diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7406 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 1052,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0019011406844106464,
+      "grad_norm": 3.5348725835093706,
+      "learning_rate": 1.886792452830189e-07,
+      "loss": 0.7021,
+      "step": 1
+    },
+    {
+      "epoch": 0.0038022813688212928,
+      "grad_norm": 3.152552789260304,
+      "learning_rate": 3.773584905660378e-07,
+      "loss": 0.7047,
+      "step": 2
+    },
+    {
+      "epoch": 0.005703422053231939,
+      "grad_norm": 3.3238379381208314,
+      "learning_rate": 5.660377358490567e-07,
+      "loss": 0.7018,
+      "step": 3
+    },
+    {
+      "epoch": 0.0076045627376425855,
+      "grad_norm": 2.968960369469816,
+      "learning_rate": 7.547169811320755e-07,
+      "loss": 0.6815,
+      "step": 4
+    },
+    {
+      "epoch": 0.009505703422053232,
+      "grad_norm": 3.3161217457536765,
+      "learning_rate": 9.433962264150944e-07,
+      "loss": 0.7269,
+      "step": 5
+    },
+    {
+      "epoch": 0.011406844106463879,
+      "grad_norm": 2.954309676235125,
+      "learning_rate": 1.1320754716981133e-06,
+      "loss": 0.6917,
+      "step": 6
+    },
+    {
+      "epoch": 0.013307984790874524,
+      "grad_norm": 3.137612440923356,
+      "learning_rate": 1.3207547169811322e-06,
+      "loss": 0.6856,
+      "step": 7
+    },
+    {
+      "epoch": 0.015209125475285171,
+      "grad_norm": 2.910877121482838,
+      "learning_rate": 1.509433962264151e-06,
+      "loss": 0.7162,
+      "step": 8
+    },
+    {
+      "epoch": 0.017110266159695818,
+      "grad_norm": 2.547930082953052,
+      "learning_rate": 1.6981132075471698e-06,
+      "loss": 0.7015,
+      "step": 9
+    },
+    {
+      "epoch": 0.019011406844106463,
+      "grad_norm": 2.2786847672972,
+      "learning_rate": 1.8867924528301889e-06,
+      "loss": 0.6814,
+      "step": 10
+    },
+    {
+      "epoch": 0.02091254752851711,
+      "grad_norm": 1.600469457594992,
+      "learning_rate": 2.075471698113208e-06,
+      "loss": 0.6553,
+      "step": 11
+    },
+    {
+      "epoch": 0.022813688212927757,
+      "grad_norm": 1.5780040255351546,
+      "learning_rate": 2.2641509433962266e-06,
+      "loss": 0.6209,
+      "step": 12
+    },
+    {
+      "epoch": 0.024714828897338403,
+      "grad_norm": 1.5491541864478167,
+      "learning_rate": 2.4528301886792453e-06,
+      "loss": 0.5966,
+      "step": 13
+    },
+    {
+      "epoch": 0.026615969581749048,
+      "grad_norm": 1.345876530387494,
+      "learning_rate": 2.6415094339622644e-06,
+      "loss": 0.5872,
+      "step": 14
+    },
+    {
+      "epoch": 0.028517110266159697,
+      "grad_norm": 2.0183197896409313,
+      "learning_rate": 2.830188679245283e-06,
+      "loss": 0.6,
+      "step": 15
+    },
+    {
+      "epoch": 0.030418250950570342,
+      "grad_norm": 2.3431760727123043,
+      "learning_rate": 3.018867924528302e-06,
+      "loss": 0.5859,
+      "step": 16
+    },
+    {
+      "epoch": 0.03231939163498099,
+      "grad_norm": 2.4293382124166873,
+      "learning_rate": 3.207547169811321e-06,
+      "loss": 0.6141,
+      "step": 17
+    },
+    {
+      "epoch": 0.034220532319391636,
+      "grad_norm": 1.7961879239794132,
+      "learning_rate": 3.3962264150943395e-06,
+      "loss": 0.5872,
+      "step": 18
+    },
+    {
+      "epoch": 0.03612167300380228,
+      "grad_norm": 1.648220509894086,
+      "learning_rate": 3.5849056603773586e-06,
+      "loss": 0.5525,
+      "step": 19
+    },
+    {
+      "epoch": 0.03802281368821293,
+      "grad_norm": 1.1905659359100127,
+      "learning_rate": 3.7735849056603777e-06,
+      "loss": 0.5605,
+      "step": 20
+    },
+    {
+      "epoch": 0.039923954372623575,
+      "grad_norm": 1.3372501971443427,
+      "learning_rate": 3.962264150943396e-06,
+      "loss": 0.5622,
+      "step": 21
+    },
+    {
+      "epoch": 0.04182509505703422,
+      "grad_norm": 1.3600743973267355,
+      "learning_rate": 4.150943396226416e-06,
+      "loss": 0.5892,
+      "step": 22
+    },
+    {
+      "epoch": 0.043726235741444866,
+      "grad_norm": 1.3341301057422068,
+      "learning_rate": 4.339622641509435e-06,
+      "loss": 0.543,
+      "step": 23
+    },
+    {
+      "epoch": 0.045627376425855515,
+      "grad_norm": 1.1534728432173944,
+      "learning_rate": 4.528301886792453e-06,
+      "loss": 0.5762,
+      "step": 24
+    },
+    {
+      "epoch": 0.04752851711026616,
+      "grad_norm": 1.0242363777805332,
+      "learning_rate": 4.716981132075472e-06,
+      "loss": 0.5418,
+      "step": 25
+    },
+    {
+      "epoch": 0.049429657794676805,
+      "grad_norm": 1.0329069109540132,
+      "learning_rate": 4.905660377358491e-06,
+      "loss": 0.5642,
+      "step": 26
+    },
+    {
+      "epoch": 0.051330798479087454,
+      "grad_norm": 0.9939502466149218,
+      "learning_rate": 5.09433962264151e-06,
+      "loss": 0.5259,
+      "step": 27
+    },
+    {
+      "epoch": 0.053231939163498096,
+      "grad_norm": 1.000828937748107,
+      "learning_rate": 5.283018867924529e-06,
+      "loss": 0.5285,
+      "step": 28
+    },
+    {
+      "epoch": 0.055133079847908745,
+      "grad_norm": 0.9931493276628309,
+      "learning_rate": 5.4716981132075475e-06,
+      "loss": 0.525,
+      "step": 29
+    },
+    {
+      "epoch": 0.057034220532319393,
+      "grad_norm": 0.8360480820091103,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 0.513,
+      "step": 30
+    },
+    {
+      "epoch": 0.058935361216730035,
+      "grad_norm": 0.9003713863257823,
+      "learning_rate": 5.849056603773585e-06,
+      "loss": 0.5131,
+      "step": 31
+    },
+    {
+      "epoch": 0.060836501901140684,
+      "grad_norm": 0.9128100714990433,
+      "learning_rate": 6.037735849056604e-06,
+      "loss": 0.4956,
+      "step": 32
+    },
+    {
+      "epoch": 0.06273764258555133,
+      "grad_norm": 0.8922211493174361,
+      "learning_rate": 6.226415094339623e-06,
+      "loss": 0.5105,
+      "step": 33
+    },
+    {
+      "epoch": 0.06463878326996197,
+      "grad_norm": 0.9151175626053105,
+      "learning_rate": 6.415094339622642e-06,
+      "loss": 0.5298,
+      "step": 34
+    },
+    {
+      "epoch": 0.06653992395437262,
+      "grad_norm": 0.988271693627707,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.5127,
+      "step": 35
+    },
+    {
+      "epoch": 0.06844106463878327,
+      "grad_norm": 0.8867707924396664,
+      "learning_rate": 6.792452830188679e-06,
+      "loss": 0.5273,
+      "step": 36
+    },
+    {
+      "epoch": 0.07034220532319392,
+      "grad_norm": 0.8968038145591141,
+      "learning_rate": 6.981132075471699e-06,
+      "loss": 0.4662,
+      "step": 37
+    },
+    {
+      "epoch": 0.07224334600760456,
+      "grad_norm": 0.8858299210038523,
+      "learning_rate": 7.169811320754717e-06,
+      "loss": 0.5228,
+      "step": 38
+    },
+    {
+      "epoch": 0.0741444866920152,
+      "grad_norm": 0.8701976625579625,
+      "learning_rate": 7.358490566037736e-06,
+      "loss": 0.5256,
+      "step": 39
+    },
+    {
+      "epoch": 0.07604562737642585,
+      "grad_norm": 0.8674654101800184,
+      "learning_rate": 7.5471698113207555e-06,
+      "loss": 0.4851,
+      "step": 40
+    },
+    {
+      "epoch": 0.0779467680608365,
+      "grad_norm": 0.835864914245775,
+      "learning_rate": 7.735849056603775e-06,
+      "loss": 0.505,
+      "step": 41
+    },
+    {
+      "epoch": 0.07984790874524715,
+      "grad_norm": 0.7861571414846842,
+      "learning_rate": 7.924528301886793e-06,
+      "loss": 0.4949,
+      "step": 42
+    },
+    {
+      "epoch": 0.0817490494296578,
+      "grad_norm": 0.840788101622193,
+      "learning_rate": 8.113207547169812e-06,
+      "loss": 0.4877,
+      "step": 43
+    },
+    {
+      "epoch": 0.08365019011406843,
+      "grad_norm": 0.85650148550159,
+      "learning_rate": 8.301886792452832e-06,
+      "loss": 0.4613,
+      "step": 44
+    },
+    {
+      "epoch": 0.08555133079847908,
+      "grad_norm": 0.8112155257170715,
+      "learning_rate": 8.49056603773585e-06,
+      "loss": 0.504,
+      "step": 45
+    },
+    {
+      "epoch": 0.08745247148288973,
+      "grad_norm": 0.8091311470938732,
+      "learning_rate": 8.67924528301887e-06,
+      "loss": 0.4864,
+      "step": 46
+    },
+    {
+      "epoch": 0.08935361216730038,
+      "grad_norm": 0.7776146998518452,
+      "learning_rate": 8.867924528301887e-06,
+      "loss": 0.5089,
+      "step": 47
+    },
+    {
+      "epoch": 0.09125475285171103,
+      "grad_norm": 0.8481375068812715,
+      "learning_rate": 9.056603773584907e-06,
+      "loss": 0.4768,
+      "step": 48
+    },
+    {
+      "epoch": 0.09315589353612168,
+      "grad_norm": 0.8253414367696812,
+      "learning_rate": 9.245283018867926e-06,
+      "loss": 0.4869,
+      "step": 49
+    },
+    {
+      "epoch": 0.09505703422053231,
+      "grad_norm": 0.8401792183587481,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 0.5062,
+      "step": 50
+    },
+    {
+      "epoch": 0.09695817490494296,
+      "grad_norm": 0.7921527138172748,
+      "learning_rate": 9.622641509433963e-06,
+      "loss": 0.4697,
+      "step": 51
+    },
+    {
+      "epoch": 0.09885931558935361,
+      "grad_norm": 0.8496042332188748,
+      "learning_rate": 9.811320754716981e-06,
+      "loss": 0.4876,
+      "step": 52
+    },
+    {
+      "epoch": 0.10076045627376426,
+      "grad_norm": 0.802064117284494,
+      "learning_rate": 1e-05,
+      "loss": 0.4758,
+      "step": 53
+    },
+    {
+      "epoch": 0.10266159695817491,
+      "grad_norm": 0.9219192567077908,
+      "learning_rate": 1.018867924528302e-05,
+      "loss": 0.4866,
+      "step": 54
+    },
+    {
+      "epoch": 0.10456273764258556,
+      "grad_norm": 0.7978521716975093,
+      "learning_rate": 1.0377358490566038e-05,
+      "loss": 0.4986,
+      "step": 55
+    },
+    {
+      "epoch": 0.10646387832699619,
+      "grad_norm": 0.8214944534493908,
+      "learning_rate": 1.0566037735849058e-05,
+      "loss": 0.4617,
+      "step": 56
+    },
+    {
+      "epoch": 0.10836501901140684,
+      "grad_norm": 0.9035849904027986,
+      "learning_rate": 1.0754716981132076e-05,
+      "loss": 0.4632,
+      "step": 57
+    },
+    {
+      "epoch": 0.11026615969581749,
+      "grad_norm": 0.8231141621156364,
+      "learning_rate": 1.0943396226415095e-05,
+      "loss": 0.4991,
+      "step": 58
+    },
+    {
+      "epoch": 0.11216730038022814,
+      "grad_norm": 0.7615541291151133,
+      "learning_rate": 1.1132075471698115e-05,
+      "loss": 0.4684,
+      "step": 59
+    },
+    {
+      "epoch": 0.11406844106463879,
+      "grad_norm": 0.8619525658796648,
+      "learning_rate": 1.1320754716981132e-05,
+      "loss": 0.4889,
+      "step": 60
+    },
+    {
+      "epoch": 0.11596958174904944,
+      "grad_norm": 0.8072015198687831,
+      "learning_rate": 1.1509433962264152e-05,
+      "loss": 0.4617,
+      "step": 61
+    },
+    {
+      "epoch": 0.11787072243346007,
+      "grad_norm": 0.8927939283698709,
+      "learning_rate": 1.169811320754717e-05,
+      "loss": 0.4796,
+      "step": 62
+    },
+    {
+      "epoch": 0.11977186311787072,
+      "grad_norm": 0.8113351051027795,
+      "learning_rate": 1.188679245283019e-05,
+      "loss": 0.4578,
+      "step": 63
+    },
+    {
+      "epoch": 0.12167300380228137,
+      "grad_norm": 0.8313670463366891,
+      "learning_rate": 1.2075471698113209e-05,
+      "loss": 0.5084,
+      "step": 64
+    },
+    {
+      "epoch": 0.12357414448669202,
+      "grad_norm": 0.8382033475084935,
+      "learning_rate": 1.2264150943396227e-05,
+      "loss": 0.4776,
+      "step": 65
+    },
+    {
+      "epoch": 0.12547528517110265,
+      "grad_norm": 0.873148439237489,
+      "learning_rate": 1.2452830188679246e-05,
+      "loss": 0.4614,
+      "step": 66
+    },
+    {
+      "epoch": 0.12737642585551331,
+      "grad_norm": 0.8033899203362915,
+      "learning_rate": 1.2641509433962264e-05,
+      "loss": 0.4517,
+      "step": 67
+    },
+    {
+      "epoch": 0.12927756653992395,
+      "grad_norm": 0.8468039539232246,
+      "learning_rate": 1.2830188679245283e-05,
+      "loss": 0.4759,
+      "step": 68
+    },
+    {
+      "epoch": 0.1311787072243346,
+      "grad_norm": 0.9114257746120382,
+      "learning_rate": 1.3018867924528303e-05,
+      "loss": 0.4972,
+      "step": 69
+    },
+    {
+      "epoch": 0.13307984790874525,
+      "grad_norm": 0.7848838019174844,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 0.4702,
+      "step": 70
+    },
+    {
+      "epoch": 0.13498098859315588,
+      "grad_norm": 0.8469439793857244,
+      "learning_rate": 1.339622641509434e-05,
+      "loss": 0.4559,
+      "step": 71
+    },
+    {
+      "epoch": 0.13688212927756654,
+      "grad_norm": 0.8476364524813007,
+      "learning_rate": 1.3584905660377358e-05,
+      "loss": 0.4758,
+      "step": 72
+    },
+    {
+      "epoch": 0.13878326996197718,
+      "grad_norm": 0.8000017650555116,
+      "learning_rate": 1.3773584905660378e-05,
+      "loss": 0.463,
+      "step": 73
+    },
+    {
+      "epoch": 0.14068441064638784,
+      "grad_norm": 0.8227147797427505,
+      "learning_rate": 1.3962264150943397e-05,
+      "loss": 0.4561,
+      "step": 74
+    },
+    {
+      "epoch": 0.14258555133079848,
+      "grad_norm": 0.8309673613968465,
+      "learning_rate": 1.4150943396226415e-05,
+      "loss": 0.4628,
+      "step": 75
+    },
+    {
+      "epoch": 0.1444866920152091,
+      "grad_norm": 0.8682211816392789,
+      "learning_rate": 1.4339622641509435e-05,
+      "loss": 0.4936,
+      "step": 76
+    },
+    {
+      "epoch": 0.14638783269961977,
+      "grad_norm": 0.8293738944592072,
+      "learning_rate": 1.4528301886792452e-05,
+      "loss": 0.472,
+      "step": 77
+    },
+    {
+      "epoch": 0.1482889733840304,
+      "grad_norm": 0.8311959083126518,
+      "learning_rate": 1.4716981132075472e-05,
+      "loss": 0.4565,
+      "step": 78
+    },
+    {
+      "epoch": 0.15019011406844107,
+      "grad_norm": 0.886596067131909,
+      "learning_rate": 1.4905660377358491e-05,
+      "loss": 0.4785,
+      "step": 79
+    },
+    {
+      "epoch": 0.1520912547528517,
+      "grad_norm": 0.8880752001108226,
+      "learning_rate": 1.5094339622641511e-05,
+      "loss": 0.4893,
+      "step": 80
+    },
+    {
+      "epoch": 0.15399239543726237,
+      "grad_norm": 0.7731681622222218,
+      "learning_rate": 1.5283018867924532e-05,
+      "loss": 0.4835,
+      "step": 81
+    },
+    {
+      "epoch": 0.155893536121673,
+      "grad_norm": 0.8682520169633279,
+      "learning_rate": 1.547169811320755e-05,
+      "loss": 0.4532,
+      "step": 82
+    },
+    {
+      "epoch": 0.15779467680608364,
+      "grad_norm": 0.8957928382030637,
+      "learning_rate": 1.5660377358490568e-05,
+      "loss": 0.4712,
+      "step": 83
+    },
+    {
+      "epoch": 0.1596958174904943,
+      "grad_norm": 0.9071350994714376,
+      "learning_rate": 1.5849056603773586e-05,
+      "loss": 0.4695,
+      "step": 84
+    },
+    {
+      "epoch": 0.16159695817490494,
+      "grad_norm": 0.9689924369755653,
+      "learning_rate": 1.6037735849056607e-05,
+      "loss": 0.4862,
+      "step": 85
+    },
+    {
+      "epoch": 0.1634980988593156,
+      "grad_norm": 0.9184630886339062,
+      "learning_rate": 1.6226415094339625e-05,
+      "loss": 0.4538,
+      "step": 86
+    },
+    {
+      "epoch": 0.16539923954372623,
+      "grad_norm": 0.9033149640262258,
+      "learning_rate": 1.6415094339622643e-05,
+      "loss": 0.4919,
+      "step": 87
+    },
+    {
+      "epoch": 0.16730038022813687,
+      "grad_norm": 0.8755433448025809,
+      "learning_rate": 1.6603773584905664e-05,
+      "loss": 0.4972,
+      "step": 88
+    },
+    {
+      "epoch": 0.16920152091254753,
+      "grad_norm": 0.9623989857102851,
+      "learning_rate": 1.679245283018868e-05,
+      "loss": 0.4767,
+      "step": 89
+    },
+    {
+      "epoch": 0.17110266159695817,
+      "grad_norm": 0.8555585233710505,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.4862,
+      "step": 90
+    },
+    {
+      "epoch": 0.17300380228136883,
+      "grad_norm": 0.9283993299739618,
+      "learning_rate": 1.716981132075472e-05,
+      "loss": 0.4661,
+      "step": 91
+    },
+    {
+      "epoch": 0.17490494296577946,
+      "grad_norm": 0.8480863383200016,
+      "learning_rate": 1.735849056603774e-05,
+      "loss": 0.4811,
+      "step": 92
+    },
+    {
+      "epoch": 0.17680608365019013,
+      "grad_norm": 0.7745282827642337,
+      "learning_rate": 1.7547169811320756e-05,
+      "loss": 0.5009,
+      "step": 93
+    },
+    {
+      "epoch": 0.17870722433460076,
+      "grad_norm": 0.8859319913529026,
+      "learning_rate": 1.7735849056603774e-05,
+      "loss": 0.4901,
+      "step": 94
+    },
+    {
+      "epoch": 0.1806083650190114,
+      "grad_norm": 0.8445364566835015,
+      "learning_rate": 1.7924528301886795e-05,
+      "loss": 0.4871,
+      "step": 95
+    },
+    {
+      "epoch": 0.18250950570342206,
+      "grad_norm": 0.7938683332258069,
+      "learning_rate": 1.8113207547169813e-05,
+      "loss": 0.4595,
+      "step": 96
+    },
+    {
+      "epoch": 0.1844106463878327,
+      "grad_norm": 0.978898944191785,
+      "learning_rate": 1.830188679245283e-05,
+      "loss": 0.4742,
+      "step": 97
+    },
+    {
+      "epoch": 0.18631178707224336,
+      "grad_norm": 0.8013217626607537,
+      "learning_rate": 1.8490566037735852e-05,
+      "loss": 0.44,
+      "step": 98
+    },
+    {
+      "epoch": 0.188212927756654,
+      "grad_norm": 1.0048972496199768,
+      "learning_rate": 1.867924528301887e-05,
+      "loss": 0.4684,
+      "step": 99
+    },
+    {
+      "epoch": 0.19011406844106463,
+      "grad_norm": 0.821923428407393,
+      "learning_rate": 1.8867924528301888e-05,
+      "loss": 0.4735,
+      "step": 100
+    },
+    {
+      "epoch": 0.1920152091254753,
+      "grad_norm": 0.9231221325177691,
+      "learning_rate": 1.905660377358491e-05,
+      "loss": 0.4986,
+      "step": 101
+    },
+    {
+      "epoch": 0.19391634980988592,
+      "grad_norm": 0.8546733264282385,
+      "learning_rate": 1.9245283018867927e-05,
+      "loss": 0.4443,
+      "step": 102
+    },
+    {
+      "epoch": 0.1958174904942966,
+      "grad_norm": 0.8531487317624786,
+      "learning_rate": 1.9433962264150945e-05,
+      "loss": 0.4797,
+      "step": 103
+    },
+    {
+      "epoch": 0.19771863117870722,
+      "grad_norm": 0.8569328197403074,
+      "learning_rate": 1.9622641509433963e-05,
+      "loss": 0.4559,
+      "step": 104
+    },
+    {
+      "epoch": 0.19961977186311788,
+      "grad_norm": 0.9034077731112596,
+      "learning_rate": 1.9811320754716984e-05,
+      "loss": 0.4652,
+      "step": 105
+    },
+    {
+      "epoch": 0.20152091254752852,
+      "grad_norm": 0.9610503645357051,
+      "learning_rate": 2e-05,
+      "loss": 0.4627,
+      "step": 106
+    },
+    {
+      "epoch": 0.20342205323193915,
+      "grad_norm": 1.0172363811845022,
+      "learning_rate": 1.9999944857420527e-05,
+      "loss": 0.4551,
+      "step": 107
+    },
+    {
+      "epoch": 0.20532319391634982,
+      "grad_norm": 0.9901514420625347,
+      "learning_rate": 1.9999779430290247e-05,
+      "loss": 0.466,
+      "step": 108
+    },
+    {
+      "epoch": 0.20722433460076045,
+      "grad_norm": 0.9758642568720305,
+      "learning_rate": 1.9999503720433575e-05,
+      "loss": 0.4913,
+      "step": 109
+    },
+    {
+      "epoch": 0.20912547528517111,
+      "grad_norm": 0.9411898461799564,
+      "learning_rate": 1.999911773089118e-05,
+      "loss": 0.4549,
+      "step": 110
+    },
+    {
+      "epoch": 0.21102661596958175,
+      "grad_norm": 1.0070983105099944,
+      "learning_rate": 1.999862146591996e-05,
+      "loss": 0.4408,
+      "step": 111
+    },
+    {
+      "epoch": 0.21292775665399238,
+      "grad_norm": 0.8550946539823405,
+      "learning_rate": 1.9998014930992976e-05,
+      "loss": 0.4608,
+      "step": 112
+    },
+    {
+      "epoch": 0.21482889733840305,
+      "grad_norm": 1.0222283114098931,
+      "learning_rate": 1.9997298132799408e-05,
+      "loss": 0.492,
+      "step": 113
+    },
+    {
+      "epoch": 0.21673003802281368,
+      "grad_norm": 0.8336696438646395,
+      "learning_rate": 1.9996471079244477e-05,
+      "loss": 0.4893,
+      "step": 114
+    },
+    {
+      "epoch": 0.21863117870722434,
+      "grad_norm": 0.9179528078733727,
+      "learning_rate": 1.999553377944936e-05,
+      "loss": 0.4726,
+      "step": 115
+    },
+    {
+      "epoch": 0.22053231939163498,
+      "grad_norm": 0.8769475416641306,
+      "learning_rate": 1.9994486243751076e-05,
+      "loss": 0.4577,
+      "step": 116
+    },
+    {
+      "epoch": 0.2224334600760456,
+      "grad_norm": 0.8056187688990881,
+      "learning_rate": 1.9993328483702393e-05,
+      "loss": 0.4719,
+      "step": 117
+    },
+    {
+      "epoch": 0.22433460076045628,
+      "grad_norm": 0.9467742648790678,
+      "learning_rate": 1.999206051207169e-05,
+      "loss": 0.4699,
+      "step": 118
+    },
+    {
+      "epoch": 0.2262357414448669,
+      "grad_norm": 0.8956052623957506,
+      "learning_rate": 1.9990682342842805e-05,
+      "loss": 0.48,
+      "step": 119
+    },
+    {
+      "epoch": 0.22813688212927757,
+      "grad_norm": 0.8988003280768448,
+      "learning_rate": 1.99891939912149e-05,
+      "loss": 0.4451,
+      "step": 120
+    },
+    {
+      "epoch": 0.2300380228136882,
+      "grad_norm": 0.8630801281689711,
+      "learning_rate": 1.9987595473602292e-05,
+      "loss": 0.4929,
+      "step": 121
+    },
+    {
+      "epoch": 0.23193916349809887,
+      "grad_norm": 0.9487161860879703,
+      "learning_rate": 1.9985886807634246e-05,
+      "loss": 0.4897,
+      "step": 122
+    },
+    {
+      "epoch": 0.2338403041825095,
+      "grad_norm": 0.8095239563393253,
+      "learning_rate": 1.9984068012154824e-05,
+      "loss": 0.4586,
+      "step": 123
+    },
+    {
+      "epoch": 0.23574144486692014,
+      "grad_norm": 0.9100607260452165,
+      "learning_rate": 1.9982139107222634e-05,
+      "loss": 0.4622,
+      "step": 124
+    },
+    {
+      "epoch": 0.2376425855513308,
+      "grad_norm": 0.8965853320304211,
+      "learning_rate": 1.9980100114110637e-05,
+      "loss": 0.4552,
+      "step": 125
+    },
+    {
+      "epoch": 0.23954372623574144,
+      "grad_norm": 0.9587796338450637,
+      "learning_rate": 1.99779510553059e-05,
+      "loss": 0.4509,
+      "step": 126
+    },
+    {
+      "epoch": 0.2414448669201521,
+      "grad_norm": 0.8128181982292385,
+      "learning_rate": 1.9975691954509347e-05,
+      "loss": 0.4735,
+      "step": 127
+    },
+    {
+      "epoch": 0.24334600760456274,
+      "grad_norm": 1.0749212411393139,
+      "learning_rate": 1.9973322836635517e-05,
+      "loss": 0.4623,
+      "step": 128
+    },
+    {
+      "epoch": 0.24524714828897337,
+      "grad_norm": 0.8408975422161333,
+      "learning_rate": 1.997084372781226e-05,
+      "loss": 0.4905,
+      "step": 129
+    },
+    {
+      "epoch": 0.24714828897338403,
+      "grad_norm": 0.8832438871438801,
+      "learning_rate": 1.9968254655380465e-05,
+      "loss": 0.4738,
+      "step": 130
+    },
+    {
+      "epoch": 0.24904942965779467,
+      "grad_norm": 0.8168040593793138,
+      "learning_rate": 1.996555564789376e-05,
+      "loss": 0.4749,
+      "step": 131
+    },
+    {
