5.0191855930511946e-06,
+      "loss": 0.3073,
+      "step": 736
+    },
+    {
+      "epoch": 1.4011406844106464,
+      "grad_norm": 0.6600856679077798,
+      "learning_rate": 4.990416388036233e-06,
+      "loss": 0.2998,
+      "step": 737
+    },
+    {
+      "epoch": 1.4030418250950571,
+      "grad_norm": 0.6474504417334683,
+      "learning_rate": 4.961702431293759e-06,
+      "loss": 0.3093,
+      "step": 738
+    },
+    {
+      "epoch": 1.4049429657794676,
+      "grad_norm": 0.6910282970602829,
+      "learning_rate": 4.933044039496107e-06,
+      "loss": 0.3096,
+      "step": 739
+    },
+    {
+      "epoch": 1.4068441064638784,
+      "grad_norm": 0.6680073344886641,
+      "learning_rate": 4.904441528702806e-06,
+      "loss": 0.3054,
+      "step": 740
+    },
+    {
+      "epoch": 1.408745247148289,
+      "grad_norm": 0.6803066629541039,
+      "learning_rate": 4.875895214357093e-06,
+      "loss": 0.3038,
+      "step": 741
+    },
+    {
+      "epoch": 1.4106463878326996,
+      "grad_norm": 0.6972877482755048,
+      "learning_rate": 4.847405411282462e-06,
+      "loss": 0.3285,
+      "step": 742
+    },
+    {
+      "epoch": 1.4125475285171103,
+      "grad_norm": 0.6913557595475017,
+      "learning_rate": 4.818972433679145e-06,
+      "loss": 0.3158,
+      "step": 743
+    },
+    {
+      "epoch": 1.414448669201521,
+      "grad_norm": 0.6851369751310428,
+      "learning_rate": 4.790596595120699e-06,
+      "loss": 0.3021,
+      "step": 744
+    },
+    {
+      "epoch": 1.4163498098859315,
+      "grad_norm": 0.6165745810507383,
+      "learning_rate": 4.762278208550505e-06,
+      "loss": 0.2815,
+      "step": 745
+    },
+    {
+      "epoch": 1.4182509505703422,
+      "grad_norm": 0.6783799631069839,
+      "learning_rate": 4.734017586278337e-06,
+      "loss": 0.2885,
+      "step": 746
+    },
+    {
+      "epoch": 1.420152091254753,
+      "grad_norm": 0.6820450791363494,
+      "learning_rate": 4.7058150399769245e-06,
+      "loss": 0.3145,
+      "step": 747
+    },
+    {
+      "epoch": 1.4220532319391634,
+      "grad_norm": 0.6698778643156899,
+      "learning_rate": 4.677670880678493e-06,
+      "loss": 0.3123,
+      "step": 748
+    },
+    {
+      "epoch": 1.4239543726235742,
+      "grad_norm": 0.6462806248452764,
+      "learning_rate": 4.649585418771348e-06,
+      "loss": 0.3021,
+      "step": 749
+    },
+    {
+      "epoch": 1.4258555133079849,
+      "grad_norm": 0.7087874573250521,
+      "learning_rate": 4.621558963996458e-06,
+      "loss": 0.3229,
+      "step": 750
+    },
+    {
+      "epoch": 1.4277566539923954,
+      "grad_norm": 0.6792498034649502,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.3017,
+      "step": 751
+    },
+    {
+      "epoch": 1.429657794676806,
+      "grad_norm": 0.6810823112500632,
+      "learning_rate": 4.565684311550077e-06,
+      "loss": 0.3131,
+      "step": 752
+    },
+    {
+      "epoch": 1.4315589353612168,
+      "grad_norm": 0.6669654687874899,
+      "learning_rate": 4.537836730093077e-06,
+      "loss": 0.3078,
+      "step": 753
+    },
+    {
+      "epoch": 1.4334600760456273,
+      "grad_norm": 0.6713378438409738,
+      "learning_rate": 4.510049388190518e-06,
+      "loss": 0.3141,
+      "step": 754
+    },
+    {
+      "epoch": 1.435361216730038,
+      "grad_norm": 0.6528802550479103,
+      "learning_rate": 4.482322592295541e-06,
+      "loss": 0.3024,
+      "step": 755
+    },
+    {
+      "epoch": 1.4372623574144487,
+      "grad_norm": 0.6333600565991427,
+      "learning_rate": 4.454656648193559e-06,
+      "loss": 0.2988,
+      "step": 756
+    },
+    {
+      "epoch": 1.4391634980988592,
+      "grad_norm": 0.6423623842982019,
+      "learning_rate": 4.427051860998877e-06,
+      "loss": 0.299,
+      "step": 757
+    },
+    {
+      "epoch": 1.44106463878327,
+      "grad_norm": 0.6449862805456914,
+      "learning_rate": 4.399508535151321e-06,
+      "loss": 0.2947,
+      "step": 758
+    },
+    {
+      "epoch": 1.4429657794676807,
+      "grad_norm": 0.6280695923490843,
+      "learning_rate": 4.372026974412907e-06,
+      "loss": 0.2833,
+      "step": 759
+    },
+    {
+      "epoch": 1.4448669201520912,
+      "grad_norm": 0.6585594745696902,
+      "learning_rate": 4.344607481864466e-06,
+      "loss": 0.2968,
+      "step": 760
+    },
+    {
+      "epoch": 1.446768060836502,
+      "grad_norm": 0.6880160224989331,
+      "learning_rate": 4.317250359902295e-06,
+      "loss": 0.3029,
+      "step": 761
+    },
+    {
+      "epoch": 1.4486692015209126,
+      "grad_norm": 0.6490095062795858,
+      "learning_rate": 4.2899559102348585e-06,
+      "loss": 0.3106,
+      "step": 762
+    },
+    {
+      "epoch": 1.450570342205323,
+      "grad_norm": 0.6817801030896319,
+      "learning_rate": 4.262724433879427e-06,
+      "loss": 0.3155,
+      "step": 763
+    },
+    {
+      "epoch": 1.4524714828897338,
+      "grad_norm": 0.6830292298795404,
+      "learning_rate": 4.235556231158765e-06,
+      "loss": 0.3039,
+      "step": 764
+    },
+    {
+      "epoch": 1.4543726235741445,
+      "grad_norm": 0.6566950334951492,
+      "learning_rate": 4.208451601697836e-06,
+      "loss": 0.3024,
+      "step": 765
+    },
+    {
+      "epoch": 1.456273764258555,
+      "grad_norm": 0.6471057927294712,
+      "learning_rate": 4.181410844420473e-06,
+      "loss": 0.3012,
+      "step": 766
+    },
+    {
+      "epoch": 1.4581749049429658,
+      "grad_norm": 0.6570402493922917,
+      "learning_rate": 4.154434257546095e-06,
+      "loss": 0.2826,
+      "step": 767
+    },
+    {
+      "epoch": 1.4600760456273765,
+      "grad_norm": 0.6067513944982301,
+      "learning_rate": 4.127522138586424e-06,
+      "loss": 0.2988,
+      "step": 768
+    },
+    {
+      "epoch": 1.461977186311787,
+      "grad_norm": 0.6540846669316159,
+      "learning_rate": 4.10067478434219e-06,
+      "loss": 0.3004,
+      "step": 769
+    },
+    {
+      "epoch": 1.4638783269961977,
+      "grad_norm": 0.6213457055665008,
+      "learning_rate": 4.073892490899865e-06,
+      "loss": 0.3011,
+      "step": 770
+    },
+    {
+      "epoch": 1.4657794676806084,
+      "grad_norm": 0.6476039996775701,
+      "learning_rate": 4.047175553628397e-06,
+      "loss": 0.3,
+      "step": 771
+    },
+    {
+      "epoch": 1.467680608365019,
+      "grad_norm": 0.6490219071418775,
+      "learning_rate": 4.020524267175954e-06,
+      "loss": 0.2939,
+      "step": 772
+    },
+    {
+      "epoch": 1.4695817490494296,
+      "grad_norm": 0.6337227292778735,
+      "learning_rate": 3.993938925466674e-06,
+      "loss": 0.2941,
+      "step": 773
+    },
+    {
+      "epoch": 1.4714828897338403,
+      "grad_norm": 0.6567469479657247,
+      "learning_rate": 3.96741982169742e-06,
+      "loss": 0.3084,
+      "step": 774
+    },
+    {
+      "epoch": 1.4733840304182508,
+      "grad_norm": 0.6769301402918131,
+      "learning_rate": 3.9409672483345465e-06,
+      "loss": 0.3013,
+      "step": 775
+    },
+    {
+      "epoch": 1.4752851711026616,
+      "grad_norm": 0.6767911475856203,
