diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4108 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 582,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.005154639175257732,
+      "grad_norm": 16.203728734181727,
+      "learning_rate": 0.0,
+      "loss": 0.782,
+      "step": 1
+    },
+    {
+      "epoch": 0.010309278350515464,
+      "grad_norm": 17.966080542459263,
+      "learning_rate": 1.6949152542372883e-07,
+      "loss": 0.8806,
+      "step": 2
+    },
+    {
+      "epoch": 0.015463917525773196,
+      "grad_norm": 19.39755361913842,
+      "learning_rate": 3.3898305084745766e-07,
+      "loss": 0.9713,
+      "step": 3
+    },
+    {
+      "epoch": 0.020618556701030927,
+      "grad_norm": 18.521681485723096,
+      "learning_rate": 5.084745762711865e-07,
+      "loss": 0.7594,
+      "step": 4
+    },
+    {
+      "epoch": 0.02577319587628866,
+      "grad_norm": 16.960097281692487,
+      "learning_rate": 6.779661016949153e-07,
+      "loss": 1.0095,
+      "step": 5
+    },
+    {
+      "epoch": 0.030927835051546393,
+      "grad_norm": 17.9303091015193,
+      "learning_rate": 8.474576271186441e-07,
+      "loss": 1.0024,
+      "step": 6
+    },
+    {
+      "epoch": 0.03608247422680412,
+      "grad_norm": 16.393493317929863,
+      "learning_rate": 1.016949152542373e-06,
+      "loss": 0.7936,
+      "step": 7
+    },
+    {
+      "epoch": 0.041237113402061855,
+      "grad_norm": 16.350575208232332,
+      "learning_rate": 1.186440677966102e-06,
+      "loss": 0.6727,
+      "step": 8
+    },
+    {
+      "epoch": 0.04639175257731959,
+      "grad_norm": 13.392733244736176,
+      "learning_rate": 1.3559322033898307e-06,
+      "loss": 0.6366,
+      "step": 9
+    },
+    {
+      "epoch": 0.05154639175257732,
+      "grad_norm": 13.99345418325563,
+      "learning_rate": 1.5254237288135596e-06,
+      "loss": 0.8352,
+      "step": 10
+    },
+    {
+      "epoch": 0.05670103092783505,
+      "grad_norm": 16.574632940102198,
+      "learning_rate": 1.6949152542372882e-06,
+      "loss": 0.8211,
+      "step": 11
+    },
+    {
+      "epoch": 0.061855670103092786,
+      "grad_norm": 7.432819145467484,
+      "learning_rate": 1.8644067796610171e-06,
+      "loss": 0.4671,
+      "step": 12
+    },
+    {
+      "epoch": 0.06701030927835051,
+      "grad_norm": 7.368792116734323,
+      "learning_rate": 2.033898305084746e-06,
+      "loss": 0.4617,
+      "step": 13
+    },
+    {
+      "epoch": 0.07216494845360824,
+      "grad_norm": 5.189955956808789,
+      "learning_rate": 2.203389830508475e-06,
+      "loss": 0.4118,
+      "step": 14
+    },
+    {
+      "epoch": 0.07731958762886598,
+      "grad_norm": 6.110418564319445,
+      "learning_rate": 2.372881355932204e-06,
+      "loss": 0.4772,
+      "step": 15
+    },
+    {
+      "epoch": 0.08247422680412371,
+      "grad_norm": 2.524741574511695,
+      "learning_rate": 2.5423728813559323e-06,
+      "loss": 0.1693,
+      "step": 16
+    },
+    {
+      "epoch": 0.08762886597938144,
+      "grad_norm": 4.093704704249702,
+      "learning_rate": 2.7118644067796613e-06,
+      "loss": 0.5946,
+      "step": 17
+    },
+    {
+      "epoch": 0.09278350515463918,
+      "grad_norm": 2.792475441128467,
+      "learning_rate": 2.8813559322033903e-06,
+      "loss": 0.3907,
+      "step": 18
+    },
+    {
+      "epoch": 0.0979381443298969,
+      "grad_norm": 2.8142697054484422,
+      "learning_rate": 3.0508474576271192e-06,
+      "loss": 0.4128,
+      "step": 19
+    },
+    {
+      "epoch": 0.10309278350515463,
+      "grad_norm": 3.0557753995059835,
+      "learning_rate": 3.2203389830508473e-06,
+      "loss": 0.4128,
+      "step": 20
+    },
+    {
+      "epoch": 0.10824742268041238,
+      "grad_norm": 2.407462983991879,
+      "learning_rate": 3.3898305084745763e-06,
+      "loss": 0.2454,
+      "step": 21
+    },
+    {
+      "epoch": 0.1134020618556701,
+      "grad_norm": 4.175164045570657,
+      "learning_rate": 3.5593220338983053e-06,
+      "loss": 0.5676,
+      "step": 22
+    },
+    {
+      "epoch": 0.11855670103092783,
+      "grad_norm": 3.2745984458679542,
+      "learning_rate": 3.7288135593220342e-06,
+      "loss": 0.3725,
+      "step": 23
+    },
+    {
+      "epoch": 0.12371134020618557,
+      "grad_norm": 2.630482014005189,
+      "learning_rate": 3.898305084745763e-06,
+      "loss": 0.284,
+      "step": 24
+    },
+    {
+      "epoch": 0.12886597938144329,
+      "grad_norm": 2.5818523655885657,
+      "learning_rate": 4.067796610169492e-06,
+      "loss": 0.3914,
+      "step": 25
+    },
+    {
+      "epoch": 0.13402061855670103,
+      "grad_norm": 2.1862545024847733,
+      "learning_rate": 4.23728813559322e-06,
+      "loss": 0.3245,
+      "step": 26
+    },
+    {
+      "epoch": 0.13917525773195877,
+      "grad_norm": 1.8029553321125102,
+      "learning_rate": 4.40677966101695e-06,
+      "loss": 0.1906,
+      "step": 27
+    },
+    {
+      "epoch": 0.14432989690721648,
+      "grad_norm": 2.1274587742776228,
+      "learning_rate": 4.576271186440678e-06,
+      "loss": 0.3129,
+      "step": 28
+    },
+    {
+      "epoch": 0.14948453608247422,
+      "grad_norm": 2.0553069128433985,
+      "learning_rate": 4.745762711864408e-06,
+      "loss": 0.362,
+      "step": 29
+    },
+    {
+      "epoch": 0.15463917525773196,
+      "grad_norm": 2.2636243270640812,
+      "learning_rate": 4.915254237288136e-06,
+      "loss": 0.3487,
+      "step": 30
+    },
+    {
+      "epoch": 0.15979381443298968,
+      "grad_norm": 2.0627428648390023,
+      "learning_rate": 5.084745762711865e-06,
+      "loss": 0.3017,
+      "step": 31
+    },
+    {
+      "epoch": 0.16494845360824742,
+      "grad_norm": 1.707507687529141,
+      "learning_rate": 5.254237288135594e-06,
+      "loss": 0.2379,
+      "step": 32
+    },
+    {
+      "epoch": 0.17010309278350516,
+      "grad_norm": 1.6433474151177994,
+      "learning_rate": 5.423728813559323e-06,
+      "loss": 0.2785,
+      "step": 33
+    },
+    {
+      "epoch": 0.17525773195876287,
+      "grad_norm": 1.8049400111968015,
+      "learning_rate": 5.593220338983051e-06,
+      "loss": 0.2046,
+      "step": 34
+    },
+    {
+      "epoch": 0.18041237113402062,
+      "grad_norm": 1.956624509674235,
+      "learning_rate": 5.7627118644067805e-06,
+      "loss": 0.2924,
+      "step": 35
+    },
+    {
+      "epoch": 0.18556701030927836,
+      "grad_norm": 2.043442546886233,
+      "learning_rate": 5.932203389830509e-06,
+      "loss": 0.2517,
+      "step": 36
+    },
+    {
+      "epoch": 0.19072164948453607,
+      "grad_norm": 1.7809412898084382,
+      "learning_rate": 6.1016949152542385e-06,
+      "loss": 0.214,
+      "step": 37
+    },
+    {
+      "epoch": 0.1958762886597938,
+      "grad_norm": 1.0871959933984234,
+      "learning_rate": 6.271186440677966e-06,
+      "loss": 0.0794,
+      "step": 38
+    },
+    {
+      "epoch": 0.20103092783505155,
+      "grad_norm": 1.6627701820242757,
+      "learning_rate": 6.440677966101695e-06,
+      "loss": 0.1529,
+      "step": 39
+    },
+    {
+      "epoch": 0.20618556701030927,
+      "grad_norm": 2.233281327693517,
+      "learning_rate": 6.610169491525424e-06,
+      "loss": 0.3918,
+      "step": 40
+    },
+    {
+      "epoch": 0.211340206185567,
+      "grad_norm": 2.4810527246233143,
+      "learning_rate": 6.779661016949153e-06,
+      "loss": 0.3276,
+      "step": 41
+    },
+    {
+      "epoch": 0.21649484536082475,
+      "grad_norm": 2.138313718866784,
+      "learning_rate": 6.949152542372882e-06,
+      "loss": 0.1874,
+      "step": 42
+    },
+    {
+      "epoch": 0.22164948453608246,
+      "grad_norm": 2.3207138440634445,
+      "learning_rate": 7.1186440677966106e-06,
+      "loss": 0.143,
+      "step": 43
+    },
+    {
+      "epoch": 0.2268041237113402,
+      "grad_norm": 1.806845922939825,
+      "learning_rate": 7.288135593220339e-06,
+      "loss": 0.2639,
+      "step": 44
+    },
+    {
+      "epoch": 0.23195876288659795,
+      "grad_norm": 1.6856465858879643,
+      "learning_rate": 7.4576271186440685e-06,
+      "loss": 0.1803,
+      "step": 45
+    },
+    {
+      "epoch": 0.23711340206185566,
+      "grad_norm": 1.8626644402225192,
+      "learning_rate": 7.627118644067797e-06,
+      "loss": 0.2046,
+      "step": 46
+    },
+    {
+      "epoch": 0.2422680412371134,
+      "grad_norm": 1.7027465927641543,
+      "learning_rate": 7.796610169491526e-06,
+      "loss": 0.124,
+      "step": 47
+    },
+    {
+      "epoch": 0.24742268041237114,
+      "grad_norm": 1.9041870166779817,
+      "learning_rate": 7.966101694915255e-06,
+      "loss": 0.1982,
+      "step": 48
+    },
+    {
+      "epoch": 0.25257731958762886,
+      "grad_norm": 1.770069442877448,
+      "learning_rate": 8.135593220338983e-06,
+      "loss": 0.2813,
+      "step": 49
+    },
+    {
+      "epoch": 0.25773195876288657,
+      "grad_norm": 1.953365961357983,
+      "learning_rate": 8.305084745762712e-06,
+      "loss": 0.2251,
+      "step": 50
+    },
+    {
+      "epoch": 0.26288659793814434,
+      "grad_norm": 2.002309230380239,
+      "learning_rate": 8.47457627118644e-06,
+      "loss": 0.2785,
+      "step": 51
+    },
+    {
+      "epoch": 0.26804123711340205,
+      "grad_norm": 1.5522236927921595,
+      "learning_rate": 8.64406779661017e-06,
+      "loss": 0.2388,
+      "step": 52
+    },
+    {
+      "epoch": 0.27319587628865977,
+      "grad_norm": 1.4065957481275269,
+      "learning_rate": 8.8135593220339e-06,
+      "loss": 0.1592,
+      "step": 53
+    },
+    {
+      "epoch": 0.27835051546391754,
+      "grad_norm": 1.344843484282986,
+      "learning_rate": 8.983050847457628e-06,
+      "loss": 0.184,
+      "step": 54
+    },
+    {
+      "epoch": 0.28350515463917525,
+      "grad_norm": 1.8809824575076326,
+      "learning_rate": 9.152542372881356e-06,
+      "loss": 0.1559,
+      "step": 55
+    },
+    {
+      "epoch": 0.28865979381443296,
+      "grad_norm": 1.554740329023379,
+      "learning_rate": 9.322033898305085e-06,
+      "loss": 0.1945,
+      "step": 56
+    },
+    {
+      "epoch": 0.29381443298969073,
+      "grad_norm": 1.7706197688983198,
+      "learning_rate": 9.491525423728815e-06,
+      "loss": 0.2799,
+      "step": 57
+    },
+    {
+      "epoch": 0.29896907216494845,
+      "grad_norm": 1.339143828253002,
+      "learning_rate": 9.661016949152544e-06,
+      "loss": 0.1094,
+      "step": 58
+    },
+    {
+      "epoch": 0.30412371134020616,
+      "grad_norm": 2.1220725285103086,
+      "learning_rate": 9.830508474576272e-06,
+      "loss": 0.3211,
+      "step": 59
+    },
+    {
+      "epoch": 0.30927835051546393,
+      "grad_norm": 1.3971936674913341,
+      "learning_rate": 1e-05,
+      "loss": 0.2081,
+      "step": 60
+    },
+    {
+      "epoch": 0.31443298969072164,
+      "grad_norm": 1.8165010693502677,
+      "learning_rate": 9.999909794073715e-06,
+      "loss": 0.191,
+      "step": 61
+    },
+    {
+      "epoch": 0.31958762886597936,
+      "grad_norm": 1.8206712917879821,
+      "learning_rate": 9.999639179549699e-06,
+      "loss": 0.1817,
+      "step": 62
+    },
+    {
+      "epoch": 0.3247422680412371,
+      "grad_norm": 2.1904370888969704,
+      "learning_rate": 9.999188166192368e-06,
+      "loss": 0.1933,
+      "step": 63
+    },
+    {
+      "epoch": 0.32989690721649484,
+      "grad_norm": 1.356123942482486,
+      "learning_rate": 9.998556770275351e-06,
+      "loss": 0.1529,
+      "step": 64
+    },
+    {
+      "epoch": 0.33505154639175255,
+      "grad_norm": 1.4167068373050817,
+      "learning_rate": 9.997745014580912e-06,
+      "loss": 0.2182,
+      "step": 65
+    },
+    {
+      "epoch": 0.3402061855670103,
+      "grad_norm": 1.2778114297184384,
+      "learning_rate": 9.996752928399121e-06,
+      "loss": 0.1231,
+      "step": 66
+    },
+    {
+      "epoch": 0.34536082474226804,
+      "grad_norm": 1.4693549048926278,
+      "learning_rate": 9.995580547526798e-06,
+      "loss": 0.2222,
+      "step": 67
+    },
+    {
+      "epoch": 0.35051546391752575,
+      "grad_norm": 1.4349642188073422,
+      "learning_rate": 9.994227914266222e-06,
+      "loss": 0.1934,
+      "step": 68
+    },
+    {
+      "epoch": 0.3556701030927835,
+      "grad_norm": 1.3639498444936125,
+      "learning_rate": 9.992695077423609e-06,
+      "loss": 0.1707,
+      "step": 69
+    },
+    {
+      "epoch": 0.36082474226804123,
+      "grad_norm": 1.3943030751682663,
+      "learning_rate": 9.990982092307347e-06,
+      "loss": 0.1716,
+      "step": 70
+    },
+    {
+      "epoch": 0.36597938144329895,
+      "grad_norm": 1.4616151628165952,
+      "learning_rate": 9.989089020725999e-06,
+      "loss": 0.2363,
+      "step": 71
+    },
+    {
+      "epoch": 0.3711340206185567,
+      "grad_norm": 1.7297954033170302,
+      "learning_rate": 9.987015930986074e-06,
+      "loss": 0.0612,
+      "step": 72
+    },
+    {
+      "epoch": 0.37628865979381443,