+      "epoch": 0.2509505703422053,
+      "grad_norm": 0.9016000415038536,
+      "learning_rate": 1.996274673511819e-05,
+      "loss": 0.4298,
+      "step": 132
+    },
+    {
+      "epoch": 0.25285171102661597,
+      "grad_norm": 0.7763368605754961,
+      "learning_rate": 1.99598279480319e-05,
+      "loss": 0.4539,
+      "step": 133
+    },
+    {
+      "epoch": 0.25475285171102663,
+      "grad_norm": 0.820424048966225,
+      "learning_rate": 1.9956799318824776e-05,
+      "loss": 0.4608,
+      "step": 134
+    },
+    {
+      "epoch": 0.25665399239543724,
+      "grad_norm": 0.8768597149203753,
+      "learning_rate": 1.99536608808981e-05,
+      "loss": 0.4654,
+      "step": 135
+    },
+    {
+      "epoch": 0.2585551330798479,
+      "grad_norm": 0.7911679404363704,
+      "learning_rate": 1.995041266886419e-05,
+      "loss": 0.4794,
+      "step": 136
+    },
+    {
+      "epoch": 0.26045627376425856,
+      "grad_norm": 0.8632487730529239,
+      "learning_rate": 1.9947054718545996e-05,
+      "loss": 0.4802,
+      "step": 137
+    },
+    {
+      "epoch": 0.2623574144486692,
+      "grad_norm": 0.7794814703146564,
+      "learning_rate": 1.994358706697674e-05,
+      "loss": 0.4483,
+      "step": 138
+    },
+    {
+      "epoch": 0.26425855513307983,
+      "grad_norm": 0.8349568357042889,
+      "learning_rate": 1.9940009752399462e-05,
+      "loss": 0.4607,
+      "step": 139
+    },
+    {
+      "epoch": 0.2661596958174905,
+      "grad_norm": 0.8484393953947692,
+      "learning_rate": 1.9936322814266634e-05,
+      "loss": 0.4834,
+      "step": 140
+    },
+    {
+      "epoch": 0.26806083650190116,
+      "grad_norm": 0.8059606713904582,
+      "learning_rate": 1.9932526293239713e-05,
+      "loss": 0.4816,
+      "step": 141
+    },
+    {
+      "epoch": 0.26996197718631176,
+      "grad_norm": 0.8351412581603349,
+      "learning_rate": 1.9928620231188694e-05,
+      "loss": 0.4388,
+      "step": 142
+    },
+    {
+      "epoch": 0.2718631178707224,
+      "grad_norm": 0.8928088131393197,
+      "learning_rate": 1.992460467119164e-05,
+      "loss": 0.4912,
+      "step": 143
+    },
+    {
+      "epoch": 0.2737642585551331,
+      "grad_norm": 0.9076694919320091,
+      "learning_rate": 1.992047965753422e-05,
+      "loss": 0.4869,
+      "step": 144
+    },
+    {
+      "epoch": 0.27566539923954375,
+      "grad_norm": 0.9025652946054866,
+      "learning_rate": 1.991624523570922e-05,
+      "loss": 0.5028,
+      "step": 145
+    },
+    {
+      "epoch": 0.27756653992395436,
+      "grad_norm": 0.9195503388546261,
+      "learning_rate": 1.9911901452416012e-05,
+      "loss": 0.4439,
+      "step": 146
+    },
+    {
+      "epoch": 0.279467680608365,
+      "grad_norm": 0.933562200992786,
+      "learning_rate": 1.9907448355560094e-05,
+      "loss": 0.4576,
+      "step": 147
+    },
+    {
+      "epoch": 0.2813688212927757,
+      "grad_norm": 0.9438238152638735,
+      "learning_rate": 1.9902885994252506e-05,
+      "loss": 0.4524,
+      "step": 148
+    },
+    {
+      "epoch": 0.2832699619771863,
+      "grad_norm": 0.863683544760749,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.4596,
+      "step": 149
+    },
+    {
+      "epoch": 0.28517110266159695,
+      "grad_norm": 1.0227347284052737,
+      "learning_rate": 1.9893433680751105e-05,
+      "loss": 0.4701,
+      "step": 150
+    },
+    {
+      "epoch": 0.2870722433460076,
+      "grad_norm": 0.7912531250132444,
+      "learning_rate": 1.9888543832802277e-05,
+      "loss": 0.4373,
+      "step": 151
+    },
+    {
+      "epoch": 0.2889733840304182,
+      "grad_norm": 0.8748540341014369,
+      "learning_rate": 1.9883544928890612e-05,
+      "loss": 0.4695,
+      "step": 152
+    },
+    {
+      "epoch": 0.2908745247148289,
+      "grad_norm": 0.8606809032132475,
+      "learning_rate": 1.9878437024146603e-05,
+      "loss": 0.5046,
+      "step": 153
+    },
+    {
+      "epoch": 0.29277566539923955,
+      "grad_norm": 0.8961713137421418,
+      "learning_rate": 1.9873220174902857e-05,
+      "loss": 0.4474,
+      "step": 154
+    },
+    {
+      "epoch": 0.2946768060836502,
+      "grad_norm": 0.8122411943583865,
+      "learning_rate": 1.986789443869348e-05,
+      "loss": 0.4591,
+      "step": 155
+    },
+    {
+      "epoch": 0.2965779467680608,
+      "grad_norm": 0.779125616681954,
+      "learning_rate": 1.9862459874253438e-05,
+      "loss": 0.43,
+      "step": 156
+    },
+    {
+      "epoch": 0.2984790874524715,
+      "grad_norm": 0.8822448946512448,
+      "learning_rate": 1.985691654151791e-05,
+      "loss": 0.4391,
+      "step": 157
+    },
+    {
+      "epoch": 0.30038022813688214,
+      "grad_norm": 0.795404824726714,
+      "learning_rate": 1.9851264501621635e-05,
+      "loss": 0.4701,
+      "step": 158
+    },
+    {
+      "epoch": 0.30228136882129275,
+      "grad_norm": 0.8655010991173399,
+      "learning_rate": 1.984550381689822e-05,
+      "loss": 0.4502,
+      "step": 159
+    },
+    {
+      "epoch": 0.3041825095057034,
+      "grad_norm": 0.7746028743142694,
+      "learning_rate": 1.983963455087946e-05,
+      "loss": 0.4614,
+      "step": 160
+    },
+    {
+      "epoch": 0.3060836501901141,
+      "grad_norm": 0.7928753727251139,
+      "learning_rate": 1.983365676829466e-05,
+      "loss": 0.4727,
+      "step": 161
+    },
+    {
+      "epoch": 0.30798479087452474,
+      "grad_norm": 0.819275816034261,
+      "learning_rate": 1.982757053506989e-05,
+      "loss": 0.4419,
+      "step": 162
+    },
+    {
+      "epoch": 0.30988593155893535,
+      "grad_norm": 0.7816057094946948,
+      "learning_rate": 1.9821375918327268e-05,
+      "loss": 0.4702,
+      "step": 163
+    },
+    {
+      "epoch": 0.311787072243346,
+      "grad_norm": 0.7675096328203197,
+      "learning_rate": 1.981507298638422e-05,
+      "loss": 0.4392,
+      "step": 164
+    },
+    {
+      "epoch": 0.31368821292775667,
+      "grad_norm": 0.8257814334127902,
+      "learning_rate": 1.9808661808752735e-05,
+      "loss": 0.4752,
+      "step": 165
+    },
+    {
+      "epoch": 0.3155893536121673,
+      "grad_norm": 0.8005285120540652,
+      "learning_rate": 1.980214245613858e-05,
+      "loss": 0.4518,
+      "step": 166
+    },
+    {
+      "epoch": 0.31749049429657794,
+      "grad_norm": 0.8188369412150304,
+      "learning_rate": 1.979551500044055e-05,
+      "loss": 0.4572,
+      "step": 167
+    },
+    {
+      "epoch": 0.3193916349809886,
+      "grad_norm": 0.7228513971968181,
+      "learning_rate": 1.9788779514749635e-05,
+      "loss": 0.4497,
+      "step": 168
+    },
+    {
+      "epoch": 0.32129277566539927,
+      "grad_norm": 0.8510544665422051,
+      "learning_rate": 1.978193607334826e-05,
+      "loss": 0.4581,
+      "step": 169
+    },
+    {
+      "epoch": 0.3231939163498099,
+      "grad_norm": 0.8257581265764004,
+      "learning_rate": 1.977498475170941e-05,
+      "loss": 0.4667,
+      "step": 170
+    },
+    {
+      "epoch": 0.32509505703422054,
+      "grad_norm": 0.724343083453893,
+      "learning_rate": 1.9767925626495857e-05,
+      "loss": 0.46,
+      "step": 171
+    },
+    {
+      "epoch": 0.3269961977186312,
+      "grad_norm": 0.7968592353850333,
+      "learning_rate": 1.9760758775559275e-05,
+      "loss": 0.4522,
+      "step": 172
+    },
+    {
+      "epoch": 0.3288973384030418,
+      "grad_norm": 0.8253473458934747,
+      "learning_rate": 1.975348427793939e-05,
+      "loss": 0.4811,
+      "step": 173
+    },
+    {
+      "epoch": 0.33079847908745247,
+      "grad_norm": 0.7097504714709826,
+      "learning_rate": 1.9746102213863113e-05,
+      "loss": 0.4579,
+      "step": 174
+    },
+    {
+      "epoch": 0.33269961977186313,
+      "grad_norm": 0.7709579255181828,
+      "learning_rate": 1.973861266474366e-05,
+      "loss": 0.4519,
+      "step": 175
+    },
+    {
+      "epoch": 0.33460076045627374,
+      "grad_norm": 0.8277981328888928,
+      "learning_rate": 1.9731015713179643e-05,
+      "loss": 0.481,
+      "step": 176
+    },
+    {
+      "epoch": 0.3365019011406844,
+      "grad_norm": 0.7753653732465653,
+      "learning_rate": 1.9723311442954163e-05,
+      "loss": 0.4821,
+      "step": 177
+    },
+    {
+      "epoch": 0.33840304182509506,
+      "grad_norm": 0.7563752441590016,
+      "learning_rate": 1.9715499939033883e-05,
+      "loss": 0.4443,
+      "step": 178
+    },
+    {
+      "epoch": 0.3403041825095057,
+      "grad_norm": 0.7965730899205258,
+      "learning_rate": 1.9707581287568094e-05,
+      "loss": 0.4274,
+      "step": 179
+    },
+    {
+      "epoch": 0.34220532319391633,
+      "grad_norm": 0.7275151268563129,
+      "learning_rate": 1.969955557588778e-05,
+      "loss": 0.4314,
+      "step": 180
+    },
+    {
+      "epoch": 0.344106463878327,
+      "grad_norm": 0.8375183480576625,
+      "learning_rate": 1.9691422892504626e-05,
+      "loss": 0.4509,
+      "step": 181
+    },
+    {
+      "epoch": 0.34600760456273766,
+      "grad_norm": 0.8048721347032028,
+      "learning_rate": 1.968318332711006e-05,
+      "loss": 0.4833,
+      "step": 182
+    },
+    {
+      "epoch": 0.34790874524714827,
+      "grad_norm": 0.7998180395088168,
+      "learning_rate": 1.9674836970574253e-05,
+      "loss": 0.4607,
+      "step": 183
+    },
+    {
+      "epoch": 0.34980988593155893,
+      "grad_norm": 0.7538344152518309,
+      "learning_rate": 1.966638391494514e-05,
+      "loss": 0.4408,
+      "step": 184
+    },
+    {
+      "epoch": 0.3517110266159696,
+      "grad_norm": 0.706894591870399,
+      "learning_rate": 1.9657824253447378e-05,
+      "loss": 0.4414,
+      "step": 185
+    },
+    {
+      "epoch": 0.35361216730038025,
+      "grad_norm": 0.7603575638803718,
+      "learning_rate": 1.9649158080481327e-05,
+      "loss": 0.4641,
+      "step": 186
+    },
+    {
+      "epoch": 0.35551330798479086,
+      "grad_norm": 0.7527366529086855,
+      "learning_rate": 1.964038549162201e-05,
+      "loss": 0.4356,
+      "step": 187
+    },
+    {
+      "epoch": 0.3574144486692015,
+      "grad_norm": 0.742246215334776,
+      "learning_rate": 1.963150658361807e-05,
+      "loss": 0.4638,
+      "step": 188
+    },
+    {
+      "epoch": 0.3593155893536122,
+      "grad_norm": 0.8352540917694925,
+      "learning_rate": 1.962252145439068e-05,
+      "loss": 0.4217,
+      "step": 189
+    },
+    {
+      "epoch": 0.3612167300380228,
+      "grad_norm": 0.8139149187540878,
+      "learning_rate": 1.9613430203032486e-05,
+      "loss": 0.4486,
+      "step": 190
+    },
+    {
+      "epoch": 0.36311787072243346,
+      "grad_norm": 0.7607053993602467,
+      "learning_rate": 1.9604232929806493e-05,
+      "loss": 0.467,
+      "step": 191
+    },
+    {
+      "epoch": 0.3650190114068441,
+      "grad_norm": 0.8034771919817714,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.4353,
+      "step": 192
+    },
+    {
+      "epoch": 0.3669201520912547,
+      "grad_norm": 0.7528937429985373,
+      "learning_rate": 1.9585520724648354e-05,
+      "loss": 0.4389,
+      "step": 193
+    },
+    {
+      "epoch": 0.3688212927756654,
+      "grad_norm": 0.6670585742782636,
+      "learning_rate": 1.957600599908406e-05,
+      "loss": 0.4489,
+      "step": 194
+    },
+    {
+      "epoch": 0.37072243346007605,
+      "grad_norm": 0.9250225839045535,
+      "learning_rate": 1.95663856643854e-05,
+      "loss": 0.4366,
+      "step": 195
+    },
+    {
+      "epoch": 0.3726235741444867,
+      "grad_norm": 0.7805834549736975,
+      "learning_rate": 1.955665982665038e-05,
+      "loss": 0.4442,
+      "step": 196
+    },
+    {
+      "epoch": 0.3745247148288973,
+      "grad_norm": 0.759959666303769,
+      "learning_rate": 1.9546828593140565e-05,
+      "loss": 0.4634,
+      "step": 197
+    },
+    {
+      "epoch": 0.376425855513308,
+      "grad_norm": 0.854464165159959,
+      "learning_rate": 1.9536892072279863e-05,
+      "loss": 0.4674,
+      "step": 198
+    },
+    {
+      "epoch": 0.37832699619771865,
+      "grad_norm": 0.7601546641187042,
+      "learning_rate": 1.9526850373653356e-05,
+      "loss": 0.4637,
+      "step": 199
+    },
+    {
+      "epoch": 0.38022813688212925,
+      "grad_norm": 0.8050328276825799,
+      "learning_rate": 1.9516703608006074e-05,
+      "loss": 0.4157,
+      "step": 200
+    },
+    {
+      "epoch": 0.3821292775665399,
+      "grad_norm": 0.7438926620885794,
+      "learning_rate": 1.9506451887241787e-05,
+      "loss": 0.4476,
+      "step": 201
+    },
+    {
+      "epoch": 0.3840304182509506,
+      "grad_norm": 1.0081869636135679,
+      "learning_rate": 1.949609532442176e-05,
+      "loss": 0.4683,
+      "step": 202
+    },
+    {
+      "epoch": 0.38593155893536124,
+      "grad_norm": 0.8100411240085761,
+      "learning_rate": 1.9485634033763507e-05,
+      "loss": 0.4367,
+      "step": 203
+    },
+    {
+      "epoch": 0.38783269961977185,
+      "grad_norm": 0.8746891174272511,
+      "learning_rate": 1.9475068130639543e-05,
+      "loss": 0.4357,
+      "step": 204
+    },
+    {
+      "epoch": 0.3897338403041825,
+      "grad_norm": 0.6935414217462783,
+      "learning_rate": 1.9464397731576093e-05,
+      "loss": 0.4342,
+      "step": 205
+    },
+    {
+      "epoch": 0.3916349809885932,
+      "grad_norm": 0.8882200579728433,
+      "learning_rate": 1.945362295425183e-05,
+      "loss": 0.4524,
+      "step": 206
+    },
+    {
+      "epoch": 0.3935361216730038,
+      "grad_norm": 0.7098826049100907,
+      "learning_rate": 1.944274391749655e-05,
+      "loss": 0.4288,
+      "step": 207
+    },
+    {
+      "epoch": 0.39543726235741444,
+      "grad_norm": 0.7899575089208016,
+      "learning_rate": 1.9431760741289886e-05,
+      "loss": 0.4404,
+      "step": 208
+    },
+    {
+      "epoch": 0.3973384030418251,
+      "grad_norm": 0.7739384675933767,
+      "learning_rate": 1.942067354675997e-05,
+      "loss": 0.4593,
+      "step": 209
+    },
+    {
+      "epoch": 0.39923954372623577,
+      "grad_norm": 0.7061033603201634,
+      "learning_rate": 1.9409482456182105e-05,
+      "loss": 0.4403,
+      "step": 210
+    },
+    {
+      "epoch": 0.4011406844106464,
+      "grad_norm": 0.784408371746286,
+      "learning_rate": 1.939818759297741e-05,
+      "loss": 0.4428,
+      "step": 211
+    },
+    {
+      "epoch": 0.40304182509505704,
+      "grad_norm": 0.7745481953996032,
+      "learning_rate": 1.9386789081711465e-05,
+      "loss": 0.4392,
+      "step": 212
+    },
+    {
+      "epoch": 0.4049429657794677,
+      "grad_norm": 0.7218520689105371,
+      "learning_rate": 1.9375287048092927e-05,
+      "loss": 0.436,
+      "step": 213
+    },
+    {
+      "epoch": 0.4068441064638783,
+      "grad_norm": 0.7621861074258811,
+      "learning_rate": 1.9363681618972166e-05,
+      "loss": 0.4513,
+      "step": 214
+    },
+    {
+      "epoch": 0.40874524714828897,
+      "grad_norm": 0.7932530408366242,
+      "learning_rate": 1.9351972922339835e-05,
+      "loss": 0.4581,
+      "step": 215
+    },
+    {
+      "epoch": 0.41064638783269963,
+      "grad_norm": 0.8945158319906973,
+      "learning_rate": 1.9340161087325483e-05,
+      "loss": 0.4523,
+      "step": 216
+    },
+    {
+      "epoch": 0.41254752851711024,
+      "grad_norm": 0.6490217060974268,
+      "learning_rate": 1.9328246244196117e-05,
+      "loss": 0.4361,
+      "step": 217
+    },
+    {
+      "epoch": 0.4144486692015209,
+      "grad_norm": 0.7750039631705506,
+      "learning_rate": 1.931622852435478e-05,
+      "loss": 0.4539,
+      "step": 218
+    },
+    {
+      "epoch": 0.41634980988593157,
+      "grad_norm": 0.7891506216193496,
+      "learning_rate": 1.930410806033908e-05,
+      "loss": 0.4311,
+      "step": 219
+    },
+    {
+      "epoch": 0.41825095057034223,
+      "grad_norm": 0.7088575022686695,
+      "learning_rate": 1.929188498581975e-05,
+      "loss": 0.4353,
+      "step": 220
+    },
+    {
+      "epoch": 0.42015209125475284,
+      "grad_norm": 0.8686661930047552,
+      "learning_rate": 1.9279559435599164e-05,
+      "loss": 0.4619,
+      "step": 221
+    },
+    {
+      "epoch": 0.4220532319391635,
+      "grad_norm": 0.8199255659017032,
+      "learning_rate": 1.926713154560984e-05,
+      "loss": 0.4583,
+      "step": 222
+    },
+    {
+      "epoch": 0.42395437262357416,
+      "grad_norm": 0.8286700725599053,
+      "learning_rate": 1.9254601452912972e-05,
+      "loss": 0.4625,
+      "step": 223
+    },
+    {
+      "epoch": 0.42585551330798477,
+      "grad_norm": 0.770190339380276,
+      "learning_rate": 1.924196929569688e-05,
+      "loss": 0.4226,
+      "step": 224
+    },
+    {
+      "epoch": 0.42775665399239543,
+      "grad_norm": 0.7531712221421968,
+      "learning_rate": 1.922923521327551e-05,
+      "loss": 0.4885,
+      "step": 225
+    },
+    {
+      "epoch": 0.4296577946768061,
+      "grad_norm": 0.8358243726895588,
+      "learning_rate": 1.9216399346086893e-05,
+      "loss": 0.4494,
+      "step": 226
+    },
+    {
+      "epoch": 0.43155893536121676,
+      "grad_norm": 0.738096236093835,
+      "learning_rate": 1.9203461835691596e-05,
+      "loss": 0.4527,
+      "step": 227
+    },
+    {
+      "epoch": 0.43346007604562736,
+      "grad_norm": 0.7502916182738939,
+      "learning_rate": 1.9190422824771158e-05,
+      "loss": 0.4436,
+      "step": 228
+    },
+    {
+      "epoch": 0.435361216730038,
+      "grad_norm": 0.7055382452809816,
+      "learning_rate": 1.9177282457126515e-05,
+      "loss": 0.4284,
+      "step": 229
+    },
+    {
+      "epoch": 0.4372623574144487,
+      "grad_norm": 0.7084095946817636,
+      "learning_rate": 1.9164040877676425e-05,
+      "loss": 0.4508,
+      "step": 230
+    },
+    {
+      "epoch": 0.4391634980988593,
+      "grad_norm": 0.7916958474551536,
+      "learning_rate": 1.9150698232455853e-05,
+      "loss": 0.4104,
+      "step": 231
+    },
+    {
+      "epoch": 0.44106463878326996,
+      "grad_norm": 0.6889180206555621,
+      "learning_rate": 1.913725466861438e-05,
+      "loss": 0.4433,
+      "step": 232
+    },
+    {
+      "epoch": 0.4429657794676806,
+      "grad_norm": 0.7144333741543766,
+      "learning_rate": 1.9123710334414552e-05,
+      "loss": 0.4524,
+      "step": 233
+    },
+    {
+      "epoch": 0.4448669201520912,
+      "grad_norm": 0.7051930859307878,
+      "learning_rate": 1.911006537923029e-05,
+      "loss": 0.4328,
+      "step": 234
+    },
+    {
+      "epoch": 0.4467680608365019,
+      "grad_norm": 0.7622820686587574,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.423,
+      "step": 235
+    },
+    {
+      "epoch": 0.44866920152091255,
+      "grad_norm": 0.6985168387427974,
+      "learning_rate": 1.908247420895089e-05,
+      "loss": 0.4326,
+      "step": 236
+    },
+    {
+      "epoch": 0.4505703422053232,
+      "grad_norm": 0.7187887370047864,
+      "learning_rate": 1.9068528298145418e-05,
+      "loss": 0.446,
+      "step": 237
+    },
+    {
+      "epoch": 0.4524714828897338,
+      "grad_norm": 0.7167170166545045,
+      "learning_rate": 1.905448237493147e-05,
+      "loss": 0.4433,
+      "step": 238
+    },
+    {
+      "epoch": 0.4543726235741445,
+      "grad_norm": 0.7408627641668901,
+      "learning_rate": 1.9040336594214727e-05,
+      "loss": 0.4382,
+      "step": 239
+    },
+    {
+      "epoch": 0.45627376425855515,
+      "grad_norm": 0.6692335966997324,
+      "learning_rate": 1.9026091112002163e-05,
+      "loss": 0.4185,
+      "step": 240
+    },
+    {
+      "epoch": 0.45817490494296575,
+      "grad_norm": 0.691267735094259,
+      "learning_rate": 1.90117460854003e-05,
+      "loss": 0.4267,
+      "step": 241
+    },
+    {
+      "epoch": 0.4600760456273764,
+      "grad_norm": 0.692939472265219,
+      "learning_rate": 1.8997301672613496e-05,
+      "loss": 0.4685,
+      "step": 242
+    },
+    {
+      "epoch": 0.4619771863117871,
+      "grad_norm": 0.7318908050941767,
+      "learning_rate": 1.8982758032942184e-05,
+      "loss": 0.4624,
+      "step": 243
+    },
+    {
+      "epoch": 0.46387832699619774,
+      "grad_norm": 0.7279742085412728,
+      "learning_rate": 1.896811532678113e-05,
+      "loss": 0.4446,
+      "step": 244
+    },
+    {
+      "epoch": 0.46577946768060835,
+      "grad_norm": 0.7165540378971268,
+      "learning_rate": 1.8953373715617646e-05,
+      "loss": 0.4439,
+      "step": 245
+    },
+    {
+      "epoch": 0.467680608365019,
+      "grad_norm": 0.6716858449552553,
+      "learning_rate": 1.893853336202983e-05,
+      "loss": 0.436,
+      "step": 246
+    },
+    {
+      "epoch": 0.4695817490494297,
+      "grad_norm": 0.7203748002910793,
+      "learning_rate": 1.892359442968475e-05,
+      "loss": 0.4396,
+      "step": 247
+    },
+    {
+      "epoch": 0.4714828897338403,
+      "grad_norm": 0.6714620924477679,
+      "learning_rate": 1.8908557083336668e-05,
+      "loss": 0.4343,
+      "step": 248
+    },
+    {
+      "epoch": 0.47338403041825095,
+      "grad_norm": 0.8475320559878834,
+      "learning_rate": 1.889342148882519e-05,
+      "loss": 0.4549,
+      "step": 249
+    },
+    {
+      "epoch": 0.4752851711026616,
+      "grad_norm": 0.7855293495109291,
+      "learning_rate": 1.8878187813073465e-05,
+      "loss": 0.4598,
+      "step": 250
+    },
+    {
+      "epoch": 0.47718631178707227,
+      "grad_norm": 0.7178290775649008,
+      "learning_rate": 1.886285622408633e-05,
+      "loss": 0.4186,
+      "step": 251
+    },
+    {
+      "epoch": 0.4790874524714829,
+      "grad_norm": 0.767461158514695,
+      "learning_rate": 1.8847426890948447e-05,
+      "loss": 0.4466,
+      "step": 252
+    },
+    {
+      "epoch": 0.48098859315589354,
+      "grad_norm": 0.7320191128235685,
+      "learning_rate": 1.8831899983822475e-05,
+      "loss": 0.45,
+      "step": 253
+    },
+    {
+      "epoch": 0.4828897338403042,
+      "grad_norm": 0.7035050611718557,
+      "learning_rate": 1.8816275673947148e-05,
+      "loss": 0.4277,
+      "step": 254
+    },
+    {
+      "epoch": 0.4847908745247148,
+      "grad_norm": 0.7312388241616493,
+      "learning_rate": 1.8800554133635417e-05,
+      "loss": 0.4429,
+      "step": 255
+    },
+    {
+      "epoch": 0.4866920152091255,
+      "grad_norm": 0.7239815441570333,
+      "learning_rate": 1.8784735536272543e-05,
+      "loss": 0.4311,
+      "step": 256
+    },
+    {
+      "epoch": 0.48859315589353614,
+      "grad_norm": 0.699901234173822,
+      "learning_rate": 1.8768820056314173e-05,
+      "loss": 0.4173,
+      "step": 257
+    },
+    {
+      "epoch": 0.49049429657794674,
+      "grad_norm": 0.7669094341999264,
+      "learning_rate": 1.875280786928444e-05,
+      "loss": 0.4499,
+      "step": 258
+    },
+    {
+      "epoch": 0.4923954372623574,
+      "grad_norm": 0.7192014356850595,
+      "learning_rate": 1.873669915177399e-05,
+      "loss": 0.458,
+      "step": 259
+    },
+    {
+      "epoch": 0.49429657794676807,
+      "grad_norm": 0.6771316775438889,
+      "learning_rate": 1.872049408143808e-05,
+      "loss": 0.4173,
+      "step": 260
+    },
+    {
+      "epoch": 0.49619771863117873,
+      "grad_norm": 0.6901831210458916,
+      "learning_rate": 1.8704192836994578e-05,
+      "loss": 0.4187,
+      "step": 261
+    },
+    {
+      "epoch": 0.49809885931558934,
+      "grad_norm": 0.7417602272853672,
+      "learning_rate": 1.8687795598222024e-05,
+      "loss": 0.4343,
+      "step": 262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7154822710078802,
+      "learning_rate": 1.8671302545957628e-05,