+      "learning_rate": 3.914581497110684e-06,
+      "loss": 0.2953,
+      "step": 776
+    },
+    {
+      "epoch": 1.4771863117870723,
+      "grad_norm": 0.6507483411066131,
+      "learning_rate": 3.888262859021508e-06,
+      "loss": 0.2926,
+      "step": 777
+    },
+    {
+      "epoch": 1.4790874524714828,
+      "grad_norm": 0.6426383102167935,
+      "learning_rate": 3.862011624322534e-06,
+      "loss": 0.2922,
+      "step": 778
+    },
+    {
+      "epoch": 1.4809885931558935,
+      "grad_norm": 0.7215316838042333,
+      "learning_rate": 3.835828082525925e-06,
+      "loss": 0.3135,
+      "step": 779
+    },
+    {
+      "epoch": 1.4828897338403042,
+      "grad_norm": 0.6615360335675595,
+      "learning_rate": 3.8097125223972864e-06,
+      "loss": 0.2898,
+      "step": 780
+    },
+    {
+      "epoch": 1.4847908745247147,
+      "grad_norm": 0.6801732538242692,
+      "learning_rate": 3.7836652319524835e-06,
+      "loss": 0.2926,
+      "step": 781
+    },
+    {
+      "epoch": 1.4866920152091254,
+      "grad_norm": 0.6757730873246335,
+      "learning_rate": 3.7576864984544814e-06,
+      "loss": 0.3132,
+      "step": 782
+    },
+    {
+      "epoch": 1.4885931558935361,
+      "grad_norm": 0.6841603700834816,
+      "learning_rate": 3.73177660841015e-06,
+      "loss": 0.3252,
+      "step": 783
+    },
+    {
+      "epoch": 1.4904942965779466,
+      "grad_norm": 0.6794648800110954,
+      "learning_rate": 3.7059358475671225e-06,
+      "loss": 0.3118,
+      "step": 784
+    },
+    {
+      "epoch": 1.4923954372623573,
+      "grad_norm": 0.6419266792305235,
+      "learning_rate": 3.680164500910646e-06,
+      "loss": 0.2999,
+      "step": 785
+    },
+    {
+      "epoch": 1.494296577946768,
+      "grad_norm": 0.6515348749594349,
+      "learning_rate": 3.654462852660423e-06,
+      "loss": 0.3157,
+      "step": 786
+    },
+    {
+      "epoch": 1.4961977186311788,
+      "grad_norm": 0.6416219717705516,
+      "learning_rate": 3.6288311862674885e-06,
+      "loss": 0.3229,
+      "step": 787
+    },
+    {
+      "epoch": 1.4980988593155893,
+      "grad_norm": 0.6520759764270343,
+      "learning_rate": 3.6032697844110896e-06,
+      "loss": 0.3102,
+      "step": 788
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6451415146836288,
+      "learning_rate": 3.5777789289955454e-06,
+      "loss": 0.3055,
+      "step": 789
+    },
+    {
+      "epoch": 1.5019011406844105,
+      "grad_norm": 0.6390330470686145,
+      "learning_rate": 3.5523589011471592e-06,
+      "loss": 0.2841,
+      "step": 790
+    },
+    {
+      "epoch": 1.5038022813688214,
+      "grad_norm": 0.7007684636203716,
+      "learning_rate": 3.527009981211119e-06,
+      "loss": 0.3146,
+      "step": 791
+    },
+    {
+      "epoch": 1.505703422053232,
+      "grad_norm": 0.5977372351511565,
+      "learning_rate": 3.5017324487483873e-06,
+      "loss": 0.2893,
+      "step": 792
+    },
+    {
+      "epoch": 1.5076045627376424,
+      "grad_norm": 0.6467232521728097,
+      "learning_rate": 3.47652658253263e-06,
+      "loss": 0.2973,
+      "step": 793
+    },
+    {
+      "epoch": 1.5095057034220534,
+      "grad_norm": 0.672042618932107,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.2915,
+      "step": 794
+    },
+    {
+      "epoch": 1.5114068441064639,
+      "grad_norm": 0.6099346805059594,
+      "learning_rate": 3.4263309599818017e-06,
+      "loss": 0.2974,
+      "step": 795
+    },
+    {
+      "epoch": 1.5133079847908744,
+      "grad_norm": 0.6634113301178954,
+      "learning_rate": 3.4013417572299446e-06,
+      "loss": 0.3208,
+      "step": 796
+    },
+    {
+      "epoch": 1.5152091254752853,
+      "grad_norm": 0.6560948529802192,
+      "learning_rate": 3.37642532788541e-06,
+      "loss": 0.3052,
+      "step": 797
+    },
+    {
+      "epoch": 1.5171102661596958,
+      "grad_norm": 0.6381856678759426,
+      "learning_rate": 3.3515819467394184e-06,
+      "loss": 0.3023,
+      "step": 798
+    },
+    {
+      "epoch": 1.5190114068441065,
+      "grad_norm": 0.6340096475046755,
+      "learning_rate": 3.326811887777607e-06,
+      "loss": 0.3008,
+      "step": 799
+    },
+    {
+      "epoch": 1.5209125475285172,
+      "grad_norm": 0.6639364359662655,
+      "learning_rate": 3.3021154241769606e-06,
+      "loss": 0.2939,
+      "step": 800
+    },
+    {
+      "epoch": 1.5228136882129277,
+      "grad_norm": 0.6656842068204994,
+      "learning_rate": 3.2774928283028153e-06,
+      "loss": 0.3131,
+      "step": 801
+    },
+    {
+      "epoch": 1.5247148288973384,
+      "grad_norm": 0.7000106433561076,
+      "learning_rate": 3.2529443717058693e-06,
+      "loss": 0.3221,
+      "step": 802
+    },
+    {
+      "epoch": 1.5266159695817492,
+      "grad_norm": 0.7795468713772266,
+      "learning_rate": 3.228470325119164e-06,
+      "loss": 0.2999,
+      "step": 803
+    },
+    {
+      "epoch": 1.5285171102661597,
+      "grad_norm": 0.6312943947894026,
+      "learning_rate": 3.20407095845511e-06,
+      "loss": 0.2929,
+      "step": 804
+    },
+    {
+      "epoch": 1.5304182509505704,
+      "grad_norm": 0.7004987843014133,
+      "learning_rate": 3.179746540802506e-06,
+      "loss": 0.3267,
+      "step": 805
+    },
+    {
+      "epoch": 1.532319391634981,
+      "grad_norm": 0.6186306426503927,
+      "learning_rate": 3.155497340423588e-06,
+      "loss": 0.2967,
+      "step": 806
+    },
+    {
+      "epoch": 1.5342205323193916,
+      "grad_norm": 0.6414129222984206,
+      "learning_rate": 3.1313236247510414e-06,
+      "loss": 0.2915,
+      "step": 807
+    },
+    {
+      "epoch": 1.5361216730038023,
+      "grad_norm": 0.7056038929404758,
+      "learning_rate": 3.107225660385077e-06,
+      "loss": 0.3233,
+      "step": 808
+    },
+    {
+      "epoch": 1.538022813688213,
+      "grad_norm": 0.6615139009923636,
+      "learning_rate": 3.0832037130904748e-06,
+      "loss": 0.3079,
+      "step": 809
+    },
+    {
+      "epoch": 1.5399239543726235,
+      "grad_norm": 0.6417345968753414,
+      "learning_rate": 3.0592580477936606e-06,
+      "loss": 0.2972,
+      "step": 810
+    },
+    {
+      "epoch": 1.5418250950570342,
+      "grad_norm": 0.6975595339345018,
+      "learning_rate": 3.035388928579792e-06,
+      "loss": 0.3283,
+      "step": 811
+    },
+    {
+      "epoch": 1.543726235741445,
+      "grad_norm": 0.7062767102068703,
+      "learning_rate": 3.011596618689825e-06,
+      "loss": 0.2972,
+      "step": 812
+    },
+    {
+      "epoch": 1.5456273764258555,
+      "grad_norm": 0.6570514913751374,
+      "learning_rate": 2.9878813805176252e-06,
+      "loss": 0.2993,
+      "step": 813
+    },
+    {
+      "epoch": 1.5475285171102662,
+      "grad_norm": 0.6467045734296413,
+      "learning_rate": 2.9642434756070793e-06,
+      "loss": 0.2934,
+      "step": 814
+    },
+    {
+      "epoch": 1.549429657794677,
+      "grad_norm": 0.6465515071482372,
+      "learning_rate": 2.940683164649194e-06,
+      "loss": 0.297,
+      "step": 815
+    },
+    {
+      "epoch": 1.5513307984790874,
+      "grad_norm": 0.6448161357590496,
