diff --git "a/checkpoint-1207/trainer_state.json" "b/checkpoint-1207/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1207/trainer_state.json"
@@ -0,0 +1,8482 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.999723909442297,
+  "eval_steps": 500,
+  "global_step": 1207,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008282716731087797,
+      "grad_norm": 2.4795358180999756,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 1.1215,
+      "step": 1
+    },
+    {
+      "epoch": 0.0016565433462175593,
+      "grad_norm": 2.6060028076171875,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 1.1219,
+      "step": 2
+    },
+    {
+      "epoch": 0.002484815019326339,
+      "grad_norm": 2.750868082046509,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 1.125,
+      "step": 3
+    },
+    {
+      "epoch": 0.0033130866924351186,
+      "grad_norm": 2.7321369647979736,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 1.1164,
+      "step": 4
+    },
+    {
+      "epoch": 0.0041413583655438985,
+      "grad_norm": 2.8193929195404053,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 1.1263,
+      "step": 5
+    },
+    {
+      "epoch": 0.004969630038652678,
+      "grad_norm": 2.648301839828491,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 1.1092,
+      "step": 6
+    },
+    {
+      "epoch": 0.005797901711761457,
+      "grad_norm": 2.751993179321289,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 1.0997,
+      "step": 7
+    },
+    {
+      "epoch": 0.006626173384870237,
+      "grad_norm": 2.592381000518799,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 1.112,
+      "step": 8
+    },
+    {
+      "epoch": 0.007454445057979017,
+      "grad_norm": 2.3135671615600586,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 1.1045,
+      "step": 9
+    },
+    {
+      "epoch": 0.008282716731087797,
+      "grad_norm": 2.60491943359375,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.0956,
+      "step": 10
+    },
+    {
+      "epoch": 0.009110988404196576,
+      "grad_norm": 2.2435271739959717,
+      "learning_rate": 5.5e-07,
+      "loss": 1.1241,
+      "step": 11
+    },
+    {
+      "epoch": 0.009939260077305357,
+      "grad_norm": 2.4412319660186768,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.11,
+      "step": 12
+    },
+    {
+      "epoch": 0.010767531750414136,
+      "grad_norm": 2.2075185775756836,
+      "learning_rate": 6.5e-07,
+      "loss": 1.0878,
+      "step": 13
+    },
+    {
+      "epoch": 0.011595803423522915,
+      "grad_norm": 2.101902723312378,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 1.0863,
+      "step": 14
+    },
+    {
+      "epoch": 0.012424075096631695,
+      "grad_norm": 2.134535074234009,
+      "learning_rate": 7.5e-07,
+      "loss": 1.0735,
+      "step": 15
+    },
+    {
+      "epoch": 0.013252346769740474,
+      "grad_norm": 2.000504493713379,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.0875,
+      "step": 16
+    },
+    {
+      "epoch": 0.014080618442849255,
+      "grad_norm": 1.9191882610321045,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 1.1039,
+      "step": 17
+    },
+    {
+      "epoch": 0.014908890115958034,
+      "grad_norm": 1.7732031345367432,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 1.0819,
+      "step": 18
+    },
+    {
+      "epoch": 0.015737161789066815,
+      "grad_norm": 1.7775251865386963,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 1.0936,
+      "step": 19
+    },
+    {
+      "epoch": 0.016565433462175594,
+      "grad_norm": 1.6648598909378052,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.0881,
+      "step": 20
+    },
+    {
+      "epoch": 0.017393705135284373,
+      "grad_norm": 1.5726720094680786,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 1.07,
+      "step": 21
+    },
+    {
+      "epoch": 0.018221976808393152,
+      "grad_norm": 1.561417818069458,
+      "learning_rate": 1.1e-06,
+      "loss": 1.0668,
+      "step": 22
+    },
+    {
+      "epoch": 0.019050248481501934,
+      "grad_norm": 1.4953842163085938,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 1.054,
+      "step": 23
+    },
+    {
+      "epoch": 0.019878520154610713,
+      "grad_norm": 1.4247829914093018,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.0363,
+      "step": 24
+    },
+    {
+      "epoch": 0.020706791827719492,
+      "grad_norm": 1.3483426570892334,
+      "learning_rate": 1.25e-06,
+      "loss": 1.0143,
+      "step": 25
+    },
+    {
+      "epoch": 0.02153506350082827,
+      "grad_norm": 1.3412944078445435,
+      "learning_rate": 1.3e-06,
+      "loss": 1.025,
+      "step": 26
+    },
+    {
+      "epoch": 0.02236333517393705,
+      "grad_norm": 1.2834237813949585,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 1.0227,
+      "step": 27
+    },
+    {
+      "epoch": 0.02319160684704583,
+      "grad_norm": 1.2128241062164307,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 1.0161,
+      "step": 28
+    },
+    {
+      "epoch": 0.024019878520154612,
+      "grad_norm": 1.1665573120117188,
+      "learning_rate": 1.45e-06,
+      "loss": 1.0034,
+      "step": 29
+    },
+    {
+      "epoch": 0.02484815019326339,
+      "grad_norm": 1.1575899124145508,
+      "learning_rate": 1.5e-06,
+      "loss": 0.9968,
+      "step": 30
+    },
+    {
+      "epoch": 0.02567642186637217,
+      "grad_norm": 1.195304274559021,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 1.0176,
+      "step": 31
+    },
+    {
+      "epoch": 0.02650469353948095,
+      "grad_norm": 1.1321678161621094,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.0113,
+      "step": 32
+    },
+    {
+      "epoch": 0.027332965212589728,
+      "grad_norm": 1.1076338291168213,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 0.9971,
+      "step": 33
+    },
+    {
+      "epoch": 0.02816123688569851,
+      "grad_norm": 1.0770610570907593,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 0.9994,
+      "step": 34
+    },
+    {
+      "epoch": 0.02898950855880729,
+      "grad_norm": 1.0767767429351807,
+      "learning_rate": 1.75e-06,
+      "loss": 0.978,
+      "step": 35
+    },
+    {
+      "epoch": 0.02981778023191607,
+      "grad_norm": 1.054800033569336,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.9829,
+      "step": 36
+    },
+    {
+      "epoch": 0.030646051905024847,
+      "grad_norm": 1.031015157699585,
+      "learning_rate": 1.85e-06,
+      "loss": 0.97,
+      "step": 37
+    },
+    {
+      "epoch": 0.03147432357813363,
+      "grad_norm": 1.0182578563690186,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.9549,
+      "step": 38
+    },
+    {
+      "epoch": 0.032302595251242405,
+      "grad_norm": 0.9796578288078308,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 0.9762,
+      "step": 39
+    },
+    {
+      "epoch": 0.03313086692435119,
+      "grad_norm": 0.9956578016281128,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.9693,
+      "step": 40
+    },
+    {
+      "epoch": 0.03395913859745997,
+      "grad_norm": 0.9383260011672974,
+      "learning_rate": 2.05e-06,
+      "loss": 0.9556,
+      "step": 41
+    },
+    {
+      "epoch": 0.034787410270568746,
+      "grad_norm": 0.9372819066047668,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.9419,
+      "step": 42
+    },
+    {
+      "epoch": 0.03561568194367753,
+      "grad_norm": 0.9334004521369934,
+      "learning_rate": 2.15e-06,
+      "loss": 0.9587,
+      "step": 43
+    },
+    {
+      "epoch": 0.036443953616786304,
+      "grad_norm": 0.8916004300117493,
+      "learning_rate": 2.2e-06,
+      "loss": 0.9227,
+      "step": 44
+    },
+    {
+      "epoch": 0.037272225289895086,
+      "grad_norm": 0.8698756694793701,
+      "learning_rate": 2.25e-06,
+      "loss": 0.9275,
+      "step": 45
+    },
+    {
+      "epoch": 0.03810049696300387,
+      "grad_norm": 0.8775100708007812,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.941,
+      "step": 46
+    },
+    {
+      "epoch": 0.038928768636112644,
+      "grad_norm": 0.8763384222984314,
+      "learning_rate": 2.35e-06,
+      "loss": 0.9253,
+      "step": 47
+    },
+    {
+      "epoch": 0.03975704030922143,
+      "grad_norm": 0.8534179925918579,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.9327,
+      "step": 48
+    },
+    {
+      "epoch": 0.0405853119823302,
+      "grad_norm": 0.8432146906852722,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.9236,
+      "step": 49
+    },
+    {
+      "epoch": 0.041413583655438985,
+      "grad_norm": 0.9899975061416626,
+      "learning_rate": 2.5e-06,
+      "loss": 0.8839,
+      "step": 50
+    },
+    {
+      "epoch": 0.04224185532854776,
+      "grad_norm": 0.8370104432106018,
+      "learning_rate": 2.55e-06,
+      "loss": 0.906,
+      "step": 51
+    },
+    {
+      "epoch": 0.04307012700165654,
+      "grad_norm": 0.8207446336746216,
+      "learning_rate": 2.6e-06,
+      "loss": 0.8755,
+      "step": 52
+    },
+    {
+      "epoch": 0.043898398674765325,
+      "grad_norm": 0.805584728717804,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.9049,
+      "step": 53
+    },
+    {
+      "epoch": 0.0447266703478741,
+      "grad_norm": 0.8162860870361328,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.8993,
+      "step": 54
+    },
+    {
+      "epoch": 0.04555494202098288,
+      "grad_norm": 0.8151668906211853,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.9282,
+      "step": 55
+    },
+    {
+      "epoch": 0.04638321369409166,
+      "grad_norm": 0.7996147274971008,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.8946,
+      "step": 56
+    },
+    {
+      "epoch": 0.04721148536720044,
+      "grad_norm": 0.8223204016685486,
+      "learning_rate": 2.85e-06,
+      "loss": 0.897,
+      "step": 57
+    },
+    {
+      "epoch": 0.048039757040309224,
+      "grad_norm": 0.8117988109588623,
+      "learning_rate": 2.9e-06,
+      "loss": 0.8845,
+      "step": 58
+    },
+    {
+      "epoch": 0.048868028713418,
+      "grad_norm": 0.7902786731719971,
+      "learning_rate": 2.95e-06,
+      "loss": 0.8846,
+      "step": 59
+    },
+    {
+      "epoch": 0.04969630038652678,
+      "grad_norm": 0.804192304611206,
+      "learning_rate": 3e-06,
+      "loss": 0.8722,
+      "step": 60
+    },
+    {
+      "epoch": 0.05052457205963556,
+      "grad_norm": 0.7852040529251099,
+      "learning_rate": 3.05e-06,
+      "loss": 0.8821,
+      "step": 61
+    },
+    {
+      "epoch": 0.05135284373274434,
+      "grad_norm": 0.8045080304145813,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.8661,
+      "step": 62
+    },
+    {
+      "epoch": 0.05218111540585312,
+      "grad_norm": 0.7920243144035339,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.8712,
+      "step": 63
+    },
+    {
+      "epoch": 0.0530093870789619,
+      "grad_norm": 0.7601844072341919,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.8792,
+      "step": 64
+    },
+    {
+      "epoch": 0.05383765875207068,
+      "grad_norm": 0.7833033800125122,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.8811,
+      "step": 65
+    },
+    {
+      "epoch": 0.054665930425179456,
+      "grad_norm": 0.7691664099693298,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.8701,
+      "step": 66
+    },
+    {
+      "epoch": 0.05549420209828824,
+      "grad_norm": 0.7845340967178345,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.8929,
+      "step": 67
+    },
+    {
+      "epoch": 0.05632247377139702,
+      "grad_norm": 0.9076205492019653,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.8352,
+      "step": 68
+    },
+    {
+      "epoch": 0.057150745444505796,
+      "grad_norm": 0.7614927887916565,
+      "learning_rate": 3.45e-06,
+      "loss": 0.8766,
+      "step": 69
+    },
+    {
+      "epoch": 0.05797901711761458,
+      "grad_norm": 0.7722934484481812,
+      "learning_rate": 3.5e-06,
+      "loss": 0.8666,
+      "step": 70
+    },
+    {
+      "epoch": 0.058807288790723354,
+      "grad_norm": 0.7603748440742493,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.84,
+      "step": 71
+    },
+    {
+      "epoch": 0.05963556046383214,
+      "grad_norm": 0.7563591003417969,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.8784,
+      "step": 72
+    },
+    {
+      "epoch": 0.06046383213694092,
+      "grad_norm": 0.7835114002227783,
+      "learning_rate": 3.65e-06,
+      "loss": 0.8516,
+      "step": 73
+    },
+    {
+      "epoch": 0.061292103810049695,
+      "grad_norm": 0.7760982513427734,
+      "learning_rate": 3.7e-06,
+      "loss": 0.8265,
+      "step": 74
+    },
+    {
+      "epoch": 0.06212037548315848,
+      "grad_norm": 0.7978456020355225,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.8564,
+      "step": 75
+    },
+    {
+      "epoch": 0.06294864715626726,
+      "grad_norm": 0.7940747141838074,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.8635,
+      "step": 76
+    },
+    {
+      "epoch": 0.06377691882937604,
+      "grad_norm": 0.8055519461631775,
+      "learning_rate": 3.85e-06,
+      "loss": 0.8445,
+      "step": 77
+    },
+    {
+      "epoch": 0.06460519050248481,
+      "grad_norm": 0.7532406449317932,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.8772,
+      "step": 78
+    },
+    {
+      "epoch": 0.06543346217559359,
+      "grad_norm": 0.763933539390564,
+      "learning_rate": 3.95e-06,
+      "loss": 0.8574,
+      "step": 79
+    },
+    {
+      "epoch": 0.06626173384870238,
+      "grad_norm": 0.7833206057548523,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.8643,
+      "step": 80
+    },
+    {
+      "epoch": 0.06709000552181116,
+      "grad_norm": 0.7825527787208557,
+      "learning_rate": 4.05e-06,
+      "loss": 0.8518,
+      "step": 81
+    },
+    {
+      "epoch": 0.06791827719491994,
+      "grad_norm": 0.7749581336975098,
+      "learning_rate": 4.1e-06,
+      "loss": 0.8146,
+      "step": 82
+    },
+    {
+      "epoch": 0.06874654886802871,
+      "grad_norm": 0.8117144703865051,
+      "learning_rate": 4.15e-06,
+      "loss": 0.8719,
+      "step": 83
+    },
+    {
+      "epoch": 0.06957482054113749,
+      "grad_norm": 0.8440507054328918,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.8665,
+      "step": 84
+    },
+    {
+      "epoch": 0.07040309221424627,
+      "grad_norm": 0.7941984534263611,
+      "learning_rate": 4.25e-06,
+      "loss": 0.8492,
+      "step": 85
+    },
+    {
+      "epoch": 0.07123136388735506,
+      "grad_norm": 0.8301989436149597,
+      "learning_rate": 4.3e-06,
+      "loss": 0.8544,
+      "step": 86
+    },
+    {
+      "epoch": 0.07205963556046384,
+      "grad_norm": 0.757798969745636,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.8426,
+      "step": 87
+    },
+    {
+      "epoch": 0.07288790723357261,
+      "grad_norm": 0.7817673683166504,
+      "learning_rate": 4.4e-06,
+      "loss": 0.8554,
+      "step": 88
+    },
+    {
+      "epoch": 0.07371617890668139,
+      "grad_norm": 0.7609818577766418,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.8473,
+      "step": 89
+    },
+    {
+      "epoch": 0.07454445057979017,
+      "grad_norm": 0.8146091103553772,
+      "learning_rate": 4.5e-06,
+      "loss": 0.8317,
+      "step": 90
+    },
+    {
+      "epoch": 0.07537272225289895,
+      "grad_norm": 0.8058239817619324,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.8411,
+      "step": 91
+    },
+    {
+      "epoch": 0.07620099392600774,
+      "grad_norm": 0.7700346112251282,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.815,
+      "step": 92
+    },
+    {
+      "epoch": 0.0770292655991165,
+      "grad_norm": 0.7949067950248718,
+      "learning_rate": 4.65e-06,
+      "loss": 0.8512,
+      "step": 93
+    },
+    {
+      "epoch": 0.07785753727222529,
+      "grad_norm": 0.8112176060676575,
+      "learning_rate": 4.7e-06,
+      "loss": 0.8398,
+      "step": 94
+    },
+    {
+      "epoch": 0.07868580894533407,
+      "grad_norm": 0.7990434765815735,
+      "learning_rate": 4.75e-06,
+      "loss": 0.8276,
+      "step": 95
+    },
+    {
+      "epoch": 0.07951408061844285,
+      "grad_norm": 0.7942920923233032,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.8333,
+      "step": 96
+    },
+    {
+      "epoch": 0.08034235229155164,
+      "grad_norm": 0.8005534410476685,
+      "learning_rate": 4.85e-06,
+      "loss": 0.8453,
+      "step": 97
+    },
+    {
+      "epoch": 0.0811706239646604,
+      "grad_norm": 0.8103216290473938,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.8303,
+      "step": 98
+    },
+    {
+      "epoch": 0.08199889563776919,
+      "grad_norm": 0.8217238187789917,
+      "learning_rate": 4.95e-06,
+      "loss": 0.8204,
+      "step": 99
+    },
+    {
+      "epoch": 0.08282716731087797,
+      "grad_norm": 0.7958654761314392,
+      "learning_rate": 5e-06,
+      "loss": 0.8237,
+      "step": 100
+    },
+    {
+      "epoch": 0.08365543898398675,
+      "grad_norm": 0.8631463050842285,
+      "learning_rate": 4.999999758136652e-06,
+      "loss": 0.821,
+      "step": 101
+    },
+    {
+      "epoch": 0.08448371065709552,
+      "grad_norm": 0.7666735053062439,
+      "learning_rate": 4.999999032546657e-06,
+      "loss": 0.8307,
+      "step": 102
+    },
+    {
+      "epoch": 0.0853119823302043,
+      "grad_norm": 0.7709226012229919,
+      "learning_rate": 4.999997823230153e-06,
+      "loss": 0.8299,
+      "step": 103
+    },
+    {
+      "epoch": 0.08614025400331309,
+      "grad_norm": 0.8020651340484619,
+      "learning_rate": 4.9999961301873736e-06,
+      "loss": 0.8075,
+      "step": 104
+    },
+    {
+      "epoch": 0.08696852567642187,
+      "grad_norm": 0.8019537329673767,
+      "learning_rate": 4.999993953418649e-06,
+      "loss": 0.8615,
+      "step": 105
+    },
+    {
+      "epoch": 0.08779679734953065,
+      "grad_norm": 0.7904282808303833,
+      "learning_rate": 4.999991292924399e-06,
+      "loss": 0.8353,
+      "step": 106
+    },
+    {
+      "epoch": 0.08862506902263942,
+      "grad_norm": 0.7854570150375366,
+      "learning_rate": 4.999988148705138e-06,
+      "loss": 0.8154,
+      "step": 107
+    },
+    {
+      "epoch": 0.0894533406957482,
+      "grad_norm": 0.8044222593307495,
+      "learning_rate": 4.999984520761475e-06,
+      "loss": 0.8117,
+      "step": 108
+    },
+    {
+      "epoch": 0.09028161236885698,
+      "grad_norm": 0.7986987829208374,
+      "learning_rate": 4.999980409094112e-06,
+      "loss": 0.8407,
+      "step": 109
+    },
+    {
+      "epoch": 0.09110988404196577,
+      "grad_norm": 0.7976786494255066,
+      "learning_rate": 4.999975813703844e-06,
+      "loss": 0.7871,
+      "step": 110
+    },
+    {
+      "epoch": 0.09193815571507455,
+      "grad_norm": 0.8330315351486206,
+      "learning_rate": 4.9999707345915605e-06,
+      "loss": 0.8129,
+      "step": 111
+    },
+    {
+      "epoch": 0.09276642738818332,
+      "grad_norm": 0.8350826501846313,
+      "learning_rate": 4.999965171758245e-06,
+      "loss": 0.8144,
+      "step": 112
+    },
+    {
+      "epoch": 0.0935946990612921,
+      "grad_norm": 0.8054768443107605,
+      "learning_rate": 4.999959125204973e-06,
+      "loss": 0.809,
+      "step": 113
+    },
+    {
+      "epoch": 0.09442297073440088,
+      "grad_norm": 0.8101645708084106,
+      "learning_rate": 4.999952594932914e-06,
+      "loss": 0.8081,
+      "step": 114
+    },
+    {
+      "epoch": 0.09525124240750966,
+      "grad_norm": 0.905028760433197,
+      "learning_rate": 4.999945580943332e-06,
+      "loss": 0.8251,
+      "step": 115
+    },
+    {
+      "epoch": 0.09607951408061845,
+      "grad_norm": 0.8529418706893921,
+      "learning_rate": 4.999938083237585e-06,
+      "loss": 0.7984,
+      "step": 116
+    },
+    {
+      "epoch": 0.09690778575372722,
+      "grad_norm": 0.8047451376914978,
+      "learning_rate": 4.9999301018171225e-06,
+      "loss": 0.8222,
+      "step": 117
+    },
+    {
+      "epoch": 0.097736057426836,
+      "grad_norm": 0.8073400259017944,
+      "learning_rate": 4.99992163668349e-06,
+      "loss": 0.8251,
+      "step": 118
+    },
+    {
+      "epoch": 0.09856432909994478,
+      "grad_norm": 0.8470693826675415,
+      "learning_rate": 4.999912687838324e-06,
+      "loss": 0.8291,
+      "step": 119
+    },
+    {
+      "epoch": 0.09939260077305356,
+      "grad_norm": 0.8261650800704956,
+      "learning_rate": 4.999903255283357e-06,
+      "loss": 0.808,
+      "step": 120
+    },
+    {
+      "epoch": 0.10022087244616235,
+      "grad_norm": 0.8421704769134521,
+      "learning_rate": 4.999893339020414e-06,
+      "loss": 0.8295,
+      "step": 121
+    },
+    {
+      "epoch": 0.10104914411927111,
+      "grad_norm": 0.8244299292564392,
+      "learning_rate": 4.9998829390514134e-06,
+      "loss": 0.8215,
+      "step": 122
+    },
+    {
+      "epoch": 0.1018774157923799,
+      "grad_norm": 0.8365658521652222,
+      "learning_rate": 4.999872055378368e-06,
+      "loss": 0.8245,
+      "step": 123
+    },
+    {
+      "epoch": 0.10270568746548868,
+      "grad_norm": 0.9089561104774475,
+      "learning_rate": 4.999860688003384e-06,
+      "loss": 0.8125,
+      "step": 124
+    },
+    {
+      "epoch": 0.10353395913859746,
+      "grad_norm": 0.8784728646278381,
+      "learning_rate": 4.9998488369286595e-06,
+      "loss": 0.8321,
+      "step": 125
+    },
+    {
+      "epoch": 0.10436223081170624,
+      "grad_norm": 0.8596073389053345,
+      "learning_rate": 4.999836502156487e-06,
+      "loss": 0.8333,
+      "step": 126
+    },
+    {
+      "epoch": 0.10519050248481501,
+      "grad_norm": 0.8472688794136047,
+      "learning_rate": 4.9998236836892566e-06,
+      "loss": 0.8068,
+      "step": 127
+    },
+    {
+      "epoch": 0.1060187741579238,
+      "grad_norm": 0.8716503381729126,
+      "learning_rate": 4.999810381529447e-06,
+      "loss": 0.8005,
+      "step": 128
+    },
+    {
+      "epoch": 0.10684704583103258,
+      "grad_norm": 0.8416318893432617,
+      "learning_rate": 4.999796595679629e-06,
+      "loss": 0.7914,
+      "step": 129
+    },
+    {
+      "epoch": 0.10767531750414136,
+      "grad_norm": 0.838101863861084,
+      "learning_rate": 4.999782326142474e-06,
+      "loss": 0.7957,
+      "step": 130
+    },
+    {
+      "epoch": 0.10850358917725014,
+      "grad_norm": 0.8640785217285156,
+      "learning_rate": 4.999767572920741e-06,
+      "loss": 0.8077,
+      "step": 131
+    },
+    {
+      "epoch": 0.10933186085035891,
+      "grad_norm": 0.8207422494888306,
+      "learning_rate": 4.999752336017286e-06,
+      "loss": 0.7973,
+      "step": 132
+    },
+    {
+      "epoch": 0.1101601325234677,
+      "grad_norm": 0.8640758395195007,
+      "learning_rate": 4.9997366154350566e-06,
+      "loss": 0.8165,
+      "step": 133
+    },
+    {
+      "epoch": 0.11098840419657648,
+      "grad_norm": 0.7965807914733887,
+      "learning_rate": 4.999720411177094e-06,
+      "loss": 0.8059,
+      "step": 134
+    },
+    {
+      "epoch": 0.11181667586968526,
+      "grad_norm": 0.8463795185089111,
+      "learning_rate": 4.999703723246533e-06,
+      "loss": 0.8106,
+      "step": 135
+    },
+    {
+      "epoch": 0.11264494754279404,
+      "grad_norm": 0.8118059635162354,
+      "learning_rate": 4.999686551646604e-06,
+      "loss": 0.8099,
+      "step": 136
+    },
+    {
+      "epoch": 0.11347321921590281,
+      "grad_norm": 0.8256771564483643,
+      "learning_rate": 4.999668896380629e-06,
+      "loss": 0.8043,
+      "step": 137
+    },
+    {
+      "epoch": 0.11430149088901159,
+      "grad_norm": 0.8203126192092896,
+      "learning_rate": 4.999650757452025e-06,
+      "loss": 0.8029,
+      "step": 138
+    },
+    {
+      "epoch": 0.11512976256212037,
+      "grad_norm": 0.8442854881286621,
+      "learning_rate": 4.9996321348643e-06,
+      "loss": 0.8004,
+      "step": 139
+    },
+    {
+      "epoch": 0.11595803423522916,
+      "grad_norm": 0.8523190021514893,
+      "learning_rate": 4.999613028621059e-06,
+      "loss": 0.8092,
+      "step": 140
+    },
+    {
+      "epoch": 0.11678630590833794,
+      "grad_norm": 0.821428656578064,
+      "learning_rate": 4.999593438725997e-06,
+      "loss": 0.8105,
+      "step": 141
+    },
+    {
+      "epoch": 0.11761457758144671,
+      "grad_norm": 0.8406105041503906,
+      "learning_rate": 4.999573365182906e-06,
+      "loss": 0.7757,
+      "step": 142
+    },
+    {
+      "epoch": 0.11844284925455549,
+      "grad_norm": 0.8620600700378418,
+      "learning_rate": 4.999552807995669e-06,
+      "loss": 0.781,
+      "step": 143
+    },
+    {
+      "epoch": 0.11927112092766427,
+      "grad_norm": 0.920437216758728,
+      "learning_rate": 4.9995317671682655e-06,
+      "loss": 0.7579,
+      "step": 144
+    },
+    {
+      "epoch": 0.12009939260077306,
+      "grad_norm": 0.8414936661720276,
+      "learning_rate": 4.999510242704765e-06,
+      "loss": 0.8031,
+      "step": 145
+    },
+    {
+      "epoch": 0.12092766427388184,
+      "grad_norm": 0.8372176289558411,
+      "learning_rate": 4.999488234609332e-06,
+      "loss": 0.7931,
+      "step": 146
+    },
+    {
+      "epoch": 0.1217559359469906,
+      "grad_norm": 0.8785244226455688,
+      "learning_rate": 4.999465742886226e-06,
+      "loss": 0.8189,
+      "step": 147
+    },
+    {
+      "epoch": 0.12258420762009939,
+      "grad_norm": 0.8404353260993958,
+      "learning_rate": 4.999442767539799e-06,
+      "loss": 0.7899,
+      "step": 148
+    },
+    {
+      "epoch": 0.12341247929320817,
+      "grad_norm": 0.8541820049285889,
+      "learning_rate": 4.999419308574494e-06,
+      "loss": 0.7764,
+      "step": 149
+    },
+    {
+      "epoch": 0.12424075096631695,
+      "grad_norm": 0.8220177292823792,
+      "learning_rate": 4.999395365994854e-06,
+      "loss": 0.7942,
+      "step": 150
+    },
+    {
+      "epoch": 0.12506902263942574,
+      "grad_norm": 0.8823731541633606,
+      "learning_rate": 4.999370939805509e-06,
+      "loss": 0.793,
+      "step": 151
+    },
+    {
+      "epoch": 0.12589729431253452,
+      "grad_norm": 0.838860034942627,
+      "learning_rate": 4.999346030011186e-06,
+      "loss": 0.7741,
+      "step": 152
+    },
+    {
+      "epoch": 0.1267255659856433,
+      "grad_norm": 0.8162702322006226,
+      "learning_rate": 4.999320636616705e-06,
+      "loss": 0.7989,
+      "step": 153
+    },
+    {
+      "epoch": 0.12755383765875208,
+      "grad_norm": 0.8328465819358826,
+      "learning_rate": 4.99929475962698e-06,
+      "loss": 0.8055,
+      "step": 154
+    },
+    {
+      "epoch": 0.12838210933186084,
+      "grad_norm": 0.8373605608940125,
+      "learning_rate": 4.999268399047016e-06,
+      "loss": 0.801,
+      "step": 155
+    },
+    {