+      "loss": 0.4371,
+      "step": 263
+    },
+    {
+      "epoch": 0.5019011406844106,
+      "grad_norm": 0.6685426183139277,
+      "learning_rate": 1.8654713862095272e-05,
+      "loss": 0.4356,
+      "step": 264
+    },
+    {
+      "epoch": 0.5038022813688213,
+      "grad_norm": 0.7121921303726694,
+      "learning_rate": 1.8638029729583524e-05,
+      "loss": 0.447,
+      "step": 265
+    },
+    {
+      "epoch": 0.5057034220532319,
+      "grad_norm": 0.7743635879030388,
+      "learning_rate": 1.8621250332423603e-05,
+      "loss": 0.4422,
+      "step": 266
+    },
+    {
+      "epoch": 0.5076045627376425,
+      "grad_norm": 0.7270303589496028,
+      "learning_rate": 1.860437585566736e-05,
+      "loss": 0.4366,
+      "step": 267
+    },
+    {
+      "epoch": 0.5095057034220533,
+      "grad_norm": 0.7137799507431448,
+      "learning_rate": 1.8587406485415226e-05,
+      "loss": 0.4431,
+      "step": 268
+    },
+    {
+      "epoch": 0.5114068441064639,
+      "grad_norm": 0.7492738756754567,
+      "learning_rate": 1.8570342408814173e-05,
+      "loss": 0.4299,
+      "step": 269
+    },
+    {
+      "epoch": 0.5133079847908745,
+      "grad_norm": 0.7285958496485917,
+      "learning_rate": 1.855318381405564e-05,
+      "loss": 0.4596,
+      "step": 270
+    },
+    {
+      "epoch": 0.5152091254752852,
+      "grad_norm": 0.7559474300743639,
+      "learning_rate": 1.8535930890373467e-05,
+      "loss": 0.4443,
+      "step": 271
+    },
+    {
+      "epoch": 0.5171102661596958,
+      "grad_norm": 0.7342284641865824,
+      "learning_rate": 1.8518583828041787e-05,
+      "loss": 0.4529,
+      "step": 272
+    },
+    {
+      "epoch": 0.5190114068441065,
+      "grad_norm": 0.8238892216238191,
+      "learning_rate": 1.8501142818372964e-05,
+      "loss": 0.4446,
+      "step": 273
+    },
+    {
+      "epoch": 0.5209125475285171,
+      "grad_norm": 0.7084039539882753,
+      "learning_rate": 1.848360805371544e-05,
+      "loss": 0.4489,
+      "step": 274
+    },
+    {
+      "epoch": 0.5228136882129277,
+      "grad_norm": 0.6720936859437973,
+      "learning_rate": 1.8465979727451653e-05,
+      "loss": 0.4165,
+      "step": 275
+    },
+    {
+      "epoch": 0.5247148288973384,
+      "grad_norm": 0.8312100781325986,
+      "learning_rate": 1.8448258033995877e-05,
+      "loss": 0.4427,
+      "step": 276
+    },
+    {
+      "epoch": 0.526615969581749,
+      "grad_norm": 0.745236533446814,
+      "learning_rate": 1.8430443168792087e-05,
+      "loss": 0.4226,
+      "step": 277
+    },
+    {
+      "epoch": 0.5285171102661597,
+      "grad_norm": 0.726994655365599,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.4482,
+      "step": 278
+    },
+    {
+      "epoch": 0.5304182509505704,
+      "grad_norm": 0.8400480988033476,
+      "learning_rate": 1.8394534710051956e-05,
+      "loss": 0.4466,
+      "step": 279
+    },
+    {
+      "epoch": 0.532319391634981,
+      "grad_norm": 0.7999779446125437,
+      "learning_rate": 1.8376441512532617e-05,
+      "loss": 0.4638,
+      "step": 280
+    },
+    {
+      "epoch": 0.5342205323193916,
+      "grad_norm": 0.7524310541334193,
+      "learning_rate": 1.835825593529492e-05,
+      "loss": 0.415,
+      "step": 281
+    },
+    {
+      "epoch": 0.5361216730038023,
+      "grad_norm": 0.7189470659927434,
+      "learning_rate": 1.833997817889878e-05,
+      "loss": 0.4204,
+      "step": 282
+    },
+    {
+      "epoch": 0.5380228136882129,
+      "grad_norm": 0.7713948153452118,
+      "learning_rate": 1.8321608444920738e-05,
+      "loss": 0.4052,
+      "step": 283
+    },
+    {
+      "epoch": 0.5399239543726235,
+      "grad_norm": 0.7535258416024757,
+      "learning_rate": 1.830314693595169e-05,
+      "loss": 0.4339,
+      "step": 284
+    },
+    {
+      "epoch": 0.5418250950570342,
+      "grad_norm": 0.8185578895013603,
+      "learning_rate": 1.828459385559468e-05,
+      "loss": 0.4184,
+      "step": 285
+    },
+    {
+      "epoch": 0.5437262357414449,
+      "grad_norm": 0.6896944341799538,
+      "learning_rate": 1.8265949408462657e-05,
+      "loss": 0.416,
+      "step": 286
+    },
+    {
+      "epoch": 0.5456273764258555,
+      "grad_norm": 0.7361573015735334,
+      "learning_rate": 1.8247213800176192e-05,
+      "loss": 0.4415,
+      "step": 287
+    },
+    {
+      "epoch": 0.5475285171102662,
+      "grad_norm": 0.7615623124912052,
+      "learning_rate": 1.8228387237361245e-05,
+      "loss": 0.4249,
+      "step": 288
+    },
+    {
+      "epoch": 0.5494296577946768,
+      "grad_norm": 0.7756588385029941,
+      "learning_rate": 1.8209469927646863e-05,
+      "loss": 0.4271,
+      "step": 289
+    },
+    {
+      "epoch": 0.5513307984790875,
+      "grad_norm": 0.6712848987270621,
+      "learning_rate": 1.8190462079662897e-05,
+      "loss": 0.4385,
+      "step": 290
+    },
+    {
+      "epoch": 0.5532319391634981,
+      "grad_norm": 0.7242430780764021,
+      "learning_rate": 1.81713639030377e-05,
+      "loss": 0.4359,
+      "step": 291
+    },
+    {
+      "epoch": 0.5551330798479087,
+      "grad_norm": 0.6876488182792876,
+      "learning_rate": 1.8152175608395814e-05,
+      "loss": 0.4193,
+      "step": 292
+    },
+    {
+      "epoch": 0.5570342205323194,
+      "grad_norm": 0.7642072303542361,
+      "learning_rate": 1.8132897407355657e-05,
+      "loss": 0.4254,
+      "step": 293
+    },
+    {
+      "epoch": 0.55893536121673,
+      "grad_norm": 0.6448148715357482,
+      "learning_rate": 1.811352951252717e-05,
+      "loss": 0.414,
+      "step": 294
+    },
+    {
+      "epoch": 0.5608365019011406,
+      "grad_norm": 0.7951094665270376,
+      "learning_rate": 1.809407213750949e-05,
+      "loss": 0.4325,
+      "step": 295
+    },
+    {
+      "epoch": 0.5627376425855514,
+      "grad_norm": 0.7127166433691607,
+      "learning_rate": 1.807452549688859e-05,
+      "loss": 0.4567,
+      "step": 296
+    },
+    {
+      "epoch": 0.564638783269962,
+      "grad_norm": 0.7629834827406596,
+      "learning_rate": 1.8054889806234906e-05,
+      "loss": 0.4589,
+      "step": 297
+    },
+    {
+      "epoch": 0.5665399239543726,
+      "grad_norm": 0.7582660806340702,
+      "learning_rate": 1.8035165282100963e-05,
+      "loss": 0.4277,
+      "step": 298
+    },
+    {
+      "epoch": 0.5684410646387833,
+      "grad_norm": 0.6901232924019108,
+      "learning_rate": 1.8015352142018984e-05,
+      "loss": 0.4207,
+      "step": 299
+    },
+    {
+      "epoch": 0.5703422053231939,
+      "grad_norm": 0.6676283890349991,
+      "learning_rate": 1.799545060449851e-05,
+      "loss": 0.4168,
+      "step": 300
+    },
+    {
+      "epoch": 0.5722433460076045,
+      "grad_norm": 0.6772322849836512,
+      "learning_rate": 1.797546088902396e-05,
+      "loss": 0.4182,
+      "step": 301
+    },
+    {
+      "epoch": 0.5741444866920152,
+      "grad_norm": 0.6965324803967498,
+      "learning_rate": 1.7955383216052224e-05,
+      "loss": 0.4327,
+      "step": 302
+    },
+    {
+      "epoch": 0.5760456273764258,
+      "grad_norm": 0.7596614386247041,
+      "learning_rate": 1.7935217807010238e-05,
+      "loss": 0.4303,
+      "step": 303
+    },
+    {
+      "epoch": 0.5779467680608364,
+      "grad_norm": 0.7522404373680283,
+      "learning_rate": 1.7914964884292543e-05,
+      "loss": 0.4373,
+      "step": 304
+    },
+    {
+      "epoch": 0.5798479087452472,
+      "grad_norm": 0.6699913144266135,
+      "learning_rate": 1.7894624671258813e-05,
+      "loss": 0.4261,
+      "step": 305
+    },
+    {
+      "epoch": 0.5817490494296578,
+      "grad_norm": 0.6544511658991378,
+      "learning_rate": 1.7874197392231414e-05,
+      "loss": 0.4236,
+      "step": 306
+    },
+    {
+      "epoch": 0.5836501901140685,
+      "grad_norm": 0.7228660901869968,
+      "learning_rate": 1.7853683272492913e-05,
+      "loss": 0.4448,
+      "step": 307
+    },
+    {
+      "epoch": 0.5855513307984791,
+      "grad_norm": 0.6520331070986572,
+      "learning_rate": 1.7833082538283615e-05,
+      "loss": 0.4158,
+      "step": 308
+    },
+    {
+      "epoch": 0.5874524714828897,
+      "grad_norm": 0.6519008846920139,
+      "learning_rate": 1.7812395416799034e-05,
+      "loss": 0.416,
+      "step": 309
+    },
+    {
+      "epoch": 0.5893536121673004,
+      "grad_norm": 0.7035814828278057,
+      "learning_rate": 1.7791622136187422e-05,
+      "loss": 0.4422,
+      "step": 310
+    },
+    {
+      "epoch": 0.591254752851711,
+      "grad_norm": 0.7326060953758017,
+      "learning_rate": 1.7770762925547235e-05,
+      "loss": 0.4466,
+      "step": 311
+    },
+    {
+      "epoch": 0.5931558935361216,
+      "grad_norm": 0.7032292071399574,
+      "learning_rate": 1.7749818014924612e-05,
+      "loss": 0.4285,
+      "step": 312
+    },
+    {
+      "epoch": 0.5950570342205324,
+      "grad_norm": 0.6814651660596601,
+      "learning_rate": 1.7728787635310828e-05,
+      "loss": 0.41,
+      "step": 313
+    },
+    {
+      "epoch": 0.596958174904943,
+      "grad_norm": 0.6669958828262627,
+      "learning_rate": 1.770767201863976e-05,
+      "loss": 0.4358,
+      "step": 314
+    },
+    {
+      "epoch": 0.5988593155893536,
+      "grad_norm": 0.681100492018271,
+      "learning_rate": 1.7686471397785322e-05,
+      "loss": 0.4189,
+      "step": 315
+    },
+    {
+      "epoch": 0.6007604562737643,
+      "grad_norm": 0.6996522951115274,
+      "learning_rate": 1.76651860065589e-05,
+      "loss": 0.4179,
+      "step": 316
+    },
+    {
+      "epoch": 0.6026615969581749,
+      "grad_norm": 0.7705590530169981,
+      "learning_rate": 1.764381607970677e-05,
+      "loss": 0.4309,
+      "step": 317
+    },
+    {
+      "epoch": 0.6045627376425855,
+      "grad_norm": 0.6464937285827252,
+      "learning_rate": 1.7622361852907506e-05,
+      "loss": 0.4166,
+      "step": 318
+    },
+    {
+      "epoch": 0.6064638783269962,
+      "grad_norm": 0.6668355632837625,
+      "learning_rate": 1.760082356276939e-05,
+      "loss": 0.4261,
+      "step": 319
+    },
+    {
+      "epoch": 0.6083650190114068,
+      "grad_norm": 0.7313408891252491,
+      "learning_rate": 1.75792014468278e-05,
+      "loss": 0.4319,
+      "step": 320
+    },
+    {
+      "epoch": 0.6102661596958175,
+      "grad_norm": 0.675982323864499,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.448,
+      "step": 321
+    },
+    {
+      "epoch": 0.6121673003802282,
+      "grad_norm": 0.7296342683554718,
+      "learning_rate": 1.7535706692295436e-05,
+      "loss": 0.4494,
+      "step": 322
+    },
+    {
+      "epoch": 0.6140684410646388,
+      "grad_norm": 0.6869721599235838,
+      "learning_rate": 1.7513834533387256e-05,
+      "loss": 0.4324,
+      "step": 323
+    },
+    {
+      "epoch": 0.6159695817490495,
+      "grad_norm": 0.6868823819524486,
+      "learning_rate": 1.7491879508035488e-05,
+      "loss": 0.4278,
+      "step": 324
+    },
+    {
+      "epoch": 0.6178707224334601,
+      "grad_norm": 0.6353307067663678,
+      "learning_rate": 1.746984185837149e-05,
+      "loss": 0.4039,
+      "step": 325
+    },
+    {
+      "epoch": 0.6197718631178707,
+      "grad_norm": 0.7113141686158224,
+      "learning_rate": 1.744772182743782e-05,
+      "loss": 0.4191,
+      "step": 326
+    },
+    {
+      "epoch": 0.6216730038022814,
+      "grad_norm": 0.6094163092370466,
+      "learning_rate": 1.7425519659185596e-05,
+      "loss": 0.4222,
+      "step": 327
+    },
+    {
+      "epoch": 0.623574144486692,
+      "grad_norm": 0.6505190038388743,
+      "learning_rate": 1.740323559847179e-05,
+      "loss": 0.4283,
+      "step": 328
+    },
+    {
+      "epoch": 0.6254752851711026,
+      "grad_norm": 0.6522405887547765,
+      "learning_rate": 1.738086989105651e-05,
+      "loss": 0.4088,
+      "step": 329
+    },
+    {
+      "epoch": 0.6273764258555133,
+      "grad_norm": 0.6486038487519525,
+      "learning_rate": 1.735842278360032e-05,
+      "loss": 0.4516,
+      "step": 330
+    },
+    {
+      "epoch": 0.629277566539924,
+      "grad_norm": 0.6359155163423266,
+      "learning_rate": 1.73358945236615e-05,
+      "loss": 0.4168,
+      "step": 331
+    },
+    {
+      "epoch": 0.6311787072243346,
+      "grad_norm": 0.6614807750716021,
+      "learning_rate": 1.7313285359693322e-05,
+      "loss": 0.4503,
+      "step": 332
+    },
+    {
+      "epoch": 0.6330798479087453,
+      "grad_norm": 0.6297516559931785,
+      "learning_rate": 1.7290595541041312e-05,
+      "loss": 0.4172,
+      "step": 333
+    },
+    {
+      "epoch": 0.6349809885931559,
+      "grad_norm": 0.6425729968704109,
+      "learning_rate": 1.7267825317940494e-05,
+      "loss": 0.4432,
+      "step": 334
+    },
+    {
+      "epoch": 0.6368821292775665,
+      "grad_norm": 0.6605166662329315,
+      "learning_rate": 1.724497494151264e-05,
+      "loss": 0.4155,
+      "step": 335
+    },
+    {
+      "epoch": 0.6387832699619772,
+      "grad_norm": 0.7193898838532637,
+      "learning_rate": 1.7222044663763484e-05,
+      "loss": 0.4224,
+      "step": 336
+    },
+    {
+      "epoch": 0.6406844106463878,
+      "grad_norm": 0.650976985325332,
+      "learning_rate": 1.7199034737579962e-05,
+      "loss": 0.4254,
+      "step": 337
+    },
+    {
+      "epoch": 0.6425855513307985,
+      "grad_norm": 0.8083689583400872,
+      "learning_rate": 1.7175945416727405e-05,
+      "loss": 0.4319,
+      "step": 338
+    },
+    {
+      "epoch": 0.6444866920152091,
+      "grad_norm": 0.6587482265073595,
+      "learning_rate": 1.7152776955846768e-05,
+      "loss": 0.4119,
+      "step": 339
+    },
+    {
+      "epoch": 0.6463878326996197,
+      "grad_norm": 0.6953249191063823,
+      "learning_rate": 1.7129529610451775e-05,
+      "loss": 0.4023,
+      "step": 340
+    },
+    {
+      "epoch": 0.6482889733840305,
+      "grad_norm": 0.7920752883874513,
+      "learning_rate": 1.7106203636926154e-05,
+      "loss": 0.4316,
+      "step": 341
+    },
+    {
+      "epoch": 0.6501901140684411,
+      "grad_norm": 0.6525508665848344,
+      "learning_rate": 1.7082799292520767e-05,
+      "loss": 0.4283,
+      "step": 342
+    },
+    {
+      "epoch": 0.6520912547528517,
+      "grad_norm": 0.8254132992624188,
+      "learning_rate": 1.7059316835350806e-05,
+      "loss": 0.4354,
+      "step": 343
+    },
+    {
+      "epoch": 0.6539923954372624,
+      "grad_norm": 0.8638491320788857,
+      "learning_rate": 1.7035756524392924e-05,
+      "loss": 0.4175,
+      "step": 344
+    },
+    {
+      "epoch": 0.655893536121673,
+      "grad_norm": 0.7507682043869717,
+      "learning_rate": 1.7012118619482376e-05,
+      "loss": 0.4397,
+      "step": 345
+    },
+    {
+      "epoch": 0.6577946768060836,
+      "grad_norm": 0.8693183793898529,
+      "learning_rate": 1.6988403381310177e-05,
+      "loss": 0.4488,
+      "step": 346
+    },
+    {
+      "epoch": 0.6596958174904943,
+      "grad_norm": 0.6663308271286326,
+      "learning_rate": 1.696461107142021e-05,
+      "loss": 0.4219,
+      "step": 347
+    },
+    {
+      "epoch": 0.6615969581749049,
+      "grad_norm": 0.7942157641443345,
+      "learning_rate": 1.6940741952206342e-05,
+      "loss": 0.4102,
+      "step": 348
+    },
+    {
+      "epoch": 0.6634980988593155,
+      "grad_norm": 0.8521552726036399,
+      "learning_rate": 1.691679628690953e-05,
+      "loss": 0.4425,
+      "step": 349
+    },
+    {
+      "epoch": 0.6653992395437263,
+      "grad_norm": 0.686632797951827,
+      "learning_rate": 1.6892774339614927e-05,
+      "loss": 0.4053,
+      "step": 350
+    },
+    {
+      "epoch": 0.6673003802281369,
+      "grad_norm": 0.8656906649258893,
+      "learning_rate": 1.686867637524896e-05,
+      "loss": 0.4394,
+      "step": 351
+    },
+    {
+      "epoch": 0.6692015209125475,
+      "grad_norm": 0.6627440644143102,
+      "learning_rate": 1.6844502659576414e-05,
+      "loss": 0.4249,
+      "step": 352
+    },
+    {
+      "epoch": 0.6711026615969582,
+      "grad_norm": 0.8638496392734872,
+      "learning_rate": 1.6820253459197493e-05,
+      "loss": 0.4527,
+      "step": 353
+    },
+    {
+      "epoch": 0.6730038022813688,
+      "grad_norm": 0.6811664559844841,
+      "learning_rate": 1.679592904154489e-05,
+      "loss": 0.4301,
+      "step": 354
+    },
+    {
+      "epoch": 0.6749049429657795,
+      "grad_norm": 0.7139175268110214,
+      "learning_rate": 1.677152967488084e-05,
+      "loss": 0.4315,
+      "step": 355
+    },
+    {
+      "epoch": 0.6768060836501901,
+      "grad_norm": 0.7715328047425238,
+      "learning_rate": 1.6747055628294134e-05,
+      "loss": 0.4357,
+      "step": 356
+    },
+    {
+      "epoch": 0.6787072243346007,
+      "grad_norm": 0.6824438804585463,
+      "learning_rate": 1.6722507171697184e-05,
+      "loss": 0.4487,
+      "step": 357
+    },
+    {
+      "epoch": 0.6806083650190115,
+      "grad_norm": 0.6489516068004965,
+      "learning_rate": 1.669788457582304e-05,
+      "loss": 0.439,
+      "step": 358
+    },
+    {
+      "epoch": 0.6825095057034221,
+      "grad_norm": 0.6775986545135927,
+      "learning_rate": 1.6673188112222394e-05,
+      "loss": 0.4143,
+      "step": 359
+    },
+    {
+      "epoch": 0.6844106463878327,
+      "grad_norm": 0.6290923010946774,
+      "learning_rate": 1.6648418053260585e-05,
+      "loss": 0.4307,
+      "step": 360
+    },
+    {
+      "epoch": 0.6863117870722434,
+      "grad_norm": 0.6876369011417524,
+      "learning_rate": 1.6623574672114596e-05,
+      "loss": 0.4457,
+      "step": 361
+    },
+    {
+      "epoch": 0.688212927756654,
+      "grad_norm": 0.687605864585501,
+      "learning_rate": 1.6598658242770054e-05,
+      "loss": 0.4419,
+      "step": 362
+    },
+    {
+      "epoch": 0.6901140684410646,
+      "grad_norm": 0.6404847706150755,
+      "learning_rate": 1.6573669040018202e-05,
+      "loss": 0.4118,
+      "step": 363
+    },
+    {
+      "epoch": 0.6920152091254753,
+      "grad_norm": 0.6488769938764549,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.3995,
+      "step": 364
+    },
+    {
+      "epoch": 0.6939163498098859,
+      "grad_norm": 0.6677484348346899,
+      "learning_rate": 1.652347341746737e-05,
+      "loss": 0.425,
+      "step": 365
+    },
+    {
+      "epoch": 0.6958174904942965,
+      "grad_norm": 0.6705187136522287,
+      "learning_rate": 1.6498267551251618e-05,
+      "loss": 0.4152,
+      "step": 366
+    },
+    {
+      "epoch": 0.6977186311787072,
+      "grad_norm": 0.6313280471294431,
+      "learning_rate": 1.6472990018788884e-05,
+      "loss": 0.4147,
+      "step": 367
+    },
+    {
+      "epoch": 0.6996197718631179,
+      "grad_norm": 0.6181349369314366,
+      "learning_rate": 1.644764109885284e-05,
+      "loss": 0.4315,
+      "step": 368
+    },
+    {
+      "epoch": 0.7015209125475285,
+      "grad_norm": 0.6571542429219354,
+      "learning_rate": 1.642222107100446e-05,
+      "loss": 0.4302,
+      "step": 369
+    },
+    {
+      "epoch": 0.7034220532319392,
+      "grad_norm": 0.6263741047023472,
+      "learning_rate": 1.6396730215588913e-05,
+      "loss": 0.4289,
+      "step": 370
+    },
+    {
+      "epoch": 0.7053231939163498,
+      "grad_norm": 0.7106834249285274,
+      "learning_rate": 1.6371168813732514e-05,
+      "loss": 0.418,
+      "step": 371
+    },
+    {
+      "epoch": 0.7072243346007605,
+      "grad_norm": 0.6661713244874843,
+      "learning_rate": 1.6345537147339578e-05,
+      "loss": 0.4253,
+      "step": 372
+    },
+    {
+      "epoch": 0.7091254752851711,
+      "grad_norm": 0.819597776040357,
+      "learning_rate": 1.6319835499089358e-05,
+      "loss": 0.4158,
+      "step": 373
+    },
+    {
+      "epoch": 0.7110266159695817,
+      "grad_norm": 0.6572672368841243,
+      "learning_rate": 1.6294064152432878e-05,
+      "loss": 0.3997,
+      "step": 374
+    },
+    {
+      "epoch": 0.7129277566539924,
+      "grad_norm": 0.6782559603916488,
+      "learning_rate": 1.626822339158985e-05,
+      "loss": 0.4167,
+      "step": 375
+    },
+    {
+      "epoch": 0.714828897338403,
+      "grad_norm": 0.6504077037280834,
+      "learning_rate": 1.6242313501545522e-05,
+      "loss": 0.4234,
+      "step": 376
+    },
+    {
+      "epoch": 0.7167300380228137,
+      "grad_norm": 0.6946779470999478,
+      "learning_rate": 1.621633476804752e-05,
+      "loss": 0.4327,
+      "step": 377
+    },
+    {
+      "epoch": 0.7186311787072244,
+      "grad_norm": 0.7060272669401575,
+      "learning_rate": 1.6190287477602716e-05,
+      "loss": 0.4368,
+      "step": 378
+    },
+    {
+      "epoch": 0.720532319391635,
+      "grad_norm": 0.6707925114410718,
+      "learning_rate": 1.6164171917474078e-05,
+      "loss": 0.4156,
+      "step": 379
+    },
+    {
+      "epoch": 0.7224334600760456,
+      "grad_norm": 0.693429892761807,
+      "learning_rate": 1.6137988375677466e-05,
+      "loss": 0.4633,
+      "step": 380
+    },
+    {
+      "epoch": 0.7243346007604563,
+      "grad_norm": 0.6778793891583361,
+      "learning_rate": 1.6111737140978495e-05,
+      "loss": 0.4239,
+      "step": 381
+    },
+    {
+      "epoch": 0.7262357414448669,
+      "grad_norm": 0.6736096122364035,
+      "learning_rate": 1.6085418502889315e-05,
+      "loss": 0.4337,
+      "step": 382
+    },
+    {
+      "epoch": 0.7281368821292775,
+      "grad_norm": 0.7594804800572929,
+      "learning_rate": 1.6059032751665454e-05,
+      "loss": 0.4529,
+      "step": 383
+    },
+    {
+      "epoch": 0.7300380228136882,
+      "grad_norm": 0.6544869302739386,
+      "learning_rate": 1.6032580178302585e-05,
+      "loss": 0.4518,
+      "step": 384
+    },
+    {
+      "epoch": 0.7319391634980988,
+      "grad_norm": 0.6998229136975362,
+      "learning_rate": 1.600606107453333e-05,
+      "loss": 0.4301,
+      "step": 385
+    },
+    {
+      "epoch": 0.7338403041825095,
+      "grad_norm": 0.7516862791476883,
+      "learning_rate": 1.597947573282405e-05,
+      "loss": 0.4223,
+      "step": 386
+    },
+    {
+      "epoch": 0.7357414448669202,
+      "grad_norm": 0.6980933774045253,
+      "learning_rate": 1.5952824446371608e-05,
+      "loss": 0.4596,
+      "step": 387
+    },
+    {
+      "epoch": 0.7376425855513308,
+      "grad_norm": 0.7430629073207451,
+      "learning_rate": 1.592610750910014e-05,
+      "loss": 0.416,
+      "step": 388
+    },
+    {
+      "epoch": 0.7395437262357415,
+      "grad_norm": 0.6639731632039884,
+      "learning_rate": 1.589932521565781e-05,
+      "loss": 0.419,
+      "step": 389
+    },
+    {
+      "epoch": 0.7414448669201521,
+      "grad_norm": 0.7259575942341968,
+      "learning_rate": 1.587247786141358e-05,
+      "loss": 0.4106,
+      "step": 390
+    },
+    {
+      "epoch": 0.7433460076045627,
+      "grad_norm": 0.7002284497582576,
+      "learning_rate": 1.5845565742453906e-05,
+      "loss": 0.4275,
+      "step": 391
+    },
+    {
+      "epoch": 0.7452471482889734,
+      "grad_norm": 0.6584481566520094,
+      "learning_rate": 1.581858915557953e-05,
+      "loss": 0.4351,
+      "step": 392
+    },
+    {
+      "epoch": 0.747148288973384,
+      "grad_norm": 0.7464663839488714,
+      "learning_rate": 1.5791548398302167e-05,
+      "loss": 0.4252,
+      "step": 393
+    },
+    {
+      "epoch": 0.7490494296577946,
+      "grad_norm": 0.6262405138106963,
+      "learning_rate": 1.5764443768841234e-05,
+      "loss": 0.394,
+      "step": 394
+    },
+    {
+      "epoch": 0.7509505703422054,