+      "grad_norm": 1.4130551566794556,
+      "learning_rate": 9.984762897889568e-06,
+      "loss": 0.2438,
+      "step": 73
+    },
+    {
+      "epoch": 0.38144329896907214,
+      "grad_norm": 1.4316589582684316,
+      "learning_rate": 9.98233000273125e-06,
+      "loss": 0.2529,
+      "step": 74
+    },
+    {
+      "epoch": 0.3865979381443299,
+      "grad_norm": 1.4921368045815033,
+      "learning_rate": 9.97971733329575e-06,
+      "loss": 0.2033,
+      "step": 75
+    },
+    {
+      "epoch": 0.3917525773195876,
+      "grad_norm": 1.6884913833837993,
+      "learning_rate": 9.97692498385437e-06,
+      "loss": 0.3251,
+      "step": 76
+    },
+    {
+      "epoch": 0.39690721649484534,
+      "grad_norm": 1.2618275142031115,
+      "learning_rate": 9.973953055161702e-06,
+      "loss": 0.2173,
+      "step": 77
+    },
+    {
+      "epoch": 0.4020618556701031,
+      "grad_norm": 1.2058891275808017,
+      "learning_rate": 9.970801654451974e-06,
+      "loss": 0.1337,
+      "step": 78
+    },
+    {
+      "epoch": 0.4072164948453608,
+      "grad_norm": 1.3014010097389481,
+      "learning_rate": 9.967470895435197e-06,
+      "loss": 0.2217,
+      "step": 79
+    },
+    {
+      "epoch": 0.41237113402061853,
+      "grad_norm": 1.0907558669955588,
+      "learning_rate": 9.963960898293049e-06,
+      "loss": 0.1166,
+      "step": 80
+    },
+    {
+      "epoch": 0.4175257731958763,
+      "grad_norm": 1.3508925279475854,
+      "learning_rate": 9.96027178967455e-06,
+      "loss": 0.2185,
+      "step": 81
+    },
+    {
+      "epoch": 0.422680412371134,
+      "grad_norm": 1.8405638710013943,
+      "learning_rate": 9.956403702691482e-06,
+      "loss": 0.2981,
+      "step": 82
+    },
+    {
+      "epoch": 0.42783505154639173,
+      "grad_norm": 0.9981831988732881,
+      "learning_rate": 9.952356776913594e-06,
+      "loss": 0.1471,
+      "step": 83
+    },
+    {
+      "epoch": 0.4329896907216495,
+      "grad_norm": 1.5438578399853147,
+      "learning_rate": 9.948131158363564e-06,
+      "loss": 0.2219,
+      "step": 84
+    },
+    {
+      "epoch": 0.4381443298969072,
+      "grad_norm": 1.373967099765683,
+      "learning_rate": 9.943726999511721e-06,
+      "loss": 0.2625,
+      "step": 85
+    },
+    {
+      "epoch": 0.44329896907216493,
+      "grad_norm": 1.5045665327970632,
+      "learning_rate": 9.939144459270557e-06,
+      "loss": 0.2378,
+      "step": 86
+    },
+    {
+      "epoch": 0.4484536082474227,
+      "grad_norm": 1.2348801218707532,
+      "learning_rate": 9.934383702988992e-06,
+      "loss": 0.1299,
+      "step": 87
+    },
+    {
+      "epoch": 0.4536082474226804,
+      "grad_norm": 1.3457854200982862,
+      "learning_rate": 9.929444902446392e-06,
+      "loss": 0.2193,
+      "step": 88
+    },
+    {
+      "epoch": 0.4587628865979381,
+      "grad_norm": 1.5974941630503592,
+      "learning_rate": 9.924328235846393e-06,
+      "loss": 0.3447,
+      "step": 89
+    },
+    {
+      "epoch": 0.4639175257731959,
+      "grad_norm": 1.7921462524347982,
+      "learning_rate": 9.919033887810451e-06,
+      "loss": 0.3081,
+      "step": 90
+    },
+    {
+      "epoch": 0.4690721649484536,
+      "grad_norm": 1.0544333428105994,
+      "learning_rate": 9.913562049371196e-06,
+      "loss": 0.1723,
+      "step": 91
+    },
+    {
+      "epoch": 0.4742268041237113,
+      "grad_norm": 1.173180659355975,
+      "learning_rate": 9.90791291796553e-06,
+      "loss": 0.1816,
+      "step": 92
+    },
+    {
+      "epoch": 0.4793814432989691,
+      "grad_norm": 1.2343501847799747,
+      "learning_rate": 9.902086697427504e-06,
+      "loss": 0.2048,
+      "step": 93
+    },
+    {
+      "epoch": 0.4845360824742268,
+      "grad_norm": 1.4166315003510848,
+      "learning_rate": 9.896083597980968e-06,
+      "loss": 0.1611,
+      "step": 94
+    },
+    {
+      "epoch": 0.4896907216494845,
+      "grad_norm": 0.9316258247144923,
+      "learning_rate": 9.88990383623198e-06,
+      "loss": 0.1307,
+      "step": 95
+    },
+    {
+      "epoch": 0.4948453608247423,
+      "grad_norm": 1.1284196528102999,
+      "learning_rate": 9.883547635160991e-06,
+      "loss": 0.1816,
+      "step": 96
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9136263992805103,
+      "learning_rate": 9.877015224114806e-06,
+      "loss": 0.1173,
+      "step": 97
+    },
+    {
+      "epoch": 0.5051546391752577,
+      "grad_norm": 1.5233730471599805,
+      "learning_rate": 9.870306838798299e-06,
+      "loss": 0.2239,
+      "step": 98
+    },
+    {
+      "epoch": 0.5103092783505154,
+      "grad_norm": 1.4031394060036828,
+      "learning_rate": 9.863422721265913e-06,
+      "loss": 0.2398,
+      "step": 99
+    },
+    {
+      "epoch": 0.5154639175257731,
+      "grad_norm": 1.12765997685746,
+      "learning_rate": 9.856363119912931e-06,
+      "loss": 0.1318,
+      "step": 100
+    },
+    {
+      "epoch": 0.520618556701031,
+      "grad_norm": 0.8420145818838017,
+      "learning_rate": 9.849128289466503e-06,
+      "loss": 0.1246,
+      "step": 101
+    },
+    {
+      "epoch": 0.5257731958762887,
+      "grad_norm": 0.946095131905409,
+      "learning_rate": 9.841718490976461e-06,
+      "loss": 0.1241,
+      "step": 102
+    },
+    {
+      "epoch": 0.5309278350515464,
+      "grad_norm": 0.9796883673872017,
+      "learning_rate": 9.8341339918059e-06,
+      "loss": 0.0955,
+      "step": 103
+    },
+    {
+      "epoch": 0.5360824742268041,
+      "grad_norm": 1.1930460993410585,
+      "learning_rate": 9.826375065621533e-06,
+      "loss": 0.1646,
+      "step": 104
+    },
+    {
+      "epoch": 0.5412371134020618,
+      "grad_norm": 1.27651192729961,
+      "learning_rate": 9.818441992383802e-06,
+      "loss": 0.1566,
+      "step": 105
+    },
+    {
+      "epoch": 0.5463917525773195,
+      "grad_norm": 1.1315939745886057,
+      "learning_rate": 9.810335058336801e-06,
+      "loss": 0.1786,
+      "step": 106
+    },
+    {
+      "epoch": 0.5515463917525774,
+      "grad_norm": 0.9355015058700618,
+      "learning_rate": 9.802054555997927e-06,
+      "loss": 0.0945,
+      "step": 107
+    },
+    {
+      "epoch": 0.5567010309278351,
+      "grad_norm": 1.7639349345635655,
+      "learning_rate": 9.79360078414733e-06,
+      "loss": 0.254,
+      "step": 108
+    },
+    {
+      "epoch": 0.5618556701030928,
+      "grad_norm": 1.3663266258683646,
+      "learning_rate": 9.784974047817142e-06,
+      "loss": 0.1953,
+      "step": 109
+    },
+    {
+      "epoch": 0.5670103092783505,
+      "grad_norm": 1.441702754636908,
+      "learning_rate": 9.776174658280458e-06,
+      "loss": 0.1841,
+      "step": 110
+    },
+    {
+      "epoch": 0.5721649484536082,
+      "grad_norm": 1.1499799893613811,
+      "learning_rate": 9.767202933040111e-06,
+      "loss": 0.1997,
+      "step": 111
+    },
+    {
+      "epoch": 0.5773195876288659,
+      "grad_norm": 1.2076330399094353,
+      "learning_rate": 9.758059195817216e-06,
+      "loss": 0.1768,
+      "step": 112
+    },
+    {
+      "epoch": 0.5824742268041238,
+      "grad_norm": 0.878277090541755,
+      "learning_rate": 9.748743776539489e-06,
+      "loss": 0.1173,
+      "step": 113
+    },
+    {
+      "epoch": 0.5876288659793815,
+      "grad_norm": 1.0228670375234183,
+      "learning_rate": 9.739257011329336e-06,
+      "loss": 0.1361,
+      "step": 114
+    },
+    {
+      "epoch": 0.5927835051546392,
+      "grad_norm": 1.4761049929099612,
+      "learning_rate": 9.729599242491738e-06,
+      "loss": 0.2792,
+      "step": 115
+    },
+    {
+      "epoch": 0.5979381443298969,
+      "grad_norm": 1.0296746075367031,
+      "learning_rate": 9.719770818501885e-06,
+      "loss": 0.1,
+      "step": 116
+    },
+    {
+      "epoch": 0.6030927835051546,
+      "grad_norm": 0.9546648084132865,
+      "learning_rate": 9.709772093992619e-06,
+      "loss": 0.1072,
+      "step": 117
+    },
+    {
+      "epoch": 0.6082474226804123,
+      "grad_norm": 0.9359977344927911,
+      "learning_rate": 9.699603429741615e-06,
+      "loss": 0.1171,
+      "step": 118
+    },
+    {
+      "epoch": 0.6134020618556701,
+      "grad_norm": 1.2912642891989952,
+      "learning_rate": 9.689265192658387e-06,
+      "loss": 0.1783,
+      "step": 119
+    },
+    {
+      "epoch": 0.6185567010309279,
+      "grad_norm": 1.1169115856391039,
+      "learning_rate": 9.67875775577104e-06,
+      "loss": 0.2051,
+      "step": 120
+    },
+    {
+      "epoch": 0.6237113402061856,
+      "grad_norm": 1.4787464277701932,
+      "learning_rate": 9.668081498212799e-06,
+      "loss": 0.2377,
+      "step": 121
+    },
+    {
+      "epoch": 0.6288659793814433,
+      "grad_norm": 0.9403827546416947,
+      "learning_rate": 9.657236805208347e-06,
+      "loss": 0.1219,
+      "step": 122
+    },
+    {
+      "epoch": 0.634020618556701,
+      "grad_norm": 0.7358413148522787,
+      "learning_rate": 9.646224068059917e-06,
+      "loss": 0.0895,
+      "step": 123
+    },
+    {
+      "epoch": 0.6391752577319587,
+      "grad_norm": 1.1530465641868872,
+      "learning_rate": 9.63504368413317e-06,
+      "loss": 0.2149,
+      "step": 124
+    },
+    {
+      "epoch": 0.6443298969072165,
+      "grad_norm": 1.4217244985998034,
+      "learning_rate": 9.62369605684286e-06,
+      "loss": 0.197,
+      "step": 125
+    },
+    {
+      "epoch": 0.6494845360824743,
+      "grad_norm": 1.910749720256578,
+      "learning_rate": 9.612181595638279e-06,
+      "loss": 0.2508,
+      "step": 126
+    },
+    {
+      "epoch": 0.654639175257732,
+      "grad_norm": 1.2757659427730983,
+      "learning_rate": 9.600500715988486e-06,
+      "loss": 0.2809,
+      "step": 127
+    },
+    {
+      "epoch": 0.6597938144329897,
+      "grad_norm": 1.0891754027329807,
+      "learning_rate": 9.588653839367304e-06,
+      "loss": 0.165,
+      "step": 128
+    },
+    {
+      "epoch": 0.6649484536082474,
+      "grad_norm": 1.293429113847632,
+      "learning_rate": 9.576641393238129e-06,
+      "loss": 0.0855,
+      "step": 129
+    },
+    {
+      "epoch": 0.6701030927835051,
+      "grad_norm": 1.4599213966596118,
+      "learning_rate": 9.564463811038489e-06,
+      "loss": 0.2503,
+      "step": 130
+    },
+    {
+      "epoch": 0.6752577319587629,
+      "grad_norm": 1.3648021125385577,
+      "learning_rate": 9.55212153216442e-06,
+      "loss": 0.1794,
+      "step": 131
+    },
+    {
+      "epoch": 0.6804123711340206,
+      "grad_norm": 0.9809648930976671,
+      "learning_rate": 9.5396150019546e-06,
+      "loss": 0.0749,
+      "step": 132
+    },
+    {
+      "epoch": 0.6855670103092784,
+      "grad_norm": 1.2479101653995877,
+      "learning_rate": 9.526944671674287e-06,
+      "loss": 0.1705,
+      "step": 133
+    },
+    {
+      "epoch": 0.6907216494845361,
+      "grad_norm": 1.180767703263976,
+      "learning_rate": 9.514110998499032e-06,
+      "loss": 0.2419,
+      "step": 134
+    },
+    {
+      "epoch": 0.6958762886597938,
+      "grad_norm": 0.9973862175809532,
+      "learning_rate": 9.501114445498183e-06,
+      "loss": 0.1545,
+      "step": 135
+    },
+    {
+      "epoch": 0.7010309278350515,
+      "grad_norm": 0.9580969711134073,
+      "learning_rate": 9.487955481618184e-06,
+      "loss": 0.1563,
+      "step": 136
+    },
+    {
+      "epoch": 0.7061855670103093,
+      "grad_norm": 1.0576903096823542,
+      "learning_rate": 9.474634581665645e-06,
+      "loss": 0.128,
+      "step": 137
+    },
+    {
+      "epoch": 0.711340206185567,
+      "grad_norm": 1.1608177511918483,
+      "learning_rate": 9.461152226290212e-06,
+      "loss": 0.2324,
+      "step": 138
+    },
+    {
+      "epoch": 0.7164948453608248,
+      "grad_norm": 0.7247855062304921,
+      "learning_rate": 9.44750890196723e-06,
+      "loss": 0.0696,
+      "step": 139
+    },
+    {
+      "epoch": 0.7216494845360825,
+      "grad_norm": 1.294941334572682,
+      "learning_rate": 9.43370510098018e-06,
+      "loss": 0.2068,
+      "step": 140
+    },
+    {
+      "epoch": 0.7268041237113402,
+      "grad_norm": 1.5307061314891508,
+      "learning_rate": 9.419741321402923e-06,
+      "loss": 0.134,
+      "step": 141
+    },
+    {
+      "epoch": 0.7319587628865979,
+      "grad_norm": 1.724248389396898,
+      "learning_rate": 9.405618067081729e-06,
+      "loss": 0.3587,
+      "step": 142
+    },
+    {
+      "epoch": 0.7371134020618557,
+      "grad_norm": 0.7880104208217124,
+      "learning_rate": 9.391335847617093e-06,
+      "loss": 0.1081,
+      "step": 143
+    },
+    {
+      "epoch": 0.7422680412371134,
+      "grad_norm": 1.176218642717777,
+      "learning_rate": 9.37689517834535e-06,
+      "loss": 0.2552,
+      "step": 144
+    },
+    {
+      "epoch": 0.7474226804123711,
+      "grad_norm": 0.9007810641742303,
+      "learning_rate": 9.362296580320078e-06,
+      "loss": 0.1411,
+      "step": 145
+    },
+    {
+      "epoch": 0.7525773195876289,
+      "grad_norm": 1.0987174332049452,
