model_pblpfs3g / checkpoint-2414 /trainer_state.json
ugaoo's picture
Upload folder using huggingface_hub
081b0d2 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9988956377691882,
"eval_steps": 500,
"global_step": 2414,
"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
},
{
"epoch": 1.0,
"grad_norm": 0.9202356338500977,
"learning_rate": 4.708904434062783e-06,
"loss": 0.6199,
"step": 1208
},
{
"epoch": 1.0008282716731087,
"grad_norm": 1.6839245557785034,
"learning_rate": 4.708389219748263e-06,
"loss": 0.643,
"step": 1209
},
{
"epoch": 1.0016565433462175,
"grad_norm": 0.8636482357978821,
"learning_rate": 4.707873578131015e-06,
"loss": 0.6467,
"step": 1210
},
{
"epoch": 1.0024848150193264,
"grad_norm": 0.8548988103866577,
"learning_rate": 4.707357509310811e-06,
"loss": 0.6574,
"step": 1211
},
{
"epoch": 1.0033130866924351,
"grad_norm": 0.8961352705955505,
"learning_rate": 4.7068410133875066e-06,
"loss": 0.6367,
"step": 1212
},
{
"epoch": 1.0041413583655439,
"grad_norm": 0.8914269208908081,
"learning_rate": 4.706324090461037e-06,
"loss": 0.6294,
"step": 1213
},
{
"epoch": 1.0049696300386526,
"grad_norm": 0.879915177822113,
"learning_rate": 4.7058067406314235e-06,
"loss": 0.6346,
"step": 1214
},
{
"epoch": 1.0057979017117615,
"grad_norm": 0.8798143863677979,
"learning_rate": 4.7052889639987674e-06,
"loss": 0.6559,
"step": 1215
},
{
"epoch": 1.0066261733848703,
"grad_norm": 0.8888561129570007,
"learning_rate": 4.704770760663255e-06,
"loss": 0.6471,
"step": 1216
},
{
"epoch": 1.007454445057979,
"grad_norm": 0.8707575798034668,
"learning_rate": 4.704252130725152e-06,
"loss": 0.6485,
"step": 1217
},
{
"epoch": 1.0082827167310877,
"grad_norm": 0.8609346151351929,
"learning_rate": 4.7037330742848105e-06,
"loss": 0.6544,
"step": 1218
},
{
"epoch": 1.0091109884041967,
"grad_norm": 0.8588053584098816,
"learning_rate": 4.703213591442663e-06,
"loss": 0.6421,
"step": 1219
},
{
"epoch": 1.0099392600773054,
"grad_norm": 0.8932695388793945,
"learning_rate": 4.702693682299222e-06,
"loss": 0.6333,
"step": 1220
},
{
"epoch": 1.0107675317504141,
"grad_norm": 0.8615652322769165,
"learning_rate": 4.7021733469550865e-06,
"loss": 0.6371,
"step": 1221
},
{
"epoch": 1.0115958034235228,
"grad_norm": 0.8876634240150452,
"learning_rate": 4.7016525855109376e-06,
"loss": 0.6368,
"step": 1222
},
{
"epoch": 1.0124240750966318,
"grad_norm": 0.895575225353241,
"learning_rate": 4.701131398067537e-06,
"loss": 0.6226,
"step": 1223
},
{
"epoch": 1.0132523467697405,
"grad_norm": 0.8836749196052551,
"learning_rate": 4.70060978472573e-06,
"loss": 0.6516,
"step": 1224
},
{
"epoch": 1.0140806184428492,
"grad_norm": 0.9471280574798584,
"learning_rate": 4.700087745586442e-06,
"loss": 0.633,
"step": 1225
},
{
"epoch": 1.014908890115958,
"grad_norm": 0.8721561431884766,
"learning_rate": 4.699565280750685e-06,
"loss": 0.6568,
"step": 1226
},
{
"epoch": 1.015737161789067,
"grad_norm": 0.8946447372436523,
"learning_rate": 4.69904239031955e-06,
"loss": 0.6503,
"step": 1227
},
{
"epoch": 1.0165654334621756,
"grad_norm": 0.8854631185531616,
"learning_rate": 4.698519074394212e-06,
"loss": 0.6341,
"step": 1228
},
{
"epoch": 1.0173937051352844,
"grad_norm": 0.9348397850990295,
"learning_rate": 4.697995333075927e-06,
"loss": 0.6317,
"step": 1229
},
{
"epoch": 1.018221976808393,
"grad_norm": 0.9037336111068726,
"learning_rate": 4.697471166466035e-06,
"loss": 0.6618,
"step": 1230
},
{
"epoch": 1.0190502484815018,
"grad_norm": 0.8866400718688965,
"learning_rate": 4.696946574665957e-06,
"loss": 0.6421,
"step": 1231
},
{
"epoch": 1.0198785201546108,
"grad_norm": 0.8897701501846313,
"learning_rate": 4.696421557777196e-06,
"loss": 0.6608,
"step": 1232
},
{
"epoch": 1.0207067918277195,
"grad_norm": 0.9046173095703125,
"learning_rate": 4.6958961159013386e-06,
"loss": 0.6464,
"step": 1233
},
{
"epoch": 1.0215350635008282,
"grad_norm": 0.9492906332015991,
"learning_rate": 4.695370249140052e-06,
"loss": 0.6578,
"step": 1234
},
{
"epoch": 1.022363335173937,
"grad_norm": 0.8906951546669006,
"learning_rate": 4.6948439575950885e-06,
"loss": 0.6276,
"step": 1235
},
{
"epoch": 1.023191606847046,
"grad_norm": 0.9218266606330872,
"learning_rate": 4.694317241368278e-06,
"loss": 0.627,
"step": 1236
},
{
"epoch": 1.0240198785201546,
"grad_norm": 0.901157557964325,
"learning_rate": 4.693790100561537e-06,
"loss": 0.6379,
"step": 1237
},
{
"epoch": 1.0248481501932634,
"grad_norm": 0.8992452025413513,
"learning_rate": 4.693262535276863e-06,
"loss": 0.6585,
"step": 1238
},
{
"epoch": 1.025676421866372,
"grad_norm": 0.9525597095489502,
"learning_rate": 4.692734545616331e-06,
"loss": 0.6625,
"step": 1239
},
{
"epoch": 1.026504693539481,
"grad_norm": 0.9229065775871277,
"learning_rate": 4.692206131682106e-06,
"loss": 0.6454,
"step": 1240
},
{
"epoch": 1.0273329652125898,
"grad_norm": 1.012373685836792,
"learning_rate": 4.691677293576431e-06,
"loss": 0.6492,
"step": 1241
},
{
"epoch": 1.0281612368856985,
"grad_norm": 0.8650060892105103,
"learning_rate": 4.69114803140163e-06,
"loss": 0.6739,
"step": 1242
},
{
"epoch": 1.0289895085588072,
"grad_norm": 0.9396429657936096,
"learning_rate": 4.69061834526011e-06,
"loss": 0.6729,
"step": 1243
},
{
"epoch": 1.0298177802319162,
"grad_norm": 0.9584783911705017,
"learning_rate": 4.6900882352543614e-06,
"loss": 0.6678,
"step": 1244
},
{
"epoch": 1.030646051905025,
"grad_norm": 0.985394299030304,
"learning_rate": 4.689557701486954e-06,
"loss": 0.6478,
"step": 1245
},
{
"epoch": 1.0314743235781336,
"grad_norm": 0.9014914631843567,
"learning_rate": 4.689026744060543e-06,
"loss": 0.6393,
"step": 1246
},
{
"epoch": 1.0323025952512423,
"grad_norm": 0.9904847145080566,
"learning_rate": 4.688495363077863e-06,
"loss": 0.6752,
"step": 1247
},
{
"epoch": 1.0331308669243513,
"grad_norm": 0.9439241886138916,
"learning_rate": 4.687963558641731e-06,
"loss": 0.6417,
"step": 1248
},
{
"epoch": 1.03395913859746,
"grad_norm": 0.9277704954147339,
"learning_rate": 4.687431330855046e-06,
"loss": 0.6493,
"step": 1249
},
{
"epoch": 1.0347874102705688,
"grad_norm": 0.9456571340560913,
"learning_rate": 4.686898679820789e-06,
"loss": 0.6774,
"step": 1250
},
{
"epoch": 1.0356156819436775,
"grad_norm": 0.9043509364128113,
"learning_rate": 4.686365605642025e-06,
"loss": 0.6198,
"step": 1251
},
{
"epoch": 1.0364439536167862,
"grad_norm": 0.9246267676353455,
"learning_rate": 4.685832108421896e-06,
"loss": 0.6513,
"step": 1252
},
{
"epoch": 1.0372722252898952,
"grad_norm": 0.9227612614631653,
"learning_rate": 4.685298188263631e-06,
"loss": 0.6597,
"step": 1253
},
{
"epoch": 1.0381004969630039,
"grad_norm": 0.8927509188652039,
"learning_rate": 4.684763845270537e-06,
"loss": 0.641,
"step": 1254
},
{
"epoch": 1.0389287686361126,
"grad_norm": 0.8812937140464783,
"learning_rate": 4.684229079546005e-06,
"loss": 0.6292,
"step": 1255
},
{
"epoch": 1.0397570403092213,
"grad_norm": 0.9186881184577942,
"learning_rate": 4.683693891193508e-06,
"loss": 0.6485,
"step": 1256
},
{
"epoch": 1.0405853119823303,
"grad_norm": 0.8928134441375732,
"learning_rate": 4.683158280316599e-06,
"loss": 0.6435,
"step": 1257
},
{
"epoch": 1.041413583655439,
"grad_norm": 0.8919920325279236,
"learning_rate": 4.682622247018913e-06,
"loss": 0.622,
"step": 1258
},
{
"epoch": 1.0422418553285477,
"grad_norm": 0.885259747505188,
"learning_rate": 4.682085791404169e-06,
"loss": 0.6521,
"step": 1259
},
{
"epoch": 1.0430701270016565,
"grad_norm": 0.9034169316291809,
"learning_rate": 4.681548913576165e-06,
"loss": 0.6437,
"step": 1260
},
{
"epoch": 1.0438983986747654,
"grad_norm": 0.884848952293396,
"learning_rate": 4.6810116136387825e-06,
"loss": 0.6733,
"step": 1261
},
{
"epoch": 1.0447266703478741,
"grad_norm": 0.9137836694717407,
"learning_rate": 4.680473891695985e-06,
"loss": 0.6415,
"step": 1262
},
{
"epoch": 1.0455549420209829,
"grad_norm": 0.9044780135154724,
"learning_rate": 4.6799357478518145e-06,
"loss": 0.6453,
"step": 1263
},
{
"epoch": 1.0463832136940916,
"grad_norm": 0.9111152291297913,
"learning_rate": 4.6793971822103985e-06,
"loss": 0.6644,
"step": 1264
},
{
"epoch": 1.0472114853672005,
"grad_norm": 0.9158084988594055,
"learning_rate": 4.678858194875944e-06,
"loss": 0.6434,
"step": 1265
},
{
"epoch": 1.0480397570403093,
"grad_norm": 0.8970723152160645,
"learning_rate": 4.678318785952739e-06,
"loss": 0.6391,
"step": 1266
},
{
"epoch": 1.048868028713418,
"grad_norm": 0.8958215117454529,
"learning_rate": 4.677778955545155e-06,
"loss": 0.6562,
"step": 1267
},
{
"epoch": 1.0496963003865267,
"grad_norm": 0.9229645729064941,
"learning_rate": 4.677238703757644e-06,
"loss": 0.646,
"step": 1268
},
{
"epoch": 1.0505245720596355,
"grad_norm": 0.8986689448356628,
"learning_rate": 4.676698030694741e-06,
"loss": 0.6241,
"step": 1269
},
{
"epoch": 1.0513528437327444,
"grad_norm": 0.8943328261375427,
"learning_rate": 4.6761569364610585e-06,
"loss": 0.6455,
"step": 1270
},
{
"epoch": 1.0521811154058531,
"grad_norm": 0.9147390127182007,
"learning_rate": 4.675615421161295e-06,
"loss": 0.6299,
"step": 1271
},
{
"epoch": 1.0530093870789619,
"grad_norm": 0.9124075174331665,
"learning_rate": 4.675073484900229e-06,
"loss": 0.6379,
"step": 1272
},
{
"epoch": 1.0538376587520706,
"grad_norm": 0.9267525672912598,
"learning_rate": 4.674531127782718e-06,
"loss": 0.6478,
"step": 1273
},
{
"epoch": 1.0546659304251795,
"grad_norm": 0.8710078001022339,
"learning_rate": 4.673988349913707e-06,
"loss": 0.6314,
"step": 1274
},
{
"epoch": 1.0554942020982883,
"grad_norm": 0.8905670642852783,
"learning_rate": 4.673445151398214e-06,
"loss": 0.6456,
"step": 1275
},
{
"epoch": 1.056322473771397,
"grad_norm": 0.9125292897224426,
"learning_rate": 4.672901532341346e-06,
"loss": 0.6439,
"step": 1276
},
{
"epoch": 1.0571507454445057,
"grad_norm": 0.9003625512123108,
"learning_rate": 4.6723574928482865e-06,
"loss": 0.6561,
"step": 1277
},
{
"epoch": 1.0579790171176147,
"grad_norm": 0.96099454164505,
"learning_rate": 4.6718130330243014e-06,
"loss": 0.6515,
"step": 1278
},
{
"epoch": 1.0588072887907234,
"grad_norm": 0.8841069340705872,
"learning_rate": 4.671268152974742e-06,
"loss": 0.6389,
"step": 1279
},
{
"epoch": 1.0596355604638321,
"grad_norm": 0.8906645178794861,
"learning_rate": 4.670722852805033e-06,
"loss": 0.6448,
"step": 1280
},
{
"epoch": 1.0604638321369408,
"grad_norm": 0.9538317918777466,
"learning_rate": 4.670177132620689e-06,
"loss": 0.6544,
"step": 1281
},
{
"epoch": 1.0612921038100498,
"grad_norm": 0.9158053994178772,
"learning_rate": 4.669630992527299e-06,
"loss": 0.6494,
"step": 1282
},
{
"epoch": 1.0621203754831585,
"grad_norm": 0.9756888151168823,
"learning_rate": 4.6690844326305374e-06,
"loss": 0.6419,
"step": 1283
},
{
"epoch": 1.0629486471562672,
"grad_norm": 0.905055820941925,
"learning_rate": 4.668537453036159e-06,
"loss": 0.6463,
"step": 1284
},
{
"epoch": 1.063776918829376,
"grad_norm": 0.9011061191558838,
"learning_rate": 4.667990053849997e-06,
"loss": 0.6353,
"step": 1285
},
{
"epoch": 1.0646051905024847,
"grad_norm": 0.955047070980072,
"learning_rate": 4.667442235177969e-06,
"loss": 0.6485,
"step": 1286
},
{
"epoch": 1.0654334621755936,
"grad_norm": 0.9253432750701904,
"learning_rate": 4.666893997126074e-06,
"loss": 0.6391,
"step": 1287
},
{
"epoch": 1.0662617338487024,
"grad_norm": 0.8840277194976807,
"learning_rate": 4.66634533980039e-06,
"loss": 0.6372,
"step": 1288
},
{
"epoch": 1.067090005521811,
"grad_norm": 1.0881837606430054,
"learning_rate": 4.6657962633070765e-06,
"loss": 0.6419,
"step": 1289
},
{
"epoch": 1.06791827719492,
"grad_norm": 0.9263116717338562,
"learning_rate": 4.665246767752376e-06,
"loss": 0.6226,
"step": 1290
},
{
"epoch": 1.0687465488680288,
"grad_norm": 0.9110594987869263,
"learning_rate": 4.66469685324261e-06,
"loss": 0.6495,
"step": 1291
},
{
"epoch": 1.0695748205411375,
"grad_norm": 0.9086888432502747,
"learning_rate": 4.664146519884182e-06,
"loss": 0.6408,
"step": 1292
},
{
"epoch": 1.0704030922142462,
"grad_norm": 0.8991205096244812,
"learning_rate": 4.663595767783575e-06,
"loss": 0.6424,
"step": 1293
},
{
"epoch": 1.071231363887355,
"grad_norm": 0.926933228969574,
"learning_rate": 4.6630445970473576e-06,
"loss": 0.6379,
"step": 1294
},
{
"epoch": 1.072059635560464,
"grad_norm": 0.8793080449104309,
"learning_rate": 4.662493007782174e-06,
"loss": 0.6254,
"step": 1295
},
{
"epoch": 1.0728879072335726,
"grad_norm": 0.9588155150413513,
"learning_rate": 4.661941000094751e-06,
"loss": 0.6307,
"step": 1296
},
{
"epoch": 1.0737161789066814,
"grad_norm": 0.9468860030174255,
"learning_rate": 4.661388574091898e-06,
"loss": 0.6308,
"step": 1297
},
{
"epoch": 1.07454445057979,
"grad_norm": 0.8930124640464783,
"learning_rate": 4.660835729880505e-06,
"loss": 0.6371,
"step": 1298
},
{
"epoch": 1.075372722252899,
"grad_norm": 0.9233878254890442,
"learning_rate": 4.660282467567541e-06,
"loss": 0.6327,
"step": 1299
},
{
"epoch": 1.0762009939260078,
"grad_norm": 0.9714951515197754,
"learning_rate": 4.659728787260057e-06,
"loss": 0.6218,
"step": 1300
},
{
"epoch": 1.0770292655991165,
"grad_norm": 0.9464428424835205,
"learning_rate": 4.659174689065185e-06,
"loss": 0.6406,
"step": 1301
},
{
"epoch": 1.0778575372722252,
"grad_norm": 0.8984559774398804,
"learning_rate": 4.6586201730901395e-06,
"loss": 0.6483,
"step": 1302
},
{
"epoch": 1.0786858089453342,
"grad_norm": 0.9007774591445923,
"learning_rate": 4.658065239442212e-06,
"loss": 0.6288,
"step": 1303
},
{
"epoch": 1.079514080618443,
"grad_norm": 0.907171368598938,
"learning_rate": 4.657509888228779e-06,
"loss": 0.6445,
"step": 1304
},
{
"epoch": 1.0803423522915516,
"grad_norm": 0.9207282066345215,
"learning_rate": 4.656954119557293e-06,
"loss": 0.6148,
"step": 1305
},
{
"epoch": 1.0811706239646603,
"grad_norm": 0.9467616677284241,
"learning_rate": 4.656397933535293e-06,
"loss": 0.6374,
"step": 1306
},
{
"epoch": 1.0819988956377693,
"grad_norm": 0.9310930967330933,
"learning_rate": 4.655841330270393e-06,
"loss": 0.6362,
"step": 1307
},
{
"epoch": 1.082827167310878,
"grad_norm": 0.890816330909729,
"learning_rate": 4.655284309870294e-06,
"loss": 0.6372,
"step": 1308
},
{
"epoch": 1.0836554389839868,
"grad_norm": 0.9206095933914185,
"learning_rate": 4.654726872442771e-06,
"loss": 0.6297,
"step": 1309
},
{
"epoch": 1.0844837106570955,
"grad_norm": 0.9015856385231018,
"learning_rate": 4.654169018095685e-06,
"loss": 0.658,
"step": 1310
},
{
"epoch": 1.0853119823302042,
"grad_norm": 0.8909441828727722,
"learning_rate": 4.653610746936975e-06,
"loss": 0.6382,
"step": 1311
},
{
"epoch": 1.0861402540033132,
"grad_norm": 0.9035969972610474,
"learning_rate": 4.653052059074661e-06,
"loss": 0.6441,
"step": 1312
},
{
"epoch": 1.0869685256764219,
"grad_norm": 0.8988869786262512,
"learning_rate": 4.652492954616844e-06,
"loss": 0.6359,
"step": 1313
},
{
"epoch": 1.0877967973495306,
"grad_norm": 0.881885826587677,
"learning_rate": 4.6519334336717055e-06,
"loss": 0.6397,
"step": 1314
},
{
"epoch": 1.0886250690226393,
"grad_norm": 0.980633020401001,
"learning_rate": 4.651373496347508e-06,
"loss": 0.6223,
"step": 1315
},
{
"epoch": 1.0894533406957483,
"grad_norm": 0.9027420878410339,
"learning_rate": 4.650813142752593e-06,
"loss": 0.6211,
"step": 1316
},
{
"epoch": 1.090281612368857,
"grad_norm": 0.9011778831481934,
"learning_rate": 4.6502523729953855e-06,
"loss": 0.6442,
"step": 1317
},
{
"epoch": 1.0911098840419657,
"grad_norm": 0.926692545413971,
"learning_rate": 4.649691187184387e-06,
"loss": 0.6661,
"step": 1318
},
{
"epoch": 1.0919381557150745,
"grad_norm": 0.9758741855621338,
"learning_rate": 4.649129585428184e-06,
"loss": 0.6461,
"step": 1319
},
{
"epoch": 1.0927664273881834,
"grad_norm": 0.967698335647583,
"learning_rate": 4.64856756783544e-06,
"loss": 0.6461,
"step": 1320
},
{
"epoch": 1.0935946990612921,
"grad_norm": 0.9372200965881348,
"learning_rate": 4.6480051345149e-06,
"loss": 0.6348,
"step": 1321
},
{
"epoch": 1.0944229707344009,
"grad_norm": 0.8976414799690247,
"learning_rate": 4.64744228557539e-06,
"loss": 0.6297,
"step": 1322
},
{
"epoch": 1.0952512424075096,
"grad_norm": 0.9163550734519958,
"learning_rate": 4.646879021125816e-06,
"loss": 0.6746,
"step": 1323
},
{
"epoch": 1.0960795140806185,
"grad_norm": 0.9064576625823975,
"learning_rate": 4.646315341275164e-06,
"loss": 0.6242,
"step": 1324
},
{
"epoch": 1.0969077857537273,
"grad_norm": 0.9352027773857117,
"learning_rate": 4.645751246132501e-06,
"loss": 0.6568,
"step": 1325
},
{
"epoch": 1.097736057426836,
"grad_norm": 0.8680023550987244,
"learning_rate": 4.645186735806976e-06,
"loss": 0.6484,
"step": 1326
},
{
"epoch": 1.0985643290999447,
"grad_norm": 0.8782095909118652,
"learning_rate": 4.644621810407813e-06,
"loss": 0.646,
"step": 1327
},
{
"epoch": 1.0993926007730535,
"grad_norm": 0.8931819796562195,
"learning_rate": 4.644056470044323e-06,
"loss": 0.6708,
"step": 1328
},
{
"epoch": 1.1002208724461624,
"grad_norm": 0.9228624105453491,
"learning_rate": 4.6434907148258915e-06,
"loss": 0.6255,
"step": 1329
},
{
"epoch": 1.1010491441192711,
"grad_norm": 0.8960729837417603,
"learning_rate": 4.642924544861988e-06,
"loss": 0.6523,
"step": 1330
},
{
"epoch": 1.1018774157923799,
"grad_norm": 0.9092081785202026,
"learning_rate": 4.642357960262162e-06,
"loss": 0.6534,
"step": 1331
},
{
"epoch": 1.1027056874654886,
"grad_norm": 0.9024838209152222,
"learning_rate": 4.641790961136041e-06,
"loss": 0.6238,
"step": 1332
},
{
"epoch": 1.1035339591385975,
"grad_norm": 0.9016388058662415,
"learning_rate": 4.641223547593334e-06,
"loss": 0.6643,
"step": 1333
},
{
"epoch": 1.1043622308117063,
"grad_norm": 0.8877143859863281,
"learning_rate": 4.6406557197438324e-06,
"loss": 0.6222,
"step": 1334
},
{
"epoch": 1.105190502484815,
"grad_norm": 0.9066653251647949,
"learning_rate": 4.640087477697403e-06,
"loss": 0.6205,
"step": 1335
},
{
"epoch": 1.1060187741579237,
"grad_norm": 0.903465747833252,
"learning_rate": 4.639518821563997e-06,
"loss": 0.6299,
"step": 1336
},
{
"epoch": 1.1068470458310327,
"grad_norm": 0.8908408880233765,
"learning_rate": 4.6389497514536415e-06,
"loss": 0.6247,
"step": 1337
},
{
"epoch": 1.1076753175041414,
"grad_norm": 0.8957250714302063,
"learning_rate": 4.63838026747645e-06,
"loss": 0.632,
"step": 1338
},
{
"epoch": 1.1085035891772501,
"grad_norm": 0.938494086265564,
"learning_rate": 4.637810369742609e-06,
"loss": 0.6309,
"step": 1339
},
{
"epoch": 1.1093318608503588,
"grad_norm": 0.9913409948348999,
"learning_rate": 4.637240058362391e-06,
"loss": 0.6571,
"step": 1340
},
{
"epoch": 1.1101601325234678,
"grad_norm": 0.9037449359893799,
"learning_rate": 4.636669333446145e-06,
"loss": 0.6467,
"step": 1341
},
{
"epoch": 1.1109884041965765,
"grad_norm": 0.911696195602417,
"learning_rate": 4.636098195104299e-06,
"loss": 0.6361,
"step": 1342
},
{
"epoch": 1.1118166758696852,
"grad_norm": 0.9105570316314697,
"learning_rate": 4.6355266434473665e-06,
"loss": 0.6418,
"step": 1343
},
{
"epoch": 1.112644947542794,
"grad_norm": 0.9101701378822327,
"learning_rate": 4.634954678585935e-06,
"loss": 0.6454,
"step": 1344
},
{
"epoch": 1.1134732192159027,
"grad_norm": 0.8980433344841003,
"learning_rate": 4.634382300630675e-06,
"loss": 0.6159,
"step": 1345
},
{
"epoch": 1.1143014908890116,
"grad_norm": 0.9253267049789429,
"learning_rate": 4.633809509692336e-06,
"loss": 0.6465,
"step": 1346
},
{
"epoch": 1.1151297625621204,
"grad_norm": 0.9150062203407288,
"learning_rate": 4.6332363058817484e-06,
"loss": 0.639,
"step": 1347
},
{
"epoch": 1.115958034235229,
"grad_norm": 0.9257630705833435,
"learning_rate": 4.632662689309821e-06,
"loss": 0.6571,
"step": 1348
},
{
"epoch": 1.116786305908338,
"grad_norm": 0.9183084964752197,
"learning_rate": 4.632088660087545e-06,
"loss": 0.6291,
"step": 1349
},
{
"epoch": 1.1176145775814468,
"grad_norm": 0.9229403138160706,
"learning_rate": 4.631514218325987e-06,
"loss": 0.6487,
"step": 1350
},
{
"epoch": 1.1184428492545555,
"grad_norm": 0.9379578828811646,
"learning_rate": 4.630939364136298e-06,
"loss": 0.6564,
"step": 1351
},
{
"epoch": 1.1192711209276642,
"grad_norm": 0.9094021320343018,
"learning_rate": 4.630364097629706e-06,
"loss": 0.6354,
"step": 1352
},
{