+      "grad_norm": 0.6780695271154578,
+      "learning_rate": 1.5737275566120577e-05,
+      "loss": 0.4244,
+      "step": 395
+    },
+    {
+      "epoch": 0.752851711026616,
+      "grad_norm": 0.6802247489906732,
+      "learning_rate": 1.5710044089765144e-05,
+      "loss": 0.4312,
+      "step": 396
+    },
+    {
+      "epoch": 0.7547528517110266,
+      "grad_norm": 0.7263152642353367,
+      "learning_rate": 1.5682749640097708e-05,
+      "loss": 0.4236,
+      "step": 397
+    },
+    {
+      "epoch": 0.7566539923954373,
+      "grad_norm": 0.7472986459494889,
+      "learning_rate": 1.565539251813554e-05,
+      "loss": 0.401,
+      "step": 398
+    },
+    {
+      "epoch": 0.7585551330798479,
+      "grad_norm": 0.6933783601104639,
+      "learning_rate": 1.5627973025587093e-05,
+      "loss": 0.4412,
+      "step": 399
+    },
+    {
+      "epoch": 0.7604562737642585,
+      "grad_norm": 0.7800601746384036,
+      "learning_rate": 1.560049146484868e-05,
+      "loss": 0.4238,
+      "step": 400
+    },
+    {
+      "epoch": 0.7623574144486692,
+      "grad_norm": 0.7415369649002813,
+      "learning_rate": 1.5572948139001128e-05,
+      "loss": 0.4338,
+      "step": 401
+    },
+    {
+      "epoch": 0.7642585551330798,
+      "grad_norm": 0.6272433071020401,
+      "learning_rate": 1.5545343351806443e-05,
+      "loss": 0.4005,
+      "step": 402
+    },
+    {
+      "epoch": 0.7661596958174905,
+      "grad_norm": 0.7236439687728006,
+      "learning_rate": 1.551767740770446e-05,
+      "loss": 0.4036,
+      "step": 403
+    },
+    {
+      "epoch": 0.7680608365019012,
+      "grad_norm": 0.7872892947079139,
+      "learning_rate": 1.5489950611809484e-05,
+      "loss": 0.4194,
+      "step": 404
+    },
+    {
+      "epoch": 0.7699619771863118,
+      "grad_norm": 0.6394055914356919,
+      "learning_rate": 1.5462163269906928e-05,
+      "loss": 0.4276,
+      "step": 405
+    },
+    {
+      "epoch": 0.7718631178707225,
+      "grad_norm": 0.8473517991183547,
+      "learning_rate": 1.5434315688449924e-05,
+      "loss": 0.4555,
+      "step": 406
+    },
+    {
+      "epoch": 0.7737642585551331,
+      "grad_norm": 0.6958515694063482,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.3891,
+      "step": 407
+    },
+    {
+      "epoch": 0.7756653992395437,
+      "grad_norm": 0.709057524722897,
+      "learning_rate": 1.5378441036003543e-05,
+      "loss": 0.4343,
+      "step": 408
+    },
+    {
+      "epoch": 0.7775665399239544,
+      "grad_norm": 0.8651037064338964,
+      "learning_rate": 1.535041458122865e-05,
+      "loss": 0.4083,
+      "step": 409
+    },
+    {
+      "epoch": 0.779467680608365,
+      "grad_norm": 0.6442972383133835,
+      "learning_rate": 1.5322329119321508e-05,
+      "loss": 0.4126,
+      "step": 410
+    },
+    {
+      "epoch": 0.7813688212927756,
+      "grad_norm": 0.8894329484157454,
+      "learning_rate": 1.529418496002308e-05,
+      "loss": 0.4207,
+      "step": 411
+    },
+    {
+      "epoch": 0.7832699619771863,
+      "grad_norm": 0.7429228133083364,
+      "learning_rate": 1.5265982413721662e-05,
+      "loss": 0.3991,
+      "step": 412
+    },
+    {
+      "epoch": 0.785171102661597,
+      "grad_norm": 0.6524047667214418,
+      "learning_rate": 1.5237721791449497e-05,
+      "loss": 0.4169,
+      "step": 413
+    },
+    {
+      "epoch": 0.7870722433460076,
+      "grad_norm": 0.7985574837191214,
+      "learning_rate": 1.5209403404879305e-05,
+      "loss": 0.4372,
+      "step": 414
+    },
+    {
+      "epoch": 0.7889733840304183,
+      "grad_norm": 0.6884078521763929,
+      "learning_rate": 1.5181027566320858e-05,
+      "loss": 0.4151,
+      "step": 415
+    },
+    {
+      "epoch": 0.7908745247148289,
+      "grad_norm": 0.6705502012815527,
+      "learning_rate": 1.5152594588717544e-05,
+      "loss": 0.4282,
+      "step": 416
+    },
+    {
+      "epoch": 0.7927756653992395,
+      "grad_norm": 0.8292952607045119,
+      "learning_rate": 1.5124104785642909e-05,
+      "loss": 0.4247,
+      "step": 417
+    },
+    {
+      "epoch": 0.7946768060836502,
+      "grad_norm": 0.6592448406740135,
+      "learning_rate": 1.5095558471297196e-05,
+      "loss": 0.4121,
+      "step": 418
+    },
+    {
+      "epoch": 0.7965779467680608,
+      "grad_norm": 0.6732939577723265,
+      "learning_rate": 1.5066955960503893e-05,
+      "loss": 0.4245,
+      "step": 419
+    },
+    {
+      "epoch": 0.7984790874524715,
+      "grad_norm": 0.8059232301749493,
+      "learning_rate": 1.5038297568706244e-05,
+      "loss": 0.4145,
+      "step": 420
+    },
+    {
+      "epoch": 0.8003802281368821,
+      "grad_norm": 0.6724625879976294,
+      "learning_rate": 1.5009583611963772e-05,
+      "loss": 0.4276,
+      "step": 421
+    },
+    {
+      "epoch": 0.8022813688212928,
+      "grad_norm": 0.7261380807087485,
+      "learning_rate": 1.4980814406948806e-05,
+      "loss": 0.4306,
+      "step": 422
+    },
+    {
+      "epoch": 0.8041825095057035,
+      "grad_norm": 0.703013374065038,
+      "learning_rate": 1.4951990270942991e-05,
+      "loss": 0.4039,
+      "step": 423
+    },
+    {
+      "epoch": 0.8060836501901141,
+      "grad_norm": 0.6331660489159426,
+      "learning_rate": 1.492311152183376e-05,
+      "loss": 0.3917,
+      "step": 424
+    },
+    {
+      "epoch": 0.8079847908745247,
+      "grad_norm": 0.7045691008861624,
+      "learning_rate": 1.4894178478110856e-05,
+      "loss": 0.4177,
+      "step": 425
+    },
+    {
+      "epoch": 0.8098859315589354,
+      "grad_norm": 0.5900848245472016,
+      "learning_rate": 1.4865191458862816e-05,
+      "loss": 0.4054,
+      "step": 426
+    },
+    {
+      "epoch": 0.811787072243346,
+      "grad_norm": 0.6622459481214953,
+      "learning_rate": 1.4836150783773442e-05,
+      "loss": 0.4002,
+      "step": 427
+    },
+    {
+      "epoch": 0.8136882129277566,
+      "grad_norm": 0.6949932285653622,
+      "learning_rate": 1.4807056773118276e-05,
+      "loss": 0.4385,
+      "step": 428
+    },
+    {
+      "epoch": 0.8155893536121673,
+      "grad_norm": 0.6190135002927535,
+      "learning_rate": 1.4777909747761085e-05,
+      "loss": 0.3973,
+      "step": 429
+    },
+    {
+      "epoch": 0.8174904942965779,
+      "grad_norm": 0.6436685879046238,
+      "learning_rate": 1.4748710029150296e-05,
+      "loss": 0.4222,
+      "step": 430
+    },
+    {
+      "epoch": 0.8193916349809885,
+      "grad_norm": 0.6386163098470623,
+      "learning_rate": 1.4719457939315468e-05,
+      "loss": 0.4284,
+      "step": 431
+    },
+    {
+      "epoch": 0.8212927756653993,
+      "grad_norm": 0.666182603128127,
+      "learning_rate": 1.4690153800863743e-05,
+      "loss": 0.4034,
+      "step": 432
+    },
+    {
+      "epoch": 0.8231939163498099,
+      "grad_norm": 0.6379481428977116,
+      "learning_rate": 1.4660797936976278e-05,
+      "loss": 0.3988,
+      "step": 433
+    },
+    {
+      "epoch": 0.8250950570342205,
+      "grad_norm": 0.6717411668705876,
+      "learning_rate": 1.4631390671404682e-05,
+      "loss": 0.4115,
+      "step": 434
+    },
+    {
+      "epoch": 0.8269961977186312,
+      "grad_norm": 0.6750236301339044,
+      "learning_rate": 1.460193232846745e-05,
+      "loss": 0.401,
+      "step": 435
+    },
+    {
+      "epoch": 0.8288973384030418,
+      "grad_norm": 0.7153132483295134,
+      "learning_rate": 1.4572423233046386e-05,
+      "loss": 0.4324,
+      "step": 436
+    },
+    {
+      "epoch": 0.8307984790874525,
+      "grad_norm": 0.7156916834621986,
+      "learning_rate": 1.4542863710583022e-05,
+      "loss": 0.44,
+      "step": 437
+    },
+    {
+      "epoch": 0.8326996197718631,
+      "grad_norm": 0.6683381333785062,
+      "learning_rate": 1.4513254087075015e-05,
+      "loss": 0.4217,
+      "step": 438
+    },
+    {
+      "epoch": 0.8346007604562737,
+      "grad_norm": 0.6702765172041102,
+      "learning_rate": 1.4483594689072571e-05,
+      "loss": 0.4051,
+      "step": 439
+    },
+    {
+      "epoch": 0.8365019011406845,
+      "grad_norm": 0.731998620493878,
+      "learning_rate": 1.4453885843674837e-05,
+      "loss": 0.406,
+      "step": 440
+    },
+    {
+      "epoch": 0.8384030418250951,
+      "grad_norm": 0.6403867699782739,
+      "learning_rate": 1.4424127878526278e-05,
+      "loss": 0.4165,
+      "step": 441
+    },
+    {
+      "epoch": 0.8403041825095057,
+      "grad_norm": 0.6884554847402573,
+      "learning_rate": 1.4394321121813093e-05,
+      "loss": 0.4235,
+      "step": 442
+    },
+    {
+      "epoch": 0.8422053231939164,
+      "grad_norm": 0.6970583308435006,
+      "learning_rate": 1.436446590225957e-05,
+      "loss": 0.4269,
+      "step": 443
+    },
+    {
+      "epoch": 0.844106463878327,
+      "grad_norm": 0.5995650193286344,
+      "learning_rate": 1.433456254912447e-05,
+      "loss": 0.3997,
+      "step": 444
+    },
+    {
+      "epoch": 0.8460076045627376,
+      "grad_norm": 0.7479644701388113,
+      "learning_rate": 1.4304611392197399e-05,
+      "loss": 0.4173,
+      "step": 445
+    },
+    {
+      "epoch": 0.8479087452471483,
+      "grad_norm": 0.6528540743262322,
+      "learning_rate": 1.427461276179517e-05,
+      "loss": 0.4263,
+      "step": 446
+    },
+    {
+      "epoch": 0.8498098859315589,
+      "grad_norm": 0.6173880840879935,
+      "learning_rate": 1.4244566988758152e-05,
+      "loss": 0.4101,
+      "step": 447
+    },
+    {
+      "epoch": 0.8517110266159695,
+      "grad_norm": 0.6199207449684375,
+      "learning_rate": 1.4214474404446633e-05,
+      "loss": 0.415,
+      "step": 448
+    },
+    {
+      "epoch": 0.8536121673003803,
+      "grad_norm": 0.6491569646490017,
+      "learning_rate": 1.4184335340737158e-05,
+      "loss": 0.4185,
+      "step": 449
+    },
+    {
+      "epoch": 0.8555133079847909,
+      "grad_norm": 0.6214992787226633,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.4199,
+      "step": 450
+    },
+    {
+      "epoch": 0.8574144486692015,
+      "grad_norm": 0.6504552307089649,
+      "learning_rate": 1.4123919105189836e-05,
+      "loss": 0.4217,
+      "step": 451
+    },
+    {
+      "epoch": 0.8593155893536122,
+      "grad_norm": 0.6655583980495458,
+      "learning_rate": 1.4093642599653406e-05,
+      "loss": 0.4263,
+      "step": 452
+    },
+    {
+      "epoch": 0.8612167300380228,
+      "grad_norm": 0.6256798679709287,
+      "learning_rate": 1.40633209473145e-05,
+      "loss": 0.4429,
+      "step": 453
+    },
+    {
+      "epoch": 0.8631178707224335,
+      "grad_norm": 0.6816148690858034,
+      "learning_rate": 1.4032954482575938e-05,
+      "loss": 0.4183,
+      "step": 454
+    },
+    {
+      "epoch": 0.8650190114068441,
+      "grad_norm": 0.6421545016228782,
+      "learning_rate": 1.4002543540334766e-05,
+      "loss": 0.4366,
+      "step": 455
+    },
+    {
+      "epoch": 0.8669201520912547,
+      "grad_norm": 0.6609345755235317,
+      "learning_rate": 1.3972088455978537e-05,
+      "loss": 0.4226,
+      "step": 456
+    },
+    {
+      "epoch": 0.8688212927756654,
+      "grad_norm": 0.7126686385801381,
+      "learning_rate": 1.3941589565381635e-05,
+      "loss": 0.4064,
+      "step": 457
+    },
+    {
+      "epoch": 0.870722433460076,
+      "grad_norm": 0.6481140353638131,
+      "learning_rate": 1.391104720490156e-05,
+      "loss": 0.4477,
+      "step": 458
+    },
+    {
+      "epoch": 0.8726235741444867,
+      "grad_norm": 0.6301361158362933,
+      "learning_rate": 1.3880461711375224e-05,
+      "loss": 0.3949,
+      "step": 459
+    },
+    {
+      "epoch": 0.8745247148288974,
+      "grad_norm": 0.6779760380630674,
+      "learning_rate": 1.3849833422115221e-05,
+      "loss": 0.4168,
+      "step": 460
+    },
+    {
+      "epoch": 0.876425855513308,
+      "grad_norm": 0.6710277551654018,
+      "learning_rate": 1.3819162674906134e-05,
+      "loss": 0.4149,
+      "step": 461
+    },
+    {
+      "epoch": 0.8783269961977186,
+      "grad_norm": 0.6364363791697419,
+      "learning_rate": 1.378844980800078e-05,
+      "loss": 0.4024,
+      "step": 462
+    },
+    {
+      "epoch": 0.8802281368821293,
+      "grad_norm": 0.6570421337044803,
+      "learning_rate": 1.3757695160116502e-05,
+      "loss": 0.433,
+      "step": 463
+    },
+    {
+      "epoch": 0.8821292775665399,
+      "grad_norm": 0.6405479516348339,
+      "learning_rate": 1.3726899070431423e-05,
+      "loss": 0.4062,
+      "step": 464
+    },
+    {
+      "epoch": 0.8840304182509505,
+      "grad_norm": 0.6696580217493422,
+      "learning_rate": 1.3696061878580707e-05,
+      "loss": 0.4269,
+      "step": 465
+    },
+    {
+      "epoch": 0.8859315589353612,
+      "grad_norm": 0.6720440343618022,
+      "learning_rate": 1.3665183924652817e-05,
+      "loss": 0.4272,
+      "step": 466
+    },
+    {
+      "epoch": 0.8878326996197718,
+      "grad_norm": 0.6663390477945006,
+      "learning_rate": 1.3634265549185755e-05,
+      "loss": 0.3892,
+      "step": 467
+    },
+    {
+      "epoch": 0.8897338403041825,
+      "grad_norm": 0.6765994461474362,
+      "learning_rate": 1.3603307093163319e-05,
+      "loss": 0.4342,
+      "step": 468
+    },
+    {
+      "epoch": 0.8916349809885932,
+      "grad_norm": 0.6657557020728164,
+      "learning_rate": 1.3572308898011328e-05,
+      "loss": 0.4064,
+      "step": 469
+    },
+    {
+      "epoch": 0.8935361216730038,
+      "grad_norm": 0.688524577354246,
+      "learning_rate": 1.3541271305593878e-05,
+      "loss": 0.4331,
+      "step": 470
+    },
+    {
+      "epoch": 0.8954372623574145,
+      "grad_norm": 0.6527483753745547,
+      "learning_rate": 1.3510194658209547e-05,
+      "loss": 0.4199,
+      "step": 471
+    },
+    {
+      "epoch": 0.8973384030418251,
+      "grad_norm": 0.6624149995106321,
+      "learning_rate": 1.3479079298587634e-05,
+      "loss": 0.3969,
+      "step": 472
+    },
+    {
+      "epoch": 0.8992395437262357,
+      "grad_norm": 0.6050892214873685,
+      "learning_rate": 1.3447925569884374e-05,
+      "loss": 0.4091,
+      "step": 473
+    },
+    {
+      "epoch": 0.9011406844106464,
+      "grad_norm": 0.7154467170912298,
+      "learning_rate": 1.3416733815679166e-05,
+      "loss": 0.4263,
+      "step": 474
+    },
+    {
+      "epoch": 0.903041825095057,
+      "grad_norm": 0.6657935173917197,
+      "learning_rate": 1.3385504379970764e-05,
+      "loss": 0.4313,
+      "step": 475
+    },
+    {
+      "epoch": 0.9049429657794676,
+      "grad_norm": 0.6405416094572317,
+      "learning_rate": 1.3354237607173494e-05,
+      "loss": 0.4115,
+      "step": 476
+    },
+    {
+      "epoch": 0.9068441064638784,
+      "grad_norm": 0.6645840153852225,
+      "learning_rate": 1.3322933842113457e-05,
+      "loss": 0.402,
+      "step": 477
+    },
+    {
+      "epoch": 0.908745247148289,
+      "grad_norm": 0.5832288804583774,
+      "learning_rate": 1.3291593430024727e-05,
+      "loss": 0.4049,
+      "step": 478
+    },
+    {
+      "epoch": 0.9106463878326996,
+      "grad_norm": 0.6332076261371975,
+      "learning_rate": 1.3260216716545534e-05,
+      "loss": 0.4179,
+      "step": 479
+    },
+    {
+      "epoch": 0.9125475285171103,
+      "grad_norm": 0.6772257736328506,
+      "learning_rate": 1.3228804047714462e-05,
+      "loss": 0.4029,
+      "step": 480
+    },
+    {
+      "epoch": 0.9144486692015209,
+      "grad_norm": 0.6541532252566259,
+      "learning_rate": 1.319735576996663e-05,
+      "loss": 0.4517,
+      "step": 481
+    },
+    {
+      "epoch": 0.9163498098859315,
+      "grad_norm": 0.6904332877152957,
+      "learning_rate": 1.3165872230129869e-05,
+      "loss": 0.4117,
+      "step": 482
+    },
+    {
+      "epoch": 0.9182509505703422,
+      "grad_norm": 0.6345490426666023,
+      "learning_rate": 1.3134353775420895e-05,
+      "loss": 0.4023,
+      "step": 483
+    },
+    {
+      "epoch": 0.9201520912547528,
+      "grad_norm": 0.6334871318858123,
+      "learning_rate": 1.3102800753441488e-05,
+      "loss": 0.4291,
+      "step": 484
+    },
+    {
+      "epoch": 0.9220532319391636,
+      "grad_norm": 0.6532062088255811,
+      "learning_rate": 1.3071213512174655e-05,
+      "loss": 0.4002,
+      "step": 485
+    },
+    {
+      "epoch": 0.9239543726235742,
+      "grad_norm": 0.6138941964993868,
+      "learning_rate": 1.3039592399980785e-05,
+      "loss": 0.3873,
+      "step": 486
+    },
+    {
+      "epoch": 0.9258555133079848,
+      "grad_norm": 0.7291808845277177,
+      "learning_rate": 1.3007937765593818e-05,
+      "loss": 0.4329,
+      "step": 487
+    },
+    {
+      "epoch": 0.9277566539923955,
+      "grad_norm": 0.6996319713613872,
+      "learning_rate": 1.2976249958117395e-05,
+      "loss": 0.4194,
+      "step": 488
+    },
+    {
+      "epoch": 0.9296577946768061,
+      "grad_norm": 0.7192151041203426,
+      "learning_rate": 1.2944529327021002e-05,
+      "loss": 0.4028,
+      "step": 489
+    },
+    {
+      "epoch": 0.9315589353612167,
+      "grad_norm": 0.6374040267871274,
+      "learning_rate": 1.291277622213612e-05,
+      "loss": 0.4079,
+      "step": 490
+    },
+    {
+      "epoch": 0.9334600760456274,
+      "grad_norm": 0.6389432737752742,
+      "learning_rate": 1.2880990993652379e-05,
+      "loss": 0.4187,
+      "step": 491
+    },
+    {
+      "epoch": 0.935361216730038,
+      "grad_norm": 0.6198767361950689,
+      "learning_rate": 1.2849173992113669e-05,
+      "loss": 0.4002,
+      "step": 492
+    },
+    {
+      "epoch": 0.9372623574144486,
+      "grad_norm": 0.6110552629674318,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.4353,
+      "step": 493
+    },
+    {
+      "epoch": 0.9391634980988594,
+      "grad_norm": 0.6374508335989014,
+      "learning_rate": 1.2785446073795118e-05,
+      "loss": 0.4184,
+      "step": 494
+    },
+    {
+      "epoch": 0.94106463878327,
+      "grad_norm": 0.612295300882452,
+      "learning_rate": 1.2753535859839638e-05,
+      "loss": 0.4253,
+      "step": 495
+    },
+    {
+      "epoch": 0.9429657794676806,
+      "grad_norm": 0.6142557329024687,
+      "learning_rate": 1.272159527847016e-05,
+      "loss": 0.4132,
+      "step": 496
+    },
+    {
+      "epoch": 0.9448669201520913,
+      "grad_norm": 0.6299748776572197,
+      "learning_rate": 1.2689624681943897e-05,
+      "loss": 0.4514,
+      "step": 497
+    },
+    {
+      "epoch": 0.9467680608365019,
+      "grad_norm": 0.6511253265365365,
+      "learning_rate": 1.2657624422849077e-05,
+      "loss": 0.428,
+      "step": 498
+    },
+    {
+      "epoch": 0.9486692015209125,
+      "grad_norm": 0.7438845353781717,
+      "learning_rate": 1.2625594854101066e-05,
+      "loss": 0.4281,
+      "step": 499
+    },
+    {
+      "epoch": 0.9505703422053232,
+      "grad_norm": 0.65685452942434,
+      "learning_rate": 1.2593536328938471e-05,
+      "loss": 0.4288,
+      "step": 500
+    },
+    {
+      "epoch": 0.9524714828897338,
+      "grad_norm": 0.6408202528965171,
+      "learning_rate": 1.2561449200919253e-05,
+      "loss": 0.3934,
+      "step": 501
+    },
+    {
+      "epoch": 0.9543726235741445,
+      "grad_norm": 0.5953660898719252,
+      "learning_rate": 1.2529333823916807e-05,
+      "loss": 0.4079,
+      "step": 502
+    },
+    {
+      "epoch": 0.9562737642585551,
+      "grad_norm": 0.6550441395846707,
+      "learning_rate": 1.2497190552116082e-05,
+      "loss": 0.4308,
+      "step": 503
+    },
+    {
+      "epoch": 0.9581749049429658,
+      "grad_norm": 0.6476356798725659,
+      "learning_rate": 1.2465019740009662e-05,
+      "loss": 0.4006,
+      "step": 504
+    },
+    {
+      "epoch": 0.9600760456273765,
+      "grad_norm": 0.6348145725893728,
+      "learning_rate": 1.2432821742393854e-05,
+      "loss": 0.4071,
+      "step": 505
+    },
+    {
+      "epoch": 0.9619771863117871,
+      "grad_norm": 0.6144705921984532,
+      "learning_rate": 1.2400596914364792e-05,
+      "loss": 0.3919,
+      "step": 506
+    },
+    {
+      "epoch": 0.9638783269961977,
+      "grad_norm": 0.6607309299818755,
+      "learning_rate": 1.2368345611314508e-05,
+      "loss": 0.4213,
+      "step": 507
+    },
+    {
+      "epoch": 0.9657794676806084,
+      "grad_norm": 0.641116774400723,
+      "learning_rate": 1.2336068188927002e-05,
+      "loss": 0.3967,
+      "step": 508
+    },
+    {
+      "epoch": 0.967680608365019,
+      "grad_norm": 0.598757156191632,
+      "learning_rate": 1.2303765003174342e-05,
+      "loss": 0.4206,
+      "step": 509
+    },
+    {
+      "epoch": 0.9695817490494296,
+      "grad_norm": 0.6774339793330069,
+      "learning_rate": 1.2271436410312727e-05,
+      "loss": 0.4428,
+      "step": 510
+    },
+    {
+      "epoch": 0.9714828897338403,
+      "grad_norm": 0.6371606098847108,
+      "learning_rate": 1.2239082766878557e-05,
+      "loss": 0.4288,
+      "step": 511
+    },
+    {
+      "epoch": 0.973384030418251,
+      "grad_norm": 0.6987166984296828,
+      "learning_rate": 1.2206704429684504e-05,
+      "loss": 0.4236,
+      "step": 512
+    },
+    {
+      "epoch": 0.9752851711026616,
+      "grad_norm": 0.766784578482961,
+      "learning_rate": 1.2174301755815572e-05,
+      "loss": 0.4258,
+      "step": 513
+    },
+    {
+      "epoch": 0.9771863117870723,
+      "grad_norm": 0.5944147433838181,
+      "learning_rate": 1.2141875102625166e-05,
+      "loss": 0.3913,
+      "step": 514
+    },
+    {
+      "epoch": 0.9790874524714829,
+      "grad_norm": 0.6294229579228068,
+      "learning_rate": 1.2109424827731144e-05,
+      "loss": 0.4194,
+      "step": 515
+    },
+    {
+      "epoch": 0.9809885931558935,
+      "grad_norm": 0.6920565540604553,
+      "learning_rate": 1.2076951289011884e-05,
+      "loss": 0.4228,
+      "step": 516
+    },
+    {
+      "epoch": 0.9828897338403042,
+      "grad_norm": 0.6337519856004178,
+      "learning_rate": 1.204445484460232e-05,
+      "loss": 0.4188,
+      "step": 517
+    },
+    {
+      "epoch": 0.9847908745247148,
+      "grad_norm": 0.70754716105521,
+      "learning_rate": 1.2011935852890004e-05,
+      "loss": 0.4267,
+      "step": 518
+    },
+    {
+      "epoch": 0.9866920152091255,
+      "grad_norm": 0.680529952509599,
+      "learning_rate": 1.1979394672511156e-05,
+      "loss": 0.3978,
+      "step": 519
+    },
+    {
+      "epoch": 0.9885931558935361,
+      "grad_norm": 0.6478251182176888,
+      "learning_rate": 1.19468316623467e-05,
+      "loss": 0.4459,
+      "step": 520
+    },
+    {
+      "epoch": 0.9904942965779467,
+      "grad_norm": 0.7225331559840614,
+      "learning_rate": 1.1914247181518312e-05,
+      "loss": 0.4059,
+      "step": 521
+    },
+    {
+      "epoch": 0.9923954372623575,
+      "grad_norm": 0.8159415444532943,
+      "learning_rate": 1.1881641589384456e-05,
+      "loss": 0.427,
+      "step": 522
+    },
+    {
+      "epoch": 0.9942965779467681,
+      "grad_norm": 0.6423332729839314,
+      "learning_rate": 1.1849015245536424e-05,
+      "loss": 0.4238,
+      "step": 523
+    },
+    {
+      "epoch": 0.9961977186311787,
+      "grad_norm": 0.732772774576064,
+      "learning_rate": 1.1816368509794365e-05,
+      "loss": 0.4239,
+      "step": 524
+    },
+    {
+      "epoch": 0.9980988593155894,
+      "grad_norm": 0.6642073585899388,
+      "learning_rate": 1.1783701742203326e-05,
+      "loss": 0.3882,
+      "step": 525
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6339019571030245,
+      "learning_rate": 1.1751015303029272e-05,
+      "loss": 0.4253,
+      "step": 526
+    },
+    {