+      "learning_rate": 9.347540580293301e-06,
+      "loss": 0.1984,
+      "step": 146
+    },
+    {
+      "epoch": 0.7577319587628866,
+      "grad_norm": 0.9593956667801066,
+      "learning_rate": 9.332627710696477e-06,
+      "loss": 0.1403,
+      "step": 147
+    },
+    {
+      "epoch": 0.7628865979381443,
+      "grad_norm": 1.5203788468340629,
+      "learning_rate": 9.317558509621297e-06,
+      "loss": 0.3035,
+      "step": 148
+    },
+    {
+      "epoch": 0.7680412371134021,
+      "grad_norm": 0.9687515693670804,
+      "learning_rate": 9.302333520800253e-06,
+      "loss": 0.1366,
+      "step": 149
+    },
+    {
+      "epoch": 0.7731958762886598,
+      "grad_norm": 0.7832678330612435,
+      "learning_rate": 9.286953293587035e-06,
+      "loss": 0.11,
+      "step": 150
+    },
+    {
+      "epoch": 0.7783505154639175,
+      "grad_norm": 0.9289801978987015,
+      "learning_rate": 9.271418382936697e-06,
+      "loss": 0.0988,
+      "step": 151
+    },
+    {
+      "epoch": 0.7835051546391752,
+      "grad_norm": 1.1356540696916577,
+      "learning_rate": 9.255729349385645e-06,
+      "loss": 0.1375,
+      "step": 152
+    },
+    {
+      "epoch": 0.788659793814433,
+      "grad_norm": 1.3495379294095782,
+      "learning_rate": 9.239886759031399e-06,
+      "loss": 0.3033,
+      "step": 153
+    },
+    {
+      "epoch": 0.7938144329896907,
+      "grad_norm": 1.2612818337900333,
+      "learning_rate": 9.223891183512174e-06,
+      "loss": 0.1364,
+      "step": 154
+    },
+    {
+      "epoch": 0.7989690721649485,
+      "grad_norm": 1.1325852394462053,
+      "learning_rate": 9.207743199986252e-06,
+      "loss": 0.1559,
+      "step": 155
+    },
+    {
+      "epoch": 0.8041237113402062,
+      "grad_norm": 0.6629329908683664,
+      "learning_rate": 9.191443391111157e-06,
+      "loss": 0.0678,
+      "step": 156
+    },
+    {
+      "epoch": 0.8092783505154639,
+      "grad_norm": 1.184280051526035,
+      "learning_rate": 9.174992345022636e-06,
+      "loss": 0.1762,
+      "step": 157
+    },
+    {
+      "epoch": 0.8144329896907216,
+      "grad_norm": 1.4363833078261825,
+      "learning_rate": 9.158390655313422e-06,
+      "loss": 0.2509,
+      "step": 158
+    },
+    {
+      "epoch": 0.8195876288659794,
+      "grad_norm": 1.4851296678460735,
+      "learning_rate": 9.141638921011842e-06,
+      "loss": 0.3053,
+      "step": 159
+    },
+    {
+      "epoch": 0.8247422680412371,
+      "grad_norm": 0.9939630731717258,
+      "learning_rate": 9.124737746560175e-06,
+      "loss": 0.0952,
+      "step": 160
+    },
+    {
+      "epoch": 0.8298969072164949,
+      "grad_norm": 1.2169101845107388,
+      "learning_rate": 9.107687741792863e-06,
+      "loss": 0.1494,
+      "step": 161
+    },
+    {
+      "epoch": 0.8350515463917526,
+      "grad_norm": 1.3295674637794006,
+      "learning_rate": 9.090489521914492e-06,
+      "loss": 0.2761,
+      "step": 162
+    },
+    {
+      "epoch": 0.8402061855670103,
+      "grad_norm": 1.2793707854001373,
+      "learning_rate": 9.073143707477607e-06,
+      "loss": 0.1433,
+      "step": 163
+    },
+    {
+      "epoch": 0.845360824742268,
+      "grad_norm": 1.2039862398086603,
+      "learning_rate": 9.055650924360308e-06,
+      "loss": 0.2154,
+      "step": 164
+    },
+    {
+      "epoch": 0.8505154639175257,
+      "grad_norm": 1.1250244597985437,
+      "learning_rate": 9.038011803743679e-06,
+      "loss": 0.1817,
+      "step": 165
+    },
+    {
+      "epoch": 0.8556701030927835,
+      "grad_norm": 1.1063531458560925,
+      "learning_rate": 9.020226982089005e-06,
+      "loss": 0.1564,
+      "step": 166
+    },
+    {
+      "epoch": 0.8608247422680413,
+      "grad_norm": 1.1306904530717008,
+      "learning_rate": 9.002297101114813e-06,
+      "loss": 0.1349,
+      "step": 167
+    },
+    {
+      "epoch": 0.865979381443299,
+      "grad_norm": 1.0467505323127504,
+      "learning_rate": 8.984222807773707e-06,
+      "loss": 0.1813,
+      "step": 168
+    },
+    {
+      "epoch": 0.8711340206185567,
+      "grad_norm": 1.715578103301857,
+      "learning_rate": 8.966004754229037e-06,
+      "loss": 0.3016,
+      "step": 169
+    },
+    {
+      "epoch": 0.8762886597938144,
+      "grad_norm": 0.9646361196653148,
+      "learning_rate": 8.947643597831365e-06,
+      "loss": 0.0727,
+      "step": 170
+    },
+    {
+      "epoch": 0.8814432989690721,
+      "grad_norm": 0.9218462690369367,
+      "learning_rate": 8.929140001094734e-06,
+      "loss": 0.1584,
+      "step": 171
+    },
+    {
+      "epoch": 0.8865979381443299,
+      "grad_norm": 0.9892292920559456,
+      "learning_rate": 8.910494631672783e-06,
+      "loss": 0.1449,
+      "step": 172
+    },
+    {
+      "epoch": 0.8917525773195877,
+      "grad_norm": 1.0579410641324796,
+      "learning_rate": 8.891708162334635e-06,
+      "loss": 0.156,
+      "step": 173
+    },
+    {
+      "epoch": 0.8969072164948454,
+      "grad_norm": 0.7647207044009656,
+      "learning_rate": 8.87278127094064e-06,
+      "loss": 0.1107,
+      "step": 174
+    },
+    {
+      "epoch": 0.9020618556701031,
+      "grad_norm": 1.1371579318359795,
+      "learning_rate": 8.853714640417906e-06,
+      "loss": 0.1844,
+      "step": 175
+    },
+    {
+      "epoch": 0.9072164948453608,
+      "grad_norm": 0.7724480629708496,
+      "learning_rate": 8.834508958735656e-06,
+      "loss": 0.0826,
+      "step": 176
+    },
+    {
+      "epoch": 0.9123711340206185,
+      "grad_norm": 1.0272763923631063,
+      "learning_rate": 8.815164918880418e-06,
+      "loss": 0.1678,
+      "step": 177
+    },
+    {
+      "epoch": 0.9175257731958762,
+      "grad_norm": 1.2865546631779095,
+      "learning_rate": 8.795683218831002e-06,
+      "loss": 0.2714,
+      "step": 178
+    },
+    {
+      "epoch": 0.9226804123711341,
+      "grad_norm": 1.8028031060416778,
+      "learning_rate": 8.776064561533329e-06,
+      "loss": 0.2045,
+      "step": 179
+    },
+    {
+      "epoch": 0.9278350515463918,
+      "grad_norm": 1.140464084681919,
+      "learning_rate": 8.756309654875059e-06,
+      "loss": 0.3075,
+      "step": 180
+    },
+    {
+      "epoch": 0.9329896907216495,
+      "grad_norm": 1.1867468428900187,
+      "learning_rate": 8.736419211660054e-06,
+      "loss": 0.2445,
+      "step": 181
+    },
+    {
+      "epoch": 0.9381443298969072,
+      "grad_norm": 1.2294904183688709,
+      "learning_rate": 8.716393949582656e-06,
+      "loss": 0.2416,
+      "step": 182
+    },
+    {
+      "epoch": 0.9432989690721649,
+      "grad_norm": 0.7483697175175423,
+      "learning_rate": 8.696234591201793e-06,
+      "loss": 0.1126,
+      "step": 183
+    },
+    {
+      "epoch": 0.9484536082474226,
+      "grad_norm": 0.9739357119214708,
+      "learning_rate": 8.6759418639149e-06,
+      "loss": 0.167,
+      "step": 184
+    },
+    {
+      "epoch": 0.9536082474226805,
+      "grad_norm": 1.4251293577602404,
+      "learning_rate": 8.655516499931684e-06,
+      "loss": 0.3314,
+      "step": 185
+    },
+    {
+      "epoch": 0.9587628865979382,
+      "grad_norm": 1.140569606339371,
+      "learning_rate": 8.634959236247695e-06,
+      "loss": 0.2206,
+      "step": 186
+    },
+    {
+      "epoch": 0.9639175257731959,
+      "grad_norm": 0.9214436194181765,
+      "learning_rate": 8.61427081461774e-06,
+      "loss": 0.1622,
+      "step": 187
+    },
+    {
+      "epoch": 0.9690721649484536,
+      "grad_norm": 1.4025795614332346,
+      "learning_rate": 8.593451981529109e-06,
+      "loss": 0.2202,
+      "step": 188
+    },
+    {
+      "epoch": 0.9742268041237113,
+      "grad_norm": 0.7327985656704634,
+      "learning_rate": 8.572503488174655e-06,
+      "loss": 0.0798,
+      "step": 189
+    },
+    {
+      "epoch": 0.979381443298969,
+      "grad_norm": 1.1641896784338794,
+      "learning_rate": 8.551426090425678e-06,
+      "loss": 0.1978,
+      "step": 190
+    },
+    {
+      "epoch": 0.9845360824742269,
+      "grad_norm": 1.086213303235639,
+      "learning_rate": 8.53022054880465e-06,
+      "loss": 0.1558,
+      "step": 191
+    },
+    {
+      "epoch": 0.9896907216494846,
+      "grad_norm": 1.0472347005656282,
+      "learning_rate": 8.508887628457783e-06,
+      "loss": 0.2053,
+      "step": 192
+    },
+    {
+      "epoch": 0.9948453608247423,
+      "grad_norm": 0.7910662337183756,
+      "learning_rate": 8.487428099127411e-06,
+      "loss": 0.0927,
+      "step": 193
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1006430197444728,
+      "learning_rate": 8.465842735124224e-06,
+      "loss": 0.1894,
+      "step": 194
+    },
+    {
+      "epoch": 1.0051546391752577,
+      "grad_norm": 0.5906742201713365,
+      "learning_rate": 8.444132315299321e-06,
+      "loss": 0.0537,
+      "step": 195
+    },
+    {
+      "epoch": 1.0103092783505154,
+      "grad_norm": 0.6751125368476019,
+      "learning_rate": 8.422297623016118e-06,
+      "loss": 0.0806,
+      "step": 196
+    },
+    {
+      "epoch": 1.0154639175257731,
+      "grad_norm": 1.0260689629426207,
+      "learning_rate": 8.40033944612207e-06,
+      "loss": 0.1132,
+      "step": 197
+    },
+    {
+      "epoch": 1.0206185567010309,
+      "grad_norm": 0.5551436381218386,
+      "learning_rate": 8.378258576920253e-06,
+      "loss": 0.0523,
+      "step": 198
+    },
+    {
+      "epoch": 1.0257731958762886,
+      "grad_norm": 0.6869421294596499,
+      "learning_rate": 8.356055812140768e-06,
+      "loss": 0.0907,
+      "step": 199
+    },
+    {
+      "epoch": 1.0309278350515463,
+      "grad_norm": 0.9886419404257938,
+      "learning_rate": 8.333731952912e-06,
+      "loss": 0.1147,
+      "step": 200
+    },
+    {
+      "epoch": 1.0360824742268042,
+      "grad_norm": 0.8201901345855567,
+      "learning_rate": 8.311287804731716e-06,
+      "loss": 0.053,
+      "step": 201
+    },
+    {
+      "epoch": 1.041237113402062,
+      "grad_norm": 0.6231211273936398,
+      "learning_rate": 8.288724177437976e-06,
+      "loss": 0.0558,
+      "step": 202
+    },
+    {
+      "epoch": 1.0463917525773196,
+      "grad_norm": 1.0979347326591262,
+      "learning_rate": 8.266041885179949e-06,
+      "loss": 0.1248,
+      "step": 203
+    },
+    {
+      "epoch": 1.0515463917525774,
+      "grad_norm": 1.0441782376710624,
+      "learning_rate": 8.243241746388504e-06,
+      "loss": 0.1024,
+      "step": 204
+    },
+    {
+      "epoch": 1.056701030927835,
+      "grad_norm": 0.8617708313835702,
+      "learning_rate": 8.220324583746697e-06,
+      "loss": 0.0864,
+      "step": 205
+    },
+    {
+      "epoch": 1.0618556701030928,
+      "grad_norm": 0.8219004811721925,
+      "learning_rate": 8.197291224160082e-06,
+      "loss": 0.0965,
+      "step": 206
+    },
+    {
+      "epoch": 1.0670103092783505,
+      "grad_norm": 1.446199673821668,
+      "learning_rate": 8.174142498726875e-06,
+      "loss": 0.1316,
+      "step": 207
+    },
+    {
+      "epoch": 1.0721649484536082,
+      "grad_norm": 0.9970004312056989,
+      "learning_rate": 8.150879242707963e-06,
+      "loss": 0.0785,
+      "step": 208
+    },
+    {
+      "epoch": 1.077319587628866,
+      "grad_norm": 0.6799947961502295,
+      "learning_rate": 8.127502295496768e-06,
+      "loss": 0.0546,
+      "step": 209
+    },
+    {
+      "epoch": 1.0824742268041236,
+      "grad_norm": 0.8459778750809145,
+      "learning_rate": 8.104012500588962e-06,
+      "loss": 0.0754,
+      "step": 210
+    },
+    {
+      "epoch": 1.0876288659793814,
+      "grad_norm": 0.8764373407801908,
+      "learning_rate": 8.080410705552028e-06,
+      "loss": 0.0725,
+      "step": 211
+    },
+    {
+      "epoch": 1.0927835051546393,
+      "grad_norm": 1.4250917004738353,
+      "learning_rate": 8.056697761994679e-06,
+      "loss": 0.1466,
+      "step": 212
+    },
+    {
+      "epoch": 1.097938144329897,
+      "grad_norm": 0.7470601320144464,
+      "learning_rate": 8.032874525536132e-06,
+      "loss": 0.0554,
+      "step": 213
+    },
+    {
+      "epoch": 1.1030927835051547,
+      "grad_norm": 0.7217791614950412,
+      "learning_rate": 8.008941855775228e-06,
+      "loss": 0.104,
+      "step": 214
+    },
+    {
+      "epoch": 1.1082474226804124,
+      "grad_norm": 1.018751301091875,
+      "learning_rate": 7.98490061625943e-06,
+      "loss": 0.1098,
+      "step": 215
+    },
+    {
+      "epoch": 1.1134020618556701,
+      "grad_norm": 0.8577713345259472,
+      "learning_rate": 7.960751674453644e-06,
+      "loss": 0.106,
+      "step": 216
+    },
+    {
+      "epoch": 1.1185567010309279,
+      "grad_norm": 0.7354671182870788,
+      "learning_rate": 7.93649590170894e-06,
+      "loss": 0.0776,
+      "step": 217
+    },
+    {
+      "epoch": 1.1237113402061856,
+      "grad_norm": 0.784080342118224,
+      "learning_rate": 7.912134173231099e-06,
+      "loss": 0.1065,
+      "step": 218
+    },
+    {
+      "epoch": 1.1288659793814433,
+      "grad_norm": 1.0447979386760058,