"epoch": 1.120099392600773,
"grad_norm": 0.9528741836547852,
"learning_rate": 4.62978841891752e-06,
"loss": 0.6259,
"step": 1353
},
{
"epoch": 1.120927664273882,
"grad_norm": 0.8859900832176208,
"learning_rate": 4.629212328111129e-06,
"loss": 0.6783,
"step": 1354
},
{
"epoch": 1.1217559359469906,
"grad_norm": 0.8739688396453857,
"learning_rate": 4.628635825322e-06,
"loss": 0.6518,
"step": 1355
},
{
"epoch": 1.1225842076200994,
"grad_norm": 0.8773418068885803,
"learning_rate": 4.628058910661683e-06,
"loss": 0.6477,
"step": 1356
},
{
"epoch": 1.123412479293208,
"grad_norm": 0.9025301933288574,
"learning_rate": 4.627481584241803e-06,
"loss": 0.6471,
"step": 1357
},
{
"epoch": 1.124240750966317,
"grad_norm": 0.89899080991745,
"learning_rate": 4.62690384617407e-06,
"loss": 0.6196,
"step": 1358
},
{
"epoch": 1.1250690226394258,
"grad_norm": 0.9514591693878174,
"learning_rate": 4.6263256965702684e-06,
"loss": 0.6495,
"step": 1359
},
{
"epoch": 1.1258972943125345,
"grad_norm": 0.908789336681366,
"learning_rate": 4.6257471355422654e-06,
"loss": 0.6606,
"step": 1360
},
{
"epoch": 1.1267255659856432,
"grad_norm": 0.9262552857398987,
"learning_rate": 4.625168163202009e-06,
"loss": 0.6308,
"step": 1361
},
{
"epoch": 1.1275538376587522,
"grad_norm": 0.9395532608032227,
"learning_rate": 4.624588779661523e-06,
"loss": 0.6196,
"step": 1362
},
{
"epoch": 1.128382109331861,
"grad_norm": 0.9126531481742859,
"learning_rate": 4.624008985032913e-06,
"loss": 0.6474,
"step": 1363
},
{
"epoch": 1.1292103810049696,
"grad_norm": 0.9105198383331299,
"learning_rate": 4.623428779428365e-06,
"loss": 0.6455,
"step": 1364
},
{
"epoch": 1.1300386526780783,
"grad_norm": 0.9048190712928772,
"learning_rate": 4.6228481629601415e-06,
"loss": 0.6647,
"step": 1365
},
{
"epoch": 1.1308669243511873,
"grad_norm": 0.9415149092674255,
"learning_rate": 4.622267135740588e-06,
"loss": 0.6403,
"step": 1366
},
{
"epoch": 1.131695196024296,
"grad_norm": 0.954728901386261,
"learning_rate": 4.6216856978821275e-06,
"loss": 0.6351,
"step": 1367
},
{
"epoch": 1.1325234676974048,
"grad_norm": 0.9392480850219727,
"learning_rate": 4.621103849497261e-06,
"loss": 0.6481,
"step": 1368
},
{
"epoch": 1.1333517393705135,
"grad_norm": 0.9199283719062805,
"learning_rate": 4.620521590698574e-06,
"loss": 0.6578,
"step": 1369
},
{
"epoch": 1.1341800110436222,
"grad_norm": 0.9175401926040649,
"learning_rate": 4.6199389215987266e-06,
"loss": 0.6508,
"step": 1370
},
{
"epoch": 1.1350082827167312,
"grad_norm": 0.9140470027923584,
"learning_rate": 4.619355842310459e-06,
"loss": 0.6536,
"step": 1371
},
{
"epoch": 1.1358365543898399,
"grad_norm": 0.8863652348518372,
"learning_rate": 4.618772352946593e-06,
"loss": 0.6376,
"step": 1372
},
{
"epoch": 1.1366648260629486,
"grad_norm": 0.9074656963348389,
"learning_rate": 4.618188453620026e-06,
"loss": 0.6253,
"step": 1373
},
{
"epoch": 1.1374930977360573,
"grad_norm": 0.891055703163147,
"learning_rate": 4.61760414444374e-06,
"loss": 0.6507,
"step": 1374
},
{
"epoch": 1.1383213694091663,
"grad_norm": 0.9153634905815125,
"learning_rate": 4.617019425530791e-06,
"loss": 0.6699,
"step": 1375
},
{
"epoch": 1.139149641082275,
"grad_norm": 0.9172940850257874,
"learning_rate": 4.6164342969943196e-06,
"loss": 0.6385,
"step": 1376
},
{
"epoch": 1.1399779127553837,
"grad_norm": 0.8978968262672424,
"learning_rate": 4.615848758947539e-06,
"loss": 0.6394,
"step": 1377
},
{
"epoch": 1.1408061844284925,
"grad_norm": 0.9372277855873108,
"learning_rate": 4.615262811503749e-06,
"loss": 0.6418,
"step": 1378
},
{
"epoch": 1.1416344561016014,
"grad_norm": 0.9596112370491028,
"learning_rate": 4.614676454776321e-06,
"loss": 0.6573,
"step": 1379
},
{
"epoch": 1.1424627277747101,
"grad_norm": 0.9243675470352173,
"learning_rate": 4.614089688878713e-06,
"loss": 0.6546,
"step": 1380
},
{
"epoch": 1.1432909994478189,
"grad_norm": 0.8839207291603088,
"learning_rate": 4.613502513924459e-06,
"loss": 0.654,
"step": 1381
},
{
"epoch": 1.1441192711209276,
"grad_norm": 0.8940262198448181,
"learning_rate": 4.612914930027168e-06,
"loss": 0.6223,
"step": 1382
},
{
"epoch": 1.1449475427940365,
"grad_norm": 0.9091162085533142,
"learning_rate": 4.612326937300535e-06,
"loss": 0.6512,
"step": 1383
},
{
"epoch": 1.1457758144671453,
"grad_norm": 0.9332976937294006,
"learning_rate": 4.6117385358583315e-06,
"loss": 0.6373,
"step": 1384
},
{
"epoch": 1.146604086140254,
"grad_norm": 0.9103875756263733,
"learning_rate": 4.611149725814406e-06,
"loss": 0.6502,
"step": 1385
},
{
"epoch": 1.1474323578133627,
"grad_norm": 0.9004862308502197,
"learning_rate": 4.610560507282688e-06,
"loss": 0.6361,
"step": 1386
},
{
"epoch": 1.1482606294864715,
"grad_norm": 0.880293071269989,
"learning_rate": 4.609970880377187e-06,
"loss": 0.655,
"step": 1387
},
{
"epoch": 1.1490889011595804,
"grad_norm": 0.9099947214126587,
"learning_rate": 4.60938084521199e-06,
"loss": 0.6517,
"step": 1388
},
{
"epoch": 1.1499171728326891,
"grad_norm": 0.8913806080818176,
"learning_rate": 4.608790401901262e-06,
"loss": 0.6429,
"step": 1389
},
{
"epoch": 1.1507454445057979,
"grad_norm": 0.9198290109634399,
"learning_rate": 4.608199550559249e-06,
"loss": 0.6537,
"step": 1390
},
{
"epoch": 1.1515737161789068,
"grad_norm": 0.9192278981208801,
"learning_rate": 4.607608291300275e-06,
"loss": 0.6531,
"step": 1391
},
{
"epoch": 1.1524019878520155,
"grad_norm": 0.8977665305137634,
"learning_rate": 4.607016624238744e-06,
"loss": 0.6358,
"step": 1392
},
{
"epoch": 1.1532302595251243,
"grad_norm": 0.9576854705810547,
"learning_rate": 4.606424549489138e-06,
"loss": 0.6493,
"step": 1393
},
{
"epoch": 1.154058531198233,
"grad_norm": 0.905611515045166,
"learning_rate": 4.605832067166017e-06,
"loss": 0.63,
"step": 1394
},
{
"epoch": 1.1548868028713417,
"grad_norm": 0.8840015530586243,
"learning_rate": 4.605239177384021e-06,
"loss": 0.6504,
"step": 1395
},
{
"epoch": 1.1557150745444507,
"grad_norm": 0.9247309565544128,
"learning_rate": 4.604645880257869e-06,
"loss": 0.6477,
"step": 1396
},
{
"epoch": 1.1565433462175594,
"grad_norm": 0.9473649263381958,
"learning_rate": 4.604052175902359e-06,
"loss": 0.6374,
"step": 1397
},
{
"epoch": 1.1573716178906681,
"grad_norm": 0.9354954361915588,
"learning_rate": 4.603458064432367e-06,
"loss": 0.6491,
"step": 1398
},
{
"epoch": 1.1581998895637768,
"grad_norm": 0.8788221478462219,
"learning_rate": 4.6028635459628476e-06,
"loss": 0.6187,
"step": 1399
},
{
"epoch": 1.1590281612368858,
"grad_norm": 0.8836075067520142,
"learning_rate": 4.602268620608834e-06,
"loss": 0.665,
"step": 1400
},
{
"epoch": 1.1598564329099945,
"grad_norm": 0.9195554256439209,
"learning_rate": 4.60167328848544e-06,
"loss": 0.6291,
"step": 1401
},
{
"epoch": 1.1606847045831032,
"grad_norm": 0.8984520435333252,
"learning_rate": 4.601077549707856e-06,
"loss": 0.6361,
"step": 1402
},
{
"epoch": 1.161512976256212,
"grad_norm": 0.8667109608650208,
"learning_rate": 4.600481404391352e-06,
"loss": 0.6345,
"step": 1403
},
{
"epoch": 1.1623412479293207,
"grad_norm": 0.9312148094177246,
"learning_rate": 4.599884852651277e-06,
"loss": 0.6388,
"step": 1404
},
{
"epoch": 1.1631695196024296,
"grad_norm": 0.9096866846084595,
"learning_rate": 4.5992878946030575e-06,
"loss": 0.6383,
"step": 1405
},
{
"epoch": 1.1639977912755384,
"grad_norm": 0.9010055661201477,
"learning_rate": 4.5986905303622e-06,
"loss": 0.6524,
"step": 1406
},
{
"epoch": 1.164826062948647,
"grad_norm": 0.9414717555046082,
"learning_rate": 4.598092760044288e-06,
"loss": 0.6825,
"step": 1407
},
{
"epoch": 1.165654334621756,
"grad_norm": 0.9162404537200928,
"learning_rate": 4.597494583764987e-06,
"loss": 0.6467,
"step": 1408
},
{
"epoch": 1.1664826062948648,
"grad_norm": 0.8935617208480835,
"learning_rate": 4.596896001640034e-06,
"loss": 0.6375,
"step": 1409
},
{
"epoch": 1.1673108779679735,
"grad_norm": 0.8976510167121887,
"learning_rate": 4.596297013785253e-06,
"loss": 0.6135,
"step": 1410
},
{
"epoch": 1.1681391496410822,
"grad_norm": 0.9503909349441528,
"learning_rate": 4.5956976203165414e-06,
"loss": 0.6305,
"step": 1411
},
{
"epoch": 1.168967421314191,
"grad_norm": 0.9147273302078247,
"learning_rate": 4.595097821349876e-06,
"loss": 0.6548,
"step": 1412
},
{
"epoch": 1.1697956929873,
"grad_norm": 0.9171034693717957,
"learning_rate": 4.594497617001312e-06,
"loss": 0.6253,
"step": 1413
},
{
"epoch": 1.1706239646604086,
"grad_norm": 0.8970120549201965,
"learning_rate": 4.593897007386985e-06,
"loss": 0.6414,
"step": 1414
},
{
"epoch": 1.1714522363335174,
"grad_norm": 0.9032952189445496,
"learning_rate": 4.593295992623105e-06,
"loss": 0.6595,
"step": 1415
},
{
"epoch": 1.172280508006626,
"grad_norm": 0.9066448211669922,
"learning_rate": 4.592694572825964e-06,
"loss": 0.6654,
"step": 1416
},
{
"epoch": 1.173108779679735,
"grad_norm": 0.8747309446334839,
"learning_rate": 4.592092748111931e-06,
"loss": 0.6619,
"step": 1417
},
{
"epoch": 1.1739370513528438,
"grad_norm": 0.8836600184440613,
"learning_rate": 4.5914905185974535e-06,
"loss": 0.6443,
"step": 1418
},
{
"epoch": 1.1747653230259525,
"grad_norm": 0.9340765476226807,
"learning_rate": 4.590887884399058e-06,
"loss": 0.6633,
"step": 1419
},
{
"epoch": 1.1755935946990612,
"grad_norm": 0.8889044523239136,
"learning_rate": 4.590284845633347e-06,
"loss": 0.6489,
"step": 1420
},
{
"epoch": 1.1764218663721702,
"grad_norm": 0.9043524265289307,
"learning_rate": 4.589681402417005e-06,
"loss": 0.6392,
"step": 1421
},
{
"epoch": 1.177250138045279,
"grad_norm": 0.9210046529769897,
"learning_rate": 4.5890775548667895e-06,
"loss": 0.6105,
"step": 1422
},
{
"epoch": 1.1780784097183876,
"grad_norm": 0.8954651951789856,
"learning_rate": 4.588473303099542e-06,
"loss": 0.6425,
"step": 1423
},
{
"epoch": 1.1789066813914963,
"grad_norm": 0.945517897605896,
"learning_rate": 4.587868647232179e-06,
"loss": 0.6404,
"step": 1424
},
{
"epoch": 1.1797349530646053,
"grad_norm": 0.9059463739395142,
"learning_rate": 4.587263587381696e-06,
"loss": 0.6521,
"step": 1425
},
{
"epoch": 1.180563224737714,
"grad_norm": 0.8876296281814575,
"learning_rate": 4.586658123665165e-06,
"loss": 0.6447,
"step": 1426
},
{
"epoch": 1.1813914964108228,
"grad_norm": 0.8862487077713013,
"learning_rate": 4.58605225619974e-06,
"loss": 0.6445,
"step": 1427
},
{
"epoch": 1.1822197680839315,
"grad_norm": 0.9066351056098938,
"learning_rate": 4.585445985102649e-06,
"loss": 0.6563,
"step": 1428
},
{
"epoch": 1.1830480397570402,
"grad_norm": 0.913724958896637,
"learning_rate": 4.584839310491199e-06,
"loss": 0.6507,
"step": 1429
},
{
"epoch": 1.1838763114301492,
"grad_norm": 0.8942010402679443,
"learning_rate": 4.584232232482778e-06,
"loss": 0.6424,
"step": 1430
},
{
"epoch": 1.1847045831032579,
"grad_norm": 0.8986000418663025,
"learning_rate": 4.58362475119485e-06,
"loss": 0.6491,
"step": 1431
},
{
"epoch": 1.1855328547763666,
"grad_norm": 0.8879220485687256,
"learning_rate": 4.5830168667449544e-06,
"loss": 0.6412,
"step": 1432
},
{
"epoch": 1.1863611264494753,
"grad_norm": 0.9497302770614624,
"learning_rate": 4.5824085792507135e-06,
"loss": 0.6453,
"step": 1433
},
{
"epoch": 1.1871893981225843,
"grad_norm": 0.8933898210525513,
"learning_rate": 4.581799888829824e-06,
"loss": 0.6381,
"step": 1434
},
{
"epoch": 1.188017669795693,
"grad_norm": 0.9034610986709595,
"learning_rate": 4.581190795600062e-06,
"loss": 0.6551,
"step": 1435
},
{
"epoch": 1.1888459414688017,
"grad_norm": 0.8968634605407715,
"learning_rate": 4.580581299679282e-06,
"loss": 0.637,
"step": 1436
},
{
"epoch": 1.1896742131419105,
"grad_norm": 0.8974384069442749,
"learning_rate": 4.579971401185416e-06,
"loss": 0.6301,
"step": 1437
},
{
"epoch": 1.1905024848150194,
"grad_norm": 0.8726125359535217,
"learning_rate": 4.579361100236473e-06,
"loss": 0.6487,
"step": 1438
},
{
"epoch": 1.1913307564881281,
"grad_norm": 0.9131139516830444,
"learning_rate": 4.5787503969505405e-06,
"loss": 0.6211,
"step": 1439
},
{
"epoch": 1.1921590281612369,
"grad_norm": 0.8924953937530518,
"learning_rate": 4.578139291445783e-06,
"loss": 0.6339,
"step": 1440
},
{
"epoch": 1.1929872998343456,
"grad_norm": 0.890883207321167,
"learning_rate": 4.577527783840447e-06,
"loss": 0.66,
"step": 1441
},
{
"epoch": 1.1938155715074545,
"grad_norm": 0.9326318502426147,
"learning_rate": 4.5769158742528494e-06,
"loss": 0.6611,
"step": 1442
},
{
"epoch": 1.1946438431805633,
"grad_norm": 0.9354713559150696,
"learning_rate": 4.576303562801392e-06,
"loss": 0.6445,
"step": 1443
},
{
"epoch": 1.195472114853672,
"grad_norm": 0.8868383169174194,
"learning_rate": 4.57569084960455e-06,
"loss": 0.6344,
"step": 1444
},
{
"epoch": 1.1963003865267807,
"grad_norm": 0.893924355506897,
"learning_rate": 4.5750777347808775e-06,
"loss": 0.602,
"step": 1445
},
{
"epoch": 1.1971286581998895,
"grad_norm": 0.9299339652061462,
"learning_rate": 4.574464218449006e-06,
"loss": 0.6206,
"step": 1446
},
{
"epoch": 1.1979569298729984,
"grad_norm": 0.9058142304420471,
"learning_rate": 4.573850300727648e-06,
"loss": 0.6073,
"step": 1447
},
{
"epoch": 1.1987852015461071,
"grad_norm": 0.8893733620643616,
"learning_rate": 4.5732359817355885e-06,
"loss": 0.6336,
"step": 1448
},
{
"epoch": 1.1996134732192159,
"grad_norm": 0.9018000960350037,
"learning_rate": 4.572621261591692e-06,
"loss": 0.6268,
"step": 1449
},
{
"epoch": 1.2004417448923248,
"grad_norm": 0.9087574481964111,
"learning_rate": 4.572006140414903e-06,
"loss": 0.6511,
"step": 1450
},
{
"epoch": 1.2012700165654335,
"grad_norm": 0.9092984795570374,
"learning_rate": 4.5713906183242404e-06,
"loss": 0.6424,
"step": 1451
},
{
"epoch": 1.2020982882385423,
"grad_norm": 0.9040535092353821,
"learning_rate": 4.5707746954388034e-06,
"loss": 0.6242,
"step": 1452
},
{
"epoch": 1.202926559911651,
"grad_norm": 0.891659677028656,
"learning_rate": 4.570158371877766e-06,
"loss": 0.643,
"step": 1453
},
{
"epoch": 1.2037548315847597,
"grad_norm": 0.9254544973373413,
"learning_rate": 4.569541647760381e-06,
"loss": 0.6191,
"step": 1454
},
{
"epoch": 1.2045831032578687,
"grad_norm": 0.8978536128997803,
"learning_rate": 4.568924523205979e-06,
"loss": 0.6632,
"step": 1455
},
{
"epoch": 1.2054113749309774,
"grad_norm": 0.9179201126098633,
"learning_rate": 4.568306998333968e-06,
"loss": 0.6485,
"step": 1456
},
{
"epoch": 1.2062396466040861,
"grad_norm": 0.9061882495880127,
"learning_rate": 4.567689073263834e-06,
"loss": 0.6704,
"step": 1457
},
{
"epoch": 1.2070679182771948,
"grad_norm": 0.9143427610397339,
"learning_rate": 4.567070748115139e-06,
"loss": 0.6453,
"step": 1458
},
{
"epoch": 1.2078961899503038,
"grad_norm": 0.8740119338035583,
"learning_rate": 4.566452023007523e-06,
"loss": 0.6453,
"step": 1459
},
{
"epoch": 1.2087244616234125,
"grad_norm": 0.8977463841438293,
"learning_rate": 4.565832898060703e-06,
"loss": 0.6451,
"step": 1460
},
{
"epoch": 1.2095527332965212,
"grad_norm": 0.9416611194610596,
"learning_rate": 4.565213373394476e-06,
"loss": 0.6265,
"step": 1461
},
{
"epoch": 1.21038100496963,
"grad_norm": 0.9132000207901001,
"learning_rate": 4.5645934491287115e-06,
"loss": 0.6412,
"step": 1462
},
{
"epoch": 1.2112092766427387,
"grad_norm": 0.9480498433113098,
"learning_rate": 4.563973125383362e-06,
"loss": 0.6363,
"step": 1463
},
{
"epoch": 1.2120375483158476,
"grad_norm": 0.9170709848403931,
"learning_rate": 4.563352402278451e-06,
"loss": 0.6201,
"step": 1464
},
{
"epoch": 1.2128658199889564,
"grad_norm": 0.9655918478965759,
"learning_rate": 4.562731279934086e-06,
"loss": 0.6497,
"step": 1465
},
{
"epoch": 1.213694091662065,
"grad_norm": 0.8885562419891357,
"learning_rate": 4.562109758470446e-06,
"loss": 0.6265,
"step": 1466
},
{
"epoch": 1.214522363335174,
"grad_norm": 0.9072791337966919,
"learning_rate": 4.561487838007792e-06,
"loss": 0.6257,
"step": 1467
},
{
"epoch": 1.2153506350082828,
"grad_norm": 0.8992362022399902,
"learning_rate": 4.560865518666458e-06,
"loss": 0.6425,
"step": 1468
},
{
"epoch": 1.2161789066813915,
"grad_norm": 0.8934991955757141,
"learning_rate": 4.560242800566855e-06,
"loss": 0.6583,
"step": 1469
},
{
"epoch": 1.2170071783545002,
"grad_norm": 0.9292325973510742,
"learning_rate": 4.559619683829477e-06,
"loss": 0.6403,
"step": 1470
},
{
"epoch": 1.217835450027609,
"grad_norm": 0.8756814002990723,
"learning_rate": 4.55899616857489e-06,
"loss": 0.6261,
"step": 1471
},
{
"epoch": 1.218663721700718,
"grad_norm": 0.9050565361976624,
"learning_rate": 4.558372254923738e-06,
"loss": 0.6325,
"step": 1472
},
{
"epoch": 1.2194919933738266,
"grad_norm": 0.9230931401252747,
"learning_rate": 4.557747942996742e-06,
"loss": 0.6275,
"step": 1473
},
{
"epoch": 1.2203202650469354,
"grad_norm": 0.894026517868042,
"learning_rate": 4.557123232914702e-06,
"loss": 0.6348,
"step": 1474
},
{
"epoch": 1.221148536720044,
"grad_norm": 0.8882637023925781,
"learning_rate": 4.556498124798492e-06,
"loss": 0.6123,
"step": 1475
},
{
"epoch": 1.221976808393153,
"grad_norm": 0.9191063642501831,
"learning_rate": 4.555872618769065e-06,
"loss": 0.6531,
"step": 1476
},
{
"epoch": 1.2228050800662618,
"grad_norm": 0.9398623704910278,
"learning_rate": 4.555246714947452e-06,
"loss": 0.6553,
"step": 1477
},
{
"epoch": 1.2236333517393705,
"grad_norm": 0.9399452805519104,
"learning_rate": 4.554620413454757e-06,
"loss": 0.6485,
"step": 1478
},
{
"epoch": 1.2244616234124792,
"grad_norm": 0.9001572728157043,
"learning_rate": 4.553993714412166e-06,
"loss": 0.6467,
"step": 1479
},
{
"epoch": 1.2252898950855882,
"grad_norm": 0.9266572594642639,
"learning_rate": 4.553366617940938e-06,
"loss": 0.6325,
"step": 1480
},
{
"epoch": 1.226118166758697,
"grad_norm": 0.928301215171814,
"learning_rate": 4.552739124162411e-06,
"loss": 0.6398,
"step": 1481
},
{
"epoch": 1.2269464384318056,
"grad_norm": 0.880954921245575,
"learning_rate": 4.552111233197999e-06,
"loss": 0.669,
"step": 1482
},
{
"epoch": 1.2277747101049143,
"grad_norm": 0.9089685082435608,
"learning_rate": 4.551482945169191e-06,
"loss": 0.62,
"step": 1483
},
{
"epoch": 1.2286029817780233,
"grad_norm": 0.9187189936637878,
"learning_rate": 4.550854260197559e-06,
"loss": 0.6381,
"step": 1484
},
{
"epoch": 1.229431253451132,
"grad_norm": 0.8782045245170593,
"learning_rate": 4.550225178404744e-06,
"loss": 0.6397,
"step": 1485
},
{
"epoch": 1.2302595251242407,
"grad_norm": 0.8798948526382446,
"learning_rate": 4.54959569991247e-06,
"loss": 0.6522,
"step": 1486
},
{
"epoch": 1.2310877967973495,
"grad_norm": 0.9257029294967651,
"learning_rate": 4.548965824842534e-06,
"loss": 0.6522,
"step": 1487
},
{
"epoch": 1.2319160684704582,
"grad_norm": 0.8923602104187012,
"learning_rate": 4.548335553316811e-06,
"loss": 0.6538,
"step": 1488
},
{
"epoch": 1.2327443401435672,
"grad_norm": 0.9071294069290161,
"learning_rate": 4.5477048854572524e-06,
"loss": 0.6305,
"step": 1489
},
{
"epoch": 1.2335726118166759,
"grad_norm": 0.9495804905891418,
"learning_rate": 4.547073821385888e-06,
"loss": 0.6398,
"step": 1490
},
{
"epoch": 1.2344008834897846,
"grad_norm": 0.9163299202919006,
"learning_rate": 4.54644236122482e-06,
"loss": 0.6454,
"step": 1491
},
{
"epoch": 1.2352291551628933,
"grad_norm": 0.9286475777626038,
"learning_rate": 4.545810505096233e-06,
"loss": 0.6345,
"step": 1492
},
{
"epoch": 1.2360574268360023,
"grad_norm": 0.9116879105567932,
"learning_rate": 4.545178253122384e-06,
"loss": 0.6451,
"step": 1493
},
{
"epoch": 1.236885698509111,
"grad_norm": 0.917826771736145,
"learning_rate": 4.544545605425607e-06,
"loss": 0.6374,
"step": 1494
},
{
"epoch": 1.2377139701822197,
"grad_norm": 0.9433108568191528,
"learning_rate": 4.543912562128316e-06,
"loss": 0.6321,
"step": 1495
},
{
"epoch": 1.2385422418553285,
"grad_norm": 0.9017132520675659,
"learning_rate": 4.543279123352996e-06,
"loss": 0.6341,
"step": 1496
},
{
"epoch": 1.2393705135284374,
"grad_norm": 0.9557158350944519,
"learning_rate": 4.542645289222214e-06,
"loss": 0.6364,
"step": 1497
},
{
"epoch": 1.2401987852015461,
"grad_norm": 0.9258891344070435,
"learning_rate": 4.542011059858609e-06,
"loss": 0.6598,
"step": 1498
},
{
"epoch": 1.2410270568746549,
"grad_norm": 0.8988100290298462,
"learning_rate": 4.541376435384899e-06,
"loss": 0.6496,
"step": 1499
},
{