+      "epoch": 0.12921038100496962,
+      "grad_norm": 0.8408295512199402,
+      "learning_rate": 4.999241554881915e-06,
+      "loss": 0.7749,
+      "step": 156
+    },
+    {
+      "epoch": 0.1300386526780784,
+      "grad_norm": 0.8326342701911926,
+      "learning_rate": 4.999214227136871e-06,
+      "loss": 0.7818,
+      "step": 157
+    },
+    {
+      "epoch": 0.13086692435118719,
+      "grad_norm": 0.8064060211181641,
+      "learning_rate": 4.99918641581717e-06,
+      "loss": 0.7755,
+      "step": 158
+    },
+    {
+      "epoch": 0.13169519602429597,
+      "grad_norm": 0.8274610042572021,
+      "learning_rate": 4.999158120928196e-06,
+      "loss": 0.8008,
+      "step": 159
+    },
+    {
+      "epoch": 0.13252346769740475,
+      "grad_norm": 0.8121576309204102,
+      "learning_rate": 4.999129342475422e-06,
+      "loss": 0.7713,
+      "step": 160
+    },
+    {
+      "epoch": 0.13335173937051353,
+      "grad_norm": 0.8227710723876953,
+      "learning_rate": 4.9991000804644176e-06,
+      "loss": 0.7902,
+      "step": 161
+    },
+    {
+      "epoch": 0.13418001104362232,
+      "grad_norm": 0.8437026739120483,
+      "learning_rate": 4.999070334900843e-06,
+      "loss": 0.8176,
+      "step": 162
+    },
+    {
+      "epoch": 0.1350082827167311,
+      "grad_norm": 0.8379645943641663,
+      "learning_rate": 4.999040105790455e-06,
+      "loss": 0.7973,
+      "step": 163
+    },
+    {
+      "epoch": 0.13583655438983988,
+      "grad_norm": 0.822542130947113,
+      "learning_rate": 4.999009393139103e-06,
+      "loss": 0.7955,
+      "step": 164
+    },
+    {
+      "epoch": 0.13666482606294864,
+      "grad_norm": 0.8141928315162659,
+      "learning_rate": 4.9989781969527274e-06,
+      "loss": 0.7766,
+      "step": 165
+    },
+    {
+      "epoch": 0.13749309773605742,
+      "grad_norm": 0.8325410485267639,
+      "learning_rate": 4.998946517237367e-06,
+      "loss": 0.7737,
+      "step": 166
+    },
+    {
+      "epoch": 0.1383213694091662,
+      "grad_norm": 0.8399263024330139,
+      "learning_rate": 4.99891435399915e-06,
+      "loss": 0.7746,
+      "step": 167
+    },
+    {
+      "epoch": 0.13914964108227498,
+      "grad_norm": 0.8694303035736084,
+      "learning_rate": 4.9988817072443e-06,
+      "loss": 0.776,
+      "step": 168
+    },
+    {
+      "epoch": 0.13997791275538377,
+      "grad_norm": 0.8650498390197754,
+      "learning_rate": 4.998848576979135e-06,
+      "loss": 0.7941,
+      "step": 169
+    },
+    {
+      "epoch": 0.14080618442849255,
+      "grad_norm": 0.8246620893478394,
+      "learning_rate": 4.998814963210062e-06,
+      "loss": 0.7604,
+      "step": 170
+    },
+    {
+      "epoch": 0.14163445610160133,
+      "grad_norm": 0.8137409090995789,
+      "learning_rate": 4.99878086594359e-06,
+      "loss": 0.7681,
+      "step": 171
+    },
+    {
+      "epoch": 0.1424627277747101,
+      "grad_norm": 0.9755353927612305,
+      "learning_rate": 4.998746285186312e-06,
+      "loss": 0.7887,
+      "step": 172
+    },
+    {
+      "epoch": 0.1432909994478189,
+      "grad_norm": 0.8673408031463623,
+      "learning_rate": 4.998711220944921e-06,
+      "loss": 0.7803,
+      "step": 173
+    },
+    {
+      "epoch": 0.14411927112092768,
+      "grad_norm": 0.8312270045280457,
+      "learning_rate": 4.998675673226202e-06,
+      "loss": 0.7709,
+      "step": 174
+    },
+    {
+      "epoch": 0.14494754279403643,
+      "grad_norm": 0.8126094341278076,
+      "learning_rate": 4.998639642037032e-06,
+      "loss": 0.7822,
+      "step": 175
+    },
+    {
+      "epoch": 0.14577581446714521,
+      "grad_norm": 0.8561585545539856,
+      "learning_rate": 4.9986031273843834e-06,
+      "loss": 0.7829,
+      "step": 176
+    },
+    {
+      "epoch": 0.146604086140254,
+      "grad_norm": 0.8787612915039062,
+      "learning_rate": 4.9985661292753214e-06,
+      "loss": 0.7913,
+      "step": 177
+    },
+    {
+      "epoch": 0.14743235781336278,
+      "grad_norm": 0.8798345327377319,
+      "learning_rate": 4.998528647717006e-06,
+      "loss": 0.797,
+      "step": 178
+    },
+    {
+      "epoch": 0.14826062948647156,
+      "grad_norm": 0.7956589460372925,
+      "learning_rate": 4.998490682716687e-06,
+      "loss": 0.7571,
+      "step": 179
+    },
+    {
+      "epoch": 0.14908890115958034,
+      "grad_norm": 0.8505003452301025,
+      "learning_rate": 4.998452234281712e-06,
+      "loss": 0.7736,
+      "step": 180
+    },
+    {
+      "epoch": 0.14991717283268913,
+      "grad_norm": 0.8857074975967407,
+      "learning_rate": 4.998413302419519e-06,
+      "loss": 0.7863,
+      "step": 181
+    },
+    {
+      "epoch": 0.1507454445057979,
+      "grad_norm": 0.8762485980987549,
+      "learning_rate": 4.998373887137642e-06,
+      "loss": 0.7588,
+      "step": 182
+    },
+    {
+      "epoch": 0.1515737161789067,
+      "grad_norm": 0.8497205972671509,
+      "learning_rate": 4.998333988443709e-06,
+      "loss": 0.7773,
+      "step": 183
+    },
+    {
+      "epoch": 0.15240198785201547,
+      "grad_norm": 0.8720596432685852,
+      "learning_rate": 4.998293606345437e-06,
+      "loss": 0.7834,
+      "step": 184
+    },
+    {
+      "epoch": 0.15323025952512423,
+      "grad_norm": 0.8567108511924744,
+      "learning_rate": 4.998252740850641e-06,
+      "loss": 0.7794,
+      "step": 185
+    },
+    {
+      "epoch": 0.154058531198233,
+      "grad_norm": 0.8227219581604004,
+      "learning_rate": 4.998211391967228e-06,
+      "loss": 0.7688,
+      "step": 186
+    },
+    {
+      "epoch": 0.1548868028713418,
+      "grad_norm": 0.8321112394332886,
+      "learning_rate": 4.998169559703199e-06,
+      "loss": 0.7747,
+      "step": 187
+    },
+    {
+      "epoch": 0.15571507454445058,
+      "grad_norm": 0.8556115627288818,
+      "learning_rate": 4.9981272440666486e-06,
+      "loss": 0.7916,
+      "step": 188
+    },
+    {
+      "epoch": 0.15654334621755936,
+      "grad_norm": 0.8473951816558838,
+      "learning_rate": 4.998084445065763e-06,
+      "loss": 0.7823,
+      "step": 189
+    },
+    {
+      "epoch": 0.15737161789066814,
+      "grad_norm": 0.8551077246665955,
+      "learning_rate": 4.998041162708823e-06,
+      "loss": 0.7816,
+      "step": 190
+    },
+    {
+      "epoch": 0.15819988956377692,
+      "grad_norm": 0.945099413394928,
+      "learning_rate": 4.997997397004206e-06,
+      "loss": 0.7828,
+      "step": 191
+    },
+    {
+      "epoch": 0.1590281612368857,
+      "grad_norm": 0.8747237920761108,
+      "learning_rate": 4.9979531479603785e-06,
+      "loss": 0.7559,
+      "step": 192
+    },
+    {
+      "epoch": 0.1598564329099945,
+      "grad_norm": 0.8479958176612854,
+      "learning_rate": 4.997908415585902e-06,
+      "loss": 0.7694,
+      "step": 193
+    },
+    {
+      "epoch": 0.16068470458310327,
+      "grad_norm": 0.8659047484397888,
+      "learning_rate": 4.997863199889433e-06,
+      "loss": 0.7534,
+      "step": 194
+    },
+    {
+      "epoch": 0.16151297625621203,
+      "grad_norm": 0.8744727969169617,
+      "learning_rate": 4.997817500879719e-06,
+      "loss": 0.7499,
+      "step": 195
+    },
+    {
+      "epoch": 0.1623412479293208,
+      "grad_norm": 0.896468997001648,
+      "learning_rate": 4.997771318565603e-06,
+      "loss": 0.7974,
+      "step": 196
+    },
+    {
+      "epoch": 0.1631695196024296,
+      "grad_norm": 0.9058924913406372,
+      "learning_rate": 4.997724652956021e-06,
+      "loss": 0.7753,
+      "step": 197
+    },
+    {
+      "epoch": 0.16399779127553837,
+      "grad_norm": 0.8904663920402527,
+      "learning_rate": 4.9976775040600014e-06,
+      "loss": 0.7569,
+      "step": 198
+    },
+    {
+      "epoch": 0.16482606294864716,
+      "grad_norm": 0.8709238767623901,
+      "learning_rate": 4.997629871886669e-06,
+      "loss": 0.763,
+      "step": 199
+    },
+    {
+      "epoch": 0.16565433462175594,
+      "grad_norm": 0.8400478959083557,
+      "learning_rate": 4.997581756445238e-06,
+      "loss": 0.7795,
+      "step": 200
+    },
+    {
+      "epoch": 0.16648260629486472,
+      "grad_norm": 0.8458966612815857,
+      "learning_rate": 4.9975331577450204e-06,
+      "loss": 0.7597,
+      "step": 201
+    },
+    {
+      "epoch": 0.1673108779679735,
+      "grad_norm": 0.8707828521728516,
+      "learning_rate": 4.997484075795417e-06,
+      "loss": 0.759,
+      "step": 202
+    },
+    {
+      "epoch": 0.16813914964108229,
+      "grad_norm": 0.8774807453155518,
+      "learning_rate": 4.997434510605927e-06,
+      "loss": 0.7706,
+      "step": 203
+    },
+    {
+      "epoch": 0.16896742131419104,
+      "grad_norm": 0.8987929821014404,
+      "learning_rate": 4.99738446218614e-06,
+      "loss": 0.7319,
+      "step": 204
+    },
+    {
+      "epoch": 0.16979569298729982,
+      "grad_norm": 0.8869186639785767,
+      "learning_rate": 4.99733393054574e-06,
+      "loss": 0.7812,
+      "step": 205
+    },
+    {
+      "epoch": 0.1706239646604086,
+      "grad_norm": 0.860407829284668,
+      "learning_rate": 4.997282915694505e-06,
+      "loss": 0.7524,
+      "step": 206
+    },
+    {
+      "epoch": 0.1714522363335174,
+      "grad_norm": 0.8719788789749146,
+      "learning_rate": 4.9972314176423035e-06,
+      "loss": 0.7796,
+      "step": 207
+    },
+    {
+      "epoch": 0.17228050800662617,
+      "grad_norm": 0.8596137762069702,
+      "learning_rate": 4.997179436399103e-06,
+      "loss": 0.7554,
+      "step": 208
+    },
+    {
+      "epoch": 0.17310877967973495,
+      "grad_norm": 0.8505175709724426,
+      "learning_rate": 4.997126971974959e-06,
+      "loss": 0.7607,
+      "step": 209
+    },
+    {
+      "epoch": 0.17393705135284374,
+      "grad_norm": 0.7858585119247437,
+      "learning_rate": 4.9970740243800245e-06,
+      "loss": 0.7622,
+      "step": 210
+    },
+    {
+      "epoch": 0.17476532302595252,
+      "grad_norm": 0.8634621500968933,
+      "learning_rate": 4.997020593624543e-06,
+      "loss": 0.7659,
+      "step": 211
+    },
+    {
+      "epoch": 0.1755935946990613,
+      "grad_norm": 0.8650951385498047,
+      "learning_rate": 4.9969666797188545e-06,
+      "loss": 0.7689,
+      "step": 212
+    },
+    {
+      "epoch": 0.17642186637217008,
+      "grad_norm": 0.8602325916290283,
+      "learning_rate": 4.996912282673389e-06,
+      "loss": 0.7695,
+      "step": 213
+    },
+    {
+      "epoch": 0.17725013804527884,
+      "grad_norm": 0.8453742265701294,
+      "learning_rate": 4.9968574024986735e-06,
+      "loss": 0.7434,
+      "step": 214
+    },
+    {
+      "epoch": 0.17807840971838762,
+      "grad_norm": 0.8542691469192505,
+      "learning_rate": 4.9968020392053255e-06,
+      "loss": 0.7298,
+      "step": 215
+    },
+    {
+      "epoch": 0.1789066813914964,
+      "grad_norm": 0.8806673288345337,
+      "learning_rate": 4.996746192804058e-06,
+      "loss": 0.7649,
+      "step": 216
+    },
+    {
+      "epoch": 0.17973495306460519,
+      "grad_norm": 0.8451183438301086,
+      "learning_rate": 4.9966898633056765e-06,
+      "loss": 0.7681,
+      "step": 217
+    },
+    {
+      "epoch": 0.18056322473771397,
+      "grad_norm": 0.8664626479148865,
+      "learning_rate": 4.99663305072108e-06,
+      "loss": 0.7753,
+      "step": 218
+    },
+    {
+      "epoch": 0.18139149641082275,
+      "grad_norm": 0.8828144669532776,
+      "learning_rate": 4.996575755061262e-06,
+      "loss": 0.7514,
+      "step": 219
+    },
+    {
+      "epoch": 0.18221976808393153,
+      "grad_norm": 0.8493948578834534,
+      "learning_rate": 4.996517976337308e-06,
+      "loss": 0.7684,
+      "step": 220
+    },
+    {
+      "epoch": 0.18304803975704032,
+      "grad_norm": 0.8561757802963257,
+      "learning_rate": 4.9964597145603975e-06,
+      "loss": 0.7723,
+      "step": 221
+    },
+    {
+      "epoch": 0.1838763114301491,
+      "grad_norm": 0.8863773941993713,
+      "learning_rate": 4.996400969741805e-06,
+      "loss": 0.7732,
+      "step": 222
+    },
+    {
+      "epoch": 0.18470458310325788,
+      "grad_norm": 0.8571644425392151,
+      "learning_rate": 4.996341741892895e-06,
+      "loss": 0.744,
+      "step": 223
+    },
+    {
+      "epoch": 0.18553285477636663,
+      "grad_norm": 0.8622035980224609,
+      "learning_rate": 4.9962820310251286e-06,
+      "loss": 0.7607,
+      "step": 224
+    },
+    {
+      "epoch": 0.18636112644947542,
+      "grad_norm": 0.8904410600662231,
+      "learning_rate": 4.9962218371500595e-06,
+      "loss": 0.7648,
+      "step": 225
+    },
+    {
+      "epoch": 0.1871893981225842,
+      "grad_norm": 0.8975846171379089,
+      "learning_rate": 4.996161160279335e-06,
+      "loss": 0.7469,
+      "step": 226
+    },
+    {
+      "epoch": 0.18801766979569298,
+      "grad_norm": 0.8788724541664124,
+      "learning_rate": 4.996100000424694e-06,
+      "loss": 0.7473,
+      "step": 227
+    },
+    {
+      "epoch": 0.18884594146880176,
+      "grad_norm": 0.9446991086006165,
+      "learning_rate": 4.996038357597971e-06,
+      "loss": 0.7635,
+      "step": 228
+    },
+    {
+      "epoch": 0.18967421314191055,
+      "grad_norm": 0.86910480260849,
+      "learning_rate": 4.995976231811094e-06,
+      "loss": 0.7539,
+      "step": 229
+    },
+    {
+      "epoch": 0.19050248481501933,
+      "grad_norm": 0.862197756767273,
+      "learning_rate": 4.995913623076084e-06,
+      "loss": 0.7546,
+      "step": 230
+    },
+    {
+      "epoch": 0.1913307564881281,
+      "grad_norm": 0.906878650188446,
+      "learning_rate": 4.995850531405054e-06,
+      "loss": 0.7737,
+      "step": 231
+    },
+    {
+      "epoch": 0.1921590281612369,
+      "grad_norm": 0.8557183742523193,
+      "learning_rate": 4.995786956810212e-06,
+      "loss": 0.7663,
+      "step": 232
+    },
+    {
+      "epoch": 0.19298729983434568,
+      "grad_norm": 0.8125383853912354,
+      "learning_rate": 4.995722899303859e-06,
+      "loss": 0.7647,
+      "step": 233
+    },
+    {
+      "epoch": 0.19381557150745443,
+      "grad_norm": 0.8610332012176514,
+      "learning_rate": 4.995658358898391e-06,
+      "loss": 0.7541,
+      "step": 234
+    },
+    {
+      "epoch": 0.19464384318056321,
+      "grad_norm": 0.8620829582214355,
+      "learning_rate": 4.995593335606294e-06,
+      "loss": 0.7625,
+      "step": 235
+    },
+    {
+      "epoch": 0.195472114853672,
+      "grad_norm": 0.8296957612037659,
+      "learning_rate": 4.9955278294401496e-06,
+      "loss": 0.7536,
+      "step": 236
+    },
+    {
+      "epoch": 0.19630038652678078,
+      "grad_norm": 0.8974814414978027,
+      "learning_rate": 4.995461840412634e-06,
+      "loss": 0.7651,
+      "step": 237
+    },
+    {
+      "epoch": 0.19712865819988956,
+      "grad_norm": 0.8701095581054688,
+      "learning_rate": 4.995395368536515e-06,
+      "loss": 0.788,
+      "step": 238
+    },
+    {
+      "epoch": 0.19795692987299834,
+      "grad_norm": 0.8556636571884155,
+      "learning_rate": 4.995328413824653e-06,
+      "loss": 0.7649,
+      "step": 239
+    },
+    {
+      "epoch": 0.19878520154610713,
+      "grad_norm": 0.836686909198761,
+      "learning_rate": 4.995260976290005e-06,
+      "loss": 0.7583,
+      "step": 240
+    },
+    {
+      "epoch": 0.1996134732192159,
+      "grad_norm": 0.8807066082954407,
+      "learning_rate": 4.995193055945618e-06,
+      "loss": 0.7766,
+      "step": 241
+    },
+    {
+      "epoch": 0.2004417448923247,
+      "grad_norm": 0.8363582491874695,
+      "learning_rate": 4.995124652804635e-06,
+      "loss": 0.7655,
+      "step": 242
+    },
+    {
+      "epoch": 0.20127001656543347,
+      "grad_norm": 0.8699461817741394,
+      "learning_rate": 4.9950557668802905e-06,
+      "loss": 0.743,
+      "step": 243
+    },
+    {
+      "epoch": 0.20209828823854223,
+      "grad_norm": 0.8571060299873352,
+      "learning_rate": 4.994986398185914e-06,
+      "loss": 0.7512,
+      "step": 244
+    },
+    {
+      "epoch": 0.202926559911651,
+      "grad_norm": 0.8429675102233887,
+      "learning_rate": 4.994916546734927e-06,
+      "loss": 0.7589,
+      "step": 245
+    },
+    {
+      "epoch": 0.2037548315847598,
+      "grad_norm": 0.874822199344635,
+      "learning_rate": 4.994846212540846e-06,
+      "loss": 0.7534,
+      "step": 246
+    },
+    {
+      "epoch": 0.20458310325786858,
+      "grad_norm": 0.878341555595398,
+      "learning_rate": 4.994775395617279e-06,
+      "loss": 0.7709,
+      "step": 247
+    },
+    {
+      "epoch": 0.20541137493097736,
+      "grad_norm": 0.897961437702179,
+      "learning_rate": 4.99470409597793e-06,
+      "loss": 0.7768,
+      "step": 248
+    },
+    {
+      "epoch": 0.20623964660408614,
+      "grad_norm": 0.8439584374427795,
+      "learning_rate": 4.994632313636593e-06,
+      "loss": 0.7558,
+      "step": 249
+    },
+    {
+      "epoch": 0.20706791827719492,
+      "grad_norm": 0.8394468426704407,
+      "learning_rate": 4.994560048607159e-06,
+      "loss": 0.7492,
+      "step": 250
+    },
+    {
+      "epoch": 0.2078961899503037,
+      "grad_norm": 0.8489627242088318,
+      "learning_rate": 4.9944873009036074e-06,
+      "loss": 0.7365,
+      "step": 251
+    },
+    {
+      "epoch": 0.2087244616234125,
+      "grad_norm": 0.8675844669342041,
+      "learning_rate": 4.994414070540018e-06,
+      "loss": 0.762,
+      "step": 252
+    },
+    {
+      "epoch": 0.20955273329652127,
+      "grad_norm": 0.8228297829627991,
+      "learning_rate": 4.994340357530558e-06,
+      "loss": 0.7475,
+      "step": 253
+    },
+    {
+      "epoch": 0.21038100496963003,
+      "grad_norm": 0.8794129490852356,
+      "learning_rate": 4.9942661618894915e-06,
+      "loss": 0.7631,
+      "step": 254
+    },
+    {
+      "epoch": 0.2112092766427388,
+      "grad_norm": 0.8616135716438293,
+      "learning_rate": 4.9941914836311735e-06,
+      "loss": 0.7749,
+      "step": 255
+    },
+    {
+      "epoch": 0.2120375483158476,
+      "grad_norm": 0.8470495939254761,
+      "learning_rate": 4.994116322770054e-06,
+      "loss": 0.7484,
+      "step": 256
+    },
+    {
+      "epoch": 0.21286581998895637,
+      "grad_norm": 0.9346716403961182,
+      "learning_rate": 4.9940406793206755e-06,
+      "loss": 0.7557,
+      "step": 257
+    },
+    {
+      "epoch": 0.21369409166206516,
+      "grad_norm": 0.8993160724639893,
+      "learning_rate": 4.993964553297676e-06,
+      "loss": 0.7422,
+      "step": 258
+    },
+    {
+      "epoch": 0.21452236333517394,
+      "grad_norm": 0.8257086277008057,
+      "learning_rate": 4.993887944715783e-06,
+      "loss": 0.7783,
+      "step": 259
+    },
+    {
+      "epoch": 0.21535063500828272,
+      "grad_norm": 0.8950703144073486,
+      "learning_rate": 4.993810853589819e-06,
+      "loss": 0.7311,
+      "step": 260
+    },
+    {
+      "epoch": 0.2161789066813915,
+      "grad_norm": 0.8733702898025513,
+      "learning_rate": 4.993733279934704e-06,
+      "loss": 0.7572,
+      "step": 261
+    },
+    {
+      "epoch": 0.21700717835450029,
+      "grad_norm": 0.8322939276695251,
+      "learning_rate": 4.993655223765444e-06,
+      "loss": 0.7632,
+      "step": 262
+    },
+    {
+      "epoch": 0.21783545002760907,
+      "grad_norm": 0.8666137456893921,
+      "learning_rate": 4.993576685097145e-06,
+      "loss": 0.7619,
+      "step": 263
+    },
+    {
+      "epoch": 0.21866372170071782,
+      "grad_norm": 0.8770918250083923,
+      "learning_rate": 4.993497663945002e-06,
+      "loss": 0.7538,
+      "step": 264
+    },
+    {
+      "epoch": 0.2194919933738266,
+      "grad_norm": 0.8940248489379883,
+      "learning_rate": 4.9934181603243045e-06,
+      "loss": 0.7512,
+      "step": 265
+    },
+    {
+      "epoch": 0.2203202650469354,
+      "grad_norm": 0.8979714512825012,
+      "learning_rate": 4.993338174250437e-06,
+      "loss": 0.7504,
+      "step": 266
+    },
+    {
+      "epoch": 0.22114853672004417,
+      "grad_norm": 0.8802001476287842,
+      "learning_rate": 4.993257705738876e-06,
+      "loss": 0.7501,
+      "step": 267
+    },
+    {
+      "epoch": 0.22197680839315295,
+      "grad_norm": 0.8734432458877563,
+      "learning_rate": 4.993176754805189e-06,
+      "loss": 0.7452,
+      "step": 268
+    },
+    {
+      "epoch": 0.22280508006626173,
+      "grad_norm": 0.9127711057662964,
+      "learning_rate": 4.993095321465042e-06,
+      "loss": 0.7404,
+      "step": 269
+    },
+    {
+      "epoch": 0.22363335173937052,
+      "grad_norm": 0.8533952832221985,
+      "learning_rate": 4.993013405734191e-06,
+      "loss": 0.7382,
+      "step": 270
+    },
+    {
+      "epoch": 0.2244616234124793,
+      "grad_norm": 0.8463877439498901,
+      "learning_rate": 4.9929310076284845e-06,
+      "loss": 0.7574,
+      "step": 271
+    },
+    {
+      "epoch": 0.22528989508558808,
+      "grad_norm": 0.8598008155822754,
+      "learning_rate": 4.992848127163868e-06,
+      "loss": 0.7624,
+      "step": 272
+    },
+    {
+      "epoch": 0.22611816675869686,
+      "grad_norm": 0.8759154081344604,
+      "learning_rate": 4.992764764356376e-06,
+      "loss": 0.7637,
+      "step": 273
+    },
+    {
+      "epoch": 0.22694643843180562,
+      "grad_norm": 0.8880444765090942,
+      "learning_rate": 4.992680919222141e-06,
+      "loss": 0.7514,
+      "step": 274
+    },
+    {
+      "epoch": 0.2277747101049144,
+      "grad_norm": 0.871366024017334,
+      "learning_rate": 4.9925965917773826e-06,
+      "loss": 0.7569,
+      "step": 275
+    },
+    {
+      "epoch": 0.22860298177802318,
+      "grad_norm": 0.862654983997345,
+      "learning_rate": 4.99251178203842e-06,
+      "loss": 0.7349,
+      "step": 276
+    },
+    {
+      "epoch": 0.22943125345113197,
+      "grad_norm": 0.8615298867225647,
+      "learning_rate": 4.992426490021662e-06,
+      "loss": 0.7438,
+      "step": 277
+    },
+    {
+      "epoch": 0.23025952512424075,
+      "grad_norm": 0.8735869526863098,
+      "learning_rate": 4.992340715743614e-06,
+      "loss": 0.7755,
+      "step": 278
+    },
+    {
+      "epoch": 0.23108779679734953,
+      "grad_norm": 0.9082722067832947,
+      "learning_rate": 4.9922544592208695e-06,
+      "loss": 0.7561,
+      "step": 279
+    },
+    {
+      "epoch": 0.23191606847045831,
+      "grad_norm": 0.8639047145843506,
+      "learning_rate": 4.992167720470119e-06,
+      "loss": 0.7493,
+      "step": 280
+    },
+    {
+      "epoch": 0.2327443401435671,
+      "grad_norm": 0.891640305519104,
+      "learning_rate": 4.9920804995081475e-06,
+      "loss": 0.75,
+      "step": 281
+    },
+    {
+      "epoch": 0.23357261181667588,
+      "grad_norm": 0.8880137205123901,
+      "learning_rate": 4.99199279635183e-06,
+      "loss": 0.7801,
+      "step": 282
+    },
+    {
+      "epoch": 0.23440088348978466,
+      "grad_norm": 0.8600473403930664,
+      "learning_rate": 4.991904611018137e-06,
+      "loss": 0.7477,
+      "step": 283
+    },
+    {
+      "epoch": 0.23522915516289342,
+      "grad_norm": 0.8685399889945984,
+      "learning_rate": 4.99181594352413e-06,
+      "loss": 0.741,
+      "step": 284
+    },
+    {
+      "epoch": 0.2360574268360022,
+      "grad_norm": 0.9006181955337524,
+      "learning_rate": 4.991726793886966e-06,
+      "loss": 0.7288,
+      "step": 285
+    },
+    {
+      "epoch": 0.23688569850911098,
+      "grad_norm": 0.863895833492279,
+      "learning_rate": 4.991637162123896e-06,
+      "loss": 0.7375,
+      "step": 286
+    },
+    {
+      "epoch": 0.23771397018221976,
+      "grad_norm": 0.8620381951332092,
+      "learning_rate": 4.9915470482522625e-06,
+      "loss": 0.7249,
+      "step": 287
+    },
+    {
+      "epoch": 0.23854224185532855,
+      "grad_norm": 0.9634964466094971,
+      "learning_rate": 4.991456452289499e-06,
+      "loss": 0.7313,
+      "step": 288
+    },
+    {
+      "epoch": 0.23937051352843733,
+      "grad_norm": 0.9154514670372009,
+      "learning_rate": 4.9913653742531385e-06,
+      "loss": 0.782,
+      "step": 289
+    },
+    {
+      "epoch": 0.2401987852015461,
+      "grad_norm": 0.9179158806800842,
+      "learning_rate": 4.991273814160802e-06,
+      "loss": 0.7388,
+      "step": 290
+    },
+    {
+      "epoch": 0.2410270568746549,
+      "grad_norm": 0.8609046936035156,
+      "learning_rate": 4.991181772030206e-06,
+      "loss": 0.7393,
+      "step": 291
+    },
+    {
+      "epoch": 0.24185532854776368,
+      "grad_norm": 0.8689826130867004,
+      "learning_rate": 4.991089247879159e-06,
+      "loss": 0.7302,
+      "step": 292
+    },
+    {
+      "epoch": 0.24268360022087246,
+      "grad_norm": 0.8414047956466675,
+      "learning_rate": 4.990996241725565e-06,
+      "loss": 0.7496,
+      "step": 293
+    },
+    {
+      "epoch": 0.2435118718939812,
+      "grad_norm": 0.8621289134025574,
+      "learning_rate": 4.990902753587418e-06,
+      "loss": 0.7317,
+      "step": 294
+    },
+    {
+      "epoch": 0.24434014356709,
+      "grad_norm": 0.856566309928894,
+      "learning_rate": 4.990808783482809e-06,
+      "loss": 0.7169,
+      "step": 295
+    },
+    {
+      "epoch": 0.24516841524019878,
+      "grad_norm": 0.8549672961235046,
+      "learning_rate": 4.990714331429919e-06,
+      "loss": 0.7357,
+      "step": 296
+    },
+    {
+      "epoch": 0.24599668691330756,
+      "grad_norm": 0.8813024759292603,
+      "learning_rate": 4.990619397447024e-06,
+      "loss": 0.7437,
+      "step": 297
+    },
+    {
+      "epoch": 0.24682495858641634,
+      "grad_norm": 0.9008268713951111,
+      "learning_rate": 4.990523981552493e-06,
+      "loss": 0.7652,
+      "step": 298
+    },
+    {
+      "epoch": 0.24765323025952513,
+      "grad_norm": 0.910277783870697,
+      "learning_rate": 4.990428083764788e-06,
+      "loss": 0.7598,
+      "step": 299
+    },
+    {
+      "epoch": 0.2484815019326339,
+      "grad_norm": 0.9261599183082581,
+      "learning_rate": 4.990331704102464e-06,
+      "loss": 0.7199,
+      "step": 300
+    },
+    {
+      "epoch": 0.2493097736057427,
+      "grad_norm": 0.9166674613952637,
+      "learning_rate": 4.99023484258417e-06,
+      "loss": 0.7414,
+      "step": 301
+    },
+    {
+      "epoch": 0.2501380452788515,
+      "grad_norm": 0.8701214790344238,
+      "learning_rate": 4.990137499228648e-06,
+      "loss": 0.7326,
+      "step": 302
+    },
+    {
+      "epoch": 0.25096631695196026,
+      "grad_norm": 0.9120466113090515,
+      "learning_rate": 4.990039674054732e-06,
+      "loss": 0.7574,
+      "step": 303
+    },
+    {
+      "epoch": 0.25179458862506904,
+      "grad_norm": 0.9449286460876465,
+      "learning_rate": 4.989941367081351e-06,
+      "loss": 0.725,
+      "step": 304
+    },
+    {
+      "epoch": 0.2526228602981778,
+      "grad_norm": 0.8989110589027405,
+      "learning_rate": 4.989842578327526e-06,
+      "loss": 0.7569,
+      "step": 305
+    },
+    {
+      "epoch": 0.2534511319712866,