+      "learning_rate": 7.887667368049028e-06,
+      "loss": 0.1501,
+      "step": 219
+    },
+    {
+      "epoch": 1.134020618556701,
+      "grad_norm": 0.8251703086624239,
+      "learning_rate": 7.863096368983061e-06,
+      "loss": 0.0888,
+      "step": 220
+    },
+    {
+      "epoch": 1.1391752577319587,
+      "grad_norm": 0.8822362088146911,
+      "learning_rate": 7.838422062613088e-06,
+      "loss": 0.093,
+      "step": 221
+    },
+    {
+      "epoch": 1.1443298969072164,
+      "grad_norm": 0.8949834524497724,
+      "learning_rate": 7.813645339246578e-06,
+      "loss": 0.0921,
+      "step": 222
+    },
+    {
+      "epoch": 1.1494845360824741,
+      "grad_norm": 0.5374131970068409,
+      "learning_rate": 7.78876709288644e-06,
+      "loss": 0.0386,
+      "step": 223
+    },
+    {
+      "epoch": 1.1546391752577319,
+      "grad_norm": 0.7564742772338368,
+      "learning_rate": 7.763788221198775e-06,
+      "loss": 0.0948,
+      "step": 224
+    },
+    {
+      "epoch": 1.1597938144329896,
+      "grad_norm": 0.8808313875725176,
+      "learning_rate": 7.738709625480494e-06,
+      "loss": 0.092,
+      "step": 225
+    },
+    {
+      "epoch": 1.1649484536082475,
+      "grad_norm": 0.6478248778597727,
+      "learning_rate": 7.713532210626771e-06,
+      "loss": 0.0458,
+      "step": 226
+    },
+    {
+      "epoch": 1.1701030927835052,
+      "grad_norm": 0.9288097880385096,
+      "learning_rate": 7.68825688509842e-06,
+      "loss": 0.0894,
+      "step": 227
+    },
+    {
+      "epoch": 1.175257731958763,
+      "grad_norm": 0.714867355071131,
+      "learning_rate": 7.662884560889106e-06,
+      "loss": 0.0541,
+      "step": 228
+    },
+    {
+      "epoch": 1.1804123711340206,
+      "grad_norm": 0.6642395918740392,
+      "learning_rate": 7.637416153492426e-06,
+      "loss": 0.0663,
+      "step": 229
+    },
+    {
+      "epoch": 1.1855670103092784,
+      "grad_norm": 0.9527549900666177,
+      "learning_rate": 7.611852581868895e-06,
+      "loss": 0.1443,
+      "step": 230
+    },
+    {
+      "epoch": 1.190721649484536,
+      "grad_norm": 0.7893883831371513,
+      "learning_rate": 7.586194768412778e-06,
+      "loss": 0.0872,
+      "step": 231
+    },
+    {
+      "epoch": 1.1958762886597938,
+      "grad_norm": 0.9884146330428806,
+      "learning_rate": 7.560443638918801e-06,
+      "loss": 0.0821,
+      "step": 232
+    },
+    {
+      "epoch": 1.2010309278350515,
+      "grad_norm": 0.7413524893904028,
+      "learning_rate": 7.534600122548766e-06,
+      "loss": 0.1001,
+      "step": 233
+    },
+    {
+      "epoch": 1.2061855670103092,
+      "grad_norm": 0.6487178792290845,
+      "learning_rate": 7.508665151798e-06,
+      "loss": 0.0669,
+      "step": 234
+    },
+    {
+      "epoch": 1.211340206185567,
+      "grad_norm": 1.2033223139705194,
+      "learning_rate": 7.482639662461731e-06,
+      "loss": 0.1102,
+      "step": 235
+    },
+    {
+      "epoch": 1.2164948453608249,
+      "grad_norm": 1.1496624331229564,
+      "learning_rate": 7.456524593601306e-06,
+      "loss": 0.0817,
+      "step": 236
+    },
+    {
+      "epoch": 1.2216494845360826,
+      "grad_norm": 0.8048271397849805,
+      "learning_rate": 7.430320887510319e-06,
+      "loss": 0.066,
+      "step": 237
+    },
+    {
+      "epoch": 1.2268041237113403,
+      "grad_norm": 1.0865257986253214,
+      "learning_rate": 7.404029489680597e-06,
+      "loss": 0.1054,
+      "step": 238
+    },
+    {
+      "epoch": 1.231958762886598,
+      "grad_norm": 0.9118222773920444,
+      "learning_rate": 7.377651348768102e-06,
+      "loss": 0.0734,
+      "step": 239
+    },
+    {
+      "epoch": 1.2371134020618557,
+      "grad_norm": 0.7111948595434698,
+      "learning_rate": 7.351187416558686e-06,
+      "loss": 0.0455,
+      "step": 240
+    },
+    {
+      "epoch": 1.2422680412371134,
+      "grad_norm": 1.1586255328358697,
+      "learning_rate": 7.324638647933756e-06,
+      "loss": 0.1247,
+      "step": 241
+    },
+    {
+      "epoch": 1.2474226804123711,
+      "grad_norm": 0.7885336165369894,
+      "learning_rate": 7.29800600083582e-06,
+      "loss": 0.0944,
+      "step": 242
+    },
+    {
+      "epoch": 1.2525773195876289,
+      "grad_norm": 0.815328204346094,
+      "learning_rate": 7.2712904362339155e-06,
+      "loss": 0.0846,
+      "step": 243
+    },
+    {
+      "epoch": 1.2577319587628866,
+      "grad_norm": 1.044831188842817,
+      "learning_rate": 7.244492918088946e-06,
+      "loss": 0.1029,
+      "step": 244
+    },
+    {
+      "epoch": 1.2628865979381443,
+      "grad_norm": 1.1074505493277451,
+      "learning_rate": 7.217614413318887e-06,
+      "loss": 0.1109,
+      "step": 245
+    },
+    {
+      "epoch": 1.268041237113402,
+      "grad_norm": 0.836136628649968,
+      "learning_rate": 7.19065589176391e-06,
+      "loss": 0.1074,
+      "step": 246
+    },
+    {
+      "epoch": 1.2731958762886597,
+      "grad_norm": 0.9806183729514004,
+      "learning_rate": 7.1636183261513784e-06,
+      "loss": 0.1046,
+      "step": 247
+    },
+    {
+      "epoch": 1.2783505154639174,
+      "grad_norm": 0.8666879067703104,
+      "learning_rate": 7.136502692060746e-06,
+      "loss": 0.0826,
+      "step": 248
+    },
+    {
+      "epoch": 1.2835051546391751,
+      "grad_norm": 0.801232774030085,
+      "learning_rate": 7.109309967888376e-06,
+      "loss": 0.0556,
+      "step": 249
+    },
+    {
+      "epoch": 1.2886597938144329,
+      "grad_norm": 1.1897623094936245,
+      "learning_rate": 7.0820411348122144e-06,
+      "loss": 0.154,
+      "step": 250
+    },
+    {
+      "epoch": 1.2938144329896908,
+      "grad_norm": 0.8276583732971441,
+      "learning_rate": 7.0546971767564e-06,
+      "loss": 0.0833,
+      "step": 251
+    },
+    {
+      "epoch": 1.2989690721649485,
+      "grad_norm": 0.7583663543944267,
+      "learning_rate": 7.027279080355756e-06,
+      "loss": 0.0499,
+      "step": 252
+    },
+    {
+      "epoch": 1.3041237113402062,
+      "grad_norm": 0.8649391169592322,
+      "learning_rate": 6.999787834920202e-06,
+      "loss": 0.0867,
+      "step": 253
+    },
+    {
+      "epoch": 1.309278350515464,
+      "grad_norm": 1.0915318471452742,
+      "learning_rate": 6.972224432399038e-06,
+      "loss": 0.0676,
+      "step": 254
+    },
+    {
+      "epoch": 1.3144329896907216,
+      "grad_norm": 1.0231154625513728,
+      "learning_rate": 6.9445898673451635e-06,
+      "loss": 0.1028,
+      "step": 255
+    },
+    {
+      "epoch": 1.3195876288659794,
+      "grad_norm": 0.933663875875191,
+      "learning_rate": 6.916885136879197e-06,
+      "loss": 0.0967,
+      "step": 256
+    },
+    {
+      "epoch": 1.324742268041237,
+      "grad_norm": 0.8912988511114369,
+      "learning_rate": 6.889111240653488e-06,
+      "loss": 0.0959,
+      "step": 257
+    },
+    {
+      "epoch": 1.3298969072164948,
+      "grad_norm": 0.8687955070867821,
+      "learning_rate": 6.861269180816052e-06,
+      "loss": 0.0977,
+      "step": 258
+    },
+    {
+      "epoch": 1.3350515463917525,
+      "grad_norm": 0.7106326271334324,
+      "learning_rate": 6.833359961974406e-06,
+      "loss": 0.0869,
+      "step": 259
+    },
+    {
+      "epoch": 1.3402061855670104,
+      "grad_norm": 0.9501736705275039,
+      "learning_rate": 6.805384591159325e-06,
+      "loss": 0.0949,
+      "step": 260
+    },
+    {
+      "epoch": 1.3453608247422681,
+      "grad_norm": 0.804663253737915,
+      "learning_rate": 6.7773440777885055e-06,
+      "loss": 0.0811,
+      "step": 261
+    },
+    {
+      "epoch": 1.3505154639175259,
+      "grad_norm": 0.6042006243936727,
+      "learning_rate": 6.749239433630137e-06,
+      "loss": 0.0537,
+      "step": 262
+    },
+    {
+      "epoch": 1.3556701030927836,
+      "grad_norm": 1.1315557610677505,
+      "learning_rate": 6.721071672766407e-06,
+      "loss": 0.1242,
+      "step": 263
+    },
+    {
+      "epoch": 1.3608247422680413,
+      "grad_norm": 0.664551606971833,
+      "learning_rate": 6.6928418115568994e-06,
+      "loss": 0.05,
+      "step": 264
+    },
+    {
+      "epoch": 1.365979381443299,
+      "grad_norm": 0.6628232247101428,
+      "learning_rate": 6.6645508686019225e-06,
+      "loss": 0.0709,
+      "step": 265
+    },
+    {
+      "epoch": 1.3711340206185567,
+      "grad_norm": 1.3704143398811302,
+      "learning_rate": 6.636199864705766e-06,
+      "loss": 0.1788,
+      "step": 266
+    },
+    {
+      "epoch": 1.3762886597938144,
+      "grad_norm": 0.653208777448595,
+      "learning_rate": 6.607789822839855e-06,
+      "loss": 0.0605,
+      "step": 267
+    },
+    {
+      "epoch": 1.3814432989690721,
+      "grad_norm": 0.7238199007998108,
+      "learning_rate": 6.579321768105845e-06,
+      "loss": 0.081,
+      "step": 268
+    },
+    {
+      "epoch": 1.3865979381443299,
+      "grad_norm": 0.7205533607672164,
+      "learning_rate": 6.550796727698639e-06,
+      "loss": 0.067,
+      "step": 269
+    },
+    {
+      "epoch": 1.3917525773195876,
+      "grad_norm": 0.6700885764195117,
+      "learning_rate": 6.52221573086931e-06,
+      "loss": 0.0503,
+      "step": 270
+    },
+    {
+      "epoch": 1.3969072164948453,
+      "grad_norm": 1.219456301028477,
+      "learning_rate": 6.493579808887976e-06,
+      "loss": 0.1329,
+      "step": 271
+    },
+    {
+      "epoch": 1.402061855670103,
+      "grad_norm": 1.327064144446766,
+      "learning_rate": 6.4648899950065865e-06,
+      "loss": 0.1581,
+      "step": 272
+    },
+    {
+      "epoch": 1.4072164948453607,
+      "grad_norm": 0.8319301363228734,
+      "learning_rate": 6.436147324421635e-06,
+      "loss": 0.0867,
+      "step": 273
+    },
+    {
+      "epoch": 1.4123711340206184,
+      "grad_norm": 1.1371748524536707,
+      "learning_rate": 6.407352834236807e-06,
+      "loss": 0.1496,
+      "step": 274
+    },
+    {
+      "epoch": 1.4175257731958764,
+      "grad_norm": 0.856128052640479,
+      "learning_rate": 6.378507563425571e-06,
+      "loss": 0.0804,
+      "step": 275
+    },
+    {
+      "epoch": 1.422680412371134,
+      "grad_norm": 1.062404018021394,
+      "learning_rate": 6.349612552793675e-06,
+      "loss": 0.1043,
+      "step": 276
+    },
+    {
+      "epoch": 1.4278350515463918,
+      "grad_norm": 0.9101560194006858,
+      "learning_rate": 6.320668844941598e-06,
+      "loss": 0.1002,
+      "step": 277
+    },
+    {
+      "epoch": 1.4329896907216495,
+      "grad_norm": 1.2757033041145538,
+      "learning_rate": 6.291677484226929e-06,
+      "loss": 0.1098,
+      "step": 278
+    },
+    {
+      "epoch": 1.4381443298969072,
+      "grad_norm": 1.2846837622920635,
+      "learning_rate": 6.26263951672669e-06,
+      "loss": 0.1197,
+      "step": 279
+    },
+    {
+      "epoch": 1.443298969072165,
+      "grad_norm": 0.6126591325904064,
+      "learning_rate": 6.233555990199583e-06,
+      "loss": 0.0526,
+      "step": 280
+    },
+    {
+      "epoch": 1.4484536082474226,
+      "grad_norm": 0.9201467393931315,
+      "learning_rate": 6.204427954048186e-06,
+      "loss": 0.095,
+      "step": 281
+    },
+    {
+      "epoch": 1.4536082474226804,
+      "grad_norm": 0.8709260663548162,
+      "learning_rate": 6.175256459281093e-06,
+      "loss": 0.0956,
+      "step": 282
+    },
+    {
+      "epoch": 1.458762886597938,
+      "grad_norm": 0.8394883872798908,
+      "learning_rate": 6.146042558474987e-06,
+      "loss": 0.0782,
+      "step": 283
+    },
+    {
+      "epoch": 1.463917525773196,
+      "grad_norm": 0.7647943393509891,
+      "learning_rate": 6.116787305736659e-06,
+      "loss": 0.0964,
+      "step": 284
+    },
+    {
+      "epoch": 1.4690721649484537,
+      "grad_norm": 1.1981279245762775,
+      "learning_rate": 6.087491756664982e-06,
+      "loss": 0.1251,
+      "step": 285
+    },
+    {
+      "epoch": 1.4742268041237114,
+      "grad_norm": 0.7281773116566167,
+      "learning_rate": 6.058156968312808e-06,
+      "loss": 0.0636,
+      "step": 286
+    },
+    {
+      "epoch": 1.4793814432989691,
+      "grad_norm": 0.9415282186895138,
+      "learning_rate": 6.028783999148841e-06,
+      "loss": 0.121,
+      "step": 287
+    },
+    {
+      "epoch": 1.4845360824742269,
+      "grad_norm": 0.8155749049438948,
+      "learning_rate": 5.999373909019437e-06,
+      "loss": 0.0607,
+      "step": 288
+    },
+    {
+      "epoch": 1.4896907216494846,
+      "grad_norm": 0.7639944315517929,
+      "learning_rate": 5.9699277591103665e-06,
+      "loss": 0.0548,
+      "step": 289
+    },
+    {
+      "epoch": 1.4948453608247423,
+      "grad_norm": 0.8636520383698243,
+      "learning_rate": 5.940446611908519e-06,
+      "loss": 0.0979,
+      "step": 290
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5326124314349575,
+      "learning_rate": 5.91093153116357e-06,
+      "loss": 0.0373,
+      "step": 291
+    },
+    {
+      "epoch": 1.5051546391752577,
+      "grad_norm": 0.8280155720382041,
+      "learning_rate": 5.881383581849601e-06,