+      "learning_rate": 2.9172007074792342e-06,
+      "loss": 0.2934,
+      "step": 816
+    },
+    {
+      "epoch": 1.553231939163498,
+      "grad_norm": 0.6770231150233177,
+      "learning_rate": 2.8937963630738517e-06,
+      "loss": 0.3174,
+      "step": 817
+    },
+    {
+      "epoch": 1.5551330798479088,
+      "grad_norm": 0.6650765765829416,
+      "learning_rate": 2.87047038954823e-06,
+      "loss": 0.3158,
+      "step": 818
+    },
+    {
+      "epoch": 1.5570342205323193,
+      "grad_norm": 0.6437231085969415,
+      "learning_rate": 2.8472230441532365e-06,
+      "loss": 0.3017,
+      "step": 819
+    },
+    {
+      "epoch": 1.55893536121673,
+      "grad_norm": 0.6555844881174364,
+      "learning_rate": 2.8240545832725963e-06,
+      "loss": 0.2985,
+      "step": 820
+    },
+    {
+      "epoch": 1.5608365019011408,
+      "grad_norm": 0.6564183147990454,
+      "learning_rate": 2.8009652624200436e-06,
+      "loss": 0.3053,
+      "step": 821
+    },
+    {
+      "epoch": 1.5627376425855513,
+      "grad_norm": 0.6445011684397202,
+      "learning_rate": 2.7779553362365184e-06,
+      "loss": 0.2889,
+      "step": 822
+    },
+    {
+      "epoch": 1.564638783269962,
+      "grad_norm": 0.6909310861345966,
+      "learning_rate": 2.755025058487364e-06,
+      "loss": 0.3078,
+      "step": 823
+    },
+    {
+      "epoch": 1.5665399239543727,
+      "grad_norm": 0.674653281176453,
+      "learning_rate": 2.7321746820595084e-06,
+      "loss": 0.3016,
+      "step": 824
+    },
+    {
+      "epoch": 1.5684410646387832,
+      "grad_norm": 0.6647913737680543,
+      "learning_rate": 2.709404458958693e-06,
+      "loss": 0.2931,
+      "step": 825
+    },
+    {
+      "epoch": 1.570342205323194,
+      "grad_norm": 0.6516005908467472,
+      "learning_rate": 2.6867146403066833e-06,
+      "loss": 0.2822,
+      "step": 826
+    },
+    {
+      "epoch": 1.5722433460076046,
+      "grad_norm": 0.6287654793860907,
+      "learning_rate": 2.6641054763385044e-06,
+      "loss": 0.2975,
+      "step": 827
+    },
+    {
+      "epoch": 1.5741444866920151,
+      "grad_norm": 0.7061945540043719,
+      "learning_rate": 2.6415772163996845e-06,
+      "loss": 0.3019,
+      "step": 828
+    },
+    {
+      "epoch": 1.5760456273764258,
+      "grad_norm": 0.6437258193736213,
+      "learning_rate": 2.619130108943494e-06,
+      "loss": 0.2961,
+      "step": 829
+    },
+    {
+      "epoch": 1.5779467680608366,
+      "grad_norm": 0.658046269857564,
+      "learning_rate": 2.5967644015282146e-06,
+      "loss": 0.2859,
+      "step": 830
+    },
+    {
+      "epoch": 1.579847908745247,
+      "grad_norm": 0.6548519257557726,
+      "learning_rate": 2.5744803408144026e-06,
+      "loss": 0.2931,
+      "step": 831
+    },
+    {
+      "epoch": 1.5817490494296578,
+      "grad_norm": 0.6754084075085008,
+      "learning_rate": 2.5522781725621814e-06,
+      "loss": 0.3157,
+      "step": 832
+    },
+    {
+      "epoch": 1.5836501901140685,
+      "grad_norm": 0.6807881444420174,
+      "learning_rate": 2.530158141628515e-06,
+      "loss": 0.3034,
+      "step": 833
+    },
+    {
+      "epoch": 1.585551330798479,
+      "grad_norm": 0.6542877718177368,
+      "learning_rate": 2.508120491964512e-06,
+      "loss": 0.2994,
+      "step": 834
+    },
+    {
+      "epoch": 1.5874524714828897,
+      "grad_norm": 0.6644972022747354,
+      "learning_rate": 2.486165466612751e-06,
+      "loss": 0.2969,
+      "step": 835
+    },
+    {
+      "epoch": 1.5893536121673004,
+      "grad_norm": 0.6761855907532135,
+      "learning_rate": 2.464293307704566e-06,
+      "loss": 0.3022,
+      "step": 836
+    },
+    {
+      "epoch": 1.591254752851711,
+      "grad_norm": 0.7079332536661452,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.2983,
+      "step": 837
+    },
+    {
+      "epoch": 1.5931558935361216,
+      "grad_norm": 0.6236686411342898,
+      "learning_rate": 2.4207985531722034e-06,
+      "loss": 0.2917,
+      "step": 838
+    },
+    {
+      "epoch": 1.5950570342205324,
+      "grad_norm": 0.6316323230978219,
+      "learning_rate": 2.3991764372306113e-06,
+      "loss": 0.2857,
+      "step": 839
+    },
+    {
+      "epoch": 1.5969581749049429,
+      "grad_norm": 0.5990237537652477,
+      "learning_rate": 2.377638147092497e-06,
+      "loss": 0.2674,
+      "step": 840
+    },
+    {
+      "epoch": 1.5988593155893536,
+      "grad_norm": 0.6666637010576122,
+      "learning_rate": 2.3561839202932344e-06,
+      "loss": 0.3016,
+      "step": 841
+    },
+    {
+      "epoch": 1.6007604562737643,
+      "grad_norm": 0.6805951701176431,
+      "learning_rate": 2.3348139934411008e-06,
+      "loss": 0.2967,
+      "step": 842
+    },
+    {
+      "epoch": 1.6026615969581748,
+      "grad_norm": 0.6448274875956888,
+      "learning_rate": 2.3135286022146785e-06,
+      "loss": 0.2903,
+      "step": 843
+    },
+    {
+      "epoch": 1.6045627376425855,
+      "grad_norm": 0.6070663317322023,
+      "learning_rate": 2.292327981360245e-06,
+      "loss": 0.288,
+      "step": 844
+    },
+    {
+      "epoch": 1.6064638783269962,
+      "grad_norm": 0.6536822181107652,
+      "learning_rate": 2.271212364689176e-06,
+      "loss": 0.2945,
+      "step": 845
+    },
+    {
+      "epoch": 1.6083650190114067,
+      "grad_norm": 0.6682161464958246,
+      "learning_rate": 2.2501819850753925e-06,
+      "loss": 0.3104,
+      "step": 846
+    },
+    {
+      "epoch": 1.6102661596958177,
+      "grad_norm": 0.7131182349175145,
+      "learning_rate": 2.229237074452768e-06,
+      "loss": 0.307,
+      "step": 847
+    },
+    {
+      "epoch": 1.6121673003802282,
+      "grad_norm": 0.6343190066900588,
+      "learning_rate": 2.2083778638125796e-06,
+      "loss": 0.2992,
+      "step": 848
+    },
+    {
+      "epoch": 1.6140684410646386,
+      "grad_norm": 0.6645537751647479,
+      "learning_rate": 2.1876045832009694e-06,
+      "loss": 0.2919,
+      "step": 849
+    },
+    {
+      "epoch": 1.6159695817490496,
+      "grad_norm": 0.6401455733750004,
+      "learning_rate": 2.16691746171639e-06,
+      "loss": 0.2951,
+      "step": 850
+    },
+    {
+      "epoch": 1.61787072243346,
+      "grad_norm": 0.6448892772748341,
+      "learning_rate": 2.1463167275070863e-06,
+      "loss": 0.2937,
+      "step": 851
+    },
+    {
+      "epoch": 1.6197718631178706,
+      "grad_norm": 0.645589027674529,
+      "learning_rate": 2.125802607768588e-06,
+      "loss": 0.2897,
+      "step": 852
+    },
+    {
+      "epoch": 1.6216730038022815,
+      "grad_norm": 0.6698967023703924,
+      "learning_rate": 2.1053753287411895e-06,
+      "loss": 0.2839,
+      "step": 853
+    },
+    {
+      "epoch": 1.623574144486692,
+      "grad_norm": 0.6861418770892869,
+      "learning_rate": 2.08503511570746e-06,
+      "loss": 0.2963,
+      "step": 854
+    },
+    {
+      "epoch": 1.6254752851711025,
+      "grad_norm": 0.6272901302628687,
+      "learning_rate": 2.064782192989765e-06,
+      "loss": 0.2925,
+      "step": 855
+    },
+    {
+      "epoch": 1.6273764258555135,