+      "grad_norm": 0.891677975654602,
+      "learning_rate": 4.989743307812373e-06,
+      "loss": 0.7135,
+      "step": 306
+    },
+    {
+      "epoch": 0.2542794036443954,
+      "grad_norm": 0.8689535856246948,
+      "learning_rate": 4.989643555555099e-06,
+      "loss": 0.7439,
+      "step": 307
+    },
+    {
+      "epoch": 0.25510767531750417,
+      "grad_norm": 0.8779630064964294,
+      "learning_rate": 4.989543321575004e-06,
+      "loss": 0.7322,
+      "step": 308
+    },
+    {
+      "epoch": 0.2559359469906129,
+      "grad_norm": 0.8977622985839844,
+      "learning_rate": 4.9894426058914856e-06,
+      "loss": 0.7544,
+      "step": 309
+    },
+    {
+      "epoch": 0.2567642186637217,
+      "grad_norm": 0.8216757774353027,
+      "learning_rate": 4.989341408524028e-06,
+      "loss": 0.7344,
+      "step": 310
+    },
+    {
+      "epoch": 0.25759249033683046,
+      "grad_norm": 0.8494759798049927,
+      "learning_rate": 4.9892397294922126e-06,
+      "loss": 0.7389,
+      "step": 311
+    },
+    {
+      "epoch": 0.25842076200993924,
+      "grad_norm": 0.881244957447052,
+      "learning_rate": 4.989137568815715e-06,
+      "loss": 0.7121,
+      "step": 312
+    },
+    {
+      "epoch": 0.259249033683048,
+      "grad_norm": 0.8679410219192505,
+      "learning_rate": 4.9890349265143005e-06,
+      "loss": 0.7469,
+      "step": 313
+    },
+    {
+      "epoch": 0.2600773053561568,
+      "grad_norm": 0.9746971130371094,
+      "learning_rate": 4.988931802607831e-06,
+      "loss": 0.7585,
+      "step": 314
+    },
+    {
+      "epoch": 0.2609055770292656,
+      "grad_norm": 0.9222471117973328,
+      "learning_rate": 4.988828197116259e-06,
+      "loss": 0.761,
+      "step": 315
+    },
+    {
+      "epoch": 0.26173384870237437,
+      "grad_norm": 0.9104124903678894,
+      "learning_rate": 4.988724110059631e-06,
+      "loss": 0.7652,
+      "step": 316
+    },
+    {
+      "epoch": 0.26256212037548315,
+      "grad_norm": 0.8876475095748901,
+      "learning_rate": 4.988619541458088e-06,
+      "loss": 0.7529,
+      "step": 317
+    },
+    {
+      "epoch": 0.26339039204859194,
+      "grad_norm": 0.8705421686172485,
+      "learning_rate": 4.988514491331861e-06,
+      "loss": 0.7437,
+      "step": 318
+    },
+    {
+      "epoch": 0.2642186637217007,
+      "grad_norm": 0.8594399690628052,
+      "learning_rate": 4.9884089597012784e-06,
+      "loss": 0.7406,
+      "step": 319
+    },
+    {
+      "epoch": 0.2650469353948095,
+      "grad_norm": 0.8807039856910706,
+      "learning_rate": 4.988302946586759e-06,
+      "loss": 0.7429,
+      "step": 320
+    },
+    {
+      "epoch": 0.2658752070679183,
+      "grad_norm": 0.8881412148475647,
+      "learning_rate": 4.9881964520088144e-06,
+      "loss": 0.7234,
+      "step": 321
+    },
+    {
+      "epoch": 0.26670347874102707,
+      "grad_norm": 0.8764941096305847,
+      "learning_rate": 4.988089475988052e-06,
+      "loss": 0.7514,
+      "step": 322
+    },
+    {
+      "epoch": 0.26753175041413585,
+      "grad_norm": 0.8992927670478821,
+      "learning_rate": 4.987982018545169e-06,
+      "loss": 0.7143,
+      "step": 323
+    },
+    {
+      "epoch": 0.26836002208724463,
+      "grad_norm": 0.8961546421051025,
+      "learning_rate": 4.9878740797009585e-06,
+      "loss": 0.7396,
+      "step": 324
+    },
+    {
+      "epoch": 0.2691882937603534,
+      "grad_norm": 0.9031578302383423,
+      "learning_rate": 4.987765659476304e-06,
+      "loss": 0.7407,
+      "step": 325
+    },
+    {
+      "epoch": 0.2700165654334622,
+      "grad_norm": 0.8693655729293823,
+      "learning_rate": 4.987656757892186e-06,
+      "loss": 0.7215,
+      "step": 326
+    },
+    {
+      "epoch": 0.270844837106571,
+      "grad_norm": 0.8739856481552124,
+      "learning_rate": 4.987547374969676e-06,
+      "loss": 0.7471,
+      "step": 327
+    },
+    {
+      "epoch": 0.27167310877967976,
+      "grad_norm": 0.8916322588920593,
+      "learning_rate": 4.987437510729936e-06,
+      "loss": 0.7459,
+      "step": 328
+    },
+    {
+      "epoch": 0.2725013804527885,
+      "grad_norm": 0.890178382396698,
+      "learning_rate": 4.987327165194225e-06,
+      "loss": 0.7322,
+      "step": 329
+    },
+    {
+      "epoch": 0.27332965212589727,
+      "grad_norm": 0.8611598014831543,
+      "learning_rate": 4.987216338383896e-06,
+      "loss": 0.733,
+      "step": 330
+    },
+    {
+      "epoch": 0.27415792379900605,
+      "grad_norm": 0.87791907787323,
+      "learning_rate": 4.98710503032039e-06,
+      "loss": 0.7445,
+      "step": 331
+    },
+    {
+      "epoch": 0.27498619547211484,
+      "grad_norm": 0.963479220867157,
+      "learning_rate": 4.986993241025245e-06,
+      "loss": 0.7376,
+      "step": 332
+    },
+    {
+      "epoch": 0.2758144671452236,
+      "grad_norm": 0.9083762764930725,
+      "learning_rate": 4.986880970520092e-06,
+      "loss": 0.7258,
+      "step": 333
+    },
+    {
+      "epoch": 0.2766427388183324,
+      "grad_norm": 1.048991084098816,
+      "learning_rate": 4.9867682188266524e-06,
+      "loss": 0.7399,
+      "step": 334
+    },
+    {
+      "epoch": 0.2774710104914412,
+      "grad_norm": 0.9404491782188416,
+      "learning_rate": 4.986654985966745e-06,
+      "loss": 0.7133,
+      "step": 335
+    },
+    {
+      "epoch": 0.27829928216454997,
+      "grad_norm": 0.9168346524238586,
+      "learning_rate": 4.986541271962278e-06,
+      "loss": 0.74,
+      "step": 336
+    },
+    {
+      "epoch": 0.27912755383765875,
+      "grad_norm": 0.8716689944267273,
+      "learning_rate": 4.986427076835253e-06,
+      "loss": 0.7249,
+      "step": 337
+    },
+    {
+      "epoch": 0.27995582551076753,
+      "grad_norm": 0.9097291827201843,
+      "learning_rate": 4.986312400607767e-06,
+      "loss": 0.7151,
+      "step": 338
+    },
+    {
+      "epoch": 0.2807840971838763,
+      "grad_norm": 0.8621960282325745,
+      "learning_rate": 4.98619724330201e-06,
+      "loss": 0.7125,
+      "step": 339
+    },
+    {
+      "epoch": 0.2816123688569851,
+      "grad_norm": 0.9823791980743408,
+      "learning_rate": 4.986081604940261e-06,
+      "loss": 0.7191,
+      "step": 340
+    },
+    {
+      "epoch": 0.2824406405300939,
+      "grad_norm": 0.8999047875404358,
+      "learning_rate": 4.9859654855448966e-06,
+      "loss": 0.7175,
+      "step": 341
+    },
+    {
+      "epoch": 0.28326891220320266,
+      "grad_norm": 0.8690859079360962,
+      "learning_rate": 4.985848885138386e-06,
+      "loss": 0.7309,
+      "step": 342
+    },
+    {
+      "epoch": 0.28409718387631144,
+      "grad_norm": 0.889500617980957,
+      "learning_rate": 4.985731803743287e-06,
+      "loss": 0.7391,
+      "step": 343
+    },
+    {
+      "epoch": 0.2849254555494202,
+      "grad_norm": 0.8992016911506653,
+      "learning_rate": 4.985614241382257e-06,
+      "loss": 0.7437,
+      "step": 344
+    },
+    {
+      "epoch": 0.285753727222529,
+      "grad_norm": 0.9100068211555481,
+      "learning_rate": 4.985496198078041e-06,
+      "loss": 0.722,
+      "step": 345
+    },
+    {
+      "epoch": 0.2865819988956378,
+      "grad_norm": 0.9013107419013977,
+      "learning_rate": 4.985377673853481e-06,
+      "loss": 0.7434,
+      "step": 346
+    },
+    {
+      "epoch": 0.2874102705687466,
+      "grad_norm": 0.8794704675674438,
+      "learning_rate": 4.9852586687315094e-06,
+      "loss": 0.7387,
+      "step": 347
+    },
+    {
+      "epoch": 0.28823854224185536,
+      "grad_norm": 0.9288201928138733,
+      "learning_rate": 4.9851391827351526e-06,
+      "loss": 0.7165,
+      "step": 348
+    },
+    {
+      "epoch": 0.2890668139149641,
+      "grad_norm": 0.8735847473144531,
+      "learning_rate": 4.98501921588753e-06,
+      "loss": 0.7479,
+      "step": 349
+    },
+    {
+      "epoch": 0.28989508558807286,
+      "grad_norm": 0.890616238117218,
+      "learning_rate": 4.984898768211853e-06,
+      "loss": 0.76,
+      "step": 350
+    },
+    {
+      "epoch": 0.29072335726118165,
+      "grad_norm": 0.9181028604507446,
+      "learning_rate": 4.98477783973143e-06,
+      "loss": 0.7366,
+      "step": 351
+    },
+    {
+      "epoch": 0.29155162893429043,
+      "grad_norm": 0.9947429299354553,
+      "learning_rate": 4.984656430469657e-06,
+      "loss": 0.7199,
+      "step": 352
+    },
+    {
+      "epoch": 0.2923799006073992,
+      "grad_norm": 0.8730769157409668,
+      "learning_rate": 4.984534540450027e-06,
+      "loss": 0.7153,
+      "step": 353
+    },
+    {
+      "epoch": 0.293208172280508,
+      "grad_norm": 0.8675997853279114,
+      "learning_rate": 4.9844121696961235e-06,
+      "loss": 0.7196,
+      "step": 354
+    },
+    {
+      "epoch": 0.2940364439536168,
+      "grad_norm": 0.9558542966842651,
+      "learning_rate": 4.984289318231624e-06,
+      "loss": 0.7311,
+      "step": 355
+    },
+    {
+      "epoch": 0.29486471562672556,
+      "grad_norm": 0.8817659616470337,
+      "learning_rate": 4.9841659860803e-06,
+      "loss": 0.7367,
+      "step": 356
+    },
+    {
+      "epoch": 0.29569298729983434,
+      "grad_norm": 0.8679526448249817,
+      "learning_rate": 4.984042173266016e-06,
+      "loss": 0.719,
+      "step": 357
+    },
+    {
+      "epoch": 0.2965212589729431,
+      "grad_norm": 0.9102020859718323,
+      "learning_rate": 4.983917879812726e-06,
+      "loss": 0.7301,
+      "step": 358
+    },
+    {
+      "epoch": 0.2973495306460519,
+      "grad_norm": 0.8987215757369995,
+      "learning_rate": 4.983793105744482e-06,
+      "loss": 0.7557,
+      "step": 359
+    },
+    {
+      "epoch": 0.2981778023191607,
+      "grad_norm": 0.8779832720756531,
+      "learning_rate": 4.983667851085426e-06,
+      "loss": 0.7471,
+      "step": 360
+    },
+    {
+      "epoch": 0.29900607399226947,
+      "grad_norm": 0.8762708306312561,
+      "learning_rate": 4.983542115859792e-06,
+      "loss": 0.7676,
+      "step": 361
+    },
+    {
+      "epoch": 0.29983434566537825,
+      "grad_norm": 0.910353422164917,
+      "learning_rate": 4.98341590009191e-06,
+      "loss": 0.7285,
+      "step": 362
+    },
+    {
+      "epoch": 0.30066261733848704,
+      "grad_norm": 0.9385405778884888,
+      "learning_rate": 4.983289203806202e-06,
+      "loss": 0.7142,
+      "step": 363
+    },
+    {
+      "epoch": 0.3014908890115958,
+      "grad_norm": 1.016781210899353,
+      "learning_rate": 4.983162027027182e-06,
+      "loss": 0.7408,
+      "step": 364
+    },
+    {
+      "epoch": 0.3023191606847046,
+      "grad_norm": 0.8642508387565613,
+      "learning_rate": 4.983034369779457e-06,
+      "loss": 0.7397,
+      "step": 365
+    },
+    {
+      "epoch": 0.3031474323578134,
+      "grad_norm": 0.9469170570373535,
+      "learning_rate": 4.982906232087728e-06,
+      "loss": 0.7158,
+      "step": 366
+    },
+    {
+      "epoch": 0.30397570403092217,
+      "grad_norm": 0.8818212151527405,
+      "learning_rate": 4.982777613976789e-06,
+      "loss": 0.7408,
+      "step": 367
+    },
+    {
+      "epoch": 0.30480397570403095,
+      "grad_norm": 0.8639838099479675,
+      "learning_rate": 4.982648515471526e-06,
+      "loss": 0.7217,
+      "step": 368
+    },
+    {
+      "epoch": 0.3056322473771397,
+      "grad_norm": 0.923780083656311,
+      "learning_rate": 4.982518936596917e-06,
+      "loss": 0.7284,
+      "step": 369
+    },
+    {
+      "epoch": 0.30646051905024846,
+      "grad_norm": 0.8992834687232971,
+      "learning_rate": 4.982388877378037e-06,
+      "loss": 0.7395,
+      "step": 370
+    },
+    {
+      "epoch": 0.30728879072335724,
+      "grad_norm": 0.8671404123306274,
+      "learning_rate": 4.982258337840049e-06,
+      "loss": 0.7157,
+      "step": 371
+    },
+    {
+      "epoch": 0.308117062396466,
+      "grad_norm": 0.8689168691635132,
+      "learning_rate": 4.982127318008212e-06,
+      "loss": 0.7398,
+      "step": 372
+    },
+    {
+      "epoch": 0.3089453340695748,
+      "grad_norm": 0.8964621424674988,
+      "learning_rate": 4.9819958179078765e-06,
+      "loss": 0.7385,
+      "step": 373
+    },
+    {
+      "epoch": 0.3097736057426836,
+      "grad_norm": 0.8984093070030212,
+      "learning_rate": 4.981863837564488e-06,
+      "loss": 0.7253,
+      "step": 374
+    },
+    {
+      "epoch": 0.31060187741579237,
+      "grad_norm": 0.8776745200157166,
+      "learning_rate": 4.981731377003583e-06,
+      "loss": 0.7126,
+      "step": 375
+    },
+    {
+      "epoch": 0.31143014908890115,
+      "grad_norm": 0.8782477378845215,
+      "learning_rate": 4.98159843625079e-06,
+      "loss": 0.7164,
+      "step": 376
+    },
+    {
+      "epoch": 0.31225842076200994,
+      "grad_norm": 0.8993703722953796,
+      "learning_rate": 4.981465015331832e-06,
+      "loss": 0.7391,
+      "step": 377
+    },
+    {
+      "epoch": 0.3130866924351187,
+      "grad_norm": 0.8975916504859924,
+      "learning_rate": 4.981331114272527e-06,
+      "loss": 0.7385,
+      "step": 378
+    },
+    {
+      "epoch": 0.3139149641082275,
+      "grad_norm": 0.8726824522018433,
+      "learning_rate": 4.9811967330987816e-06,
+      "loss": 0.7426,
+      "step": 379
+    },
+    {
+      "epoch": 0.3147432357813363,
+      "grad_norm": 0.8741514682769775,
+      "learning_rate": 4.981061871836597e-06,
+      "loss": 0.7246,
+      "step": 380
+    },
+    {
+      "epoch": 0.31557150745444507,
+      "grad_norm": 0.8807122707366943,
+      "learning_rate": 4.98092653051207e-06,
+      "loss": 0.7364,
+      "step": 381
+    },
+    {
+      "epoch": 0.31639977912755385,
+      "grad_norm": 0.835006058216095,
+      "learning_rate": 4.980790709151385e-06,
+      "loss": 0.7248,
+      "step": 382
+    },
+    {
+      "epoch": 0.31722805080066263,
+      "grad_norm": 0.9473997354507446,
+      "learning_rate": 4.980654407780824e-06,
+      "loss": 0.7331,
+      "step": 383
+    },
+    {
+      "epoch": 0.3180563224737714,
+      "grad_norm": 0.8332815170288086,
+      "learning_rate": 4.980517626426758e-06,
+      "loss": 0.7299,
+      "step": 384
+    },
+    {
+      "epoch": 0.3188845941468802,
+      "grad_norm": 0.8536360859870911,
+      "learning_rate": 4.980380365115656e-06,
+      "loss": 0.7095,
+      "step": 385
+    },
+    {
+      "epoch": 0.319712865819989,
+      "grad_norm": 0.9488157629966736,
+      "learning_rate": 4.980242623874075e-06,
+      "loss": 0.7382,
+      "step": 386
+    },
+    {
+      "epoch": 0.32054113749309776,
+      "grad_norm": 0.8623067736625671,
+      "learning_rate": 4.980104402728666e-06,
+      "loss": 0.7187,
+      "step": 387
+    },
+    {
+      "epoch": 0.32136940916620654,
+      "grad_norm": 0.9110252857208252,
+      "learning_rate": 4.979965701706175e-06,
+      "loss": 0.7354,
+      "step": 388
+    },
+    {
+      "epoch": 0.32219768083931527,
+      "grad_norm": 0.8966642618179321,
+      "learning_rate": 4.979826520833439e-06,
+      "loss": 0.7251,
+      "step": 389
+    },
+    {
+      "epoch": 0.32302595251242405,
+      "grad_norm": 0.856674075126648,
+      "learning_rate": 4.979686860137387e-06,
+      "loss": 0.723,
+      "step": 390
+    },
+    {
+      "epoch": 0.32385422418553284,
+      "grad_norm": 0.8623953461647034,
+      "learning_rate": 4.979546719645043e-06,
+      "loss": 0.7224,
+      "step": 391
+    },
+    {
+      "epoch": 0.3246824958586416,
+      "grad_norm": 0.8685624599456787,
+      "learning_rate": 4.979406099383523e-06,
+      "loss": 0.7283,
+      "step": 392
+    },
+    {
+      "epoch": 0.3255107675317504,
+      "grad_norm": 0.8610525727272034,
+      "learning_rate": 4.979264999380035e-06,
+      "loss": 0.7385,
+      "step": 393
+    },
+    {
+      "epoch": 0.3263390392048592,
+      "grad_norm": 0.8860796093940735,
+      "learning_rate": 4.979123419661882e-06,
+      "loss": 0.743,
+      "step": 394
+    },
+    {
+      "epoch": 0.32716731087796797,
+      "grad_norm": 0.8748682737350464,
+      "learning_rate": 4.978981360256457e-06,
+      "loss": 0.7447,
+      "step": 395
+    },
+    {
+      "epoch": 0.32799558255107675,
+      "grad_norm": 0.8750934600830078,
+      "learning_rate": 4.978838821191247e-06,
+      "loss": 0.7259,
+      "step": 396
+    },
+    {
+      "epoch": 0.32882385422418553,
+      "grad_norm": 0.9335777163505554,
+      "learning_rate": 4.978695802493834e-06,
+      "loss": 0.7271,
+      "step": 397
+    },
+    {
+      "epoch": 0.3296521258972943,
+      "grad_norm": 0.8823036551475525,
+      "learning_rate": 4.978552304191888e-06,
+      "loss": 0.723,
+      "step": 398
+    },
+    {
+      "epoch": 0.3304803975704031,
+      "grad_norm": 0.9503542184829712,
+      "learning_rate": 4.9784083263131764e-06,
+      "loss": 0.752,
+      "step": 399
+    },
+    {
+      "epoch": 0.3313086692435119,
+      "grad_norm": 0.8779938220977783,
+      "learning_rate": 4.978263868885557e-06,
+      "loss": 0.7098,
+      "step": 400
+    },
+    {
+      "epoch": 0.33213694091662066,
+      "grad_norm": 0.873857855796814,
+      "learning_rate": 4.978118931936981e-06,
+      "loss": 0.7143,
+      "step": 401
+    },
+    {
+      "epoch": 0.33296521258972944,
+      "grad_norm": 0.8437919020652771,
+      "learning_rate": 4.9779735154954924e-06,
+      "loss": 0.7275,
+      "step": 402
+    },
+    {
+      "epoch": 0.3337934842628382,
+      "grad_norm": 0.8914555311203003,
+      "learning_rate": 4.977827619589229e-06,
+      "loss": 0.722,
+      "step": 403
+    },
+    {
+      "epoch": 0.334621755935947,
+      "grad_norm": 0.9062597751617432,
+      "learning_rate": 4.977681244246417e-06,
+      "loss": 0.7067,
+      "step": 404
+    },
+    {
+      "epoch": 0.3354500276090558,
+      "grad_norm": 0.8555092811584473,
+      "learning_rate": 4.977534389495382e-06,
+      "loss": 0.7318,
+      "step": 405
+    },
+    {
+      "epoch": 0.33627829928216457,
+      "grad_norm": 0.8717684745788574,
+      "learning_rate": 4.9773870553645386e-06,
+      "loss": 0.7398,
+      "step": 406
+    },
+    {
+      "epoch": 0.33710657095527335,
+      "grad_norm": 0.8746877908706665,
+      "learning_rate": 4.977239241882393e-06,
+      "loss": 0.7154,
+      "step": 407
+    },
+    {
+      "epoch": 0.3379348426283821,
+      "grad_norm": 0.8905707597732544,
+      "learning_rate": 4.977090949077547e-06,
+      "loss": 0.7345,
+      "step": 408
+    },
+    {
+      "epoch": 0.33876311430149086,
+      "grad_norm": 0.9037347435951233,
+      "learning_rate": 4.9769421769786935e-06,
+      "loss": 0.7271,
+      "step": 409
+    },
+    {
+      "epoch": 0.33959138597459965,
+      "grad_norm": 0.8588722944259644,
+      "learning_rate": 4.976792925614619e-06,
+      "loss": 0.7335,
+      "step": 410
+    },
+    {
+      "epoch": 0.34041965764770843,
+      "grad_norm": 0.8786173462867737,
+      "learning_rate": 4.9766431950142005e-06,
+      "loss": 0.7254,
+      "step": 411
+    },
+    {
+      "epoch": 0.3412479293208172,
+      "grad_norm": 0.9019079804420471,
+      "learning_rate": 4.976492985206412e-06,
+      "loss": 0.715,
+      "step": 412
+    },
+    {
+      "epoch": 0.342076200993926,
+      "grad_norm": 0.878132700920105,
+      "learning_rate": 4.976342296220316e-06,
+      "loss": 0.7063,
+      "step": 413
+    },
+    {
+      "epoch": 0.3429044726670348,
+      "grad_norm": 0.8722684979438782,
+      "learning_rate": 4.976191128085069e-06,
+      "loss": 0.7481,
+      "step": 414
+    },
+    {
+      "epoch": 0.34373274434014356,
+      "grad_norm": 0.9076220393180847,
+      "learning_rate": 4.976039480829922e-06,
+      "loss": 0.7197,
+      "step": 415
+    },
+    {
+      "epoch": 0.34456101601325234,
+      "grad_norm": 0.8674967288970947,
+      "learning_rate": 4.975887354484217e-06,
+      "loss": 0.7282,
+      "step": 416
+    },
+    {
+      "epoch": 0.3453892876863611,
+      "grad_norm": 0.9406816959381104,
+      "learning_rate": 4.9757347490773885e-06,
+      "loss": 0.7222,
+      "step": 417
+    },
+    {
+      "epoch": 0.3462175593594699,
+      "grad_norm": 0.8747051358222961,
+      "learning_rate": 4.975581664638965e-06,
+      "loss": 0.7283,
+      "step": 418
+    },
+    {
+      "epoch": 0.3470458310325787,
+      "grad_norm": 1.0587687492370605,
+      "learning_rate": 4.9754281011985666e-06,
+      "loss": 0.7272,
+      "step": 419
+    },
+    {
+      "epoch": 0.34787410270568747,
+      "grad_norm": 0.8596946597099304,
+      "learning_rate": 4.975274058785906e-06,
+      "loss": 0.7407,
+      "step": 420
+    },
+    {
+      "epoch": 0.34870237437879625,
+      "grad_norm": 0.9230688214302063,
+      "learning_rate": 4.975119537430789e-06,
+      "loss": 0.7303,
+      "step": 421
+    },
+    {
+      "epoch": 0.34953064605190504,
+      "grad_norm": 0.9276477694511414,
+      "learning_rate": 4.974964537163115e-06,
+      "loss": 0.7296,
+      "step": 422
+    },
+    {
+      "epoch": 0.3503589177250138,
+      "grad_norm": 0.8969418406486511,
+      "learning_rate": 4.974809058012874e-06,
+      "loss": 0.7275,
+      "step": 423
+    },
+    {
+      "epoch": 0.3511871893981226,
+      "grad_norm": 0.8454467058181763,
+      "learning_rate": 4.97465310001015e-06,
+      "loss": 0.7144,
+      "step": 424
+    },
+    {
+      "epoch": 0.3520154610712314,
+      "grad_norm": 0.8833656907081604,
+      "learning_rate": 4.97449666318512e-06,
+      "loss": 0.7073,
+      "step": 425
+    },
+    {
+      "epoch": 0.35284373274434017,
+      "grad_norm": 0.9251653552055359,
+      "learning_rate": 4.974339747568052e-06,
+      "loss": 0.7351,
+      "step": 426
+    },
+    {
+      "epoch": 0.35367200441744895,
+      "grad_norm": 0.8938575983047485,
+      "learning_rate": 4.974182353189309e-06,
+      "loss": 0.7031,
+      "step": 427
+    },
+    {
+      "epoch": 0.3545002760905577,
+      "grad_norm": 0.8487107753753662,
+      "learning_rate": 4.974024480079343e-06,
+      "loss": 0.704,
+      "step": 428
+    },
+    {
+      "epoch": 0.35532854776366646,
+      "grad_norm": 0.8515727519989014,
+      "learning_rate": 4.973866128268705e-06,
+      "loss": 0.7092,
+      "step": 429
+    },
+    {
+      "epoch": 0.35615681943677524,
+      "grad_norm": 0.8869542479515076,
+      "learning_rate": 4.97370729778803e-06,
+      "loss": 0.7099,
+      "step": 430
+    },
+    {
+      "epoch": 0.356985091109884,
+      "grad_norm": 0.8830757141113281,
+      "learning_rate": 4.973547988668055e-06,
+      "loss": 0.7294,
+      "step": 431
+    },
+    {
+      "epoch": 0.3578133627829928,
+      "grad_norm": 0.9321587681770325,
+      "learning_rate": 4.973388200939601e-06,
+      "loss": 0.7276,
+      "step": 432
+    },
+    {
+      "epoch": 0.3586416344561016,
+      "grad_norm": 0.9408143758773804,
+      "learning_rate": 4.9732279346335866e-06,
+      "loss": 0.7216,
+      "step": 433
+    },
+    {
+      "epoch": 0.35946990612921037,
+      "grad_norm": 0.9189308285713196,
+      "learning_rate": 4.973067189781022e-06,
+      "loss": 0.7079,
+      "step": 434
+    },
+    {
+      "epoch": 0.36029817780231915,
+      "grad_norm": 0.9198120832443237,
+      "learning_rate": 4.97290596641301e-06,
+      "loss": 0.7188,
+      "step": 435
+    },
+    {
+      "epoch": 0.36112644947542794,
+      "grad_norm": 0.9123285412788391,
+      "learning_rate": 4.972744264560746e-06,
+      "loss": 0.7328,
+      "step": 436
+    },
+    {
+      "epoch": 0.3619547211485367,
+      "grad_norm": 0.8722008466720581,
+      "learning_rate": 4.972582084255517e-06,
+      "loss": 0.7228,
+      "step": 437
+    },
+    {
+      "epoch": 0.3627829928216455,
+      "grad_norm": 1.035325288772583,
+      "learning_rate": 4.972419425528704e-06,
+      "loss": 0.72,
+      "step": 438
+    },
+    {
+      "epoch": 0.3636112644947543,
+      "grad_norm": 0.9577400088310242,
+      "learning_rate": 4.97225628841178e-06,
+      "loss": 0.7382,
+      "step": 439
+    },
+    {
+      "epoch": 0.36443953616786307,
+      "grad_norm": 0.9206854104995728,
+      "learning_rate": 4.972092672936309e-06,
+      "loss": 0.7294,
+      "step": 440
+    },
+    {
+      "epoch": 0.36526780784097185,
+      "grad_norm": 0.9281194806098938,
+      "learning_rate": 4.971928579133952e-06,
+      "loss": 0.7137,
+      "step": 441
+    },
+    {
+      "epoch": 0.36609607951408063,
+      "grad_norm": 0.951049268245697,
+      "learning_rate": 4.971764007036457e-06,
+      "loss": 0.7354,
+      "step": 442
+    },
+    {
+      "epoch": 0.3669243511871894,
+      "grad_norm": 0.8779566287994385,
+      "learning_rate": 4.971598956675668e-06,
+      "loss": 0.6858,
+      "step": 443
+    },
+    {
+      "epoch": 0.3677526228602982,
+      "grad_norm": 0.9221998453140259,
+      "learning_rate": 4.971433428083522e-06,
+      "loss": 0.7116,
+      "step": 444
+    },
+    {
+      "epoch": 0.368580894533407,
+      "grad_norm": 0.9608840346336365,
+      "learning_rate": 4.971267421292045e-06,
+      "loss": 0.7129,
+      "step": 445
+    },
+    {
+      "epoch": 0.36940916620651576,
+      "grad_norm": 0.8761363625526428,
+      "learning_rate": 4.971100936333359e-06,
+      "loss": 0.718,
+      "step": 446
+    },
+    {
+      "epoch": 0.37023743787962454,
+      "grad_norm": 0.8808798789978027,
+      "learning_rate": 4.970933973239677e-06,
+      "loss": 0.6986,
+      "step": 447
+    },
+    {
+      "epoch": 0.37106570955273327,
+      "grad_norm": 0.8818857669830322,
+      "learning_rate": 4.9707665320433055e-06,
+      "loss": 0.7465,
+      "step": 448
+    },
+    {
+      "epoch": 0.37189398122584205,
+      "grad_norm": 0.9144098162651062,
+      "learning_rate": 4.970598612776641e-06,
+      "loss": 0.7187,
+      "step": 449
+    },
+    {
+      "epoch": 0.37272225289895083,
+      "grad_norm": 0.8439158797264099,
+      "learning_rate": 4.9704302154721776e-06,
+      "loss": 0.7195,
+      "step": 450
+    },
+    {
+      "epoch": 0.3735505245720596,
+      "grad_norm": 0.8661074638366699,
+      "learning_rate": 4.970261340162495e-06,
+      "loss": 0.712,
+      "step": 451
+    },
+    {
+      "epoch": 0.3743787962451684,
+      "grad_norm": 0.8818789124488831,
+      "learning_rate": 4.970091986880271e-06,
+      "loss": 0.7199,
+      "step": 452
+    },
+    {
+      "epoch": 0.3752070679182772,
+      "grad_norm": 0.9560503363609314,
+      "learning_rate": 4.969922155658273e-06,
+      "loss": 0.7145,
+      "step": 453
+    },
+    {
+      "epoch": 0.37603533959138596,
+      "grad_norm": 0.8723775744438171,
+      "learning_rate": 4.969751846529363e-06,
+      "loss": 0.7152,
+      "step": 454
+    },
+    {
+      "epoch": 0.37686361126449475,
+      "grad_norm": 0.8612253069877625,
+      "learning_rate": 4.9695810595264935e-06,
+      "loss": 0.728,
+      "step": 455
+    },
+    {
+      "epoch": 0.37769188293760353,
+      "grad_norm": 0.8995444178581238,