"epoch": 1.2418553285477636,
"grad_norm": 0.898154079914093,
"learning_rate": 4.540741415923878e-06,
"loss": 0.6486,
"step": 1500
},
{
"epoch": 1.2426836002208725,
"grad_norm": 0.9216037392616272,
"learning_rate": 4.540106001598418e-06,
"loss": 0.6436,
"step": 1501
},
{
"epoch": 1.2435118718939813,
"grad_norm": 0.9086095690727234,
"learning_rate": 4.539470192531463e-06,
"loss": 0.6287,
"step": 1502
},
{
"epoch": 1.24434014356709,
"grad_norm": 0.921291172504425,
"learning_rate": 4.538833988846039e-06,
"loss": 0.629,
"step": 1503
},
{
"epoch": 1.2451684152401987,
"grad_norm": 0.8621019721031189,
"learning_rate": 4.538197390665242e-06,
"loss": 0.6543,
"step": 1504
},
{
"epoch": 1.2459966869133074,
"grad_norm": 0.9523105025291443,
"learning_rate": 4.537560398112251e-06,
"loss": 0.6482,
"step": 1505
},
{
"epoch": 1.2468249585864164,
"grad_norm": 0.971931517124176,
"learning_rate": 4.536923011310317e-06,
"loss": 0.6463,
"step": 1506
},
{
"epoch": 1.2476532302595251,
"grad_norm": 0.927219033241272,
"learning_rate": 4.536285230382767e-06,
"loss": 0.6443,
"step": 1507
},
{
"epoch": 1.2484815019326339,
"grad_norm": 0.9433056116104126,
"learning_rate": 4.535647055453007e-06,
"loss": 0.6301,
"step": 1508
},
{
"epoch": 1.2493097736057428,
"grad_norm": 0.8992047905921936,
"learning_rate": 4.5350084866445195e-06,
"loss": 0.6177,
"step": 1509
},
{
"epoch": 1.2501380452788515,
"grad_norm": 0.8946895003318787,
"learning_rate": 4.534369524080858e-06,
"loss": 0.6275,
"step": 1510
},
{
"epoch": 1.2509663169519603,
"grad_norm": 0.8732156157493591,
"learning_rate": 4.5337301678856595e-06,
"loss": 0.6572,
"step": 1511
},
{
"epoch": 1.251794588625069,
"grad_norm": 0.9505333304405212,
"learning_rate": 4.533090418182631e-06,
"loss": 0.6352,
"step": 1512
},
{
"epoch": 1.2526228602981777,
"grad_norm": 0.8931122422218323,
"learning_rate": 4.532450275095558e-06,
"loss": 0.649,
"step": 1513
},
{
"epoch": 1.2534511319712867,
"grad_norm": 0.8781880140304565,
"learning_rate": 4.531809738748304e-06,
"loss": 0.651,
"step": 1514
},
{
"epoch": 1.2542794036443954,
"grad_norm": 0.9201871156692505,
"learning_rate": 4.531168809264805e-06,
"loss": 0.6451,
"step": 1515
},
{
"epoch": 1.2551076753175041,
"grad_norm": 0.9318148493766785,
"learning_rate": 4.530527486769077e-06,
"loss": 0.6458,
"step": 1516
},
{
"epoch": 1.2559359469906128,
"grad_norm": 0.9131407141685486,
"learning_rate": 4.529885771385208e-06,
"loss": 0.6679,
"step": 1517
},
{
"epoch": 1.2567642186637218,
"grad_norm": 0.8942316770553589,
"learning_rate": 4.529243663237365e-06,
"loss": 0.6262,
"step": 1518
},
{
"epoch": 1.2575924903368305,
"grad_norm": 0.9058274030685425,
"learning_rate": 4.528601162449788e-06,
"loss": 0.635,
"step": 1519
},
{
"epoch": 1.2584207620099392,
"grad_norm": 0.9129453897476196,
"learning_rate": 4.527958269146798e-06,
"loss": 0.6375,
"step": 1520
},
{
"epoch": 1.259249033683048,
"grad_norm": 0.8737504482269287,
"learning_rate": 4.527314983452787e-06,
"loss": 0.6437,
"step": 1521
},
{
"epoch": 1.2600773053561567,
"grad_norm": 0.9680130481719971,
"learning_rate": 4.526671305492225e-06,
"loss": 0.6546,
"step": 1522
},
{
"epoch": 1.2609055770292656,
"grad_norm": 0.9186084270477295,
"learning_rate": 4.526027235389658e-06,
"loss": 0.6505,
"step": 1523
},
{
"epoch": 1.2617338487023744,
"grad_norm": 0.9164871573448181,
"learning_rate": 4.525382773269706e-06,
"loss": 0.6541,
"step": 1524
},
{
"epoch": 1.262562120375483,
"grad_norm": 0.8994123935699463,
"learning_rate": 4.5247379192570695e-06,
"loss": 0.6313,
"step": 1525
},
{
"epoch": 1.263390392048592,
"grad_norm": 0.8780331015586853,
"learning_rate": 4.524092673476519e-06,
"loss": 0.6496,
"step": 1526
},
{
"epoch": 1.2642186637217008,
"grad_norm": 0.8670293092727661,
"learning_rate": 4.5234470360529045e-06,
"loss": 0.6422,
"step": 1527
},
{
"epoch": 1.2650469353948095,
"grad_norm": 0.9305918216705322,
"learning_rate": 4.522801007111152e-06,
"loss": 0.6352,
"step": 1528
},
{
"epoch": 1.2658752070679182,
"grad_norm": 0.9444066286087036,
"learning_rate": 4.52215458677626e-06,
"loss": 0.6357,
"step": 1529
},
{
"epoch": 1.266703478741027,
"grad_norm": 0.8987369537353516,
"learning_rate": 4.521507775173305e-06,
"loss": 0.6407,
"step": 1530
},
{
"epoch": 1.267531750414136,
"grad_norm": 0.9200116395950317,
"learning_rate": 4.5208605724274415e-06,
"loss": 0.6366,
"step": 1531
},
{
"epoch": 1.2683600220872446,
"grad_norm": 0.8680561780929565,
"learning_rate": 4.520212978663894e-06,
"loss": 0.6389,
"step": 1532
},
{
"epoch": 1.2691882937603534,
"grad_norm": 0.8928949236869812,
"learning_rate": 4.519564994007968e-06,
"loss": 0.6552,
"step": 1533
},
{
"epoch": 1.2700165654334623,
"grad_norm": 0.8986184000968933,
"learning_rate": 4.518916618585042e-06,
"loss": 0.6554,
"step": 1534
},
{
"epoch": 1.270844837106571,
"grad_norm": 0.9239423274993896,
"learning_rate": 4.51826785252057e-06,
"loss": 0.6448,
"step": 1535
},
{
"epoch": 1.2716731087796798,
"grad_norm": 0.9534674286842346,
"learning_rate": 4.517618695940082e-06,
"loss": 0.6303,
"step": 1536
},
{
"epoch": 1.2725013804527885,
"grad_norm": 0.9341598749160767,
"learning_rate": 4.516969148969186e-06,
"loss": 0.6335,
"step": 1537
},
{
"epoch": 1.2733296521258972,
"grad_norm": 0.9264217615127563,
"learning_rate": 4.51631921173356e-06,
"loss": 0.6543,
"step": 1538
},
{
"epoch": 1.274157923799006,
"grad_norm": 0.9460370540618896,
"learning_rate": 4.515668884358963e-06,
"loss": 0.6531,
"step": 1539
},
{
"epoch": 1.274986195472115,
"grad_norm": 0.9690954089164734,
"learning_rate": 4.515018166971227e-06,
"loss": 0.6495,
"step": 1540
},
{
"epoch": 1.2758144671452236,
"grad_norm": 0.8746740818023682,
"learning_rate": 4.514367059696259e-06,
"loss": 0.6169,
"step": 1541
},
{
"epoch": 1.2766427388183323,
"grad_norm": 0.9288576245307922,
"learning_rate": 4.513715562660043e-06,
"loss": 0.6459,
"step": 1542
},
{
"epoch": 1.2774710104914413,
"grad_norm": 0.9935420751571655,
"learning_rate": 4.513063675988637e-06,
"loss": 0.6471,
"step": 1543
},
{
"epoch": 1.27829928216455,
"grad_norm": 0.9584968686103821,
"learning_rate": 4.512411399808175e-06,
"loss": 0.6473,
"step": 1544
},
{
"epoch": 1.2791275538376587,
"grad_norm": 0.9049419164657593,
"learning_rate": 4.511758734244867e-06,
"loss": 0.6667,
"step": 1545
},
{
"epoch": 1.2799558255107675,
"grad_norm": 0.9456910490989685,
"learning_rate": 4.511105679424998e-06,
"loss": 0.625,
"step": 1546
},
{
"epoch": 1.2807840971838762,
"grad_norm": 0.9267060160636902,
"learning_rate": 4.510452235474926e-06,
"loss": 0.6239,
"step": 1547
},
{
"epoch": 1.2816123688569852,
"grad_norm": 0.9363065958023071,
"learning_rate": 4.509798402521088e-06,
"loss": 0.6416,
"step": 1548
},
{
"epoch": 1.2824406405300939,
"grad_norm": 0.85664302110672,
"learning_rate": 4.509144180689995e-06,
"loss": 0.6297,
"step": 1549
},
{
"epoch": 1.2832689122032026,
"grad_norm": 0.9298888444900513,
"learning_rate": 4.508489570108231e-06,
"loss": 0.6305,
"step": 1550
},
{
"epoch": 1.2840971838763116,
"grad_norm": 0.9485051035881042,
"learning_rate": 4.507834570902459e-06,
"loss": 0.6409,
"step": 1551
},
{
"epoch": 1.2849254555494203,
"grad_norm": 0.9129849076271057,
"learning_rate": 4.507179183199414e-06,
"loss": 0.6215,
"step": 1552
},
{
"epoch": 1.285753727222529,
"grad_norm": 0.9630280137062073,
"learning_rate": 4.506523407125907e-06,
"loss": 0.6229,
"step": 1553
},
{
"epoch": 1.2865819988956377,
"grad_norm": 0.9047848582267761,
"learning_rate": 4.505867242808826e-06,
"loss": 0.6434,
"step": 1554
},
{
"epoch": 1.2874102705687465,
"grad_norm": 0.9219448566436768,
"learning_rate": 4.5052106903751315e-06,
"loss": 0.6449,
"step": 1555
},
{
"epoch": 1.2882385422418554,
"grad_norm": 0.9347684383392334,
"learning_rate": 4.504553749951861e-06,
"loss": 0.6272,
"step": 1556
},
{
"epoch": 1.2890668139149641,
"grad_norm": 0.8930262923240662,
"learning_rate": 4.503896421666126e-06,
"loss": 0.6364,
"step": 1557
},
{
"epoch": 1.2898950855880729,
"grad_norm": 0.9067763090133667,
"learning_rate": 4.503238705645114e-06,
"loss": 0.6317,
"step": 1558
},
{
"epoch": 1.2907233572611816,
"grad_norm": 0.9524715542793274,
"learning_rate": 4.5025806020160865e-06,
"loss": 0.6525,
"step": 1559
},
{
"epoch": 1.2915516289342905,
"grad_norm": 0.927255392074585,
"learning_rate": 4.5019221109063795e-06,
"loss": 0.6343,
"step": 1560
},
{
"epoch": 1.2923799006073993,
"grad_norm": 0.904295802116394,
"learning_rate": 4.5012632324434065e-06,
"loss": 0.6484,
"step": 1561
},
{
"epoch": 1.293208172280508,
"grad_norm": 0.9138978123664856,
"learning_rate": 4.500603966754653e-06,
"loss": 0.6374,
"step": 1562
},
{
"epoch": 1.2940364439536167,
"grad_norm": 0.9038779139518738,
"learning_rate": 4.499944313967681e-06,
"loss": 0.6395,
"step": 1563
},
{
"epoch": 1.2948647156267254,
"grad_norm": 0.9218693971633911,
"learning_rate": 4.499284274210128e-06,
"loss": 0.6325,
"step": 1564
},
{
"epoch": 1.2956929872998344,
"grad_norm": 0.9210799932479858,
"learning_rate": 4.4986238476097055e-06,
"loss": 0.6476,
"step": 1565
},
{
"epoch": 1.2965212589729431,
"grad_norm": 0.8977149724960327,
"learning_rate": 4.4979630342941994e-06,
"loss": 0.6489,
"step": 1566
},
{
"epoch": 1.2973495306460519,
"grad_norm": 0.9423681497573853,
"learning_rate": 4.49730183439147e-06,
"loss": 0.6391,
"step": 1567
},
{
"epoch": 1.2981778023191608,
"grad_norm": 0.9404032826423645,
"learning_rate": 4.4966402480294545e-06,
"loss": 0.6233,
"step": 1568
},
{
"epoch": 1.2990060739922695,
"grad_norm": 0.9104741215705872,
"learning_rate": 4.495978275336164e-06,
"loss": 0.6369,
"step": 1569
},
{
"epoch": 1.2998343456653783,
"grad_norm": 0.8869790434837341,
"learning_rate": 4.4953159164396835e-06,
"loss": 0.6331,
"step": 1570
},
{
"epoch": 1.300662617338487,
"grad_norm": 0.9269986152648926,
"learning_rate": 4.494653171468173e-06,
"loss": 0.6361,
"step": 1571
},
{
"epoch": 1.3014908890115957,
"grad_norm": 0.9210803508758545,
"learning_rate": 4.4939900405498675e-06,
"loss": 0.6314,
"step": 1572
},
{
"epoch": 1.3023191606847047,
"grad_norm": 0.9426820874214172,
"learning_rate": 4.4933265238130775e-06,
"loss": 0.6625,
"step": 1573
},
{
"epoch": 1.3031474323578134,
"grad_norm": 0.9198150634765625,
"learning_rate": 4.492662621386186e-06,
"loss": 0.627,
"step": 1574
},
{
"epoch": 1.3039757040309221,
"grad_norm": 0.9312534928321838,
"learning_rate": 4.4919983333976525e-06,
"loss": 0.6386,
"step": 1575
},
{
"epoch": 1.304803975704031,
"grad_norm": 0.8926329016685486,
"learning_rate": 4.491333659976011e-06,
"loss": 0.6559,
"step": 1576
},
{
"epoch": 1.3056322473771398,
"grad_norm": 0.9172425270080566,
"learning_rate": 4.4906686012498705e-06,
"loss": 0.6254,
"step": 1577
},
{
"epoch": 1.3064605190502485,
"grad_norm": 0.9896116256713867,
"learning_rate": 4.490003157347911e-06,
"loss": 0.6331,
"step": 1578
},
{
"epoch": 1.3072887907233572,
"grad_norm": 0.9267039895057678,
"learning_rate": 4.489337328398891e-06,
"loss": 0.6296,
"step": 1579
},
{
"epoch": 1.308117062396466,
"grad_norm": 0.9033195376396179,
"learning_rate": 4.488671114531643e-06,
"loss": 0.6354,
"step": 1580
},
{
"epoch": 1.3089453340695747,
"grad_norm": 0.9047731161117554,
"learning_rate": 4.488004515875072e-06,
"loss": 0.6351,
"step": 1581
},
{
"epoch": 1.3097736057426836,
"grad_norm": 0.9096142649650574,
"learning_rate": 4.4873375325581596e-06,
"loss": 0.6504,
"step": 1582
},
{
"epoch": 1.3106018774157924,
"grad_norm": 0.8848153352737427,
"learning_rate": 4.48667016470996e-06,
"loss": 0.6354,
"step": 1583
},
{
"epoch": 1.311430149088901,
"grad_norm": 0.9036188721656799,
"learning_rate": 4.486002412459603e-06,
"loss": 0.6431,
"step": 1584
},
{
"epoch": 1.31225842076201,
"grad_norm": 0.8881173133850098,
"learning_rate": 4.4853342759362924e-06,
"loss": 0.6509,
"step": 1585
},
{
"epoch": 1.3130866924351188,
"grad_norm": 0.9195591807365417,
"learning_rate": 4.484665755269307e-06,
"loss": 0.6522,
"step": 1586
},
{
"epoch": 1.3139149641082275,
"grad_norm": 0.9021509885787964,
"learning_rate": 4.483996850587998e-06,
"loss": 0.6532,
"step": 1587
},
{
"epoch": 1.3147432357813362,
"grad_norm": 0.8902403116226196,
"learning_rate": 4.483327562021794e-06,
"loss": 0.5942,
"step": 1588
},
{
"epoch": 1.315571507454445,
"grad_norm": 0.8933504819869995,
"learning_rate": 4.482657889700194e-06,
"loss": 0.6305,
"step": 1589
},
{
"epoch": 1.316399779127554,
"grad_norm": 0.9233856201171875,
"learning_rate": 4.481987833752776e-06,
"loss": 0.6275,
"step": 1590
},
{
"epoch": 1.3172280508006626,
"grad_norm": 0.9482141733169556,
"learning_rate": 4.4813173943091865e-06,
"loss": 0.6232,
"step": 1591
},
{
"epoch": 1.3180563224737714,
"grad_norm": 0.94732266664505,
"learning_rate": 4.480646571499151e-06,
"loss": 0.6297,
"step": 1592
},
{
"epoch": 1.3188845941468803,
"grad_norm": 0.9066787958145142,
"learning_rate": 4.479975365452468e-06,
"loss": 0.6334,
"step": 1593
},
{
"epoch": 1.319712865819989,
"grad_norm": 0.9063938856124878,
"learning_rate": 4.479303776299008e-06,
"loss": 0.6415,
"step": 1594
},
{
"epoch": 1.3205411374930978,
"grad_norm": 0.9286819696426392,
"learning_rate": 4.478631804168719e-06,
"loss": 0.6534,
"step": 1595
},
{
"epoch": 1.3213694091662065,
"grad_norm": 0.9161240458488464,
"learning_rate": 4.47795944919162e-06,
"loss": 0.6401,
"step": 1596
},
{
"epoch": 1.3221976808393152,
"grad_norm": 0.956538200378418,
"learning_rate": 4.4772867114978056e-06,
"loss": 0.6649,
"step": 1597
},
{
"epoch": 1.323025952512424,
"grad_norm": 0.910504162311554,
"learning_rate": 4.4766135912174455e-06,
"loss": 0.6456,
"step": 1598
},
{
"epoch": 1.323854224185533,
"grad_norm": 0.9054026007652283,
"learning_rate": 4.4759400884807805e-06,
"loss": 0.6458,
"step": 1599
},
{
"epoch": 1.3246824958586416,
"grad_norm": 0.933506429195404,
"learning_rate": 4.475266203418127e-06,
"loss": 0.626,
"step": 1600
},
{
"epoch": 1.3255107675317503,
"grad_norm": 0.9042014479637146,
"learning_rate": 4.474591936159878e-06,
"loss": 0.6376,
"step": 1601
},
{
"epoch": 1.3263390392048593,
"grad_norm": 0.9005138278007507,
"learning_rate": 4.473917286836496e-06,
"loss": 0.6093,
"step": 1602
},
{
"epoch": 1.327167310877968,
"grad_norm": 0.9096264839172363,
"learning_rate": 4.4732422555785196e-06,
"loss": 0.6344,
"step": 1603
},
{
"epoch": 1.3279955825510767,
"grad_norm": 0.9171631336212158,
"learning_rate": 4.4725668425165605e-06,
"loss": 0.6454,
"step": 1604
},
{
"epoch": 1.3288238542241855,
"grad_norm": 0.8821735382080078,
"learning_rate": 4.4718910477813055e-06,
"loss": 0.6357,
"step": 1605
},
{
"epoch": 1.3296521258972942,
"grad_norm": 0.918499231338501,
"learning_rate": 4.471214871503514e-06,
"loss": 0.6361,
"step": 1606
},
{
"epoch": 1.3304803975704032,
"grad_norm": 0.9187904596328735,
"learning_rate": 4.470538313814021e-06,
"loss": 0.6362,
"step": 1607
},
{
"epoch": 1.3313086692435119,
"grad_norm": 0.9272632598876953,
"learning_rate": 4.469861374843734e-06,
"loss": 0.6659,
"step": 1608
},
{
"epoch": 1.3321369409166206,
"grad_norm": 0.8950392603874207,
"learning_rate": 4.469184054723632e-06,
"loss": 0.6551,
"step": 1609
},
{
"epoch": 1.3329652125897296,
"grad_norm": 0.9604961276054382,
"learning_rate": 4.468506353584773e-06,
"loss": 0.643,
"step": 1610
},
{
"epoch": 1.3337934842628383,
"grad_norm": 0.933073103427887,
"learning_rate": 4.467828271558286e-06,
"loss": 0.6193,
"step": 1611
},
{
"epoch": 1.334621755935947,
"grad_norm": 0.890964686870575,
"learning_rate": 4.467149808775371e-06,
"loss": 0.6423,
"step": 1612
},
{
"epoch": 1.3354500276090557,
"grad_norm": 0.8916820883750916,
"learning_rate": 4.466470965367306e-06,
"loss": 0.6399,
"step": 1613
},
{
"epoch": 1.3362782992821645,
"grad_norm": 0.9151459336280823,
"learning_rate": 4.465791741465439e-06,
"loss": 0.6439,
"step": 1614
},
{
"epoch": 1.3371065709552734,
"grad_norm": 0.9155884385108948,
"learning_rate": 4.465112137201197e-06,
"loss": 0.6516,
"step": 1615
},
{
"epoch": 1.3379348426283821,
"grad_norm": 0.9187855124473572,
"learning_rate": 4.464432152706073e-06,
"loss": 0.6263,
"step": 1616
},
{
"epoch": 1.3387631143014909,
"grad_norm": 0.9506946206092834,
"learning_rate": 4.46375178811164e-06,
"loss": 0.6586,
"step": 1617
},
{
"epoch": 1.3395913859745996,
"grad_norm": 0.9056413769721985,
"learning_rate": 4.463071043549543e-06,
"loss": 0.6301,
"step": 1618
},
{
"epoch": 1.3404196576477085,
"grad_norm": 0.9056457877159119,
"learning_rate": 4.462389919151498e-06,
"loss": 0.6352,
"step": 1619
},
{
"epoch": 1.3412479293208173,
"grad_norm": 0.9204022288322449,
"learning_rate": 4.461708415049297e-06,
"loss": 0.6446,
"step": 1620
},
{
"epoch": 1.342076200993926,
"grad_norm": 0.9411336183547974,
"learning_rate": 4.461026531374804e-06,
"loss": 0.6594,
"step": 1621
},
{
"epoch": 1.3429044726670347,
"grad_norm": 0.8956605792045593,
"learning_rate": 4.460344268259958e-06,
"loss": 0.6492,
"step": 1622
},
{
"epoch": 1.3437327443401434,
"grad_norm": 0.9010135531425476,
"learning_rate": 4.45966162583677e-06,
"loss": 0.6233,
"step": 1623
},
{
"epoch": 1.3445610160132524,
"grad_norm": 0.8948563933372498,
"learning_rate": 4.458978604237325e-06,
"loss": 0.6447,
"step": 1624
},
{
"epoch": 1.3453892876863611,
"grad_norm": 0.9227688312530518,
"learning_rate": 4.458295203593782e-06,
"loss": 0.6459,
"step": 1625
},
{
"epoch": 1.3462175593594699,
"grad_norm": 0.9532939195632935,
"learning_rate": 4.4576114240383725e-06,
"loss": 0.6724,
"step": 1626
},
{
"epoch": 1.3470458310325788,
"grad_norm": 0.9089633226394653,
"learning_rate": 4.4569272657034005e-06,
"loss": 0.6303,
"step": 1627
},
{
"epoch": 1.3478741027056875,
"grad_norm": 0.8903496861457825,
"learning_rate": 4.456242728721244e-06,
"loss": 0.6655,
"step": 1628
},
{
"epoch": 1.3487023743787963,
"grad_norm": 0.9009859561920166,
"learning_rate": 4.455557813224356e-06,
"loss": 0.6268,
"step": 1629
},
{
"epoch": 1.349530646051905,
"grad_norm": 0.9265439510345459,
"learning_rate": 4.454872519345261e-06,
"loss": 0.6586,
"step": 1630
},
{
"epoch": 1.3503589177250137,
"grad_norm": 0.8945439457893372,
"learning_rate": 4.454186847216556e-06,
"loss": 0.62,
"step": 1631
},
{
"epoch": 1.3511871893981227,
"grad_norm": 0.9313004016876221,
"learning_rate": 4.453500796970913e-06,
"loss": 0.6452,
"step": 1632
},
{
"epoch": 1.3520154610712314,
"grad_norm": 0.9005258083343506,
"learning_rate": 4.452814368741076e-06,
"loss": 0.6436,
"step": 1633
},
{
"epoch": 1.35284373274434,
"grad_norm": 0.8860014081001282,
"learning_rate": 4.452127562659864e-06,
"loss": 0.6423,
"step": 1634
},
{
"epoch": 1.353672004417449,
"grad_norm": 0.9049930572509766,
"learning_rate": 4.4514403788601645e-06,
"loss": 0.6343,
"step": 1635
},
{
"epoch": 1.3545002760905578,
"grad_norm": 0.8849244117736816,
"learning_rate": 4.4507528174749425e-06,
"loss": 0.6481,
"step": 1636
},
{
"epoch": 1.3553285477636665,
"grad_norm": 0.9214544892311096,
"learning_rate": 4.450064878637235e-06,
"loss": 0.6166,
"step": 1637
},
{
"epoch": 1.3561568194367752,
"grad_norm": 0.8948760628700256,
"learning_rate": 4.449376562480153e-06,
"loss": 0.6195,
"step": 1638
},
{
"epoch": 1.356985091109884,
"grad_norm": 0.9000245928764343,
"learning_rate": 4.448687869136878e-06,
"loss": 0.6378,
"step": 1639
},
{
"epoch": 1.3578133627829927,
"grad_norm": 0.8963801264762878,
"learning_rate": 4.447998798740665e-06,
"loss": 0.6118,
"step": 1640
},
{
"epoch": 1.3586416344561016,
"grad_norm": 0.8774341344833374,
"learning_rate": 4.447309351424843e-06,
"loss": 0.6184,
"step": 1641
},
{
"epoch": 1.3594699061292104,
"grad_norm": 0.8901870846748352,
"learning_rate": 4.446619527322815e-06,
"loss": 0.6499,
"step": 1642
},
{
"epoch": 1.360298177802319,
"grad_norm": 0.9108647108078003,
"learning_rate": 4.445929326568055e-06,
"loss": 0.6243,
"step": 1643
},
{
"epoch": 1.361126449475428,
"grad_norm": 0.9178787469863892,
"learning_rate": 4.445238749294109e-06,
"loss": 0.6458,
"step": 1644
},
{
"epoch": 1.3619547211485368,
"grad_norm": 0.9278320670127869,
"learning_rate": 4.444547795634598e-06,
"loss": 0.6281,
"step": 1645
},
{
"epoch": 1.3627829928216455,
"grad_norm": 0.8769298195838928,
"learning_rate": 4.443856465723216e-06,
"loss": 0.6387,
"step": 1646