+      "grad_norm": 0.6897136778800828,
+      "learning_rate": 2.0446167839477815e-06,
+      "loss": 0.3015,
+      "step": 856
+    },
+    {
+      "epoch": 1.629277566539924,
+      "grad_norm": 0.6828784305969161,
+      "learning_rate": 2.0245391109760437e-06,
+      "loss": 0.3037,
+      "step": 857
+    },
+    {
+      "epoch": 1.6311787072243344,
+      "grad_norm": 0.660337539580518,
+      "learning_rate": 2.0045493955014915e-06,
+      "loss": 0.2966,
+      "step": 858
+    },
+    {
+      "epoch": 1.6330798479087454,
+      "grad_norm": 0.650340724546804,
+      "learning_rate": 1.984647857981017e-06,
+      "loss": 0.314,
+      "step": 859
+    },
+    {
+      "epoch": 1.6349809885931559,
+      "grad_norm": 0.6242692077411451,
+      "learning_rate": 1.96483471789904e-06,
+      "loss": 0.3127,
+      "step": 860
+    },
+    {
+      "epoch": 1.6368821292775664,
+      "grad_norm": 0.6410869390323634,
+      "learning_rate": 1.9451101937650963e-06,
+      "loss": 0.3001,
+      "step": 861
+    },
+    {
+      "epoch": 1.6387832699619773,
+      "grad_norm": 0.6612442152627421,
+      "learning_rate": 1.925474503111412e-06,
+      "loss": 0.2957,
+      "step": 862
+    },
+    {
+      "epoch": 1.6406844106463878,
+      "grad_norm": 0.6464488009250702,
+      "learning_rate": 1.905927862490512e-06,
+      "loss": 0.3077,
+      "step": 863
+    },
+    {
+      "epoch": 1.6425855513307985,
+      "grad_norm": 0.6354239925628147,
+      "learning_rate": 1.8864704874728346e-06,
+      "loss": 0.286,
+      "step": 864
+    },
+    {
+      "epoch": 1.6444866920152093,
+      "grad_norm": 0.6758647177127244,
+      "learning_rate": 1.8671025926443464e-06,
+      "loss": 0.322,
+      "step": 865
+    },
+    {
+      "epoch": 1.6463878326996197,
+      "grad_norm": 0.6540101957476091,
+      "learning_rate": 1.8478243916041882e-06,
+      "loss": 0.29,
+      "step": 866
+    },
+    {
+      "epoch": 1.6482889733840305,
+      "grad_norm": 0.6231219022072507,
+      "learning_rate": 1.828636096962304e-06,
+      "loss": 0.2934,
+      "step": 867
+    },
+    {
+      "epoch": 1.6501901140684412,
+      "grad_norm": 0.65843318282095,
+      "learning_rate": 1.8095379203371044e-06,
+      "loss": 0.2966,
+      "step": 868
+    },
+    {
+      "epoch": 1.6520912547528517,
+      "grad_norm": 0.6157756858812776,
+      "learning_rate": 1.7905300723531393e-06,
+      "loss": 0.2904,
+      "step": 869
+    },
+    {
+      "epoch": 1.6539923954372624,
+      "grad_norm": 0.6857592345707993,
+      "learning_rate": 1.771612762638758e-06,
+      "loss": 0.303,
+      "step": 870
+    },
+    {
+      "epoch": 1.6558935361216731,
+      "grad_norm": 0.6395489145703676,
+      "learning_rate": 1.7527861998238094e-06,
+      "loss": 0.2868,
+      "step": 871
+    },
+    {
+      "epoch": 1.6577946768060836,
+      "grad_norm": 0.6636040683593372,
+      "learning_rate": 1.7340505915373495e-06,
+      "loss": 0.294,
+      "step": 872
+    },
+    {
+      "epoch": 1.6596958174904943,
+      "grad_norm": 0.6695972877238711,
+      "learning_rate": 1.7154061444053239e-06,
+      "loss": 0.3106,
+      "step": 873
+    },
+    {
+      "epoch": 1.661596958174905,
+      "grad_norm": 0.6247106595584331,
+      "learning_rate": 1.6968530640483126e-06,
+      "loss": 0.2843,
+      "step": 874
+    },
+    {
+      "epoch": 1.6634980988593155,
+      "grad_norm": 0.6388367686361557,
+      "learning_rate": 1.6783915550792652e-06,
+      "loss": 0.2937,
+      "step": 875
+    },
+    {
+      "epoch": 1.6653992395437263,
+      "grad_norm": 0.6424611949583836,
+      "learning_rate": 1.660021821101222e-06,
+      "loss": 0.2828,
+      "step": 876
+    },
+    {
+      "epoch": 1.667300380228137,
+      "grad_norm": 0.6098258817209667,
+      "learning_rate": 1.6417440647050853e-06,
+      "loss": 0.2837,
+      "step": 877
+    },
+    {
+      "epoch": 1.6692015209125475,
+      "grad_norm": 0.6336273919162337,
+      "learning_rate": 1.6235584874673848e-06,
+      "loss": 0.3087,
+      "step": 878
+    },
+    {
+      "epoch": 1.6711026615969582,
+      "grad_norm": 0.6204916551328592,
+      "learning_rate": 1.6054652899480472e-06,
+      "loss": 0.2783,
+      "step": 879
+    },
+    {
+      "epoch": 1.673003802281369,
+      "grad_norm": 0.6103016208913884,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.3013,
+      "step": 880
+    },
+    {
+      "epoch": 1.6749049429657794,
+      "grad_norm": 0.6438911400395089,
+      "learning_rate": 1.5695568312079156e-06,
+      "loss": 0.304,
+      "step": 881
+    },
+    {
+      "epoch": 1.6768060836501901,
+      "grad_norm": 0.6755352998787497,
+      "learning_rate": 1.5517419660041277e-06,
+      "loss": 0.3033,
+      "step": 882
+    },
+    {
+      "epoch": 1.6787072243346008,
+      "grad_norm": 0.6517579532046254,
+      "learning_rate": 1.534020272548349e-06,
+      "loss": 0.2979,
+      "step": 883
+    },
+    {
+      "epoch": 1.6806083650190113,
+      "grad_norm": 0.7196754779857704,
+      "learning_rate": 1.5163919462845622e-06,
+      "loss": 0.306,
+      "step": 884
+    },
+    {
+      "epoch": 1.682509505703422,
+      "grad_norm": 0.6751002818997694,
+      "learning_rate": 1.4988571816270402e-06,
+      "loss": 0.3098,
+      "step": 885
+    },
+    {
+      "epoch": 1.6844106463878328,
+      "grad_norm": 0.6447090573798387,
+      "learning_rate": 1.4814161719582132e-06,
+      "loss": 0.2763,
+      "step": 886
+    },
+    {
+      "epoch": 1.6863117870722433,
+      "grad_norm": 0.6797377347516157,
+      "learning_rate": 1.4640691096265358e-06,
+      "loss": 0.309,
+      "step": 887
+    },
+    {
+      "epoch": 1.688212927756654,
+      "grad_norm": 0.6600841805375199,
+      "learning_rate": 1.4468161859443609e-06,
+      "loss": 0.3002,
+      "step": 888
+    },
+    {
+      "epoch": 1.6901140684410647,
+      "grad_norm": 0.6476474908309581,
+      "learning_rate": 1.4296575911858268e-06,
+      "loss": 0.2951,
+      "step": 889
+    },
+    {
+      "epoch": 1.6920152091254752,
+      "grad_norm": 0.6673349198365441,
+      "learning_rate": 1.412593514584777e-06,
+      "loss": 0.2942,
+      "step": 890
+    },
+    {
+      "epoch": 1.693916349809886,
+      "grad_norm": 0.7036357487473542,
+      "learning_rate": 1.3956241443326423e-06,
+      "loss": 0.3119,
+      "step": 891
+    },
+    {
+      "epoch": 1.6958174904942966,
+      "grad_norm": 0.6581490967745297,
+      "learning_rate": 1.378749667576399e-06,
+      "loss": 0.3182,
+      "step": 892
+    },
+    {
+      "epoch": 1.6977186311787071,
+      "grad_norm": 0.6834109849224165,
+      "learning_rate": 1.3619702704164783e-06,
+      "loss": 0.3113,
+      "step": 893
+    },
+    {
+      "epoch": 1.6996197718631179,
+      "grad_norm": 0.6690538456103786,
+      "learning_rate": 1.3452861379047289e-06,
+      "loss": 0.3028,
+      "step": 894
+    },
+    {
+      "epoch": 1.7015209125475286,
+      "grad_norm": 0.682273545877122,
+      "learning_rate": 1.3286974540423747e-06,
+      "loss": 0.2985,
+      "step": 895
+    },
+    {