+      "learning_rate": 4.969409794682709e-06,
+      "loss": 0.678,
+      "step": 456
+    },
+    {
+      "epoch": 0.3785201546107123,
+      "grad_norm": 0.8770595788955688,
+      "learning_rate": 4.96923805203115e-06,
+      "loss": 0.7354,
+      "step": 457
+    },
+    {
+      "epoch": 0.3793484262838211,
+      "grad_norm": 0.893329918384552,
+      "learning_rate": 4.969065831605045e-06,
+      "loss": 0.7102,
+      "step": 458
+    },
+    {
+      "epoch": 0.3801766979569299,
+      "grad_norm": 0.954058051109314,
+      "learning_rate": 4.968893133437718e-06,
+      "loss": 0.7144,
+      "step": 459
+    },
+    {
+      "epoch": 0.38100496963003866,
+      "grad_norm": 0.9711002707481384,
+      "learning_rate": 4.968719957562585e-06,
+      "loss": 0.7452,
+      "step": 460
+    },
+    {
+      "epoch": 0.38183324130314744,
+      "grad_norm": 0.9036396741867065,
+      "learning_rate": 4.968546304013153e-06,
+      "loss": 0.7535,
+      "step": 461
+    },
+    {
+      "epoch": 0.3826615129762562,
+      "grad_norm": 0.8948055505752563,
+      "learning_rate": 4.968372172823023e-06,
+      "loss": 0.7284,
+      "step": 462
+    },
+    {
+      "epoch": 0.383489784649365,
+      "grad_norm": 0.9134189486503601,
+      "learning_rate": 4.968197564025886e-06,
+      "loss": 0.723,
+      "step": 463
+    },
+    {
+      "epoch": 0.3843180563224738,
+      "grad_norm": 0.9191608428955078,
+      "learning_rate": 4.9680224776555305e-06,
+      "loss": 0.7078,
+      "step": 464
+    },
+    {
+      "epoch": 0.38514632799558257,
+      "grad_norm": 0.879683792591095,
+      "learning_rate": 4.9678469137458315e-06,
+      "loss": 0.6805,
+      "step": 465
+    },
+    {
+      "epoch": 0.38597459966869135,
+      "grad_norm": 0.8887293338775635,
+      "learning_rate": 4.96767087233076e-06,
+      "loss": 0.7296,
+      "step": 466
+    },
+    {
+      "epoch": 0.38680287134180014,
+      "grad_norm": 0.889991044998169,
+      "learning_rate": 4.967494353444377e-06,
+      "loss": 0.702,
+      "step": 467
+    },
+    {
+      "epoch": 0.38763114301490886,
+      "grad_norm": 0.9667506814002991,
+      "learning_rate": 4.967317357120839e-06,
+      "loss": 0.7305,
+      "step": 468
+    },
+    {
+      "epoch": 0.38845941468801765,
+      "grad_norm": 0.9297274351119995,
+      "learning_rate": 4.967139883394392e-06,
+      "loss": 0.7344,
+      "step": 469
+    },
+    {
+      "epoch": 0.38928768636112643,
+      "grad_norm": 0.9167365431785583,
+      "learning_rate": 4.966961932299377e-06,
+      "loss": 0.7016,
+      "step": 470
+    },
+    {
+      "epoch": 0.3901159580342352,
+      "grad_norm": 0.877378523349762,
+      "learning_rate": 4.966783503870224e-06,
+      "loss": 0.7313,
+      "step": 471
+    },
+    {
+      "epoch": 0.390944229707344,
+      "grad_norm": 0.9625933766365051,
+      "learning_rate": 4.966604598141458e-06,
+      "loss": 0.7245,
+      "step": 472
+    },
+    {
+      "epoch": 0.3917725013804528,
+      "grad_norm": 0.8859156966209412,
+      "learning_rate": 4.9664252151476955e-06,
+      "loss": 0.7094,
+      "step": 473
+    },
+    {
+      "epoch": 0.39260077305356156,
+      "grad_norm": 0.9172984957695007,
+      "learning_rate": 4.966245354923645e-06,
+      "loss": 0.6966,
+      "step": 474
+    },
+    {
+      "epoch": 0.39342904472667034,
+      "grad_norm": 0.9040864109992981,
+      "learning_rate": 4.96606501750411e-06,
+      "loss": 0.722,
+      "step": 475
+    },
+    {
+      "epoch": 0.3942573163997791,
+      "grad_norm": 0.8900951147079468,
+      "learning_rate": 4.965884202923982e-06,
+      "loss": 0.7263,
+      "step": 476
+    },
+    {
+      "epoch": 0.3950855880728879,
+      "grad_norm": 0.9096625447273254,
+      "learning_rate": 4.965702911218247e-06,
+      "loss": 0.702,
+      "step": 477
+    },
+    {
+      "epoch": 0.3959138597459967,
+      "grad_norm": 0.8801544904708862,
+      "learning_rate": 4.965521142421984e-06,
+      "loss": 0.7113,
+      "step": 478
+    },
+    {
+      "epoch": 0.39674213141910547,
+      "grad_norm": 0.9054175615310669,
+      "learning_rate": 4.965338896570363e-06,
+      "loss": 0.7135,
+      "step": 479
+    },
+    {
+      "epoch": 0.39757040309221425,
+      "grad_norm": 0.8734445571899414,
+      "learning_rate": 4.965156173698648e-06,
+      "loss": 0.7181,
+      "step": 480
+    },
+    {
+      "epoch": 0.39839867476532304,
+      "grad_norm": 0.9113215208053589,
+      "learning_rate": 4.964972973842191e-06,
+      "loss": 0.7008,
+      "step": 481
+    },
+    {
+      "epoch": 0.3992269464384318,
+      "grad_norm": 0.8941736221313477,
+      "learning_rate": 4.964789297036444e-06,
+      "loss": 0.7201,
+      "step": 482
+    },
+    {
+      "epoch": 0.4000552181115406,
+      "grad_norm": 0.8555464148521423,
+      "learning_rate": 4.964605143316944e-06,
+      "loss": 0.7173,
+      "step": 483
+    },
+    {
+      "epoch": 0.4008834897846494,
+      "grad_norm": 0.8719691634178162,
+      "learning_rate": 4.9644205127193225e-06,
+      "loss": 0.7211,
+      "step": 484
+    },
+    {
+      "epoch": 0.40171176145775817,
+      "grad_norm": 0.901626467704773,
+      "learning_rate": 4.964235405279306e-06,
+      "loss": 0.7223,
+      "step": 485
+    },
+    {
+      "epoch": 0.40254003313086695,
+      "grad_norm": 0.856769859790802,
+      "learning_rate": 4.9640498210327095e-06,
+      "loss": 0.7103,
+      "step": 486
+    },
+    {
+      "epoch": 0.40336830480397573,
+      "grad_norm": 0.9053338766098022,
+      "learning_rate": 4.963863760015442e-06,
+      "loss": 0.7301,
+      "step": 487
+    },
+    {
+      "epoch": 0.40419657647708446,
+      "grad_norm": 0.8989829421043396,
+      "learning_rate": 4.963677222263505e-06,
+      "loss": 0.7391,
+      "step": 488
+    },
+    {
+      "epoch": 0.40502484815019324,
+      "grad_norm": 0.9187886118888855,
+      "learning_rate": 4.963490207812991e-06,
+      "loss": 0.7179,
+      "step": 489
+    },
+    {
+      "epoch": 0.405853119823302,
+      "grad_norm": 0.8712778687477112,
+      "learning_rate": 4.963302716700087e-06,
+      "loss": 0.7088,
+      "step": 490
+    },
+    {
+      "epoch": 0.4066813914964108,
+      "grad_norm": 1.0485447645187378,
+      "learning_rate": 4.9631147489610695e-06,
+      "loss": 0.7068,
+      "step": 491
+    },
+    {
+      "epoch": 0.4075096631695196,
+      "grad_norm": 0.9332602024078369,
+      "learning_rate": 4.962926304632309e-06,
+      "loss": 0.7201,
+      "step": 492
+    },
+    {
+      "epoch": 0.40833793484262837,
+      "grad_norm": 0.9264400601387024,
+      "learning_rate": 4.962737383750268e-06,
+      "loss": 0.7136,
+      "step": 493
+    },
+    {
+      "epoch": 0.40916620651573715,
+      "grad_norm": 0.8812651634216309,
+      "learning_rate": 4.962547986351501e-06,
+      "loss": 0.7315,
+      "step": 494
+    },
+    {
+      "epoch": 0.40999447818884593,
+      "grad_norm": 0.8905262351036072,
+      "learning_rate": 4.962358112472653e-06,
+      "loss": 0.7127,
+      "step": 495
+    },
+    {
+      "epoch": 0.4108227498619547,
+      "grad_norm": 0.887514591217041,
+      "learning_rate": 4.962167762150465e-06,
+      "loss": 0.7602,
+      "step": 496
+    },
+    {
+      "epoch": 0.4116510215350635,
+      "grad_norm": 0.8683431148529053,
+      "learning_rate": 4.9619769354217665e-06,
+      "loss": 0.7091,
+      "step": 497
+    },
+    {
+      "epoch": 0.4124792932081723,
+      "grad_norm": 0.944595217704773,
+      "learning_rate": 4.961785632323481e-06,
+      "loss": 0.708,
+      "step": 498
+    },
+    {
+      "epoch": 0.41330756488128106,
+      "grad_norm": 0.8834611773490906,
+      "learning_rate": 4.961593852892625e-06,
+      "loss": 0.7308,
+      "step": 499
+    },
+    {
+      "epoch": 0.41413583655438985,
+      "grad_norm": 0.8479524850845337,
+      "learning_rate": 4.961401597166304e-06,
+      "loss": 0.7061,
+      "step": 500
+    },
+    {
+      "epoch": 0.41496410822749863,
+      "grad_norm": 0.9092981219291687,
+      "learning_rate": 4.96120886518172e-06,
+      "loss": 0.7176,
+      "step": 501
+    },
+    {
+      "epoch": 0.4157923799006074,
+      "grad_norm": 0.9024932384490967,
+      "learning_rate": 4.961015656976163e-06,
+      "loss": 0.6944,
+      "step": 502
+    },
+    {
+      "epoch": 0.4166206515737162,
+      "grad_norm": 0.8945576548576355,
+      "learning_rate": 4.960821972587018e-06,
+      "loss": 0.6958,
+      "step": 503
+    },
+    {
+      "epoch": 0.417448923246825,
+      "grad_norm": 0.8668336868286133,
+      "learning_rate": 4.9606278120517606e-06,
+      "loss": 0.7219,
+      "step": 504
+    },
+    {
+      "epoch": 0.41827719491993376,
+      "grad_norm": 0.8831227421760559,
+      "learning_rate": 4.96043317540796e-06,
+      "loss": 0.7124,
+      "step": 505
+    },
+    {
+      "epoch": 0.41910546659304254,
+      "grad_norm": 0.9637073874473572,
+      "learning_rate": 4.960238062693274e-06,
+      "loss": 0.7337,
+      "step": 506
+    },
+    {
+      "epoch": 0.4199337382661513,
+      "grad_norm": 0.8895899653434753,
+      "learning_rate": 4.960042473945459e-06,
+      "loss": 0.7459,
+      "step": 507
+    },
+    {
+      "epoch": 0.42076200993926005,
+      "grad_norm": 0.8762213587760925,
+      "learning_rate": 4.959846409202356e-06,
+      "loss": 0.7319,
+      "step": 508
+    },
+    {
+      "epoch": 0.42159028161236883,
+      "grad_norm": 0.9177393317222595,
+      "learning_rate": 4.959649868501903e-06,
+      "loss": 0.7114,
+      "step": 509
+    },
+    {
+      "epoch": 0.4224185532854776,
+      "grad_norm": 0.88541179895401,
+      "learning_rate": 4.9594528518821305e-06,
+      "loss": 0.716,
+      "step": 510
+    },
+    {
+      "epoch": 0.4232468249585864,
+      "grad_norm": 0.865566611289978,
+      "learning_rate": 4.959255359381158e-06,
+      "loss": 0.7098,
+      "step": 511
+    },
+    {
+      "epoch": 0.4240750966316952,
+      "grad_norm": 0.8629707098007202,
+      "learning_rate": 4.959057391037197e-06,
+      "loss": 0.7146,
+      "step": 512
+    },
+    {
+      "epoch": 0.42490336830480396,
+      "grad_norm": 0.8642273545265198,
+      "learning_rate": 4.958858946888554e-06,
+      "loss": 0.7345,
+      "step": 513
+    },
+    {
+      "epoch": 0.42573163997791275,
+      "grad_norm": 0.9157605171203613,
+      "learning_rate": 4.958660026973626e-06,
+      "loss": 0.7177,
+      "step": 514
+    },
+    {
+      "epoch": 0.42655991165102153,
+      "grad_norm": 0.9108191728591919,
+      "learning_rate": 4.958460631330903e-06,
+      "loss": 0.7317,
+      "step": 515
+    },
+    {
+      "epoch": 0.4273881833241303,
+      "grad_norm": 0.8524913191795349,
+      "learning_rate": 4.958260759998965e-06,
+      "loss": 0.7193,
+      "step": 516
+    },
+    {
+      "epoch": 0.4282164549972391,
+      "grad_norm": 0.9197820425033569,
+      "learning_rate": 4.958060413016485e-06,
+      "loss": 0.7127,
+      "step": 517
+    },
+    {
+      "epoch": 0.4290447266703479,
+      "grad_norm": 0.8744186758995056,
+      "learning_rate": 4.9578595904222296e-06,
+      "loss": 0.6968,
+      "step": 518
+    },
+    {
+      "epoch": 0.42987299834345666,
+      "grad_norm": 0.9071037769317627,
+      "learning_rate": 4.957658292255055e-06,
+      "loss": 0.6963,
+      "step": 519
+    },
+    {
+      "epoch": 0.43070127001656544,
+      "grad_norm": 0.937796950340271,
+      "learning_rate": 4.95745651855391e-06,
+      "loss": 0.7193,
+      "step": 520
+    },
+    {
+      "epoch": 0.4315295416896742,
+      "grad_norm": 0.883471667766571,
+      "learning_rate": 4.9572542693578375e-06,
+      "loss": 0.6886,
+      "step": 521
+    },
+    {
+      "epoch": 0.432357813362783,
+      "grad_norm": 0.8738589882850647,
+      "learning_rate": 4.9570515447059705e-06,
+      "loss": 0.7072,
+      "step": 522
+    },
+    {
+      "epoch": 0.4331860850358918,
+      "grad_norm": 0.9039497971534729,
+      "learning_rate": 4.9568483446375335e-06,
+      "loss": 0.7251,
+      "step": 523
+    },
+    {
+      "epoch": 0.43401435670900057,
+      "grad_norm": 0.896789014339447,
+      "learning_rate": 4.956644669191845e-06,
+      "loss": 0.7021,
+      "step": 524
+    },
+    {
+      "epoch": 0.43484262838210935,
+      "grad_norm": 0.8883570432662964,
+      "learning_rate": 4.9564405184083126e-06,
+      "loss": 0.7153,
+      "step": 525
+    },
+    {
+      "epoch": 0.43567090005521814,
+      "grad_norm": 0.8869197368621826,
+      "learning_rate": 4.956235892326438e-06,
+      "loss": 0.7061,
+      "step": 526
+    },
+    {
+      "epoch": 0.4364991717283269,
+      "grad_norm": 0.9237920641899109,
+      "learning_rate": 4.9560307909858165e-06,
+      "loss": 0.6988,
+      "step": 527
+    },
+    {
+      "epoch": 0.43732744340143564,
+      "grad_norm": 0.9359283447265625,
+      "learning_rate": 4.955825214426131e-06,
+      "loss": 0.738,
+      "step": 528
+    },
+    {
+      "epoch": 0.4381557150745444,
+      "grad_norm": 0.8602650761604309,
+      "learning_rate": 4.95561916268716e-06,
+      "loss": 0.709,
+      "step": 529
+    },
+    {
+      "epoch": 0.4389839867476532,
+      "grad_norm": 0.9013513922691345,
+      "learning_rate": 4.955412635808771e-06,
+      "loss": 0.6836,
+      "step": 530
+    },
+    {
+      "epoch": 0.439812258420762,
+      "grad_norm": 0.9136670827865601,
+      "learning_rate": 4.955205633830927e-06,
+      "loss": 0.7263,
+      "step": 531
+    },
+    {
+      "epoch": 0.4406405300938708,
+      "grad_norm": 0.9013793468475342,
+      "learning_rate": 4.95499815679368e-06,
+      "loss": 0.6937,
+      "step": 532
+    },
+    {
+      "epoch": 0.44146880176697956,
+      "grad_norm": 0.9017112851142883,
+      "learning_rate": 4.954790204737174e-06,
+      "loss": 0.6913,
+      "step": 533
+    },
+    {
+      "epoch": 0.44229707344008834,
+      "grad_norm": 0.8901436924934387,
+      "learning_rate": 4.954581777701647e-06,
+      "loss": 0.7241,
+      "step": 534
+    },
+    {
+      "epoch": 0.4431253451131971,
+      "grad_norm": 0.8882759809494019,
+      "learning_rate": 4.954372875727427e-06,
+      "loss": 0.7115,
+      "step": 535
+    },
+    {
+      "epoch": 0.4439536167863059,
+      "grad_norm": 0.8909885883331299,
+      "learning_rate": 4.954163498854936e-06,
+      "loss": 0.7146,
+      "step": 536
+    },
+    {
+      "epoch": 0.4447818884594147,
+      "grad_norm": 0.8951786756515503,
+      "learning_rate": 4.953953647124684e-06,
+      "loss": 0.699,
+      "step": 537
+    },
+    {
+      "epoch": 0.44561016013252347,
+      "grad_norm": 0.8890933990478516,
+      "learning_rate": 4.953743320577278e-06,
+      "loss": 0.7018,
+      "step": 538
+    },
+    {
+      "epoch": 0.44643843180563225,
+      "grad_norm": 0.910393238067627,
+      "learning_rate": 4.953532519253412e-06,
+      "loss": 0.7299,
+      "step": 539
+    },
+    {
+      "epoch": 0.44726670347874103,
+      "grad_norm": 0.9152857065200806,
+      "learning_rate": 4.953321243193876e-06,
+      "loss": 0.6984,
+      "step": 540
+    },
+    {
+      "epoch": 0.4480949751518498,
+      "grad_norm": 0.9001837372779846,
+      "learning_rate": 4.9531094924395485e-06,
+      "loss": 0.727,
+      "step": 541
+    },
+    {
+      "epoch": 0.4489232468249586,
+      "grad_norm": 0.882060170173645,
+      "learning_rate": 4.952897267031401e-06,
+      "loss": 0.7016,
+      "step": 542
+    },
+    {
+      "epoch": 0.4497515184980674,
+      "grad_norm": 0.8665159344673157,
+      "learning_rate": 4.9526845670104995e-06,
+      "loss": 0.7139,
+      "step": 543
+    },
+    {
+      "epoch": 0.45057979017117616,
+      "grad_norm": 0.9200844764709473,
+      "learning_rate": 4.952471392417997e-06,
+      "loss": 0.7016,
+      "step": 544
+    },
+    {
+      "epoch": 0.45140806184428495,
+      "grad_norm": 0.8852037191390991,
+      "learning_rate": 4.9522577432951416e-06,
+      "loss": 0.7048,
+      "step": 545
+    },
+    {
+      "epoch": 0.45223633351739373,
+      "grad_norm": 0.9003865122795105,
+      "learning_rate": 4.952043619683274e-06,
+      "loss": 0.7106,
+      "step": 546
+    },
+    {
+      "epoch": 0.4530646051905025,
+      "grad_norm": 0.8888446092605591,
+      "learning_rate": 4.951829021623822e-06,
+      "loss": 0.7202,
+      "step": 547
+    },
+    {
+      "epoch": 0.45389287686361124,
+      "grad_norm": 0.8940277099609375,
+      "learning_rate": 4.9516139491583094e-06,
+      "loss": 0.7156,
+      "step": 548
+    },
+    {
+      "epoch": 0.45472114853672,
+      "grad_norm": 0.9066211581230164,
+      "learning_rate": 4.951398402328353e-06,
+      "loss": 0.6865,
+      "step": 549
+    },
+    {
+      "epoch": 0.4555494202098288,
+      "grad_norm": 0.9352865815162659,
+      "learning_rate": 4.9511823811756565e-06,
+      "loss": 0.7117,
+      "step": 550
+    },
+    {
+      "epoch": 0.4563776918829376,
+      "grad_norm": 0.8672269582748413,
+      "learning_rate": 4.95096588574202e-06,
+      "loss": 0.7166,
+      "step": 551
+    },
+    {
+      "epoch": 0.45720596355604637,
+      "grad_norm": 0.8730685114860535,
+      "learning_rate": 4.95074891606933e-06,
+      "loss": 0.6998,
+      "step": 552
+    },
+    {
+      "epoch": 0.45803423522915515,
+      "grad_norm": 0.896821916103363,
+      "learning_rate": 4.950531472199572e-06,
+      "loss": 0.6839,
+      "step": 553
+    },
+    {
+      "epoch": 0.45886250690226393,
+      "grad_norm": 0.918572187423706,
+      "learning_rate": 4.950313554174817e-06,
+      "loss": 0.7155,
+      "step": 554
+    },
+    {
+      "epoch": 0.4596907785753727,
+      "grad_norm": 0.8670752048492432,
+      "learning_rate": 4.950095162037232e-06,
+      "loss": 0.708,
+      "step": 555
+    },
+    {
+      "epoch": 0.4605190502484815,
+      "grad_norm": 0.875618040561676,
+      "learning_rate": 4.9498762958290715e-06,
+      "loss": 0.6956,
+      "step": 556
+    },
+    {
+      "epoch": 0.4613473219215903,
+      "grad_norm": 0.9120676517486572,
+      "learning_rate": 4.949656955592685e-06,
+      "loss": 0.7327,
+      "step": 557
+    },
+    {
+      "epoch": 0.46217559359469906,
+      "grad_norm": 0.9670370817184448,
+      "learning_rate": 4.949437141370514e-06,
+      "loss": 0.705,
+      "step": 558
+    },
+    {
+      "epoch": 0.46300386526780785,
+      "grad_norm": 0.9155322313308716,
+      "learning_rate": 4.949216853205089e-06,
+      "loss": 0.7126,
+      "step": 559
+    },
+    {
+      "epoch": 0.46383213694091663,
+      "grad_norm": 0.8945392370223999,
+      "learning_rate": 4.948996091139034e-06,
+      "loss": 0.7083,
+      "step": 560
+    },
+    {
+      "epoch": 0.4646604086140254,
+      "grad_norm": 0.8922162055969238,
+      "learning_rate": 4.948774855215065e-06,
+      "loss": 0.7189,
+      "step": 561
+    },
+    {
+      "epoch": 0.4654886802871342,
+      "grad_norm": 0.8940657377243042,
+      "learning_rate": 4.948553145475989e-06,
+      "loss": 0.6934,
+      "step": 562
+    },
+    {
+      "epoch": 0.466316951960243,
+      "grad_norm": 0.8946855068206787,
+      "learning_rate": 4.948330961964703e-06,
+      "loss": 0.6835,
+      "step": 563
+    },
+    {
+      "epoch": 0.46714522363335176,
+      "grad_norm": 0.9002820253372192,
+      "learning_rate": 4.948108304724201e-06,
+      "loss": 0.712,
+      "step": 564
+    },
+    {
+      "epoch": 0.46797349530646054,
+      "grad_norm": 0.8715711832046509,
+      "learning_rate": 4.947885173797561e-06,
+      "loss": 0.7197,
+      "step": 565
+    },
+    {
+      "epoch": 0.4688017669795693,
+      "grad_norm": 0.9269874095916748,
+      "learning_rate": 4.94766156922796e-06,
+      "loss": 0.7311,
+      "step": 566
+    },
+    {
+      "epoch": 0.4696300386526781,
+      "grad_norm": 0.8999423980712891,
+      "learning_rate": 4.947437491058662e-06,
+      "loss": 0.7144,
+      "step": 567
+    },
+    {
+      "epoch": 0.47045831032578683,
+      "grad_norm": 0.8704255223274231,
+      "learning_rate": 4.947212939333024e-06,
+      "loss": 0.7156,
+      "step": 568
+    },
+    {
+      "epoch": 0.4712865819988956,
+      "grad_norm": 0.8770347237586975,
+      "learning_rate": 4.946987914094495e-06,
+      "loss": 0.6969,
+      "step": 569
+    },
+    {
+      "epoch": 0.4721148536720044,
+      "grad_norm": 0.8865682482719421,
+      "learning_rate": 4.946762415386615e-06,
+      "loss": 0.7089,
+      "step": 570
+    },
+    {
+      "epoch": 0.4729431253451132,
+      "grad_norm": 0.8756279349327087,
+      "learning_rate": 4.946536443253016e-06,
+      "loss": 0.7159,
+      "step": 571
+    },
+    {
+      "epoch": 0.47377139701822196,
+      "grad_norm": 0.8844839930534363,
+      "learning_rate": 4.946309997737422e-06,
+      "loss": 0.7234,
+      "step": 572
+    },
+    {
+      "epoch": 0.47459966869133074,
+      "grad_norm": 0.8549885153770447,
+      "learning_rate": 4.946083078883647e-06,
+      "loss": 0.6927,
+      "step": 573
+    },
+    {
+      "epoch": 0.4754279403644395,
+      "grad_norm": 0.8967456817626953,
+      "learning_rate": 4.945855686735599e-06,
+      "loss": 0.7082,
+      "step": 574
+    },
+    {
+      "epoch": 0.4762562120375483,
+      "grad_norm": 0.9120200276374817,
+      "learning_rate": 4.9456278213372755e-06,
+      "loss": 0.7135,
+      "step": 575
+    },
+    {
+      "epoch": 0.4770844837106571,
+      "grad_norm": 0.8780766725540161,
+      "learning_rate": 4.945399482732765e-06,
+      "loss": 0.7139,
+      "step": 576
+    },
+    {
+      "epoch": 0.4779127553837659,
+      "grad_norm": 0.8900999426841736,
+      "learning_rate": 4.945170670966253e-06,
+      "loss": 0.7132,
+      "step": 577
+    },
+    {
+      "epoch": 0.47874102705687466,
+      "grad_norm": 0.9402151107788086,
+      "learning_rate": 4.944941386082008e-06,
+      "loss": 0.7139,
+      "step": 578
+    },
+    {
+      "epoch": 0.47956929872998344,
+      "grad_norm": 0.8769465088844299,
+      "learning_rate": 4.944711628124396e-06,
+      "loss": 0.7236,
+      "step": 579
+    },
+    {
+      "epoch": 0.4803975704030922,
+      "grad_norm": 0.9185771346092224,
+      "learning_rate": 4.944481397137874e-06,
+      "loss": 0.6884,
+      "step": 580
+    },
+    {
+      "epoch": 0.481225842076201,
+      "grad_norm": 0.8775897026062012,
+      "learning_rate": 4.944250693166988e-06,
+      "loss": 0.731,
+      "step": 581
+    },
+    {
+      "epoch": 0.4820541137493098,
+      "grad_norm": 0.8751668930053711,
+      "learning_rate": 4.94401951625638e-06,
+      "loss": 0.7007,
+      "step": 582
+    },
+    {
+      "epoch": 0.48288238542241857,
+      "grad_norm": 0.9042515754699707,
+      "learning_rate": 4.943787866450776e-06,
+      "loss": 0.7064,
+      "step": 583
+    },
+    {
+      "epoch": 0.48371065709552735,
+      "grad_norm": 0.8845478892326355,
+      "learning_rate": 4.943555743795002e-06,
+      "loss": 0.7069,
+      "step": 584
+    },
+    {
+      "epoch": 0.48453892876863613,
+      "grad_norm": 0.8959096670150757,
+      "learning_rate": 4.943323148333969e-06,
+      "loss": 0.6871,
+      "step": 585
+    },
+    {
+      "epoch": 0.4853672004417449,
+      "grad_norm": 0.9021669626235962,
+      "learning_rate": 4.943090080112684e-06,
+      "loss": 0.7156,
+      "step": 586
+    },
+    {
+      "epoch": 0.4861954721148537,
+      "grad_norm": 0.8768396973609924,
+      "learning_rate": 4.942856539176243e-06,
+      "loss": 0.681,
+      "step": 587
+    },
+    {
+      "epoch": 0.4870237437879624,
+      "grad_norm": 0.8734706044197083,
+      "learning_rate": 4.942622525569832e-06,
+      "loss": 0.7371,
+      "step": 588
+    },
+    {
+      "epoch": 0.4878520154610712,
+      "grad_norm": 0.9155406355857849,
+      "learning_rate": 4.942388039338733e-06,
+      "loss": 0.7012,
+      "step": 589
+    },
+    {
+      "epoch": 0.48868028713418,
+      "grad_norm": 0.8778123259544373,
+      "learning_rate": 4.942153080528317e-06,
+      "loss": 0.7201,
+      "step": 590
+    },
+    {
+      "epoch": 0.4895085588072888,
+      "grad_norm": 0.9102092981338501,
+      "learning_rate": 4.941917649184045e-06,
+      "loss": 0.6951,
+      "step": 591
+    },
+    {
+      "epoch": 0.49033683048039756,
+      "grad_norm": 0.8855768442153931,
+      "learning_rate": 4.941681745351471e-06,
+      "loss": 0.6997,
+      "step": 592
+    },
+    {
+      "epoch": 0.49116510215350634,
+      "grad_norm": 0.8770707845687866,
+      "learning_rate": 4.94144536907624e-06,
+      "loss": 0.7092,
+      "step": 593
+    },
+    {
+      "epoch": 0.4919933738266151,
+      "grad_norm": 0.900458574295044,
+      "learning_rate": 4.9412085204040896e-06,
+      "loss": 0.6909,
+      "step": 594
+    },
+    {
+      "epoch": 0.4928216454997239,
+      "grad_norm": 0.903489887714386,
+      "learning_rate": 4.940971199380846e-06,
+      "loss": 0.7066,
+      "step": 595
+    },
+    {
+      "epoch": 0.4936499171728327,
+      "grad_norm": 0.9331871271133423,
+      "learning_rate": 4.940733406052431e-06,
+      "loss": 0.6924,
+      "step": 596
+    },
+    {
+      "epoch": 0.49447818884594147,
+      "grad_norm": 0.8855593800544739,
+      "learning_rate": 4.9404951404648535e-06,
+      "loss": 0.6991,
+      "step": 597
+    },
+    {
+      "epoch": 0.49530646051905025,
+      "grad_norm": 0.9075731039047241,
+      "learning_rate": 4.940256402664216e-06,
+      "loss": 0.6973,
+      "step": 598
+    },
+    {
+      "epoch": 0.49613473219215903,
+      "grad_norm": 0.8970743417739868,
+      "learning_rate": 4.940017192696713e-06,
+      "loss": 0.6767,
+      "step": 599
+    },
+    {
+      "epoch": 0.4969630038652678,
+      "grad_norm": 0.8644358515739441,
+      "learning_rate": 4.939777510608628e-06,
+      "loss": 0.7284,
+      "step": 600
+    },
+    {
+      "epoch": 0.4977912755383766,
+      "grad_norm": 0.8807653784751892,
+      "learning_rate": 4.939537356446339e-06,
+      "loss": 0.7143,
+      "step": 601
+    },
+    {
+      "epoch": 0.4986195472114854,
+      "grad_norm": 0.892946183681488,
+      "learning_rate": 4.939296730256312e-06,
+      "loss": 0.7021,
+      "step": 602
+    },
+    {
+      "epoch": 0.49944781888459416,
+      "grad_norm": 0.9661110639572144,
+      "learning_rate": 4.939055632085107e-06,
+      "loss": 0.7198,
+      "step": 603
+    },
+    {
+      "epoch": 0.500276090557703,
+      "grad_norm": 0.8638496398925781,
+      "learning_rate": 4.9388140619793735e-06,
+      "loss": 0.6943,
+      "step": 604
+    },
+    {
+      "epoch": 0.5011043622308117,
+      "grad_norm": 0.9025457501411438,
+      "learning_rate": 4.938572019985853e-06,
+      "loss": 0.7114,
+      "step": 605
+    },
+    {
+      "epoch": 0.5019326339039205,
+      "grad_norm": 0.9622734785079956,
+      "learning_rate": 4.93832950615138e-06,
+      "loss": 0.7086,