},
{
"epoch": 1.3636112644947542,
"grad_norm": 0.9313247203826904,
"learning_rate": 4.443164759693729e-06,
"loss": 0.6304,
"step": 1647
},
{
"epoch": 1.364439536167863,
"grad_norm": 1.0037778615951538,
"learning_rate": 4.4424726776799745e-06,
"loss": 0.6378,
"step": 1648
},
{
"epoch": 1.365267807840972,
"grad_norm": 0.9090721011161804,
"learning_rate": 4.441780219815863e-06,
"loss": 0.6456,
"step": 1649
},
{
"epoch": 1.3660960795140806,
"grad_norm": 0.8968444466590881,
"learning_rate": 4.441087386235382e-06,
"loss": 0.6377,
"step": 1650
},
{
"epoch": 1.3669243511871894,
"grad_norm": 0.9638763070106506,
"learning_rate": 4.440394177072586e-06,
"loss": 0.651,
"step": 1651
},
{
"epoch": 1.3677526228602983,
"grad_norm": 0.9224078059196472,
"learning_rate": 4.439700592461604e-06,
"loss": 0.6492,
"step": 1652
},
{
"epoch": 1.368580894533407,
"grad_norm": 0.891180157661438,
"learning_rate": 4.439006632536639e-06,
"loss": 0.6402,
"step": 1653
},
{
"epoch": 1.3694091662065158,
"grad_norm": 0.8817652463912964,
"learning_rate": 4.438312297431967e-06,
"loss": 0.6095,
"step": 1654
},
{
"epoch": 1.3702374378796245,
"grad_norm": 0.8814468383789062,
"learning_rate": 4.437617587281932e-06,
"loss": 0.6334,
"step": 1655
},
{
"epoch": 1.3710657095527332,
"grad_norm": 0.8895065784454346,
"learning_rate": 4.436922502220958e-06,
"loss": 0.6163,
"step": 1656
},
{
"epoch": 1.371893981225842,
"grad_norm": 0.9043355584144592,
"learning_rate": 4.436227042383535e-06,
"loss": 0.6625,
"step": 1657
},
{
"epoch": 1.372722252898951,
"grad_norm": 0.942014753818512,
"learning_rate": 4.435531207904229e-06,
"loss": 0.6298,
"step": 1658
},
{
"epoch": 1.3735505245720596,
"grad_norm": 0.8830228447914124,
"learning_rate": 4.434834998917677e-06,
"loss": 0.6215,
"step": 1659
},
{
"epoch": 1.3743787962451683,
"grad_norm": 0.9138136506080627,
"learning_rate": 4.434138415558588e-06,
"loss": 0.63,
"step": 1660
},
{
"epoch": 1.3752070679182773,
"grad_norm": 0.8880951404571533,
"learning_rate": 4.433441457961746e-06,
"loss": 0.6411,
"step": 1661
},
{
"epoch": 1.376035339591386,
"grad_norm": 0.9435725808143616,
"learning_rate": 4.432744126262005e-06,
"loss": 0.6279,
"step": 1662
},
{
"epoch": 1.3768636112644947,
"grad_norm": 0.894296407699585,
"learning_rate": 4.432046420594292e-06,
"loss": 0.6298,
"step": 1663
},
{
"epoch": 1.3776918829376035,
"grad_norm": 0.8850218653678894,
"learning_rate": 4.431348341093608e-06,
"loss": 0.627,
"step": 1664
},
{
"epoch": 1.3785201546107122,
"grad_norm": 0.8956788182258606,
"learning_rate": 4.430649887895025e-06,
"loss": 0.6412,
"step": 1665
},
{
"epoch": 1.3793484262838211,
"grad_norm": 0.9128378033638,
"learning_rate": 4.429951061133683e-06,
"loss": 0.6208,
"step": 1666
},
{
"epoch": 1.3801766979569299,
"grad_norm": 0.8763598203659058,
"learning_rate": 4.429251860944803e-06,
"loss": 0.6317,
"step": 1667
},
{
"epoch": 1.3810049696300386,
"grad_norm": 0.9210475087165833,
"learning_rate": 4.428552287463672e-06,
"loss": 0.6321,
"step": 1668
},
{
"epoch": 1.3818332413031476,
"grad_norm": 0.9203745126724243,
"learning_rate": 4.427852340825651e-06,
"loss": 0.6442,
"step": 1669
},
{
"epoch": 1.3826615129762563,
"grad_norm": 0.9229859709739685,
"learning_rate": 4.4271520211661735e-06,
"loss": 0.6402,
"step": 1670
},
{
"epoch": 1.383489784649365,
"grad_norm": 0.9455426335334778,
"learning_rate": 4.426451328620744e-06,
"loss": 0.6402,
"step": 1671
},
{
"epoch": 1.3843180563224737,
"grad_norm": 0.8894345760345459,
"learning_rate": 4.425750263324942e-06,
"loss": 0.6429,
"step": 1672
},
{
"epoch": 1.3851463279955825,
"grad_norm": 0.871449887752533,
"learning_rate": 4.425048825414414e-06,
"loss": 0.6449,
"step": 1673
},
{
"epoch": 1.3859745996686914,
"grad_norm": 0.9102624654769897,
"learning_rate": 4.424347015024884e-06,
"loss": 0.634,
"step": 1674
},
{
"epoch": 1.3868028713418001,
"grad_norm": 0.881172776222229,
"learning_rate": 4.4236448322921445e-06,
"loss": 0.621,
"step": 1675
},
{
"epoch": 1.3876311430149089,
"grad_norm": 0.9143398404121399,
"learning_rate": 4.422942277352063e-06,
"loss": 0.6379,
"step": 1676
},
{
"epoch": 1.3884594146880176,
"grad_norm": 0.9155046343803406,
"learning_rate": 4.4222393503405754e-06,
"loss": 0.6168,
"step": 1677
},
{
"epoch": 1.3892876863611265,
"grad_norm": 0.9501062631607056,
"learning_rate": 4.421536051393694e-06,
"loss": 0.6238,
"step": 1678
},
{
"epoch": 1.3901159580342353,
"grad_norm": 0.9055907130241394,
"learning_rate": 4.420832380647498e-06,
"loss": 0.634,
"step": 1679
},
{
"epoch": 1.390944229707344,
"grad_norm": 0.9199142456054688,
"learning_rate": 4.420128338238141e-06,
"loss": 0.6283,
"step": 1680
},
{
"epoch": 1.3917725013804527,
"grad_norm": 0.8932124376296997,
"learning_rate": 4.419423924301851e-06,
"loss": 0.6146,
"step": 1681
},
{
"epoch": 1.3926007730535614,
"grad_norm": 0.9010130167007446,
"learning_rate": 4.418719138974924e-06,
"loss": 0.6409,
"step": 1682
},
{
"epoch": 1.3934290447266704,
"grad_norm": 0.9154212474822998,
"learning_rate": 4.41801398239373e-06,
"loss": 0.6115,
"step": 1683
},
{
"epoch": 1.3942573163997791,
"grad_norm": 0.9208158850669861,
"learning_rate": 4.417308454694709e-06,
"loss": 0.6242,
"step": 1684
},
{
"epoch": 1.3950855880728878,
"grad_norm": 0.932996928691864,
"learning_rate": 4.4166025560143756e-06,
"loss": 0.6277,
"step": 1685
},
{
"epoch": 1.3959138597459968,
"grad_norm": 0.8990349769592285,
"learning_rate": 4.415896286489314e-06,
"loss": 0.6128,
"step": 1686
},
{
"epoch": 1.3967421314191055,
"grad_norm": 0.926906168460846,
"learning_rate": 4.4151896462561805e-06,
"loss": 0.6442,
"step": 1687
},
{
"epoch": 1.3975704030922143,
"grad_norm": 0.9330859184265137,
"learning_rate": 4.414482635451704e-06,
"loss": 0.5995,
"step": 1688
},
{
"epoch": 1.398398674765323,
"grad_norm": 0.9039576649665833,
"learning_rate": 4.413775254212683e-06,
"loss": 0.6363,
"step": 1689
},
{
"epoch": 1.3992269464384317,
"grad_norm": 0.9228506684303284,
"learning_rate": 4.413067502675992e-06,
"loss": 0.6294,
"step": 1690
},
{
"epoch": 1.4000552181115407,
"grad_norm": 0.921620786190033,
"learning_rate": 4.412359380978571e-06,
"loss": 0.64,
"step": 1691
},
{
"epoch": 1.4008834897846494,
"grad_norm": 0.9379149079322815,
"learning_rate": 4.4116508892574375e-06,
"loss": 0.6207,
"step": 1692
},
{
"epoch": 1.401711761457758,
"grad_norm": 1.0528016090393066,
"learning_rate": 4.410942027649677e-06,
"loss": 0.6204,
"step": 1693
},
{
"epoch": 1.402540033130867,
"grad_norm": 0.9510579705238342,
"learning_rate": 4.410232796292447e-06,
"loss": 0.6314,
"step": 1694
},
{
"epoch": 1.4033683048039758,
"grad_norm": 0.8922870755195618,
"learning_rate": 4.409523195322979e-06,
"loss": 0.6056,
"step": 1695
},
{
"epoch": 1.4041965764770845,
"grad_norm": 0.9239864349365234,
"learning_rate": 4.408813224878573e-06,
"loss": 0.6404,
"step": 1696
},
{
"epoch": 1.4050248481501932,
"grad_norm": 0.8696756362915039,
"learning_rate": 4.408102885096601e-06,
"loss": 0.6553,
"step": 1697
},
{
"epoch": 1.405853119823302,
"grad_norm": 0.9176343679428101,
"learning_rate": 4.407392176114509e-06,
"loss": 0.6286,
"step": 1698
},
{
"epoch": 1.4066813914964107,
"grad_norm": 0.9825496077537537,
"learning_rate": 4.40668109806981e-06,
"loss": 0.6421,
"step": 1699
},
{
"epoch": 1.4075096631695196,
"grad_norm": 0.9334873557090759,
"learning_rate": 4.405969651100093e-06,
"loss": 0.6347,
"step": 1700
},
{
"epoch": 1.4083379348426284,
"grad_norm": 0.8881192803382874,
"learning_rate": 4.405257835343015e-06,
"loss": 0.6377,
"step": 1701
},
{
"epoch": 1.409166206515737,
"grad_norm": 0.8896775245666504,
"learning_rate": 4.404545650936307e-06,
"loss": 0.624,
"step": 1702
},
{
"epoch": 1.409994478188846,
"grad_norm": 0.9361454844474792,
"learning_rate": 4.40383309801777e-06,
"loss": 0.6561,
"step": 1703
},
{
"epoch": 1.4108227498619548,
"grad_norm": 0.9082345962524414,
"learning_rate": 4.403120176725276e-06,
"loss": 0.6415,
"step": 1704
},
{
"epoch": 1.4116510215350635,
"grad_norm": 0.9453962445259094,
"learning_rate": 4.402406887196767e-06,
"loss": 0.6443,
"step": 1705
},
{
"epoch": 1.4124792932081722,
"grad_norm": 0.9420861005783081,
"learning_rate": 4.401693229570261e-06,
"loss": 0.6544,
"step": 1706
},
{
"epoch": 1.413307564881281,
"grad_norm": 0.869742214679718,
"learning_rate": 4.400979203983841e-06,
"loss": 0.6072,
"step": 1707
},
{
"epoch": 1.41413583655439,
"grad_norm": 0.915322482585907,
"learning_rate": 4.4002648105756665e-06,
"loss": 0.6657,
"step": 1708
},
{
"epoch": 1.4149641082274986,
"grad_norm": 0.9029238224029541,
"learning_rate": 4.3995500494839656e-06,
"loss": 0.6157,
"step": 1709
},
{
"epoch": 1.4157923799006074,
"grad_norm": 0.8992076516151428,
"learning_rate": 4.398834920847037e-06,
"loss": 0.6321,
"step": 1710
},
{
"epoch": 1.4166206515737163,
"grad_norm": 0.9099612236022949,
"learning_rate": 4.398119424803252e-06,
"loss": 0.6123,
"step": 1711
},
{
"epoch": 1.417448923246825,
"grad_norm": 0.9081803560256958,
"learning_rate": 4.397403561491052e-06,
"loss": 0.6369,
"step": 1712
},
{
"epoch": 1.4182771949199338,
"grad_norm": 0.8873599767684937,
"learning_rate": 4.3966873310489505e-06,
"loss": 0.6503,
"step": 1713
},
{
"epoch": 1.4191054665930425,
"grad_norm": 0.8841976523399353,
"learning_rate": 4.395970733615531e-06,
"loss": 0.6303,
"step": 1714
},
{
"epoch": 1.4199337382661512,
"grad_norm": 0.8922518491744995,
"learning_rate": 4.395253769329449e-06,
"loss": 0.6105,
"step": 1715
},
{
"epoch": 1.42076200993926,
"grad_norm": 0.9023404717445374,
"learning_rate": 4.39453643832943e-06,
"loss": 0.6292,
"step": 1716
},
{
"epoch": 1.421590281612369,
"grad_norm": 0.9042630791664124,
"learning_rate": 4.39381874075427e-06,
"loss": 0.6675,
"step": 1717
},
{
"epoch": 1.4224185532854776,
"grad_norm": 0.8933060169219971,
"learning_rate": 4.393100676742838e-06,
"loss": 0.6125,
"step": 1718
},
{
"epoch": 1.4232468249585863,
"grad_norm": 0.8821312189102173,
"learning_rate": 4.392382246434073e-06,
"loss": 0.6266,
"step": 1719
},
{
"epoch": 1.4240750966316953,
"grad_norm": 0.904460608959198,
"learning_rate": 4.391663449966983e-06,
"loss": 0.5984,
"step": 1720
},
{
"epoch": 1.424903368304804,
"grad_norm": 0.8999350666999817,
"learning_rate": 4.39094428748065e-06,
"loss": 0.6653,
"step": 1721
},
{
"epoch": 1.4257316399779127,
"grad_norm": 0.9061364531517029,
"learning_rate": 4.390224759114224e-06,
"loss": 0.6306,
"step": 1722
},
{
"epoch": 1.4265599116510215,
"grad_norm": 0.8995294570922852,
"learning_rate": 4.389504865006927e-06,
"loss": 0.6656,
"step": 1723
},
{
"epoch": 1.4273881833241302,
"grad_norm": 0.904737114906311,
"learning_rate": 4.388784605298053e-06,
"loss": 0.6234,
"step": 1724
},
{
"epoch": 1.4282164549972391,
"grad_norm": 0.8827239871025085,
"learning_rate": 4.388063980126966e-06,
"loss": 0.6418,
"step": 1725
},
{
"epoch": 1.4290447266703479,
"grad_norm": 0.8791747689247131,
"learning_rate": 4.387342989633097e-06,
"loss": 0.6174,
"step": 1726
},
{
"epoch": 1.4298729983434566,
"grad_norm": 0.8831204175949097,
"learning_rate": 4.386621633955955e-06,
"loss": 0.6222,
"step": 1727
},
{
"epoch": 1.4307012700165656,
"grad_norm": 0.908511221408844,
"learning_rate": 4.3858999132351124e-06,
"loss": 0.6373,
"step": 1728
},
{
"epoch": 1.4315295416896743,
"grad_norm": 0.900497555732727,
"learning_rate": 4.385177827610218e-06,
"loss": 0.6291,
"step": 1729
},
{
"epoch": 1.432357813362783,
"grad_norm": 0.930695652961731,
"learning_rate": 4.384455377220986e-06,
"loss": 0.6408,
"step": 1730
},
{
"epoch": 1.4331860850358917,
"grad_norm": 0.9583060145378113,
"learning_rate": 4.383732562207206e-06,
"loss": 0.6452,
"step": 1731
},
{
"epoch": 1.4340143567090005,
"grad_norm": 0.9173291921615601,
"learning_rate": 4.383009382708734e-06,
"loss": 0.6491,
"step": 1732
},
{
"epoch": 1.4348426283821094,
"grad_norm": 0.9054985642433167,
"learning_rate": 4.382285838865501e-06,
"loss": 0.6179,
"step": 1733
},
{
"epoch": 1.4356709000552181,
"grad_norm": 0.9162827730178833,
"learning_rate": 4.381561930817504e-06,
"loss": 0.6428,
"step": 1734
},
{
"epoch": 1.4364991717283269,
"grad_norm": 0.9013632535934448,
"learning_rate": 4.380837658704813e-06,
"loss": 0.6385,
"step": 1735
},
{
"epoch": 1.4373274434014356,
"grad_norm": 0.9274044632911682,
"learning_rate": 4.3801130226675685e-06,
"loss": 0.6508,
"step": 1736
},
{
"epoch": 1.4381557150745445,
"grad_norm": 0.8884710669517517,
"learning_rate": 4.37938802284598e-06,
"loss": 0.6675,
"step": 1737
},
{
"epoch": 1.4389839867476533,
"grad_norm": 0.9156724214553833,
"learning_rate": 4.378662659380329e-06,
"loss": 0.6424,
"step": 1738
},
{
"epoch": 1.439812258420762,
"grad_norm": 0.9267123341560364,
"learning_rate": 4.377936932410966e-06,
"loss": 0.6333,
"step": 1739
},
{
"epoch": 1.4406405300938707,
"grad_norm": 0.8859686851501465,
"learning_rate": 4.377210842078311e-06,
"loss": 0.6343,
"step": 1740
},
{
"epoch": 1.4414688017669794,
"grad_norm": 0.92928147315979,
"learning_rate": 4.376484388522858e-06,
"loss": 0.6124,
"step": 1741
},
{
"epoch": 1.4422970734400884,
"grad_norm": 0.8908365964889526,
"learning_rate": 4.375757571885169e-06,
"loss": 0.631,
"step": 1742
},
{
"epoch": 1.4431253451131971,
"grad_norm": 0.8956666588783264,
"learning_rate": 4.375030392305875e-06,
"loss": 0.6314,
"step": 1743
},
{
"epoch": 1.4439536167863058,
"grad_norm": 0.8836198449134827,
"learning_rate": 4.374302849925679e-06,
"loss": 0.6179,
"step": 1744
},
{
"epoch": 1.4447818884594148,
"grad_norm": 0.87999027967453,
"learning_rate": 4.373574944885354e-06,
"loss": 0.6226,
"step": 1745
},
{
"epoch": 1.4456101601325235,
"grad_norm": 0.9118010997772217,
"learning_rate": 4.3728466773257414e-06,
"loss": 0.6319,
"step": 1746
},
{
"epoch": 1.4464384318056323,
"grad_norm": 0.8989576101303101,
"learning_rate": 4.3721180473877555e-06,
"loss": 0.6049,
"step": 1747
},
{
"epoch": 1.447266703478741,
"grad_norm": 0.8896577954292297,
"learning_rate": 4.371389055212379e-06,
"loss": 0.6242,
"step": 1748
},
{
"epoch": 1.4480949751518497,
"grad_norm": 0.8993588089942932,
"learning_rate": 4.370659700940665e-06,
"loss": 0.6255,
"step": 1749
},
{
"epoch": 1.4489232468249587,
"grad_norm": 0.896486759185791,
"learning_rate": 4.369929984713738e-06,
"loss": 0.6354,
"step": 1750
},
{
"epoch": 1.4497515184980674,
"grad_norm": 0.9003145694732666,
"learning_rate": 4.369199906672789e-06,
"loss": 0.6464,
"step": 1751
},
{
"epoch": 1.450579790171176,
"grad_norm": 0.8885106444358826,
"learning_rate": 4.368469466959083e-06,
"loss": 0.6567,
"step": 1752
},
{
"epoch": 1.451408061844285,
"grad_norm": 0.8914943933486938,
"learning_rate": 4.367738665713953e-06,
"loss": 0.6784,
"step": 1753
},
{
"epoch": 1.4522363335173938,
"grad_norm": 0.933029055595398,
"learning_rate": 4.367007503078802e-06,
"loss": 0.6495,
"step": 1754
},
{
"epoch": 1.4530646051905025,
"grad_norm": 0.8788006901741028,
"learning_rate": 4.366275979195102e-06,
"loss": 0.6352,
"step": 1755
},
{
"epoch": 1.4538928768636112,
"grad_norm": 0.8768455982208252,
"learning_rate": 4.365544094204399e-06,
"loss": 0.6329,
"step": 1756
},
{
"epoch": 1.45472114853672,
"grad_norm": 0.9021796584129333,
"learning_rate": 4.364811848248303e-06,
"loss": 0.6616,
"step": 1757
},
{
"epoch": 1.4555494202098287,
"grad_norm": 0.9197986125946045,
"learning_rate": 4.364079241468499e-06,
"loss": 0.6248,
"step": 1758
},
{
"epoch": 1.4563776918829376,
"grad_norm": 0.9095743894577026,
"learning_rate": 4.3633462740067375e-06,
"loss": 0.6722,
"step": 1759
},
{
"epoch": 1.4572059635560464,
"grad_norm": 0.9114678502082825,
"learning_rate": 4.362612946004843e-06,
"loss": 0.6278,
"step": 1760
},
{
"epoch": 1.458034235229155,
"grad_norm": 0.8978061676025391,
"learning_rate": 4.3618792576047054e-06,
"loss": 0.633,
"step": 1761
},
{
"epoch": 1.458862506902264,
"grad_norm": 0.8578571677207947,
"learning_rate": 4.361145208948287e-06,
"loss": 0.6263,
"step": 1762
},
{
"epoch": 1.4596907785753728,
"grad_norm": 0.9333001375198364,
"learning_rate": 4.360410800177622e-06,
"loss": 0.6383,
"step": 1763
},
{
"epoch": 1.4605190502484815,
"grad_norm": 0.9448907971382141,
"learning_rate": 4.359676031434808e-06,
"loss": 0.6324,
"step": 1764
},
{
"epoch": 1.4613473219215902,
"grad_norm": 0.8806003928184509,
"learning_rate": 4.358940902862019e-06,
"loss": 0.6365,
"step": 1765
},
{
"epoch": 1.462175593594699,
"grad_norm": 0.9975590109825134,
"learning_rate": 4.358205414601493e-06,
"loss": 0.6436,
"step": 1766
},
{
"epoch": 1.463003865267808,
"grad_norm": 1.0076019763946533,
"learning_rate": 4.357469566795542e-06,
"loss": 0.6234,
"step": 1767
},
{
"epoch": 1.4638321369409166,
"grad_norm": 0.9254529476165771,
"learning_rate": 4.356733359586545e-06,
"loss": 0.6244,
"step": 1768
},
{
"epoch": 1.4646604086140254,
"grad_norm": 0.9129828810691833,
"learning_rate": 4.35599679311695e-06,
"loss": 0.6519,
"step": 1769
},
{
"epoch": 1.4654886802871343,
"grad_norm": 0.9141498804092407,
"learning_rate": 4.355259867529278e-06,
"loss": 0.6343,
"step": 1770
},
{
"epoch": 1.466316951960243,
"grad_norm": 0.920790433883667,
"learning_rate": 4.354522582966116e-06,
"loss": 0.6505,
"step": 1771
},
{
"epoch": 1.4671452236333518,
"grad_norm": 0.9683033227920532,
"learning_rate": 4.353784939570121e-06,
"loss": 0.6238,
"step": 1772
},
{
"epoch": 1.4679734953064605,
"grad_norm": 0.882957935333252,
"learning_rate": 4.3530469374840226e-06,
"loss": 0.6358,
"step": 1773
},
{
"epoch": 1.4688017669795692,
"grad_norm": 0.943648099899292,
"learning_rate": 4.352308576850615e-06,
"loss": 0.6274,
"step": 1774
},
{
"epoch": 1.4696300386526782,
"grad_norm": 0.9609524607658386,
"learning_rate": 4.351569857812764e-06,
"loss": 0.6368,
"step": 1775
},
{
"epoch": 1.4704583103257869,
"grad_norm": 0.9170458912849426,
"learning_rate": 4.3508307805134055e-06,
"loss": 0.6515,
"step": 1776
},
{
"epoch": 1.4712865819988956,
"grad_norm": 0.8956430554389954,
"learning_rate": 4.350091345095545e-06,
"loss": 0.6394,
"step": 1777
},
{
"epoch": 1.4721148536720043,
"grad_norm": 0.9145321846008301,
"learning_rate": 4.349351551702256e-06,
"loss": 0.6572,
"step": 1778
},
{
"epoch": 1.4729431253451133,
"grad_norm": 0.8807089328765869,
"learning_rate": 4.34861140047668e-06,
"loss": 0.6391,
"step": 1779
},
{
"epoch": 1.473771397018222,
"grad_norm": 0.8828510642051697,
"learning_rate": 4.347870891562032e-06,
"loss": 0.6583,
"step": 1780
},
{
"epoch": 1.4745996686913307,
"grad_norm": 0.9017962217330933,
"learning_rate": 4.34713002510159e-06,
"loss": 0.6337,
"step": 1781
},
{
"epoch": 1.4754279403644395,
"grad_norm": 0.8738356232643127,
"learning_rate": 4.346388801238708e-06,
"loss": 0.637,
"step": 1782
},
{
"epoch": 1.4762562120375482,
"grad_norm": 0.8805137872695923,
"learning_rate": 4.345647220116805e-06,
"loss": 0.6306,
"step": 1783
},
{
"epoch": 1.4770844837106571,
"grad_norm": 0.923599362373352,
"learning_rate": 4.34490528187937e-06,
"loss": 0.6254,
"step": 1784
},
{
"epoch": 1.4779127553837659,
"grad_norm": 0.9181063175201416,
"learning_rate": 4.344162986669961e-06,
"loss": 0.6257,
"step": 1785
},
{
"epoch": 1.4787410270568746,
"grad_norm": 0.9060764312744141,
"learning_rate": 4.343420334632204e-06,
"loss": 0.6326,
"step": 1786
},
{
"epoch": 1.4795692987299836,
"grad_norm": 0.9101755023002625,
"learning_rate": 4.342677325909798e-06,
"loss": 0.6319,
"step": 1787
},
{
"epoch": 1.4803975704030923,
"grad_norm": 0.9417787790298462,
"learning_rate": 4.341933960646505e-06,
"loss": 0.6147,
"step": 1788
},
{
"epoch": 1.481225842076201,
"grad_norm": 0.9138630628585815,
"learning_rate": 4.341190238986162e-06,
"loss": 0.6256,
"step": 1789
},
{
"epoch": 1.4820541137493097,
"grad_norm": 0.9004842638969421,
"learning_rate": 4.340446161072671e-06,
"loss": 0.6443,
"step": 1790
},
{
"epoch": 1.4828823854224185,
"grad_norm": 0.9181780219078064,
"learning_rate": 4.339701727050004e-06,
"loss": 0.6408,
"step": 1791
},
{
"epoch": 1.4837106570955274,
"grad_norm": 0.880736768245697,
"learning_rate": 4.338956937062203e-06,
"loss": 0.6436,
"step": 1792
},
{
"epoch": 1.4845389287686361,
"grad_norm": 0.9168913960456848,
"learning_rate": 4.338211791253376e-06,
"loss": 0.6391,
"step": 1793
},
{
"epoch": 1.4853672004417449,