+      "epoch": 1.703422053231939,
+      "grad_norm": 0.6967822524905104,
+      "learning_rate": 1.3122044017779768e-06,
+      "loss": 0.3084,
+      "step": 896
+    },
+    {
+      "epoch": 1.7053231939163498,
+      "grad_norm": 0.642314343816989,
+      "learning_rate": 1.2958071630054214e-06,
+      "loss": 0.3043,
+      "step": 897
+    },
+    {
+      "epoch": 1.7072243346007605,
+      "grad_norm": 0.6743087846371169,
+      "learning_rate": 1.279505918561923e-06,
+      "loss": 0.307,
+      "step": 898
+    },
+    {
+      "epoch": 1.709125475285171,
+      "grad_norm": 0.6669759948109761,
+      "learning_rate": 1.2633008482260146e-06,
+      "loss": 0.3077,
+      "step": 899
+    },
+    {
+      "epoch": 1.7110266159695817,
+      "grad_norm": 0.6531652814850945,
+      "learning_rate": 1.2471921307155655e-06,
+      "loss": 0.2859,
+      "step": 900
+    },
+    {
+      "epoch": 1.7129277566539924,
+      "grad_norm": 0.6559755592380898,
+      "learning_rate": 1.2311799436858275e-06,
+      "loss": 0.3068,
+      "step": 901
+    },
+    {
+      "epoch": 1.714828897338403,
+      "grad_norm": 0.6704115592949618,
+      "learning_rate": 1.2152644637274603e-06,
+      "loss": 0.3044,
+      "step": 902
+    },
+    {
+      "epoch": 1.7167300380228137,
+      "grad_norm": 0.6799327837381822,
+      "learning_rate": 1.1994458663645836e-06,
+      "loss": 0.3111,
+      "step": 903
+    },
+    {
+      "epoch": 1.7186311787072244,
+      "grad_norm": 0.6977999866043937,
+      "learning_rate": 1.1837243260528542e-06,
+      "loss": 0.3065,
+      "step": 904
+    },
+    {
+      "epoch": 1.7205323193916349,
+      "grad_norm": 0.6599889555492974,
+      "learning_rate": 1.168100016177528e-06,
+      "loss": 0.3061,
+      "step": 905
+    },
+    {
+      "epoch": 1.7224334600760456,
+      "grad_norm": 0.6543929817583138,
+      "learning_rate": 1.1525731090515536e-06,
+      "loss": 0.2986,
+      "step": 906
+    },
+    {
+      "epoch": 1.7243346007604563,
+      "grad_norm": 0.6779085338106259,
+      "learning_rate": 1.137143775913675e-06,
+      "loss": 0.2958,
+      "step": 907
+    },
+    {
+      "epoch": 1.7262357414448668,
+      "grad_norm": 0.6392901161897956,
+      "learning_rate": 1.1218121869265365e-06,
+      "loss": 0.2884,
+      "step": 908
+    },
+    {
+      "epoch": 1.7281368821292775,
+      "grad_norm": 0.6037523098853499,
+      "learning_rate": 1.1065785111748117e-06,
+      "loss": 0.2737,
+      "step": 909
+    },
+    {
+      "epoch": 1.7300380228136882,
+      "grad_norm": 0.6450983967043572,
+      "learning_rate": 1.0914429166633355e-06,
+      "loss": 0.2899,
+      "step": 910
+    },
+    {
+      "epoch": 1.7319391634980987,
+      "grad_norm": 0.6698671844473164,
+      "learning_rate": 1.076405570315252e-06,
+      "loss": 0.306,
+      "step": 911
+    },
+    {
+      "epoch": 1.7338403041825095,
+      "grad_norm": 0.7188604786078552,
+      "learning_rate": 1.0614666379701732e-06,
+      "loss": 0.3007,
+      "step": 912
+    },
+    {
+      "epoch": 1.7357414448669202,
+      "grad_norm": 0.6648295515873993,
+      "learning_rate": 1.046626284382356e-06,
+      "loss": 0.3005,
+      "step": 913
+    },
+    {
+      "epoch": 1.7376425855513307,
+      "grad_norm": 0.6708475854741267,
+      "learning_rate": 1.0318846732188737e-06,
+      "loss": 0.3046,
+      "step": 914
+    },
+    {
+      "epoch": 1.7395437262357416,
+      "grad_norm": 0.6634112247710785,
+      "learning_rate": 1.017241967057816e-06,
+      "loss": 0.2981,
+      "step": 915
+    },
+    {
+      "epoch": 1.741444866920152,
+      "grad_norm": 0.6546747010623375,
+      "learning_rate": 1.0026983273865055e-06,
+      "loss": 0.3025,
+      "step": 916
+    },
+    {
+      "epoch": 1.7433460076045626,
+      "grad_norm": 0.6524121726208203,
+      "learning_rate": 9.882539145997027e-07,
+      "loss": 0.2928,
+      "step": 917
+    },
+    {
+      "epoch": 1.7452471482889735,
+      "grad_norm": 0.6857144698473752,
+      "learning_rate": 9.739088879978409e-07,
+      "loss": 0.3077,
+      "step": 918
+    },
+    {
+      "epoch": 1.747148288973384,
+      "grad_norm": 0.6511842648754607,
+      "learning_rate": 9.59663405785277e-07,
+      "loss": 0.2851,
+      "step": 919
+    },
+    {
+      "epoch": 1.7490494296577945,
+      "grad_norm": 0.6849602716535552,
+      "learning_rate": 9.455176250685338e-07,
+      "loss": 0.305,
+      "step": 920
+    },
+    {
+      "epoch": 1.7509505703422055,
+      "grad_norm": 0.6300916616065773,
+      "learning_rate": 9.314717018545838e-07,
+      "loss": 0.2931,
+      "step": 921
+    },
+    {
+      "epoch": 1.752851711026616,
+      "grad_norm": 0.6820918031412718,
+      "learning_rate": 9.17525791049112e-07,
+      "loss": 0.2996,
+      "step": 922
+    },
+    {
+      "epoch": 1.7547528517110265,
+      "grad_norm": 0.65653675468943,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.2922,
+      "step": 923
+    },
+    {
+      "epoch": 1.7566539923954374,
+      "grad_norm": 0.6887277922923002,
+      "learning_rate": 8.899346207697135e-07,
+      "loss": 0.2866,
+      "step": 924
+    },
+    {
+      "epoch": 1.758555133079848,
+      "grad_norm": 0.6921015006246054,
+      "learning_rate": 8.762896655854481e-07,
+      "loss": 0.3307,
+      "step": 925
+    },
+    {
+      "epoch": 1.7604562737642584,
+      "grad_norm": 0.6793023435167824,
+      "learning_rate": 8.627453313856249e-07,
+      "loss": 0.2963,
+      "step": 926
+    },
+    {
+      "epoch": 1.7623574144486693,
+      "grad_norm": 0.6299697631546028,
+      "learning_rate": 8.493017675441495e-07,
+      "loss": 0.3016,
+      "step": 927
+    },
+    {
+      "epoch": 1.7642585551330798,
+      "grad_norm": 0.7079498688941728,
+      "learning_rate": 8.359591223235785e-07,
+      "loss": 0.3136,
+      "step": 928
+    },
+    {
+      "epoch": 1.7661596958174905,
+      "grad_norm": 0.6650865662388512,
+      "learning_rate": 8.227175428734868e-07,
+      "loss": 0.3097,
+      "step": 929
+    },
+    {
+      "epoch": 1.7680608365019013,
+      "grad_norm": 0.6773783599829766,
+      "learning_rate": 8.095771752288451e-07,
+      "loss": 0.3007,
+      "step": 930
+    },
+    {
+      "epoch": 1.7699619771863118,
+      "grad_norm": 0.6522082016752995,
+      "learning_rate": 7.965381643084069e-07,
+      "loss": 0.2867,
+      "step": 931
+    },
+    {
+      "epoch": 1.7718631178707225,
+      "grad_norm": 0.6235179676796069,
+      "learning_rate": 7.83600653913108e-07,
+      "loss": 0.3012,
+      "step": 932
+    },
+    {
+      "epoch": 1.7737642585551332,
+      "grad_norm": 0.6381827982361415,
+      "learning_rate": 7.707647867244927e-07,
+      "loss": 0.2847,
+      "step": 933
+    },
+    {
+      "epoch": 1.7756653992395437,
+      "grad_norm": 0.6782433186902315,
+      "learning_rate": 7.580307043031232e-07,
+      "loss": 0.3075,
+      "step": 934
+    },
+    {
+      "epoch": 1.7775665399239544,
+      "grad_norm": 0.6486531141437399,
+      "learning_rate": 7.453985470870284e-07,
+      "loss": 0.2872,
+      "step": 935