+      "loss": 0.0698,
+      "step": 292
+    },
+    {
+      "epoch": 1.5103092783505154,
+      "grad_norm": 0.9092824338400287,
+      "learning_rate": 5.851803830126666e-06,
+      "loss": 0.1016,
+      "step": 293
+    },
+    {
+      "epoch": 1.5154639175257731,
+      "grad_norm": 1.2564418341193255,
+      "learning_rate": 5.822193343302328e-06,
+      "loss": 0.1473,
+      "step": 294
+    },
+    {
+      "epoch": 1.5206185567010309,
+      "grad_norm": 0.8983314520281567,
+      "learning_rate": 5.792553189793141e-06,
+      "loss": 0.1071,
+      "step": 295
+    },
+    {
+      "epoch": 1.5257731958762886,
+      "grad_norm": 0.9871176927471953,
+      "learning_rate": 5.762884439086108e-06,
+      "loss": 0.1235,
+      "step": 296
+    },
+    {
+      "epoch": 1.5309278350515463,
+      "grad_norm": 1.0212569217738547,
+      "learning_rate": 5.733188161700084e-06,
+      "loss": 0.1144,
+      "step": 297
+    },
+    {
+      "epoch": 1.536082474226804,
+      "grad_norm": 1.0505577277639713,
+      "learning_rate": 5.703465429147153e-06,
+      "loss": 0.0856,
+      "step": 298
+    },
+    {
+      "epoch": 1.5412371134020617,
+      "grad_norm": 0.9198501514769984,
+      "learning_rate": 5.673717313893963e-06,
+      "loss": 0.0857,
+      "step": 299
+    },
+    {
+      "epoch": 1.5463917525773194,
+      "grad_norm": 0.4413128663709364,
+      "learning_rate": 5.643944889323031e-06,
+      "loss": 0.0339,
+      "step": 300
+    },
+    {
+      "epoch": 1.5515463917525774,
+      "grad_norm": 0.9176541299396651,
+      "learning_rate": 5.6141492296940104e-06,
+      "loss": 0.0883,
+      "step": 301
+    },
+    {
+      "epoch": 1.556701030927835,
+      "grad_norm": 0.4947875542348694,
+      "learning_rate": 5.584331410104934e-06,
+      "loss": 0.0282,
+      "step": 302
+    },
+    {
+      "epoch": 1.5618556701030928,
+      "grad_norm": 1.1769881441147931,
+      "learning_rate": 5.554492506453415e-06,
+      "loss": 0.1418,
+      "step": 303
+    },
+    {
+      "epoch": 1.5670103092783505,
+      "grad_norm": 0.8273504464851464,
+      "learning_rate": 5.524633595397829e-06,
+      "loss": 0.0639,
+      "step": 304
+    },
+    {
+      "epoch": 1.5721649484536082,
+      "grad_norm": 1.104084036749989,
+      "learning_rate": 5.494755754318472e-06,
+      "loss": 0.1496,
+      "step": 305
+    },
+    {
+      "epoch": 1.577319587628866,
+      "grad_norm": 0.722149475333966,
+      "learning_rate": 5.464860061278673e-06,
+      "loss": 0.06,
+      "step": 306
+    },
+    {
+      "epoch": 1.5824742268041239,
+      "grad_norm": 0.765678640808434,
+      "learning_rate": 5.434947594985903e-06,
+      "loss": 0.064,
+      "step": 307
+    },
+    {
+      "epoch": 1.5876288659793816,
+      "grad_norm": 0.8133799235994497,
+      "learning_rate": 5.40501943475286e-06,
+      "loss": 0.0834,
+      "step": 308
+    },
+    {
+      "epoch": 1.5927835051546393,
+      "grad_norm": 1.0133768144130846,
+      "learning_rate": 5.375076660458503e-06,
+      "loss": 0.1211,
+      "step": 309
+    },
+    {
+      "epoch": 1.597938144329897,
+      "grad_norm": 0.8911686618887108,
+      "learning_rate": 5.345120352509114e-06,
+      "loss": 0.1135,
+      "step": 310
+    },
+    {
+      "epoch": 1.6030927835051547,
+      "grad_norm": 0.9837172309407369,
+      "learning_rate": 5.315151591799293e-06,
+      "loss": 0.0939,
+      "step": 311
+    },
+    {
+      "epoch": 1.6082474226804124,
+      "grad_norm": 0.9603201029901375,
+      "learning_rate": 5.28517145967297e-06,
+      "loss": 0.1191,
+      "step": 312
+    },
+    {
+      "epoch": 1.6134020618556701,
+      "grad_norm": 0.7568038226377672,
+      "learning_rate": 5.255181037884377e-06,
+      "loss": 0.0878,
+      "step": 313
+    },
+    {
+      "epoch": 1.6185567010309279,
+      "grad_norm": 0.7396794148405889,
+      "learning_rate": 5.225181408559028e-06,
+      "loss": 0.0713,
+      "step": 314
+    },
+    {
+      "epoch": 1.6237113402061856,
+      "grad_norm": 0.8098683954087241,
+      "learning_rate": 5.195173654154662e-06,
+      "loss": 0.06,
+      "step": 315
+    },
+    {
+      "epoch": 1.6288659793814433,
+      "grad_norm": 0.9687294980139846,
+      "learning_rate": 5.165158857422191e-06,
+      "loss": 0.0856,
+      "step": 316
+    },
+    {
+      "epoch": 1.634020618556701,
+      "grad_norm": 0.991898312261934,
+      "learning_rate": 5.135138101366633e-06,
+      "loss": 0.1205,
+      "step": 317
+    },
+    {
+      "epoch": 1.6391752577319587,
+      "grad_norm": 0.5573665413209947,
+      "learning_rate": 5.105112469208032e-06,
+      "loss": 0.0497,
+      "step": 318
+    },
+    {
+      "epoch": 1.6443298969072164,
+      "grad_norm": 1.0382601563430964,
+      "learning_rate": 5.075083044342371e-06,
+      "loss": 0.0852,
+      "step": 319
+    },
+    {
+      "epoch": 1.6494845360824741,
+      "grad_norm": 0.9967396087063322,
+      "learning_rate": 5.045050910302485e-06,
+      "loss": 0.0877,
+      "step": 320
+    },
+    {
+      "epoch": 1.6546391752577319,
+      "grad_norm": 0.7962058373120292,
+      "learning_rate": 5.015017150718961e-06,
+      "loss": 0.0837,
+      "step": 321
+    },
+    {
+      "epoch": 1.6597938144329896,
+      "grad_norm": 0.6489956586111506,
+      "learning_rate": 4.984982849281041e-06,
+      "loss": 0.0394,
+      "step": 322
+    },
+    {
+      "epoch": 1.6649484536082473,
+      "grad_norm": 0.6615424807101872,
+      "learning_rate": 4.9549490896975175e-06,
+      "loss": 0.0507,
+      "step": 323
+    },
+    {
+      "epoch": 1.670103092783505,
+      "grad_norm": 1.0065188138819108,
+      "learning_rate": 4.92491695565763e-06,
+      "loss": 0.0913,
+      "step": 324
+    },
+    {
+      "epoch": 1.675257731958763,
+      "grad_norm": 0.9338664154455895,
+      "learning_rate": 4.894887530791968e-06,
+      "loss": 0.0829,
+      "step": 325
+    },
+    {
+      "epoch": 1.6804123711340206,
+      "grad_norm": 0.8619735255759078,
+      "learning_rate": 4.864861898633367e-06,
+      "loss": 0.1015,
+      "step": 326
+    },
+    {
+      "epoch": 1.6855670103092784,
+      "grad_norm": 1.1629690007707925,
+      "learning_rate": 4.8348411425778105e-06,
+      "loss": 0.1093,
+      "step": 327
+    },
+    {
+      "epoch": 1.690721649484536,
+      "grad_norm": 0.8508681932219984,
+      "learning_rate": 4.80482634584534e-06,
+      "loss": 0.0787,
+      "step": 328
+    },
+    {
+      "epoch": 1.6958762886597938,
+      "grad_norm": 0.7692449841056024,
+      "learning_rate": 4.774818591440974e-06,
+      "loss": 0.0958,
+      "step": 329
+    },
+    {
+      "epoch": 1.7010309278350515,
+      "grad_norm": 0.8553151177506546,
+      "learning_rate": 4.744818962115624e-06,
+      "loss": 0.082,
+      "step": 330
+    },
+    {
+      "epoch": 1.7061855670103094,
+      "grad_norm": 0.5903484968058315,
+      "learning_rate": 4.714828540327033e-06,
+      "loss": 0.0462,
+      "step": 331
+    },
+    {
+      "epoch": 1.7113402061855671,
+      "grad_norm": 0.8864451372917644,
+      "learning_rate": 4.684848408200707e-06,
+      "loss": 0.0714,
+      "step": 332
+    },
+    {
+      "epoch": 1.7164948453608249,
+      "grad_norm": 0.8860344350682171,
+      "learning_rate": 4.654879647490887e-06,
+      "loss": 0.0691,
+      "step": 333
+    },
+    {
+      "epoch": 1.7216494845360826,
+      "grad_norm": 0.8183534311660873,
+      "learning_rate": 4.624923339541498e-06,
+      "loss": 0.0633,
+      "step": 334
+    },
+    {
+      "epoch": 1.7268041237113403,
+      "grad_norm": 0.614568204141328,
+      "learning_rate": 4.594980565247143e-06,
+      "loss": 0.0495,
+      "step": 335
+    },
+    {
+      "epoch": 1.731958762886598,
+      "grad_norm": 0.9930144229307591,
+      "learning_rate": 4.565052405014098e-06,
+      "loss": 0.1059,
+      "step": 336
+    },
+    {
+      "epoch": 1.7371134020618557,
+      "grad_norm": 0.7592312136391701,
+      "learning_rate": 4.5351399387213305e-06,
+      "loss": 0.0618,
+      "step": 337
+    },
+    {
+      "epoch": 1.7422680412371134,
+      "grad_norm": 1.4380591077584624,
+      "learning_rate": 4.5052442456815294e-06,
+      "loss": 0.0997,
+      "step": 338
+    },
+    {
+      "epoch": 1.7474226804123711,
+      "grad_norm": 0.9459809494967879,
+      "learning_rate": 4.47536640460217e-06,
+      "loss": 0.1004,
+      "step": 339
+    },
+    {
+      "epoch": 1.7525773195876289,
+      "grad_norm": 0.7417698139276238,
+      "learning_rate": 4.445507493546586e-06,
+      "loss": 0.0805,
+      "step": 340
+    },
+    {
+      "epoch": 1.7577319587628866,
+      "grad_norm": 0.8768339149771931,
+      "learning_rate": 4.4156685898950676e-06,
+      "loss": 0.0892,
+      "step": 341
+    },
+    {
+      "epoch": 1.7628865979381443,
+      "grad_norm": 0.5753560085161841,
+      "learning_rate": 4.385850770305991e-06,
+      "loss": 0.0546,
+      "step": 342
+    },
+    {
+      "epoch": 1.768041237113402,
+      "grad_norm": 0.6203625417079575,
+      "learning_rate": 4.356055110676971e-06,
+      "loss": 0.0557,
+      "step": 343
+    },
+    {
+      "epoch": 1.7731958762886597,
+      "grad_norm": 1.290592433705904,
+      "learning_rate": 4.326282686106039e-06,
+      "loss": 0.0924,
+      "step": 344
+    },
+    {
+      "epoch": 1.7783505154639174,
+      "grad_norm": 0.8501802343729535,
+      "learning_rate": 4.296534570852848e-06,
+      "loss": 0.0976,
+      "step": 345
+    },
+    {
+      "epoch": 1.7835051546391751,
+      "grad_norm": 0.7589958862583852,
+      "learning_rate": 4.266811838299916e-06,
+      "loss": 0.0616,
+      "step": 346
+    },
+    {
+      "epoch": 1.7886597938144329,
+      "grad_norm": 0.6634970203515521,
+      "learning_rate": 4.237115560913894e-06,
+      "loss": 0.0547,
+      "step": 347
+    },
+    {
+      "epoch": 1.7938144329896906,
+      "grad_norm": 0.7377540659991918,
+      "learning_rate": 4.20744681020686e-06,
+      "loss": 0.0562,
+      "step": 348
+    },
+    {
+      "epoch": 1.7989690721649485,
+      "grad_norm": 1.0154222900360976,
+      "learning_rate": 4.1778066566976736e-06,
+      "loss": 0.0889,
+      "step": 349
+    },
+    {
+      "epoch": 1.8041237113402062,
+      "grad_norm": 1.1675640094076947,
+      "learning_rate": 4.148196169873335e-06,
+      "loss": 0.1241,
+      "step": 350
+    },
+    {
+      "epoch": 1.809278350515464,
+      "grad_norm": 0.5783006867375596,
+      "learning_rate": 4.118616418150398e-06,
+      "loss": 0.0371,
+      "step": 351
+    },
+    {
+      "epoch": 1.8144329896907216,
+      "grad_norm": 1.067759818874075,
+      "learning_rate": 4.089068468836431e-06,
+      "loss": 0.1045,
+      "step": 352
+    },
+    {
+      "epoch": 1.8195876288659794,
+      "grad_norm": 0.8301529919273785,
+      "learning_rate": 4.059553388091483e-06,
+      "loss": 0.0907,
+      "step": 353
+    },
+    {
+      "epoch": 1.824742268041237,
+      "grad_norm": 0.8633283627168938,
+      "learning_rate": 4.030072240889635e-06,
+      "loss": 0.0883,
+      "step": 354
+    },
+    {
+      "epoch": 1.829896907216495,
+      "grad_norm": 1.0045407842411636,
+      "learning_rate": 4.000626090980564e-06,
+      "loss": 0.109,
+      "step": 355
+    },
+    {
+      "epoch": 1.8350515463917527,
+      "grad_norm": 0.6176720353760472,
+      "learning_rate": 3.971216000851161e-06,
+      "loss": 0.046,
+      "step": 356
+    },
+    {
+      "epoch": 1.8402061855670104,
+      "grad_norm": 0.6741212952732083,
+      "learning_rate": 3.941843031687194e-06,
+      "loss": 0.0616,
+      "step": 357
+    },
+    {
+      "epoch": 1.8453608247422681,
+      "grad_norm": 0.786429607982951,
+      "learning_rate": 3.912508243335019e-06,
+      "loss": 0.0814,
+      "step": 358
+    },
+    {
+      "epoch": 1.8505154639175259,
+      "grad_norm": 0.7681340177017846,
+      "learning_rate": 3.883212694263341e-06,
+      "loss": 0.0413,
+      "step": 359
+    },
+    {
+      "epoch": 1.8556701030927836,
+      "grad_norm": 0.9188584907010531,
+      "learning_rate": 3.853957441525014e-06,
+      "loss": 0.0917,
+      "step": 360
+    },
+    {
+      "epoch": 1.8608247422680413,
+      "grad_norm": 0.7339463471541188,
+      "learning_rate": 3.824743540718909e-06,
+      "loss": 0.0518,
+      "step": 361
+    },
+    {
+      "epoch": 1.865979381443299,
+      "grad_norm": 0.7505002518495965,
+      "learning_rate": 3.7955720459518163e-06,
+      "loss": 0.0611,
+      "step": 362
+    },
+    {
+      "epoch": 1.8711340206185567,
+      "grad_norm": 0.6977481350680906,
+      "learning_rate": 3.7664440098004194e-06,
+      "loss": 0.0662,
+      "step": 363
+    },
+    {
+      "epoch": 1.8762886597938144,
+      "grad_norm": 0.9305995047119137,
+      "learning_rate": 3.7373604832733103e-06,
+      "loss": 0.1134,
+      "step": 364
+    },
+    {
+      "epoch": 1.8814432989690721,
+      "grad_norm": 0.838945380538262,