"grad_norm": 0.942925751209259,
"learning_rate": 4.337466289767704e-06,
"loss": 0.6303,
"step": 1794
},
{
"epoch": 1.4861954721148538,
"grad_norm": 0.8768817186355591,
"learning_rate": 4.336720432749433e-06,
"loss": 0.6239,
"step": 1795
},
{
"epoch": 1.4870237437879625,
"grad_norm": 0.9388988614082336,
"learning_rate": 4.33597422034288e-06,
"loss": 0.6441,
"step": 1796
},
{
"epoch": 1.4878520154610713,
"grad_norm": 0.9257978200912476,
"learning_rate": 4.3352276526924295e-06,
"loss": 0.6373,
"step": 1797
},
{
"epoch": 1.48868028713418,
"grad_norm": 0.9225389957427979,
"learning_rate": 4.334480729942537e-06,
"loss": 0.6555,
"step": 1798
},
{
"epoch": 1.4895085588072887,
"grad_norm": 0.8992138504981995,
"learning_rate": 4.3337334522377235e-06,
"loss": 0.6375,
"step": 1799
},
{
"epoch": 1.4903368304803974,
"grad_norm": 0.9207738041877747,
"learning_rate": 4.332985819722582e-06,
"loss": 0.6439,
"step": 1800
},
{
"epoch": 1.4911651021535064,
"grad_norm": 0.8866941332817078,
"learning_rate": 4.3322378325417705e-06,
"loss": 0.6326,
"step": 1801
},
{
"epoch": 1.4919933738266151,
"grad_norm": 0.9069377779960632,
"learning_rate": 4.3314894908400175e-06,
"loss": 0.6193,
"step": 1802
},
{
"epoch": 1.4928216454997238,
"grad_norm": 0.8918569087982178,
"learning_rate": 4.3307407947621216e-06,
"loss": 0.6374,
"step": 1803
},
{
"epoch": 1.4936499171728328,
"grad_norm": 0.8993483781814575,
"learning_rate": 4.329991744452948e-06,
"loss": 0.6251,
"step": 1804
},
{
"epoch": 1.4944781888459415,
"grad_norm": 0.918445885181427,
"learning_rate": 4.32924234005743e-06,
"loss": 0.6494,
"step": 1805
},
{
"epoch": 1.4953064605190503,
"grad_norm": 0.9245757460594177,
"learning_rate": 4.328492581720573e-06,
"loss": 0.6172,
"step": 1806
},
{
"epoch": 1.496134732192159,
"grad_norm": 0.9166370034217834,
"learning_rate": 4.327742469587445e-06,
"loss": 0.622,
"step": 1807
},
{
"epoch": 1.4969630038652677,
"grad_norm": 0.9083957076072693,
"learning_rate": 4.326992003803187e-06,
"loss": 0.6205,
"step": 1808
},
{
"epoch": 1.4977912755383767,
"grad_norm": 0.8895595073699951,
"learning_rate": 4.326241184513008e-06,
"loss": 0.6245,
"step": 1809
},
{
"epoch": 1.4986195472114854,
"grad_norm": 0.9845629334449768,
"learning_rate": 4.325490011862184e-06,
"loss": 0.6163,
"step": 1810
},
{
"epoch": 1.499447818884594,
"grad_norm": 0.9501281976699829,
"learning_rate": 4.324738485996058e-06,
"loss": 0.6218,
"step": 1811
},
{
"epoch": 1.500276090557703,
"grad_norm": 0.9685792922973633,
"learning_rate": 4.323986607060047e-06,
"loss": 0.634,
"step": 1812
},
{
"epoch": 1.5011043622308118,
"grad_norm": 0.8959722518920898,
"learning_rate": 4.32323437519963e-06,
"loss": 0.6316,
"step": 1813
},
{
"epoch": 1.5019326339039205,
"grad_norm": 0.914463222026825,
"learning_rate": 4.322481790560356e-06,
"loss": 0.6258,
"step": 1814
},
{
"epoch": 1.5027609055770292,
"grad_norm": 0.9052320718765259,
"learning_rate": 4.321728853287846e-06,
"loss": 0.6345,
"step": 1815
},
{
"epoch": 1.503589177250138,
"grad_norm": 0.9306650161743164,
"learning_rate": 4.320975563527785e-06,
"loss": 0.5976,
"step": 1816
},
{
"epoch": 1.5044174489232467,
"grad_norm": 0.8930769562721252,
"learning_rate": 4.320221921425926e-06,
"loss": 0.6427,
"step": 1817
},
{
"epoch": 1.5052457205963556,
"grad_norm": 0.9205043911933899,
"learning_rate": 4.3194679271280935e-06,
"loss": 0.617,
"step": 1818
},
{
"epoch": 1.5060739922694644,
"grad_norm": 1.0371367931365967,
"learning_rate": 4.318713580780179e-06,
"loss": 0.6301,
"step": 1819
},
{
"epoch": 1.5069022639425733,
"grad_norm": 0.9442588686943054,
"learning_rate": 4.317958882528139e-06,
"loss": 0.6252,
"step": 1820
},
{
"epoch": 1.507730535615682,
"grad_norm": 0.9214493036270142,
"learning_rate": 4.317203832518003e-06,
"loss": 0.6326,
"step": 1821
},
{
"epoch": 1.5085588072887908,
"grad_norm": 0.9270440340042114,
"learning_rate": 4.316448430895866e-06,
"loss": 0.6377,
"step": 1822
},
{
"epoch": 1.5093870789618995,
"grad_norm": 0.8822277188301086,
"learning_rate": 4.315692677807889e-06,
"loss": 0.6277,
"step": 1823
},
{
"epoch": 1.5102153506350082,
"grad_norm": 0.8926153779029846,
"learning_rate": 4.314936573400305e-06,
"loss": 0.6439,
"step": 1824
},
{
"epoch": 1.511043622308117,
"grad_norm": 0.9214072227478027,
"learning_rate": 4.314180117819412e-06,
"loss": 0.6519,
"step": 1825
},
{
"epoch": 1.5118718939812257,
"grad_norm": 0.881522536277771,
"learning_rate": 4.313423311211578e-06,
"loss": 0.6213,
"step": 1826
},
{
"epoch": 1.5127001656543346,
"grad_norm": 0.8960263729095459,
"learning_rate": 4.312666153723238e-06,
"loss": 0.6592,
"step": 1827
},
{
"epoch": 1.5135284373274434,
"grad_norm": 0.9176173210144043,
"learning_rate": 4.3119086455008956e-06,
"loss": 0.6569,
"step": 1828
},
{
"epoch": 1.5143567090005523,
"grad_norm": 0.9312153458595276,
"learning_rate": 4.311150786691119e-06,
"loss": 0.6324,
"step": 1829
},
{
"epoch": 1.515184980673661,
"grad_norm": 0.9202299118041992,
"learning_rate": 4.310392577440549e-06,
"loss": 0.6381,
"step": 1830
},
{
"epoch": 1.5160132523467698,
"grad_norm": 0.8895224928855896,
"learning_rate": 4.309634017895893e-06,
"loss": 0.6491,
"step": 1831
},
{
"epoch": 1.5168415240198785,
"grad_norm": 0.9424769878387451,
"learning_rate": 4.308875108203922e-06,
"loss": 0.6342,
"step": 1832
},
{
"epoch": 1.5176697956929872,
"grad_norm": 0.8676947951316833,
"learning_rate": 4.30811584851148e-06,
"loss": 0.6287,
"step": 1833
},
{
"epoch": 1.518498067366096,
"grad_norm": 0.8764673471450806,
"learning_rate": 4.3073562389654764e-06,
"loss": 0.6123,
"step": 1834
},
{
"epoch": 1.5193263390392049,
"grad_norm": 0.9222663640975952,
"learning_rate": 4.306596279712889e-06,
"loss": 0.6433,
"step": 1835
},
{
"epoch": 1.5201546107123136,
"grad_norm": 0.8839907050132751,
"learning_rate": 4.305835970900762e-06,
"loss": 0.6416,
"step": 1836
},
{
"epoch": 1.5209828823854226,
"grad_norm": 0.9095149636268616,
"learning_rate": 4.305075312676209e-06,
"loss": 0.6417,
"step": 1837
},
{
"epoch": 1.5218111540585313,
"grad_norm": 0.8718581199645996,
"learning_rate": 4.304314305186409e-06,
"loss": 0.6174,
"step": 1838
},
{
"epoch": 1.52263942573164,
"grad_norm": 0.9424081444740295,
"learning_rate": 4.303552948578611e-06,
"loss": 0.6064,
"step": 1839
},
{
"epoch": 1.5234676974047487,
"grad_norm": 0.9170297384262085,
"learning_rate": 4.30279124300013e-06,
"loss": 0.6078,
"step": 1840
},
{
"epoch": 1.5242959690778575,
"grad_norm": 0.8814254999160767,
"learning_rate": 4.302029188598348e-06,
"loss": 0.6194,
"step": 1841
},
{
"epoch": 1.5251242407509662,
"grad_norm": 0.9248694181442261,
"learning_rate": 4.301266785520718e-06,
"loss": 0.6452,
"step": 1842
},
{
"epoch": 1.5259525124240751,
"grad_norm": 0.8757627010345459,
"learning_rate": 4.300504033914754e-06,
"loss": 0.6486,
"step": 1843
},
{
"epoch": 1.5267807840971839,
"grad_norm": 0.9229605793952942,
"learning_rate": 4.299740933928045e-06,
"loss": 0.6363,
"step": 1844
},
{
"epoch": 1.5276090557702926,
"grad_norm": 0.9192973971366882,
"learning_rate": 4.298977485708243e-06,
"loss": 0.6383,
"step": 1845
},
{
"epoch": 1.5284373274434015,
"grad_norm": 0.9071361422538757,
"learning_rate": 4.2982136894030655e-06,
"loss": 0.6541,
"step": 1846
},
{
"epoch": 1.5292655991165103,
"grad_norm": 0.9067169427871704,
"learning_rate": 4.297449545160303e-06,
"loss": 0.6317,
"step": 1847
},
{
"epoch": 1.530093870789619,
"grad_norm": 0.9047722816467285,
"learning_rate": 4.2966850531278086e-06,
"loss": 0.6365,
"step": 1848
},
{
"epoch": 1.5309221424627277,
"grad_norm": 0.8948541879653931,
"learning_rate": 4.295920213453504e-06,
"loss": 0.6268,
"step": 1849
},
{
"epoch": 1.5317504141358365,
"grad_norm": 0.9061904549598694,
"learning_rate": 4.295155026285382e-06,
"loss": 0.6271,
"step": 1850
},
{
"epoch": 1.5325786858089452,
"grad_norm": 0.9092041850090027,
"learning_rate": 4.2943894917714946e-06,
"loss": 0.6148,
"step": 1851
},
{
"epoch": 1.5334069574820541,
"grad_norm": 0.932331919670105,
"learning_rate": 4.293623610059968e-06,
"loss": 0.647,
"step": 1852
},
{
"epoch": 1.5342352291551629,
"grad_norm": 0.9136486649513245,
"learning_rate": 4.292857381298993e-06,
"loss": 0.6326,
"step": 1853
},
{
"epoch": 1.5350635008282718,
"grad_norm": 0.9760140180587769,
"learning_rate": 4.292090805636827e-06,
"loss": 0.6324,
"step": 1854
},
{
"epoch": 1.5358917725013805,
"grad_norm": 0.9075965881347656,
"learning_rate": 4.291323883221797e-06,
"loss": 0.6194,
"step": 1855
},
{
"epoch": 1.5367200441744893,
"grad_norm": 0.8804707527160645,
"learning_rate": 4.290556614202292e-06,
"loss": 0.6246,
"step": 1856
},
{
"epoch": 1.537548315847598,
"grad_norm": 0.9123021960258484,
"learning_rate": 4.2897889987267755e-06,
"loss": 0.6156,
"step": 1857
},
{
"epoch": 1.5383765875207067,
"grad_norm": 0.9396944642066956,
"learning_rate": 4.289021036943771e-06,
"loss": 0.6268,
"step": 1858
},
{
"epoch": 1.5392048591938154,
"grad_norm": 0.858566164970398,
"learning_rate": 4.288252729001874e-06,
"loss": 0.6237,
"step": 1859
},
{
"epoch": 1.5400331308669244,
"grad_norm": 0.899341344833374,
"learning_rate": 4.287484075049743e-06,
"loss": 0.5973,
"step": 1860
},
{
"epoch": 1.5408614025400331,
"grad_norm": 0.9021979570388794,
"learning_rate": 4.286715075236106e-06,
"loss": 0.6164,
"step": 1861
},
{
"epoch": 1.541689674213142,
"grad_norm": 0.9693799018859863,
"learning_rate": 4.285945729709757e-06,
"loss": 0.6409,
"step": 1862
},
{
"epoch": 1.5425179458862508,
"grad_norm": 0.9177312254905701,
"learning_rate": 4.285176038619559e-06,
"loss": 0.6276,
"step": 1863
},
{
"epoch": 1.5433462175593595,
"grad_norm": 0.9382041692733765,
"learning_rate": 4.284406002114438e-06,
"loss": 0.643,
"step": 1864
},
{
"epoch": 1.5441744892324683,
"grad_norm": 0.8778985142707825,
"learning_rate": 4.28363562034339e-06,
"loss": 0.6119,
"step": 1865
},
{
"epoch": 1.545002760905577,
"grad_norm": 0.8890271782875061,
"learning_rate": 4.282864893455476e-06,
"loss": 0.6388,
"step": 1866
},
{
"epoch": 1.5458310325786857,
"grad_norm": 0.9081838130950928,
"learning_rate": 4.282093821599824e-06,
"loss": 0.6464,
"step": 1867
},
{
"epoch": 1.5466593042517944,
"grad_norm": 0.8805630803108215,
"learning_rate": 4.281322404925631e-06,
"loss": 0.6533,
"step": 1868
},
{
"epoch": 1.5474875759249034,
"grad_norm": 0.8817034363746643,
"learning_rate": 4.280550643582158e-06,
"loss": 0.643,
"step": 1869
},
{
"epoch": 1.548315847598012,
"grad_norm": 0.9003176689147949,
"learning_rate": 4.279778537718733e-06,
"loss": 0.6322,
"step": 1870
},
{
"epoch": 1.549144119271121,
"grad_norm": 0.8982899785041809,
"learning_rate": 4.2790060874847525e-06,
"loss": 0.6456,
"step": 1871
},
{
"epoch": 1.5499723909442298,
"grad_norm": 0.8730716109275818,
"learning_rate": 4.278233293029677e-06,
"loss": 0.6428,
"step": 1872
},
{
"epoch": 1.5508006626173385,
"grad_norm": 0.9135062098503113,
"learning_rate": 4.277460154503036e-06,
"loss": 0.645,
"step": 1873
},
{
"epoch": 1.5516289342904472,
"grad_norm": 0.9091652631759644,
"learning_rate": 4.276686672054424e-06,
"loss": 0.6256,
"step": 1874
},
{
"epoch": 1.552457205963556,
"grad_norm": 0.9090970754623413,
"learning_rate": 4.275912845833504e-06,
"loss": 0.6404,
"step": 1875
},
{
"epoch": 1.5532854776366647,
"grad_norm": 0.91478031873703,
"learning_rate": 4.275138675990003e-06,
"loss": 0.6498,
"step": 1876
},
{
"epoch": 1.5541137493097736,
"grad_norm": 0.8979254961013794,
"learning_rate": 4.274364162673715e-06,
"loss": 0.6275,
"step": 1877
},
{
"epoch": 1.5549420209828824,
"grad_norm": 0.92198646068573,
"learning_rate": 4.273589306034503e-06,
"loss": 0.6585,
"step": 1878
},
{
"epoch": 1.5557702926559913,
"grad_norm": 0.9540958404541016,
"learning_rate": 4.272814106222293e-06,
"loss": 0.6111,
"step": 1879
},
{
"epoch": 1.5565985643291,
"grad_norm": 0.8989667892456055,
"learning_rate": 4.272038563387079e-06,
"loss": 0.6255,
"step": 1880
},
{
"epoch": 1.5574268360022088,
"grad_norm": 0.9551357626914978,
"learning_rate": 4.2712626776789224e-06,
"loss": 0.6164,
"step": 1881
},
{
"epoch": 1.5582551076753175,
"grad_norm": 0.9278994798660278,
"learning_rate": 4.270486449247949e-06,
"loss": 0.6376,
"step": 1882
},
{
"epoch": 1.5590833793484262,
"grad_norm": 0.927180826663971,
"learning_rate": 4.269709878244352e-06,
"loss": 0.6602,
"step": 1883
},
{
"epoch": 1.559911651021535,
"grad_norm": 0.932837724685669,
"learning_rate": 4.268932964818389e-06,
"loss": 0.6229,
"step": 1884
},
{
"epoch": 1.5607399226946437,
"grad_norm": 0.895219087600708,
"learning_rate": 4.26815570912039e-06,
"loss": 0.6192,
"step": 1885
},
{
"epoch": 1.5615681943677526,
"grad_norm": 0.9075414538383484,
"learning_rate": 4.2673781113007415e-06,
"loss": 0.6418,
"step": 1886
},
{
"epoch": 1.5623964660408614,
"grad_norm": 0.9130772948265076,
"learning_rate": 4.266600171509905e-06,
"loss": 0.6402,
"step": 1887
},
{
"epoch": 1.5632247377139703,
"grad_norm": 0.8892185688018799,
"learning_rate": 4.265821889898402e-06,
"loss": 0.6213,
"step": 1888
},
{
"epoch": 1.564053009387079,
"grad_norm": 0.8853991031646729,
"learning_rate": 4.265043266616825e-06,
"loss": 0.6357,
"step": 1889
},
{
"epoch": 1.5648812810601878,
"grad_norm": 0.9273191690444946,
"learning_rate": 4.26426430181583e-06,
"loss": 0.6311,
"step": 1890
},
{
"epoch": 1.5657095527332965,
"grad_norm": 0.8890771865844727,
"learning_rate": 4.263484995646138e-06,
"loss": 0.6053,
"step": 1891
},
{
"epoch": 1.5665378244064052,
"grad_norm": 0.9022220373153687,
"learning_rate": 4.262705348258538e-06,
"loss": 0.6256,
"step": 1892
},
{
"epoch": 1.567366096079514,
"grad_norm": 0.9201255440711975,
"learning_rate": 4.261925359803886e-06,
"loss": 0.6302,
"step": 1893
},
{
"epoch": 1.5681943677526229,
"grad_norm": 0.9675379991531372,
"learning_rate": 4.261145030433101e-06,
"loss": 0.602,
"step": 1894
},
{
"epoch": 1.5690226394257316,
"grad_norm": 0.9039328098297119,
"learning_rate": 4.26036436029717e-06,
"loss": 0.6477,
"step": 1895
},
{
"epoch": 1.5698509110988406,
"grad_norm": 0.911361813545227,
"learning_rate": 4.259583349547144e-06,
"loss": 0.6211,
"step": 1896
},
{
"epoch": 1.5706791827719493,
"grad_norm": 0.8901238441467285,
"learning_rate": 4.258801998334143e-06,
"loss": 0.6442,
"step": 1897
},
{
"epoch": 1.571507454445058,
"grad_norm": 0.9290407299995422,
"learning_rate": 4.258020306809351e-06,
"loss": 0.6382,
"step": 1898
},
{
"epoch": 1.5723357261181667,
"grad_norm": 0.9179307222366333,
"learning_rate": 4.257238275124017e-06,
"loss": 0.6513,
"step": 1899
},
{
"epoch": 1.5731639977912755,
"grad_norm": 0.9212455153465271,
"learning_rate": 4.256455903429459e-06,
"loss": 0.6257,
"step": 1900
},
{
"epoch": 1.5739922694643842,
"grad_norm": 0.9126132130622864,
"learning_rate": 4.2556731918770565e-06,
"loss": 0.6384,
"step": 1901
},
{
"epoch": 1.5748205411374931,
"grad_norm": 0.8735355734825134,
"learning_rate": 4.254890140618257e-06,
"loss": 0.6031,
"step": 1902
},
{
"epoch": 1.5756488128106019,
"grad_norm": 0.9375724196434021,
"learning_rate": 4.254106749804575e-06,
"loss": 0.6333,
"step": 1903
},
{
"epoch": 1.5764770844837108,
"grad_norm": 0.9451910853385925,
"learning_rate": 4.253323019587588e-06,
"loss": 0.6186,
"step": 1904
},
{
"epoch": 1.5773053561568195,
"grad_norm": 0.9468234777450562,
"learning_rate": 4.25253895011894e-06,
"loss": 0.6023,
"step": 1905
},
{
"epoch": 1.5781336278299283,
"grad_norm": 0.8941681385040283,
"learning_rate": 4.251754541550343e-06,
"loss": 0.634,
"step": 1906
},
{
"epoch": 1.578961899503037,
"grad_norm": 0.9508136510848999,
"learning_rate": 4.250969794033573e-06,
"loss": 0.6346,
"step": 1907
},
{
"epoch": 1.5797901711761457,
"grad_norm": 0.8964453935623169,
"learning_rate": 4.2501847077204685e-06,
"loss": 0.6298,
"step": 1908
},
{
"epoch": 1.5806184428492545,
"grad_norm": 0.8885358572006226,
"learning_rate": 4.249399282762938e-06,
"loss": 0.6319,
"step": 1909
},
{
"epoch": 1.5814467145223632,
"grad_norm": 0.8929858207702637,
"learning_rate": 4.248613519312955e-06,
"loss": 0.6272,
"step": 1910
},
{
"epoch": 1.5822749861954721,
"grad_norm": 0.9664214849472046,
"learning_rate": 4.247827417522557e-06,
"loss": 0.6336,
"step": 1911
},
{
"epoch": 1.5831032578685809,
"grad_norm": 0.890846312046051,
"learning_rate": 4.247040977543845e-06,
"loss": 0.6242,
"step": 1912
},
{
"epoch": 1.5839315295416898,
"grad_norm": 0.8973182439804077,
"learning_rate": 4.246254199528991e-06,
"loss": 0.6336,
"step": 1913
},
{
"epoch": 1.5847598012147985,
"grad_norm": 0.9789074659347534,
"learning_rate": 4.245467083630227e-06,
"loss": 0.6191,
"step": 1914
},
{
"epoch": 1.5855880728879073,
"grad_norm": 0.9195513129234314,
"learning_rate": 4.244679629999853e-06,
"loss": 0.6333,
"step": 1915
},
{
"epoch": 1.586416344561016,
"grad_norm": 0.950316846370697,
"learning_rate": 4.243891838790235e-06,
"loss": 0.635,
"step": 1916
},
{
"epoch": 1.5872446162341247,
"grad_norm": 0.8971677422523499,
"learning_rate": 4.243103710153802e-06,
"loss": 0.6352,
"step": 1917
},
{
"epoch": 1.5880728879072334,
"grad_norm": 0.9198834896087646,
"learning_rate": 4.24231524424305e-06,
"loss": 0.6252,
"step": 1918
},
{
"epoch": 1.5889011595803424,
"grad_norm": 0.9535729885101318,
"learning_rate": 4.241526441210541e-06,
"loss": 0.6177,
"step": 1919
},
{
"epoch": 1.5897294312534511,
"grad_norm": 0.8910471200942993,
"learning_rate": 4.240737301208899e-06,
"loss": 0.6408,
"step": 1920
},
{
"epoch": 1.59055770292656,
"grad_norm": 0.862991452217102,
"learning_rate": 4.2399478243908164e-06,
"loss": 0.6221,
"step": 1921
},
{
"epoch": 1.5913859745996688,
"grad_norm": 0.8935141563415527,
"learning_rate": 4.239158010909049e-06,
"loss": 0.608,
"step": 1922
},
{
"epoch": 1.5922142462727775,
"grad_norm": 0.9177968502044678,
"learning_rate": 4.23836786091642e-06,
"loss": 0.6442,
"step": 1923
},
{
"epoch": 1.5930425179458862,
"grad_norm": 0.9157758951187134,
"learning_rate": 4.237577374565813e-06,
"loss": 0.6419,
"step": 1924
},
{
"epoch": 1.593870789618995,
"grad_norm": 0.9484630227088928,
"learning_rate": 4.236786552010182e-06,
"loss": 0.6733,
"step": 1925
},
{
"epoch": 1.5946990612921037,
"grad_norm": 0.9401251077651978,
"learning_rate": 4.235995393402544e-06,
"loss": 0.6325,
"step": 1926
},
{
"epoch": 1.5955273329652124,
"grad_norm": 0.9104607105255127,
"learning_rate": 4.23520389889598e-06,
"loss": 0.6313,
"step": 1927
},
{
"epoch": 1.5963556046383214,
"grad_norm": 0.9173511862754822,
"learning_rate": 4.234412068643637e-06,
"loss": 0.6044,
"step": 1928
},
{
"epoch": 1.59718387631143,
"grad_norm": 0.9189788699150085,
"learning_rate": 4.233619902798727e-06,
"loss": 0.6182,
"step": 1929
},
{
"epoch": 1.598012147984539,
"grad_norm": 0.910804808139801,
"learning_rate": 4.232827401514525e-06,
"loss": 0.6435,
"step": 1930
},
{
"epoch": 1.5988404196576478,
"grad_norm": 0.8986726403236389,
"learning_rate": 4.232034564944375e-06,
"loss": 0.6268,
"step": 1931
},
{
"epoch": 1.5996686913307565,
"grad_norm": 0.8782910108566284,
"learning_rate": 4.231241393241682e-06,
"loss": 0.6381,
"step": 1932
},
{
"epoch": 1.6004969630038652,
"grad_norm": 0.9176427721977234,
"learning_rate": 4.230447886559919e-06,
"loss": 0.6131,
"step": 1933
},
{
"epoch": 1.601325234676974,
"grad_norm": 0.8755510449409485,
"learning_rate": 4.22965404505262e-06,
"loss": 0.654,
"step": 1934
},
{
"epoch": 1.6021535063500827,
"grad_norm": 0.8644832968711853,
"learning_rate": 4.228859868873387e-06,
"loss": 0.627,
"step": 1935
},
{
"epoch": 1.6029817780231916,
"grad_norm": 0.9181885719299316,
"learning_rate": 4.228065358175886e-06,
"loss": 0.6062,
"step": 1936
},
{
"epoch": 1.6038100496963004,
"grad_norm": 0.9515942335128784,
"learning_rate": 4.227270513113846e-06,
"loss": 0.6262,
"step": 1937
},
{
"epoch": 1.6046383213694093,
"grad_norm": 0.9062270522117615,
"learning_rate": 4.226475333841063e-06,
"loss": 0.6139,
"step": 1938
},
{
"epoch": 1.605466593042518,
"grad_norm": 0.9040980339050293,
"learning_rate": 4.225679820511398e-06,
"loss": 0.64,
"step": 1939
},
{
"epoch": 1.6062948647156268,
"grad_norm": 0.8910935521125793,
"learning_rate": 4.224883973278775e-06,
"loss": 0.6239,
"step": 1940
},
{
"epoch": 1.6071231363887355,