+    },
+    {
+      "epoch": 1.7794676806083651,
+      "grad_norm": 0.6297554230903384,
+      "learning_rate": 7.328684543901598e-07,
+      "loss": 0.3045,
+      "step": 936
+    },
+    {
+      "epoch": 1.7813688212927756,
+      "grad_norm": 0.6224697516537832,
+      "learning_rate": 7.204405644008416e-07,
+      "loss": 0.2943,
+      "step": 937
+    },
+    {
+      "epoch": 1.7832699619771863,
+      "grad_norm": 0.624452199306809,
+      "learning_rate": 7.081150141802518e-07,
+      "loss": 0.2895,
+      "step": 938
+    },
+    {
+      "epoch": 1.785171102661597,
+      "grad_norm": 0.6197252909452271,
+      "learning_rate": 6.958919396609231e-07,
+      "loss": 0.2992,
+      "step": 939
+    },
+    {
+      "epoch": 1.7870722433460076,
+      "grad_norm": 0.6575601895095748,
+      "learning_rate": 6.837714756452241e-07,
+      "loss": 0.2931,
+      "step": 940
+    },
+    {
+      "epoch": 1.7889733840304183,
+      "grad_norm": 0.6349315838973804,
+      "learning_rate": 6.717537558038845e-07,
+      "loss": 0.2954,
+      "step": 941
+    },
+    {
+      "epoch": 1.790874524714829,
+      "grad_norm": 0.6474000975584557,
+      "learning_rate": 6.598389126745209e-07,
+      "loss": 0.3047,
+      "step": 942
+    },
+    {
+      "epoch": 1.7927756653992395,
+      "grad_norm": 0.6268917324240604,
+      "learning_rate": 6.480270776601682e-07,
+      "loss": 0.2877,
+      "step": 943
+    },
+    {
+      "epoch": 1.7946768060836502,
+      "grad_norm": 0.6320558359819795,
+      "learning_rate": 6.36318381027835e-07,
+      "loss": 0.2948,
+      "step": 944
+    },
+    {
+      "epoch": 1.796577946768061,
+      "grad_norm": 0.6252826712570154,
+      "learning_rate": 6.247129519070728e-07,
+      "loss": 0.2856,
+      "step": 945
+    },
+    {
+      "epoch": 1.7984790874524714,
+      "grad_norm": 0.6369321370908779,
+      "learning_rate": 6.132109182885382e-07,
+      "loss": 0.2999,
+      "step": 946
+    },
+    {
+      "epoch": 1.8003802281368821,
+      "grad_norm": 0.680328571307375,
+      "learning_rate": 6.018124070225928e-07,
+      "loss": 0.2854,
+      "step": 947
+    },
+    {
+      "epoch": 1.8022813688212929,
+      "grad_norm": 0.595233430152232,
+      "learning_rate": 5.905175438178979e-07,
+      "loss": 0.2983,
+      "step": 948
+    },
+    {
+      "epoch": 1.8041825095057034,
+      "grad_norm": 0.6342684302005492,
+      "learning_rate": 5.793264532400311e-07,
+      "loss": 0.3078,
+      "step": 949
+    },
+    {
+      "epoch": 1.806083650190114,
+      "grad_norm": 0.6563193316365448,
+      "learning_rate": 5.68239258710116e-07,
+      "loss": 0.3005,
+      "step": 950
+    },
+    {
+      "epoch": 1.8079847908745248,
+      "grad_norm": 0.6395725995365743,
+      "learning_rate": 5.572560825034523e-07,
+      "loss": 0.2939,
+      "step": 951
+    },
+    {
+      "epoch": 1.8098859315589353,
+      "grad_norm": 0.6934516478877907,
+      "learning_rate": 5.463770457481732e-07,
+      "loss": 0.3065,
+      "step": 952
+    },
+    {
+      "epoch": 1.811787072243346,
+      "grad_norm": 0.6473710945850893,
+      "learning_rate": 5.35602268423906e-07,
+      "loss": 0.3057,
+      "step": 953
+    },
+    {
+      "epoch": 1.8136882129277567,
+      "grad_norm": 0.6650519112805277,
+      "learning_rate": 5.249318693604577e-07,
+      "loss": 0.307,
+      "step": 954
+    },
+    {
+      "epoch": 1.8155893536121672,
+      "grad_norm": 0.6549396626250665,
+      "learning_rate": 5.143659662364931e-07,
+      "loss": 0.3025,
+      "step": 955
+    },
+    {
+      "epoch": 1.817490494296578,
+      "grad_norm": 0.66690605724442,
+      "learning_rate": 5.039046755782417e-07,
+      "loss": 0.2926,
+      "step": 956
+    },
+    {
+      "epoch": 1.8193916349809887,
+      "grad_norm": 0.6296814221306245,
+      "learning_rate": 4.935481127582131e-07,
+      "loss": 0.2787,
+      "step": 957
+    },
+    {
+      "epoch": 1.8212927756653992,
+      "grad_norm": 0.6861140106899881,
+      "learning_rate": 4.83296391993926e-07,
+      "loss": 0.3167,
+      "step": 958
+    },
+    {
+      "epoch": 1.8231939163498099,
+      "grad_norm": 0.6244342338887253,
+      "learning_rate": 4.7314962634664616e-07,
+      "loss": 0.2923,
+      "step": 959
+    },
+    {
+      "epoch": 1.8250950570342206,
+      "grad_norm": 0.6107334303789421,
+      "learning_rate": 4.631079277201389e-07,
+      "loss": 0.2798,
+      "step": 960
+    },
+    {
+      "epoch": 1.826996197718631,
+      "grad_norm": 0.6358962173489247,
+      "learning_rate": 4.5317140685943726e-07,
+      "loss": 0.2955,
+      "step": 961
+    },
+    {
+      "epoch": 1.8288973384030418,
+      "grad_norm": 1.090562260478516,
+      "learning_rate": 4.433401733496201e-07,
+      "loss": 0.2976,
+      "step": 962
+    },
+    {
+      "epoch": 1.8307984790874525,
+      "grad_norm": 0.6433615000028192,
+      "learning_rate": 4.3361433561460274e-07,
+      "loss": 0.3028,
+      "step": 963
+    },
+    {
+      "epoch": 1.832699619771863,
+      "grad_norm": 0.6349156921369634,
+      "learning_rate": 4.2399400091594154e-07,
+      "loss": 0.28,
+      "step": 964
+    },
+    {
+      "epoch": 1.8346007604562737,
+      "grad_norm": 0.6073828695101741,
+      "learning_rate": 4.14479275351648e-07,
+      "loss": 0.279,
+      "step": 965
+    },
+    {
+      "epoch": 1.8365019011406845,
+      "grad_norm": 0.6415782505981967,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.286,
+      "step": 966
+    },
+    {
+      "epoch": 1.838403041825095,
+      "grad_norm": 0.6492098737245536,
+      "learning_rate": 3.9576707019350903e-07,
+      "loss": 0.3011,
+      "step": 967
+    },
+    {
+      "epoch": 1.8403041825095057,
+      "grad_norm": 0.6922497966354111,
+      "learning_rate": 3.865697969675164e-07,
+      "loss": 0.3115,
+      "step": 968
+    },
+    {
+      "epoch": 1.8422053231939164,
+      "grad_norm": 0.6646766364736871,
+      "learning_rate": 3.7747854560931996e-07,
+      "loss": 0.2984,
+      "step": 969
+    },
+    {
+      "epoch": 1.8441064638783269,
+      "grad_norm": 0.6437735096565248,
+      "learning_rate": 3.684934163819309e-07,
+      "loss": 0.2982,
+      "step": 970
+    },
+    {
+      "epoch": 1.8460076045627376,
+      "grad_norm": 0.6580071735611409,
+      "learning_rate": 3.596145083779912e-07,
+      "loss": 0.2872,
+      "step": 971
+    },
+    {
+      "epoch": 1.8479087452471483,
+      "grad_norm": 0.6552366466020833,
+      "learning_rate": 3.508419195186774e-07,
+      "loss": 0.2887,
+      "step": 972
+    },
+    {
+      "epoch": 1.8498098859315588,
+      "grad_norm": 0.6379237491369711,
+      "learning_rate": 3.421757465526243e-07,
+      "loss": 0.2979,
+      "step": 973
+    },
+    {
+      "epoch": 1.8517110266159695,
+      "grad_norm": 0.6712156119188715,
+      "learning_rate": 3.33616085054862e-07,
+      "loss": 0.3007,
+      "step": 974
+    },
+    {
+      "epoch": 1.8536121673003803,
+      "grad_norm": 0.6938088811002897,
+      "learning_rate": 3.2516302942574794e-07,
+      "loss": 0.3019,