+      "epoch": 1.0019011406844107,
+      "grad_norm": 0.8239085685640279,
+      "learning_rate": 1.1718309552755118e-05,
+      "loss": 0.3266,
+      "step": 527
+    },
+    {
+      "epoch": 1.0038022813688212,
+      "grad_norm": 0.7355822102448376,
+      "learning_rate": 1.1685584852076746e-05,
+      "loss": 0.3359,
+      "step": 528
+    },
+    {
+      "epoch": 1.005703422053232,
+      "grad_norm": 0.630421432244059,
+      "learning_rate": 1.1652841561899042e-05,
+      "loss": 0.322,
+      "step": 529
+    },
+    {
+      "epoch": 1.0076045627376427,
+      "grad_norm": 0.6121749709891758,
+      "learning_rate": 1.1620080043331901e-05,
+      "loss": 0.3107,
+      "step": 530
+    },
+    {
+      "epoch": 1.0095057034220531,
+      "grad_norm": 0.7416239576210724,
+      "learning_rate": 1.1587300657686254e-05,
+      "loss": 0.3165,
+      "step": 531
+    },
+    {
+      "epoch": 1.0114068441064639,
+      "grad_norm": 0.8168067318335017,
+      "learning_rate": 1.1554503766470069e-05,
+      "loss": 0.32,
+      "step": 532
+    },
+    {
+      "epoch": 1.0133079847908746,
+      "grad_norm": 0.7786299165633455,
+      "learning_rate": 1.1521689731384391e-05,
+      "loss": 0.3231,
+      "step": 533
+    },
+    {
+      "epoch": 1.015209125475285,
+      "grad_norm": 0.7741237001317054,
+      "learning_rate": 1.1488858914319321e-05,
+      "loss": 0.3121,
+      "step": 534
+    },
+    {
+      "epoch": 1.0171102661596958,
+      "grad_norm": 0.7267180368735672,
+      "learning_rate": 1.1456011677350052e-05,
+      "loss": 0.3015,
+      "step": 535
+    },
+    {
+      "epoch": 1.0190114068441065,
+      "grad_norm": 0.7114524508818363,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.327,
+      "step": 536
+    },
+    {
+      "epoch": 1.020912547528517,
+      "grad_norm": 0.7507911819851849,
+      "learning_rate": 1.1390269392901096e-05,
+      "loss": 0.2997,
+      "step": 537
+    },
+    {
+      "epoch": 1.0228136882129277,
+      "grad_norm": 0.7164312252640419,
+      "learning_rate": 1.1357375070461241e-05,
+      "loss": 0.2963,
+      "step": 538
+    },
+    {
+      "epoch": 1.0247148288973384,
+      "grad_norm": 0.6897195972386778,
+      "learning_rate": 1.1324465778188846e-05,
+      "loss": 0.3263,
+      "step": 539
+    },
+    {
+      "epoch": 1.026615969581749,
+      "grad_norm": 0.6828722333901289,
+      "learning_rate": 1.1291541879024568e-05,
+      "loss": 0.3247,
+      "step": 540
+    },
+    {
+      "epoch": 1.0285171102661597,
+      "grad_norm": 0.6734226171859484,
+      "learning_rate": 1.1258603736070145e-05,
+      "loss": 0.2889,
+      "step": 541
+    },
+    {
+      "epoch": 1.0304182509505704,
+      "grad_norm": 0.6604300292551003,
+      "learning_rate": 1.1225651712584413e-05,
+      "loss": 0.3144,
+      "step": 542
+    },
+    {
+      "epoch": 1.032319391634981,
+      "grad_norm": 0.6854430020701753,
+      "learning_rate": 1.1192686171979288e-05,
+      "loss": 0.317,
+      "step": 543
+    },
+    {
+      "epoch": 1.0342205323193916,
+      "grad_norm": 0.6899530939415073,
+      "learning_rate": 1.1159707477815756e-05,
+      "loss": 0.2954,
+      "step": 544
+    },
+    {
+      "epoch": 1.0361216730038023,
+      "grad_norm": 0.7248002314170193,
+      "learning_rate": 1.1126715993799875e-05,
+      "loss": 0.305,
+      "step": 545
+    },
+    {
+      "epoch": 1.038022813688213,
+      "grad_norm": 0.6706190198686904,
+      "learning_rate": 1.1093712083778748e-05,
+      "loss": 0.3111,
+      "step": 546
+    },
+    {
+      "epoch": 1.0399239543726235,
+      "grad_norm": 0.6729668449571125,
+      "learning_rate": 1.1060696111736515e-05,
+      "loss": 0.3223,
+      "step": 547
+    },
+    {
+      "epoch": 1.0418250950570342,
+      "grad_norm": 0.7291151051765314,
+      "learning_rate": 1.1027668441790358e-05,
+      "loss": 0.2977,
+      "step": 548
+    },
+    {
+      "epoch": 1.043726235741445,
+      "grad_norm": 0.6678563579676036,
+      "learning_rate": 1.099462943818646e-05,
+      "loss": 0.3226,
+      "step": 549
+    },
+    {
+      "epoch": 1.0456273764258555,
+      "grad_norm": 0.6516439653461512,
+      "learning_rate": 1.0961579465295987e-05,
+      "loss": 0.3007,
+      "step": 550
+    },
+    {
+      "epoch": 1.0475285171102662,
+      "grad_norm": 0.7388516148486083,
+      "learning_rate": 1.0928518887611099e-05,
+      "loss": 0.3176,
+      "step": 551
+    },
+    {
+      "epoch": 1.049429657794677,
+      "grad_norm": 0.7357088911294506,
+      "learning_rate": 1.0895448069740902e-05,
+      "loss": 0.3104,
+      "step": 552
+    },
+    {
+      "epoch": 1.0513307984790874,
+      "grad_norm": 0.6679301979965167,
+      "learning_rate": 1.0862367376407433e-05,
+      "loss": 0.3045,
+      "step": 553
+    },
+    {
+      "epoch": 1.053231939163498,
+      "grad_norm": 0.7317499764200953,
+      "learning_rate": 1.0829277172441648e-05,
+      "loss": 0.316,
+      "step": 554
+    },
+    {
+      "epoch": 1.0551330798479088,
+      "grad_norm": 0.604387376523591,
+      "learning_rate": 1.0796177822779384e-05,
+      "loss": 0.3056,
+      "step": 555
+    },
+    {
+      "epoch": 1.0570342205323193,
+      "grad_norm": 0.7076392777013988,
+      "learning_rate": 1.0763069692457346e-05,
+      "loss": 0.3248,
+      "step": 556
+    },
+    {
+      "epoch": 1.05893536121673,
+      "grad_norm": 0.7537917590791791,
+      "learning_rate": 1.0729953146609076e-05,
+      "loss": 0.3185,
+      "step": 557
+    },
+    {
+      "epoch": 1.0608365019011408,
+      "grad_norm": 0.6840036590775822,
+      "learning_rate": 1.0696828550460928e-05,
+      "loss": 0.3209,
+      "step": 558
+    },
+    {
+      "epoch": 1.0627376425855513,
+      "grad_norm": 0.6925716243596572,
+      "learning_rate": 1.0663696269328034e-05,
+      "loss": 0.3169,
+      "step": 559
+    },
+    {
+      "epoch": 1.064638783269962,
+      "grad_norm": 0.6225564181707167,
+      "learning_rate": 1.0630556668610286e-05,
+      "loss": 0.2865,
+      "step": 560
+    },
+    {
+      "epoch": 1.0665399239543727,
+      "grad_norm": 0.6712066776547839,
+      "learning_rate": 1.059741011378829e-05,
+      "loss": 0.3051,
+      "step": 561
+    },
+    {
+      "epoch": 1.0684410646387832,
+      "grad_norm": 0.6614781310313652,
+      "learning_rate": 1.0564256970419367e-05,
+      "loss": 0.3057,
+      "step": 562
+    },
+    {
+      "epoch": 1.070342205323194,
+      "grad_norm": 0.724613447872154,
+      "learning_rate": 1.0531097604133473e-05,
+      "loss": 0.3072,
+      "step": 563
+    },
+    {
+      "epoch": 1.0722433460076046,
+      "grad_norm": 0.695659983130449,
+      "learning_rate": 1.0497932380629207e-05,
+      "loss": 0.3274,
+      "step": 564
+    },
+    {
+      "epoch": 1.0741444866920151,
+      "grad_norm": 0.6217907364272548,
+      "learning_rate": 1.0464761665669771e-05,
+      "loss": 0.3089,
+      "step": 565
+    },
+    {
+      "epoch": 1.0760456273764258,
+      "grad_norm": 0.7283902958211673,
+      "learning_rate": 1.0431585825078916e-05,
+      "loss": 0.3141,
+      "step": 566
+    },
+    {
+      "epoch": 1.0779467680608366,
+      "grad_norm": 0.737058547373869,
+      "learning_rate": 1.0398405224736927e-05,
+      "loss": 0.3095,
+      "step": 567
+    },
+    {
+      "epoch": 1.079847908745247,
+      "grad_norm": 0.6485179118211635,
+      "learning_rate": 1.0365220230576592e-05,
+      "loss": 0.2917,
+      "step": 568
+    },
+    {
+      "epoch": 1.0817490494296578,
+      "grad_norm": 0.7434968188364466,
+      "learning_rate": 1.0332031208579133e-05,
+      "loss": 0.321,
+      "step": 569
+    },
+    {
+      "epoch": 1.0836501901140685,
+      "grad_norm": 0.6716488675118957,
+      "learning_rate": 1.0298838524770212e-05,
+      "loss": 0.2989,
+      "step": 570
+    },
+    {
+      "epoch": 1.085551330798479,
+      "grad_norm": 0.6363564003818519,
+      "learning_rate": 1.0265642545215872e-05,
+      "loss": 0.2973,
+      "step": 571
+    },
+    {
+      "epoch": 1.0874524714828897,
+      "grad_norm": 0.6544298318179953,
+      "learning_rate": 1.0232443636018502e-05,
+      "loss": 0.3188,
+      "step": 572
+    },
+    {
+      "epoch": 1.0893536121673004,
+      "grad_norm": 0.6994160556406729,
+      "learning_rate": 1.0199242163312794e-05,
+      "loss": 0.2914,
+      "step": 573
+    },
+    {
+      "epoch": 1.091254752851711,
+      "grad_norm": 0.6504286706974884,
+      "learning_rate": 1.0166038493261723e-05,
+      "loss": 0.2904,
+      "step": 574
+    },
+    {
+      "epoch": 1.0931558935361216,
+      "grad_norm": 0.7000133547980442,
+      "learning_rate": 1.013283299205249e-05,
+      "loss": 0.2989,
+      "step": 575
+    },
+    {
+      "epoch": 1.0950570342205324,
+      "grad_norm": 0.6452939855020527,
+      "learning_rate": 1.0099626025892491e-05,
+      "loss": 0.2936,
+      "step": 576
+    },
+    {
+      "epoch": 1.0969581749049429,
+      "grad_norm": 0.7291773519340569,
+      "learning_rate": 1.0066417961005283e-05,
+      "loss": 0.3115,
+      "step": 577
+    },
+    {
+      "epoch": 1.0988593155893536,
+      "grad_norm": 0.6956090466113257,
+      "learning_rate": 1.0033209163626539e-05,
+      "loss": 0.3052,
+      "step": 578
+    },
+    {
+      "epoch": 1.1007604562737643,
+      "grad_norm": 0.6910070043006833,
+      "learning_rate": 1e-05,
+      "loss": 0.3066,
+      "step": 579
+    },
+    {
+      "epoch": 1.102661596958175,
+      "grad_norm": 0.6451305031987221,
+      "learning_rate": 9.966790836373465e-06,
+      "loss": 0.2996,
+      "step": 580
+    },
+    {
+      "epoch": 1.1045627376425855,
+      "grad_norm": 0.6978879307153206,
+      "learning_rate": 9.933582038994719e-06,
+      "loss": 0.3068,
+      "step": 581
+    },
+    {
+      "epoch": 1.1064638783269962,
+      "grad_norm": 0.6465859220949993,
+      "learning_rate": 9.90037397410751e-06,
+      "loss": 0.3217,
+      "step": 582
+    },
+    {
+      "epoch": 1.108365019011407,
+      "grad_norm": 0.6936796253130603,
+      "learning_rate": 9.867167007947511e-06,
+      "loss": 0.3082,
+      "step": 583
+    },
+    {
+      "epoch": 1.1102661596958174,
+      "grad_norm": 0.6724611244553041,
+      "learning_rate": 9.833961506738282e-06,
+      "loss": 0.3042,
+      "step": 584
+    },
+    {
+      "epoch": 1.1121673003802282,
+      "grad_norm": 0.6293147033807189,
+      "learning_rate": 9.80075783668721e-06,
+      "loss": 0.2936,
+      "step": 585
+    },
+    {
+      "epoch": 1.1140684410646389,
+      "grad_norm": 0.7092154760280038,
+      "learning_rate": 9.767556363981503e-06,
+      "loss": 0.3254,
+      "step": 586
+    },
+    {
+      "epoch": 1.1159695817490494,
+      "grad_norm": 0.6478261321935254,
+      "learning_rate": 9.734357454784131e-06,
+      "loss": 0.3105,
+      "step": 587
+    },
+    {
+      "epoch": 1.11787072243346,
+      "grad_norm": 0.6818119649452757,
+      "learning_rate": 9.701161475229791e-06,
+      "loss": 0.3076,
+      "step": 588
+    },
+    {
+      "epoch": 1.1197718631178708,
+      "grad_norm": 0.6678839023714132,
+      "learning_rate": 9.66796879142087e-06,
+      "loss": 0.2985,
+      "step": 589
+    },
+    {
+      "epoch": 1.1216730038022813,
+      "grad_norm": 0.668010853059842,
+      "learning_rate": 9.634779769423412e-06,
+      "loss": 0.295,
+      "step": 590
+    },
+    {
+      "epoch": 1.123574144486692,
+      "grad_norm": 0.6528962962633305,
+      "learning_rate": 9.601594775263073e-06,
+      "loss": 0.3028,
+      "step": 591
+    },
+    {
+      "epoch": 1.1254752851711027,
+      "grad_norm": 0.6624218057685479,
+      "learning_rate": 9.568414174921085e-06,
+      "loss": 0.3117,
+      "step": 592
+    },
+    {
+      "epoch": 1.1273764258555132,
+      "grad_norm": 0.6906647146965358,
+      "learning_rate": 9.535238334330234e-06,
+      "loss": 0.3122,
+      "step": 593
+    },
+    {
+      "epoch": 1.129277566539924,
+      "grad_norm": 0.6940521851319968,
+      "learning_rate": 9.502067619370794e-06,
+      "loss": 0.314,
+      "step": 594
+    },
+    {
+      "epoch": 1.1311787072243347,
+      "grad_norm": 0.6681529422341516,
+      "learning_rate": 9.468902395866532e-06,
+      "loss": 0.3122,
+      "step": 595
+    },
+    {
+      "epoch": 1.1330798479087452,
+      "grad_norm": 0.6916290301464028,
+      "learning_rate": 9.435743029580638e-06,
+      "loss": 0.3125,
+      "step": 596
+    },
+    {
+      "epoch": 1.1349809885931559,
+      "grad_norm": 0.6922400153102771,
+      "learning_rate": 9.402589886211711e-06,
+      "loss": 0.2984,
+      "step": 597
+    },
+    {
+      "epoch": 1.1368821292775666,
+      "grad_norm": 0.6711998361923937,
+      "learning_rate": 9.369443331389718e-06,
+      "loss": 0.3071,
+      "step": 598
+    },
+    {
+      "epoch": 1.138783269961977,
+      "grad_norm": 0.7409213872877687,
+      "learning_rate": 9.336303730671968e-06,
+      "loss": 0.3109,
+      "step": 599
+    },
+    {
+      "epoch": 1.1406844106463878,
+      "grad_norm": 0.7288395663639611,
+      "learning_rate": 9.303171449539074e-06,
+      "loss": 0.3205,
+      "step": 600
+    },
+    {
+      "epoch": 1.1425855513307985,
+      "grad_norm": 0.6710727969695096,
+      "learning_rate": 9.270046853390924e-06,
+      "loss": 0.3072,
+      "step": 601
+    },
+    {
+      "epoch": 1.144486692015209,
+      "grad_norm": 0.6974241062215073,
+      "learning_rate": 9.236930307542654e-06,
+      "loss": 0.318,
+      "step": 602
+    },
+    {
+      "epoch": 1.1463878326996197,
+      "grad_norm": 0.6382783585641221,
+      "learning_rate": 9.203822177220621e-06,
+      "loss": 0.2904,
+      "step": 603
+    },
+    {
+      "epoch": 1.1482889733840305,
+      "grad_norm": 0.7722959203492658,
+      "learning_rate": 9.170722827558357e-06,
+      "loss": 0.3278,
+      "step": 604
+    },
+    {
+      "epoch": 1.1501901140684412,
+      "grad_norm": 0.6534977117383473,
+      "learning_rate": 9.13763262359257e-06,
+      "loss": 0.3166,
+      "step": 605
+    },
+    {
+      "epoch": 1.1520912547528517,
+      "grad_norm": 0.6822499374993315,
+      "learning_rate": 9.104551930259101e-06,
+      "loss": 0.3064,
+      "step": 606
+    },
+    {
+      "epoch": 1.1539923954372624,
+      "grad_norm": 0.7941313737772939,
+      "learning_rate": 9.071481112388905e-06,
+      "loss": 0.3287,
+      "step": 607
+    },
+    {
+      "epoch": 1.1558935361216731,
+      "grad_norm": 0.629270285848771,
+      "learning_rate": 9.038420534704015e-06,
+      "loss": 0.3,
+      "step": 608
+    },
+    {
+      "epoch": 1.1577946768060836,
+      "grad_norm": 0.6712943348476993,
+      "learning_rate": 9.005370561813545e-06,
+      "loss": 0.3256,
+      "step": 609
+    },
+    {
+      "epoch": 1.1596958174904943,
+      "grad_norm": 0.727838970892581,
+      "learning_rate": 8.972331558209644e-06,
+      "loss": 0.321,
+      "step": 610
+    },
+    {
+      "epoch": 1.161596958174905,
+      "grad_norm": 0.6545082811531332,
+      "learning_rate": 8.939303888263485e-06,
+      "loss": 0.2986,
+      "step": 611
+    },
+    {
+      "epoch": 1.1634980988593155,
+      "grad_norm": 0.6524884352806564,
+      "learning_rate": 8.906287916221259e-06,
+      "loss": 0.3145,
+      "step": 612
+    },
+    {
+      "epoch": 1.1653992395437263,
+      "grad_norm": 0.6596279203582796,
+      "learning_rate": 8.873284006200129e-06,
+      "loss": 0.3097,
+      "step": 613
+    },
+    {
+      "epoch": 1.167300380228137,
+      "grad_norm": 0.7150285137952024,
+      "learning_rate": 8.840292522184247e-06,
+      "loss": 0.2904,
+      "step": 614
+    },
+    {
+      "epoch": 1.1692015209125475,
+      "grad_norm": 0.6783323201564836,
+      "learning_rate": 8.807313828020715e-06,
+      "loss": 0.3279,
+      "step": 615
+    },
+    {
+      "epoch": 1.1711026615969582,
+      "grad_norm": 0.6856014928544015,
+      "learning_rate": 8.774348287415589e-06,
+      "loss": 0.3212,
+      "step": 616
+    },
+    {
+      "epoch": 1.173003802281369,
+      "grad_norm": 0.7029328447050676,
+      "learning_rate": 8.74139626392986e-06,
+      "loss": 0.3255,
+      "step": 617
+    },
+    {
+      "epoch": 1.1749049429657794,
+      "grad_norm": 0.7433860212778325,
+      "learning_rate": 8.708458120975436e-06,
+      "loss": 0.3215,
+      "step": 618
+    },
+    {
+      "epoch": 1.1768060836501901,
+      "grad_norm": 0.6539975215500312,
+      "learning_rate": 8.675534221811156e-06,
+      "loss": 0.3026,
+      "step": 619
+    },
+    {
+      "epoch": 1.1787072243346008,
+      "grad_norm": 0.6710983761746697,
+      "learning_rate": 8.64262492953876e-06,
+      "loss": 0.3117,
+      "step": 620
+    },
+    {
+      "epoch": 1.1806083650190113,
+      "grad_norm": 0.7046823956510755,
+      "learning_rate": 8.60973060709891e-06,
+      "loss": 0.3295,
+      "step": 621
+    },
+    {
+      "epoch": 1.182509505703422,
+      "grad_norm": 0.656668131565369,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.3086,
+      "step": 622
+    },
+    {
+      "epoch": 1.1844106463878328,
+      "grad_norm": 0.7082446119091285,
+      "learning_rate": 8.543988322649954e-06,
+      "loss": 0.3262,
+      "step": 623
+    },
+    {
+      "epoch": 1.1863117870722433,
+      "grad_norm": 0.7131654480343576,
+      "learning_rate": 8.511141085680684e-06,
+      "loss": 0.3264,
+      "step": 624
+    },
+    {
+      "epoch": 1.188212927756654,
+      "grad_norm": 0.6482616128863535,
+      "learning_rate": 8.478310268615612e-06,
+      "loss": 0.3054,
+      "step": 625
+    },
+    {
+      "epoch": 1.1901140684410647,
+      "grad_norm": 0.6704440024566604,
+      "learning_rate": 8.445496233529934e-06,
+      "loss": 0.3134,
+      "step": 626
+    },
+    {
+      "epoch": 1.1920152091254752,
+      "grad_norm": 0.6482355086913613,
+      "learning_rate": 8.41269934231375e-06,
+      "loss": 0.3057,
+      "step": 627
+    },
+    {
+      "epoch": 1.193916349809886,
+      "grad_norm": 0.6602982487727822,
+      "learning_rate": 8.3799199566681e-06,
+      "loss": 0.2908,
+      "step": 628
+    },
+    {
+      "epoch": 1.1958174904942966,
+      "grad_norm": 0.6342704197613757,
+      "learning_rate": 8.34715843810096e-06,
+      "loss": 0.2898,
+      "step": 629
+    },
+    {
+      "epoch": 1.1977186311787071,
+      "grad_norm": 0.6781832465530957,
+      "learning_rate": 8.314415147923254e-06,
+      "loss": 0.3133,
+      "step": 630
+    },
+    {
+      "epoch": 1.1996197718631179,
+      "grad_norm": 0.6964554751137004,
+      "learning_rate": 8.281690447244887e-06,
+      "loss": 0.3045,
+      "step": 631
+    },
+    {
+      "epoch": 1.2015209125475286,
+      "grad_norm": 0.7222448212086929,
+      "learning_rate": 8.248984696970732e-06,
+      "loss": 0.3121,
+      "step": 632
+    },
+    {
+      "epoch": 1.203422053231939,
+      "grad_norm": 0.6588115697988763,
+      "learning_rate": 8.216298257796677e-06,
+      "loss": 0.3173,
+      "step": 633
+    },
+    {
+      "epoch": 1.2053231939163498,
+      "grad_norm": 0.6720461140272046,
+      "learning_rate": 8.183631490205636e-06,
+      "loss": 0.3022,
+      "step": 634
+    },
+    {
+      "epoch": 1.2072243346007605,
+      "grad_norm": 0.700059179608859,
+      "learning_rate": 8.150984754463578e-06,
+      "loss": 0.3125,
+      "step": 635
+    },
+    {
+      "epoch": 1.209125475285171,
+      "grad_norm": 0.6433722531994444,
+      "learning_rate": 8.118358410615545e-06,
+      "loss": 0.2987,
+      "step": 636
+    },
+    {
+      "epoch": 1.2110266159695817,
+      "grad_norm": 0.6801759707901847,
+      "learning_rate": 8.08575281848169e-06,
+      "loss": 0.3044,
+      "step": 637
+    },
+    {
+      "epoch": 1.2129277566539924,
+      "grad_norm": 0.6782708763117696,
+      "learning_rate": 8.0531683376533e-06,
+      "loss": 0.2921,
+      "step": 638
+    },
+    {
+      "epoch": 1.214828897338403,
+      "grad_norm": 0.6641498904928511,
+      "learning_rate": 8.020605327488846e-06,
+      "loss": 0.3106,
+      "step": 639
+    },
+    {
+      "epoch": 1.2167300380228137,
+      "grad_norm": 0.6790114094181067,
+      "learning_rate": 7.988064147110001e-06,
+      "loss": 0.3159,
+      "step": 640
+    },
+    {
+      "epoch": 1.2186311787072244,
+      "grad_norm": 0.7064577224878649,
+      "learning_rate": 7.955545155397684e-06,
+      "loss": 0.3134,
+      "step": 641
+    },
+    {
+      "epoch": 1.2205323193916349,
+      "grad_norm": 0.675780099507623,
+      "learning_rate": 7.923048710988119e-06,
+      "loss": 0.3077,
+      "step": 642
+    },
+    {
+      "epoch": 1.2224334600760456,
+      "grad_norm": 0.6868545603511439,
+      "learning_rate": 7.890575172268858e-06,
+      "loss": 0.324,
+      "step": 643
+    },
+    {
+      "epoch": 1.2243346007604563,
+      "grad_norm": 0.7113017364518008,
+      "learning_rate": 7.858124897374837e-06,
+      "loss": 0.3078,
+      "step": 644
+    },
+    {
+      "epoch": 1.2262357414448668,
+      "grad_norm": 0.6994960307833665,
+      "learning_rate": 7.825698244184432e-06,
+      "loss": 0.2923,
+      "step": 645
+    },
+    {
+      "epoch": 1.2281368821292775,
+      "grad_norm": 0.6560477034727196,
+      "learning_rate": 7.7932955703155e-06,
+      "loss": 0.3018,
+      "step": 646
+    },
+    {
+      "epoch": 1.2300380228136882,
+      "grad_norm": 0.6748498510550248,
+      "learning_rate": 7.760917233121443e-06,
+      "loss": 0.2992,
+      "step": 647
+    },
+    {
+      "epoch": 1.231939163498099,
+      "grad_norm": 0.6933336722344913,
+      "learning_rate": 7.728563589687275e-06,
+      "loss": 0.3012,
+      "step": 648
+    },
+    {
+      "epoch": 1.2338403041825095,
+      "grad_norm": 0.6625868555319298,
+      "learning_rate": 7.696234996825663e-06,
+      "loss": 0.3102,
+      "step": 649
+    },
+    {
+      "epoch": 1.2357414448669202,
+      "grad_norm": 0.6655288361273267,
+      "learning_rate": 7.663931811073003e-06,
+      "loss": 0.3013,
+      "step": 650
+    },
+    {
+      "epoch": 1.2376425855513309,
+      "grad_norm": 0.6923213154413905,
+      "learning_rate": 7.631654388685496e-06,
+      "loss": 0.3056,
+      "step": 651
+    },
+    {
+      "epoch": 1.2395437262357414,
+      "grad_norm": 0.6689466317715584,
+      "learning_rate": 7.599403085635208e-06,
+      "loss": 0.3122,
+      "step": 652
+    },
+    {
+      "epoch": 1.241444866920152,
+      "grad_norm": 0.708287217073541,
+      "learning_rate": 7.567178257606147e-06,
+      "loss": 0.3308,
+      "step": 653
+    },
+    {
+      "epoch": 1.2433460076045628,
+      "grad_norm": 0.666459954289038,
+      "learning_rate": 7.534980259990341e-06,
+      "loss": 0.3118,
+      "step": 654
+    },
+    {
+      "epoch": 1.2452471482889733,
+      "grad_norm": 0.6635883065911933,
+      "learning_rate": 7.50280944788392e-06,
+      "loss": 0.2984,
+      "step": 655
+    },
+    {
+      "epoch": 1.247148288973384,
+      "grad_norm": 0.6660653549199904,
+      "learning_rate": 7.470666176083193e-06,
+      "loss": 0.3218,
+      "step": 656
+    },
+    {
+      "epoch": 1.2490494296577948,
+      "grad_norm": 0.6791258347971807,
+      "learning_rate": 7.438550799080746e-06,
+      "loss": 0.3015,
+      "step": 657
+    },
+    {
+      "epoch": 1.2509505703422052,
+      "grad_norm": 0.6595120929136269,
+      "learning_rate": 7.40646367106153e-06,
+      "loss": 0.3082,
+      "step": 658
+    },
+    {
+      "epoch": 1.252851711026616,