"grad_norm": 0.9150440692901611,
"learning_rate": 4.224087792297181e-06,
"loss": 0.6283,
"step": 1941
},
{
"epoch": 1.6079514080618442,
"grad_norm": 0.9175113439559937,
"learning_rate": 4.223291277720671e-06,
"loss": 0.6498,
"step": 1942
},
{
"epoch": 1.608779679734953,
"grad_norm": 0.8720189332962036,
"learning_rate": 4.222494429703363e-06,
"loss": 0.623,
"step": 1943
},
{
"epoch": 1.609607951408062,
"grad_norm": 0.8868989944458008,
"learning_rate": 4.22169724839944e-06,
"loss": 0.6357,
"step": 1944
},
{
"epoch": 1.6104362230811706,
"grad_norm": 0.9084488153457642,
"learning_rate": 4.22089973396315e-06,
"loss": 0.6607,
"step": 1945
},
{
"epoch": 1.6112644947542794,
"grad_norm": 0.9061319231987,
"learning_rate": 4.220101886548802e-06,
"loss": 0.6309,
"step": 1946
},
{
"epoch": 1.6120927664273883,
"grad_norm": 0.895646870136261,
"learning_rate": 4.219303706310775e-06,
"loss": 0.6276,
"step": 1947
},
{
"epoch": 1.612921038100497,
"grad_norm": 0.8863298892974854,
"learning_rate": 4.218505193403506e-06,
"loss": 0.6444,
"step": 1948
},
{
"epoch": 1.6137493097736058,
"grad_norm": 0.9005767107009888,
"learning_rate": 4.2177063479815025e-06,
"loss": 0.6356,
"step": 1949
},
{
"epoch": 1.6145775814467145,
"grad_norm": 0.9094089865684509,
"learning_rate": 4.216907170199334e-06,
"loss": 0.6053,
"step": 1950
},
{
"epoch": 1.6154058531198232,
"grad_norm": 0.8869677782058716,
"learning_rate": 4.216107660211632e-06,
"loss": 0.6308,
"step": 1951
},
{
"epoch": 1.616234124792932,
"grad_norm": 0.9285000562667847,
"learning_rate": 4.215307818173095e-06,
"loss": 0.6188,
"step": 1952
},
{
"epoch": 1.6170623964660409,
"grad_norm": 0.9039749503135681,
"learning_rate": 4.214507644238485e-06,
"loss": 0.63,
"step": 1953
},
{
"epoch": 1.6178906681391496,
"grad_norm": 0.9202860593795776,
"learning_rate": 4.213707138562628e-06,
"loss": 0.632,
"step": 1954
},
{
"epoch": 1.6187189398122586,
"grad_norm": 0.8759778141975403,
"learning_rate": 4.212906301300415e-06,
"loss": 0.6253,
"step": 1955
},
{
"epoch": 1.6195472114853673,
"grad_norm": 0.8918818831443787,
"learning_rate": 4.212105132606801e-06,
"loss": 0.625,
"step": 1956
},
{
"epoch": 1.620375483158476,
"grad_norm": 0.9184549450874329,
"learning_rate": 4.211303632636804e-06,
"loss": 0.6293,
"step": 1957
},
{
"epoch": 1.6212037548315847,
"grad_norm": 0.8887206315994263,
"learning_rate": 4.210501801545506e-06,
"loss": 0.6118,
"step": 1958
},
{
"epoch": 1.6220320265046935,
"grad_norm": 0.9010420441627502,
"learning_rate": 4.209699639488054e-06,
"loss": 0.6209,
"step": 1959
},
{
"epoch": 1.6228602981778022,
"grad_norm": 0.8888309001922607,
"learning_rate": 4.20889714661966e-06,
"loss": 0.6301,
"step": 1960
},
{
"epoch": 1.6236885698509111,
"grad_norm": 0.9068322777748108,
"learning_rate": 4.208094323095597e-06,
"loss": 0.6234,
"step": 1961
},
{
"epoch": 1.6245168415240199,
"grad_norm": 0.8940340876579285,
"learning_rate": 4.207291169071207e-06,
"loss": 0.6076,
"step": 1962
},
{
"epoch": 1.6253451131971288,
"grad_norm": 0.907363772392273,
"learning_rate": 4.20648768470189e-06,
"loss": 0.6522,
"step": 1963
},
{
"epoch": 1.6261733848702375,
"grad_norm": 0.9104205965995789,
"learning_rate": 4.205683870143114e-06,
"loss": 0.6177,
"step": 1964
},
{
"epoch": 1.6270016565433463,
"grad_norm": 0.9051296710968018,
"learning_rate": 4.204879725550409e-06,
"loss": 0.6069,
"step": 1965
},
{
"epoch": 1.627829928216455,
"grad_norm": 0.9034865498542786,
"learning_rate": 4.20407525107937e-06,
"loss": 0.6099,
"step": 1966
},
{
"epoch": 1.6286581998895637,
"grad_norm": 0.9055590033531189,
"learning_rate": 4.203270446885655e-06,
"loss": 0.6226,
"step": 1967
},
{
"epoch": 1.6294864715626725,
"grad_norm": 0.897927463054657,
"learning_rate": 4.202465313124986e-06,
"loss": 0.6342,
"step": 1968
},
{
"epoch": 1.6303147432357812,
"grad_norm": 0.8995800018310547,
"learning_rate": 4.2016598499531495e-06,
"loss": 0.6408,
"step": 1969
},
{
"epoch": 1.6311430149088901,
"grad_norm": 0.8727264404296875,
"learning_rate": 4.200854057525995e-06,
"loss": 0.6263,
"step": 1970
},
{
"epoch": 1.6319712865819989,
"grad_norm": 0.9289891719818115,
"learning_rate": 4.200047935999435e-06,
"loss": 0.6351,
"step": 1971
},
{
"epoch": 1.6327995582551078,
"grad_norm": 0.8936645984649658,
"learning_rate": 4.199241485529446e-06,
"loss": 0.6205,
"step": 1972
},
{
"epoch": 1.6336278299282165,
"grad_norm": 0.9054183959960938,
"learning_rate": 4.198434706272072e-06,
"loss": 0.6101,
"step": 1973
},
{
"epoch": 1.6344561016013253,
"grad_norm": 0.9181169271469116,
"learning_rate": 4.1976275983834135e-06,
"loss": 0.6437,
"step": 1974
},
{
"epoch": 1.635284373274434,
"grad_norm": 0.9325777888298035,
"learning_rate": 4.1968201620196404e-06,
"loss": 0.6343,
"step": 1975
},
{
"epoch": 1.6361126449475427,
"grad_norm": 0.878094494342804,
"learning_rate": 4.196012397336983e-06,
"loss": 0.6298,
"step": 1976
},
{
"epoch": 1.6369409166206514,
"grad_norm": 0.9336429238319397,
"learning_rate": 4.195204304491737e-06,
"loss": 0.6232,
"step": 1977
},
{
"epoch": 1.6377691882937604,
"grad_norm": 0.9159499406814575,
"learning_rate": 4.19439588364026e-06,
"loss": 0.621,
"step": 1978
},
{
"epoch": 1.6385974599668691,
"grad_norm": 0.9075011610984802,
"learning_rate": 4.193587134938975e-06,
"loss": 0.6375,
"step": 1979
},
{
"epoch": 1.639425731639978,
"grad_norm": 0.8912638425827026,
"learning_rate": 4.192778058544367e-06,
"loss": 0.6494,
"step": 1980
},
{
"epoch": 1.6402540033130868,
"grad_norm": 0.9125227928161621,
"learning_rate": 4.191968654612984e-06,
"loss": 0.6238,
"step": 1981
},
{
"epoch": 1.6410822749861955,
"grad_norm": 0.906326174736023,
"learning_rate": 4.191158923301439e-06,
"loss": 0.5904,
"step": 1982
},
{
"epoch": 1.6419105466593042,
"grad_norm": 0.8858581781387329,
"learning_rate": 4.190348864766407e-06,
"loss": 0.6222,
"step": 1983
},
{
"epoch": 1.642738818332413,
"grad_norm": 0.864389955997467,
"learning_rate": 4.189538479164627e-06,
"loss": 0.6218,
"step": 1984
},
{
"epoch": 1.6435670900055217,
"grad_norm": 0.8983911871910095,
"learning_rate": 4.188727766652901e-06,
"loss": 0.6511,
"step": 1985
},
{
"epoch": 1.6443953616786304,
"grad_norm": 0.9633772373199463,
"learning_rate": 4.187916727388094e-06,
"loss": 0.623,
"step": 1986
},
{
"epoch": 1.6452236333517394,
"grad_norm": 0.9191408753395081,
"learning_rate": 4.187105361527135e-06,
"loss": 0.6245,
"step": 1987
},
{
"epoch": 1.646051905024848,
"grad_norm": 0.907792866230011,
"learning_rate": 4.186293669227016e-06,
"loss": 0.5994,
"step": 1988
},
{
"epoch": 1.646880176697957,
"grad_norm": 0.906218409538269,
"learning_rate": 4.18548165064479e-06,
"loss": 0.653,
"step": 1989
},
{
"epoch": 1.6477084483710658,
"grad_norm": 0.9408275485038757,
"learning_rate": 4.184669305937577e-06,
"loss": 0.5978,
"step": 1990
},
{
"epoch": 1.6485367200441745,
"grad_norm": 0.9278435111045837,
"learning_rate": 4.183856635262558e-06,
"loss": 0.6167,
"step": 1991
},
{
"epoch": 1.6493649917172832,
"grad_norm": 0.91252601146698,
"learning_rate": 4.183043638776978e-06,
"loss": 0.6229,
"step": 1992
},
{
"epoch": 1.650193263390392,
"grad_norm": 0.9178313612937927,
"learning_rate": 4.182230316638142e-06,
"loss": 0.6318,
"step": 1993
},
{
"epoch": 1.6510215350635007,
"grad_norm": 0.9056826233863831,
"learning_rate": 4.181416669003421e-06,
"loss": 0.5932,
"step": 1994
},
{
"epoch": 1.6518498067366096,
"grad_norm": 0.9139112830162048,
"learning_rate": 4.180602696030248e-06,
"loss": 0.6407,
"step": 1995
},
{
"epoch": 1.6526780784097184,
"grad_norm": 0.9195094108581543,
"learning_rate": 4.179788397876121e-06,
"loss": 0.6371,
"step": 1996
},
{
"epoch": 1.6535063500828273,
"grad_norm": 0.9221140146255493,
"learning_rate": 4.178973774698597e-06,
"loss": 0.611,
"step": 1997
},
{
"epoch": 1.654334621755936,
"grad_norm": 0.9180639386177063,
"learning_rate": 4.178158826655299e-06,
"loss": 0.6138,
"step": 1998
},
{
"epoch": 1.6551628934290448,
"grad_norm": 0.9255500435829163,
"learning_rate": 4.177343553903912e-06,
"loss": 0.6425,
"step": 1999
},
{
"epoch": 1.6559911651021535,
"grad_norm": 0.9378511905670166,
"learning_rate": 4.176527956602182e-06,
"loss": 0.6357,
"step": 2000
},
{
"epoch": 1.6568194367752622,
"grad_norm": 0.9212007522583008,
"learning_rate": 4.175712034907922e-06,
"loss": 0.6307,
"step": 2001
},
{
"epoch": 1.657647708448371,
"grad_norm": 0.9185839891433716,
"learning_rate": 4.174895788979004e-06,
"loss": 0.6173,
"step": 2002
},
{
"epoch": 1.65847598012148,
"grad_norm": 0.9658777117729187,
"learning_rate": 4.174079218973364e-06,
"loss": 0.622,
"step": 2003
},
{
"epoch": 1.6593042517945886,
"grad_norm": 0.9310017824172974,
"learning_rate": 4.173262325049e-06,
"loss": 0.6407,
"step": 2004
},
{
"epoch": 1.6601325234676974,
"grad_norm": 0.9048784375190735,
"learning_rate": 4.172445107363974e-06,
"loss": 0.6557,
"step": 2005
},
{
"epoch": 1.6609607951408063,
"grad_norm": 0.8820594549179077,
"learning_rate": 4.17162756607641e-06,
"loss": 0.6077,
"step": 2006
},
{
"epoch": 1.661789066813915,
"grad_norm": 0.8961437344551086,
"learning_rate": 4.170809701344496e-06,
"loss": 0.6151,
"step": 2007
},
{
"epoch": 1.6626173384870238,
"grad_norm": 0.9773615598678589,
"learning_rate": 4.169991513326479e-06,
"loss": 0.6339,
"step": 2008
},
{
"epoch": 1.6634456101601325,
"grad_norm": 0.9309290647506714,
"learning_rate": 4.169173002180672e-06,
"loss": 0.6235,
"step": 2009
},
{
"epoch": 1.6642738818332412,
"grad_norm": 1.0016266107559204,
"learning_rate": 4.168354168065448e-06,
"loss": 0.6355,
"step": 2010
},
{
"epoch": 1.66510215350635,
"grad_norm": 0.8820549249649048,
"learning_rate": 4.167535011139246e-06,
"loss": 0.6258,
"step": 2011
},
{
"epoch": 1.6659304251794589,
"grad_norm": 0.9030908942222595,
"learning_rate": 4.166715531560564e-06,
"loss": 0.6217,
"step": 2012
},
{
"epoch": 1.6667586968525676,
"grad_norm": 0.983755350112915,
"learning_rate": 4.165895729487962e-06,
"loss": 0.6281,
"step": 2013
},
{
"epoch": 1.6675869685256766,
"grad_norm": 0.948001503944397,
"learning_rate": 4.165075605080066e-06,
"loss": 0.6242,
"step": 2014
},
{
"epoch": 1.6684152401987853,
"grad_norm": 0.9468982219696045,
"learning_rate": 4.164255158495563e-06,
"loss": 0.6563,
"step": 2015
},
{
"epoch": 1.669243511871894,
"grad_norm": 0.9012095332145691,
"learning_rate": 4.1634343898932e-06,
"loss": 0.618,
"step": 2016
},
{
"epoch": 1.6700717835450027,
"grad_norm": 0.9554924964904785,
"learning_rate": 4.162613299431789e-06,
"loss": 0.6306,
"step": 2017
},
{
"epoch": 1.6709000552181115,
"grad_norm": 0.9587515592575073,
"learning_rate": 4.161791887270203e-06,
"loss": 0.6305,
"step": 2018
},
{
"epoch": 1.6717283268912202,
"grad_norm": 0.8792864084243774,
"learning_rate": 4.160970153567379e-06,
"loss": 0.6181,
"step": 2019
},
{
"epoch": 1.6725565985643291,
"grad_norm": 0.8958392143249512,
"learning_rate": 4.160148098482313e-06,
"loss": 0.623,
"step": 2020
},
{
"epoch": 1.6733848702374379,
"grad_norm": 0.9501842260360718,
"learning_rate": 4.159325722174065e-06,
"loss": 0.6454,
"step": 2021
},
{
"epoch": 1.6742131419105468,
"grad_norm": 0.9107666015625,
"learning_rate": 4.158503024801759e-06,
"loss": 0.6237,
"step": 2022
},
{
"epoch": 1.6750414135836555,
"grad_norm": 0.9405720233917236,
"learning_rate": 4.157680006524577e-06,
"loss": 0.6423,
"step": 2023
},
{
"epoch": 1.6758696852567643,
"grad_norm": 0.9008631706237793,
"learning_rate": 4.1568566675017675e-06,
"loss": 0.6385,
"step": 2024
},
{
"epoch": 1.676697956929873,
"grad_norm": 0.9056677222251892,
"learning_rate": 4.156033007892637e-06,
"loss": 0.6225,
"step": 2025
},
{
"epoch": 1.6775262286029817,
"grad_norm": 0.899982750415802,
"learning_rate": 4.155209027856557e-06,
"loss": 0.6268,
"step": 2026
},
{
"epoch": 1.6783545002760905,
"grad_norm": 0.928506076335907,
"learning_rate": 4.15438472755296e-06,
"loss": 0.6343,
"step": 2027
},
{
"epoch": 1.6791827719491992,
"grad_norm": 0.8831804394721985,
"learning_rate": 4.15356010714134e-06,
"loss": 0.6404,
"step": 2028
},
{
"epoch": 1.6800110436223081,
"grad_norm": 0.8637033700942993,
"learning_rate": 4.152735166781253e-06,
"loss": 0.6305,
"step": 2029
},
{
"epoch": 1.6808393152954169,
"grad_norm": 0.9164902567863464,
"learning_rate": 4.151909906632319e-06,
"loss": 0.6171,
"step": 2030
},
{
"epoch": 1.6816675869685258,
"grad_norm": 0.9124466776847839,
"learning_rate": 4.151084326854216e-06,
"loss": 0.6446,
"step": 2031
},
{
"epoch": 1.6824958586416345,
"grad_norm": 0.9212199449539185,
"learning_rate": 4.150258427606687e-06,
"loss": 0.6074,
"step": 2032
},
{
"epoch": 1.6833241303147433,
"grad_norm": 0.9536449313163757,
"learning_rate": 4.149432209049536e-06,
"loss": 0.6149,
"step": 2033
},
{
"epoch": 1.684152401987852,
"grad_norm": 0.9232591390609741,
"learning_rate": 4.148605671342629e-06,
"loss": 0.6238,
"step": 2034
},
{
"epoch": 1.6849806736609607,
"grad_norm": 0.8844353556632996,
"learning_rate": 4.147778814645892e-06,
"loss": 0.6434,
"step": 2035
},
{
"epoch": 1.6858089453340694,
"grad_norm": 0.8500795960426331,
"learning_rate": 4.1469516391193155e-06,
"loss": 0.6304,
"step": 2036
},
{
"epoch": 1.6866372170071784,
"grad_norm": 0.8676083087921143,
"learning_rate": 4.146124144922949e-06,
"loss": 0.6158,
"step": 2037
},
{
"epoch": 1.6874654886802871,
"grad_norm": 0.8914940357208252,
"learning_rate": 4.1452963322169046e-06,
"loss": 0.6336,
"step": 2038
},
{
"epoch": 1.688293760353396,
"grad_norm": 0.9245527386665344,
"learning_rate": 4.144468201161358e-06,
"loss": 0.6385,
"step": 2039
},
{
"epoch": 1.6891220320265048,
"grad_norm": 0.8946509957313538,
"learning_rate": 4.143639751916544e-06,
"loss": 0.6243,
"step": 2040
},
{
"epoch": 1.6899503036996135,
"grad_norm": 0.882095217704773,
"learning_rate": 4.14281098464276e-06,
"loss": 0.6171,
"step": 2041
},
{
"epoch": 1.6907785753727222,
"grad_norm": 0.8981113433837891,
"learning_rate": 4.141981899500364e-06,
"loss": 0.6389,
"step": 2042
},
{
"epoch": 1.691606847045831,
"grad_norm": 0.928701639175415,
"learning_rate": 4.1411524966497775e-06,
"loss": 0.6488,
"step": 2043
},
{
"epoch": 1.6924351187189397,
"grad_norm": 0.8937084674835205,
"learning_rate": 4.140322776251481e-06,
"loss": 0.6415,
"step": 2044
},
{
"epoch": 1.6932633903920484,
"grad_norm": 0.8940730094909668,
"learning_rate": 4.1394927384660185e-06,
"loss": 0.6419,
"step": 2045
},
{
"epoch": 1.6940916620651574,
"grad_norm": 0.9067767262458801,
"learning_rate": 4.138662383453994e-06,
"loss": 0.608,
"step": 2046
},
{
"epoch": 1.694919933738266,
"grad_norm": 0.9123063087463379,
"learning_rate": 4.137831711376074e-06,
"loss": 0.633,
"step": 2047
},
{
"epoch": 1.695748205411375,
"grad_norm": 0.89399653673172,
"learning_rate": 4.137000722392985e-06,
"loss": 0.6338,
"step": 2048
},
{
"epoch": 1.6965764770844838,
"grad_norm": 0.9146568775177002,
"learning_rate": 4.136169416665517e-06,
"loss": 0.6214,
"step": 2049
},
{
"epoch": 1.6974047487575925,
"grad_norm": 0.9402440786361694,
"learning_rate": 4.135337794354518e-06,
"loss": 0.6266,
"step": 2050
},
{
"epoch": 1.6982330204307012,
"grad_norm": 0.8790574669837952,
"learning_rate": 4.134505855620902e-06,
"loss": 0.5989,
"step": 2051
},
{
"epoch": 1.69906129210381,
"grad_norm": 0.9421536326408386,
"learning_rate": 4.133673600625637e-06,
"loss": 0.611,
"step": 2052
},
{
"epoch": 1.6998895637769187,
"grad_norm": 0.9365546703338623,
"learning_rate": 4.132841029529762e-06,
"loss": 0.6201,
"step": 2053
},
{
"epoch": 1.7007178354500276,
"grad_norm": 0.9213638305664062,
"learning_rate": 4.132008142494367e-06,
"loss": 0.632,
"step": 2054
},
{
"epoch": 1.7015461071231364,
"grad_norm": 0.9204041957855225,
"learning_rate": 4.13117493968061e-06,
"loss": 0.6086,
"step": 2055
},
{
"epoch": 1.7023743787962453,
"grad_norm": 0.8972902297973633,
"learning_rate": 4.130341421249708e-06,
"loss": 0.656,
"step": 2056
},
{
"epoch": 1.703202650469354,
"grad_norm": 0.9163963794708252,
"learning_rate": 4.129507587362939e-06,
"loss": 0.623,
"step": 2057
},
{
"epoch": 1.7040309221424628,
"grad_norm": 0.9178527593612671,
"learning_rate": 4.128673438181642e-06,
"loss": 0.6327,
"step": 2058
},
{
"epoch": 1.7048591938155715,
"grad_norm": 0.8982755541801453,
"learning_rate": 4.127838973867217e-06,
"loss": 0.6165,
"step": 2059
},
{
"epoch": 1.7056874654886802,
"grad_norm": 0.8995897769927979,
"learning_rate": 4.127004194581125e-06,
"loss": 0.6204,
"step": 2060
},
{
"epoch": 1.706515737161789,
"grad_norm": 0.9148748517036438,
"learning_rate": 4.126169100484888e-06,
"loss": 0.6177,
"step": 2061
},
{
"epoch": 1.707344008834898,
"grad_norm": 0.9249931573867798,
"learning_rate": 4.12533369174009e-06,
"loss": 0.6113,
"step": 2062
},
{
"epoch": 1.7081722805080066,
"grad_norm": 0.9162423610687256,
"learning_rate": 4.124497968508372e-06,
"loss": 0.6119,
"step": 2063
},
{
"epoch": 1.7090005521811154,
"grad_norm": 0.8919503688812256,
"learning_rate": 4.123661930951441e-06,
"loss": 0.6194,
"step": 2064
},
{
"epoch": 1.7098288238542243,
"grad_norm": 0.8884986639022827,
"learning_rate": 4.122825579231063e-06,
"loss": 0.656,
"step": 2065
},
{
"epoch": 1.710657095527333,
"grad_norm": 0.9072293043136597,
"learning_rate": 4.121988913509062e-06,
"loss": 0.6351,
"step": 2066
},
{
"epoch": 1.7114853672004418,
"grad_norm": 0.9255547523498535,
"learning_rate": 4.121151933947327e-06,
"loss": 0.601,
"step": 2067
},
{
"epoch": 1.7123136388735505,
"grad_norm": 0.9141303300857544,
"learning_rate": 4.120314640707804e-06,
"loss": 0.6257,
"step": 2068
},
{
"epoch": 1.7131419105466592,
"grad_norm": 0.8791571855545044,
"learning_rate": 4.119477033952504e-06,
"loss": 0.6355,
"step": 2069
},
{
"epoch": 1.713970182219768,
"grad_norm": 0.8910380005836487,
"learning_rate": 4.118639113843493e-06,
"loss": 0.6405,
"step": 2070
},
{
"epoch": 1.7147984538928769,
"grad_norm": 0.9583408236503601,
"learning_rate": 4.117800880542904e-06,
"loss": 0.6408,
"step": 2071
},
{
"epoch": 1.7156267255659856,
"grad_norm": 0.9684842824935913,
"learning_rate": 4.116962334212924e-06,
"loss": 0.6155,
"step": 2072
},
{
"epoch": 1.7164549972390946,
"grad_norm": 0.8882633447647095,
"learning_rate": 4.116123475015806e-06,
"loss": 0.6332,
"step": 2073
},
{
"epoch": 1.7172832689122033,
"grad_norm": 0.8599907755851746,
"learning_rate": 4.1152843031138604e-06,
"loss": 0.6416,
"step": 2074
},
{
"epoch": 1.718111540585312,
"grad_norm": 0.9216133952140808,
"learning_rate": 4.114444818669461e-06,
"loss": 0.61,
"step": 2075
},
{
"epoch": 1.7189398122584207,
"grad_norm": 0.943729817867279,
"learning_rate": 4.113605021845037e-06,
"loss": 0.643,
"step": 2076
},
{
"epoch": 1.7197680839315295,
"grad_norm": 0.9158332943916321,
"learning_rate": 4.112764912803084e-06,
"loss": 0.6353,
"step": 2077
},
{
"epoch": 1.7205963556046382,
"grad_norm": 0.9036465287208557,
"learning_rate": 4.111924491706154e-06,
"loss": 0.6143,
"step": 2078
},
{
"epoch": 1.7214246272777471,
"grad_norm": 0.9264538884162903,
"learning_rate": 4.111083758716862e-06,
"loss": 0.6302,
"step": 2079
},
{
"epoch": 1.7222528989508559,
"grad_norm": 0.9309650659561157,
"learning_rate": 4.11024271399788e-06,
"loss": 0.6394,
"step": 2080
},
{
"epoch": 1.7230811706239648,
"grad_norm": 0.9261552095413208,
"learning_rate": 4.109401357711944e-06,
"loss": 0.6338,
"step": 2081
},
{
"epoch": 1.7239094422970735,
"grad_norm": 0.9027671217918396,
"learning_rate": 4.1085596900218476e-06,
"loss": 0.633,
"step": 2082
},
{
"epoch": 1.7247377139701823,
"grad_norm": 0.8899479508399963,
"learning_rate": 4.107717711090446e-06,
"loss": 0.6372,
"step": 2083
},
{
"epoch": 1.725565985643291,
"grad_norm": 0.9513740539550781,
"learning_rate": 4.106875421080654e-06,
"loss": 0.6057,
"step": 2084
},
{
"epoch": 1.7263942573163997,
"grad_norm": 0.8926359415054321,
"learning_rate": 4.106032820155447e-06,
"loss": 0.6219,
"step": 2085
},
{
"epoch": 1.7272225289895085,
"grad_norm": 0.8961042165756226,
"learning_rate": 4.105189908477862e-06,
"loss": 0.5926,
"step": 2086
},
{
"epoch": 1.7280508006626172,
"grad_norm": 0.8870651125907898,
"learning_rate": 4.104346686210992e-06,
"loss": 0.6298,
"step": 2087
},
{
"epoch": 1.7288790723357261,
"grad_norm": 0.9134538173675537,