+      "step": 606
+    },
+    {
+      "epoch": 0.5027609055770292,
+      "grad_norm": 0.9565468430519104,
+      "learning_rate": 4.938086520522876e-06,
+      "loss": 0.7125,
+      "step": 607
+    },
+    {
+      "epoch": 0.5035891772501381,
+      "grad_norm": 0.8970094323158264,
+      "learning_rate": 4.937843063147359e-06,
+      "loss": 0.7023,
+      "step": 608
+    },
+    {
+      "epoch": 0.5044174489232468,
+      "grad_norm": 0.8920996785163879,
+      "learning_rate": 4.937599134071933e-06,
+      "loss": 0.7041,
+      "step": 609
+    },
+    {
+      "epoch": 0.5052457205963556,
+      "grad_norm": 0.9304755926132202,
+      "learning_rate": 4.937354733343799e-06,
+      "loss": 0.7165,
+      "step": 610
+    },
+    {
+      "epoch": 0.5060739922694644,
+      "grad_norm": 0.8995933532714844,
+      "learning_rate": 4.937109861010244e-06,
+      "loss": 0.6907,
+      "step": 611
+    },
+    {
+      "epoch": 0.5069022639425732,
+      "grad_norm": 0.8648850321769714,
+      "learning_rate": 4.936864517118649e-06,
+      "loss": 0.6994,
+      "step": 612
+    },
+    {
+      "epoch": 0.5077305356156819,
+      "grad_norm": 0.8414506316184998,
+      "learning_rate": 4.936618701716487e-06,
+      "loss": 0.7081,
+      "step": 613
+    },
+    {
+      "epoch": 0.5085588072887908,
+      "grad_norm": 0.9282650947570801,
+      "learning_rate": 4.936372414851321e-06,
+      "loss": 0.7303,
+      "step": 614
+    },
+    {
+      "epoch": 0.5093870789618995,
+      "grad_norm": 0.877757728099823,
+      "learning_rate": 4.936125656570802e-06,
+      "loss": 0.7132,
+      "step": 615
+    },
+    {
+      "epoch": 0.5102153506350083,
+      "grad_norm": 0.8645135164260864,
+      "learning_rate": 4.935878426922679e-06,
+      "loss": 0.723,
+      "step": 616
+    },
+    {
+      "epoch": 0.5110436223081171,
+      "grad_norm": 0.9014747738838196,
+      "learning_rate": 4.935630725954787e-06,
+      "loss": 0.7167,
+      "step": 617
+    },
+    {
+      "epoch": 0.5118718939812258,
+      "grad_norm": 0.9295386672019958,
+      "learning_rate": 4.935382553715055e-06,
+      "loss": 0.705,
+      "step": 618
+    },
+    {
+      "epoch": 0.5127001656543346,
+      "grad_norm": 0.8910616040229797,
+      "learning_rate": 4.935133910251499e-06,
+      "loss": 0.6781,
+      "step": 619
+    },
+    {
+      "epoch": 0.5135284373274434,
+      "grad_norm": 0.8804931640625,
+      "learning_rate": 4.934884795612234e-06,
+      "loss": 0.6984,
+      "step": 620
+    },
+    {
+      "epoch": 0.5143567090005522,
+      "grad_norm": 0.8223545551300049,
+      "learning_rate": 4.934635209845458e-06,
+      "loss": 0.6892,
+      "step": 621
+    },
+    {
+      "epoch": 0.5151849806736609,
+      "grad_norm": 0.8509450554847717,
+      "learning_rate": 4.934385152999463e-06,
+      "loss": 0.7139,
+      "step": 622
+    },
+    {
+      "epoch": 0.5160132523467698,
+      "grad_norm": 0.8944748044013977,
+      "learning_rate": 4.934134625122636e-06,
+      "loss": 0.6976,
+      "step": 623
+    },
+    {
+      "epoch": 0.5168415240198785,
+      "grad_norm": 0.9021086692810059,
+      "learning_rate": 4.933883626263448e-06,
+      "loss": 0.7189,
+      "step": 624
+    },
+    {
+      "epoch": 0.5176697956929873,
+      "grad_norm": 0.8939547538757324,
+      "learning_rate": 4.933632156470467e-06,
+      "loss": 0.704,
+      "step": 625
+    },
+    {
+      "epoch": 0.518498067366096,
+      "grad_norm": 0.8793498277664185,
+      "learning_rate": 4.93338021579235e-06,
+      "loss": 0.7046,
+      "step": 626
+    },
+    {
+      "epoch": 0.5193263390392049,
+      "grad_norm": 0.8728756308555603,
+      "learning_rate": 4.933127804277845e-06,
+      "loss": 0.6826,
+      "step": 627
+    },
+    {
+      "epoch": 0.5201546107123136,
+      "grad_norm": 0.8989412188529968,
+      "learning_rate": 4.932874921975792e-06,
+      "loss": 0.6842,
+      "step": 628
+    },
+    {
+      "epoch": 0.5209828823854225,
+      "grad_norm": 0.8836191296577454,
+      "learning_rate": 4.93262156893512e-06,
+      "loss": 0.7061,
+      "step": 629
+    },
+    {
+      "epoch": 0.5218111540585312,
+      "grad_norm": 0.9429040551185608,
+      "learning_rate": 4.9323677452048505e-06,
+      "loss": 0.7053,
+      "step": 630
+    },
+    {
+      "epoch": 0.52263942573164,
+      "grad_norm": 0.879463255405426,
+      "learning_rate": 4.932113450834097e-06,
+      "loss": 0.6926,
+      "step": 631
+    },
+    {
+      "epoch": 0.5234676974047487,
+      "grad_norm": 0.8566175699234009,
+      "learning_rate": 4.931858685872062e-06,
+      "loss": 0.7126,
+      "step": 632
+    },
+    {
+      "epoch": 0.5242959690778576,
+      "grad_norm": 0.8694538474082947,
+      "learning_rate": 4.931603450368042e-06,
+      "loss": 0.7153,
+      "step": 633
+    },
+    {
+      "epoch": 0.5251242407509663,
+      "grad_norm": 0.8754484057426453,
+      "learning_rate": 4.931347744371421e-06,
+      "loss": 0.711,
+      "step": 634
+    },
+    {
+      "epoch": 0.5259525124240751,
+      "grad_norm": 0.8982204794883728,
+      "learning_rate": 4.931091567931675e-06,
+      "loss": 0.7168,
+      "step": 635
+    },
+    {
+      "epoch": 0.5267807840971839,
+      "grad_norm": 0.9262725710868835,
+      "learning_rate": 4.930834921098374e-06,
+      "loss": 0.7315,
+      "step": 636
+    },
+    {
+      "epoch": 0.5276090557702926,
+      "grad_norm": 0.9271771311759949,
+      "learning_rate": 4.930577803921176e-06,
+      "loss": 0.6968,
+      "step": 637
+    },
+    {
+      "epoch": 0.5284373274434014,
+      "grad_norm": 0.8855197429656982,
+      "learning_rate": 4.93032021644983e-06,
+      "loss": 0.6986,
+      "step": 638
+    },
+    {
+      "epoch": 0.5292655991165102,
+      "grad_norm": 0.8615745902061462,
+      "learning_rate": 4.930062158734178e-06,
+      "loss": 0.7014,
+      "step": 639
+    },
+    {
+      "epoch": 0.530093870789619,
+      "grad_norm": 0.9004185795783997,
+      "learning_rate": 4.92980363082415e-06,
+      "loss": 0.7045,
+      "step": 640
+    },
+    {
+      "epoch": 0.5309221424627277,
+      "grad_norm": 0.8533529043197632,
+      "learning_rate": 4.92954463276977e-06,
+      "loss": 0.6911,
+      "step": 641
+    },
+    {
+      "epoch": 0.5317504141358366,
+      "grad_norm": 0.881901204586029,
+      "learning_rate": 4.929285164621152e-06,
+      "loss": 0.7148,
+      "step": 642
+    },
+    {
+      "epoch": 0.5325786858089453,
+      "grad_norm": 0.8778982162475586,
+      "learning_rate": 4.9290252264285e-06,
+      "loss": 0.6879,
+      "step": 643
+    },
+    {
+      "epoch": 0.5334069574820541,
+      "grad_norm": 0.9455735087394714,
+      "learning_rate": 4.928764818242109e-06,
+      "loss": 0.7088,
+      "step": 644
+    },
+    {
+      "epoch": 0.5342352291551629,
+      "grad_norm": 0.9255897998809814,
+      "learning_rate": 4.928503940112367e-06,
+      "loss": 0.6817,
+      "step": 645
+    },
+    {
+      "epoch": 0.5350635008282717,
+      "grad_norm": 0.8854040503501892,
+      "learning_rate": 4.928242592089752e-06,
+      "loss": 0.7074,
+      "step": 646
+    },
+    {
+      "epoch": 0.5358917725013804,
+      "grad_norm": 0.8824400901794434,
+      "learning_rate": 4.927980774224829e-06,
+      "loss": 0.7092,
+      "step": 647
+    },
+    {
+      "epoch": 0.5367200441744893,
+      "grad_norm": 0.8873158693313599,
+      "learning_rate": 4.927718486568261e-06,
+      "loss": 0.7214,
+      "step": 648
+    },
+    {
+      "epoch": 0.537548315847598,
+      "grad_norm": 0.9013757109642029,
+      "learning_rate": 4.927455729170796e-06,
+      "loss": 0.6955,
+      "step": 649
+    },
+    {
+      "epoch": 0.5383765875207068,
+      "grad_norm": 0.9100024104118347,
+      "learning_rate": 4.927192502083276e-06,
+      "loss": 0.7089,
+      "step": 650
+    },
+    {
+      "epoch": 0.5392048591938156,
+      "grad_norm": 0.8935851454734802,
+      "learning_rate": 4.9269288053566335e-06,
+      "loss": 0.7018,
+      "step": 651
+    },
+    {
+      "epoch": 0.5400331308669244,
+      "grad_norm": 0.8726267218589783,
+      "learning_rate": 4.92666463904189e-06,
+      "loss": 0.6829,
+      "step": 652
+    },
+    {
+      "epoch": 0.5408614025400331,
+      "grad_norm": 0.9320974349975586,
+      "learning_rate": 4.926400003190161e-06,
+      "loss": 0.6766,
+      "step": 653
+    },
+    {
+      "epoch": 0.541689674213142,
+      "grad_norm": 0.9257067441940308,
+      "learning_rate": 4.926134897852648e-06,
+      "loss": 0.6913,
+      "step": 654
+    },
+    {
+      "epoch": 0.5425179458862507,
+      "grad_norm": 0.8642577528953552,
+      "learning_rate": 4.92586932308065e-06,
+      "loss": 0.7155,
+      "step": 655
+    },
+    {
+      "epoch": 0.5433462175593595,
+      "grad_norm": 0.8861536383628845,
+      "learning_rate": 4.9256032789255505e-06,
+      "loss": 0.6695,
+      "step": 656
+    },
+    {
+      "epoch": 0.5441744892324683,
+      "grad_norm": 0.8901777267456055,
+      "learning_rate": 4.925336765438829e-06,
+      "loss": 0.6847,
+      "step": 657
+    },
+    {
+      "epoch": 0.545002760905577,
+      "grad_norm": 0.8810168504714966,
+      "learning_rate": 4.9250697826720515e-06,
+      "loss": 0.7038,
+      "step": 658
+    },
+    {
+      "epoch": 0.5458310325786858,
+      "grad_norm": 0.8655433654785156,
+      "learning_rate": 4.924802330676877e-06,
+      "loss": 0.6908,
+      "step": 659
+    },
+    {
+      "epoch": 0.5466593042517945,
+      "grad_norm": 0.8786777853965759,
+      "learning_rate": 4.924534409505055e-06,
+      "loss": 0.663,
+      "step": 660
+    },
+    {
+      "epoch": 0.5474875759249034,
+      "grad_norm": 0.8837904334068298,
+      "learning_rate": 4.924266019208427e-06,
+      "loss": 0.6843,
+      "step": 661
+    },
+    {
+      "epoch": 0.5483158475980121,
+      "grad_norm": 0.8715303540229797,
+      "learning_rate": 4.923997159838923e-06,
+      "loss": 0.6841,
+      "step": 662
+    },
+    {
+      "epoch": 0.549144119271121,
+      "grad_norm": 0.8825753927230835,
+      "learning_rate": 4.923727831448564e-06,
+      "loss": 0.7053,
+      "step": 663
+    },
+    {
+      "epoch": 0.5499723909442297,
+      "grad_norm": 0.877668559551239,
+      "learning_rate": 4.923458034089463e-06,
+      "loss": 0.6624,
+      "step": 664
+    },
+    {
+      "epoch": 0.5508006626173385,
+      "grad_norm": 0.8721411824226379,
+      "learning_rate": 4.923187767813825e-06,
+      "loss": 0.7339,
+      "step": 665
+    },
+    {
+      "epoch": 0.5516289342904472,
+      "grad_norm": 0.8607308268547058,
+      "learning_rate": 4.922917032673941e-06,
+      "loss": 0.7184,
+      "step": 666
+    },
+    {
+      "epoch": 0.5524572059635561,
+      "grad_norm": 0.8456798791885376,
+      "learning_rate": 4.922645828722199e-06,
+      "loss": 0.7056,
+      "step": 667
+    },
+    {
+      "epoch": 0.5532854776366648,
+      "grad_norm": 0.8557775020599365,
+      "learning_rate": 4.922374156011071e-06,
+      "loss": 0.6942,
+      "step": 668
+    },
+    {
+      "epoch": 0.5541137493097736,
+      "grad_norm": 0.8877674341201782,
+      "learning_rate": 4.922102014593126e-06,
+      "loss": 0.7173,
+      "step": 669
+    },
+    {
+      "epoch": 0.5549420209828824,
+      "grad_norm": 0.8662481904029846,
+      "learning_rate": 4.921829404521019e-06,
+      "loss": 0.7113,
+      "step": 670
+    },
+    {
+      "epoch": 0.5557702926559912,
+      "grad_norm": 0.8887649774551392,
+      "learning_rate": 4.921556325847499e-06,
+      "loss": 0.7028,
+      "step": 671
+    },
+    {
+      "epoch": 0.5565985643290999,
+      "grad_norm": 0.9837973713874817,
+      "learning_rate": 4.921282778625403e-06,
+      "loss": 0.6975,
+      "step": 672
+    },
+    {
+      "epoch": 0.5574268360022088,
+      "grad_norm": 0.8754701018333435,
+      "learning_rate": 4.921008762907661e-06,
+      "loss": 0.6983,
+      "step": 673
+    },
+    {
+      "epoch": 0.5582551076753175,
+      "grad_norm": 0.9125426411628723,
+      "learning_rate": 4.920734278747291e-06,
+      "loss": 0.6982,
+      "step": 674
+    },
+    {
+      "epoch": 0.5590833793484263,
+      "grad_norm": 0.9424439072608948,
+      "learning_rate": 4.920459326197403e-06,
+      "loss": 0.7304,
+      "step": 675
+    },
+    {
+      "epoch": 0.5599116510215351,
+      "grad_norm": 0.8718393445014954,
+      "learning_rate": 4.9201839053112e-06,
+      "loss": 0.6644,
+      "step": 676
+    },
+    {
+      "epoch": 0.5607399226946438,
+      "grad_norm": 0.9002220630645752,
+      "learning_rate": 4.919908016141972e-06,
+      "loss": 0.6966,
+      "step": 677
+    },
+    {
+      "epoch": 0.5615681943677526,
+      "grad_norm": 0.8925678133964539,
+      "learning_rate": 4.9196316587431e-06,
+      "loss": 0.713,
+      "step": 678
+    },
+    {
+      "epoch": 0.5623964660408614,
+      "grad_norm": 0.8854464292526245,
+      "learning_rate": 4.919354833168059e-06,
+      "loss": 0.7075,
+      "step": 679
+    },
+    {
+      "epoch": 0.5632247377139702,
+      "grad_norm": 0.8867958188056946,
+      "learning_rate": 4.919077539470409e-06,
+      "loss": 0.7166,
+      "step": 680
+    },
+    {
+      "epoch": 0.5640530093870789,
+      "grad_norm": 0.862531840801239,
+      "learning_rate": 4.918799777703806e-06,
+      "loss": 0.7119,
+      "step": 681
+    },
+    {
+      "epoch": 0.5648812810601878,
+      "grad_norm": 0.8623220920562744,
+      "learning_rate": 4.9185215479219944e-06,
+      "loss": 0.6758,
+      "step": 682
+    },
+    {
+      "epoch": 0.5657095527332965,
+      "grad_norm": 0.9171786904335022,
+      "learning_rate": 4.918242850178808e-06,
+      "loss": 0.7049,
+      "step": 683
+    },
+    {
+      "epoch": 0.5665378244064053,
+      "grad_norm": 0.8947017192840576,
+      "learning_rate": 4.9179636845281735e-06,
+      "loss": 0.7144,
+      "step": 684
+    },
+    {
+      "epoch": 0.567366096079514,
+      "grad_norm": 0.9343388080596924,
+      "learning_rate": 4.917684051024104e-06,
+      "loss": 0.7024,
+      "step": 685
+    },
+    {
+      "epoch": 0.5681943677526229,
+      "grad_norm": 0.8929025530815125,
+      "learning_rate": 4.917403949720709e-06,
+      "loss": 0.7162,
+      "step": 686
+    },
+    {
+      "epoch": 0.5690226394257316,
+      "grad_norm": 0.9000421762466431,
+      "learning_rate": 4.9171233806721854e-06,
+      "loss": 0.7085,
+      "step": 687
+    },
+    {
+      "epoch": 0.5698509110988405,
+      "grad_norm": 0.9404364228248596,
+      "learning_rate": 4.91684234393282e-06,
+      "loss": 0.6996,
+      "step": 688
+    },
+    {
+      "epoch": 0.5706791827719492,
+      "grad_norm": 0.9566411375999451,
+      "learning_rate": 4.916560839556989e-06,
+      "loss": 0.7241,
+      "step": 689
+    },
+    {
+      "epoch": 0.571507454445058,
+      "grad_norm": 0.9066181182861328,
+      "learning_rate": 4.916278867599163e-06,
+      "loss": 0.724,
+      "step": 690
+    },
+    {
+      "epoch": 0.5723357261181667,
+      "grad_norm": 0.8914163708686829,
+      "learning_rate": 4.915996428113901e-06,
+      "loss": 0.7046,
+      "step": 691
+    },
+    {
+      "epoch": 0.5731639977912756,
+      "grad_norm": 0.8977548480033875,
+      "learning_rate": 4.91571352115585e-06,
+      "loss": 0.6823,
+      "step": 692
+    },
+    {
+      "epoch": 0.5739922694643843,
+      "grad_norm": 0.8902036547660828,
+      "learning_rate": 4.915430146779754e-06,
+      "loss": 0.702,
+      "step": 693
+    },
+    {
+      "epoch": 0.5748205411374931,
+      "grad_norm": 0.8859509229660034,
+      "learning_rate": 4.915146305040439e-06,
+      "loss": 0.6996,
+      "step": 694
+    },
+    {
+      "epoch": 0.5756488128106019,
+      "grad_norm": 0.8916864991188049,
+      "learning_rate": 4.914861995992829e-06,
+      "loss": 0.6978,
+      "step": 695
+    },
+    {
+      "epoch": 0.5764770844837107,
+      "grad_norm": 0.8961358070373535,
+      "learning_rate": 4.9145772196919336e-06,
+      "loss": 0.6803,
+      "step": 696
+    },
+    {
+      "epoch": 0.5773053561568194,
+      "grad_norm": 0.8817684650421143,
+      "learning_rate": 4.914291976192854e-06,
+      "loss": 0.6891,
+      "step": 697
+    },
+    {
+      "epoch": 0.5781336278299282,
+      "grad_norm": 0.8821113705635071,
+      "learning_rate": 4.914006265550783e-06,
+      "loss": 0.7009,
+      "step": 698
+    },
+    {
+      "epoch": 0.578961899503037,
+      "grad_norm": 0.9553062319755554,
+      "learning_rate": 4.913720087821003e-06,
+      "loss": 0.7118,
+      "step": 699
+    },
+    {
+      "epoch": 0.5797901711761457,
+      "grad_norm": 0.8997507691383362,
+      "learning_rate": 4.913433443058887e-06,
+      "loss": 0.7036,
+      "step": 700
+    },
+    {
+      "epoch": 0.5806184428492546,
+      "grad_norm": 0.8659473657608032,
+      "learning_rate": 4.9131463313198964e-06,
+      "loss": 0.6766,
+      "step": 701
+    },
+    {
+      "epoch": 0.5814467145223633,
+      "grad_norm": 0.8916030526161194,
+      "learning_rate": 4.912858752659586e-06,
+      "loss": 0.6976,
+      "step": 702
+    },
+    {
+      "epoch": 0.5822749861954721,
+      "grad_norm": 0.8726498484611511,
+      "learning_rate": 4.912570707133599e-06,
+      "loss": 0.6989,
+      "step": 703
+    },
+    {
+      "epoch": 0.5831032578685809,
+      "grad_norm": 0.8725529909133911,
+      "learning_rate": 4.91228219479767e-06,
+      "loss": 0.6862,
+      "step": 704
+    },
+    {
+      "epoch": 0.5839315295416897,
+      "grad_norm": 0.8807168006896973,
+      "learning_rate": 4.911993215707624e-06,
+      "loss": 0.6964,
+      "step": 705
+    },
+    {
+      "epoch": 0.5847598012147984,
+      "grad_norm": 0.9244167804718018,
+      "learning_rate": 4.911703769919375e-06,
+      "loss": 0.6742,
+      "step": 706
+    },
+    {
+      "epoch": 0.5855880728879073,
+      "grad_norm": 0.9174355268478394,
+      "learning_rate": 4.9114138574889285e-06,
+      "loss": 0.6852,
+      "step": 707
+    },
+    {
+      "epoch": 0.586416344561016,
+      "grad_norm": 0.8894474506378174,
+      "learning_rate": 4.9111234784723785e-06,
+      "loss": 0.689,
+      "step": 708
+    },
+    {
+      "epoch": 0.5872446162341248,
+      "grad_norm": 0.8856050968170166,
+      "learning_rate": 4.910832632925912e-06,
+      "loss": 0.6742,
+      "step": 709
+    },
+    {
+      "epoch": 0.5880728879072336,
+      "grad_norm": 0.9120550751686096,
+      "learning_rate": 4.910541320905804e-06,
+      "loss": 0.7016,
+      "step": 710
+    },
+    {
+      "epoch": 0.5889011595803424,
+      "grad_norm": 0.8925642967224121,
+      "learning_rate": 4.910249542468422e-06,
+      "loss": 0.7165,
+      "step": 711
+    },
+    {
+      "epoch": 0.5897294312534511,
+      "grad_norm": 0.8786419630050659,
+      "learning_rate": 4.909957297670222e-06,
+      "loss": 0.7207,
+      "step": 712
+    },
+    {
+      "epoch": 0.59055770292656,
+      "grad_norm": 0.89600670337677,
+      "learning_rate": 4.909664586567749e-06,
+      "loss": 0.672,
+      "step": 713
+    },
+    {
+      "epoch": 0.5913859745996687,
+      "grad_norm": 0.889835000038147,
+      "learning_rate": 4.909371409217642e-06,
+      "loss": 0.7089,
+      "step": 714
+    },
+    {
+      "epoch": 0.5922142462727775,
+      "grad_norm": 0.9029288291931152,
+      "learning_rate": 4.909077765676627e-06,
+      "loss": 0.7065,
+      "step": 715
+    },
+    {
+      "epoch": 0.5930425179458862,
+      "grad_norm": 0.9212722778320312,
+      "learning_rate": 4.908783656001521e-06,
+      "loss": 0.6787,
+      "step": 716
+    },
+    {
+      "epoch": 0.593870789618995,
+      "grad_norm": 0.9079681038856506,
+      "learning_rate": 4.9084890802492325e-06,
+      "loss": 0.6996,
+      "step": 717
+    },
+    {
+      "epoch": 0.5946990612921038,
+      "grad_norm": 0.9110735654830933,
+      "learning_rate": 4.908194038476758e-06,
+      "loss": 0.6962,
+      "step": 718
+    },
+    {
+      "epoch": 0.5955273329652125,
+      "grad_norm": 0.8945547938346863,
+      "learning_rate": 4.9078985307411865e-06,
+      "loss": 0.6648,
+      "step": 719
+    },
+    {
+      "epoch": 0.5963556046383214,
+      "grad_norm": 0.8831511735916138,
+      "learning_rate": 4.907602557099696e-06,
+      "loss": 0.7077,
+      "step": 720
+    },
+    {
+      "epoch": 0.5971838763114301,
+      "grad_norm": 0.9393531680107117,
+      "learning_rate": 4.9073061176095525e-06,
+      "loss": 0.6722,
+      "step": 721
+    },
+    {
+      "epoch": 0.5980121479845389,
+      "grad_norm": 0.8807207942008972,
+      "learning_rate": 4.907009212328117e-06,
+      "loss": 0.69,
+      "step": 722
+    },
+    {
+      "epoch": 0.5988404196576477,
+      "grad_norm": 0.8811002969741821,
+      "learning_rate": 4.906711841312836e-06,
+      "loss": 0.6902,
+      "step": 723
+    },
+    {
+      "epoch": 0.5996686913307565,
+      "grad_norm": 0.8953952789306641,
+      "learning_rate": 4.906414004621249e-06,
+      "loss": 0.6747,
+      "step": 724
+    },
+    {
+      "epoch": 0.6004969630038652,
+      "grad_norm": 0.8661350607872009,
+      "learning_rate": 4.9061157023109855e-06,
+      "loss": 0.7047,
+      "step": 725
+    },
+    {
+      "epoch": 0.6013252346769741,
+      "grad_norm": 0.8722146153450012,
+      "learning_rate": 4.905816934439762e-06,
+      "loss": 0.7231,
+      "step": 726
+    },
+    {
+      "epoch": 0.6021535063500828,
+      "grad_norm": 0.8933189511299133,
+      "learning_rate": 4.905517701065389e-06,
+      "loss": 0.7029,
+      "step": 727
+    },
+    {
+      "epoch": 0.6029817780231916,
+      "grad_norm": 0.8863442540168762,
+      "learning_rate": 4.905218002245764e-06,
+      "loss": 0.6886,
+      "step": 728
+    },
+    {
+      "epoch": 0.6038100496963004,
+      "grad_norm": 0.8672969341278076,
+      "learning_rate": 4.904917838038877e-06,
+      "loss": 0.7048,
+      "step": 729
+    },
+    {
+      "epoch": 0.6046383213694092,
+      "grad_norm": 0.8892956972122192,
+      "learning_rate": 4.904617208502807e-06,
+      "loss": 0.7009,
+      "step": 730
+    },
+    {
+      "epoch": 0.6054665930425179,
+      "grad_norm": 0.9083442687988281,
+      "learning_rate": 4.904316113695723e-06,
+      "loss": 0.7169,
+      "step": 731
+    },
+    {
+      "epoch": 0.6062948647156268,
+      "grad_norm": 0.8885907530784607,
+      "learning_rate": 4.904014553675883e-06,
+      "loss": 0.6939,
+      "step": 732
+    },
+    {
+      "epoch": 0.6071231363887355,
+      "grad_norm": 0.8859131336212158,
+      "learning_rate": 4.903712528501637e-06,
+      "loss": 0.6881,
+      "step": 733
+    },
+    {
+      "epoch": 0.6079514080618443,
+      "grad_norm": 0.874146580696106,
+      "learning_rate": 4.903410038231425e-06,
+      "loss": 0.6857,
+      "step": 734
+    },
+    {
+      "epoch": 0.6087796797349531,
+      "grad_norm": 0.8719701170921326,
+      "learning_rate": 4.903107082923774e-06,
+      "loss": 0.6834,
+      "step": 735
+    },
+    {
+      "epoch": 0.6096079514080619,
+      "grad_norm": 0.9656700491905212,
+      "learning_rate": 4.902803662637304e-06,
+      "loss": 0.6933,
+      "step": 736
+    },
+    {
+      "epoch": 0.6104362230811706,
+      "grad_norm": 0.9022018313407898,
+      "learning_rate": 4.9024997774307235e-06,
+      "loss": 0.683,
+      "step": 737
+    },
+    {
+      "epoch": 0.6112644947542794,
+      "grad_norm": 0.8694130182266235,
+      "learning_rate": 4.902195427362832e-06,
+      "loss": 0.642,
+      "step": 738
+    },
+    {
+      "epoch": 0.6120927664273882,
+      "grad_norm": 0.9153549075126648,
+      "learning_rate": 4.901890612492519e-06,
+      "loss": 0.6674,
+      "step": 739
+    },
+    {
+      "epoch": 0.6129210381004969,
+      "grad_norm": 0.9431132674217224,
+      "learning_rate": 4.901585332878762e-06,
+      "loss": 0.6918,
+      "step": 740
+    },
+    {
+      "epoch": 0.6137493097736058,
+      "grad_norm": 0.881587028503418,
+      "learning_rate": 4.901279588580631e-06,
+      "loss": 0.687,
+      "step": 741
+    },
+    {
+      "epoch": 0.6145775814467145,
+      "grad_norm": 0.9289383888244629,
+      "learning_rate": 4.9009733796572834e-06,
+      "loss": 0.6789,
+      "step": 742
+    },
+    {
+      "epoch": 0.6154058531198233,
+      "grad_norm": 0.9058337211608887,
+      "learning_rate": 4.900666706167968e-06,
+      "loss": 0.6856,
+      "step": 743
+    },
+    {
+      "epoch": 0.616234124792932,
+      "grad_norm": 0.8999397158622742,
+      "learning_rate": 4.900359568172024e-06,
+      "loss": 0.7159,
+      "step": 744
+    },
+    {
+      "epoch": 0.6170623964660409,
+      "grad_norm": 0.895321249961853,
+      "learning_rate": 4.90005196572888e-06,
+      "loss": 0.7071,
+      "step": 745
+    },
+    {
+      "epoch": 0.6178906681391496,
+      "grad_norm": 0.9142102599143982,
+      "learning_rate": 4.899743898898054e-06,
+      "loss": 0.6918,
+      "step": 746
+    },
+    {
+      "epoch": 0.6187189398122585,
+      "grad_norm": 0.8867097496986389,
+      "learning_rate": 4.899435367739152e-06,
+      "loss": 0.6991,
+      "step": 747
+    },
+    {
+      "epoch": 0.6195472114853672,
+      "grad_norm": 0.8594791293144226,
+      "learning_rate": 4.899126372311876e-06,
+      "loss": 0.7153,
+      "step": 748
+    },
+    {
+      "epoch": 0.620375483158476,
+      "grad_norm": 0.8740563988685608,
+      "learning_rate": 4.8988169126760085e-06,
+      "loss": 0.6786,
+      "step": 749
+    },
+    {
+      "epoch": 0.6212037548315847,
+      "grad_norm": 0.8818353414535522,
+      "learning_rate": 4.898506988891432e-06,
+      "loss": 0.6965,
+      "step": 750
+    },
+    {
+      "epoch": 0.6220320265046936,
+      "grad_norm": 0.9459843635559082,
+      "learning_rate": 4.898196601018111e-06,
+      "loss": 0.6821,
+      "step": 751
+    },
+    {
+      "epoch": 0.6228602981778023,
+      "grad_norm": 0.923457682132721,
+      "learning_rate": 4.897885749116104e-06,
+      "loss": 0.6822,
+      "step": 752
+    },
+    {
+      "epoch": 0.6236885698509111,
+      "grad_norm": 0.9342297315597534,
+      "learning_rate": 4.897574433245557e-06,
+      "loss": 0.7001,
+      "step": 753
+    },
+    {
+      "epoch": 0.6245168415240199,
+      "grad_norm": 0.9327735900878906,
+      "learning_rate": 4.897262653466706e-06,
+      "loss": 0.6921,
+      "step": 754
+    },
+    {
+      "epoch": 0.6253451131971287,
+      "grad_norm": 0.8876991868019104,
+      "learning_rate": 4.8969504098398805e-06,
+      "loss": 0.7087,
+      "step": 755
+    },
+    {
+      "epoch": 0.6261733848702374,
+      "grad_norm": 0.8983528017997742,
+      "learning_rate": 4.896637702425493e-06,
+      "loss": 0.6824,
+      "step": 756
+    },
+    {
+      "epoch": 0.6270016565433462,
+      "grad_norm": 0.9074965715408325,
+      "learning_rate": 4.896324531284053e-06,