+      "grad_norm": 0.659029778938644,
+      "learning_rate": 7.3744051458989395e-06,
+      "loss": 0.3104,
+      "step": 659
+    },
+    {
+      "epoch": 1.2547528517110267,
+      "grad_norm": 0.683018986550129,
+      "learning_rate": 7.342375577150928e-06,
+      "loss": 0.3204,
+      "step": 660
+    },
+    {
+      "epoch": 1.2566539923954372,
+      "grad_norm": 0.6857923879949662,
+      "learning_rate": 7.310375318056107e-06,
+      "loss": 0.3085,
+      "step": 661
+    },
+    {
+      "epoch": 1.258555133079848,
+      "grad_norm": 0.6553915267650084,
+      "learning_rate": 7.278404721529843e-06,
+      "loss": 0.3124,
+      "step": 662
+    },
+    {
+      "epoch": 1.2604562737642586,
+      "grad_norm": 0.6439873239933867,
+      "learning_rate": 7.246464140160365e-06,
+      "loss": 0.317,
+      "step": 663
+    },
+    {
+      "epoch": 1.2623574144486693,
+      "grad_norm": 0.6735288889682733,
+      "learning_rate": 7.214553926204884e-06,
+      "loss": 0.3052,
+      "step": 664
+    },
+    {
+      "epoch": 1.2642585551330798,
+      "grad_norm": 0.6868619382571414,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.3059,
+      "step": 665
+    },
+    {
+      "epoch": 1.2661596958174905,
+      "grad_norm": 0.6824819862187799,
+      "learning_rate": 7.150826007886334e-06,
+      "loss": 0.3065,
+      "step": 666
+    },
+    {
+      "epoch": 1.2680608365019013,
+      "grad_norm": 0.6409503102681645,
+      "learning_rate": 7.119009006347625e-06,
+      "loss": 0.2914,
+      "step": 667
+    },
+    {
+      "epoch": 1.2699619771863118,
+      "grad_norm": 0.6821925317312917,
+      "learning_rate": 7.087223777863883e-06,
+      "loss": 0.3025,
+      "step": 668
+    },
+    {
+      "epoch": 1.2718631178707225,
+      "grad_norm": 0.7749259886203776,
+      "learning_rate": 7.055470672979003e-06,
+      "loss": 0.3316,
+      "step": 669
+    },
+    {
+      "epoch": 1.2737642585551332,
+      "grad_norm": 0.6759450156406435,
+      "learning_rate": 7.023750041882609e-06,
+      "loss": 0.3104,
+      "step": 670
+    },
+    {
+      "epoch": 1.2756653992395437,
+      "grad_norm": 0.676114953937676,
+      "learning_rate": 6.992062234406185e-06,
+      "loss": 0.3173,
+      "step": 671
+    },
+    {
+      "epoch": 1.2775665399239544,
+      "grad_norm": 0.7062518128760734,
+      "learning_rate": 6.960407600019217e-06,
+      "loss": 0.2919,
+      "step": 672
+    },
+    {
+      "epoch": 1.2794676806083651,
+      "grad_norm": 0.6658105084688699,
+      "learning_rate": 6.9287864878253475e-06,
+      "loss": 0.2977,
+      "step": 673
+    },
+    {
+      "epoch": 1.2813688212927756,
+      "grad_norm": 0.6857889827678252,
+      "learning_rate": 6.897199246558515e-06,
+      "loss": 0.3165,
+      "step": 674
+    },
+    {
+      "epoch": 1.2832699619771863,
+      "grad_norm": 0.6291326161554626,
+      "learning_rate": 6.865646224579108e-06,
+      "loss": 0.2942,
+      "step": 675
+    },
+    {
+      "epoch": 1.285171102661597,
+      "grad_norm": 0.7168017940337101,
+      "learning_rate": 6.834127769870134e-06,
+      "loss": 0.3097,
+      "step": 676
+    },
+    {
+      "epoch": 1.2870722433460076,
+      "grad_norm": 0.7047453826592799,
+      "learning_rate": 6.802644230033373e-06,
+      "loss": 0.3058,
+      "step": 677
+    },
+    {
+      "epoch": 1.2889733840304183,
+      "grad_norm": 0.6862387213993515,
+      "learning_rate": 6.771195952285541e-06,
+      "loss": 0.3,
+      "step": 678
+    },
+    {
+      "epoch": 1.290874524714829,
+      "grad_norm": 0.6807940830301819,
+      "learning_rate": 6.739783283454469e-06,
+      "loss": 0.3092,
+      "step": 679
+    },
+    {
+      "epoch": 1.2927756653992395,
+      "grad_norm": 0.660267405023725,
+      "learning_rate": 6.708406569975274e-06,
+      "loss": 0.2899,
+      "step": 680
+    },
+    {
+      "epoch": 1.2946768060836502,
+      "grad_norm": 0.6700291606951794,
+      "learning_rate": 6.6770661578865444e-06,
+      "loss": 0.3021,
+      "step": 681
+    },
+    {
+      "epoch": 1.296577946768061,
+      "grad_norm": 0.6987544790989073,
+      "learning_rate": 6.645762392826509e-06,
+      "loss": 0.3063,
+      "step": 682
+    },
+    {
+      "epoch": 1.2984790874524714,
+      "grad_norm": 0.7064464464617494,
+      "learning_rate": 6.614495620029238e-06,
+      "loss": 0.3164,
+      "step": 683
+    },
+    {
+      "epoch": 1.3003802281368821,
+      "grad_norm": 0.6517172370036731,
+      "learning_rate": 6.583266184320836e-06,
+      "loss": 0.3007,
+      "step": 684
+    },
+    {
+      "epoch": 1.3022813688212929,
+      "grad_norm": 0.7028107636567705,
+      "learning_rate": 6.552074430115624e-06,
+      "loss": 0.3209,
+      "step": 685
+    },
+    {
+      "epoch": 1.3041825095057034,
+      "grad_norm": 0.716432409595219,
+      "learning_rate": 6.520920701412371e-06,
+      "loss": 0.2885,
+      "step": 686
+    },
+    {
+      "epoch": 1.306083650190114,
+      "grad_norm": 0.7046990053274319,
+      "learning_rate": 6.489805341790456e-06,
+      "loss": 0.302,
+      "step": 687
+    },
+    {
+      "epoch": 1.3079847908745248,
+      "grad_norm": 0.6679369412402957,
+      "learning_rate": 6.458728694406124e-06,
+      "loss": 0.313,
+      "step": 688
+    },
+    {
+      "epoch": 1.3098859315589353,
+      "grad_norm": 0.6548242682031429,
+      "learning_rate": 6.427691101988673e-06,
+      "loss": 0.3024,
+      "step": 689
+    },
+    {
+      "epoch": 1.311787072243346,
+      "grad_norm": 0.7151513698454964,
+      "learning_rate": 6.396692906836686e-06,
+      "loss": 0.3282,
+      "step": 690
+    },
+    {
+      "epoch": 1.3136882129277567,
+      "grad_norm": 0.66242821878352,
+      "learning_rate": 6.3657344508142495e-06,
+      "loss": 0.293,
+      "step": 691
+    },
+    {
+      "epoch": 1.3155893536121672,
+      "grad_norm": 0.6677253426448723,
+      "learning_rate": 6.334816075347185e-06,
+      "loss": 0.2965,
+      "step": 692
+    },
+    {
+      "epoch": 1.317490494296578,
+      "grad_norm": 0.6643894079114213,
+      "learning_rate": 6.303938121419295e-06,
+      "loss": 0.2892,
+      "step": 693
+    },
+    {
+      "epoch": 1.3193916349809887,
+      "grad_norm": 0.6882569637798754,
+      "learning_rate": 6.273100929568579e-06,
+      "loss": 0.3028,
+      "step": 694
+    },
+    {
+      "epoch": 1.3212927756653992,
+      "grad_norm": 0.6890694849821942,
+      "learning_rate": 6.242304839883502e-06,
+      "loss": 0.3068,
+      "step": 695
+    },
+    {
+      "epoch": 1.3231939163498099,
+      "grad_norm": 0.6988825652365313,
+      "learning_rate": 6.211550191999223e-06,
+      "loss": 0.3227,
+      "step": 696
+    },
+    {
+      "epoch": 1.3250950570342206,
+      "grad_norm": 0.7260264267924778,
+      "learning_rate": 6.18083732509387e-06,
+      "loss": 0.3217,
+      "step": 697
+    },
+    {
+      "epoch": 1.326996197718631,
+      "grad_norm": 0.6834527977890457,
+      "learning_rate": 6.150166577884781e-06,
+      "loss": 0.3023,
+      "step": 698
+    },
+    {
+      "epoch": 1.3288973384030418,
+      "grad_norm": 0.7384960400255748,
+      "learning_rate": 6.119538288624778e-06,
+      "loss": 0.2927,
+      "step": 699
+    },
+    {
+      "epoch": 1.3307984790874525,
+      "grad_norm": 0.6758397253707514,
+      "learning_rate": 6.088952795098442e-06,
+      "loss": 0.2936,
+      "step": 700
+    },
+    {
+      "epoch": 1.332699619771863,
+      "grad_norm": 0.6869310961732931,
+      "learning_rate": 6.058410434618367e-06,
+      "loss": 0.3048,
+      "step": 701
+    },
+    {
+      "epoch": 1.3346007604562737,
+      "grad_norm": 0.6698342707024305,
+      "learning_rate": 6.027911544021465e-06,
+      "loss": 0.3017,
+      "step": 702
+    },
+    {
+      "epoch": 1.3365019011406845,
+      "grad_norm": 0.6569768988215678,
+      "learning_rate": 5.997456459665237e-06,
+      "loss": 0.2976,
+      "step": 703
+    },
+    {
+      "epoch": 1.338403041825095,
+      "grad_norm": 0.6592666688904228,
+      "learning_rate": 5.967045517424062e-06,
+      "loss": 0.3077,
+      "step": 704
+    },
+    {
+      "epoch": 1.3403041825095057,
+      "grad_norm": 0.7748930983605694,
+      "learning_rate": 5.936679052685505e-06,
+      "loss": 0.3233,
+      "step": 705
+    },
+    {
+      "epoch": 1.3422053231939164,
+      "grad_norm": 0.740363556308047,
+      "learning_rate": 5.906357400346596e-06,
+      "loss": 0.3113,
+      "step": 706
+    },
+    {
+      "epoch": 1.3441064638783269,
+      "grad_norm": 0.7070584442662466,
+      "learning_rate": 5.876080894810167e-06,
+      "loss": 0.3129,
+      "step": 707
+    },
+    {
+      "epoch": 1.3460076045627376,
+      "grad_norm": 0.7391264007310618,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.3025,
+      "step": 708
+    },
+    {
+      "epoch": 1.3479087452471483,
+      "grad_norm": 0.6956913431662224,
+      "learning_rate": 5.815664659262845e-06,
+      "loss": 0.3227,
+      "step": 709
+    },
+    {
+      "epoch": 1.3498098859315588,
+      "grad_norm": 0.6891894423428386,
+      "learning_rate": 5.78552559555337e-06,
+      "loss": 0.3009,
+      "step": 710
+    },
+    {
+      "epoch": 1.3517110266159695,
+      "grad_norm": 0.6518191046885317,
+      "learning_rate": 5.755433011241851e-06,
+      "loss": 0.3026,
+      "step": 711
+    },
+    {
+      "epoch": 1.3536121673003803,
+      "grad_norm": 0.7421902123934244,
+      "learning_rate": 5.725387238204831e-06,
+      "loss": 0.3247,
+      "step": 712
+    },
+    {
+      "epoch": 1.3555133079847907,
+      "grad_norm": 0.725770432434914,
+      "learning_rate": 5.695388607802603e-06,
+      "loss": 0.3149,
+      "step": 713
+    },
+    {
+      "epoch": 1.3574144486692015,
+      "grad_norm": 0.6680042126047062,
+      "learning_rate": 5.665437450875534e-06,
+      "loss": 0.3159,
+      "step": 714
+    },
+    {
+      "epoch": 1.3593155893536122,
+      "grad_norm": 0.7013960685257498,
+      "learning_rate": 5.635534097740435e-06,
+      "loss": 0.3097,
+      "step": 715
+    },
+    {
+      "epoch": 1.3612167300380227,
+      "grad_norm": 0.6858063864716573,
+      "learning_rate": 5.605678878186911e-06,
+      "loss": 0.299,
+      "step": 716
+    },
+    {
+      "epoch": 1.3631178707224334,
+      "grad_norm": 0.6660006620912544,
+      "learning_rate": 5.575872121473722e-06,
+      "loss": 0.2884,
+      "step": 717
+    },
+    {
+      "epoch": 1.3650190114068441,
+      "grad_norm": 0.7111085662932116,
+      "learning_rate": 5.546114156325166e-06,
+      "loss": 0.2977,
+      "step": 718
+    },
+    {
+      "epoch": 1.3669201520912546,
+      "grad_norm": 0.7045185991016167,
+      "learning_rate": 5.516405310927431e-06,
+      "loss": 0.3096,
+      "step": 719
+    },
+    {
+      "epoch": 1.3688212927756653,
+      "grad_norm": 0.7318016872130918,
+      "learning_rate": 5.4867459129249846e-06,
+      "loss": 0.2906,
+      "step": 720
+    },
+    {
+      "epoch": 1.370722433460076,
+      "grad_norm": 0.7044531807506517,
+      "learning_rate": 5.4571362894169795e-06,
+      "loss": 0.3125,
+      "step": 721
+    },
+    {
+      "epoch": 1.3726235741444868,
+      "grad_norm": 0.6812403352558687,
+      "learning_rate": 5.427576766953615e-06,
+      "loss": 0.3058,
+      "step": 722
+    },
+    {
+      "epoch": 1.3745247148288973,
+      "grad_norm": 0.6737933644525289,
+      "learning_rate": 5.398067671532554e-06,
+      "loss": 0.2997,
+      "step": 723
+    },
+    {
+      "epoch": 1.376425855513308,
+      "grad_norm": 0.6646475624269411,
+      "learning_rate": 5.368609328595323e-06,
+      "loss": 0.2922,
+      "step": 724
+    },
+    {
+      "epoch": 1.3783269961977187,
+      "grad_norm": 0.6599470736244435,
+      "learning_rate": 5.339202063023727e-06,
+      "loss": 0.2917,
+      "step": 725
+    },
+    {
+      "epoch": 1.3802281368821292,
+      "grad_norm": 0.674590805947845,
+      "learning_rate": 5.309846199136258e-06,
+      "loss": 0.2988,
+      "step": 726
+    },
+    {
+      "epoch": 1.38212927756654,
+      "grad_norm": 0.680596592970524,
+      "learning_rate": 5.280542060684535e-06,
+      "loss": 0.3124,
+      "step": 727
+    },
+    {
+      "epoch": 1.3840304182509506,
+      "grad_norm": 0.6627100491614774,
+      "learning_rate": 5.2512899708497086e-06,
+      "loss": 0.2921,
+      "step": 728
+    },
+    {
+      "epoch": 1.3859315589353614,
+      "grad_norm": 0.6352255167236072,
+      "learning_rate": 5.222090252238916e-06,
+      "loss": 0.3013,
+      "step": 729
+    },
+    {
+      "epoch": 1.3878326996197718,
+      "grad_norm": 0.6455194781673255,
+      "learning_rate": 5.192943226881724e-06,
+      "loss": 0.2743,
+      "step": 730
+    },
+    {
+      "epoch": 1.3897338403041826,
+      "grad_norm": 0.7205739652808391,
+      "learning_rate": 5.163849216226562e-06,
+      "loss": 0.3035,
+      "step": 731
+    },
+    {
+      "epoch": 1.3916349809885933,
+      "grad_norm": 0.6869360042375092,
+      "learning_rate": 5.134808541137183e-06,
+      "loss": 0.3206,
+      "step": 732
+    },
+    {
+      "epoch": 1.3935361216730038,
+      "grad_norm": 0.6632078708595743,
+      "learning_rate": 5.105821521889147e-06,
+      "loss": 0.3049,
+      "step": 733
+    },
+    {
+      "epoch": 1.3954372623574145,
+      "grad_norm": 0.6552948058318698,
+      "learning_rate": 5.076888478166247e-06,
+      "loss": 0.3045,
+      "step": 734
+    },
+    {
+      "epoch": 1.3973384030418252,
+      "grad_norm": 0.6944385942040264,
+      "learning_rate": 5.048009729057012e-06,
+      "loss": 0.3086,
+      "step": 735
+    },
+    {
+      "epoch": 1.3992395437262357,
+      "grad_norm": 0.6673920862692374,
+      "learning_rate": 5.0191855930511946e-06,
+      "loss": 0.2956,
+      "step": 736
+    },
+    {
+      "epoch": 1.4011406844106464,
+      "grad_norm": 0.6736016378688329,
+      "learning_rate": 4.990416388036233e-06,
+      "loss": 0.2867,
+      "step": 737
+    },
+    {
+      "epoch": 1.4030418250950571,
+      "grad_norm": 0.6569022097008016,
+      "learning_rate": 4.961702431293759e-06,
+      "loss": 0.2929,
+      "step": 738
+    },
+    {
+      "epoch": 1.4049429657794676,
+      "grad_norm": 0.6974029235939001,
+      "learning_rate": 4.933044039496107e-06,
+      "loss": 0.3042,
+      "step": 739
+    },
+    {
+      "epoch": 1.4068441064638784,
+      "grad_norm": 0.6737023052024314,
+      "learning_rate": 4.904441528702806e-06,
+      "loss": 0.298,
+      "step": 740
+    },
+    {
+      "epoch": 1.408745247148289,
+      "grad_norm": 0.7098671155105646,
+      "learning_rate": 4.875895214357093e-06,
+      "loss": 0.3048,
+      "step": 741
+    },
+    {
+      "epoch": 1.4106463878326996,
+      "grad_norm": 0.7551463219000628,
+      "learning_rate": 4.847405411282462e-06,
+      "loss": 0.3228,
+      "step": 742
+    },
+    {
+      "epoch": 1.4125475285171103,
+      "grad_norm": 0.7110332500993938,
+      "learning_rate": 4.818972433679145e-06,
+      "loss": 0.3019,
+      "step": 743
+    },
+    {
+      "epoch": 1.414448669201521,
+      "grad_norm": 0.6868072740970074,
+      "learning_rate": 4.790596595120699e-06,
+      "loss": 0.3017,
+      "step": 744
+    },
+    {
+      "epoch": 1.4163498098859315,
+      "grad_norm": 0.6799365348753907,
+      "learning_rate": 4.762278208550505e-06,
+      "loss": 0.282,
+      "step": 745
+    },
+    {
+      "epoch": 1.4182509505703422,
+      "grad_norm": 0.6727516354009891,
+      "learning_rate": 4.734017586278337e-06,
+      "loss": 0.2883,
+      "step": 746
+    },
+    {
+      "epoch": 1.420152091254753,
+      "grad_norm": 0.6690305179126783,
+      "learning_rate": 4.7058150399769245e-06,
+      "loss": 0.3089,
+      "step": 747
+    },
+    {
+      "epoch": 1.4220532319391634,
+      "grad_norm": 0.698330626264572,
+      "learning_rate": 4.677670880678493e-06,
+      "loss": 0.3058,
+      "step": 748
+    },
+    {
+      "epoch": 1.4239543726235742,
+      "grad_norm": 0.6764178900019842,
+      "learning_rate": 4.649585418771348e-06,
+      "loss": 0.2986,
+      "step": 749
+    },
+    {
+      "epoch": 1.4258555133079849,
+      "grad_norm": 0.7309181910001722,
+      "learning_rate": 4.621558963996458e-06,
+      "loss": 0.3128,
+      "step": 750
+    },
+    {
+      "epoch": 1.4277566539923954,
+      "grad_norm": 0.6892096283537845,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.2892,
+      "step": 751
+    },
+    {
+      "epoch": 1.429657794676806,
+      "grad_norm": 0.7013554581585089,
+      "learning_rate": 4.565684311550077e-06,
+      "loss": 0.3125,
+      "step": 752
+    },
+    {
+      "epoch": 1.4315589353612168,
+      "grad_norm": 0.6965843750129259,
+      "learning_rate": 4.537836730093077e-06,
+      "loss": 0.3051,
+      "step": 753
+    },
+    {
+      "epoch": 1.4334600760456273,
+      "grad_norm": 0.6936550848331486,
+      "learning_rate": 4.510049388190518e-06,
+      "loss": 0.3142,
+      "step": 754
+    },
+    {
+      "epoch": 1.435361216730038,
+      "grad_norm": 0.6802532387815458,
+      "learning_rate": 4.482322592295541e-06,
+      "loss": 0.2968,
+      "step": 755
+    },
+    {
+      "epoch": 1.4372623574144487,
+      "grad_norm": 0.6363475710614847,
+      "learning_rate": 4.454656648193559e-06,
+      "loss": 0.2801,
+      "step": 756
+    },
+    {
+      "epoch": 1.4391634980988592,
+      "grad_norm": 0.651651589298049,
+      "learning_rate": 4.427051860998877e-06,
+      "loss": 0.2925,
+      "step": 757
+    },
+    {
+      "epoch": 1.44106463878327,
+      "grad_norm": 0.6507707958385838,
+      "learning_rate": 4.399508535151321e-06,
+      "loss": 0.2916,
+      "step": 758
+    },
+    {
+      "epoch": 1.4429657794676807,
+      "grad_norm": 0.6356978572956105,
+      "learning_rate": 4.372026974412907e-06,
+      "loss": 0.2769,
+      "step": 759
+    },
+    {
+      "epoch": 1.4448669201520912,
+      "grad_norm": 0.7073288471632508,
+      "learning_rate": 4.344607481864466e-06,
+      "loss": 0.2985,
+      "step": 760
+    },
+    {
+      "epoch": 1.446768060836502,
+      "grad_norm": 0.6869649220094196,
+      "learning_rate": 4.317250359902295e-06,
+      "loss": 0.2903,
+      "step": 761
+    },
+    {
+      "epoch": 1.4486692015209126,
+      "grad_norm": 0.6657911325135003,
+      "learning_rate": 4.2899559102348585e-06,
+      "loss": 0.3062,
+      "step": 762
+    },
+    {
+      "epoch": 1.450570342205323,
+      "grad_norm": 0.6659994951630063,
+      "learning_rate": 4.262724433879427e-06,
+      "loss": 0.2942,
+      "step": 763
+    },
+    {
+      "epoch": 1.4524714828897338,
+      "grad_norm": 0.6754738828912081,
+      "learning_rate": 4.235556231158765e-06,
+      "loss": 0.2986,
+      "step": 764
+    },
+    {
+      "epoch": 1.4543726235741445,
+      "grad_norm": 0.6601453059808033,
+      "learning_rate": 4.208451601697836e-06,
+      "loss": 0.2901,
+      "step": 765
+    },
+    {
+      "epoch": 1.456273764258555,
+      "grad_norm": 0.6760829403247115,
+      "learning_rate": 4.181410844420473e-06,
+      "loss": 0.3081,
+      "step": 766
+    },
+    {
+      "epoch": 1.4581749049429658,
+      "grad_norm": 0.6593896610789374,
+      "learning_rate": 4.154434257546095e-06,
+      "loss": 0.2777,
+      "step": 767
+    },
+    {
+      "epoch": 1.4600760456273765,
+      "grad_norm": 0.6288291470875572,
+      "learning_rate": 4.127522138586424e-06,
+      "loss": 0.2926,
+      "step": 768
+    },
+    {
+      "epoch": 1.461977186311787,
+      "grad_norm": 0.6543872832090505,
+      "learning_rate": 4.10067478434219e-06,
+      "loss": 0.2901,
+      "step": 769
+    },
+    {
+      "epoch": 1.4638783269961977,
+      "grad_norm": 0.6722663921711005,
+      "learning_rate": 4.073892490899865e-06,
+      "loss": 0.3072,
+      "step": 770
+    },
+    {
+      "epoch": 1.4657794676806084,
+      "grad_norm": 0.656876924800257,
+      "learning_rate": 4.047175553628397e-06,
+      "loss": 0.2852,
+      "step": 771
+    },
+    {
+      "epoch": 1.467680608365019,
+      "grad_norm": 0.7043130056565048,
+      "learning_rate": 4.020524267175954e-06,
+      "loss": 0.3058,
+      "step": 772
+    },
+    {
+      "epoch": 1.4695817490494296,
+      "grad_norm": 0.6475618795822702,
+      "learning_rate": 3.993938925466674e-06,
+      "loss": 0.2867,
+      "step": 773
+    },
+    {
+      "epoch": 1.4714828897338403,
+      "grad_norm": 0.6602995753053854,
+      "learning_rate": 3.96741982169742e-06,
+      "loss": 0.3096,
+      "step": 774
+    },
+    {
+      "epoch": 1.4733840304182508,
+      "grad_norm": 0.6767831322273554,
+      "learning_rate": 3.9409672483345465e-06,
+      "loss": 0.2976,
+      "step": 775
+    },
+    {
+      "epoch": 1.4752851711026616,
+      "grad_norm": 0.6838179944513524,
+      "learning_rate": 3.914581497110684e-06,
+      "loss": 0.2947,
+      "step": 776
+    },
+    {
+      "epoch": 1.4771863117870723,
+      "grad_norm": 0.6443057908177426,
+      "learning_rate": 3.888262859021508e-06,
+      "loss": 0.2827,
+      "step": 777
+    },
+    {
+      "epoch": 1.4790874524714828,
+      "grad_norm": 0.6598559562084777,
+      "learning_rate": 3.862011624322534e-06,
+      "loss": 0.2851,
+      "step": 778
+    },
+    {
+      "epoch": 1.4809885931558935,
+      "grad_norm": 0.7425767378724061,
+      "learning_rate": 3.835828082525925e-06,
+      "loss": 0.3177,
+      "step": 779
+    },
+    {
+      "epoch": 1.4828897338403042,
+      "grad_norm": 0.7186686647540542,
+      "learning_rate": 3.8097125223972864e-06,
+      "loss": 0.3098,
+      "step": 780
+    },
+    {
+      "epoch": 1.4847908745247147,
+      "grad_norm": 0.6669516224564478,
+      "learning_rate": 3.7836652319524835e-06,
+      "loss": 0.2963,
+      "step": 781
+    },
+    {
+      "epoch": 1.4866920152091254,
+      "grad_norm": 0.6829923670452973,
+      "learning_rate": 3.7576864984544814e-06,
+      "loss": 0.2998,
+      "step": 782
+    },
+    {
+      "epoch": 1.4885931558935361,
+      "grad_norm": 0.6696468535573125,
+      "learning_rate": 3.73177660841015e-06,
+      "loss": 0.3164,
+      "step": 783
+    },
+    {
+      "epoch": 1.4904942965779466,
+      "grad_norm": 0.6923355890043177,
+      "learning_rate": 3.7059358475671225e-06,
+      "loss": 0.3058,
+      "step": 784
+    },
+    {
+      "epoch": 1.4923954372623573,
+      "grad_norm": 0.6755089757832773,
+      "learning_rate": 3.680164500910646e-06,
+      "loss": 0.2946,
+      "step": 785
+    },
+    {
+      "epoch": 1.494296577946768,
+      "grad_norm": 0.65335473021462,
+      "learning_rate": 3.654462852660423e-06,
+      "loss": 0.3019,
+      "step": 786
+    },
+    {
+      "epoch": 1.4961977186311788,
+      "grad_norm": 0.6611878368310804,
+      "learning_rate": 3.6288311862674885e-06,
+      "loss": 0.3132,
+      "step": 787
+    },
+    {
+      "epoch": 1.4980988593155893,
+      "grad_norm": 0.6847735908887931,
+      "learning_rate": 3.6032697844110896e-06,
+      "loss": 0.3102,
+      "step": 788
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6569314403707336,
+      "learning_rate": 3.5777789289955454e-06,
+      "loss": 0.2974,
+      "step": 789
+    },
+    {
+      "epoch": 1.5019011406844105,
+      "grad_norm": 0.631725287566082,
+      "learning_rate": 3.5523589011471592e-06,
+      "loss": 0.2843,
+      "step": 790
+    },
+    {
+      "epoch": 1.5038022813688214,
+      "grad_norm": 0.710296725896784,