+      "learning_rate": 3.708322515773071e-06,
+      "loss": 0.0848,
+      "step": 365
+    },
+    {
+      "epoch": 1.8865979381443299,
+      "grad_norm": 0.950916164653827,
+      "learning_rate": 3.6793311550584043e-06,
+      "loss": 0.0848,
+      "step": 366
+    },
+    {
+      "epoch": 1.8917525773195876,
+      "grad_norm": 0.8218742627036616,
+      "learning_rate": 3.6503874472063268e-06,
+      "loss": 0.1023,
+      "step": 367
+    },
+    {
+      "epoch": 1.8969072164948453,
+      "grad_norm": 0.9078783194850608,
+      "learning_rate": 3.62149243657443e-06,
+      "loss": 0.0931,
+      "step": 368
+    },
+    {
+      "epoch": 1.902061855670103,
+      "grad_norm": 0.688660913780332,
+      "learning_rate": 3.5926471657631945e-06,
+      "loss": 0.0589,
+      "step": 369
+    },
+    {
+      "epoch": 1.9072164948453607,
+      "grad_norm": 0.6952675588410017,
+      "learning_rate": 3.563852675578368e-06,
+      "loss": 0.0502,
+      "step": 370
+    },
+    {
+      "epoch": 1.9123711340206184,
+      "grad_norm": 1.3884427251376699,
+      "learning_rate": 3.535110004993414e-06,
+      "loss": 0.1733,
+      "step": 371
+    },
+    {
+      "epoch": 1.9175257731958761,
+      "grad_norm": 1.0258997166309576,
+      "learning_rate": 3.506420191112023e-06,
+      "loss": 0.1182,
+      "step": 372
+    },
+    {
+      "epoch": 1.922680412371134,
+      "grad_norm": 0.6919148043225545,
+      "learning_rate": 3.477784269130691e-06,
+      "loss": 0.07,
+      "step": 373
+    },
+    {
+      "epoch": 1.9278350515463918,
+      "grad_norm": 0.9093312963775791,
+      "learning_rate": 3.449203272301362e-06,
+      "loss": 0.0967,
+      "step": 374
+    },
+    {
+      "epoch": 1.9329896907216495,
+      "grad_norm": 0.8167616489115295,
+      "learning_rate": 3.4206782318941556e-06,
+      "loss": 0.0586,
+      "step": 375
+    },
+    {
+      "epoch": 1.9381443298969072,
+      "grad_norm": 0.8922256610004783,
+      "learning_rate": 3.3922101771601475e-06,
+      "loss": 0.0716,
+      "step": 376
+    },
+    {
+      "epoch": 1.943298969072165,
+      "grad_norm": 1.0215655504364405,
+      "learning_rate": 3.363800135294236e-06,
+      "loss": 0.1191,
+      "step": 377
+    },
+    {
+      "epoch": 1.9484536082474226,
+      "grad_norm": 1.3202130028849257,
+      "learning_rate": 3.3354491313980774e-06,
+      "loss": 0.1322,
+      "step": 378
+    },
+    {
+      "epoch": 1.9536082474226806,
+      "grad_norm": 1.0653863680738789,
+      "learning_rate": 3.3071581884431014e-06,
+      "loss": 0.1362,
+      "step": 379
+    },
+    {
+      "epoch": 1.9587628865979383,
+      "grad_norm": 0.7846807384773636,
+      "learning_rate": 3.2789283272335935e-06,
+      "loss": 0.0823,
+      "step": 380
+    },
+    {
+      "epoch": 1.963917525773196,
+      "grad_norm": 1.005891449118457,
+      "learning_rate": 3.250760566369864e-06,
+      "loss": 0.1152,
+      "step": 381
+    },
+    {
+      "epoch": 1.9690721649484537,
+      "grad_norm": 0.8386229662220965,
+      "learning_rate": 3.2226559222114974e-06,
+      "loss": 0.0743,
+      "step": 382
+    },
+    {
+      "epoch": 1.9742268041237114,
+      "grad_norm": 0.8015266394615227,
+      "learning_rate": 3.194615408840678e-06,
+      "loss": 0.0818,
+      "step": 383
+    },
+    {
+      "epoch": 1.9793814432989691,
+      "grad_norm": 1.1193048769312743,
+      "learning_rate": 3.1666400380255944e-06,
+      "loss": 0.156,
+      "step": 384
+    },
+    {
+      "epoch": 1.9845360824742269,
+      "grad_norm": 0.8507066467744703,
+      "learning_rate": 3.1387308191839495e-06,
+      "loss": 0.0967,
+      "step": 385
+    },
+    {
+      "epoch": 1.9896907216494846,
+      "grad_norm": 0.6565071516482492,
+      "learning_rate": 3.110888759346512e-06,
+      "loss": 0.0469,
+      "step": 386
+    },
+    {
+      "epoch": 1.9948453608247423,
+      "grad_norm": 0.7510076309546407,
+      "learning_rate": 3.0831148631208043e-06,
+      "loss": 0.0607,
+      "step": 387
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9275588340131792,
+      "learning_rate": 3.055410132654838e-06,
+      "loss": 0.0939,
+      "step": 388
+    },
+    {
+      "epoch": 2.0051546391752577,
+      "grad_norm": 0.7961170755671138,
+      "learning_rate": 3.0277755676009657e-06,
+      "loss": 0.0571,
+      "step": 389
+    },
+    {
+      "epoch": 2.0103092783505154,
+      "grad_norm": 1.0079482379255948,
+      "learning_rate": 3.0002121650798e-06,
+      "loss": 0.0623,
+      "step": 390
+    },
+    {
+      "epoch": 2.015463917525773,
+      "grad_norm": 0.6835130843362455,
+      "learning_rate": 2.9727209196442436e-06,
+      "loss": 0.053,
+      "step": 391
+    },
+    {
+      "epoch": 2.020618556701031,
+      "grad_norm": 0.5487177394190583,
+      "learning_rate": 2.945302823243602e-06,
+      "loss": 0.0344,
+      "step": 392
+    },
+    {
+      "epoch": 2.0257731958762886,
+      "grad_norm": 0.6405255254402394,
+      "learning_rate": 2.9179588651877877e-06,
+      "loss": 0.0404,
+      "step": 393
+    },
+    {
+      "epoch": 2.0309278350515463,
+      "grad_norm": 0.6455275267472669,
+      "learning_rate": 2.890690032111625e-06,
+      "loss": 0.0438,
+      "step": 394
+    },
+    {
+      "epoch": 2.036082474226804,
+      "grad_norm": 0.632307449036581,
+      "learning_rate": 2.863497307939255e-06,
+      "loss": 0.0438,
+      "step": 395
+    },
+    {
+      "epoch": 2.0412371134020617,
+      "grad_norm": 0.4896989853032177,
+      "learning_rate": 2.836381673848626e-06,
+      "loss": 0.0306,
+      "step": 396
+    },
+    {
+      "epoch": 2.0463917525773194,
+      "grad_norm": 0.4887400794591483,
+      "learning_rate": 2.809344108236089e-06,
+      "loss": 0.0418,
+      "step": 397
+    },
+    {
+      "epoch": 2.051546391752577,
+      "grad_norm": 0.6784658966637123,
+      "learning_rate": 2.782385586681112e-06,
+      "loss": 0.0332,
+      "step": 398
+    },
+    {
+      "epoch": 2.056701030927835,
+      "grad_norm": 1.0902259652603303,
+      "learning_rate": 2.7555070819110554e-06,
+      "loss": 0.0654,
+      "step": 399
+    },
+    {
+      "epoch": 2.0618556701030926,
+      "grad_norm": 0.6516911660495184,
+      "learning_rate": 2.7287095637660844e-06,
+      "loss": 0.0507,
+      "step": 400
+    },
+    {
+      "epoch": 2.0670103092783507,
+      "grad_norm": 0.5685061203950251,
+      "learning_rate": 2.7019939991641817e-06,
+      "loss": 0.0379,
+      "step": 401
+    },
+    {
+      "epoch": 2.0721649484536084,
+      "grad_norm": 0.6332607919424238,
+      "learning_rate": 2.675361352066246e-06,
+      "loss": 0.0341,
+      "step": 402
+    },
+    {
+      "epoch": 2.077319587628866,
+      "grad_norm": 0.5325338132827512,
+      "learning_rate": 2.648812583441316e-06,
+      "loss": 0.0236,
+      "step": 403
+    },
+    {
+      "epoch": 2.082474226804124,
+      "grad_norm": 0.7261655399729875,
+      "learning_rate": 2.6223486512318984e-06,
+      "loss": 0.042,
+      "step": 404
+    },
+    {
+      "epoch": 2.0876288659793816,
+      "grad_norm": 0.435792336105894,
+      "learning_rate": 2.5959705103194044e-06,
+      "loss": 0.0236,
+      "step": 405
+    },
+    {
+      "epoch": 2.0927835051546393,
+      "grad_norm": 0.6891313001342396,
+      "learning_rate": 2.5696791124896843e-06,
+      "loss": 0.0342,
+      "step": 406
+    },
+    {
+      "epoch": 2.097938144329897,
+      "grad_norm": 0.45724741270721986,
+      "learning_rate": 2.5434754063986944e-06,
+      "loss": 0.0273,
+      "step": 407
+    },
+    {
+      "epoch": 2.1030927835051547,
+      "grad_norm": 1.272124005413923,
+      "learning_rate": 2.5173603375382715e-06,
+      "loss": 0.0522,
+      "step": 408
+    },
+    {
+      "epoch": 2.1082474226804124,
+      "grad_norm": 0.8289742713601923,
+      "learning_rate": 2.4913348482020017e-06,
+      "loss": 0.0424,
+      "step": 409
+    },
+    {
+      "epoch": 2.11340206185567,
+      "grad_norm": 0.7874488954689518,
+      "learning_rate": 2.4653998774512356e-06,
+      "loss": 0.0453,
+      "step": 410
+    },
+    {
+      "epoch": 2.118556701030928,
+      "grad_norm": 0.9440880659612745,
+      "learning_rate": 2.4395563610812003e-06,
+      "loss": 0.0454,
+      "step": 411
+    },
+    {
+      "epoch": 2.1237113402061856,
+      "grad_norm": 1.0587599130365977,
+      "learning_rate": 2.413805231587225e-06,
+      "loss": 0.0536,
+      "step": 412
+    },
+    {
+      "epoch": 2.1288659793814433,
+      "grad_norm": 0.7575668449843368,
+      "learning_rate": 2.3881474181311054e-06,
+      "loss": 0.0346,
+      "step": 413
+    },
+    {
+      "epoch": 2.134020618556701,
+      "grad_norm": 0.7258410232800275,
+      "learning_rate": 2.3625838465075757e-06,
+      "loss": 0.0329,
+      "step": 414
+    },
+    {
+      "epoch": 2.1391752577319587,
+      "grad_norm": 0.8982683491629246,
+      "learning_rate": 2.3371154391108968e-06,
+      "loss": 0.0373,
+      "step": 415
+    },
+    {
+      "epoch": 2.1443298969072164,
+      "grad_norm": 0.9718369828761707,
+      "learning_rate": 2.31174311490158e-06,
+      "loss": 0.0585,
+      "step": 416
+    },
+    {
+      "epoch": 2.149484536082474,
+      "grad_norm": 0.7060391066658174,
+      "learning_rate": 2.2864677893732302e-06,
+      "loss": 0.0392,
+      "step": 417
+    },
+    {
+      "epoch": 2.154639175257732,
+      "grad_norm": 1.0197737574698842,
+      "learning_rate": 2.2612903745195085e-06,
+      "loss": 0.0441,
+      "step": 418
+    },
+    {
+      "epoch": 2.1597938144329896,
+      "grad_norm": 0.781194740237935,
+      "learning_rate": 2.2362117788012237e-06,
+      "loss": 0.0401,
+      "step": 419
+    },
+    {
+      "epoch": 2.1649484536082473,
+      "grad_norm": 0.8215547635000949,
+      "learning_rate": 2.211232907113562e-06,
+      "loss": 0.0412,
+      "step": 420
+    },
+    {
+      "epoch": 2.170103092783505,
+      "grad_norm": 0.5478388174570431,
+      "learning_rate": 2.186354660753424e-06,
+      "loss": 0.0281,
+      "step": 421
+    },
+    {
+      "epoch": 2.1752577319587627,
+      "grad_norm": 0.7326277955471945,
+      "learning_rate": 2.1615779373869116e-06,
+      "loss": 0.0422,
+      "step": 422
+    },
+    {
+      "epoch": 2.1804123711340204,
+      "grad_norm": 0.5930888725441763,
+      "learning_rate": 2.1369036310169394e-06,
+      "loss": 0.0286,
+      "step": 423
+    },
+    {
+      "epoch": 2.1855670103092786,
+      "grad_norm": 0.7463894181675266,
+      "learning_rate": 2.1123326319509736e-06,
+      "loss": 0.0362,
+      "step": 424
+    },
+    {
+      "epoch": 2.1907216494845363,
+      "grad_norm": 0.7685247934977517,
+      "learning_rate": 2.087865826768902e-06,
+      "loss": 0.0357,
+      "step": 425
+    },
+    {
+      "epoch": 2.195876288659794,
+      "grad_norm": 0.6968892457824759,
+      "learning_rate": 2.0635040982910596e-06,
+      "loss": 0.0372,
+      "step": 426
+    },
+    {
+      "epoch": 2.2010309278350517,
+      "grad_norm": 0.8893661323544511,
+      "learning_rate": 2.0392483255463568e-06,
+      "loss": 0.0458,
+      "step": 427
+    },
+    {
+      "epoch": 2.2061855670103094,
+      "grad_norm": 0.8451595069777674,
+      "learning_rate": 2.0150993837405715e-06,
+      "loss": 0.0307,
+      "step": 428
+    },
+    {
+      "epoch": 2.211340206185567,
+      "grad_norm": 0.753815598720654,
+      "learning_rate": 1.991058144224773e-06,
+      "loss": 0.0465,
+      "step": 429
+    },
+    {
+      "epoch": 2.216494845360825,
+      "grad_norm": 1.439374052545699,
+      "learning_rate": 1.9671254744638696e-06,
+      "loss": 0.0444,
+      "step": 430
+    },
+    {
+      "epoch": 2.2216494845360826,
+      "grad_norm": 0.7273050060533928,
+      "learning_rate": 1.9433022380053223e-06,
+      "loss": 0.0429,
+      "step": 431
+    },
+    {
+      "epoch": 2.2268041237113403,
+      "grad_norm": 0.9062307253651583,
+      "learning_rate": 1.9195892944479726e-06,
+      "loss": 0.0413,
+      "step": 432
+    },
+    {
+      "epoch": 2.231958762886598,
+      "grad_norm": 0.6309173620859928,
+      "learning_rate": 1.8959874994110395e-06,
+      "loss": 0.0263,
+      "step": 433
+    },
+    {
+      "epoch": 2.2371134020618557,
+      "grad_norm": 0.5542210116343528,
+      "learning_rate": 1.8724977045032343e-06,
+      "loss": 0.0291,
+      "step": 434
+    },
+    {
+      "epoch": 2.2422680412371134,
+      "grad_norm": 1.0172406554175537,
+      "learning_rate": 1.849120757292039e-06,
+      "loss": 0.0417,
+      "step": 435
+    },
+    {
+      "epoch": 2.247422680412371,
+      "grad_norm": 0.6996581336646202,
+      "learning_rate": 1.8258575012731256e-06,
+      "loss": 0.0321,
+      "step": 436
+    },
+    {
+      "epoch": 2.252577319587629,
+      "grad_norm": 0.6713360166170348,
+      "learning_rate": 1.8027087758399186e-06,
+      "loss": 0.028,
+      "step": 437
+    },
+    {
+      "epoch": 2.2577319587628866,
+      "grad_norm": 0.7330856423922832,