+      "step": 975
+    },
+    {
+      "epoch": 1.8555133079847907,
+      "grad_norm": 0.6486081219171044,
+      "learning_rate": 3.1681667288994353e-07,
+      "loss": 0.2979,
+      "step": 976
+    },
+    {
+      "epoch": 1.8574144486692015,
+      "grad_norm": 0.6213783951876463,
+      "learning_rate": 3.0857710749537585e-07,
+      "loss": 0.2842,
+      "step": 977
+    },
+    {
+      "epoch": 1.8593155893536122,
+      "grad_norm": 0.636682209888363,
+      "learning_rate": 3.0044442411222066e-07,
+      "loss": 0.2909,
+      "step": 978
+    },
+    {
+      "epoch": 1.8612167300380227,
+      "grad_norm": 0.6863881898837885,
+      "learning_rate": 2.9241871243190555e-07,
+      "loss": 0.3036,
+      "step": 979
+    },
+    {
+      "epoch": 1.8631178707224336,
+      "grad_norm": 0.6525914300599576,
+      "learning_rate": 2.845000609661208e-07,
+      "loss": 0.2856,
+      "step": 980
+    },
+    {
+      "epoch": 1.8650190114068441,
+      "grad_norm": 0.6406882149369793,
+      "learning_rate": 2.7668855704583997e-07,
+      "loss": 0.2916,
+      "step": 981
+    },
+    {
+      "epoch": 1.8669201520912546,
+      "grad_norm": 0.6220648985590173,
+      "learning_rate": 2.689842868203563e-07,
+      "loss": 0.2876,
+      "step": 982
+    },
+    {
+      "epoch": 1.8688212927756656,
+      "grad_norm": 0.7013901258210652,
+      "learning_rate": 2.6138733525633896e-07,
+      "loss": 0.3011,
+      "step": 983
+    },
+    {
+      "epoch": 1.870722433460076,
+      "grad_norm": 0.6589907998271488,
+      "learning_rate": 2.5389778613688744e-07,
+      "loss": 0.3033,
+      "step": 984
+    },
+    {
+      "epoch": 1.8726235741444865,
+      "grad_norm": 0.6362889501668817,
+      "learning_rate": 2.46515722060614e-07,
+      "loss": 0.3019,
+      "step": 985
+    },
+    {
+      "epoch": 1.8745247148288975,
+      "grad_norm": 0.6390381552592813,
+      "learning_rate": 2.392412244407294e-07,
+      "loss": 0.2688,
+      "step": 986
+    },
+    {
+      "epoch": 1.876425855513308,
+      "grad_norm": 0.6508716061948319,
+      "learning_rate": 2.3207437350414418e-07,
+      "loss": 0.2829,
+      "step": 987
+    },
+    {
+      "epoch": 1.8783269961977185,
+      "grad_norm": 0.6512543928057634,
+      "learning_rate": 2.2501524829059208e-07,
+      "loss": 0.2949,
+      "step": 988
+    },
+    {
+      "epoch": 1.8802281368821294,
+      "grad_norm": 0.6769977707451778,
+      "learning_rate": 2.180639266517448e-07,
+      "loss": 0.2923,
+      "step": 989
+    },
+    {
+      "epoch": 1.88212927756654,
+      "grad_norm": 0.6869567061645191,
+      "learning_rate": 2.1122048525036409e-07,
+      "loss": 0.3003,
+      "step": 990
+    },
+    {
+      "epoch": 1.8840304182509504,
+      "grad_norm": 0.6430624599490663,
+      "learning_rate": 2.0448499955945223e-07,
+      "loss": 0.2975,
+      "step": 991
+    },
+    {
+      "epoch": 1.8859315589353614,
+      "grad_norm": 0.7010410458067073,
+      "learning_rate": 1.9785754386142164e-07,
+      "loss": 0.2994,
+      "step": 992
+    },
+    {
+      "epoch": 1.8878326996197718,
+      "grad_norm": 0.6299372927644749,
+      "learning_rate": 1.9133819124727003e-07,
+      "loss": 0.2938,
+      "step": 993
+    },
+    {
+      "epoch": 1.8897338403041823,
+      "grad_norm": 0.6926910051931238,
+      "learning_rate": 1.8492701361578326e-07,
+      "loss": 0.3211,
+      "step": 994
+    },
+    {
+      "epoch": 1.8916349809885933,
+      "grad_norm": 0.6453307382724512,
+      "learning_rate": 1.7862408167273472e-07,
+      "loss": 0.293,
+      "step": 995
+    },
+    {
+      "epoch": 1.8935361216730038,
+      "grad_norm": 0.6332735087033571,
+      "learning_rate": 1.724294649301095e-07,
+      "loss": 0.3071,
+      "step": 996
+    },
+    {
+      "epoch": 1.8954372623574145,
+      "grad_norm": 0.6783361936956762,
+      "learning_rate": 1.6634323170533928e-07,
+      "loss": 0.3057,
+      "step": 997
+    },
+    {
+      "epoch": 1.8973384030418252,
+      "grad_norm": 0.6541600071840108,
+      "learning_rate": 1.6036544912054087e-07,
+      "loss": 0.3028,
+      "step": 998
+    },
+    {
+      "epoch": 1.8992395437262357,
+      "grad_norm": 0.6493509675046257,
+      "learning_rate": 1.544961831017855e-07,
+      "loss": 0.2927,
+      "step": 999
+    },
+    {
+      "epoch": 1.9011406844106464,
+      "grad_norm": 0.6824336123713493,
+      "learning_rate": 1.487354983783673e-07,
+      "loss": 0.3088,
+      "step": 1000
+    },
+    {
+      "epoch": 1.9030418250950571,
+      "grad_norm": 0.6607958461237431,
+      "learning_rate": 1.430834584820895e-07,
+      "loss": 0.2921,
+      "step": 1001
+    },
+    {
+      "epoch": 1.9049429657794676,
+      "grad_norm": 0.6711877757901952,
+      "learning_rate": 1.375401257465625e-07,
+      "loss": 0.3053,
+      "step": 1002
+    },
+    {
+      "epoch": 1.9068441064638784,
+      "grad_norm": 0.6305806973766593,
+      "learning_rate": 1.3210556130652031e-07,
+      "loss": 0.2934,
+      "step": 1003
+    },
+    {
+      "epoch": 1.908745247148289,
+      "grad_norm": 0.6426180867272963,
+      "learning_rate": 1.2677982509714415e-07,
+      "loss": 0.2948,
+      "step": 1004
+    },
+    {
+      "epoch": 1.9106463878326996,
+      "grad_norm": 0.643476698028188,
+      "learning_rate": 1.2156297585339872e-07,
+      "loss": 0.2872,
+      "step": 1005
+    },
+    {
+      "epoch": 1.9125475285171103,
+      "grad_norm": 0.6423443018918694,
+      "learning_rate": 1.1645507110938925e-07,
+      "loss": 0.2882,
+      "step": 1006
+    },
+    {
+      "epoch": 1.914448669201521,
+      "grad_norm": 0.6340494380935477,
+      "learning_rate": 1.1145616719772545e-07,
+      "loss": 0.3063,
+      "step": 1007
+    },
+    {
+      "epoch": 1.9163498098859315,
+      "grad_norm": 0.6336286660968592,
+      "learning_rate": 1.0656631924889749e-07,
+      "loss": 0.2923,
+      "step": 1008
+    },
+    {
+      "epoch": 1.9182509505703422,
+      "grad_norm": 0.6151212732962817,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.2947,
+      "step": 1009
+    },
+    {
+      "epoch": 1.920152091254753,
+      "grad_norm": 0.6849394764737524,
+      "learning_rate": 9.711400574749507e-08,
+      "loss": 0.2907,
+      "step": 1010
+    },
+    {
+      "epoch": 1.9220532319391634,
+      "grad_norm": 0.6162552921195816,
+      "learning_rate": 9.255164443990994e-08,
+      "loss": 0.2858,
+      "step": 1011
+    },
+    {
+      "epoch": 1.9239543726235742,
+      "grad_norm": 0.6665486759426869,
+      "learning_rate": 8.809854758399017e-08,
+      "loss": 0.3088,
+      "step": 1012
+    },
+    {
+      "epoch": 1.9258555133079849,
+      "grad_norm": 0.6310602114888915,
+      "learning_rate": 8.375476429078543e-08,
+      "loss": 0.2992,
+      "step": 1013
+    },
+    {
+      "epoch": 1.9277566539923954,
+      "grad_norm": 0.6254121124143176,
+      "learning_rate": 7.952034246577977e-08,
+      "loss": 0.2833,
+      "step": 1014
+    },
+    {
+      "epoch": 1.929657794676806,
+      "grad_norm": 0.6291638219381042,