"learning_rate": 4.103503153517993e-06,
"loss": 0.6544,
"step": 2088
},
{
"epoch": 1.7297073440088349,
"grad_norm": 0.900499165058136,
"learning_rate": 4.102659310562082e-06,
"loss": 0.6182,
"step": 2089
},
{
"epoch": 1.7305356156819438,
"grad_norm": 0.9146167039871216,
"learning_rate": 4.101815157506534e-06,
"loss": 0.6297,
"step": 2090
},
{
"epoch": 1.7313638873550525,
"grad_norm": 0.909670352935791,
"learning_rate": 4.100970694514687e-06,
"loss": 0.6272,
"step": 2091
},
{
"epoch": 1.7321921590281613,
"grad_norm": 1.127607822418213,
"learning_rate": 4.100125921749932e-06,
"loss": 0.6416,
"step": 2092
},
{
"epoch": 1.73302043070127,
"grad_norm": 0.8950872421264648,
"learning_rate": 4.099280839375729e-06,
"loss": 0.6411,
"step": 2093
},
{
"epoch": 1.7338487023743787,
"grad_norm": 0.9612913727760315,
"learning_rate": 4.098435447555591e-06,
"loss": 0.626,
"step": 2094
},
{
"epoch": 1.7346769740474874,
"grad_norm": 0.927638053894043,
"learning_rate": 4.097589746453094e-06,
"loss": 0.6197,
"step": 2095
},
{
"epoch": 1.7355052457205964,
"grad_norm": 0.9428266286849976,
"learning_rate": 4.096743736231875e-06,
"loss": 0.6407,
"step": 2096
},
{
"epoch": 1.7363335173937051,
"grad_norm": 0.8619174361228943,
"learning_rate": 4.095897417055627e-06,
"loss": 0.6386,
"step": 2097
},
{
"epoch": 1.737161789066814,
"grad_norm": 0.8943668007850647,
"learning_rate": 4.095050789088106e-06,
"loss": 0.6206,
"step": 2098
},
{
"epoch": 1.7379900607399228,
"grad_norm": 0.9190096259117126,
"learning_rate": 4.0942038524931245e-06,
"loss": 0.6395,
"step": 2099
},
{
"epoch": 1.7388183324130315,
"grad_norm": 0.9055671095848083,
"learning_rate": 4.09335660743456e-06,
"loss": 0.6384,
"step": 2100
},
{
"epoch": 1.7396466040861402,
"grad_norm": 0.9204139709472656,
"learning_rate": 4.092509054076345e-06,
"loss": 0.6438,
"step": 2101
},
{
"epoch": 1.740474875759249,
"grad_norm": 0.9572716355323792,
"learning_rate": 4.091661192582473e-06,
"loss": 0.6233,
"step": 2102
},
{
"epoch": 1.7413031474323577,
"grad_norm": 0.9479691982269287,
"learning_rate": 4.0908130231169975e-06,
"loss": 0.6284,
"step": 2103
},
{
"epoch": 1.7421314191054664,
"grad_norm": 0.9081270694732666,
"learning_rate": 4.089964545844032e-06,
"loss": 0.6284,
"step": 2104
},
{
"epoch": 1.7429596907785754,
"grad_norm": 0.9227046966552734,
"learning_rate": 4.0891157609277475e-06,
"loss": 0.6611,
"step": 2105
},
{
"epoch": 1.743787962451684,
"grad_norm": 0.9033496975898743,
"learning_rate": 4.088266668532377e-06,
"loss": 0.6122,
"step": 2106
},
{
"epoch": 1.744616234124793,
"grad_norm": 0.933539628982544,
"learning_rate": 4.087417268822211e-06,
"loss": 0.6167,
"step": 2107
},
{
"epoch": 1.7454445057979018,
"grad_norm": 0.9211486577987671,
"learning_rate": 4.086567561961602e-06,
"loss": 0.6364,
"step": 2108
},
{
"epoch": 1.7462727774710105,
"grad_norm": 0.8782284259796143,
"learning_rate": 4.08571754811496e-06,
"loss": 0.6164,
"step": 2109
},
{
"epoch": 1.7471010491441192,
"grad_norm": 0.8936499357223511,
"learning_rate": 4.084867227446753e-06,
"loss": 0.6155,
"step": 2110
},
{
"epoch": 1.747929320817228,
"grad_norm": 0.8931904435157776,
"learning_rate": 4.084016600121513e-06,
"loss": 0.6512,
"step": 2111
},
{
"epoch": 1.7487575924903367,
"grad_norm": 0.9070253372192383,
"learning_rate": 4.083165666303826e-06,
"loss": 0.6197,
"step": 2112
},
{
"epoch": 1.7495858641634456,
"grad_norm": 0.9232584238052368,
"learning_rate": 4.08231442615834e-06,
"loss": 0.6332,
"step": 2113
},
{
"epoch": 1.7504141358365544,
"grad_norm": 0.92015540599823,
"learning_rate": 4.081462879849765e-06,
"loss": 0.6351,
"step": 2114
},
{
"epoch": 1.7512424075096633,
"grad_norm": 0.9353263974189758,
"learning_rate": 4.080611027542863e-06,
"loss": 0.6479,
"step": 2115
},
{
"epoch": 1.752070679182772,
"grad_norm": 0.8943057060241699,
"learning_rate": 4.079758869402462e-06,
"loss": 0.6179,
"step": 2116
},
{
"epoch": 1.7528989508558808,
"grad_norm": 0.9062668681144714,
"learning_rate": 4.078906405593446e-06,
"loss": 0.6166,
"step": 2117
},
{
"epoch": 1.7537272225289895,
"grad_norm": 0.9368826150894165,
"learning_rate": 4.078053636280759e-06,
"loss": 0.6206,
"step": 2118
},
{
"epoch": 1.7545554942020982,
"grad_norm": 0.9012559652328491,
"learning_rate": 4.0772005616294034e-06,
"loss": 0.6445,
"step": 2119
},
{
"epoch": 1.755383765875207,
"grad_norm": 0.9061569571495056,
"learning_rate": 4.076347181804442e-06,
"loss": 0.6312,
"step": 2120
},
{
"epoch": 1.756212037548316,
"grad_norm": 0.892900288105011,
"learning_rate": 4.075493496970996e-06,
"loss": 0.64,
"step": 2121
},
{
"epoch": 1.7570403092214246,
"grad_norm": 0.9200263023376465,
"learning_rate": 4.074639507294243e-06,
"loss": 0.6435,
"step": 2122
},
{
"epoch": 1.7578685808945336,
"grad_norm": 0.921215295791626,
"learning_rate": 4.0737852129394265e-06,
"loss": 0.6135,
"step": 2123
},
{
"epoch": 1.7586968525676423,
"grad_norm": 0.8980262279510498,
"learning_rate": 4.07293061407184e-06,
"loss": 0.635,
"step": 2124
},
{
"epoch": 1.759525124240751,
"grad_norm": 0.9192932844161987,
"learning_rate": 4.072075710856843e-06,
"loss": 0.6317,
"step": 2125
},
{
"epoch": 1.7603533959138598,
"grad_norm": 0.9387120008468628,
"learning_rate": 4.07122050345985e-06,
"loss": 0.6324,
"step": 2126
},
{
"epoch": 1.7611816675869685,
"grad_norm": 0.8787287473678589,
"learning_rate": 4.070364992046338e-06,
"loss": 0.6177,
"step": 2127
},
{
"epoch": 1.7620099392600772,
"grad_norm": 0.8963474631309509,
"learning_rate": 4.069509176781838e-06,
"loss": 0.6158,
"step": 2128
},
{
"epoch": 1.762838210933186,
"grad_norm": 0.8926964402198792,
"learning_rate": 4.068653057831942e-06,
"loss": 0.6528,
"step": 2129
},
{
"epoch": 1.7636664826062949,
"grad_norm": 0.9290505647659302,
"learning_rate": 4.067796635362303e-06,
"loss": 0.6274,
"step": 2130
},
{
"epoch": 1.7644947542794036,
"grad_norm": 0.8848921060562134,
"learning_rate": 4.06693990953863e-06,
"loss": 0.6507,
"step": 2131
},
{
"epoch": 1.7653230259525126,
"grad_norm": 0.8977338075637817,
"learning_rate": 4.0660828805266915e-06,
"loss": 0.6437,
"step": 2132
},
{
"epoch": 1.7661512976256213,
"grad_norm": 0.9125422239303589,
"learning_rate": 4.065225548492314e-06,
"loss": 0.6221,
"step": 2133
},
{
"epoch": 1.76697956929873,
"grad_norm": 0.9186370968818665,
"learning_rate": 4.064367913601384e-06,
"loss": 0.6174,
"step": 2134
},
{
"epoch": 1.7678078409718387,
"grad_norm": 0.9084877967834473,
"learning_rate": 4.063509976019845e-06,
"loss": 0.6228,
"step": 2135
},
{
"epoch": 1.7686361126449475,
"grad_norm": 0.9180804491043091,
"learning_rate": 4.0626517359137e-06,
"loss": 0.6126,
"step": 2136
},
{
"epoch": 1.7694643843180562,
"grad_norm": 0.9099741578102112,
"learning_rate": 4.061793193449013e-06,
"loss": 0.6027,
"step": 2137
},
{
"epoch": 1.7702926559911651,
"grad_norm": 0.9060433506965637,
"learning_rate": 4.0609343487919005e-06,
"loss": 0.632,
"step": 2138
},
{
"epoch": 1.7711209276642739,
"grad_norm": 0.9404996037483215,
"learning_rate": 4.060075202108542e-06,
"loss": 0.6405,
"step": 2139
},
{
"epoch": 1.7719491993373828,
"grad_norm": 0.9031956791877747,
"learning_rate": 4.059215753565176e-06,
"loss": 0.6137,
"step": 2140
},
{
"epoch": 1.7727774710104915,
"grad_norm": 0.8913756012916565,
"learning_rate": 4.058356003328095e-06,
"loss": 0.6387,
"step": 2141
},
{
"epoch": 1.7736057426836003,
"grad_norm": 0.8986163139343262,
"learning_rate": 4.057495951563655e-06,
"loss": 0.6151,
"step": 2142
},
{
"epoch": 1.774434014356709,
"grad_norm": 0.9017037153244019,
"learning_rate": 4.0566355984382676e-06,
"loss": 0.6466,
"step": 2143
},
{
"epoch": 1.7752622860298177,
"grad_norm": 0.9211136698722839,
"learning_rate": 4.0557749441184024e-06,
"loss": 0.638,
"step": 2144
},
{
"epoch": 1.7760905577029265,
"grad_norm": 0.9097506999969482,
"learning_rate": 4.054913988770588e-06,
"loss": 0.6529,
"step": 2145
},
{
"epoch": 1.7769188293760352,
"grad_norm": 0.8985908627510071,
"learning_rate": 4.054052732561411e-06,
"loss": 0.6388,
"step": 2146
},
{
"epoch": 1.7777471010491441,
"grad_norm": 0.9066522121429443,
"learning_rate": 4.053191175657518e-06,
"loss": 0.6429,
"step": 2147
},
{
"epoch": 1.7785753727222529,
"grad_norm": 0.9080630540847778,
"learning_rate": 4.052329318225611e-06,
"loss": 0.619,
"step": 2148
},
{
"epoch": 1.7794036443953618,
"grad_norm": 0.8701719641685486,
"learning_rate": 4.0514671604324505e-06,
"loss": 0.6245,
"step": 2149
},
{
"epoch": 1.7802319160684705,
"grad_norm": 0.9059606194496155,
"learning_rate": 4.0506047024448585e-06,
"loss": 0.6124,
"step": 2150
},
{
"epoch": 1.7810601877415793,
"grad_norm": 0.8912079334259033,
"learning_rate": 4.04974194442971e-06,
"loss": 0.6295,
"step": 2151
},
{
"epoch": 1.781888459414688,
"grad_norm": 0.9063137769699097,
"learning_rate": 4.048878886553942e-06,
"loss": 0.6299,
"step": 2152
},
{
"epoch": 1.7827167310877967,
"grad_norm": 0.8850398659706116,
"learning_rate": 4.048015528984548e-06,
"loss": 0.6251,
"step": 2153
},
{
"epoch": 1.7835450027609054,
"grad_norm": 0.9059123992919922,
"learning_rate": 4.04715187188858e-06,
"loss": 0.6278,
"step": 2154
},
{
"epoch": 1.7843732744340144,
"grad_norm": 0.8777562379837036,
"learning_rate": 4.046287915433147e-06,
"loss": 0.6003,
"step": 2155
},
{
"epoch": 1.7852015461071231,
"grad_norm": 0.9048421382904053,
"learning_rate": 4.045423659785417e-06,
"loss": 0.6143,
"step": 2156
},
{
"epoch": 1.786029817780232,
"grad_norm": 0.9388517737388611,
"learning_rate": 4.044559105112614e-06,
"loss": 0.6314,
"step": 2157
},
{
"epoch": 1.7868580894533408,
"grad_norm": 0.9405019283294678,
"learning_rate": 4.043694251582023e-06,
"loss": 0.6381,
"step": 2158
},
{
"epoch": 1.7876863611264495,
"grad_norm": 0.8942162990570068,
"learning_rate": 4.0428290993609845e-06,
"loss": 0.6318,
"step": 2159
},
{
"epoch": 1.7885146327995582,
"grad_norm": 0.9065560698509216,
"learning_rate": 4.041963648616897e-06,
"loss": 0.6258,
"step": 2160
},
{
"epoch": 1.789342904472667,
"grad_norm": 0.8766914010047913,
"learning_rate": 4.041097899517218e-06,
"loss": 0.6292,
"step": 2161
},
{
"epoch": 1.7901711761457757,
"grad_norm": 0.8942935466766357,
"learning_rate": 4.040231852229462e-06,
"loss": 0.662,
"step": 2162
},
{
"epoch": 1.7909994478188846,
"grad_norm": 0.8892539143562317,
"learning_rate": 4.039365506921198e-06,
"loss": 0.6251,
"step": 2163
},
{
"epoch": 1.7918277194919934,
"grad_norm": 0.9359427690505981,
"learning_rate": 4.038498863760061e-06,
"loss": 0.6222,
"step": 2164
},
{
"epoch": 1.792655991165102,
"grad_norm": 0.8899942636489868,
"learning_rate": 4.037631922913733e-06,
"loss": 0.6318,
"step": 2165
},
{
"epoch": 1.793484262838211,
"grad_norm": 0.9233553409576416,
"learning_rate": 4.0367646845499635e-06,
"loss": 0.6188,
"step": 2166
},
{
"epoch": 1.7943125345113198,
"grad_norm": 0.8946877121925354,
"learning_rate": 4.035897148836553e-06,
"loss": 0.6247,
"step": 2167
},
{
"epoch": 1.7951408061844285,
"grad_norm": 0.9200116991996765,
"learning_rate": 4.03502931594136e-06,
"loss": 0.5981,
"step": 2168
},
{
"epoch": 1.7959690778575372,
"grad_norm": 0.9193991422653198,
"learning_rate": 4.034161186032306e-06,
"loss": 0.6144,
"step": 2169
},
{
"epoch": 1.796797349530646,
"grad_norm": 0.9637252688407898,
"learning_rate": 4.0332927592773615e-06,
"loss": 0.6094,
"step": 2170
},
{
"epoch": 1.7976256212037547,
"grad_norm": 0.8958486318588257,
"learning_rate": 4.032424035844564e-06,
"loss": 0.6304,
"step": 2171
},
{
"epoch": 1.7984538928768636,
"grad_norm": 0.9411646127700806,
"learning_rate": 4.031555015901998e-06,
"loss": 0.5984,
"step": 2172
},
{
"epoch": 1.7992821645499724,
"grad_norm": 0.8594862222671509,
"learning_rate": 4.030685699617815e-06,
"loss": 0.6176,
"step": 2173
},
{
"epoch": 1.8001104362230813,
"grad_norm": 0.8969242572784424,
"learning_rate": 4.029816087160217e-06,
"loss": 0.6294,
"step": 2174
},
{
"epoch": 1.80093870789619,
"grad_norm": 0.8950173854827881,
"learning_rate": 4.028946178697467e-06,
"loss": 0.6441,
"step": 2175
},
{
"epoch": 1.8017669795692988,
"grad_norm": 0.9084720611572266,
"learning_rate": 4.0280759743978835e-06,
"loss": 0.6245,
"step": 2176
},
{
"epoch": 1.8025952512424075,
"grad_norm": 0.9174644947052002,
"learning_rate": 4.027205474429845e-06,
"loss": 0.6136,
"step": 2177
},
{
"epoch": 1.8034235229155162,
"grad_norm": 0.9103249907493591,
"learning_rate": 4.026334678961783e-06,
"loss": 0.6341,
"step": 2178
},
{
"epoch": 1.804251794588625,
"grad_norm": 0.8916360139846802,
"learning_rate": 4.025463588162189e-06,
"loss": 0.6424,
"step": 2179
},
{
"epoch": 1.805080066261734,
"grad_norm": 0.9347325563430786,
"learning_rate": 4.024592202199611e-06,
"loss": 0.6425,
"step": 2180
},
{
"epoch": 1.8059083379348426,
"grad_norm": 0.9135183691978455,
"learning_rate": 4.023720521242653e-06,
"loss": 0.6265,
"step": 2181
},
{
"epoch": 1.8067366096079516,
"grad_norm": 0.8996950387954712,
"learning_rate": 4.022848545459979e-06,
"loss": 0.6083,
"step": 2182
},
{
"epoch": 1.8075648812810603,
"grad_norm": 0.9104750752449036,
"learning_rate": 4.021976275020307e-06,
"loss": 0.6087,
"step": 2183
},
{
"epoch": 1.808393152954169,
"grad_norm": 0.9272193908691406,
"learning_rate": 4.0211037100924135e-06,
"loss": 0.6254,
"step": 2184
},
{
"epoch": 1.8092214246272778,
"grad_norm": 0.8605275750160217,
"learning_rate": 4.020230850845132e-06,
"loss": 0.6244,
"step": 2185
},
{
"epoch": 1.8100496963003865,
"grad_norm": 0.8817747831344604,
"learning_rate": 4.019357697447352e-06,
"loss": 0.6226,
"step": 2186
},
{
"epoch": 1.8108779679734952,
"grad_norm": 0.8978784084320068,
"learning_rate": 4.018484250068021e-06,
"loss": 0.615,
"step": 2187
},
{
"epoch": 1.811706239646604,
"grad_norm": 0.8996117115020752,
"learning_rate": 4.017610508876143e-06,
"loss": 0.6248,
"step": 2188
},
{
"epoch": 1.8125345113197129,
"grad_norm": 0.9133120179176331,
"learning_rate": 4.016736474040778e-06,
"loss": 0.6364,
"step": 2189
},
{
"epoch": 1.8133627829928216,
"grad_norm": 0.903982400894165,
"learning_rate": 4.015862145731044e-06,
"loss": 0.6078,
"step": 2190
},
{
"epoch": 1.8141910546659306,
"grad_norm": 0.9343582987785339,
"learning_rate": 4.014987524116115e-06,
"loss": 0.6254,
"step": 2191
},
{
"epoch": 1.8150193263390393,
"grad_norm": 0.883305549621582,
"learning_rate": 4.014112609365224e-06,
"loss": 0.6306,
"step": 2192
},
{
"epoch": 1.815847598012148,
"grad_norm": 0.9044662714004517,
"learning_rate": 4.013237401647656e-06,
"loss": 0.6288,
"step": 2193
},
{
"epoch": 1.8166758696852567,
"grad_norm": 0.9016374349594116,
"learning_rate": 4.012361901132758e-06,
"loss": 0.6254,
"step": 2194
},
{
"epoch": 1.8175041413583655,
"grad_norm": 0.9198315739631653,
"learning_rate": 4.01148610798993e-06,
"loss": 0.621,
"step": 2195
},
{
"epoch": 1.8183324130314742,
"grad_norm": 0.8884013891220093,
"learning_rate": 4.01061002238863e-06,
"loss": 0.6195,
"step": 2196
},
{
"epoch": 1.8191606847045831,
"grad_norm": 0.901606023311615,
"learning_rate": 4.009733644498372e-06,
"loss": 0.646,
"step": 2197
},
{
"epoch": 1.8199889563776919,
"grad_norm": 0.9003238677978516,
"learning_rate": 4.008856974488728e-06,
"loss": 0.6276,
"step": 2198
},
{
"epoch": 1.8208172280508008,
"grad_norm": 0.8835832476615906,
"learning_rate": 4.007980012529324e-06,
"loss": 0.6232,
"step": 2199
},
{
"epoch": 1.8216454997239095,
"grad_norm": 0.8751473426818848,
"learning_rate": 4.007102758789846e-06,
"loss": 0.635,
"step": 2200
},
{
"epoch": 1.8224737713970183,
"grad_norm": 0.8990851640701294,
"learning_rate": 4.006225213440032e-06,
"loss": 0.605,
"step": 2201
},
{
"epoch": 1.823302043070127,
"grad_norm": 0.8766496181488037,
"learning_rate": 4.0053473766496805e-06,
"loss": 0.6396,
"step": 2202
},
{
"epoch": 1.8241303147432357,
"grad_norm": 0.914273202419281,
"learning_rate": 4.004469248588644e-06,
"loss": 0.6435,
"step": 2203
},
{
"epoch": 1.8249585864163445,
"grad_norm": 0.9063234925270081,
"learning_rate": 4.003590829426832e-06,
"loss": 0.6144,
"step": 2204
},
{
"epoch": 1.8257868580894532,
"grad_norm": 0.9479252696037292,
"learning_rate": 4.002712119334212e-06,
"loss": 0.6584,
"step": 2205
},
{
"epoch": 1.8266151297625621,
"grad_norm": 0.9253218770027161,
"learning_rate": 4.001833118480804e-06,
"loss": 0.6203,
"step": 2206
},
{
"epoch": 1.8274434014356709,
"grad_norm": 0.930874764919281,
"learning_rate": 4.000953827036689e-06,
"loss": 0.6254,
"step": 2207
},
{
"epoch": 1.8282716731087798,
"grad_norm": 0.9202263951301575,
"learning_rate": 4.000074245171999e-06,
"loss": 0.6351,
"step": 2208
},
{
"epoch": 1.8290999447818885,
"grad_norm": 0.8993806838989258,
"learning_rate": 3.999194373056926e-06,
"loss": 0.6439,
"step": 2209
},
{
"epoch": 1.8299282164549973,
"grad_norm": 0.9138032793998718,
"learning_rate": 3.998314210861718e-06,
"loss": 0.6011,
"step": 2210
},
{
"epoch": 1.830756488128106,
"grad_norm": 0.8995500206947327,
"learning_rate": 3.9974337587566765e-06,
"loss": 0.6285,
"step": 2211
},
{
"epoch": 1.8315847598012147,
"grad_norm": 0.9092652797698975,
"learning_rate": 3.996553016912161e-06,
"loss": 0.6364,
"step": 2212
},
{
"epoch": 1.8324130314743234,
"grad_norm": 0.9085352420806885,
"learning_rate": 3.995671985498589e-06,
"loss": 0.6245,
"step": 2213
},
{
"epoch": 1.8332413031474324,
"grad_norm": 0.9546533226966858,
"learning_rate": 3.994790664686429e-06,
"loss": 0.6512,
"step": 2214
},
{
"epoch": 1.8340695748205411,
"grad_norm": 0.8606089353561401,
"learning_rate": 3.99390905464621e-06,
"loss": 0.6135,
"step": 2215
},
{
"epoch": 1.83489784649365,
"grad_norm": 0.9450663328170776,
"learning_rate": 3.993027155548515e-06,
"loss": 0.6193,
"step": 2216
},
{
"epoch": 1.8357261181667588,
"grad_norm": 0.9496508836746216,
"learning_rate": 3.992144967563984e-06,
"loss": 0.5978,
"step": 2217
},
{
"epoch": 1.8365543898398675,
"grad_norm": 0.8837412595748901,
"learning_rate": 3.99126249086331e-06,
"loss": 0.6104,
"step": 2218
},
{
"epoch": 1.8373826615129762,
"grad_norm": 0.8774031400680542,
"learning_rate": 3.9903797256172464e-06,
"loss": 0.6302,
"step": 2219
},
{
"epoch": 1.838210933186085,
"grad_norm": 0.8978205919265747,
"learning_rate": 3.989496671996599e-06,
"loss": 0.6411,
"step": 2220
},
{
"epoch": 1.8390392048591937,
"grad_norm": 0.9087855815887451,
"learning_rate": 3.98861333017223e-06,
"loss": 0.592,
"step": 2221
},
{
"epoch": 1.8398674765323026,
"grad_norm": 0.9040684700012207,
"learning_rate": 3.987729700315059e-06,
"loss": 0.6121,
"step": 2222
},
{
"epoch": 1.8406957482054114,
"grad_norm": 0.8980532288551331,
"learning_rate": 3.986845782596058e-06,
"loss": 0.6231,
"step": 2223
},
{
"epoch": 1.84152401987852,
"grad_norm": 0.9104854464530945,
"learning_rate": 3.9859615771862595e-06,
"loss": 0.6228,
"step": 2224
},
{
"epoch": 1.842352291551629,
"grad_norm": 0.9314625263214111,
"learning_rate": 3.985077084256747e-06,
"loss": 0.6136,
"step": 2225
},
{
"epoch": 1.8431805632247378,
"grad_norm": 0.9033634662628174,
"learning_rate": 3.984192303978663e-06,
"loss": 0.6408,
"step": 2226
},
{
"epoch": 1.8440088348978465,
"grad_norm": 0.9022733569145203,
"learning_rate": 3.983307236523203e-06,
"loss": 0.5954,
"step": 2227
},
{
"epoch": 1.8448371065709552,
"grad_norm": 0.8900256156921387,
"learning_rate": 3.982421882061621e-06,
"loss": 0.631,
"step": 2228
},
{
"epoch": 1.845665378244064,
"grad_norm": 0.8882872462272644,
"learning_rate": 3.981536240765223e-06,
"loss": 0.6204,
"step": 2229
},
{
"epoch": 1.8464936499171727,
"grad_norm": 0.894923985004425,
"learning_rate": 3.980650312805373e-06,
"loss": 0.6013,
"step": 2230
},
{
"epoch": 1.8473219215902816,
"grad_norm": 0.8910638093948364,
"learning_rate": 3.979764098353491e-06,
"loss": 0.6373,
"step": 2231
},
{
"epoch": 1.8481501932633904,
"grad_norm": 0.8952053189277649,
"learning_rate": 3.978877597581049e-06,
"loss": 0.6374,
"step": 2232
},
{
"epoch": 1.8489784649364993,
"grad_norm": 0.9100720882415771,
"learning_rate": 3.977990810659579e-06,
"loss": 0.6417,
"step": 2233
},
{
"epoch": 1.849806736609608,
"grad_norm": 0.8954150676727295,
"learning_rate": 3.9771037377606634e-06,
"loss": 0.6027,
"step": 2234
},
{
"epoch": 1.8506350082827168,
"grad_norm": 0.8989905118942261,
"learning_rate": 3.976216379055944e-06,