+      "loss": 0.6533,
+      "step": 757
+    },
+    {
+      "epoch": 0.627829928216455,
+      "grad_norm": 0.8938930630683899,
+      "learning_rate": 4.896010896476153e-06,
+      "loss": 0.6859,
+      "step": 758
+    },
+    {
+      "epoch": 0.6286581998895637,
+      "grad_norm": 0.9036867022514343,
+      "learning_rate": 4.8956967980624806e-06,
+      "loss": 0.6904,
+      "step": 759
+    },
+    {
+      "epoch": 0.6294864715626726,
+      "grad_norm": 0.9162903428077698,
+      "learning_rate": 4.895382236103809e-06,
+      "loss": 0.6995,
+      "step": 760
+    },
+    {
+      "epoch": 0.6303147432357813,
+      "grad_norm": 1.0000487565994263,
+      "learning_rate": 4.895067210661005e-06,
+      "loss": 0.671,
+      "step": 761
+    },
+    {
+      "epoch": 0.6311430149088901,
+      "grad_norm": 0.895163893699646,
+      "learning_rate": 4.894751721795022e-06,
+      "loss": 0.6871,
+      "step": 762
+    },
+    {
+      "epoch": 0.6319712865819989,
+      "grad_norm": 0.9401512742042542,
+      "learning_rate": 4.894435769566905e-06,
+      "loss": 0.6811,
+      "step": 763
+    },
+    {
+      "epoch": 0.6327995582551077,
+      "grad_norm": 0.924578845500946,
+      "learning_rate": 4.894119354037787e-06,
+      "loss": 0.692,
+      "step": 764
+    },
+    {
+      "epoch": 0.6336278299282164,
+      "grad_norm": 0.9114750623703003,
+      "learning_rate": 4.893802475268892e-06,
+      "loss": 0.6977,
+      "step": 765
+    },
+    {
+      "epoch": 0.6344561016013253,
+      "grad_norm": 1.1472550630569458,
+      "learning_rate": 4.893485133321533e-06,
+      "loss": 0.679,
+      "step": 766
+    },
+    {
+      "epoch": 0.635284373274434,
+      "grad_norm": 0.9460884928703308,
+      "learning_rate": 4.893167328257111e-06,
+      "loss": 0.6988,
+      "step": 767
+    },
+    {
+      "epoch": 0.6361126449475428,
+      "grad_norm": 0.9040783643722534,
+      "learning_rate": 4.8928490601371216e-06,
+      "loss": 0.6706,
+      "step": 768
+    },
+    {
+      "epoch": 0.6369409166206516,
+      "grad_norm": 0.9785428047180176,
+      "learning_rate": 4.892530329023144e-06,
+      "loss": 0.7185,
+      "step": 769
+    },
+    {
+      "epoch": 0.6377691882937604,
+      "grad_norm": 0.9795847535133362,
+      "learning_rate": 4.892211134976851e-06,
+      "loss": 0.703,
+      "step": 770
+    },
+    {
+      "epoch": 0.6385974599668691,
+      "grad_norm": 0.9539455771446228,
+      "learning_rate": 4.8918914780600025e-06,
+      "loss": 0.6985,
+      "step": 771
+    },
+    {
+      "epoch": 0.639425731639978,
+      "grad_norm": 0.9304025173187256,
+      "learning_rate": 4.8915713583344505e-06,
+      "loss": 0.6731,
+      "step": 772
+    },
+    {
+      "epoch": 0.6402540033130867,
+      "grad_norm": 0.9466667771339417,
+      "learning_rate": 4.891250775862134e-06,
+      "loss": 0.6972,
+      "step": 773
+    },
+    {
+      "epoch": 0.6410822749861955,
+      "grad_norm": 0.9522298574447632,
+      "learning_rate": 4.890929730705084e-06,
+      "loss": 0.6793,
+      "step": 774
+    },
+    {
+      "epoch": 0.6419105466593042,
+      "grad_norm": 0.9077293276786804,
+      "learning_rate": 4.890608222925419e-06,
+      "loss": 0.6892,
+      "step": 775
+    },
+    {
+      "epoch": 0.6427388183324131,
+      "grad_norm": 0.8763419389724731,
+      "learning_rate": 4.890286252585348e-06,
+      "loss": 0.6985,
+      "step": 776
+    },
+    {
+      "epoch": 0.6435670900055218,
+      "grad_norm": 0.9089516401290894,
+      "learning_rate": 4.8899638197471685e-06,
+      "loss": 0.6778,
+      "step": 777
+    },
+    {
+      "epoch": 0.6443953616786305,
+      "grad_norm": 0.8733325600624084,
+      "learning_rate": 4.88964092447327e-06,
+      "loss": 0.6915,
+      "step": 778
+    },
+    {
+      "epoch": 0.6452236333517394,
+      "grad_norm": 0.9472198486328125,
+      "learning_rate": 4.889317566826128e-06,
+      "loss": 0.6984,
+      "step": 779
+    },
+    {
+      "epoch": 0.6460519050248481,
+      "grad_norm": 0.8729666471481323,
+      "learning_rate": 4.888993746868309e-06,
+      "loss": 0.6724,
+      "step": 780
+    },
+    {
+      "epoch": 0.6468801766979569,
+      "grad_norm": 0.9395233988761902,
+      "learning_rate": 4.88866946466247e-06,
+      "loss": 0.6697,
+      "step": 781
+    },
+    {
+      "epoch": 0.6477084483710657,
+      "grad_norm": 0.9158744812011719,
+      "learning_rate": 4.888344720271357e-06,
+      "loss": 0.6838,
+      "step": 782
+    },
+    {
+      "epoch": 0.6485367200441745,
+      "grad_norm": 0.8961113095283508,
+      "learning_rate": 4.888019513757804e-06,
+      "loss": 0.6488,
+      "step": 783
+    },
+    {
+      "epoch": 0.6493649917172832,
+      "grad_norm": 0.8911786079406738,
+      "learning_rate": 4.8876938451847364e-06,
+      "loss": 0.7089,
+      "step": 784
+    },
+    {
+      "epoch": 0.6501932633903921,
+      "grad_norm": 0.9214697480201721,
+      "learning_rate": 4.887367714615167e-06,
+      "loss": 0.6929,
+      "step": 785
+    },
+    {
+      "epoch": 0.6510215350635008,
+      "grad_norm": 0.9161441326141357,
+      "learning_rate": 4.8870411221122e-06,
+      "loss": 0.6761,
+      "step": 786
+    },
+    {
+      "epoch": 0.6518498067366096,
+      "grad_norm": 0.9161947965621948,
+      "learning_rate": 4.886714067739028e-06,
+      "loss": 0.6804,
+      "step": 787
+    },
+    {
+      "epoch": 0.6526780784097184,
+      "grad_norm": 0.8706585168838501,
+      "learning_rate": 4.886386551558933e-06,
+      "loss": 0.6836,
+      "step": 788
+    },
+    {
+      "epoch": 0.6535063500828272,
+      "grad_norm": 0.9037661552429199,
+      "learning_rate": 4.886058573635284e-06,
+      "loss": 0.6786,
+      "step": 789
+    },
+    {
+      "epoch": 0.6543346217559359,
+      "grad_norm": 0.8977832794189453,
+      "learning_rate": 4.885730134031545e-06,
+      "loss": 0.6938,
+      "step": 790
+    },
+    {
+      "epoch": 0.6551628934290448,
+      "grad_norm": 0.9106955528259277,
+      "learning_rate": 4.8854012328112645e-06,
+      "loss": 0.678,
+      "step": 791
+    },
+    {
+      "epoch": 0.6559911651021535,
+      "grad_norm": 0.8769396543502808,
+      "learning_rate": 4.885071870038082e-06,
+      "loss": 0.6969,
+      "step": 792
+    },
+    {
+      "epoch": 0.6568194367752623,
+      "grad_norm": 0.9946926832199097,
+      "learning_rate": 4.884742045775725e-06,
+      "loss": 0.6799,
+      "step": 793
+    },
+    {
+      "epoch": 0.6576477084483711,
+      "grad_norm": 0.9399042725563049,
+      "learning_rate": 4.884411760088012e-06,
+      "loss": 0.7038,
+      "step": 794
+    },
+    {
+      "epoch": 0.6584759801214799,
+      "grad_norm": 0.8914628624916077,
+      "learning_rate": 4.8840810130388526e-06,
+      "loss": 0.6867,
+      "step": 795
+    },
+    {
+      "epoch": 0.6593042517945886,
+      "grad_norm": 0.8900558352470398,
+      "learning_rate": 4.88374980469224e-06,
+      "loss": 0.6823,
+      "step": 796
+    },
+    {
+      "epoch": 0.6601325234676974,
+      "grad_norm": 0.8608913421630859,
+      "learning_rate": 4.8834181351122615e-06,
+      "loss": 0.6705,
+      "step": 797
+    },
+    {
+      "epoch": 0.6609607951408062,
+      "grad_norm": 0.900460422039032,
+      "learning_rate": 4.883086004363092e-06,
+      "loss": 0.7029,
+      "step": 798
+    },
+    {
+      "epoch": 0.6617890668139149,
+      "grad_norm": 0.8932703733444214,
+      "learning_rate": 4.882753412508996e-06,
+      "loss": 0.6845,
+      "step": 799
+    },
+    {
+      "epoch": 0.6626173384870238,
+      "grad_norm": 0.8737874627113342,
+      "learning_rate": 4.882420359614325e-06,
+      "loss": 0.6979,
+      "step": 800
+    },
+    {
+      "epoch": 0.6634456101601325,
+      "grad_norm": 0.8805396556854248,
+      "learning_rate": 4.882086845743524e-06,
+      "loss": 0.6935,
+      "step": 801
+    },
+    {
+      "epoch": 0.6642738818332413,
+      "grad_norm": 0.8952041268348694,
+      "learning_rate": 4.881752870961123e-06,
+      "loss": 0.7137,
+      "step": 802
+    },
+    {
+      "epoch": 0.66510215350635,
+      "grad_norm": 0.889869213104248,
+      "learning_rate": 4.881418435331745e-06,
+      "loss": 0.6643,
+      "step": 803
+    },
+    {
+      "epoch": 0.6659304251794589,
+      "grad_norm": 0.8912904858589172,
+      "learning_rate": 4.881083538920098e-06,
+      "loss": 0.6712,
+      "step": 804
+    },
+    {
+      "epoch": 0.6667586968525676,
+      "grad_norm": 0.8943006992340088,
+      "learning_rate": 4.880748181790984e-06,
+      "loss": 0.7001,
+      "step": 805
+    },
+    {
+      "epoch": 0.6675869685256764,
+      "grad_norm": 0.9042482376098633,
+      "learning_rate": 4.880412364009289e-06,
+      "loss": 0.6744,
+      "step": 806
+    },
+    {
+      "epoch": 0.6684152401987852,
+      "grad_norm": 0.9052988886833191,
+      "learning_rate": 4.880076085639991e-06,
+      "loss": 0.72,
+      "step": 807
+    },
+    {
+      "epoch": 0.669243511871894,
+      "grad_norm": 0.9702905416488647,
+      "learning_rate": 4.879739346748158e-06,
+      "loss": 0.6871,
+      "step": 808
+    },
+    {
+      "epoch": 0.6700717835450027,
+      "grad_norm": 0.8741766810417175,
+      "learning_rate": 4.879402147398944e-06,
+      "loss": 0.7174,
+      "step": 809
+    },
+    {
+      "epoch": 0.6709000552181116,
+      "grad_norm": 0.9226665496826172,
+      "learning_rate": 4.879064487657596e-06,
+      "loss": 0.6991,
+      "step": 810
+    },
+    {
+      "epoch": 0.6717283268912203,
+      "grad_norm": 0.878304123878479,
+      "learning_rate": 4.878726367589446e-06,
+      "loss": 0.6925,
+      "step": 811
+    },
+    {
+      "epoch": 0.6725565985643291,
+      "grad_norm": 0.893649160861969,
+      "learning_rate": 4.878387787259919e-06,
+      "loss": 0.6689,
+      "step": 812
+    },
+    {
+      "epoch": 0.6733848702374379,
+      "grad_norm": 0.917861819267273,
+      "learning_rate": 4.878048746734526e-06,
+      "loss": 0.7171,
+      "step": 813
+    },
+    {
+      "epoch": 0.6742131419105467,
+      "grad_norm": 0.879375696182251,
+      "learning_rate": 4.877709246078869e-06,
+      "loss": 0.6662,
+      "step": 814
+    },
+    {
+      "epoch": 0.6750414135836554,
+      "grad_norm": 0.9098542928695679,
+      "learning_rate": 4.877369285358637e-06,
+      "loss": 0.6793,
+      "step": 815
+    },
+    {
+      "epoch": 0.6758696852567642,
+      "grad_norm": 0.9465262293815613,
+      "learning_rate": 4.877028864639609e-06,
+      "loss": 0.7065,
+      "step": 816
+    },
+    {
+      "epoch": 0.676697956929873,
+      "grad_norm": 0.9029957056045532,
+      "learning_rate": 4.876687983987654e-06,
+      "loss": 0.7008,
+      "step": 817
+    },
+    {
+      "epoch": 0.6775262286029817,
+      "grad_norm": 0.9073920845985413,
+      "learning_rate": 4.876346643468731e-06,
+      "loss": 0.6836,
+      "step": 818
+    },
+    {
+      "epoch": 0.6783545002760906,
+      "grad_norm": 0.9028745889663696,
+      "learning_rate": 4.876004843148884e-06,
+      "loss": 0.6965,
+      "step": 819
+    },
+    {
+      "epoch": 0.6791827719491993,
+      "grad_norm": 0.8745572566986084,
+      "learning_rate": 4.875662583094249e-06,
+      "loss": 0.6974,
+      "step": 820
+    },
+    {
+      "epoch": 0.6800110436223081,
+      "grad_norm": 0.9203384518623352,
+      "learning_rate": 4.875319863371048e-06,
+      "loss": 0.686,
+      "step": 821
+    },
+    {
+      "epoch": 0.6808393152954169,
+      "grad_norm": 0.9087005853652954,
+      "learning_rate": 4.874976684045597e-06,
+      "loss": 0.683,
+      "step": 822
+    },
+    {
+      "epoch": 0.6816675869685257,
+      "grad_norm": 0.9007741808891296,
+      "learning_rate": 4.874633045184297e-06,
+      "loss": 0.6938,
+      "step": 823
+    },
+    {
+      "epoch": 0.6824958586416344,
+      "grad_norm": 0.9854089617729187,
+      "learning_rate": 4.874288946853639e-06,
+      "loss": 0.6945,
+      "step": 824
+    },
+    {
+      "epoch": 0.6833241303147433,
+      "grad_norm": 0.9044802784919739,
+      "learning_rate": 4.8739443891202025e-06,
+      "loss": 0.714,
+      "step": 825
+    },
+    {
+      "epoch": 0.684152401987852,
+      "grad_norm": 0.9388754963874817,
+      "learning_rate": 4.873599372050656e-06,
+      "loss": 0.6838,
+      "step": 826
+    },
+    {
+      "epoch": 0.6849806736609608,
+      "grad_norm": 0.9215866327285767,
+      "learning_rate": 4.873253895711757e-06,
+      "loss": 0.6994,
+      "step": 827
+    },
+    {
+      "epoch": 0.6858089453340696,
+      "grad_norm": 0.8786854147911072,
+      "learning_rate": 4.872907960170353e-06,
+      "loss": 0.6883,
+      "step": 828
+    },
+    {
+      "epoch": 0.6866372170071784,
+      "grad_norm": 0.9032605290412903,
+      "learning_rate": 4.872561565493379e-06,
+      "loss": 0.6847,
+      "step": 829
+    },
+    {
+      "epoch": 0.6874654886802871,
+      "grad_norm": 0.8940309882164001,
+      "learning_rate": 4.872214711747858e-06,
+      "loss": 0.6877,
+      "step": 830
+    },
+    {
+      "epoch": 0.688293760353396,
+      "grad_norm": 0.8703469634056091,
+      "learning_rate": 4.8718673990009045e-06,
+      "loss": 0.6829,
+      "step": 831
+    },
+    {
+      "epoch": 0.6891220320265047,
+      "grad_norm": 0.8618217706680298,
+      "learning_rate": 4.871519627319719e-06,
+      "loss": 0.6987,
+      "step": 832
+    },
+    {
+      "epoch": 0.6899503036996135,
+      "grad_norm": 0.8871967196464539,
+      "learning_rate": 4.871171396771594e-06,
+      "loss": 0.6824,
+      "step": 833
+    },
+    {
+      "epoch": 0.6907785753727222,
+      "grad_norm": 0.9162507057189941,
+      "learning_rate": 4.8708227074239064e-06,
+      "loss": 0.7158,
+      "step": 834
+    },
+    {
+      "epoch": 0.6916068470458311,
+      "grad_norm": 0.8829584717750549,
+      "learning_rate": 4.8704735593441255e-06,
+      "loss": 0.6714,
+      "step": 835
+    },
+    {
+      "epoch": 0.6924351187189398,
+      "grad_norm": 0.9130700826644897,
+      "learning_rate": 4.8701239525998076e-06,
+      "loss": 0.6841,
+      "step": 836
+    },
+    {
+      "epoch": 0.6932633903920485,
+      "grad_norm": 0.8706396222114563,
+      "learning_rate": 4.8697738872586e-06,
+      "loss": 0.7162,
+      "step": 837
+    },
+    {
+      "epoch": 0.6940916620651574,
+      "grad_norm": 0.8906163573265076,
+      "learning_rate": 4.869423363388235e-06,
+      "loss": 0.6586,
+      "step": 838
+    },
+    {
+      "epoch": 0.6949199337382661,
+      "grad_norm": 0.9134309887886047,
+      "learning_rate": 4.869072381056537e-06,
+      "loss": 0.6842,
+      "step": 839
+    },
+    {
+      "epoch": 0.6957482054113749,
+      "grad_norm": 0.869674563407898,
+      "learning_rate": 4.8687209403314175e-06,
+      "loss": 0.7049,
+      "step": 840
+    },
+    {
+      "epoch": 0.6965764770844837,
+      "grad_norm": 0.8405192494392395,
+      "learning_rate": 4.8683690412808775e-06,
+      "loss": 0.6677,
+      "step": 841
+    },
+    {
+      "epoch": 0.6974047487575925,
+      "grad_norm": 0.9173605442047119,
+      "learning_rate": 4.868016683973005e-06,
+      "loss": 0.6784,
+      "step": 842
+    },
+    {
+      "epoch": 0.6982330204307012,
+      "grad_norm": 0.9220597743988037,
+      "learning_rate": 4.867663868475978e-06,
+      "loss": 0.6774,
+      "step": 843
+    },
+    {
+      "epoch": 0.6990612921038101,
+      "grad_norm": 0.8687174320220947,
+      "learning_rate": 4.867310594858065e-06,
+      "loss": 0.7094,
+      "step": 844
+    },
+    {
+      "epoch": 0.6998895637769188,
+      "grad_norm": 0.9161131381988525,
+      "learning_rate": 4.866956863187619e-06,
+      "loss": 0.6867,
+      "step": 845
+    },
+    {
+      "epoch": 0.7007178354500276,
+      "grad_norm": 0.9338458776473999,
+      "learning_rate": 4.866602673533084e-06,
+      "loss": 0.6783,
+      "step": 846
+    },
+    {
+      "epoch": 0.7015461071231364,
+      "grad_norm": 0.9268187284469604,
+      "learning_rate": 4.866248025962993e-06,
+      "loss": 0.6681,
+      "step": 847
+    },
+    {
+      "epoch": 0.7023743787962452,
+      "grad_norm": 0.8843579292297363,
+      "learning_rate": 4.865892920545967e-06,
+      "loss": 0.6854,
+      "step": 848
+    },
+    {
+      "epoch": 0.7032026504693539,
+      "grad_norm": 0.889299213886261,
+      "learning_rate": 4.865537357350716e-06,
+      "loss": 0.7055,
+      "step": 849
+    },
+    {
+      "epoch": 0.7040309221424628,
+      "grad_norm": 0.9062727093696594,
+      "learning_rate": 4.865181336446036e-06,
+      "loss": 0.6806,
+      "step": 850
+    },
+    {
+      "epoch": 0.7048591938155715,
+      "grad_norm": 0.9007996320724487,
+      "learning_rate": 4.864824857900817e-06,
+      "loss": 0.6937,
+      "step": 851
+    },
+    {
+      "epoch": 0.7056874654886803,
+      "grad_norm": 0.8883542418479919,
+      "learning_rate": 4.864467921784032e-06,
+      "loss": 0.6957,
+      "step": 852
+    },
+    {
+      "epoch": 0.7065157371617891,
+      "grad_norm": 0.9422078132629395,
+      "learning_rate": 4.864110528164746e-06,
+      "loss": 0.7025,
+      "step": 853
+    },
+    {
+      "epoch": 0.7073440088348979,
+      "grad_norm": 0.8961433172225952,
+      "learning_rate": 4.86375267711211e-06,
+      "loss": 0.6795,
+      "step": 854
+    },
+    {
+      "epoch": 0.7081722805080066,
+      "grad_norm": 0.925815761089325,
+      "learning_rate": 4.863394368695366e-06,
+      "loss": 0.6824,
+      "step": 855
+    },
+    {
+      "epoch": 0.7090005521811154,
+      "grad_norm": 0.9326804280281067,
+      "learning_rate": 4.863035602983843e-06,
+      "loss": 0.6829,
+      "step": 856
+    },
+    {
+      "epoch": 0.7098288238542242,
+      "grad_norm": 0.8784539699554443,
+      "learning_rate": 4.862676380046959e-06,
+      "loss": 0.6776,
+      "step": 857
+    },
+    {
+      "epoch": 0.7106570955273329,
+      "grad_norm": 0.8577290177345276,
+      "learning_rate": 4.86231669995422e-06,
+      "loss": 0.6573,
+      "step": 858
+    },
+    {
+      "epoch": 0.7114853672004418,
+      "grad_norm": 0.9261177778244019,
+      "learning_rate": 4.86195656277522e-06,
+      "loss": 0.6891,
+      "step": 859
+    },
+    {
+      "epoch": 0.7123136388735505,
+      "grad_norm": 0.9008646607398987,
+      "learning_rate": 4.861595968579643e-06,
+      "loss": 0.6714,
+      "step": 860
+    },
+    {
+      "epoch": 0.7131419105466593,
+      "grad_norm": 0.8729333877563477,
+      "learning_rate": 4.861234917437261e-06,
+      "loss": 0.6819,
+      "step": 861
+    },
+    {
+      "epoch": 0.713970182219768,
+      "grad_norm": 0.9006103873252869,
+      "learning_rate": 4.860873409417934e-06,
+      "loss": 0.6702,
+      "step": 862
+    },
+    {
+      "epoch": 0.7147984538928769,
+      "grad_norm": 0.8699921369552612,
+      "learning_rate": 4.860511444591609e-06,
+      "loss": 0.6681,
+      "step": 863
+    },
+    {
+      "epoch": 0.7156267255659856,
+      "grad_norm": 0.9359689950942993,
+      "learning_rate": 4.860149023028325e-06,
+      "loss": 0.6834,
+      "step": 864
+    },
+    {
+      "epoch": 0.7164549972390944,
+      "grad_norm": 0.874977707862854,
+      "learning_rate": 4.859786144798205e-06,
+      "loss": 0.6692,
+      "step": 865
+    },
+    {
+      "epoch": 0.7172832689122032,
+      "grad_norm": 0.8821915984153748,
+      "learning_rate": 4.859422809971464e-06,
+      "loss": 0.6865,
+      "step": 866
+    },
+    {
+      "epoch": 0.718111540585312,
+      "grad_norm": 0.905248761177063,
+      "learning_rate": 4.859059018618405e-06,
+      "loss": 0.6705,
+      "step": 867
+    },
+    {
+      "epoch": 0.7189398122584207,
+      "grad_norm": 0.9134542346000671,
+      "learning_rate": 4.858694770809415e-06,
+      "loss": 0.7039,
+      "step": 868
+    },
+    {
+      "epoch": 0.7197680839315296,
+      "grad_norm": 0.9048055410385132,
+      "learning_rate": 4.858330066614975e-06,
+      "loss": 0.6848,
+      "step": 869
+    },
+    {
+      "epoch": 0.7205963556046383,
+      "grad_norm": 0.8780005574226379,
+      "learning_rate": 4.857964906105651e-06,
+      "loss": 0.689,
+      "step": 870
+    },
+    {
+      "epoch": 0.7214246272777471,
+      "grad_norm": 0.8826956152915955,
+      "learning_rate": 4.857599289352098e-06,
+      "loss": 0.6799,
+      "step": 871
+    },
+    {
+      "epoch": 0.7222528989508559,
+      "grad_norm": 0.919905960559845,
+      "learning_rate": 4.857233216425061e-06,
+      "loss": 0.6587,
+      "step": 872
+    },
+    {
+      "epoch": 0.7230811706239647,
+      "grad_norm": 0.8711065053939819,
+      "learning_rate": 4.8568666873953696e-06,
+      "loss": 0.6754,
+      "step": 873
+    },
+    {
+      "epoch": 0.7239094422970734,
+      "grad_norm": 0.8740426898002625,
+      "learning_rate": 4.856499702333944e-06,
+      "loss": 0.6837,
+      "step": 874
+    },
+    {
+      "epoch": 0.7247377139701823,
+      "grad_norm": 0.921416699886322,
+      "learning_rate": 4.8561322613117945e-06,
+      "loss": 0.6709,
+      "step": 875
+    },
+    {
+      "epoch": 0.725565985643291,
+      "grad_norm": 0.8988221287727356,
+      "learning_rate": 4.855764364400016e-06,
+      "loss": 0.7111,
+      "step": 876
+    },
+    {
+      "epoch": 0.7263942573163997,
+      "grad_norm": 0.8838291168212891,
+      "learning_rate": 4.8553960116697915e-06,
+      "loss": 0.6821,
+      "step": 877
+    },
+    {
+      "epoch": 0.7272225289895086,
+      "grad_norm": 0.9051623940467834,
+      "learning_rate": 4.8550272031923965e-06,
+      "loss": 0.6848,
+      "step": 878
+    },
+    {
+      "epoch": 0.7280508006626173,
+      "grad_norm": 0.9308082461357117,
+      "learning_rate": 4.85465793903919e-06,
+      "loss": 0.7087,
+      "step": 879
+    },
+    {
+      "epoch": 0.7288790723357261,
+      "grad_norm": 0.873214066028595,
+      "learning_rate": 4.854288219281623e-06,
+      "loss": 0.6877,
+      "step": 880
+    },
+    {
+      "epoch": 0.7297073440088349,
+      "grad_norm": 0.8890054225921631,
+      "learning_rate": 4.853918043991232e-06,
+      "loss": 0.6719,
+      "step": 881
+    },
+    {
+      "epoch": 0.7305356156819437,
+      "grad_norm": 0.9580398201942444,
+      "learning_rate": 4.853547413239642e-06,
+      "loss": 0.6439,
+      "step": 882
+    },
+    {
+      "epoch": 0.7313638873550524,
+      "grad_norm": 0.875035285949707,
+      "learning_rate": 4.853176327098567e-06,
+      "loss": 0.6645,
+      "step": 883
+    },
+    {
+      "epoch": 0.7321921590281613,
+      "grad_norm": 0.9299217462539673,
+      "learning_rate": 4.85280478563981e-06,
+      "loss": 0.707,
+      "step": 884
+    },
+    {
+      "epoch": 0.73302043070127,
+      "grad_norm": 0.8840609192848206,
+      "learning_rate": 4.8524327889352585e-06,
+      "loss": 0.6789,
+      "step": 885
+    },
+    {
+      "epoch": 0.7338487023743788,
+      "grad_norm": 0.8983805179595947,
+      "learning_rate": 4.852060337056891e-06,
+      "loss": 0.6814,
+      "step": 886
+    },
+    {
+      "epoch": 0.7346769740474876,
+      "grad_norm": 0.8730597496032715,
+      "learning_rate": 4.851687430076775e-06,
+      "loss": 0.7006,
+      "step": 887
+    },
+    {
+      "epoch": 0.7355052457205964,
+      "grad_norm": 0.8980515003204346,
+      "learning_rate": 4.851314068067063e-06,
+      "loss": 0.6729,
+      "step": 888
+    },
+    {
+      "epoch": 0.7363335173937051,
+      "grad_norm": 0.8919468522071838,
+      "learning_rate": 4.850940251099997e-06,
+      "loss": 0.6764,
+      "step": 889
+    },
+    {
+      "epoch": 0.737161789066814,
+      "grad_norm": 0.8987070918083191,
+      "learning_rate": 4.850565979247908e-06,
+      "loss": 0.6997,
+      "step": 890
+    },
+    {
+      "epoch": 0.7379900607399227,
+      "grad_norm": 0.9056184887886047,
+      "learning_rate": 4.850191252583214e-06,
+      "loss": 0.6887,
+      "step": 891
+    },
+    {
+      "epoch": 0.7388183324130315,
+      "grad_norm": 0.8714621067047119,
+      "learning_rate": 4.84981607117842e-06,
+      "loss": 0.6946,
+      "step": 892
+    },
+    {
+      "epoch": 0.7396466040861402,
+      "grad_norm": 0.8845617175102234,
+      "learning_rate": 4.849440435106122e-06,
+      "loss": 0.6845,
+      "step": 893
+    },
+    {
+      "epoch": 0.7404748757592491,
+      "grad_norm": 0.8926048278808594,
+      "learning_rate": 4.849064344439e-06,
+      "loss": 0.6922,
+      "step": 894
+    },
+    {
+      "epoch": 0.7413031474323578,
+      "grad_norm": 0.8731569647789001,
+      "learning_rate": 4.848687799249826e-06,
+      "loss": 0.6837,
+      "step": 895
+    },
+    {
+      "epoch": 0.7421314191054665,
+      "grad_norm": 0.8801183104515076,
+      "learning_rate": 4.848310799611457e-06,
+      "loss": 0.6932,
+      "step": 896
+    },
+    {
+      "epoch": 0.7429596907785754,
+      "grad_norm": 0.8818139433860779,
+      "learning_rate": 4.847933345596838e-06,
+      "loss": 0.6758,
+      "step": 897
+    },
+    {
+      "epoch": 0.7437879624516841,
+      "grad_norm": 0.9117923974990845,
+      "learning_rate": 4.847555437279004e-06,
+      "loss": 0.6628,
+      "step": 898
+    },
+    {
+      "epoch": 0.7446162341247929,
+      "grad_norm": 0.891360342502594,
+      "learning_rate": 4.847177074731078e-06,
+      "loss": 0.6876,
+      "step": 899
+    },
+    {
+      "epoch": 0.7454445057979017,
+      "grad_norm": 0.8905263543128967,
+      "learning_rate": 4.846798258026267e-06,
+      "loss": 0.6862,
+      "step": 900
+    },
+    {
+      "epoch": 0.7462727774710105,
+      "grad_norm": 0.8853707909584045,
+      "learning_rate": 4.846418987237868e-06,
+      "loss": 0.654,
+      "step": 901
+    },
+    {
+      "epoch": 0.7471010491441192,
+      "grad_norm": 0.8782089948654175,
+      "learning_rate": 4.846039262439271e-06,
+      "loss": 0.7005,
+      "step": 902
+    },
+    {
+      "epoch": 0.7479293208172281,
+      "grad_norm": 0.8774383068084717,
+      "learning_rate": 4.8456590837039445e-06,
+      "loss": 0.7044,
+      "step": 903
+    },
+    {
+      "epoch": 0.7487575924903368,
+      "grad_norm": 0.9385423064231873,
+      "learning_rate": 4.845278451105451e-06,
+      "loss": 0.676,
+      "step": 904
+    },
+    {
+      "epoch": 0.7495858641634456,
+      "grad_norm": 0.9211432337760925,
+      "learning_rate": 4.8448973647174405e-06,
+      "loss": 0.697,
+      "step": 905
+    },
+    {
+      "epoch": 0.7504141358365544,
+      "grad_norm": 0.887610912322998,
+      "learning_rate": 4.844515824613649e-06,
+      "loss": 0.6965,
+      "step": 906
+    },
+    {
+      "epoch": 0.7512424075096632,
+      "grad_norm": 0.885761559009552,
+      "learning_rate": 4.8441338308679e-06,
+      "loss": 0.6851,
+      "step": 907
+    },
+    {
+      "epoch": 0.7520706791827719,
+      "grad_norm": 0.9209930300712585,