+      "learning_rate": 7.539532880836087e-08,
+      "loss": 0.2978,
+      "step": 1015
+    },
+    {
+      "epoch": 1.9315589353612168,
+      "grad_norm": 0.6252249469286902,
+      "learning_rate": 7.137976881130826e-08,
+      "loss": 0.3042,
+      "step": 1016
+    },
+    {
+      "epoch": 1.9334600760456273,
+      "grad_norm": 0.663625695068365,
+      "learning_rate": 6.747370676028819e-08,
+      "loss": 0.3033,
+      "step": 1017
+    },
+    {
+      "epoch": 1.935361216730038,
+      "grad_norm": 0.654142469857814,
+      "learning_rate": 6.367718573336845e-08,
+      "loss": 0.294,
+      "step": 1018
+    },
+    {
+      "epoch": 1.9372623574144487,
+      "grad_norm": 0.688037008690913,
+      "learning_rate": 5.999024760054095e-08,
+      "loss": 0.2974,
+      "step": 1019
+    },
+    {
+      "epoch": 1.9391634980988592,
+      "grad_norm": 0.6616747263155708,
+      "learning_rate": 5.641293302326323e-08,
+      "loss": 0.299,
+      "step": 1020
+    },
+    {
+      "epoch": 1.94106463878327,
+      "grad_norm": 0.6677750834904783,
+      "learning_rate": 5.2945281454003236e-08,
+      "loss": 0.3045,
+      "step": 1021
+    },
+    {
+      "epoch": 1.9429657794676807,
+      "grad_norm": 0.6380391903927578,
+      "learning_rate": 4.958733113581415e-08,
+      "loss": 0.3102,
+      "step": 1022
+    },
+    {
+      "epoch": 1.9448669201520912,
+      "grad_norm": 0.6456797084549314,
+      "learning_rate": 4.6339119101902475e-08,
+      "loss": 0.2968,
+      "step": 1023
+    },
+    {
+      "epoch": 1.946768060836502,
+      "grad_norm": 0.6487714220119952,
+      "learning_rate": 4.320068117522835e-08,
+      "loss": 0.3007,
+      "step": 1024
+    },
+    {
+      "epoch": 1.9486692015209126,
+      "grad_norm": 0.6562217606239844,
+      "learning_rate": 4.0172051968101474e-08,
+      "loss": 0.296,
+      "step": 1025
+    },
+    {
+      "epoch": 1.950570342205323,
+      "grad_norm": 0.6012376958942862,
+      "learning_rate": 3.7253264881809137e-08,
+      "loss": 0.286,
+      "step": 1026
+    },
+    {
+      "epoch": 1.9524714828897338,
+      "grad_norm": 0.6271445455693571,
+      "learning_rate": 3.4444352106242086e-08,
+      "loss": 0.2913,
+      "step": 1027
+    },
+    {
+      "epoch": 1.9543726235741445,
+      "grad_norm": 0.6463585635143356,
+      "learning_rate": 3.174534461953593e-08,
+      "loss": 0.2938,
+      "step": 1028
+    },
+    {
+      "epoch": 1.956273764258555,
+      "grad_norm": 0.6686713905555918,
+      "learning_rate": 2.915627218774142e-08,
+      "loss": 0.2922,
+      "step": 1029
+    },
+    {
+      "epoch": 1.9581749049429658,
+      "grad_norm": 0.679768785795892,
+      "learning_rate": 2.667716336448356e-08,
+      "loss": 0.3123,
+      "step": 1030
+    },
+    {
+      "epoch": 1.9600760456273765,
+      "grad_norm": 0.643451937545961,
+      "learning_rate": 2.430804549065302e-08,
+      "loss": 0.3051,
+      "step": 1031
+    },
+    {
+      "epoch": 1.961977186311787,
+      "grad_norm": 0.6410106394326839,
+      "learning_rate": 2.2048944694104123e-08,
+      "loss": 0.305,
+      "step": 1032
+    },
+    {
+      "epoch": 1.9638783269961977,
+      "grad_norm": 0.6759529034641186,
+      "learning_rate": 1.989988588936509e-08,
+      "loss": 0.3208,
+      "step": 1033
+    },
+    {
+      "epoch": 1.9657794676806084,
+      "grad_norm": 0.6325279872811366,
+      "learning_rate": 1.7860892777367133e-08,
+      "loss": 0.291,
+      "step": 1034
+    },
+    {
+      "epoch": 1.967680608365019,
+      "grad_norm": 0.6283507210427595,
+      "learning_rate": 1.5931987845176912e-08,
+      "loss": 0.2699,
+      "step": 1035
+    },
+    {
+      "epoch": 1.9695817490494296,
+      "grad_norm": 0.6436336038628104,
+      "learning_rate": 1.411319236575337e-08,
+      "loss": 0.2947,
+      "step": 1036
+    },
+    {
+      "epoch": 1.9714828897338403,
+      "grad_norm": 0.6770641786569078,
+      "learning_rate": 1.2404526397711281e-08,
+      "loss": 0.3123,
+      "step": 1037
+    },
+    {
+      "epoch": 1.9733840304182508,
+      "grad_norm": 0.6567153462010384,
+      "learning_rate": 1.0806008785100297e-08,
+      "loss": 0.2822,
+      "step": 1038
+    },
+    {
+      "epoch": 1.9752851711026616,
+      "grad_norm": 0.7348843813145909,
+      "learning_rate": 9.317657157197347e-09,
+      "loss": 0.3201,
+      "step": 1039
+    },
+    {
+      "epoch": 1.9771863117870723,
+      "grad_norm": 0.678891700136706,
+      "learning_rate": 7.93948792831234e-09,
+      "loss": 0.2854,
+      "step": 1040
+    },
+    {
+      "epoch": 1.9790874524714828,
+      "grad_norm": 0.6472270301589406,
+      "learning_rate": 6.671516297606095e-09,
+      "loss": 0.3025,
+      "step": 1041
+    },
+    {
+      "epoch": 1.9809885931558935,
+      "grad_norm": 0.6369877455101876,
+      "learning_rate": 5.513756248924917e-09,
+      "loss": 0.2997,
+      "step": 1042
+    },
+    {
+      "epoch": 1.9828897338403042,
+      "grad_norm": 0.6343939877094746,
+      "learning_rate": 4.466220550641831e-09,
+      "loss": 0.2999,
+      "step": 1043
+    },
+    {
+      "epoch": 1.9847908745247147,
+      "grad_norm": 0.6874717833323032,
+      "learning_rate": 3.528920755523357e-09,
+      "loss": 0.3039,
+      "step": 1044
+    },
+    {
+      "epoch": 1.9866920152091256,
+      "grad_norm": 0.6489440690816634,
+      "learning_rate": 2.701867200592956e-09,
+      "loss": 0.2955,
+      "step": 1045
+    },
+    {
+      "epoch": 1.9885931558935361,
+      "grad_norm": 0.6328025499853304,
+      "learning_rate": 1.9850690070266633e-09,
+      "loss": 0.2953,
+      "step": 1046
+    },
+    {
+      "epoch": 1.9904942965779466,
+      "grad_norm": 0.6710730927086966,
+      "learning_rate": 1.378534080042071e-09,
+      "loss": 0.2981,
+      "step": 1047
+    },
+    {
+      "epoch": 1.9923954372623576,
+      "grad_norm": 0.6344502048586423,
+      "learning_rate": 8.822691088195001e-10,
+      "loss": 0.291,
+      "step": 1048
+    },
+    {
+      "epoch": 1.994296577946768,
+      "grad_norm": 0.648672823551679,
+      "learning_rate": 4.962795664265052e-10,
+      "loss": 0.2887,
+      "step": 1049
+    },
+    {
+      "epoch": 1.9961977186311786,
+      "grad_norm": 0.6298060409016122,
+      "learning_rate": 2.2056970975459223e-10,
+      "loss": 0.2851,
+      "step": 1050
+    },
+    {
+      "epoch": 1.9980988593155895,
+      "grad_norm": 0.6585889234116388,
+      "learning_rate": 5.514257947369928e-11,
+      "loss": 0.3096,
+      "step": 1051
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6134089912729347,
+      "learning_rate": 0.0,
+      "loss": 0.2954,
+      "step": 1052
+    },
+    {
+      "epoch": 2.0,
+      "step": 1052,
+      "total_flos": 2.234764237115228e+17,
+      "train_loss": 0.3796640654083894,
+      "train_runtime": 2677.3586,
+      "train_samples_per_second": 50.294,
+      "train_steps_per_second": 0.393
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1052,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.234764237115228e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}