+      "learning_rate": 3.527009981211119e-06,
+      "loss": 0.3057,
+      "step": 791
+    },
+    {
+      "epoch": 1.505703422053232,
+      "grad_norm": 0.620032196175074,
+      "learning_rate": 3.5017324487483873e-06,
+      "loss": 0.2923,
+      "step": 792
+    },
+    {
+      "epoch": 1.5076045627376424,
+      "grad_norm": 0.6775969477208345,
+      "learning_rate": 3.47652658253263e-06,
+      "loss": 0.2997,
+      "step": 793
+    },
+    {
+      "epoch": 1.5095057034220534,
+      "grad_norm": 0.698601542811014,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.2876,
+      "step": 794
+    },
+    {
+      "epoch": 1.5114068441064639,
+      "grad_norm": 0.6385192145495687,
+      "learning_rate": 3.4263309599818017e-06,
+      "loss": 0.2902,
+      "step": 795
+    },
+    {
+      "epoch": 1.5133079847908744,
+      "grad_norm": 0.6717579502587994,
+      "learning_rate": 3.4013417572299446e-06,
+      "loss": 0.3079,
+      "step": 796
+    },
+    {
+      "epoch": 1.5152091254752853,
+      "grad_norm": 0.6878665125603728,
+      "learning_rate": 3.37642532788541e-06,
+      "loss": 0.2838,
+      "step": 797
+    },
+    {
+      "epoch": 1.5171102661596958,
+      "grad_norm": 0.6991043184966474,
+      "learning_rate": 3.3515819467394184e-06,
+      "loss": 0.3018,
+      "step": 798
+    },
+    {
+      "epoch": 1.5190114068441065,
+      "grad_norm": 0.6605683028376355,
+      "learning_rate": 3.326811887777607e-06,
+      "loss": 0.2922,
+      "step": 799
+    },
+    {
+      "epoch": 1.5209125475285172,
+      "grad_norm": 0.6713783171015599,
+      "learning_rate": 3.3021154241769606e-06,
+      "loss": 0.2922,
+      "step": 800
+    },
+    {
+      "epoch": 1.5228136882129277,
+      "grad_norm": 0.6961215941076575,
+      "learning_rate": 3.2774928283028153e-06,
+      "loss": 0.3003,
+      "step": 801
+    },
+    {
+      "epoch": 1.5247148288973384,
+      "grad_norm": 0.7261312237805675,
+      "learning_rate": 3.2529443717058693e-06,
+      "loss": 0.3139,
+      "step": 802
+    },
+    {
+      "epoch": 1.5266159695817492,
+      "grad_norm": 0.6811533119895746,
+      "learning_rate": 3.228470325119164e-06,
+      "loss": 0.2959,
+      "step": 803
+    },
+    {
+      "epoch": 1.5285171102661597,
+      "grad_norm": 0.6733949937642703,
+      "learning_rate": 3.20407095845511e-06,
+      "loss": 0.2884,
+      "step": 804
+    },
+    {
+      "epoch": 1.5304182509505704,
+      "grad_norm": 0.7406888020469802,
+      "learning_rate": 3.179746540802506e-06,
+      "loss": 0.3179,
+      "step": 805
+    },
+    {
+      "epoch": 1.532319391634981,
+      "grad_norm": 0.6168063745803833,
+      "learning_rate": 3.155497340423588e-06,
+      "loss": 0.2903,
+      "step": 806
+    },
+    {
+      "epoch": 1.5342205323193916,
+      "grad_norm": 0.6990251509069143,
+      "learning_rate": 3.1313236247510414e-06,
+      "loss": 0.2953,
+      "step": 807
+    },
+    {
+      "epoch": 1.5361216730038023,
+      "grad_norm": 0.7108610419180795,
+      "learning_rate": 3.107225660385077e-06,
+      "loss": 0.3111,
+      "step": 808
+    },
+    {
+      "epoch": 1.538022813688213,
+      "grad_norm": 0.6734569033713736,
+      "learning_rate": 3.0832037130904748e-06,
+      "loss": 0.2997,
+      "step": 809
+    },
+    {
+      "epoch": 1.5399239543726235,
+      "grad_norm": 0.6827586324520255,
+      "learning_rate": 3.0592580477936606e-06,
+      "loss": 0.2848,
+      "step": 810
+    },
+    {
+      "epoch": 1.5418250950570342,
+      "grad_norm": 0.7006777021679021,
+      "learning_rate": 3.035388928579792e-06,
+      "loss": 0.3143,
+      "step": 811
+    },
+    {
+      "epoch": 1.543726235741445,
+      "grad_norm": 0.6710320406616938,
+      "learning_rate": 3.011596618689825e-06,
+      "loss": 0.2975,
+      "step": 812
+    },
+    {
+      "epoch": 1.5456273764258555,
+      "grad_norm": 0.6692722919698081,
+      "learning_rate": 2.9878813805176252e-06,
+      "loss": 0.2874,
+      "step": 813
+    },
+    {
+      "epoch": 1.5475285171102662,
+      "grad_norm": 0.6628634074075862,
+      "learning_rate": 2.9642434756070793e-06,
+      "loss": 0.2889,
+      "step": 814
+    },
+    {
+      "epoch": 1.549429657794677,
+      "grad_norm": 0.6750613208415934,
+      "learning_rate": 2.940683164649194e-06,
+      "loss": 0.3017,
+      "step": 815
+    },
+    {
+      "epoch": 1.5513307984790874,
+      "grad_norm": 0.6840567854552813,
+      "learning_rate": 2.9172007074792342e-06,
+      "loss": 0.2911,
+      "step": 816
+    },
+    {
+      "epoch": 1.553231939163498,
+      "grad_norm": 0.6922672699589506,
+      "learning_rate": 2.8937963630738517e-06,
+      "loss": 0.3181,
+      "step": 817
+    },
+    {
+      "epoch": 1.5551330798479088,
+      "grad_norm": 0.6803485583075909,
+      "learning_rate": 2.87047038954823e-06,
+      "loss": 0.3034,
+      "step": 818
+    },
+    {
+      "epoch": 1.5570342205323193,
+      "grad_norm": 0.7002731653220754,
+      "learning_rate": 2.8472230441532365e-06,
+      "loss": 0.3071,
+      "step": 819
+    },
+    {
+      "epoch": 1.55893536121673,
+      "grad_norm": 0.6601907776791074,
+      "learning_rate": 2.8240545832725963e-06,
+      "loss": 0.2935,
+      "step": 820
+    },
+    {
+      "epoch": 1.5608365019011408,
+      "grad_norm": 0.6629669814912943,
+      "learning_rate": 2.8009652624200436e-06,
+      "loss": 0.3043,
+      "step": 821
+    },
+    {
+      "epoch": 1.5627376425855513,
+      "grad_norm": 0.6463670252050109,
+      "learning_rate": 2.7779553362365184e-06,
+      "loss": 0.2847,
+      "step": 822
+    },
+    {
+      "epoch": 1.564638783269962,
+      "grad_norm": 0.6880279741910543,
+      "learning_rate": 2.755025058487364e-06,
+      "loss": 0.305,
+      "step": 823
+    },
+    {
+      "epoch": 1.5665399239543727,
+      "grad_norm": 0.6969330350619346,
+      "learning_rate": 2.7321746820595084e-06,
+      "loss": 0.3061,
+      "step": 824
+    },
+    {
+      "epoch": 1.5684410646387832,
+      "grad_norm": 0.6808094529073682,
+      "learning_rate": 2.709404458958693e-06,
+      "loss": 0.3003,
+      "step": 825
+    },
+    {
+      "epoch": 1.570342205323194,
+      "grad_norm": 0.6737208492250114,
+      "learning_rate": 2.6867146403066833e-06,
+      "loss": 0.2825,
+      "step": 826
+    },
+    {
+      "epoch": 1.5722433460076046,
+      "grad_norm": 0.6594071937774411,
+      "learning_rate": 2.6641054763385044e-06,
+      "loss": 0.29,
+      "step": 827
+    },
+    {
+      "epoch": 1.5741444866920151,
+      "grad_norm": 0.7074877218570444,
+      "learning_rate": 2.6415772163996845e-06,
+      "loss": 0.2986,
+      "step": 828
+    },
+    {
+      "epoch": 1.5760456273764258,
+      "grad_norm": 0.6719412255934047,
+      "learning_rate": 2.619130108943494e-06,
+      "loss": 0.2929,
+      "step": 829
+    },
+    {
+      "epoch": 1.5779467680608366,
+      "grad_norm": 0.6644759135068584,
+      "learning_rate": 2.5967644015282146e-06,
+      "loss": 0.2795,
+      "step": 830
+    },
+    {
+      "epoch": 1.579847908745247,
+      "grad_norm": 0.6520706337473319,
+      "learning_rate": 2.5744803408144026e-06,
+      "loss": 0.2898,
+      "step": 831
+    },
+    {
+      "epoch": 1.5817490494296578,
+      "grad_norm": 0.6788959927692927,
+      "learning_rate": 2.5522781725621814e-06,
+      "loss": 0.3115,
+      "step": 832
+    },
+    {
+      "epoch": 1.5836501901140685,
+      "grad_norm": 0.6702927305597481,
+      "learning_rate": 2.530158141628515e-06,
+      "loss": 0.296,
+      "step": 833
+    },
+    {
+      "epoch": 1.585551330798479,
+      "grad_norm": 0.7437327351065274,
+      "learning_rate": 2.508120491964512e-06,
+      "loss": 0.3055,
+      "step": 834
+    },
+    {
+      "epoch": 1.5874524714828897,
+      "grad_norm": 0.7556826554440722,
+      "learning_rate": 2.486165466612751e-06,
+      "loss": 0.2898,
+      "step": 835
+    },
+    {
+      "epoch": 1.5893536121673004,
+      "grad_norm": 0.6661383592504586,
+      "learning_rate": 2.464293307704566e-06,
+      "loss": 0.2994,
+      "step": 836
+    },
+    {
+      "epoch": 1.591254752851711,
+      "grad_norm": 0.6997398646799898,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.2925,
+      "step": 837
+    },
+    {
+      "epoch": 1.5931558935361216,
+      "grad_norm": 0.653863924304275,
+      "learning_rate": 2.4207985531722034e-06,
+      "loss": 0.2879,
+      "step": 838
+    },
+    {
+      "epoch": 1.5950570342205324,
+      "grad_norm": 0.6474604053826938,
+      "learning_rate": 2.3991764372306113e-06,
+      "loss": 0.2823,
+      "step": 839
+    },
+    {
+      "epoch": 1.5969581749049429,
+      "grad_norm": 0.6187710954259721,
+      "learning_rate": 2.377638147092497e-06,
+      "loss": 0.2692,
+      "step": 840
+    },
+    {
+      "epoch": 1.5988593155893536,
+      "grad_norm": 0.7048352726288213,
+      "learning_rate": 2.3561839202932344e-06,
+      "loss": 0.3068,
+      "step": 841
+    },
+    {
+      "epoch": 1.6007604562737643,
+      "grad_norm": 0.6841855158333704,
+      "learning_rate": 2.3348139934411008e-06,
+      "loss": 0.299,
+      "step": 842
+    },
+    {
+      "epoch": 1.6026615969581748,
+      "grad_norm": 0.6551472138877459,
+      "learning_rate": 2.3135286022146785e-06,
+      "loss": 0.289,
+      "step": 843
+    },
+    {
+      "epoch": 1.6045627376425855,
+      "grad_norm": 0.6121229182829925,
+      "learning_rate": 2.292327981360245e-06,
+      "loss": 0.2726,
+      "step": 844
+    },
+    {
+      "epoch": 1.6064638783269962,
+      "grad_norm": 0.6717503404129097,
+      "learning_rate": 2.271212364689176e-06,
+      "loss": 0.2857,
+      "step": 845
+    },
+    {
+      "epoch": 1.6083650190114067,
+      "grad_norm": 0.6845008148291871,
+      "learning_rate": 2.2501819850753925e-06,
+      "loss": 0.3118,
+      "step": 846
+    },
+    {
+      "epoch": 1.6102661596958177,
+      "grad_norm": 0.6728832747385453,
+      "learning_rate": 2.229237074452768e-06,
+      "loss": 0.2884,
+      "step": 847
+    },
+    {
+      "epoch": 1.6121673003802282,
+      "grad_norm": 0.6545234838769938,
+      "learning_rate": 2.2083778638125796e-06,
+      "loss": 0.2913,
+      "step": 848
+    },
+    {
+      "epoch": 1.6140684410646386,
+      "grad_norm": 0.7328509474713253,
+      "learning_rate": 2.1876045832009694e-06,
+      "loss": 0.2927,
+      "step": 849
+    },
+    {
+      "epoch": 1.6159695817490496,
+      "grad_norm": 0.6740897609555396,
+      "learning_rate": 2.16691746171639e-06,
+      "loss": 0.3027,
+      "step": 850
+    },
+    {
+      "epoch": 1.61787072243346,
+      "grad_norm": 0.6364767429850742,
+      "learning_rate": 2.1463167275070863e-06,
+      "loss": 0.2981,
+      "step": 851
+    },
+    {
+      "epoch": 1.6197718631178706,
+      "grad_norm": 0.6301750981029648,
+      "learning_rate": 2.125802607768588e-06,
+      "loss": 0.2906,
+      "step": 852
+    },
+    {
+      "epoch": 1.6216730038022815,
+      "grad_norm": 0.6617620371058369,
+      "learning_rate": 2.1053753287411895e-06,
+      "loss": 0.2759,
+      "step": 853
+    },
+    {
+      "epoch": 1.623574144486692,
+      "grad_norm": 0.6854896364710342,
+      "learning_rate": 2.08503511570746e-06,
+      "loss": 0.2887,
+      "step": 854
+    },
+    {
+      "epoch": 1.6254752851711025,
+      "grad_norm": 0.6473013079627202,
+      "learning_rate": 2.064782192989765e-06,
+      "loss": 0.283,
+      "step": 855
+    },
+    {
+      "epoch": 1.6273764258555135,
+      "grad_norm": 0.6806669656816828,
+      "learning_rate": 2.0446167839477815e-06,
+      "loss": 0.2926,
+      "step": 856
+    },
+    {
+      "epoch": 1.629277566539924,
+      "grad_norm": 0.6866118946871398,
+      "learning_rate": 2.0245391109760437e-06,
+      "loss": 0.3008,
+      "step": 857
+    },
+    {
+      "epoch": 1.6311787072243344,
+      "grad_norm": 0.67856619010215,
+      "learning_rate": 2.0045493955014915e-06,
+      "loss": 0.2977,
+      "step": 858
+    },
+    {
+      "epoch": 1.6330798479087454,
+      "grad_norm": 0.6556859224996296,
+      "learning_rate": 1.984647857981017e-06,
+      "loss": 0.3098,
+      "step": 859
+    },
+    {
+      "epoch": 1.6349809885931559,
+      "grad_norm": 0.6409500452068347,
+      "learning_rate": 1.96483471789904e-06,
+      "loss": 0.3001,
+      "step": 860
+    },
+    {
+      "epoch": 1.6368821292775664,
+      "grad_norm": 0.6634371980140289,
+      "learning_rate": 1.9451101937650963e-06,
+      "loss": 0.3004,
+      "step": 861
+    },
+    {
+      "epoch": 1.6387832699619773,
+      "grad_norm": 0.7255107785611558,
+      "learning_rate": 1.925474503111412e-06,
+      "loss": 0.2892,
+      "step": 862
+    },
+    {
+      "epoch": 1.6406844106463878,
+      "grad_norm": 0.6976679974024178,
+      "learning_rate": 1.905927862490512e-06,
+      "loss": 0.2971,
+      "step": 863
+    },
+    {
+      "epoch": 1.6425855513307985,
+      "grad_norm": 0.6425306078492332,
+      "learning_rate": 1.8864704874728346e-06,
+      "loss": 0.2755,
+      "step": 864
+    },
+    {
+      "epoch": 1.6444866920152093,
+      "grad_norm": 0.6744738897005956,
+      "learning_rate": 1.8671025926443464e-06,
+      "loss": 0.3152,
+      "step": 865
+    },
+    {
+      "epoch": 1.6463878326996197,
+      "grad_norm": 0.6746392434330984,
+      "learning_rate": 1.8478243916041882e-06,
+      "loss": 0.2913,
+      "step": 866
+    },
+    {
+      "epoch": 1.6482889733840305,
+      "grad_norm": 0.6628698315654836,
+      "learning_rate": 1.828636096962304e-06,
+      "loss": 0.2946,
+      "step": 867
+    },
+    {
+      "epoch": 1.6501901140684412,
+      "grad_norm": 0.673605637229709,
+      "learning_rate": 1.8095379203371044e-06,
+      "loss": 0.2887,
+      "step": 868
+    },
+    {
+      "epoch": 1.6520912547528517,
+      "grad_norm": 0.6622912510981255,
+      "learning_rate": 1.7905300723531393e-06,
+      "loss": 0.2869,
+      "step": 869
+    },
+    {
+      "epoch": 1.6539923954372624,
+      "grad_norm": 0.6791673894966933,
+      "learning_rate": 1.771612762638758e-06,
+      "loss": 0.3,
+      "step": 870
+    },
+    {
+      "epoch": 1.6558935361216731,
+      "grad_norm": 0.63786704213842,
+      "learning_rate": 1.7527861998238094e-06,
+      "loss": 0.287,
+      "step": 871
+    },
+    {
+      "epoch": 1.6577946768060836,
+      "grad_norm": 0.6615590186861462,
+      "learning_rate": 1.7340505915373495e-06,
+      "loss": 0.2787,
+      "step": 872
+    },
+    {
+      "epoch": 1.6596958174904943,
+      "grad_norm": 0.6691075234738113,
+      "learning_rate": 1.7154061444053239e-06,
+      "loss": 0.2933,
+      "step": 873
+    },
+    {
+      "epoch": 1.661596958174905,
+      "grad_norm": 0.6295231269322634,
+      "learning_rate": 1.6968530640483126e-06,
+      "loss": 0.2764,
+      "step": 874
+    },
+    {
+      "epoch": 1.6634980988593155,
+      "grad_norm": 0.6542264579941419,
+      "learning_rate": 1.6783915550792652e-06,
+      "loss": 0.2793,
+      "step": 875
+    },
+    {
+      "epoch": 1.6653992395437263,
+      "grad_norm": 0.6465369949087939,
+      "learning_rate": 1.660021821101222e-06,
+      "loss": 0.2765,
+      "step": 876
+    },
+    {
+      "epoch": 1.667300380228137,
+      "grad_norm": 0.6336787863229356,
+      "learning_rate": 1.6417440647050853e-06,
+      "loss": 0.2836,
+      "step": 877
+    },
+    {
+      "epoch": 1.6692015209125475,
+      "grad_norm": 0.6608679969098998,
+      "learning_rate": 1.6235584874673848e-06,
+      "loss": 0.2913,
+      "step": 878
+    },
+    {
+      "epoch": 1.6711026615969582,
+      "grad_norm": 0.637880164433468,
+      "learning_rate": 1.6054652899480472e-06,
+      "loss": 0.272,
+      "step": 879
+    },
+    {
+      "epoch": 1.673003802281369,
+      "grad_norm": 0.6194905037857826,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.2899,
+      "step": 880
+    },
+    {
+      "epoch": 1.6749049429657794,
+      "grad_norm": 0.6160578342244919,
+      "learning_rate": 1.5695568312079156e-06,
+      "loss": 0.2892,
+      "step": 881
+    },
+    {
+      "epoch": 1.6768060836501901,
+      "grad_norm": 0.6581157701118096,
+      "learning_rate": 1.5517419660041277e-06,
+      "loss": 0.2985,
+      "step": 882
+    },
+    {
+      "epoch": 1.6787072243346008,
+      "grad_norm": 0.6743151055930919,
+      "learning_rate": 1.534020272548349e-06,
+      "loss": 0.2979,
+      "step": 883
+    },
+    {
+      "epoch": 1.6806083650190113,
+      "grad_norm": 0.7042648135024562,
+      "learning_rate": 1.5163919462845622e-06,
+      "loss": 0.3059,
+      "step": 884
+    },
+    {
+      "epoch": 1.682509505703422,
+      "grad_norm": 0.7045196914184306,
+      "learning_rate": 1.4988571816270402e-06,
+      "loss": 0.3138,
+      "step": 885
+    },
+    {
+      "epoch": 1.6844106463878328,
+      "grad_norm": 0.655691564050441,
+      "learning_rate": 1.4814161719582132e-06,
+      "loss": 0.2795,
+      "step": 886
+    },
+    {
+      "epoch": 1.6863117870722433,
+      "grad_norm": 0.6597369568759405,
+      "learning_rate": 1.4640691096265358e-06,
+      "loss": 0.2977,
+      "step": 887
+    },
+    {
+      "epoch": 1.688212927756654,
+      "grad_norm": 0.6921664258562352,
+      "learning_rate": 1.4468161859443609e-06,
+      "loss": 0.2984,
+      "step": 888
+    },
+    {
+      "epoch": 1.6901140684410647,
+      "grad_norm": 0.6603090342618605,
+      "learning_rate": 1.4296575911858268e-06,
+      "loss": 0.2963,
+      "step": 889
+    },
+    {
+      "epoch": 1.6920152091254752,
+      "grad_norm": 0.6754310569706267,
+      "learning_rate": 1.412593514584777e-06,
+      "loss": 0.2943,
+      "step": 890
+    },
+    {
+      "epoch": 1.693916349809886,
+      "grad_norm": 0.6770548082276626,
+      "learning_rate": 1.3956241443326423e-06,
+      "loss": 0.3065,
+      "step": 891
+    },
+    {
+      "epoch": 1.6958174904942966,
+      "grad_norm": 0.6845514391697731,
+      "learning_rate": 1.378749667576399e-06,
+      "loss": 0.3245,
+      "step": 892
+    },
+    {
+      "epoch": 1.6977186311787071,
+      "grad_norm": 0.6589640159523908,
+      "learning_rate": 1.3619702704164783e-06,
+      "loss": 0.3023,
+      "step": 893
+    },
+    {
+      "epoch": 1.6996197718631179,
+      "grad_norm": 0.6920678399012251,
+      "learning_rate": 1.3452861379047289e-06,
+      "loss": 0.2959,
+      "step": 894
+    },
+    {
+      "epoch": 1.7015209125475286,
+      "grad_norm": 0.6461485139481378,
+      "learning_rate": 1.3286974540423747e-06,
+      "loss": 0.2865,
+      "step": 895
+    },
+    {
+      "epoch": 1.703422053231939,
+      "grad_norm": 0.6808559296758336,
+      "learning_rate": 1.3122044017779768e-06,
+      "loss": 0.3005,
+      "step": 896
+    },
+    {
+      "epoch": 1.7053231939163498,
+      "grad_norm": 0.6489271104318259,
+      "learning_rate": 1.2958071630054214e-06,
+      "loss": 0.2953,
+      "step": 897
+    },
+    {
+      "epoch": 1.7072243346007605,
+      "grad_norm": 0.6514889206542622,
+      "learning_rate": 1.279505918561923e-06,
+      "loss": 0.3038,
+      "step": 898
+    },
+    {
+      "epoch": 1.709125475285171,
+      "grad_norm": 0.6698184648794909,
+      "learning_rate": 1.2633008482260146e-06,
+      "loss": 0.298,
+      "step": 899
+    },
+    {
+      "epoch": 1.7110266159695817,
+      "grad_norm": 0.7035406840886368,
+      "learning_rate": 1.2471921307155655e-06,
+      "loss": 0.2877,
+      "step": 900
+    },
+    {
+      "epoch": 1.7129277566539924,
+      "grad_norm": 0.6800324843144723,
+      "learning_rate": 1.2311799436858275e-06,
+      "loss": 0.303,
+      "step": 901
+    },
+    {
+      "epoch": 1.714828897338403,
+      "grad_norm": 0.6364746840223059,
+      "learning_rate": 1.2152644637274603e-06,
+      "loss": 0.2856,
+      "step": 902
+    },
+    {
+      "epoch": 1.7167300380228137,
+      "grad_norm": 0.6807974237410163,
+      "learning_rate": 1.1994458663645836e-06,
+      "loss": 0.3031,
+      "step": 903
+    },
+    {
+      "epoch": 1.7186311787072244,
+      "grad_norm": 0.6905991990680014,
+      "learning_rate": 1.1837243260528542e-06,
+      "loss": 0.3014,
+      "step": 904
+    },
+    {
+      "epoch": 1.7205323193916349,
+      "grad_norm": 0.6505491314027649,
+      "learning_rate": 1.168100016177528e-06,
+      "loss": 0.2883,
+      "step": 905
+    },
+    {
+      "epoch": 1.7224334600760456,
+      "grad_norm": 0.6673548341602854,
+      "learning_rate": 1.1525731090515536e-06,
+      "loss": 0.2927,
+      "step": 906
+    },
+    {
+      "epoch": 1.7243346007604563,
+      "grad_norm": 0.7118106905750516,
+      "learning_rate": 1.137143775913675e-06,
+      "loss": 0.2949,
+      "step": 907
+    },
+    {
+      "epoch": 1.7262357414448668,
+      "grad_norm": 0.6431272088355892,
+      "learning_rate": 1.1218121869265365e-06,
+      "loss": 0.2922,
+      "step": 908
+    },
+    {
+      "epoch": 1.7281368821292775,
+      "grad_norm": 0.6118567590545564,
+      "learning_rate": 1.1065785111748117e-06,
+      "loss": 0.275,
+      "step": 909
+    },
+    {
+      "epoch": 1.7300380228136882,
+      "grad_norm": 0.640477713903404,
+      "learning_rate": 1.0914429166633355e-06,
+      "loss": 0.281,
+      "step": 910
+    },
+    {
+      "epoch": 1.7319391634980987,
+      "grad_norm": 0.6681452722326737,
+      "learning_rate": 1.076405570315252e-06,
+      "loss": 0.2991,
+      "step": 911
+    },
+    {
+      "epoch": 1.7338403041825095,
+      "grad_norm": 0.6849634054735122,
+      "learning_rate": 1.0614666379701732e-06,
+      "loss": 0.3004,
+      "step": 912
+    },
+    {
+      "epoch": 1.7357414448669202,
+      "grad_norm": 0.6772662133763292,
+      "learning_rate": 1.046626284382356e-06,
+      "loss": 0.292,
+      "step": 913
+    },
+    {
+      "epoch": 1.7376425855513307,
+      "grad_norm": 0.6725808212144174,
+      "learning_rate": 1.0318846732188737e-06,
+      "loss": 0.294,
+      "step": 914
+    },
+    {
+      "epoch": 1.7395437262357416,
+      "grad_norm": 0.6485066978478725,
+      "learning_rate": 1.017241967057816e-06,
+      "loss": 0.2909,
+      "step": 915
+    },
+    {
+      "epoch": 1.741444866920152,
+      "grad_norm": 0.6625054782730969,
+      "learning_rate": 1.0026983273865055e-06,
+      "loss": 0.2968,
+      "step": 916
+    },
+    {
+      "epoch": 1.7433460076045626,
+      "grad_norm": 0.6469806174440148,
+      "learning_rate": 9.882539145997027e-07,
+      "loss": 0.2892,
+      "step": 917
+    },
+    {
+      "epoch": 1.7452471482889735,
+      "grad_norm": 0.6561459302146309,
+      "learning_rate": 9.739088879978409e-07,
+      "loss": 0.304,
+      "step": 918
+    },
+    {
+      "epoch": 1.747148288973384,
+      "grad_norm": 0.6493021776003873,
+      "learning_rate": 9.59663405785277e-07,
+      "loss": 0.2867,
+      "step": 919
+    },
+    {
+      "epoch": 1.7490494296577945,
+      "grad_norm": 0.7110499619130468,
+      "learning_rate": 9.455176250685338e-07,
+      "loss": 0.3054,
+      "step": 920
+    },
+    {
+      "epoch": 1.7509505703422055,
+      "grad_norm": 0.670800146275233,
+      "learning_rate": 9.314717018545838e-07,
+      "loss": 0.2854,
+      "step": 921
+    },
+    {
+      "epoch": 1.752851711026616,
+      "grad_norm": 0.6713162487633502,
+      "learning_rate": 9.17525791049112e-07,
+      "loss": 0.2982,
+      "step": 922
+    },
+    {
+      "epoch": 1.7547528517110265,