+      "learning_rate": 1.7796754162533031e-06,
+      "loss": 0.0352,
+      "step": 438
+    },
+    {
+      "epoch": 2.2628865979381443,
+      "grad_norm": 0.5395581155148059,
+      "learning_rate": 1.7567582536114974e-06,
+      "loss": 0.0347,
+      "step": 439
+    },
+    {
+      "epoch": 2.268041237113402,
+      "grad_norm": 0.648257580104865,
+      "learning_rate": 1.733958114820053e-06,
+      "loss": 0.0366,
+      "step": 440
+    },
+    {
+      "epoch": 2.2731958762886597,
+      "grad_norm": 0.8632515630774079,
+      "learning_rate": 1.7112758225620241e-06,
+      "loss": 0.0315,
+      "step": 441
+    },
+    {
+      "epoch": 2.2783505154639174,
+      "grad_norm": 0.5888956902163126,
+      "learning_rate": 1.6887121952682872e-06,
+      "loss": 0.0307,
+      "step": 442
+    },
+    {
+      "epoch": 2.283505154639175,
+      "grad_norm": 0.8546230795478369,
+      "learning_rate": 1.6662680470879994e-06,
+      "loss": 0.0379,
+      "step": 443
+    },
+    {
+      "epoch": 2.288659793814433,
+      "grad_norm": 0.6354466796154322,
+      "learning_rate": 1.6439441878592327e-06,
+      "loss": 0.0304,
+      "step": 444
+    },
+    {
+      "epoch": 2.2938144329896906,
+      "grad_norm": 0.7021144673842427,
+      "learning_rate": 1.621741423079749e-06,
+      "loss": 0.0265,
+      "step": 445
+    },
+    {
+      "epoch": 2.2989690721649483,
+      "grad_norm": 0.6972215933611017,
+      "learning_rate": 1.5996605538779315e-06,
+      "loss": 0.0409,
+      "step": 446
+    },
+    {
+      "epoch": 2.304123711340206,
+      "grad_norm": 0.7820256017166072,
+      "learning_rate": 1.5777023769838829e-06,
+      "loss": 0.0434,
+      "step": 447
+    },
+    {
+      "epoch": 2.3092783505154637,
+      "grad_norm": 1.0039869369381016,
+      "learning_rate": 1.5558676847006804e-06,
+      "loss": 0.0356,
+      "step": 448
+    },
+    {
+      "epoch": 2.3144329896907214,
+      "grad_norm": 0.7755710074571075,
+      "learning_rate": 1.534157264875778e-06,
+      "loss": 0.045,
+      "step": 449
+    },
+    {
+      "epoch": 2.319587628865979,
+      "grad_norm": 0.8163030212240976,
+      "learning_rate": 1.5125719008725914e-06,
+      "loss": 0.0376,
+      "step": 450
+    },
+    {
+      "epoch": 2.3247422680412373,
+      "grad_norm": 0.9974216797385139,
+      "learning_rate": 1.4911123715422187e-06,
+      "loss": 0.0472,
+      "step": 451
+    },
+    {
+      "epoch": 2.329896907216495,
+      "grad_norm": 0.7320306461711172,
+      "learning_rate": 1.4697794511953512e-06,
+      "loss": 0.0234,
+      "step": 452
+    },
+    {
+      "epoch": 2.3350515463917527,
+      "grad_norm": 0.5551288224279005,
+      "learning_rate": 1.448573909574323e-06,
+      "loss": 0.032,
+      "step": 453
+    },
+    {
+      "epoch": 2.3402061855670104,
+      "grad_norm": 0.6010629184530717,
+      "learning_rate": 1.4274965118253453e-06,
+      "loss": 0.0335,
+      "step": 454
+    },
+    {
+      "epoch": 2.345360824742268,
+      "grad_norm": 0.9144825600431091,
+      "learning_rate": 1.406548018470893e-06,
+      "loss": 0.0417,
+      "step": 455
+    },
+    {
+      "epoch": 2.350515463917526,
+      "grad_norm": 0.62509219056783,
+      "learning_rate": 1.3857291853822624e-06,
+      "loss": 0.0371,
+      "step": 456
+    },
+    {
+      "epoch": 2.3556701030927836,
+      "grad_norm": 1.0245042407320217,
+      "learning_rate": 1.3650407637523055e-06,
+      "loss": 0.0464,
+      "step": 457
+    },
+    {
+      "epoch": 2.3608247422680413,
+      "grad_norm": 0.460836672459896,
+      "learning_rate": 1.344483500068317e-06,
+      "loss": 0.0231,
+      "step": 458
+    },
+    {
+      "epoch": 2.365979381443299,
+      "grad_norm": 0.703040438234448,
+      "learning_rate": 1.3240581360851017e-06,
+      "loss": 0.0445,
+      "step": 459
+    },
+    {
+      "epoch": 2.3711340206185567,
+      "grad_norm": 0.3479413871407806,
+      "learning_rate": 1.3037654087982083e-06,
+      "loss": 0.0146,
+      "step": 460
+    },
+    {
+      "epoch": 2.3762886597938144,
+      "grad_norm": 0.6144775547807693,
+      "learning_rate": 1.2836060504173448e-06,
+      "loss": 0.0264,
+      "step": 461
+    },
+    {
+      "epoch": 2.381443298969072,
+      "grad_norm": 0.6117334170509691,
+      "learning_rate": 1.2635807883399487e-06,
+      "loss": 0.032,
+      "step": 462
+    },
+    {
+      "epoch": 2.38659793814433,
+      "grad_norm": 0.9895965753292417,
+      "learning_rate": 1.2436903451249417e-06,
+      "loss": 0.0353,
+      "step": 463
+    },
+    {
+      "epoch": 2.3917525773195876,
+      "grad_norm": 0.805324192811629,
+      "learning_rate": 1.2239354384666729e-06,
+      "loss": 0.0384,
+      "step": 464
+    },
+    {
+      "epoch": 2.3969072164948453,
+      "grad_norm": 0.7642731807536111,
+      "learning_rate": 1.2043167811689999e-06,
+      "loss": 0.0407,
+      "step": 465
+    },
+    {
+      "epoch": 2.402061855670103,
+      "grad_norm": 0.7282861704316356,
+      "learning_rate": 1.184835081119583e-06,
+      "loss": 0.0272,
+      "step": 466
+    },
+    {
+      "epoch": 2.4072164948453607,
+      "grad_norm": 0.7248189934139107,
+      "learning_rate": 1.165491041264345e-06,
+      "loss": 0.0394,
+      "step": 467
+    },
+    {
+      "epoch": 2.4123711340206184,
+      "grad_norm": 0.6132554336490188,
+      "learning_rate": 1.1462853595820976e-06,
+      "loss": 0.028,
+      "step": 468
+    },
+    {
+      "epoch": 2.417525773195876,
+      "grad_norm": 0.4481366887208859,
+      "learning_rate": 1.1272187290593601e-06,
+      "loss": 0.0198,
+      "step": 469
+    },
+    {
+      "epoch": 2.422680412371134,
+      "grad_norm": 0.6423378262075967,
+      "learning_rate": 1.108291837665365e-06,
+      "loss": 0.0379,
+      "step": 470
+    },
+    {
+      "epoch": 2.4278350515463916,
+      "grad_norm": 0.6949665105410578,
+      "learning_rate": 1.0895053683272182e-06,
+      "loss": 0.0294,
+      "step": 471
+    },
+    {
+      "epoch": 2.4329896907216497,
+      "grad_norm": 0.5203342004591067,
+      "learning_rate": 1.0708599989052653e-06,
+      "loss": 0.0219,
+      "step": 472
+    },
+    {
+      "epoch": 2.4381443298969074,
+      "grad_norm": 0.6585762086070828,
+      "learning_rate": 1.0523564021686367e-06,
+      "loss": 0.0316,
+      "step": 473
+    },
+    {
+      "epoch": 2.443298969072165,
+      "grad_norm": 0.5287292409234337,
+      "learning_rate": 1.0339952457709645e-06,
+      "loss": 0.0263,
+      "step": 474
+    },
+    {
+      "epoch": 2.448453608247423,
+      "grad_norm": 0.672655997320893,
+      "learning_rate": 1.0157771922262954e-06,
+      "loss": 0.032,
+      "step": 475
+    },
+    {
+      "epoch": 2.4536082474226806,
+      "grad_norm": 0.9056980944300012,
+      "learning_rate": 9.97702898885189e-07,
+      "loss": 0.0355,
+      "step": 476
+    },
+    {
+      "epoch": 2.4587628865979383,
+      "grad_norm": 0.9727772351982882,
+      "learning_rate": 9.797730179109953e-07,
+      "loss": 0.0462,
+      "step": 477
+    },
+    {
+      "epoch": 2.463917525773196,
+      "grad_norm": 0.471860289660639,
+      "learning_rate": 9.619881962563226e-07,
+      "loss": 0.0213,
+      "step": 478
+    },
+    {
+      "epoch": 2.4690721649484537,
+      "grad_norm": 0.8540053890635061,
+      "learning_rate": 9.443490756396934e-07,
+      "loss": 0.0337,
+      "step": 479
+    },
+    {
+      "epoch": 2.4742268041237114,
+      "grad_norm": 0.7393373284547952,
+      "learning_rate": 9.268562925223956e-07,
+      "loss": 0.0291,
+      "step": 480
+    },
+    {
+      "epoch": 2.479381443298969,
+      "grad_norm": 1.0178262863520333,
+      "learning_rate": 9.095104780855085e-07,
+      "loss": 0.06,
+      "step": 481
+    },
+    {
+      "epoch": 2.484536082474227,
+      "grad_norm": 0.6419818956203113,
+      "learning_rate": 8.923122582071375e-07,
+      "loss": 0.0262,
+      "step": 482
+    },
+    {
+      "epoch": 2.4896907216494846,
+      "grad_norm": 0.595800378742964,
+      "learning_rate": 8.752622534398253e-07,
+      "loss": 0.0226,
+      "step": 483
+    },
+    {
+      "epoch": 2.4948453608247423,
+      "grad_norm": 0.5855135906779485,
+      "learning_rate": 8.583610789881597e-07,
+      "loss": 0.0224,
+      "step": 484
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.8382231702975446,
+      "learning_rate": 8.41609344686578e-07,
+      "loss": 0.041,
+      "step": 485
+    },
+    {
+      "epoch": 2.5051546391752577,
+      "grad_norm": 1.1568369617447913,
+      "learning_rate": 8.250076549773672e-07,
+      "loss": 0.0399,
+      "step": 486
+    },
+    {
+      "epoch": 2.5103092783505154,
+      "grad_norm": 0.7049565289339409,
+      "learning_rate": 8.085566088888447e-07,
+      "loss": 0.0266,
+      "step": 487
+    },
+    {
+      "epoch": 2.515463917525773,
+      "grad_norm": 0.9021686507801038,
+      "learning_rate": 7.922568000137498e-07,
+      "loss": 0.0437,
+      "step": 488
+    },
+    {
+      "epoch": 2.520618556701031,
+      "grad_norm": 0.5262135746894241,
+      "learning_rate": 7.761088164878272e-07,
+      "loss": 0.0289,
+      "step": 489
+    },
+    {
+      "epoch": 2.5257731958762886,
+      "grad_norm": 0.7544552659389241,
+      "learning_rate": 7.601132409686019e-07,
+      "loss": 0.0257,
+      "step": 490
+    },
+    {
+      "epoch": 2.5309278350515463,
+      "grad_norm": 0.931143398887618,
+      "learning_rate": 7.442706506143549e-07,
+      "loss": 0.0327,
+      "step": 491
+    },
+    {
+      "epoch": 2.536082474226804,
+      "grad_norm": 0.5273275385274795,
+      "learning_rate": 7.285816170633037e-07,
+      "loss": 0.0281,
+      "step": 492
+    },
+    {
+      "epoch": 2.5412371134020617,
+      "grad_norm": 0.6918194060651451,
+      "learning_rate": 7.130467064129675e-07,
+      "loss": 0.0399,
+      "step": 493
+    },
+    {
+      "epoch": 2.5463917525773194,
+      "grad_norm": 0.5632783189164381,
+      "learning_rate": 6.976664791997484e-07,
+      "loss": 0.0308,
+      "step": 494
+    },
+    {
+      "epoch": 2.551546391752577,
+      "grad_norm": 0.4667969942058232,
+      "learning_rate": 6.824414903787041e-07,
+      "loss": 0.0208,
+      "step": 495
+    },
+    {
+      "epoch": 2.556701030927835,
+      "grad_norm": 0.3651029418094621,
+      "learning_rate": 6.673722893035228e-07,
+      "loss": 0.0147,
+      "step": 496
+    },
+    {
+      "epoch": 2.5618556701030926,
+      "grad_norm": 0.8235621709812485,
+      "learning_rate": 6.524594197067002e-07,
+      "loss": 0.0321,
+      "step": 497
+    },
+    {
+      "epoch": 2.5670103092783503,
+      "grad_norm": 0.47110921045669607,
+      "learning_rate": 6.377034196799232e-07,
+      "loss": 0.0179,
+      "step": 498
+    },
+    {
+      "epoch": 2.572164948453608,
+      "grad_norm": 1.1301569692968019,
+      "learning_rate": 6.23104821654652e-07,
+      "loss": 0.0469,
+      "step": 499
+    },
+    {
+      "epoch": 2.5773195876288657,
+      "grad_norm": 0.5466735219913834,
+      "learning_rate": 6.086641523829079e-07,
+      "loss": 0.0279,
+      "step": 500
+    },
+    {
+      "epoch": 2.582474226804124,
+      "grad_norm": 0.7232728109542946,
+      "learning_rate": 5.943819329182726e-07,
+      "loss": 0.0289,
+      "step": 501
+    },
+    {
+      "epoch": 2.5876288659793816,
+      "grad_norm": 0.7342931394967595,
+      "learning_rate": 5.802586785970782e-07,
+      "loss": 0.0344,
+      "step": 502
+    },
+    {
+      "epoch": 2.5927835051546393,
+      "grad_norm": 0.7998011121158775,
+      "learning_rate": 5.662948990198225e-07,
+      "loss": 0.035,
+      "step": 503
+    },
+    {
+      "epoch": 2.597938144329897,
+      "grad_norm": 0.780521929367139,
+      "learning_rate": 5.524910980327714e-07,
+      "loss": 0.0378,
+      "step": 504
+    },
+    {
+      "epoch": 2.6030927835051547,
+      "grad_norm": 0.3753251347849975,
+      "learning_rate": 5.388477737097892e-07,
+      "loss": 0.0129,
+      "step": 505
+    },
+    {
+      "epoch": 2.6082474226804124,
+      "grad_norm": 0.8067566714210145,
+      "learning_rate": 5.25365418334357e-07,
+      "loss": 0.0405,
+      "step": 506
+    },
+    {
+      "epoch": 2.61340206185567,
+      "grad_norm": 0.6718120073753768,
+      "learning_rate": 5.120445183818173e-07,
+      "loss": 0.0286,
+      "step": 507
+    },
+    {
+      "epoch": 2.618556701030928,
+      "grad_norm": 0.6935793608986474,
+      "learning_rate": 4.988855545018184e-07,
+      "loss": 0.0402,
+      "step": 508
+    },
+    {
+      "epoch": 2.6237113402061856,
+      "grad_norm": 0.45521262118768585,
+      "learning_rate": 4.85889001500971e-07,
+      "loss": 0.0204,
+      "step": 509
+    },
+    {
+      "epoch": 2.6288659793814433,
+      "grad_norm": 0.9118444591169176,
+      "learning_rate": 4.730553283257144e-07,
+      "loss": 0.0426,
+      "step": 510
+    },
+    {
+      "epoch": 2.634020618556701,