+      "learning_rate": 4.843751383554106e-06,
+      "loss": 0.6718,
+      "step": 908
+    },
+    {
+      "epoch": 0.7528989508558808,
+      "grad_norm": 0.9308252334594727,
+      "learning_rate": 4.843368482746269e-06,
+      "loss": 0.6802,
+      "step": 909
+    },
+    {
+      "epoch": 0.7537272225289895,
+      "grad_norm": 0.8931897282600403,
+      "learning_rate": 4.842985128518473e-06,
+      "loss": 0.7076,
+      "step": 910
+    },
+    {
+      "epoch": 0.7545554942020983,
+      "grad_norm": 0.9102733731269836,
+      "learning_rate": 4.842601320944898e-06,
+      "loss": 0.6898,
+      "step": 911
+    },
+    {
+      "epoch": 0.7553837658752071,
+      "grad_norm": 0.9336893558502197,
+      "learning_rate": 4.842217060099804e-06,
+      "loss": 0.6709,
+      "step": 912
+    },
+    {
+      "epoch": 0.7562120375483159,
+      "grad_norm": 0.9366991519927979,
+      "learning_rate": 4.8418323460575425e-06,
+      "loss": 0.6665,
+      "step": 913
+    },
+    {
+      "epoch": 0.7570403092214246,
+      "grad_norm": 0.9371415376663208,
+      "learning_rate": 4.841447178892552e-06,
+      "loss": 0.7013,
+      "step": 914
+    },
+    {
+      "epoch": 0.7578685808945335,
+      "grad_norm": 1.031123399734497,
+      "learning_rate": 4.84106155867936e-06,
+      "loss": 0.687,
+      "step": 915
+    },
+    {
+      "epoch": 0.7586968525676422,
+      "grad_norm": 0.9319218397140503,
+      "learning_rate": 4.8406754854925785e-06,
+      "loss": 0.669,
+      "step": 916
+    },
+    {
+      "epoch": 0.7595251242407509,
+      "grad_norm": 0.8563560843467712,
+      "learning_rate": 4.840288959406911e-06,
+      "loss": 0.6905,
+      "step": 917
+    },
+    {
+      "epoch": 0.7603533959138598,
+      "grad_norm": 0.9424378275871277,
+      "learning_rate": 4.839901980497145e-06,
+      "loss": 0.6685,
+      "step": 918
+    },
+    {
+      "epoch": 0.7611816675869685,
+      "grad_norm": 0.9202269315719604,
+      "learning_rate": 4.839514548838158e-06,
+      "loss": 0.6728,
+      "step": 919
+    },
+    {
+      "epoch": 0.7620099392600773,
+      "grad_norm": 0.9152582287788391,
+      "learning_rate": 4.839126664504915e-06,
+      "loss": 0.6842,
+      "step": 920
+    },
+    {
+      "epoch": 0.762838210933186,
+      "grad_norm": 0.9457129836082458,
+      "learning_rate": 4.838738327572468e-06,
+      "loss": 0.6822,
+      "step": 921
+    },
+    {
+      "epoch": 0.7636664826062949,
+      "grad_norm": 0.93124920129776,
+      "learning_rate": 4.838349538115955e-06,
+      "loss": 0.691,
+      "step": 922
+    },
+    {
+      "epoch": 0.7644947542794036,
+      "grad_norm": 0.9249301552772522,
+      "learning_rate": 4.837960296210605e-06,
+      "loss": 0.6574,
+      "step": 923
+    },
+    {
+      "epoch": 0.7653230259525124,
+      "grad_norm": 1.0551042556762695,
+      "learning_rate": 4.837570601931731e-06,
+      "loss": 0.6782,
+      "step": 924
+    },
+    {
+      "epoch": 0.7661512976256212,
+      "grad_norm": 0.9404964447021484,
+      "learning_rate": 4.8371804553547365e-06,
+      "loss": 0.6703,
+      "step": 925
+    },
+    {
+      "epoch": 0.76697956929873,
+      "grad_norm": 0.909015417098999,
+      "learning_rate": 4.836789856555111e-06,
+      "loss": 0.6812,
+      "step": 926
+    },
+    {
+      "epoch": 0.7678078409718387,
+      "grad_norm": 0.8572261929512024,
+      "learning_rate": 4.836398805608431e-06,
+      "loss": 0.6831,
+      "step": 927
+    },
+    {
+      "epoch": 0.7686361126449476,
+      "grad_norm": 0.951753556728363,
+      "learning_rate": 4.836007302590362e-06,
+      "loss": 0.6672,
+      "step": 928
+    },
+    {
+      "epoch": 0.7694643843180563,
+      "grad_norm": 0.8875932693481445,
+      "learning_rate": 4.8356153475766564e-06,
+      "loss": 0.6819,
+      "step": 929
+    },
+    {
+      "epoch": 0.7702926559911651,
+      "grad_norm": 0.9405681490898132,
+      "learning_rate": 4.835222940643153e-06,
+      "loss": 0.6917,
+      "step": 930
+    },
+    {
+      "epoch": 0.7711209276642739,
+      "grad_norm": 0.942520022392273,
+      "learning_rate": 4.834830081865779e-06,
+      "loss": 0.6657,
+      "step": 931
+    },
+    {
+      "epoch": 0.7719491993373827,
+      "grad_norm": 0.9063259363174438,
+      "learning_rate": 4.83443677132055e-06,
+      "loss": 0.6857,
+      "step": 932
+    },
+    {
+      "epoch": 0.7727774710104914,
+      "grad_norm": 0.8778199553489685,
+      "learning_rate": 4.834043009083566e-06,
+      "loss": 0.7028,
+      "step": 933
+    },
+    {
+      "epoch": 0.7736057426836003,
+      "grad_norm": 0.9215186834335327,
+      "learning_rate": 4.833648795231018e-06,
+      "loss": 0.6712,
+      "step": 934
+    },
+    {
+      "epoch": 0.774434014356709,
+      "grad_norm": 0.8992031812667847,
+      "learning_rate": 4.8332541298391825e-06,
+      "loss": 0.6827,
+      "step": 935
+    },
+    {
+      "epoch": 0.7752622860298177,
+      "grad_norm": 0.9203833341598511,
+      "learning_rate": 4.8328590129844224e-06,
+      "loss": 0.6835,
+      "step": 936
+    },
+    {
+      "epoch": 0.7760905577029266,
+      "grad_norm": 0.9109310507774353,
+      "learning_rate": 4.832463444743191e-06,
+      "loss": 0.673,
+      "step": 937
+    },
+    {
+      "epoch": 0.7769188293760353,
+      "grad_norm": 0.978349506855011,
+      "learning_rate": 4.832067425192024e-06,
+      "loss": 0.6638,
+      "step": 938
+    },
+    {
+      "epoch": 0.7777471010491441,
+      "grad_norm": 0.9552261829376221,
+      "learning_rate": 4.831670954407551e-06,
+      "loss": 0.6919,
+      "step": 939
+    },
+    {
+      "epoch": 0.7785753727222529,
+      "grad_norm": 0.9106539487838745,
+      "learning_rate": 4.831274032466484e-06,
+      "loss": 0.6964,
+      "step": 940
+    },
+    {
+      "epoch": 0.7794036443953617,
+      "grad_norm": 0.9195299744606018,
+      "learning_rate": 4.830876659445623e-06,
+      "loss": 0.6793,
+      "step": 941
+    },
+    {
+      "epoch": 0.7802319160684704,
+      "grad_norm": 0.9075512290000916,
+      "learning_rate": 4.8304788354218564e-06,
+      "loss": 0.6649,
+      "step": 942
+    },
+    {
+      "epoch": 0.7810601877415793,
+      "grad_norm": 0.9142526984214783,
+      "learning_rate": 4.83008056047216e-06,
+      "loss": 0.6811,
+      "step": 943
+    },
+    {
+      "epoch": 0.781888459414688,
+      "grad_norm": 0.8898900151252747,
+      "learning_rate": 4.829681834673597e-06,
+      "loss": 0.6906,
+      "step": 944
+    },
+    {
+      "epoch": 0.7827167310877968,
+      "grad_norm": 0.8884983062744141,
+      "learning_rate": 4.8292826581033135e-06,
+      "loss": 0.6903,
+      "step": 945
+    },
+    {
+      "epoch": 0.7835450027609056,
+      "grad_norm": 0.9075689315795898,
+      "learning_rate": 4.828883030838551e-06,
+      "loss": 0.6794,
+      "step": 946
+    },
+    {
+      "epoch": 0.7843732744340144,
+      "grad_norm": 0.8970300555229187,
+      "learning_rate": 4.828482952956631e-06,
+      "loss": 0.6715,
+      "step": 947
+    },
+    {
+      "epoch": 0.7852015461071231,
+      "grad_norm": 0.8906447887420654,
+      "learning_rate": 4.828082424534965e-06,
+      "loss": 0.6878,
+      "step": 948
+    },
+    {
+      "epoch": 0.786029817780232,
+      "grad_norm": 0.9083748459815979,
+      "learning_rate": 4.827681445651052e-06,
+      "loss": 0.6787,
+      "step": 949
+    },
+    {
+      "epoch": 0.7868580894533407,
+      "grad_norm": 0.8975276947021484,
+      "learning_rate": 4.8272800163824784e-06,
+      "loss": 0.687,
+      "step": 950
+    },
+    {
+      "epoch": 0.7876863611264495,
+      "grad_norm": 0.9573839902877808,
+      "learning_rate": 4.8268781368069156e-06,
+      "loss": 0.6996,
+      "step": 951
+    },
+    {
+      "epoch": 0.7885146327995582,
+      "grad_norm": 0.913536787033081,
+      "learning_rate": 4.826475807002125e-06,
+      "loss": 0.6844,
+      "step": 952
+    },
+    {
+      "epoch": 0.7893429044726671,
+      "grad_norm": 0.9202503561973572,
+      "learning_rate": 4.826073027045952e-06,
+      "loss": 0.6702,
+      "step": 953
+    },
+    {
+      "epoch": 0.7901711761457758,
+      "grad_norm": 0.8800237774848938,
+      "learning_rate": 4.8256697970163315e-06,
+      "loss": 0.685,
+      "step": 954
+    },
+    {
+      "epoch": 0.7909994478188846,
+      "grad_norm": 0.8923048377037048,
+      "learning_rate": 4.825266116991285e-06,
+      "loss": 0.6839,
+      "step": 955
+    },
+    {
+      "epoch": 0.7918277194919934,
+      "grad_norm": 0.879743754863739,
+      "learning_rate": 4.824861987048922e-06,
+      "loss": 0.7112,
+      "step": 956
+    },
+    {
+      "epoch": 0.7926559911651021,
+      "grad_norm": 0.8575169444084167,
+      "learning_rate": 4.824457407267436e-06,
+      "loss": 0.671,
+      "step": 957
+    },
+    {
+      "epoch": 0.7934842628382109,
+      "grad_norm": 0.8851709961891174,
+      "learning_rate": 4.824052377725109e-06,
+      "loss": 0.669,
+      "step": 958
+    },
+    {
+      "epoch": 0.7943125345113197,
+      "grad_norm": 0.9084329605102539,
+      "learning_rate": 4.823646898500311e-06,
+      "loss": 0.6731,
+      "step": 959
+    },
+    {
+      "epoch": 0.7951408061844285,
+      "grad_norm": 0.908256471157074,
+      "learning_rate": 4.823240969671501e-06,
+      "loss": 0.688,
+      "step": 960
+    },
+    {
+      "epoch": 0.7959690778575372,
+      "grad_norm": 0.8993384838104248,
+      "learning_rate": 4.822834591317219e-06,
+      "loss": 0.6736,
+      "step": 961
+    },
+    {
+      "epoch": 0.7967973495306461,
+      "grad_norm": 0.893530011177063,
+      "learning_rate": 4.822427763516098e-06,
+      "loss": 0.6787,
+      "step": 962
+    },
+    {
+      "epoch": 0.7976256212037548,
+      "grad_norm": 0.9479735493659973,
+      "learning_rate": 4.822020486346852e-06,
+      "loss": 0.6842,
+      "step": 963
+    },
+    {
+      "epoch": 0.7984538928768636,
+      "grad_norm": 0.8716689348220825,
+      "learning_rate": 4.821612759888289e-06,
+      "loss": 0.6726,
+      "step": 964
+    },
+    {
+      "epoch": 0.7992821645499724,
+      "grad_norm": 0.8902738690376282,
+      "learning_rate": 4.821204584219299e-06,
+      "loss": 0.6672,
+      "step": 965
+    },
+    {
+      "epoch": 0.8001104362230812,
+      "grad_norm": 0.8713969588279724,
+      "learning_rate": 4.820795959418859e-06,
+      "loss": 0.6783,
+      "step": 966
+    },
+    {
+      "epoch": 0.8009387078961899,
+      "grad_norm": 0.891788899898529,
+      "learning_rate": 4.820386885566036e-06,
+      "loss": 0.6691,
+      "step": 967
+    },
+    {
+      "epoch": 0.8017669795692988,
+      "grad_norm": 0.8484700918197632,
+      "learning_rate": 4.81997736273998e-06,
+      "loss": 0.6766,
+      "step": 968
+    },
+    {
+      "epoch": 0.8025952512424075,
+      "grad_norm": 0.8931223154067993,
+      "learning_rate": 4.819567391019931e-06,
+      "loss": 0.6944,
+      "step": 969
+    },
+    {
+      "epoch": 0.8034235229155163,
+      "grad_norm": 0.9707916975021362,
+      "learning_rate": 4.819156970485216e-06,
+      "loss": 0.7076,
+      "step": 970
+    },
+    {
+      "epoch": 0.8042517945886251,
+      "grad_norm": 0.9172869920730591,
+      "learning_rate": 4.818746101215245e-06,
+      "loss": 0.6864,
+      "step": 971
+    },
+    {
+      "epoch": 0.8050800662617339,
+      "grad_norm": 0.8982006311416626,
+      "learning_rate": 4.818334783289518e-06,
+      "loss": 0.6685,
+      "step": 972
+    },
+    {
+      "epoch": 0.8059083379348426,
+      "grad_norm": 0.8960995078086853,
+      "learning_rate": 4.817923016787623e-06,
+      "loss": 0.6959,
+      "step": 973
+    },
+    {
+      "epoch": 0.8067366096079515,
+      "grad_norm": 0.8806949257850647,
+      "learning_rate": 4.81751080178923e-06,
+      "loss": 0.7028,
+      "step": 974
+    },
+    {
+      "epoch": 0.8075648812810602,
+      "grad_norm": 0.8798978924751282,
+      "learning_rate": 4.817098138374102e-06,
+      "loss": 0.6718,
+      "step": 975
+    },
+    {
+      "epoch": 0.8083931529541689,
+      "grad_norm": 0.9217897653579712,
+      "learning_rate": 4.8166850266220835e-06,
+      "loss": 0.6782,
+      "step": 976
+    },
+    {
+      "epoch": 0.8092214246272778,
+      "grad_norm": 0.8915667533874512,
+      "learning_rate": 4.8162714666131074e-06,
+      "loss": 0.6917,
+      "step": 977
+    },
+    {
+      "epoch": 0.8100496963003865,
+      "grad_norm": 0.8895093202590942,
+      "learning_rate": 4.815857458427195e-06,
+      "loss": 0.698,
+      "step": 978
+    },
+    {
+      "epoch": 0.8108779679734953,
+      "grad_norm": 0.9167448282241821,
+      "learning_rate": 4.815443002144453e-06,
+      "loss": 0.6796,
+      "step": 979
+    },
+    {
+      "epoch": 0.811706239646604,
+      "grad_norm": 0.913827657699585,
+      "learning_rate": 4.815028097845074e-06,
+      "loss": 0.6731,
+      "step": 980
+    },
+    {
+      "epoch": 0.8125345113197129,
+      "grad_norm": 0.9183399081230164,
+      "learning_rate": 4.814612745609338e-06,
+      "loss": 0.7076,
+      "step": 981
+    },
+    {
+      "epoch": 0.8133627829928216,
+      "grad_norm": 0.8812996745109558,
+      "learning_rate": 4.814196945517613e-06,
+      "loss": 0.6416,
+      "step": 982
+    },
+    {
+      "epoch": 0.8141910546659304,
+      "grad_norm": 0.9032955169677734,
+      "learning_rate": 4.813780697650351e-06,
+      "loss": 0.6799,
+      "step": 983
+    },
+    {
+      "epoch": 0.8150193263390392,
+      "grad_norm": 0.9380100965499878,
+      "learning_rate": 4.813364002088093e-06,
+      "loss": 0.6771,
+      "step": 984
+    },
+    {
+      "epoch": 0.815847598012148,
+      "grad_norm": 0.9470316171646118,
+      "learning_rate": 4.8129468589114655e-06,
+      "loss": 0.6851,
+      "step": 985
+    },
+    {
+      "epoch": 0.8166758696852567,
+      "grad_norm": 0.9401133060455322,
+      "learning_rate": 4.812529268201183e-06,
+      "loss": 0.676,
+      "step": 986
+    },
+    {
+      "epoch": 0.8175041413583656,
+      "grad_norm": 0.9379092454910278,
+      "learning_rate": 4.812111230038043e-06,
+      "loss": 0.672,
+      "step": 987
+    },
+    {
+      "epoch": 0.8183324130314743,
+      "grad_norm": 0.9591637253761292,
+      "learning_rate": 4.811692744502933e-06,
+      "loss": 0.6684,
+      "step": 988
+    },
+    {
+      "epoch": 0.8191606847045831,
+      "grad_norm": 0.8911624550819397,
+      "learning_rate": 4.8112738116768276e-06,
+      "loss": 0.6448,
+      "step": 989
+    },
+    {
+      "epoch": 0.8199889563776919,
+      "grad_norm": 0.9036282896995544,
+      "learning_rate": 4.810854431640784e-06,
+      "loss": 0.703,
+      "step": 990
+    },
+    {
+      "epoch": 0.8208172280508007,
+      "grad_norm": 0.9197635650634766,
+      "learning_rate": 4.81043460447595e-06,
+      "loss": 0.6655,
+      "step": 991
+    },
+    {
+      "epoch": 0.8216454997239094,
+      "grad_norm": 0.9548643827438354,
+      "learning_rate": 4.8100143302635575e-06,
+      "loss": 0.6892,
+      "step": 992
+    },
+    {
+      "epoch": 0.8224737713970183,
+      "grad_norm": 0.9139766097068787,
+      "learning_rate": 4.809593609084925e-06,
+      "loss": 0.6723,
+      "step": 993
+    },
+    {
+      "epoch": 0.823302043070127,
+      "grad_norm": 0.9067809581756592,
+      "learning_rate": 4.8091724410214595e-06,
+      "loss": 0.6649,
+      "step": 994
+    },
+    {
+      "epoch": 0.8241303147432358,
+      "grad_norm": 0.9705583453178406,
+      "learning_rate": 4.808750826154652e-06,
+      "loss": 0.6889,
+      "step": 995
+    },
+    {
+      "epoch": 0.8249585864163446,
+      "grad_norm": 0.9664806127548218,
+      "learning_rate": 4.808328764566082e-06,
+      "loss": 0.6758,
+      "step": 996
+    },
+    {
+      "epoch": 0.8257868580894533,
+      "grad_norm": 0.8840530514717102,
+      "learning_rate": 4.807906256337414e-06,
+      "loss": 0.6719,
+      "step": 997
+    },
+    {
+      "epoch": 0.8266151297625621,
+      "grad_norm": 0.8803462386131287,
+      "learning_rate": 4.8074833015504e-06,
+      "loss": 0.6903,
+      "step": 998
+    },
+    {
+      "epoch": 0.8274434014356709,
+      "grad_norm": 0.9404571652412415,
+      "learning_rate": 4.8070599002868756e-06,
+      "loss": 0.6917,
+      "step": 999
+    },
+    {
+      "epoch": 0.8282716731087797,
+      "grad_norm": 1.0206997394561768,
+      "learning_rate": 4.806636052628768e-06,
+      "loss": 0.6854,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8290999447818884,
+      "grad_norm": 0.9294790029525757,
+      "learning_rate": 4.806211758658086e-06,
+      "loss": 0.6791,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8299282164549973,
+      "grad_norm": 0.874011754989624,
+      "learning_rate": 4.805787018456927e-06,
+      "loss": 0.6915,
+      "step": 1002
+    },
+    {
+      "epoch": 0.830756488128106,
+      "grad_norm": 0.9098061323165894,
+      "learning_rate": 4.805361832107474e-06,
+      "loss": 0.6643,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8315847598012148,
+      "grad_norm": 0.9783986210823059,
+      "learning_rate": 4.804936199691997e-06,
+      "loss": 0.6853,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8324130314743236,
+      "grad_norm": 0.9316959381103516,
+      "learning_rate": 4.804510121292852e-06,
+      "loss": 0.6696,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8332413031474324,
+      "grad_norm": 0.8930666446685791,
+      "learning_rate": 4.80408359699248e-06,
+      "loss": 0.6805,
+      "step": 1006
+    },
+    {
+      "epoch": 0.8340695748205411,
+      "grad_norm": 0.9417555332183838,
+      "learning_rate": 4.803656626873412e-06,
+      "loss": 0.66,
+      "step": 1007
+    },
+    {
+      "epoch": 0.83489784649365,
+      "grad_norm": 0.9435075521469116,
+      "learning_rate": 4.80322921101826e-06,
+      "loss": 0.6664,
+      "step": 1008
+    },
+    {
+      "epoch": 0.8357261181667587,
+      "grad_norm": 0.9090076684951782,
+      "learning_rate": 4.8028013495097255e-06,
+      "loss": 0.6595,
+      "step": 1009
+    },
+    {
+      "epoch": 0.8365543898398675,
+      "grad_norm": 0.9024788737297058,
+      "learning_rate": 4.802373042430597e-06,
+      "loss": 0.6672,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8373826615129762,
+      "grad_norm": 0.899813711643219,
+      "learning_rate": 4.801944289863747e-06,
+      "loss": 0.6743,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8382109331860851,
+      "grad_norm": 0.887106716632843,
+      "learning_rate": 4.801515091892136e-06,
+      "loss": 0.6823,
+      "step": 1012
+    },
+    {
+      "epoch": 0.8390392048591938,
+      "grad_norm": 0.8690978288650513,
+      "learning_rate": 4.801085448598809e-06,
+      "loss": 0.6642,
+      "step": 1013
+    },
+    {
+      "epoch": 0.8398674765323026,
+      "grad_norm": 0.9131479859352112,
+      "learning_rate": 4.800655360066898e-06,
+      "loss": 0.6841,
+      "step": 1014
+    },
+    {
+      "epoch": 0.8406957482054114,
+      "grad_norm": 0.9031343460083008,
+      "learning_rate": 4.800224826379622e-06,
+      "loss": 0.6879,
+      "step": 1015
+    },
+    {
+      "epoch": 0.8415240198785201,
+      "grad_norm": 0.8914244174957275,
+      "learning_rate": 4.799793847620283e-06,
+      "loss": 0.6562,
+      "step": 1016
+    },
+    {
+      "epoch": 0.8423522915516289,
+      "grad_norm": 0.8692408800125122,
+      "learning_rate": 4.799362423872274e-06,
+      "loss": 0.6759,
+      "step": 1017
+    },
+    {
+      "epoch": 0.8431805632247377,
+      "grad_norm": 0.9511806964874268,
+      "learning_rate": 4.798930555219069e-06,
+      "loss": 0.6637,
+      "step": 1018
+    },
+    {
+      "epoch": 0.8440088348978465,
+      "grad_norm": 0.9556595683097839,
+      "learning_rate": 4.7984982417442325e-06,
+      "loss": 0.6925,
+      "step": 1019
+    },
+    {
+      "epoch": 0.8448371065709552,
+      "grad_norm": 0.9107236862182617,
+      "learning_rate": 4.798065483531413e-06,
+      "loss": 0.67,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8456653782440641,
+      "grad_norm": 0.875293493270874,
+      "learning_rate": 4.797632280664345e-06,
+      "loss": 0.6755,
+      "step": 1021
+    },
+    {
+      "epoch": 0.8464936499171728,
+      "grad_norm": 0.8972271084785461,
+      "learning_rate": 4.797198633226849e-06,
+      "loss": 0.6637,
+      "step": 1022
+    },
+    {
+      "epoch": 0.8473219215902816,
+      "grad_norm": 0.8903595805168152,
+      "learning_rate": 4.796764541302831e-06,
+      "loss": 0.6997,
+      "step": 1023
+    },
+    {
+      "epoch": 0.8481501932633904,
+      "grad_norm": 0.8948636651039124,
+      "learning_rate": 4.796330004976285e-06,
+      "loss": 0.6579,
+      "step": 1024
+    },
+    {
+      "epoch": 0.8489784649364992,
+      "grad_norm": 0.8590925335884094,
+      "learning_rate": 4.795895024331289e-06,
+      "loss": 0.6651,
+      "step": 1025
+    },
+    {
+      "epoch": 0.8498067366096079,
+      "grad_norm": 0.870811402797699,
+      "learning_rate": 4.795459599452007e-06,
+      "loss": 0.6707,
+      "step": 1026
+    },
+    {
+      "epoch": 0.8506350082827168,
+      "grad_norm": 0.8807471394538879,
+      "learning_rate": 4.795023730422692e-06,
+      "loss": 0.6787,
+      "step": 1027
+    },
+    {
+      "epoch": 0.8514632799558255,
+      "grad_norm": 0.8817671537399292,
+      "learning_rate": 4.794587417327678e-06,
+      "loss": 0.66,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8522915516289343,
+      "grad_norm": 0.8924570679664612,
+      "learning_rate": 4.794150660251389e-06,
+      "loss": 0.6842,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8531198233020431,
+      "grad_norm": 0.8746559619903564,
+      "learning_rate": 4.793713459278333e-06,
+      "loss": 0.6582,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8539480949751519,
+      "grad_norm": 0.8885277509689331,
+      "learning_rate": 4.793275814493104e-06,
+      "loss": 0.6784,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8547763666482606,
+      "grad_norm": 0.8797710537910461,
+      "learning_rate": 4.792837725980383e-06,
+      "loss": 0.6638,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8556046383213695,
+      "grad_norm": 0.8932235240936279,
+      "learning_rate": 4.7923991938249355e-06,
+      "loss": 0.6918,
+      "step": 1033
+    },
+    {
+      "epoch": 0.8564329099944782,
+      "grad_norm": 0.9317487478256226,
+      "learning_rate": 4.791960218111613e-06,
+      "loss": 0.6905,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8572611816675869,
+      "grad_norm": 0.8760019540786743,
+      "learning_rate": 4.791520798925354e-06,
+      "loss": 0.673,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8580894533406958,
+      "grad_norm": 0.8644204139709473,
+      "learning_rate": 4.7910809363511825e-06,
+      "loss": 0.6755,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8589177250138045,
+      "grad_norm": 0.8761067390441895,
+      "learning_rate": 4.790640630474206e-06,
+      "loss": 0.699,
+      "step": 1037
+    },
+    {
+      "epoch": 0.8597459966869133,
+      "grad_norm": 0.9135014414787292,
+      "learning_rate": 4.790199881379621e-06,
+      "loss": 0.6659,
+      "step": 1038
+    },
+    {
+      "epoch": 0.860574268360022,
+      "grad_norm": 0.9065434336662292,
+      "learning_rate": 4.789758689152708e-06,
+      "loss": 0.6664,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8614025400331309,
+      "grad_norm": 0.8814099431037903,
+      "learning_rate": 4.789317053878833e-06,
+      "loss": 0.681,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8622308117062396,
+      "grad_norm": 0.872502863407135,
+      "learning_rate": 4.788874975643449e-06,
+      "loss": 0.7044,
+      "step": 1041
+    },
+    {
+      "epoch": 0.8630590833793484,
+      "grad_norm": 0.9532367587089539,
+      "learning_rate": 4.788432454532093e-06,
+      "loss": 0.655,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8638873550524572,
+      "grad_norm": 0.9009373784065247,
+      "learning_rate": 4.787989490630391e-06,
+      "loss": 0.6946,
+      "step": 1043
+    },
+    {
+      "epoch": 0.864715626725566,
+      "grad_norm": 0.889933168888092,
+      "learning_rate": 4.787546084024051e-06,
+      "loss": 0.6595,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8655438983986747,
+      "grad_norm": 0.9179873466491699,
+      "learning_rate": 4.787102234798867e-06,
+      "loss": 0.658,
+      "step": 1045
+    },
+    {
+      "epoch": 0.8663721700717836,
+      "grad_norm": 0.8873292207717896,
+      "learning_rate": 4.786657943040721e-06,
+      "loss": 0.7023,
+      "step": 1046
+    },
+    {
+      "epoch": 0.8672004417448923,
+      "grad_norm": 0.8388610482215881,
+      "learning_rate": 4.78621320883558e-06,
+      "loss": 0.6644,
+      "step": 1047
+    },
+    {
+      "epoch": 0.8680287134180011,
+      "grad_norm": 0.8971349596977234,
+      "learning_rate": 4.785768032269493e-06,
+      "loss": 0.6552,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8688569850911099,
+      "grad_norm": 0.8789131045341492,
+      "learning_rate": 4.785322413428601e-06,
+      "loss": 0.6681,
+      "step": 1049
+    },
+    {
+      "epoch": 0.8696852567642187,
+      "grad_norm": 0.8867958784103394,
+      "learning_rate": 4.784876352399125e-06,
+      "loss": 0.6859,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8705135284373274,
+      "grad_norm": 0.8584237694740295,
+      "learning_rate": 4.784429849267373e-06,
+      "loss": 0.6655,
+      "step": 1051
+    },
+    {
+      "epoch": 0.8713418001104363,
+      "grad_norm": 0.9230811595916748,
+      "learning_rate": 4.783982904119743e-06,
+      "loss": 0.6648,
+      "step": 1052
+    },
+    {
+      "epoch": 0.872170071783545,
+      "grad_norm": 0.9217789769172668,
+      "learning_rate": 4.783535517042709e-06,
+      "loss": 0.6867,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8729983434566538,
+      "grad_norm": 0.9209116697311401,
+      "learning_rate": 4.783087688122842e-06,
+      "loss": 0.6972,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8738266151297626,
+      "grad_norm": 0.9493632316589355,
+      "learning_rate": 4.782639417446789e-06,
+      "loss": 0.6785,
+      "step": 1055
+    },
+    {
+      "epoch": 0.8746548868028713,
+      "grad_norm": 0.8790786862373352,
+      "learning_rate": 4.782190705101286e-06,
+      "loss": 0.6925,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8754831584759801,
+      "grad_norm": 0.8662691712379456,
+      "learning_rate": 4.781741551173157e-06,
+      "loss": 0.6686,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8763114301490889,
+      "grad_norm": 0.919658899307251,
+      "learning_rate": 4.781291955749307e-06,