+      "grad_norm": 0.6644745340642867,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.295,
+      "step": 923
+    },
+    {
+      "epoch": 1.7566539923954374,
+      "grad_norm": 0.7166081889739494,
+      "learning_rate": 8.899346207697135e-07,
+      "loss": 0.2838,
+      "step": 924
+    },
+    {
+      "epoch": 1.758555133079848,
+      "grad_norm": 0.683384726931666,
+      "learning_rate": 8.762896655854481e-07,
+      "loss": 0.3203,
+      "step": 925
+    },
+    {
+      "epoch": 1.7604562737642584,
+      "grad_norm": 0.6721039524668614,
+      "learning_rate": 8.627453313856249e-07,
+      "loss": 0.2934,
+      "step": 926
+    },
+    {
+      "epoch": 1.7623574144486693,
+      "grad_norm": 0.6760763839296361,
+      "learning_rate": 8.493017675441495e-07,
+      "loss": 0.3109,
+      "step": 927
+    },
+    {
+      "epoch": 1.7642585551330798,
+      "grad_norm": 0.7052182829307825,
+      "learning_rate": 8.359591223235785e-07,
+      "loss": 0.3167,
+      "step": 928
+    },
+    {
+      "epoch": 1.7661596958174905,
+      "grad_norm": 0.6745551181339163,
+      "learning_rate": 8.227175428734868e-07,
+      "loss": 0.3,
+      "step": 929
+    },
+    {
+      "epoch": 1.7680608365019013,
+      "grad_norm": 0.6946436531792836,
+      "learning_rate": 8.095771752288451e-07,
+      "loss": 0.2908,
+      "step": 930
+    },
+    {
+      "epoch": 1.7699619771863118,
+      "grad_norm": 0.6741742995049886,
+      "learning_rate": 7.965381643084069e-07,
+      "loss": 0.2807,
+      "step": 931
+    },
+    {
+      "epoch": 1.7718631178707225,
+      "grad_norm": 0.6481305639587837,
+      "learning_rate": 7.83600653913108e-07,
+      "loss": 0.3008,
+      "step": 932
+    },
+    {
+      "epoch": 1.7737642585551332,
+      "grad_norm": 0.6237742866741086,
+      "learning_rate": 7.707647867244927e-07,
+      "loss": 0.2881,
+      "step": 933
+    },
+    {
+      "epoch": 1.7756653992395437,
+      "grad_norm": 0.6737915203240475,
+      "learning_rate": 7.580307043031232e-07,
+      "loss": 0.2915,
+      "step": 934
+    },
+    {
+      "epoch": 1.7775665399239544,
+      "grad_norm": 0.6746743637794105,
+      "learning_rate": 7.453985470870284e-07,
+      "loss": 0.2973,
+      "step": 935
+    },
+    {
+      "epoch": 1.7794676806083651,
+      "grad_norm": 0.6486450251398976,
+      "learning_rate": 7.328684543901598e-07,
+      "loss": 0.2877,
+      "step": 936
+    },
+    {
+      "epoch": 1.7813688212927756,
+      "grad_norm": 0.631293029847571,
+      "learning_rate": 7.204405644008416e-07,
+      "loss": 0.2943,
+      "step": 937
+    },
+    {
+      "epoch": 1.7832699619771863,
+      "grad_norm": 0.6614604902048281,
+      "learning_rate": 7.081150141802518e-07,
+      "loss": 0.2964,
+      "step": 938
+    },
+    {
+      "epoch": 1.785171102661597,
+      "grad_norm": 0.6378753065973165,
+      "learning_rate": 6.958919396609231e-07,
+      "loss": 0.2956,
+      "step": 939
+    },
+    {
+      "epoch": 1.7870722433460076,
+      "grad_norm": 0.6747241774492733,
+      "learning_rate": 6.837714756452241e-07,
+      "loss": 0.2915,
+      "step": 940
+    },
+    {
+      "epoch": 1.7889733840304183,
+      "grad_norm": 0.6640195809305564,
+      "learning_rate": 6.717537558038845e-07,
+      "loss": 0.2948,
+      "step": 941
+    },
+    {
+      "epoch": 1.790874524714829,
+      "grad_norm": 0.6608500940226224,
+      "learning_rate": 6.598389126745209e-07,
+      "loss": 0.2967,
+      "step": 942
+    },
+    {
+      "epoch": 1.7927756653992395,
+      "grad_norm": 0.6261259248410527,
+      "learning_rate": 6.480270776601682e-07,
+      "loss": 0.2875,
+      "step": 943
+    },
+    {
+      "epoch": 1.7946768060836502,
+      "grad_norm": 0.6365978890028503,
+      "learning_rate": 6.36318381027835e-07,
+      "loss": 0.2802,
+      "step": 944
+    },
+    {
+      "epoch": 1.796577946768061,
+      "grad_norm": 0.6411372055147753,
+      "learning_rate": 6.247129519070728e-07,
+      "loss": 0.2824,
+      "step": 945
+    },
+    {
+      "epoch": 1.7984790874524714,
+      "grad_norm": 0.6557490512196504,
+      "learning_rate": 6.132109182885382e-07,
+      "loss": 0.2984,
+      "step": 946
+    },
+    {
+      "epoch": 1.8003802281368821,
+      "grad_norm": 0.7483746428838404,
+      "learning_rate": 6.018124070225928e-07,
+      "loss": 0.2833,
+      "step": 947
+    },
+    {
+      "epoch": 1.8022813688212929,
+      "grad_norm": 0.6549818169905346,
+      "learning_rate": 5.905175438178979e-07,
+      "loss": 0.2945,
+      "step": 948
+    },
+    {
+      "epoch": 1.8041825095057034,
+      "grad_norm": 0.6500712324988541,
+      "learning_rate": 5.793264532400311e-07,
+      "loss": 0.3028,
+      "step": 949
+    },
+    {
+      "epoch": 1.806083650190114,
+      "grad_norm": 0.666840783682001,
+      "learning_rate": 5.68239258710116e-07,
+      "loss": 0.2852,
+      "step": 950
+    },
+    {
+      "epoch": 1.8079847908745248,
+      "grad_norm": 0.674527254381119,
+      "learning_rate": 5.572560825034523e-07,
+      "loss": 0.2918,
+      "step": 951
+    },
+    {
+      "epoch": 1.8098859315589353,
+      "grad_norm": 0.6887783396936947,
+      "learning_rate": 5.463770457481732e-07,
+      "loss": 0.2988,
+      "step": 952
+    },
+    {
+      "epoch": 1.811787072243346,
+      "grad_norm": 0.665770820084528,
+      "learning_rate": 5.35602268423906e-07,
+      "loss": 0.3065,
+      "step": 953
+    },
+    {
+      "epoch": 1.8136882129277567,
+      "grad_norm": 0.8153919236218995,
+      "learning_rate": 5.249318693604577e-07,
+      "loss": 0.3139,
+      "step": 954
+    },
+    {
+      "epoch": 1.8155893536121672,
+      "grad_norm": 0.6966125859460793,
+      "learning_rate": 5.143659662364931e-07,
+      "loss": 0.3065,
+      "step": 955
+    },
+    {
+      "epoch": 1.817490494296578,
+      "grad_norm": 0.6554205004803845,
+      "learning_rate": 5.039046755782417e-07,
+      "loss": 0.2865,
+      "step": 956
+    },
+    {
+      "epoch": 1.8193916349809887,
+      "grad_norm": 0.6372543931497199,
+      "learning_rate": 4.935481127582131e-07,
+      "loss": 0.2757,
+      "step": 957
+    },
+    {
+      "epoch": 1.8212927756653992,
+      "grad_norm": 0.7194541264322672,
+      "learning_rate": 4.83296391993926e-07,
+      "loss": 0.3125,
+      "step": 958
+    },
+    {
+      "epoch": 1.8231939163498099,
+      "grad_norm": 0.6701275919877344,
+      "learning_rate": 4.7314962634664616e-07,
+      "loss": 0.2906,
+      "step": 959
+    },
+    {
+      "epoch": 1.8250950570342206,
+      "grad_norm": 0.6404564337254511,
+      "learning_rate": 4.631079277201389e-07,
+      "loss": 0.2808,
+      "step": 960
+    },
+    {
+      "epoch": 1.826996197718631,
+      "grad_norm": 0.6385329740832869,
+      "learning_rate": 4.5317140685943726e-07,
+      "loss": 0.2952,
+      "step": 961
+    },
+    {
+      "epoch": 1.8288973384030418,
+      "grad_norm": 0.6770460144244378,
+      "learning_rate": 4.433401733496201e-07,
+      "loss": 0.2826,
+      "step": 962
+    },
+    {
+      "epoch": 1.8307984790874525,
+      "grad_norm": 0.6742642347090128,
+      "learning_rate": 4.3361433561460274e-07,
+      "loss": 0.3035,
+      "step": 963
+    },
+    {
+      "epoch": 1.832699619771863,
+      "grad_norm": 0.6514289271500927,
+      "learning_rate": 4.2399400091594154e-07,
+      "loss": 0.2791,
+      "step": 964
+    },
+    {
+      "epoch": 1.8346007604562737,
+      "grad_norm": 0.6460948136624116,
+      "learning_rate": 4.14479275351648e-07,
+      "loss": 0.2788,
+      "step": 965
+    },
+    {
+      "epoch": 1.8365019011406845,
+      "grad_norm": 0.6630913406085343,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.2841,
+      "step": 966
+    },
+    {
+      "epoch": 1.838403041825095,
+      "grad_norm": 0.6647798528176125,
+      "learning_rate": 3.9576707019350903e-07,
+      "loss": 0.2949,
+      "step": 967
+    },
+    {
+      "epoch": 1.8403041825095057,
+      "grad_norm": 0.7126780335019108,
+      "learning_rate": 3.865697969675164e-07,
+      "loss": 0.3005,
+      "step": 968
+    },
+    {
+      "epoch": 1.8422053231939164,
+      "grad_norm": 0.6935661294871793,
+      "learning_rate": 3.7747854560931996e-07,
+      "loss": 0.29,
+      "step": 969
+    },
+    {
+      "epoch": 1.8441064638783269,
+      "grad_norm": 0.6812109588842511,
+      "learning_rate": 3.684934163819309e-07,
+      "loss": 0.3007,
+      "step": 970
+    },
+    {
+      "epoch": 1.8460076045627376,
+      "grad_norm": 0.6704146357876831,
+      "learning_rate": 3.596145083779912e-07,
+      "loss": 0.2859,
+      "step": 971
+    },
+    {
+      "epoch": 1.8479087452471483,
+      "grad_norm": 0.668204999358934,
+      "learning_rate": 3.508419195186774e-07,
+      "loss": 0.2881,
+      "step": 972
+    },
+    {
+      "epoch": 1.8498098859315588,
+      "grad_norm": 0.6390523472465605,
+      "learning_rate": 3.421757465526243e-07,
+      "loss": 0.2847,
+      "step": 973
+    },
+    {
+      "epoch": 1.8517110266159695,
+      "grad_norm": 0.6899070576454754,
+      "learning_rate": 3.33616085054862e-07,
+      "loss": 0.2975,
+      "step": 974
+    },
+    {
+      "epoch": 1.8536121673003803,
+      "grad_norm": 0.6664676685489681,
+      "learning_rate": 3.2516302942574794e-07,
+      "loss": 0.2992,
+      "step": 975
+    },
+    {
+      "epoch": 1.8555133079847907,
+      "grad_norm": 0.6568553549161974,
+      "learning_rate": 3.1681667288994353e-07,
+      "loss": 0.2921,
+      "step": 976
+    },
+    {
+      "epoch": 1.8574144486692015,
+      "grad_norm": 0.6465587319519815,
+      "learning_rate": 3.0857710749537585e-07,
+      "loss": 0.2835,
+      "step": 977
+    },
+    {
+      "epoch": 1.8593155893536122,
+      "grad_norm": 0.632982481293556,
+      "learning_rate": 3.0044442411222066e-07,
+      "loss": 0.2812,
+      "step": 978
+    },
+    {
+      "epoch": 1.8612167300380227,
+      "grad_norm": 0.6858999084225645,
+      "learning_rate": 2.9241871243190555e-07,
+      "loss": 0.2913,
+      "step": 979
+    },
+    {
+      "epoch": 1.8631178707224336,
+      "grad_norm": 0.6577617502631472,
+      "learning_rate": 2.845000609661208e-07,
+      "loss": 0.2849,
+      "step": 980
+    },
+    {
+      "epoch": 1.8650190114068441,
+      "grad_norm": 0.6538991396966982,
+      "learning_rate": 2.7668855704583997e-07,
+      "loss": 0.2748,
+      "step": 981
+    },
+    {
+      "epoch": 1.8669201520912546,
+      "grad_norm": 0.6338528869601765,
+      "learning_rate": 2.689842868203563e-07,
+      "loss": 0.2804,
+      "step": 982
+    },
+    {
+      "epoch": 1.8688212927756656,
+      "grad_norm": 0.7106964276835831,
+      "learning_rate": 2.6138733525633896e-07,
+      "loss": 0.2871,
+      "step": 983
+    },
+    {
+      "epoch": 1.870722433460076,
+      "grad_norm": 0.6784650089770894,
+      "learning_rate": 2.5389778613688744e-07,
+      "loss": 0.2971,
+      "step": 984
+    },
+    {
+      "epoch": 1.8726235741444865,
+      "grad_norm": 0.6475074164475771,
+      "learning_rate": 2.46515722060614e-07,
+      "loss": 0.2888,
+      "step": 985
+    },
+    {
+      "epoch": 1.8745247148288975,
+      "grad_norm": 0.6403643065350106,
+      "learning_rate": 2.392412244407294e-07,
+      "loss": 0.2665,
+      "step": 986
+    },
+    {
+      "epoch": 1.876425855513308,
+      "grad_norm": 0.6528874722807578,
+      "learning_rate": 2.3207437350414418e-07,
+      "loss": 0.2786,
+      "step": 987
+    },
+    {
+      "epoch": 1.8783269961977185,
+      "grad_norm": 0.651554678294312,
+      "learning_rate": 2.2501524829059208e-07,
+      "loss": 0.2878,
+      "step": 988
+    },
+    {
+      "epoch": 1.8802281368821294,
+      "grad_norm": 0.7178511822646201,
+      "learning_rate": 2.180639266517448e-07,
+      "loss": 0.2901,
+      "step": 989
+    },
+    {
+      "epoch": 1.88212927756654,
+      "grad_norm": 0.684217476872845,
+      "learning_rate": 2.1122048525036409e-07,
+      "loss": 0.2957,
+      "step": 990
+    },
+    {
+      "epoch": 1.8840304182509504,
+      "grad_norm": 0.6614468226998176,
+      "learning_rate": 2.0448499955945223e-07,
+      "loss": 0.3003,
+      "step": 991
+    },
+    {
+      "epoch": 1.8859315589353614,
+      "grad_norm": 0.7272061238527506,
+      "learning_rate": 1.9785754386142164e-07,
+      "loss": 0.3009,
+      "step": 992
+    },
+    {
+      "epoch": 1.8878326996197718,
+      "grad_norm": 0.6538922978993922,
+      "learning_rate": 1.9133819124727003e-07,
+      "loss": 0.2833,
+      "step": 993
+    },
+    {
+      "epoch": 1.8897338403041823,
+      "grad_norm": 0.7012939793646762,
+      "learning_rate": 1.8492701361578326e-07,
+      "loss": 0.3039,
+      "step": 994
+    },
+    {
+      "epoch": 1.8916349809885933,
+      "grad_norm": 0.6597554080602823,
+      "learning_rate": 1.7862408167273472e-07,
+      "loss": 0.296,
+      "step": 995
+    },
+    {
+      "epoch": 1.8935361216730038,
+      "grad_norm": 0.633603863248875,
+      "learning_rate": 1.724294649301095e-07,
+      "loss": 0.2914,
+      "step": 996
+    },
+    {
+      "epoch": 1.8954372623574145,
+      "grad_norm": 0.6731180020774293,
+      "learning_rate": 1.6634323170533928e-07,
+      "loss": 0.3034,
+      "step": 997
+    },
+    {
+      "epoch": 1.8973384030418252,
+      "grad_norm": 0.6518009847666095,
+      "learning_rate": 1.6036544912054087e-07,
+      "loss": 0.2963,
+      "step": 998
+    },
+    {
+      "epoch": 1.8992395437262357,
+      "grad_norm": 0.670224496673991,
+      "learning_rate": 1.544961831017855e-07,
+      "loss": 0.2944,
+      "step": 999
+    },
+    {
+      "epoch": 1.9011406844106464,
+      "grad_norm": 0.6988249089378911,
+      "learning_rate": 1.487354983783673e-07,
+      "loss": 0.307,
+      "step": 1000
+    },
+    {
+      "epoch": 1.9030418250950571,
+      "grad_norm": 0.6584505538191063,
+      "learning_rate": 1.430834584820895e-07,
+      "loss": 0.288,
+      "step": 1001
+    },
+    {
+      "epoch": 1.9049429657794676,
+      "grad_norm": 0.6864842243791344,
+      "learning_rate": 1.375401257465625e-07,
+      "loss": 0.303,
+      "step": 1002
+    },
+    {
+      "epoch": 1.9068441064638784,
+      "grad_norm": 0.6457043555603755,
+      "learning_rate": 1.3210556130652031e-07,
+      "loss": 0.2984,
+      "step": 1003
+    },
+    {
+      "epoch": 1.908745247148289,
+      "grad_norm": 0.6495170125924791,
+      "learning_rate": 1.2677982509714415e-07,
+      "loss": 0.2942,
+      "step": 1004
+    },
+    {
+      "epoch": 1.9106463878326996,
+      "grad_norm": 0.6651073599722429,
+      "learning_rate": 1.2156297585339872e-07,
+      "loss": 0.2819,
+      "step": 1005
+    },
+    {
+      "epoch": 1.9125475285171103,
+      "grad_norm": 0.6609709769387301,
+      "learning_rate": 1.1645507110938925e-07,
+      "loss": 0.2819,
+      "step": 1006
+    },
+    {
+      "epoch": 1.914448669201521,
+      "grad_norm": 0.6455735331011627,
+      "learning_rate": 1.1145616719772545e-07,
+      "loss": 0.3053,
+      "step": 1007
+    },
+    {
+      "epoch": 1.9163498098859315,
+      "grad_norm": 0.6704207903250858,
+      "learning_rate": 1.0656631924889749e-07,
+      "loss": 0.2943,
+      "step": 1008
+    },
+    {
+      "epoch": 1.9182509505703422,
+      "grad_norm": 0.6412859037920807,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.2921,
+      "step": 1009
+    },
+    {
+      "epoch": 1.920152091254753,
+      "grad_norm": 0.6446700638609452,
+      "learning_rate": 9.711400574749507e-08,
+      "loss": 0.2926,
+      "step": 1010
+    },
+    {
+      "epoch": 1.9220532319391634,
+      "grad_norm": 0.6262812527544427,
+      "learning_rate": 9.255164443990994e-08,
+      "loss": 0.2828,
+      "step": 1011
+    },
+    {
+      "epoch": 1.9239543726235742,
+      "grad_norm": 0.681070601767275,
+      "learning_rate": 8.809854758399017e-08,
+      "loss": 0.3009,
+      "step": 1012
+    },
+    {
+      "epoch": 1.9258555133079849,
+      "grad_norm": 0.6602842550236299,
+      "learning_rate": 8.375476429078543e-08,
+      "loss": 0.2993,
+      "step": 1013
+    },
+    {
+      "epoch": 1.9277566539923954,
+      "grad_norm": 0.6650499571488279,
+      "learning_rate": 7.952034246577977e-08,
+      "loss": 0.2933,
+      "step": 1014
+    },
+    {
+      "epoch": 1.929657794676806,
+      "grad_norm": 0.672449498641842,
+      "learning_rate": 7.539532880836087e-08,
+      "loss": 0.2993,
+      "step": 1015
+    },
+    {
+      "epoch": 1.9315589353612168,
+      "grad_norm": 0.629285471228973,
+      "learning_rate": 7.137976881130826e-08,
+      "loss": 0.2944,
+      "step": 1016
+    },
+    {
+      "epoch": 1.9334600760456273,
+      "grad_norm": 0.6562804111095816,
+      "learning_rate": 6.747370676028819e-08,
+      "loss": 0.2941,
+      "step": 1017
+    },
+    {
+      "epoch": 1.935361216730038,
+      "grad_norm": 0.6779580589351197,
+      "learning_rate": 6.367718573336845e-08,
+      "loss": 0.2917,
+      "step": 1018
+    },
+    {
+      "epoch": 1.9372623574144487,
+      "grad_norm": 0.726254212243231,
+      "learning_rate": 5.999024760054095e-08,
+      "loss": 0.2994,
+      "step": 1019
+    },
+    {
+      "epoch": 1.9391634980988592,
+      "grad_norm": 0.6625413405470804,
+      "learning_rate": 5.641293302326323e-08,
+      "loss": 0.2996,
+      "step": 1020
+    },
+    {
+      "epoch": 1.94106463878327,
+      "grad_norm": 0.6782464515899784,
+      "learning_rate": 5.2945281454003236e-08,
+      "loss": 0.2881,
+      "step": 1021
+    },
+    {
+      "epoch": 1.9429657794676807,
+      "grad_norm": 0.6659169873277234,
+      "learning_rate": 4.958733113581415e-08,
+      "loss": 0.3013,
+      "step": 1022
+    },
+    {
+      "epoch": 1.9448669201520912,
+      "grad_norm": 0.6652586144101917,
+      "learning_rate": 4.6339119101902475e-08,
+      "loss": 0.2971,
+      "step": 1023
+    },
+    {
+      "epoch": 1.946768060836502,
+      "grad_norm": 0.6623959414474274,
+      "learning_rate": 4.320068117522835e-08,
+      "loss": 0.2926,
+      "step": 1024
+    },
+    {
+      "epoch": 1.9486692015209126,
+      "grad_norm": 0.6779580927112054,
+      "learning_rate": 4.0172051968101474e-08,
+      "loss": 0.2973,
+      "step": 1025
+    },
+    {
+      "epoch": 1.950570342205323,
+      "grad_norm": 0.6234988596942616,
+      "learning_rate": 3.7253264881809137e-08,
+      "loss": 0.2791,
+      "step": 1026
+    },
+    {
+      "epoch": 1.9524714828897338,
+      "grad_norm": 0.6368436322896437,
+      "learning_rate": 3.4444352106242086e-08,
+      "loss": 0.2807,
+      "step": 1027
+    },
+    {
+      "epoch": 1.9543726235741445,
+      "grad_norm": 0.6575580136715466,
+      "learning_rate": 3.174534461953593e-08,
+      "loss": 0.2864,
+      "step": 1028
+    },
+    {
+      "epoch": 1.956273764258555,
+      "grad_norm": 0.653299992797501,
+      "learning_rate": 2.915627218774142e-08,
+      "loss": 0.2873,
+      "step": 1029
+    },
+    {
+      "epoch": 1.9581749049429658,
+      "grad_norm": 0.6974363536630532,
+      "learning_rate": 2.667716336448356e-08,
+      "loss": 0.3036,
+      "step": 1030
+    },
+    {
+      "epoch": 1.9600760456273765,
+      "grad_norm": 0.6316643182250802,
+      "learning_rate": 2.430804549065302e-08,
+      "loss": 0.2966,
+      "step": 1031
+    },
+    {
+      "epoch": 1.961977186311787,
+      "grad_norm": 0.6546907254072903,
+      "learning_rate": 2.2048944694104123e-08,
+      "loss": 0.2939,
+      "step": 1032
+    },
+    {
+      "epoch": 1.9638783269961977,
+      "grad_norm": 0.6728406675509413,
+      "learning_rate": 1.989988588936509e-08,
+      "loss": 0.3153,
+      "step": 1033
+    },
+    {
+      "epoch": 1.9657794676806084,
+      "grad_norm": 0.678688708491006,
+      "learning_rate": 1.7860892777367133e-08,
+      "loss": 0.2883,
+      "step": 1034
+    },
+    {
+      "epoch": 1.967680608365019,
+      "grad_norm": 0.6351759171410115,
+      "learning_rate": 1.5931987845176912e-08,
+      "loss": 0.2603,
+      "step": 1035
+    },
+    {
+      "epoch": 1.9695817490494296,
+      "grad_norm": 0.6485160029826156,
+      "learning_rate": 1.411319236575337e-08,
+      "loss": 0.2872,
+      "step": 1036
+    },
+    {
+      "epoch": 1.9714828897338403,
+      "grad_norm": 0.6824212524832303,
+      "learning_rate": 1.2404526397711281e-08,
+      "loss": 0.3114,
+      "step": 1037
+    },
+    {
+      "epoch": 1.9733840304182508,
+      "grad_norm": 0.6426680134773441,
+      "learning_rate": 1.0806008785100297e-08,
+      "loss": 0.2802,
+      "step": 1038
+    },
+    {
+      "epoch": 1.9752851711026616,
+      "grad_norm": 0.7307513538636258,
+      "learning_rate": 9.317657157197347e-09,
+      "loss": 0.3131,
+      "step": 1039
+    },
+    {
+      "epoch": 1.9771863117870723,
+      "grad_norm": 0.6884487536431106,
+      "learning_rate": 7.93948792831234e-09,
+      "loss": 0.2932,
+      "step": 1040
+    },
+    {
+      "epoch": 1.9790874524714828,
+      "grad_norm": 0.680295931784895,
+      "learning_rate": 6.671516297606095e-09,
+      "loss": 0.3032,
+      "step": 1041
+    },
+    {
+      "epoch": 1.9809885931558935,
+      "grad_norm": 0.6443008079418515,
+      "learning_rate": 5.513756248924917e-09,
+      "loss": 0.2978,
+      "step": 1042
+    },
+    {
+      "epoch": 1.9828897338403042,
+      "grad_norm": 0.6445665724612897,
+      "learning_rate": 4.466220550641831e-09,
+      "loss": 0.2934,
+      "step": 1043
+    },
+    {
+      "epoch": 1.9847908745247147,
+      "grad_norm": 0.7126505087014392,
+      "learning_rate": 3.528920755523357e-09,
+      "loss": 0.2954,
+      "step": 1044
+    },
+    {
+      "epoch": 1.9866920152091256,
+      "grad_norm": 0.679656273046371,
+      "learning_rate": 2.701867200592956e-09,
+      "loss": 0.2958,
+      "step": 1045
+    },
+    {
+      "epoch": 1.9885931558935361,
+      "grad_norm": 0.6327799367751911,
+      "learning_rate": 1.9850690070266633e-09,
+      "loss": 0.2942,
+      "step": 1046
+    },
+    {
+      "epoch": 1.9904942965779466,
+      "grad_norm": 0.6655175669835449,
+      "learning_rate": 1.378534080042071e-09,
+      "loss": 0.2907,
+      "step": 1047
+    },
+    {
+      "epoch": 1.9923954372623576,
+      "grad_norm": 0.6392903639247768,
+      "learning_rate": 8.822691088195001e-10,
+      "loss": 0.2839,
+      "step": 1048
+    },
+    {
+      "epoch": 1.994296577946768,
+      "grad_norm": 0.6595707456495274,
+      "learning_rate": 4.962795664265052e-10,
+      "loss": 0.288,
+      "step": 1049
+    },
+    {
+      "epoch": 1.9961977186311786,
+      "grad_norm": 0.6524297253883753,
+      "learning_rate": 2.2056970975459223e-10,
+      "loss": 0.2885,
+      "step": 1050
+    },
+    {
+      "epoch": 1.9980988593155895,
+      "grad_norm": 0.6756428270032067,
+      "learning_rate": 5.514257947369928e-11,
+      "loss": 0.301,
+      "step": 1051
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.614461283730078,
+      "learning_rate": 0.0,
+      "loss": 0.2925,
+      "step": 1052
+    },
+    {
+      "epoch": 2.0,
+      "step": 1052,
+      "total_flos": 2.1309492080463053e+17,
+      "train_loss": 0.37581125442859337,
+      "train_runtime": 2626.4209,
+      "train_samples_per_second": 51.27,
+      "train_steps_per_second": 0.401
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1052,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.1309492080463053e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}