+      "grad_norm": 0.4872973349736019,
+      "learning_rate": 4.603849980454006e-07,
+      "loss": 0.0177,
+      "step": 511
+    },
+    {
+      "epoch": 2.6391752577319587,
+      "grad_norm": 0.5157144934784639,
+      "learning_rate": 4.4787846783558134e-07,
+      "loss": 0.0162,
+      "step": 512
+    },
+    {
+      "epoch": 2.6443298969072164,
+      "grad_norm": 0.7669146783828849,
+      "learning_rate": 4.355361889615117e-07,
+      "loss": 0.0387,
+      "step": 513
+    },
+    {
+      "epoch": 2.649484536082474,
+      "grad_norm": 0.741624793469597,
+      "learning_rate": 4.2335860676187334e-07,
+      "loss": 0.0378,
+      "step": 514
+    },
+    {
+      "epoch": 2.654639175257732,
+      "grad_norm": 0.7705165559277041,
+      "learning_rate": 4.1134616063269685e-07,
+      "loss": 0.0355,
+      "step": 515
+    },
+    {
+      "epoch": 2.6597938144329896,
+      "grad_norm": 0.8220524914863429,
+      "learning_rate": 3.994992840115153e-07,
+      "loss": 0.041,
+      "step": 516
+    },
+    {
+      "epoch": 2.6649484536082473,
+      "grad_norm": 0.5352266924175778,
+      "learning_rate": 3.878184043617217e-07,
+      "loss": 0.0265,
+      "step": 517
+    },
+    {
+      "epoch": 2.670103092783505,
+      "grad_norm": 0.947338719017043,
+      "learning_rate": 3.7630394315714246e-07,
+      "loss": 0.0445,
+      "step": 518
+    },
+    {
+      "epoch": 2.675257731958763,
+      "grad_norm": 0.723238897545102,
+      "learning_rate": 3.649563158668318e-07,
+      "loss": 0.0368,
+      "step": 519
+    },
+    {
+      "epoch": 2.680412371134021,
+      "grad_norm": 0.835416098910731,
+      "learning_rate": 3.5377593194008406e-07,
+      "loss": 0.0384,
+      "step": 520
+    },
+    {
+      "epoch": 2.6855670103092786,
+      "grad_norm": 0.8873839704436575,
+      "learning_rate": 3.427631947916532e-07,
+      "loss": 0.0404,
+      "step": 521
+    },
+    {
+      "epoch": 2.6907216494845363,
+      "grad_norm": 0.9502292877504235,
+      "learning_rate": 3.319185017872023e-07,
+      "loss": 0.0459,
+      "step": 522
+    },
+    {
+      "epoch": 2.695876288659794,
+      "grad_norm": 0.6395104911003975,
+      "learning_rate": 3.2124224422896145e-07,
+      "loss": 0.0335,
+      "step": 523
+    },
+    {
+      "epoch": 2.7010309278350517,
+      "grad_norm": 0.6249461812750059,
+      "learning_rate": 3.107348073416128e-07,
+      "loss": 0.0249,
+      "step": 524
+    },
+    {
+      "epoch": 2.7061855670103094,
+      "grad_norm": 0.8557720308902851,
+      "learning_rate": 3.003965702583861e-07,
+      "loss": 0.0446,
+      "step": 525
+    },
+    {
+      "epoch": 2.711340206185567,
+      "grad_norm": 0.6766334639256003,
+      "learning_rate": 2.902279060073837e-07,
+      "loss": 0.0182,
+      "step": 526
+    },
+    {
+      "epoch": 2.716494845360825,
+      "grad_norm": 0.831048167311916,
+      "learning_rate": 2.8022918149811563e-07,
+      "loss": 0.0373,
+      "step": 527
+    },
+    {
+      "epoch": 2.7216494845360826,
+      "grad_norm": 0.8686114936097588,
+      "learning_rate": 2.704007575082634e-07,
+      "loss": 0.039,
+      "step": 528
+    },
+    {
+      "epoch": 2.7268041237113403,
+      "grad_norm": 0.7920316016495633,
+      "learning_rate": 2.6074298867066473e-07,
+      "loss": 0.0331,
+      "step": 529
+    },
+    {
+      "epoch": 2.731958762886598,
+      "grad_norm": 0.6155600496425835,
+      "learning_rate": 2.512562234605131e-07,
+      "loss": 0.0212,
+      "step": 530
+    },
+    {
+      "epoch": 2.7371134020618557,
+      "grad_norm": 0.39040910496765024,
+      "learning_rate": 2.4194080418278475e-07,
+      "loss": 0.0181,
+      "step": 531
+    },
+    {
+      "epoch": 2.7422680412371134,
+      "grad_norm": 0.6426304269936253,
+      "learning_rate": 2.3279706695989002e-07,
+      "loss": 0.024,
+      "step": 532
+    },
+    {
+      "epoch": 2.747422680412371,
+      "grad_norm": 0.8245244375929415,
+      "learning_rate": 2.238253417195435e-07,
+      "loss": 0.0338,
+      "step": 533
+    },
+    {
+      "epoch": 2.752577319587629,
+      "grad_norm": 0.5880833696637757,
+      "learning_rate": 2.1502595218285905e-07,
+      "loss": 0.0268,
+      "step": 534
+    },
+    {
+      "epoch": 2.7577319587628866,
+      "grad_norm": 0.5867228286206739,
+      "learning_rate": 2.063992158526701e-07,
+      "loss": 0.0334,
+      "step": 535
+    },
+    {
+      "epoch": 2.7628865979381443,
+      "grad_norm": 0.8677251644818563,
+      "learning_rate": 1.979454440020745e-07,
+      "loss": 0.0444,
+      "step": 536
+    },
+    {
+      "epoch": 2.768041237113402,
+      "grad_norm": 1.260938972819265,
+      "learning_rate": 1.896649416631996e-07,
+      "loss": 0.0369,
+      "step": 537
+    },
+    {
+      "epoch": 2.7731958762886597,
+      "grad_norm": 0.6079978206952025,
+      "learning_rate": 1.8155800761619825e-07,
+      "loss": 0.0367,
+      "step": 538
+    },
+    {
+      "epoch": 2.7783505154639174,
+      "grad_norm": 0.6942989169764706,
+      "learning_rate": 1.7362493437846906e-07,
+      "loss": 0.0404,
+      "step": 539
+    },
+    {
+      "epoch": 2.783505154639175,
+      "grad_norm": 0.6712489244767057,
+      "learning_rate": 1.6586600819409992e-07,
+      "loss": 0.0343,
+      "step": 540
+    },
+    {
+      "epoch": 2.788659793814433,
+      "grad_norm": 0.6329822911791473,
+      "learning_rate": 1.582815090235401e-07,
+      "loss": 0.0267,
+      "step": 541
+    },
+    {
+      "epoch": 2.7938144329896906,
+      "grad_norm": 0.8353013781048685,
+      "learning_rate": 1.5087171053349836e-07,
+      "loss": 0.0471,
+      "step": 542
+    },
+    {
+      "epoch": 2.7989690721649483,
+      "grad_norm": 0.6876307970310361,
+      "learning_rate": 1.4363688008707021e-07,
+      "loss": 0.0338,
+      "step": 543
+    },
+    {
+      "epoch": 2.804123711340206,
+      "grad_norm": 0.691468327037597,
+      "learning_rate": 1.3657727873408798e-07,
+      "loss": 0.0269,
+      "step": 544
+    },
+    {
+      "epoch": 2.8092783505154637,
+      "grad_norm": 0.5595080989035071,
+      "learning_rate": 1.2969316120170326e-07,
+      "loss": 0.0287,
+      "step": 545
+    },
+    {
+      "epoch": 2.8144329896907214,
+      "grad_norm": 0.8241095438340851,
+      "learning_rate": 1.2298477588519598e-07,
+      "loss": 0.032,
+      "step": 546
+    },
+    {
+      "epoch": 2.819587628865979,
+      "grad_norm": 0.46352885981179276,
+      "learning_rate": 1.1645236483900978e-07,
+      "loss": 0.0252,
+      "step": 547
+    },
+    {
+      "epoch": 2.824742268041237,
+      "grad_norm": 1.015807161902545,
+      "learning_rate": 1.1009616376802145e-07,
+      "loss": 0.0436,
+      "step": 548
+    },
+    {
+      "epoch": 2.829896907216495,
+      "grad_norm": 0.4914191666668514,
+      "learning_rate": 1.0391640201903253e-07,
+      "loss": 0.0185,
+      "step": 549
+    },
+    {
+      "epoch": 2.8350515463917527,
+      "grad_norm": 0.4945832357939433,
+      "learning_rate": 9.791330257249599e-08,
+      "loss": 0.0236,
+      "step": 550
+    },
+    {
+      "epoch": 2.8402061855670104,
+      "grad_norm": 0.7932657531697902,
+      "learning_rate": 9.208708203447047e-08,
+      "loss": 0.0324,
+      "step": 551
+    },
+    {
+      "epoch": 2.845360824742268,
+      "grad_norm": 0.5895443936626965,
+      "learning_rate": 8.643795062880433e-08,
+      "loss": 0.0295,
+      "step": 552
+    },
+    {
+      "epoch": 2.850515463917526,
+      "grad_norm": 0.7452320196861348,
+      "learning_rate": 8.096611218954941e-08,
+      "loss": 0.0468,
+      "step": 553
+    },
+    {
+      "epoch": 2.8556701030927836,
+      "grad_norm": 0.6662054784828428,
+      "learning_rate": 7.567176415360811e-08,
+      "loss": 0.0285,
+      "step": 554
+    },
+    {
+      "epoch": 2.8608247422680413,
+      "grad_norm": 0.7194136910796446,
+      "learning_rate": 7.055509755360846e-08,
+      "loss": 0.0302,
+      "step": 555
+    },
+    {
+      "epoch": 2.865979381443299,
+      "grad_norm": 0.9809765386826904,
+      "learning_rate": 6.561629701101024e-08,
+      "loss": 0.0462,
+      "step": 556
+    },
+    {
+      "epoch": 2.8711340206185567,
+      "grad_norm": 0.8531811367409373,
+      "learning_rate": 6.085554072944366e-08,
+      "loss": 0.046,
+      "step": 557
+    },
+    {
+      "epoch": 2.8762886597938144,
+      "grad_norm": 0.7203492707718298,
+      "learning_rate": 5.627300048828166e-08,
+      "loss": 0.0382,
+      "step": 558
+    },
+    {
+      "epoch": 2.881443298969072,
+      "grad_norm": 0.5263471079242228,
+      "learning_rate": 5.18688416364388e-08,
+      "loss": 0.0265,
+      "step": 559
+    },
+    {
+      "epoch": 2.88659793814433,
+      "grad_norm": 0.6161772951986925,
+      "learning_rate": 4.764322308640601e-08,
+      "loss": 0.0259,
+      "step": 560
+    },
+    {
+      "epoch": 2.8917525773195876,
+      "grad_norm": 0.7038077277693001,
+      "learning_rate": 4.3596297308518557e-08,
+      "loss": 0.0278,
+      "step": 561
+    },
+    {
+      "epoch": 2.8969072164948453,
+      "grad_norm": 1.6304164778117418,
+      "learning_rate": 3.972821032545149e-08,
+      "loss": 0.0432,
+      "step": 562
+    },
+    {
+      "epoch": 2.902061855670103,
+      "grad_norm": 0.5756860245564832,
+      "learning_rate": 3.603910170695168e-08,
+      "loss": 0.0279,
+      "step": 563
+    },
+    {
+      "epoch": 2.9072164948453607,
+      "grad_norm": 0.5564216825615497,
+      "learning_rate": 3.2529104564804626e-08,
+      "loss": 0.0265,
+      "step": 564
+    },
+    {
+      "epoch": 2.9123711340206184,
+      "grad_norm": 0.5859936807288393,
+      "learning_rate": 2.9198345548026587e-08,
+      "loss": 0.0379,
+      "step": 565
+    },
+    {
+      "epoch": 2.917525773195876,
+      "grad_norm": 0.7182530071325729,
+      "learning_rate": 2.6046944838298837e-08,
+      "loss": 0.0343,
+      "step": 566
+    },
+    {
+      "epoch": 2.9226804123711343,
+      "grad_norm": 0.8749233277497677,
+      "learning_rate": 2.307501614562946e-08,
+      "loss": 0.0343,
+      "step": 567
+    },
+    {
+      "epoch": 2.927835051546392,
+      "grad_norm": 0.6515021617294287,
+      "learning_rate": 2.028266670425105e-08,
+      "loss": 0.0369,
+      "step": 568
+    },
+    {
+      "epoch": 2.9329896907216497,
+      "grad_norm": 0.7247850456909103,
+      "learning_rate": 1.766999726874996e-08,
+      "loss": 0.0295,
+      "step": 569
+    },
+    {
+      "epoch": 2.9381443298969074,
+      "grad_norm": 0.5536332761399269,
+      "learning_rate": 1.5237102110433054e-08,
+      "loss": 0.0283,
+      "step": 570
+    },
+    {
+      "epoch": 2.943298969072165,
+      "grad_norm": 1.012719317576713,
+      "learning_rate": 1.298406901392546e-08,
+      "loss": 0.0469,
+      "step": 571
+    },
+    {
+      "epoch": 2.948453608247423,
+      "grad_norm": 0.6042020113465656,
+      "learning_rate": 1.0910979274001421e-08,
+      "loss": 0.0225,
+      "step": 572
+    },
+    {
+      "epoch": 2.9536082474226806,
+      "grad_norm": 0.777387593039972,
+      "learning_rate": 9.017907692653316e-09,
+      "loss": 0.0385,
+      "step": 573
+    },
+    {
+      "epoch": 2.9587628865979383,
+      "grad_norm": 0.6232080232360308,
+      "learning_rate": 7.304922576391593e-09,
+      "loss": 0.0246,
+      "step": 574
+    },
+    {
+      "epoch": 2.963917525773196,
+      "grad_norm": 0.6601226646157042,
+      "learning_rate": 5.77208573377952e-09,
+      "loss": 0.035,
+      "step": 575
+    },
+    {
+      "epoch": 2.9690721649484537,
+      "grad_norm": 0.9416673958502236,
+      "learning_rate": 4.4194524732038555e-09,
+      "loss": 0.0391,
+      "step": 576
+    },
+    {
+      "epoch": 2.9742268041237114,
+      "grad_norm": 0.6831145636613678,
+      "learning_rate": 3.2470716008797807e-09,
+      "loss": 0.0335,
+      "step": 577
+    },
+    {
+      "epoch": 2.979381443298969,
+      "grad_norm": 0.6371631671174869,
+      "learning_rate": 2.2549854190873077e-09,
+      "loss": 0.027,
+      "step": 578
+    },
+    {
+      "epoch": 2.984536082474227,
+      "grad_norm": 0.7228273689129697,
+      "learning_rate": 1.4432297246486092e-09,
+      "loss": 0.0312,
+      "step": 579
+    },
+    {
+      "epoch": 2.9896907216494846,
+      "grad_norm": 0.5772782243503212,
+      "learning_rate": 8.118338076329446e-10,
+      "loss": 0.0255,
+      "step": 580
+    },
+    {
+      "epoch": 2.9948453608247423,
+      "grad_norm": 0.7055691487424043,
+      "learning_rate": 3.6082045030194635e-10,
+      "loss": 0.0305,
+      "step": 581
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7595103115281251,
+      "learning_rate": 9.020592628639079e-11,
+      "loss": 0.0313,
+      "step": 582
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 582,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "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": 192630308503552.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}