+      "loss": 0.6814,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8771397018221977,
+      "grad_norm": 0.9639708399772644,
+      "learning_rate": 4.780841918916731e-06,
+      "loss": 0.6629,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8779679734953064,
+      "grad_norm": 0.8694306015968323,
+      "learning_rate": 4.780391440762505e-06,
+      "loss": 0.6893,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8787962451684153,
+      "grad_norm": 0.9725603461265564,
+      "learning_rate": 4.779940521373793e-06,
+      "loss": 0.67,
+      "step": 1061
+    },
+    {
+      "epoch": 0.879624516841524,
+      "grad_norm": 0.93159419298172,
+      "learning_rate": 4.779489160837842e-06,
+      "loss": 0.686,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8804527885146328,
+      "grad_norm": 0.9366991519927979,
+      "learning_rate": 4.779037359241989e-06,
+      "loss": 0.649,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8812810601877415,
+      "grad_norm": 0.9052289724349976,
+      "learning_rate": 4.7785851166736505e-06,
+      "loss": 0.6677,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8821093318608504,
+      "grad_norm": 0.882251501083374,
+      "learning_rate": 4.778132433220334e-06,
+      "loss": 0.6886,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8829376035339591,
+      "grad_norm": 0.8867698907852173,
+      "learning_rate": 4.777679308969627e-06,
+      "loss": 0.6645,
+      "step": 1066
+    },
+    {
+      "epoch": 0.883765875207068,
+      "grad_norm": 0.890396773815155,
+      "learning_rate": 4.777225744009207e-06,
+      "loss": 0.6684,
+      "step": 1067
+    },
+    {
+      "epoch": 0.8845941468801767,
+      "grad_norm": 0.8827376365661621,
+      "learning_rate": 4.776771738426833e-06,
+      "loss": 0.6738,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8854224185532855,
+      "grad_norm": 0.8953343033790588,
+      "learning_rate": 4.776317292310352e-06,
+      "loss": 0.684,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8862506902263942,
+      "grad_norm": 0.8751050233840942,
+      "learning_rate": 4.775862405747694e-06,
+      "loss": 0.6779,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8870789618995031,
+      "grad_norm": 0.9136618971824646,
+      "learning_rate": 4.775407078826876e-06,
+      "loss": 0.6939,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8879072335726118,
+      "grad_norm": 0.8887941837310791,
+      "learning_rate": 4.774951311635999e-06,
+      "loss": 0.6547,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8887355052457206,
+      "grad_norm": 0.8873274326324463,
+      "learning_rate": 4.77449510426325e-06,
+      "loss": 0.6769,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8895637769188294,
+      "grad_norm": 0.9307074546813965,
+      "learning_rate": 4.774038456796901e-06,
+      "loss": 0.6984,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8903920485919381,
+      "grad_norm": 0.9022747874259949,
+      "learning_rate": 4.77358136932531e-06,
+      "loss": 0.6733,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8912203202650469,
+      "grad_norm": 0.8779094219207764,
+      "learning_rate": 4.773123841936916e-06,
+      "loss": 0.6434,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8920485919381557,
+      "grad_norm": 0.9465747475624084,
+      "learning_rate": 4.7726658747202505e-06,
+      "loss": 0.6761,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8928768636112645,
+      "grad_norm": 0.8694362044334412,
+      "learning_rate": 4.772207467763923e-06,
+      "loss": 0.6756,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8937051352843732,
+      "grad_norm": 0.8839150071144104,
+      "learning_rate": 4.771748621156631e-06,
+      "loss": 0.6714,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8945334069574821,
+      "grad_norm": 0.9220975041389465,
+      "learning_rate": 4.7712893349871585e-06,
+      "loss": 0.6724,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8953616786305908,
+      "grad_norm": 0.9234760999679565,
+      "learning_rate": 4.770829609344372e-06,
+      "loss": 0.67,
+      "step": 1081
+    },
+    {
+      "epoch": 0.8961899503036996,
+      "grad_norm": 0.9262274503707886,
+      "learning_rate": 4.770369444317225e-06,
+      "loss": 0.6619,
+      "step": 1082
+    },
+    {
+      "epoch": 0.8970182219768084,
+      "grad_norm": 0.8889190554618835,
+      "learning_rate": 4.769908839994753e-06,
+      "loss": 0.688,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8978464936499172,
+      "grad_norm": 0.9028982520103455,
+      "learning_rate": 4.769447796466082e-06,
+      "loss": 0.6633,
+      "step": 1084
+    },
+    {
+      "epoch": 0.8986747653230259,
+      "grad_norm": 0.9155063033103943,
+      "learning_rate": 4.7689863138204185e-06,
+      "loss": 0.6709,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8995030369961348,
+      "grad_norm": 0.9066855311393738,
+      "learning_rate": 4.768524392147052e-06,
+      "loss": 0.6713,
+      "step": 1086
+    },
+    {
+      "epoch": 0.9003313086692435,
+      "grad_norm": 0.9365159273147583,
+      "learning_rate": 4.768062031535365e-06,
+      "loss": 0.6675,
+      "step": 1087
+    },
+    {
+      "epoch": 0.9011595803423523,
+      "grad_norm": 0.9106162190437317,
+      "learning_rate": 4.767599232074816e-06,
+      "loss": 0.6804,
+      "step": 1088
+    },
+    {
+      "epoch": 0.9019878520154611,
+      "grad_norm": 0.9183253645896912,
+      "learning_rate": 4.767135993854955e-06,
+      "loss": 0.658,
+      "step": 1089
+    },
+    {
+      "epoch": 0.9028161236885699,
+      "grad_norm": 0.9543409943580627,
+      "learning_rate": 4.766672316965413e-06,
+      "loss": 0.696,
+      "step": 1090
+    },
+    {
+      "epoch": 0.9036443953616786,
+      "grad_norm": 0.8894531726837158,
+      "learning_rate": 4.766208201495906e-06,
+      "loss": 0.6781,
+      "step": 1091
+    },
+    {
+      "epoch": 0.9044726670347875,
+      "grad_norm": 0.915542721748352,
+      "learning_rate": 4.765743647536238e-06,
+      "loss": 0.6804,
+      "step": 1092
+    },
+    {
+      "epoch": 0.9053009387078962,
+      "grad_norm": 0.8959763646125793,
+      "learning_rate": 4.765278655176296e-06,
+      "loss": 0.6873,
+      "step": 1093
+    },
+    {
+      "epoch": 0.906129210381005,
+      "grad_norm": 0.881188690662384,
+      "learning_rate": 4.76481322450605e-06,
+      "loss": 0.6582,
+      "step": 1094
+    },
+    {
+      "epoch": 0.9069574820541138,
+      "grad_norm": 0.9096754193305969,
+      "learning_rate": 4.764347355615558e-06,
+      "loss": 0.693,
+      "step": 1095
+    },
+    {
+      "epoch": 0.9077857537272225,
+      "grad_norm": 0.891132652759552,
+      "learning_rate": 4.76388104859496e-06,
+      "loss": 0.6781,
+      "step": 1096
+    },
+    {
+      "epoch": 0.9086140254003313,
+      "grad_norm": 0.893624484539032,
+      "learning_rate": 4.763414303534484e-06,
+      "loss": 0.6636,
+      "step": 1097
+    },
+    {
+      "epoch": 0.90944229707344,
+      "grad_norm": 0.8795678019523621,
+      "learning_rate": 4.762947120524438e-06,
+      "loss": 0.6604,
+      "step": 1098
+    },
+    {
+      "epoch": 0.9102705687465489,
+      "grad_norm": 0.9177265763282776,
+      "learning_rate": 4.76247949965522e-06,
+      "loss": 0.6557,
+      "step": 1099
+    },
+    {
+      "epoch": 0.9110988404196576,
+      "grad_norm": 0.9416985511779785,
+      "learning_rate": 4.762011441017309e-06,
+      "loss": 0.7008,
+      "step": 1100
+    },
+    {
+      "epoch": 0.9119271120927664,
+      "grad_norm": 0.9255479574203491,
+      "learning_rate": 4.7615429447012704e-06,
+      "loss": 0.6665,
+      "step": 1101
+    },
+    {
+      "epoch": 0.9127553837658752,
+      "grad_norm": 0.8948953151702881,
+      "learning_rate": 4.761074010797754e-06,
+      "loss": 0.6699,
+      "step": 1102
+    },
+    {
+      "epoch": 0.913583655438984,
+      "grad_norm": 0.9027442336082458,
+      "learning_rate": 4.760604639397493e-06,
+      "loss": 0.6649,
+      "step": 1103
+    },
+    {
+      "epoch": 0.9144119271120927,
+      "grad_norm": 0.9094238877296448,
+      "learning_rate": 4.7601348305913085e-06,
+      "loss": 0.6753,
+      "step": 1104
+    },
+    {
+      "epoch": 0.9152401987852016,
+      "grad_norm": 0.9098204374313354,
+      "learning_rate": 4.759664584470103e-06,
+      "loss": 0.6845,
+      "step": 1105
+    },
+    {
+      "epoch": 0.9160684704583103,
+      "grad_norm": 0.9339573383331299,
+      "learning_rate": 4.759193901124864e-06,
+      "loss": 0.6285,
+      "step": 1106
+    },
+    {
+      "epoch": 0.9168967421314191,
+      "grad_norm": 0.8812451958656311,
+      "learning_rate": 4.758722780646666e-06,
+      "loss": 0.6764,
+      "step": 1107
+    },
+    {
+      "epoch": 0.9177250138045279,
+      "grad_norm": 0.8832173347473145,
+      "learning_rate": 4.758251223126665e-06,
+      "loss": 0.6908,
+      "step": 1108
+    },
+    {
+      "epoch": 0.9185532854776367,
+      "grad_norm": 0.9124787449836731,
+      "learning_rate": 4.757779228656103e-06,
+      "loss": 0.6855,
+      "step": 1109
+    },
+    {
+      "epoch": 0.9193815571507454,
+      "grad_norm": 0.8778328895568848,
+      "learning_rate": 4.757306797326309e-06,
+      "loss": 0.6404,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9202098288238543,
+      "grad_norm": 0.8831816911697388,
+      "learning_rate": 4.75683392922869e-06,
+      "loss": 0.6425,
+      "step": 1111
+    },
+    {
+      "epoch": 0.921038100496963,
+      "grad_norm": 0.9060006141662598,
+      "learning_rate": 4.756360624454744e-06,
+      "loss": 0.6738,
+      "step": 1112
+    },
+    {
+      "epoch": 0.9218663721700718,
+      "grad_norm": 0.8746777176856995,
+      "learning_rate": 4.755886883096051e-06,
+      "loss": 0.6665,
+      "step": 1113
+    },
+    {
+      "epoch": 0.9226946438431806,
+      "grad_norm": 0.9185945987701416,
+      "learning_rate": 4.755412705244274e-06,
+      "loss": 0.6564,
+      "step": 1114
+    },
+    {
+      "epoch": 0.9235229155162893,
+      "grad_norm": 0.9648581743240356,
+      "learning_rate": 4.754938090991165e-06,
+      "loss": 0.6597,
+      "step": 1115
+    },
+    {
+      "epoch": 0.9243511871893981,
+      "grad_norm": 0.8800092935562134,
+      "learning_rate": 4.754463040428555e-06,
+      "loss": 0.6884,
+      "step": 1116
+    },
+    {
+      "epoch": 0.9251794588625069,
+      "grad_norm": 0.8659892082214355,
+      "learning_rate": 4.753987553648363e-06,
+      "loss": 0.68,
+      "step": 1117
+    },
+    {
+      "epoch": 0.9260077305356157,
+      "grad_norm": 0.88991379737854,
+      "learning_rate": 4.75351163074259e-06,
+      "loss": 0.668,
+      "step": 1118
+    },
+    {
+      "epoch": 0.9268360022087244,
+      "grad_norm": 0.9683054089546204,
+      "learning_rate": 4.753035271803323e-06,
+      "loss": 0.6789,
+      "step": 1119
+    },
+    {
+      "epoch": 0.9276642738818333,
+      "grad_norm": 0.928001880645752,
+      "learning_rate": 4.752558476922735e-06,
+      "loss": 0.6786,
+      "step": 1120
+    },
+    {
+      "epoch": 0.928492545554942,
+      "grad_norm": 0.8760973215103149,
+      "learning_rate": 4.752081246193078e-06,
+      "loss": 0.6716,
+      "step": 1121
+    },
+    {
+      "epoch": 0.9293208172280508,
+      "grad_norm": 0.9171329140663147,
+      "learning_rate": 4.751603579706695e-06,
+      "loss": 0.662,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9301490889011595,
+      "grad_norm": 0.9380231499671936,
+      "learning_rate": 4.751125477556008e-06,
+      "loss": 0.6836,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9309773605742684,
+      "grad_norm": 0.9471802711486816,
+      "learning_rate": 4.750646939833526e-06,
+      "loss": 0.668,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9318056322473771,
+      "grad_norm": 0.8860809206962585,
+      "learning_rate": 4.750167966631842e-06,
+      "loss": 0.6635,
+      "step": 1125
+    },
+    {
+      "epoch": 0.932633903920486,
+      "grad_norm": 0.8694644570350647,
+      "learning_rate": 4.7496885580436305e-06,
+      "loss": 0.6505,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9334621755935947,
+      "grad_norm": 0.9602686166763306,
+      "learning_rate": 4.749208714161655e-06,
+      "loss": 0.6874,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9342904472667035,
+      "grad_norm": 0.9198856353759766,
+      "learning_rate": 4.748728435078761e-06,
+      "loss": 0.697,
+      "step": 1128
+    },
+    {
+      "epoch": 0.9351187189398122,
+      "grad_norm": 0.9162499904632568,
+      "learning_rate": 4.748247720887876e-06,
+      "loss": 0.6521,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9359469906129211,
+      "grad_norm": 0.9064576029777527,
+      "learning_rate": 4.747766571682016e-06,
+      "loss": 0.6525,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9367752622860298,
+      "grad_norm": 0.9518057703971863,
+      "learning_rate": 4.747284987554277e-06,
+      "loss": 0.6745,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9376035339591386,
+      "grad_norm": 0.8963000178337097,
+      "learning_rate": 4.746802968597842e-06,
+      "loss": 0.6999,
+      "step": 1132
+    },
+    {
+      "epoch": 0.9384318056322474,
+      "grad_norm": 0.8687212467193604,
+      "learning_rate": 4.746320514905977e-06,
+      "loss": 0.6353,
+      "step": 1133
+    },
+    {
+      "epoch": 0.9392600773053562,
+      "grad_norm": 0.8598489165306091,
+      "learning_rate": 4.745837626572033e-06,
+      "loss": 0.6897,
+      "step": 1134
+    },
+    {
+      "epoch": 0.9400883489784649,
+      "grad_norm": 0.9425498843193054,
+      "learning_rate": 4.745354303689444e-06,
+      "loss": 0.667,
+      "step": 1135
+    },
+    {
+      "epoch": 0.9409166206515737,
+      "grad_norm": 0.9026349782943726,
+      "learning_rate": 4.744870546351727e-06,
+      "loss": 0.676,
+      "step": 1136
+    },
+    {
+      "epoch": 0.9417448923246825,
+      "grad_norm": 0.8824150562286377,
+      "learning_rate": 4.744386354652486e-06,
+      "loss": 0.6941,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9425731639977912,
+      "grad_norm": 0.8910468816757202,
+      "learning_rate": 4.743901728685406e-06,
+      "loss": 0.6701,
+      "step": 1138
+    },
+    {
+      "epoch": 0.9434014356709001,
+      "grad_norm": 0.9114013314247131,
+      "learning_rate": 4.74341666854426e-06,
+      "loss": 0.6735,
+      "step": 1139
+    },
+    {
+      "epoch": 0.9442297073440088,
+      "grad_norm": 0.9107307195663452,
+      "learning_rate": 4.742931174322902e-06,
+      "loss": 0.6505,
+      "step": 1140
+    },
+    {
+      "epoch": 0.9450579790171176,
+      "grad_norm": 0.8906923532485962,
+      "learning_rate": 4.742445246115271e-06,
+      "loss": 0.6648,
+      "step": 1141
+    },
+    {
+      "epoch": 0.9458862506902264,
+      "grad_norm": 0.902644157409668,
+      "learning_rate": 4.741958884015387e-06,
+      "loss": 0.6674,
+      "step": 1142
+    },
+    {
+      "epoch": 0.9467145223633352,
+      "grad_norm": 0.9522087574005127,
+      "learning_rate": 4.741472088117358e-06,
+      "loss": 0.6825,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9475427940364439,
+      "grad_norm": 0.9152255654335022,
+      "learning_rate": 4.740984858515376e-06,
+      "loss": 0.6708,
+      "step": 1144
+    },
+    {
+      "epoch": 0.9483710657095528,
+      "grad_norm": 0.8779835104942322,
+      "learning_rate": 4.740497195303713e-06,
+      "loss": 0.6562,
+      "step": 1145
+    },
+    {
+      "epoch": 0.9491993373826615,
+      "grad_norm": 0.8842780590057373,
+      "learning_rate": 4.740009098576729e-06,
+      "loss": 0.6769,
+      "step": 1146
+    },
+    {
+      "epoch": 0.9500276090557703,
+      "grad_norm": 0.9137415289878845,
+      "learning_rate": 4.739520568428866e-06,
+      "loss": 0.6705,
+      "step": 1147
+    },
+    {
+      "epoch": 0.950855880728879,
+      "grad_norm": 0.8867765665054321,
+      "learning_rate": 4.739031604954648e-06,
+      "loss": 0.659,
+      "step": 1148
+    },
+    {
+      "epoch": 0.9516841524019879,
+      "grad_norm": 0.8965260982513428,
+      "learning_rate": 4.7385422082486874e-06,
+      "loss": 0.6744,
+      "step": 1149
+    },
+    {
+      "epoch": 0.9525124240750966,
+      "grad_norm": 0.8861697912216187,
+      "learning_rate": 4.738052378405677e-06,
+      "loss": 0.6575,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9533406957482055,
+      "grad_norm": 0.9442310333251953,
+      "learning_rate": 4.737562115520394e-06,
+      "loss": 0.7164,
+      "step": 1151
+    },
+    {
+      "epoch": 0.9541689674213142,
+      "grad_norm": 0.9340727925300598,
+      "learning_rate": 4.7370714196877e-06,
+      "loss": 0.6577,
+      "step": 1152
+    },
+    {
+      "epoch": 0.954997239094423,
+      "grad_norm": 0.9337487816810608,
+      "learning_rate": 4.7365802910025405e-06,
+      "loss": 0.6664,
+      "step": 1153
+    },
+    {
+      "epoch": 0.9558255107675317,
+      "grad_norm": 0.8849001526832581,
+      "learning_rate": 4.736088729559943e-06,
+      "loss": 0.6721,
+      "step": 1154
+    },
+    {
+      "epoch": 0.9566537824406405,
+      "grad_norm": 0.917323648929596,
+      "learning_rate": 4.735596735455022e-06,
+      "loss": 0.6715,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9574820541137493,
+      "grad_norm": 0.9480959177017212,
+      "learning_rate": 4.735104308782972e-06,
+      "loss": 0.6784,
+      "step": 1156
+    },
+    {
+      "epoch": 0.958310325786858,
+      "grad_norm": 0.9519251585006714,
+      "learning_rate": 4.734611449639073e-06,
+      "loss": 0.6617,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9591385974599669,
+      "grad_norm": 0.8652079105377197,
+      "learning_rate": 4.734118158118689e-06,
+      "loss": 0.6846,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9599668691330756,
+      "grad_norm": 0.8860392570495605,
+      "learning_rate": 4.733624434317269e-06,
+      "loss": 0.6729,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9607951408061844,
+      "grad_norm": 0.9467882513999939,
+      "learning_rate": 4.7331302783303416e-06,
+      "loss": 0.6362,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9616234124792932,
+      "grad_norm": 0.9138529896736145,
+      "learning_rate": 4.732635690253523e-06,
+      "loss": 0.6613,
+      "step": 1161
+    },
+    {
+      "epoch": 0.962451684152402,
+      "grad_norm": 0.9401054978370667,
+      "learning_rate": 4.7321406701825105e-06,
+      "loss": 0.6845,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9632799558255107,
+      "grad_norm": 0.9049288630485535,
+      "learning_rate": 4.731645218213086e-06,
+      "loss": 0.6857,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9641082274986196,
+      "grad_norm": 0.8972534537315369,
+      "learning_rate": 4.731149334441114e-06,
+      "loss": 0.6745,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9649364991717283,
+      "grad_norm": 1.0799483060836792,
+      "learning_rate": 4.730653018962545e-06,
+      "loss": 0.6754,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9657647708448371,
+      "grad_norm": 0.9841833710670471,
+      "learning_rate": 4.730156271873412e-06,
+      "loss": 0.6718,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9665930425179459,
+      "grad_norm": 0.9388216733932495,
+      "learning_rate": 4.729659093269828e-06,
+      "loss": 0.6622,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9674213141910547,
+      "grad_norm": 0.8879573345184326,
+      "learning_rate": 4.729161483247994e-06,
+      "loss": 0.6909,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9682495858641634,
+      "grad_norm": 0.9713517427444458,
+      "learning_rate": 4.728663441904194e-06,
+      "loss": 0.6956,
+      "step": 1169
+    },
+    {
+      "epoch": 0.9690778575372723,
+      "grad_norm": 0.9200823903083801,
+      "learning_rate": 4.7281649693347925e-06,
+      "loss": 0.6766,
+      "step": 1170
+    },
+    {
+      "epoch": 0.969906129210381,
+      "grad_norm": 0.8983714580535889,
+      "learning_rate": 4.727666065636241e-06,
+      "loss": 0.6626,
+      "step": 1171
+    },
+    {
+      "epoch": 0.9707344008834898,
+      "grad_norm": 0.8792792558670044,
+      "learning_rate": 4.72716673090507e-06,
+      "loss": 0.666,
+      "step": 1172
+    },
+    {
+      "epoch": 0.9715626725565986,
+      "grad_norm": 1.0116846561431885,
+      "learning_rate": 4.7266669652379005e-06,
+      "loss": 0.6574,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9723909442297074,
+      "grad_norm": 0.9433228373527527,
+      "learning_rate": 4.7261667687314285e-06,
+      "loss": 0.6972,
+      "step": 1174
+    },
+    {
+      "epoch": 0.9732192159028161,
+      "grad_norm": 0.8979770541191101,
+      "learning_rate": 4.725666141482439e-06,
+      "loss": 0.6523,
+      "step": 1175
+    },
+    {
+      "epoch": 0.9740474875759249,
+      "grad_norm": 0.8990604281425476,
+      "learning_rate": 4.725165083587798e-06,
+      "loss": 0.6535,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9748757592490337,
+      "grad_norm": 0.9411956667900085,
+      "learning_rate": 4.724663595144457e-06,
+      "loss": 0.6947,
+      "step": 1177
+    },
+    {
+      "epoch": 0.9757040309221424,
+      "grad_norm": 1.0203242301940918,
+      "learning_rate": 4.724161676249448e-06,
+      "loss": 0.6894,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9765323025952513,
+      "grad_norm": 0.8938170075416565,
+      "learning_rate": 4.723659326999888e-06,
+      "loss": 0.6648,
+      "step": 1179
+    },
+    {
+      "epoch": 0.97736057426836,
+      "grad_norm": 0.8810412883758545,
+      "learning_rate": 4.7231565474929765e-06,
+      "loss": 0.6761,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9781888459414688,
+      "grad_norm": 0.8903076648712158,
+      "learning_rate": 4.722653337825998e-06,
+      "loss": 0.6738,
+      "step": 1181
+    },
+    {
+      "epoch": 0.9790171176145775,
+      "grad_norm": 0.905060350894928,
+      "learning_rate": 4.7221496980963175e-06,
+      "loss": 0.6625,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9798453892876864,
+      "grad_norm": 0.8791100978851318,
+      "learning_rate": 4.721645628401385e-06,
+      "loss": 0.6884,
+      "step": 1183
+    },
+    {
+      "epoch": 0.9806736609607951,
+      "grad_norm": 0.897101104259491,
+      "learning_rate": 4.721141128838733e-06,
+      "loss": 0.6693,
+      "step": 1184
+    },
+    {
+      "epoch": 0.981501932633904,
+      "grad_norm": 0.9102833867073059,
+      "learning_rate": 4.7206361995059775e-06,
+      "loss": 0.6682,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9823302043070127,
+      "grad_norm": 0.9492197632789612,
+      "learning_rate": 4.720130840500819e-06,
+      "loss": 0.6501,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9831584759801215,
+      "grad_norm": 0.9783265590667725,
+      "learning_rate": 4.719625051921038e-06,
+      "loss": 0.6454,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9839867476532302,
+      "grad_norm": 0.8938109874725342,
+      "learning_rate": 4.719118833864501e-06,
+      "loss": 0.6671,
+      "step": 1188
+    },
+    {
+      "epoch": 0.9848150193263391,
+      "grad_norm": 0.897484302520752,
+      "learning_rate": 4.718612186429156e-06,
+      "loss": 0.6664,
+      "step": 1189
+    },
+    {
+      "epoch": 0.9856432909994478,
+      "grad_norm": 0.9640376567840576,
+      "learning_rate": 4.718105109713034e-06,
+      "loss": 0.6845,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9864715626725566,
+      "grad_norm": 0.9495493769645691,
+      "learning_rate": 4.7175976038142505e-06,
+      "loss": 0.6864,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9872998343456654,
+      "grad_norm": 0.9390130639076233,
+      "learning_rate": 4.717089668831002e-06,
+      "loss": 0.6734,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9881281060187742,
+      "grad_norm": 0.8854318261146545,
+      "learning_rate": 4.716581304861571e-06,
+      "loss": 0.6546,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9889563776918829,
+      "grad_norm": 0.9242726564407349,
+      "learning_rate": 4.71607251200432e-06,
+      "loss": 0.6663,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9897846493649917,
+      "grad_norm": 0.9302003383636475,
+      "learning_rate": 4.715563290357696e-06,
+      "loss": 0.6844,
+      "step": 1195
+    },
+    {
+      "epoch": 0.9906129210381005,
+      "grad_norm": 0.9279932975769043,
+      "learning_rate": 4.715053640020228e-06,
+      "loss": 0.6731,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9914411927112092,
+      "grad_norm": 0.8753612041473389,
+      "learning_rate": 4.71454356109053e-06,
+      "loss": 0.6738,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9922694643843181,
+      "grad_norm": 0.890133261680603,
+      "learning_rate": 4.714033053667296e-06,
+      "loss": 0.6638,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9930977360574268,
+      "grad_norm": 0.9343729615211487,
+      "learning_rate": 4.713522117849305e-06,
+      "loss": 0.6826,
+      "step": 1199
+    },
+    {
+      "epoch": 0.9939260077305356,
+      "grad_norm": 0.9052755832672119,
+      "learning_rate": 4.7130107537354185e-06,
+      "loss": 0.6897,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9947542794036444,
+      "grad_norm": 0.8683924674987793,
+      "learning_rate": 4.7124989614245805e-06,
+      "loss": 0.6687,
+      "step": 1201
+    },
+    {
+      "epoch": 0.9955825510767532,
+      "grad_norm": 0.8677132725715637,
+      "learning_rate": 4.711986741015818e-06,
+      "loss": 0.6917,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9964108227498619,
+      "grad_norm": 0.8840690851211548,
+      "learning_rate": 4.711474092608243e-06,
+      "loss": 0.6841,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9972390944229708,
+      "grad_norm": 0.9750391840934753,
+      "learning_rate": 4.7109610163010435e-06,
+      "loss": 0.6574,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9980673660960795,
+      "grad_norm": 0.8752943873405457,
+      "learning_rate": 4.710447512193499e-06,
+      "loss": 0.636,
+      "step": 1205
+    },
+    {
+      "epoch": 0.9988956377691883,
+      "grad_norm": 0.9100313782691956,
+      "learning_rate": 4.709933580384965e-06,
+      "loss": 0.6712,
+      "step": 1206
+    },
+    {
+      "epoch": 0.999723909442297,
+      "grad_norm": 0.9202356338500977,
+      "learning_rate": 4.709419220974886e-06,
+      "loss": 0.6641,
+      "step": 1207
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 7242,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 1207,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.048926292989706e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}