"loss": 0.6228,
"step": 2235
},
{
"epoch": 1.8514632799558255,
"grad_norm": 0.9060820937156677,
"learning_rate": 3.975328734717117e-06,
"loss": 0.6291,
"step": 2236
},
{
"epoch": 1.8522915516289342,
"grad_norm": 0.9164367318153381,
"learning_rate": 3.974440804915932e-06,
"loss": 0.6249,
"step": 2237
},
{
"epoch": 1.853119823302043,
"grad_norm": 0.9227840304374695,
"learning_rate": 3.973552589824196e-06,
"loss": 0.6328,
"step": 2238
},
{
"epoch": 1.853948094975152,
"grad_norm": 0.8612478971481323,
"learning_rate": 3.972664089613772e-06,
"loss": 0.5794,
"step": 2239
},
{
"epoch": 1.8547763666482606,
"grad_norm": 0.8819253444671631,
"learning_rate": 3.971775304456573e-06,
"loss": 0.6145,
"step": 2240
},
{
"epoch": 1.8556046383213696,
"grad_norm": 0.9216543436050415,
"learning_rate": 3.970886234524573e-06,
"loss": 0.609,
"step": 2241
},
{
"epoch": 1.8564329099944783,
"grad_norm": 0.9114699959754944,
"learning_rate": 3.969996879989797e-06,
"loss": 0.6105,
"step": 2242
},
{
"epoch": 1.857261181667587,
"grad_norm": 0.9464704990386963,
"learning_rate": 3.969107241024328e-06,
"loss": 0.6389,
"step": 2243
},
{
"epoch": 1.8580894533406958,
"grad_norm": 0.8921800255775452,
"learning_rate": 3.968217317800303e-06,
"loss": 0.6338,
"step": 2244
},
{
"epoch": 1.8589177250138045,
"grad_norm": 0.897452175617218,
"learning_rate": 3.967327110489913e-06,
"loss": 0.6248,
"step": 2245
},
{
"epoch": 1.8597459966869132,
"grad_norm": 0.9093567132949829,
"learning_rate": 3.9664366192654066e-06,
"loss": 0.6258,
"step": 2246
},
{
"epoch": 1.860574268360022,
"grad_norm": 0.894188642501831,
"learning_rate": 3.965545844299084e-06,
"loss": 0.6095,
"step": 2247
},
{
"epoch": 1.8614025400331309,
"grad_norm": 0.9238452911376953,
"learning_rate": 3.964654785763301e-06,
"loss": 0.6237,
"step": 2248
},
{
"epoch": 1.8622308117062396,
"grad_norm": 0.8715214133262634,
"learning_rate": 3.9637634438304715e-06,
"loss": 0.6153,
"step": 2249
},
{
"epoch": 1.8630590833793486,
"grad_norm": 0.8790848255157471,
"learning_rate": 3.962871818673061e-06,
"loss": 0.6389,
"step": 2250
},
{
"epoch": 1.8638873550524573,
"grad_norm": 0.9859224557876587,
"learning_rate": 3.961979910463589e-06,
"loss": 0.6075,
"step": 2251
},
{
"epoch": 1.864715626725566,
"grad_norm": 0.9127727150917053,
"learning_rate": 3.9610877193746335e-06,
"loss": 0.6332,
"step": 2252
},
{
"epoch": 1.8655438983986747,
"grad_norm": 0.9248382449150085,
"learning_rate": 3.960195245578825e-06,
"loss": 0.622,
"step": 2253
},
{
"epoch": 1.8663721700717835,
"grad_norm": 0.9030913710594177,
"learning_rate": 3.959302489248848e-06,
"loss": 0.6326,
"step": 2254
},
{
"epoch": 1.8672004417448922,
"grad_norm": 0.8922805190086365,
"learning_rate": 3.958409450557441e-06,
"loss": 0.6099,
"step": 2255
},
{
"epoch": 1.8680287134180011,
"grad_norm": 0.9165307879447937,
"learning_rate": 3.957516129677402e-06,
"loss": 0.6245,
"step": 2256
},
{
"epoch": 1.8688569850911099,
"grad_norm": 0.9445623159408569,
"learning_rate": 3.956622526781579e-06,
"loss": 0.6317,
"step": 2257
},
{
"epoch": 1.8696852567642188,
"grad_norm": 0.8833973407745361,
"learning_rate": 3.955728642042874e-06,
"loss": 0.6264,
"step": 2258
},
{
"epoch": 1.8705135284373275,
"grad_norm": 0.9512470364570618,
"learning_rate": 3.9548344756342486e-06,
"loss": 0.6137,
"step": 2259
},
{
"epoch": 1.8713418001104363,
"grad_norm": 0.8834348320960999,
"learning_rate": 3.9539400277287134e-06,
"loss": 0.6597,
"step": 2260
},
{
"epoch": 1.872170071783545,
"grad_norm": 0.8881180882453918,
"learning_rate": 3.953045298499335e-06,
"loss": 0.6502,
"step": 2261
},
{
"epoch": 1.8729983434566537,
"grad_norm": 0.8921699523925781,
"learning_rate": 3.952150288119239e-06,
"loss": 0.6216,
"step": 2262
},
{
"epoch": 1.8738266151297625,
"grad_norm": 0.91098552942276,
"learning_rate": 3.9512549967615965e-06,
"loss": 0.6286,
"step": 2263
},
{
"epoch": 1.8746548868028712,
"grad_norm": 0.9343264698982239,
"learning_rate": 3.950359424599642e-06,
"loss": 0.6368,
"step": 2264
},
{
"epoch": 1.8754831584759801,
"grad_norm": 0.9168746471405029,
"learning_rate": 3.949463571806658e-06,
"loss": 0.6355,
"step": 2265
},
{
"epoch": 1.8763114301490889,
"grad_norm": 0.9206068515777588,
"learning_rate": 3.948567438555986e-06,
"loss": 0.607,
"step": 2266
},
{
"epoch": 1.8771397018221978,
"grad_norm": 0.9262584447860718,
"learning_rate": 3.947671025021017e-06,
"loss": 0.6197,
"step": 2267
},
{
"epoch": 1.8779679734953065,
"grad_norm": 0.8699168562889099,
"learning_rate": 3.946774331375199e-06,
"loss": 0.6181,
"step": 2268
},
{
"epoch": 1.8787962451684153,
"grad_norm": 0.8892862200737,
"learning_rate": 3.9458773577920365e-06,
"loss": 0.6245,
"step": 2269
},
{
"epoch": 1.879624516841524,
"grad_norm": 0.9038919806480408,
"learning_rate": 3.944980104445082e-06,
"loss": 0.6209,
"step": 2270
},
{
"epoch": 1.8804527885146327,
"grad_norm": 0.8780872821807861,
"learning_rate": 3.9440825715079475e-06,
"loss": 0.6355,
"step": 2271
},
{
"epoch": 1.8812810601877414,
"grad_norm": 0.9379895925521851,
"learning_rate": 3.943184759154297e-06,
"loss": 0.6381,
"step": 2272
},
{
"epoch": 1.8821093318608504,
"grad_norm": 0.9112582802772522,
"learning_rate": 3.942286667557849e-06,
"loss": 0.6273,
"step": 2273
},
{
"epoch": 1.8829376035339591,
"grad_norm": 0.9367797374725342,
"learning_rate": 3.9413882968923754e-06,
"loss": 0.6111,
"step": 2274
},
{
"epoch": 1.883765875207068,
"grad_norm": 0.8923424482345581,
"learning_rate": 3.940489647331703e-06,
"loss": 0.6187,
"step": 2275
},
{
"epoch": 1.8845941468801768,
"grad_norm": 0.8797968029975891,
"learning_rate": 3.939590719049712e-06,
"loss": 0.603,
"step": 2276
},
{
"epoch": 1.8854224185532855,
"grad_norm": 0.8977354168891907,
"learning_rate": 3.938691512220336e-06,
"loss": 0.614,
"step": 2277
},
{
"epoch": 1.8862506902263942,
"grad_norm": 0.9207090139389038,
"learning_rate": 3.937792027017564e-06,
"loss": 0.6162,
"step": 2278
},
{
"epoch": 1.887078961899503,
"grad_norm": 0.8815562725067139,
"learning_rate": 3.936892263615438e-06,
"loss": 0.6315,
"step": 2279
},
{
"epoch": 1.8879072335726117,
"grad_norm": 0.9983104467391968,
"learning_rate": 3.935992222188054e-06,
"loss": 0.6337,
"step": 2280
},
{
"epoch": 1.8887355052457206,
"grad_norm": 0.8987579941749573,
"learning_rate": 3.93509190290956e-06,
"loss": 0.644,
"step": 2281
},
{
"epoch": 1.8895637769188294,
"grad_norm": 0.9132242798805237,
"learning_rate": 3.934191305954161e-06,
"loss": 0.6063,
"step": 2282
},
{
"epoch": 1.890392048591938,
"grad_norm": 0.8940858840942383,
"learning_rate": 3.933290431496114e-06,
"loss": 0.6204,
"step": 2283
},
{
"epoch": 1.891220320265047,
"grad_norm": 0.8862229585647583,
"learning_rate": 3.932389279709729e-06,
"loss": 0.6257,
"step": 2284
},
{
"epoch": 1.8920485919381558,
"grad_norm": 0.8841557502746582,
"learning_rate": 3.931487850769371e-06,
"loss": 0.6329,
"step": 2285
},
{
"epoch": 1.8928768636112645,
"grad_norm": 0.9075404405593872,
"learning_rate": 3.930586144849459e-06,
"loss": 0.6195,
"step": 2286
},
{
"epoch": 1.8937051352843732,
"grad_norm": 0.9349889755249023,
"learning_rate": 3.929684162124462e-06,
"loss": 0.6159,
"step": 2287
},
{
"epoch": 1.894533406957482,
"grad_norm": 0.9378470182418823,
"learning_rate": 3.928781902768908e-06,
"loss": 0.6221,
"step": 2288
},
{
"epoch": 1.8953616786305907,
"grad_norm": 0.8877297043800354,
"learning_rate": 3.927879366957374e-06,
"loss": 0.6359,
"step": 2289
},
{
"epoch": 1.8961899503036996,
"grad_norm": 0.9335882067680359,
"learning_rate": 3.926976554864495e-06,
"loss": 0.6397,
"step": 2290
},
{
"epoch": 1.8970182219768084,
"grad_norm": 0.8982004523277283,
"learning_rate": 3.926073466664953e-06,
"loss": 0.6303,
"step": 2291
},
{
"epoch": 1.8978464936499173,
"grad_norm": 0.8909676671028137,
"learning_rate": 3.92517010253349e-06,
"loss": 0.6476,
"step": 2292
},
{
"epoch": 1.898674765323026,
"grad_norm": 0.9127813577651978,
"learning_rate": 3.924266462644896e-06,
"loss": 0.6078,
"step": 2293
},
{
"epoch": 1.8995030369961348,
"grad_norm": 0.8923580050468445,
"learning_rate": 3.923362547174019e-06,
"loss": 0.6371,
"step": 2294
},
{
"epoch": 1.9003313086692435,
"grad_norm": 0.912458062171936,
"learning_rate": 3.922458356295757e-06,
"loss": 0.6162,
"step": 2295
},
{
"epoch": 1.9011595803423522,
"grad_norm": 0.8866751790046692,
"learning_rate": 3.921553890185064e-06,
"loss": 0.6011,
"step": 2296
},
{
"epoch": 1.901987852015461,
"grad_norm": 0.9114490151405334,
"learning_rate": 3.920649149016944e-06,
"loss": 0.6322,
"step": 2297
},
{
"epoch": 1.90281612368857,
"grad_norm": 0.912030816078186,
"learning_rate": 3.919744132966457e-06,
"loss": 0.6293,
"step": 2298
},
{
"epoch": 1.9036443953616786,
"grad_norm": 0.887301504611969,
"learning_rate": 3.918838842208715e-06,
"loss": 0.6288,
"step": 2299
},
{
"epoch": 1.9044726670347876,
"grad_norm": 0.9039818048477173,
"learning_rate": 3.917933276918884e-06,
"loss": 0.6154,
"step": 2300
},
{
"epoch": 1.9053009387078963,
"grad_norm": 0.9260784983634949,
"learning_rate": 3.917027437272182e-06,
"loss": 0.6111,
"step": 2301
},
{
"epoch": 1.906129210381005,
"grad_norm": 0.8966069221496582,
"learning_rate": 3.916121323443879e-06,
"loss": 0.6334,
"step": 2302
},
{
"epoch": 1.9069574820541138,
"grad_norm": 0.8983698487281799,
"learning_rate": 3.9152149356093e-06,
"loss": 0.6235,
"step": 2303
},
{
"epoch": 1.9077857537272225,
"grad_norm": 0.9150595664978027,
"learning_rate": 3.914308273943826e-06,
"loss": 0.6118,
"step": 2304
},
{
"epoch": 1.9086140254003312,
"grad_norm": 0.8768419027328491,
"learning_rate": 3.913401338622883e-06,
"loss": 0.5935,
"step": 2305
},
{
"epoch": 1.90944229707344,
"grad_norm": 0.887562096118927,
"learning_rate": 3.912494129821958e-06,
"loss": 0.6232,
"step": 2306
},
{
"epoch": 1.9102705687465489,
"grad_norm": 0.8916515111923218,
"learning_rate": 3.911586647716585e-06,
"loss": 0.6409,
"step": 2307
},
{
"epoch": 1.9110988404196576,
"grad_norm": 0.9207882881164551,
"learning_rate": 3.910678892482354e-06,
"loss": 0.6155,
"step": 2308
},
{
"epoch": 1.9119271120927666,
"grad_norm": 0.9300670027732849,
"learning_rate": 3.9097708642949085e-06,
"loss": 0.6389,
"step": 2309
},
{
"epoch": 1.9127553837658753,
"grad_norm": 0.8685980439186096,
"learning_rate": 3.908862563329941e-06,
"loss": 0.6392,
"step": 2310
},
{
"epoch": 1.913583655438984,
"grad_norm": 0.9150967001914978,
"learning_rate": 3.907953989763202e-06,
"loss": 0.6387,
"step": 2311
},
{
"epoch": 1.9144119271120927,
"grad_norm": 0.9094190001487732,
"learning_rate": 3.90704514377049e-06,
"loss": 0.6055,
"step": 2312
},
{
"epoch": 1.9152401987852015,
"grad_norm": 0.8863343596458435,
"learning_rate": 3.90613602552766e-06,
"loss": 0.6073,
"step": 2313
},
{
"epoch": 1.9160684704583102,
"grad_norm": 0.9878994822502136,
"learning_rate": 3.905226635210616e-06,
"loss": 0.613,
"step": 2314
},
{
"epoch": 1.9168967421314191,
"grad_norm": 0.8863592147827148,
"learning_rate": 3.904316972995318e-06,
"loss": 0.6102,
"step": 2315
},
{
"epoch": 1.9177250138045279,
"grad_norm": 0.8859980702400208,
"learning_rate": 3.9034070390577765e-06,
"loss": 0.6182,
"step": 2316
},
{
"epoch": 1.9185532854776368,
"grad_norm": 0.8953589200973511,
"learning_rate": 3.9024968335740556e-06,
"loss": 0.6212,
"step": 2317
},
{
"epoch": 1.9193815571507455,
"grad_norm": 0.873466968536377,
"learning_rate": 3.90158635672027e-06,
"loss": 0.616,
"step": 2318
},
{
"epoch": 1.9202098288238543,
"grad_norm": 0.8589543700218201,
"learning_rate": 3.900675608672592e-06,
"loss": 0.6141,
"step": 2319
},
{
"epoch": 1.921038100496963,
"grad_norm": 0.8717133402824402,
"learning_rate": 3.899764589607239e-06,
"loss": 0.6281,
"step": 2320
},
{
"epoch": 1.9218663721700717,
"grad_norm": 0.893227219581604,
"learning_rate": 3.898853299700486e-06,
"loss": 0.6139,
"step": 2321
},
{
"epoch": 1.9226946438431805,
"grad_norm": 0.8736181259155273,
"learning_rate": 3.89794173912866e-06,
"loss": 0.5987,
"step": 2322
},
{
"epoch": 1.9235229155162892,
"grad_norm": 0.8975626826286316,
"learning_rate": 3.897029908068139e-06,
"loss": 0.6417,
"step": 2323
},
{
"epoch": 1.9243511871893981,
"grad_norm": 0.8916004300117493,
"learning_rate": 3.896117806695355e-06,
"loss": 0.6259,
"step": 2324
},
{
"epoch": 1.9251794588625069,
"grad_norm": 0.8739155530929565,
"learning_rate": 3.8952054351867885e-06,
"loss": 0.6252,
"step": 2325
},
{
"epoch": 1.9260077305356158,
"grad_norm": 0.8667280673980713,
"learning_rate": 3.894292793718976e-06,
"loss": 0.6083,
"step": 2326
},
{
"epoch": 1.9268360022087245,
"grad_norm": 0.8749732375144958,
"learning_rate": 3.893379882468506e-06,
"loss": 0.6294,
"step": 2327
},
{
"epoch": 1.9276642738818333,
"grad_norm": 0.9085330963134766,
"learning_rate": 3.8924667016120175e-06,
"loss": 0.6101,
"step": 2328
},
{
"epoch": 1.928492545554942,
"grad_norm": 0.9394292235374451,
"learning_rate": 3.891553251326203e-06,
"loss": 0.6182,
"step": 2329
},
{
"epoch": 1.9293208172280507,
"grad_norm": 0.8756951093673706,
"learning_rate": 3.890639531787805e-06,
"loss": 0.6159,
"step": 2330
},
{
"epoch": 1.9301490889011594,
"grad_norm": 0.8765501379966736,
"learning_rate": 3.8897255431736225e-06,
"loss": 0.6182,
"step": 2331
},
{
"epoch": 1.9309773605742684,
"grad_norm": 0.907835841178894,
"learning_rate": 3.888811285660501e-06,
"loss": 0.6336,
"step": 2332
},
{
"epoch": 1.9318056322473771,
"grad_norm": 1.0617926120758057,
"learning_rate": 3.887896759425342e-06,
"loss": 0.6336,
"step": 2333
},
{
"epoch": 1.932633903920486,
"grad_norm": 0.8745515942573547,
"learning_rate": 3.886981964645099e-06,
"loss": 0.6244,
"step": 2334
},
{
"epoch": 1.9334621755935948,
"grad_norm": 0.8918715119361877,
"learning_rate": 3.886066901496775e-06,
"loss": 0.6289,
"step": 2335
},
{
"epoch": 1.9342904472667035,
"grad_norm": 0.8779492378234863,
"learning_rate": 3.885151570157426e-06,
"loss": 0.6335,
"step": 2336
},
{
"epoch": 1.9351187189398122,
"grad_norm": 0.8955790400505066,
"learning_rate": 3.88423597080416e-06,
"loss": 0.6125,
"step": 2337
},
{
"epoch": 1.935946990612921,
"grad_norm": 0.8905174136161804,
"learning_rate": 3.883320103614137e-06,
"loss": 0.5913,
"step": 2338
},
{
"epoch": 1.9367752622860297,
"grad_norm": 0.9078717231750488,
"learning_rate": 3.88240396876457e-06,
"loss": 0.6356,
"step": 2339
},
{
"epoch": 1.9376035339591386,
"grad_norm": 0.9211397171020508,
"learning_rate": 3.881487566432722e-06,
"loss": 0.6305,
"step": 2340
},
{
"epoch": 1.9384318056322474,
"grad_norm": 0.9131335020065308,
"learning_rate": 3.880570896795908e-06,
"loss": 0.6184,
"step": 2341
},
{
"epoch": 1.9392600773053563,
"grad_norm": 0.8874300718307495,
"learning_rate": 3.879653960031494e-06,
"loss": 0.6191,
"step": 2342
},
{
"epoch": 1.940088348978465,
"grad_norm": 0.9244474768638611,
"learning_rate": 3.8787367563169e-06,
"loss": 0.5994,
"step": 2343
},
{
"epoch": 1.9409166206515738,
"grad_norm": 0.8986138701438904,
"learning_rate": 3.877819285829596e-06,
"loss": 0.6544,
"step": 2344
},
{
"epoch": 1.9417448923246825,
"grad_norm": 0.9084672927856445,
"learning_rate": 3.876901548747104e-06,
"loss": 0.6248,
"step": 2345
},
{
"epoch": 1.9425731639977912,
"grad_norm": 0.9052600264549255,
"learning_rate": 3.875983545246997e-06,
"loss": 0.6112,
"step": 2346
},
{
"epoch": 1.9434014356709,
"grad_norm": 0.9029438495635986,
"learning_rate": 3.8750652755069025e-06,
"loss": 0.6157,
"step": 2347
},
{
"epoch": 1.9442297073440087,
"grad_norm": 0.882449746131897,
"learning_rate": 3.874146739704494e-06,
"loss": 0.5886,
"step": 2348
},
{
"epoch": 1.9450579790171176,
"grad_norm": 0.8628272414207458,
"learning_rate": 3.873227938017502e-06,
"loss": 0.608,
"step": 2349
},
{
"epoch": 1.9458862506902264,
"grad_norm": 0.8977845907211304,
"learning_rate": 3.872308870623704e-06,
"loss": 0.6475,
"step": 2350
},
{
"epoch": 1.9467145223633353,
"grad_norm": 0.8801501393318176,
"learning_rate": 3.871389537700933e-06,
"loss": 0.6246,
"step": 2351
},
{
"epoch": 1.947542794036444,
"grad_norm": 0.9245358109474182,
"learning_rate": 3.87046993942707e-06,
"loss": 0.6317,
"step": 2352
},
{
"epoch": 1.9483710657095528,
"grad_norm": 0.8959249258041382,
"learning_rate": 3.869550075980049e-06,
"loss": 0.6157,
"step": 2353
},
{
"epoch": 1.9491993373826615,
"grad_norm": 0.861315906047821,
"learning_rate": 3.868629947537855e-06,
"loss": 0.6251,
"step": 2354
},
{
"epoch": 1.9500276090557702,
"grad_norm": 0.8839966654777527,
"learning_rate": 3.867709554278524e-06,
"loss": 0.6289,
"step": 2355
},
{
"epoch": 1.950855880728879,
"grad_norm": 0.8878759741783142,
"learning_rate": 3.8667888963801445e-06,
"loss": 0.6113,
"step": 2356
},
{
"epoch": 1.951684152401988,
"grad_norm": 0.8725037574768066,
"learning_rate": 3.865867974020854e-06,
"loss": 0.6324,
"step": 2357
},
{
"epoch": 1.9525124240750966,
"grad_norm": 0.9059397578239441,
"learning_rate": 3.8649467873788436e-06,
"loss": 0.5981,
"step": 2358
},
{
"epoch": 1.9533406957482056,
"grad_norm": 0.8883296847343445,
"learning_rate": 3.8640253366323535e-06,
"loss": 0.598,
"step": 2359
},
{
"epoch": 1.9541689674213143,
"grad_norm": 0.8987225890159607,
"learning_rate": 3.863103621959676e-06,
"loss": 0.6253,
"step": 2360
},
{
"epoch": 1.954997239094423,
"grad_norm": 0.8977320194244385,
"learning_rate": 3.862181643539155e-06,
"loss": 0.6424,
"step": 2361
},
{
"epoch": 1.9558255107675317,
"grad_norm": 0.9267589449882507,
"learning_rate": 3.861259401549183e-06,
"loss": 0.647,
"step": 2362
},
{
"epoch": 1.9566537824406405,
"grad_norm": 0.9297959804534912,
"learning_rate": 3.860336896168207e-06,
"loss": 0.6448,
"step": 2363
},
{
"epoch": 1.9574820541137492,
"grad_norm": 0.9627453088760376,
"learning_rate": 3.859414127574722e-06,
"loss": 0.6154,
"step": 2364
},
{
"epoch": 1.958310325786858,
"grad_norm": 0.8945866227149963,
"learning_rate": 3.858491095947275e-06,
"loss": 0.6201,
"step": 2365
},
{
"epoch": 1.9591385974599669,
"grad_norm": 0.8859817981719971,
"learning_rate": 3.857567801464466e-06,
"loss": 0.6171,
"step": 2366
},
{
"epoch": 1.9599668691330756,
"grad_norm": 0.9104642271995544,
"learning_rate": 3.8566442443049425e-06,
"loss": 0.6273,
"step": 2367
},
{
"epoch": 1.9607951408061846,
"grad_norm": 0.8821743726730347,
"learning_rate": 3.855720424647403e-06,
"loss": 0.5993,
"step": 2368
},
{
"epoch": 1.9616234124792933,
"grad_norm": 0.903322160243988,
"learning_rate": 3.854796342670601e-06,
"loss": 0.6471,
"step": 2369
},
{
"epoch": 1.962451684152402,
"grad_norm": 0.8744645714759827,
"learning_rate": 3.853871998553334e-06,
"loss": 0.6444,
"step": 2370
},
{
"epoch": 1.9632799558255107,
"grad_norm": 0.9538072347640991,
"learning_rate": 3.852947392474457e-06,
"loss": 0.656,
"step": 2371
},
{
"epoch": 1.9641082274986195,
"grad_norm": 0.875407874584198,
"learning_rate": 3.852022524612872e-06,
"loss": 0.6479,
"step": 2372
},
{
"epoch": 1.9649364991717282,
"grad_norm": 0.8908282518386841,
"learning_rate": 3.8510973951475305e-06,
"loss": 0.6262,
"step": 2373
},
{
"epoch": 1.9657647708448371,
"grad_norm": 0.8698679208755493,
"learning_rate": 3.850172004257438e-06,
"loss": 0.6431,
"step": 2374
},
{
"epoch": 1.9665930425179459,
"grad_norm": 0.8997004628181458,
"learning_rate": 3.84924635212165e-06,
"loss": 0.6349,
"step": 2375
},
{
"epoch": 1.9674213141910548,
"grad_norm": 0.8815768957138062,
"learning_rate": 3.8483204389192695e-06,
"loss": 0.6449,
"step": 2376
},
{
"epoch": 1.9682495858641635,
"grad_norm": 0.9148927927017212,
"learning_rate": 3.847394264829453e-06,
"loss": 0.6106,
"step": 2377
},
{
"epoch": 1.9690778575372723,
"grad_norm": 0.8647779226303101,
"learning_rate": 3.846467830031407e-06,
"loss": 0.5954,
"step": 2378
},
{
"epoch": 1.969906129210381,
"grad_norm": 0.8877570033073425,
"learning_rate": 3.845541134704387e-06,
"loss": 0.6006,
"step": 2379
},
{
"epoch": 1.9707344008834897,
"grad_norm": 0.899742603302002,
"learning_rate": 3.844614179027701e-06,
"loss": 0.6069,
"step": 2380
},
{
"epoch": 1.9715626725565985,
"grad_norm": 0.9085001349449158,
"learning_rate": 3.843686963180705e-06,
"loss": 0.6332,
"step": 2381
},
{
"epoch": 1.9723909442297074,
"grad_norm": 0.8954899311065674,
"learning_rate": 3.842759487342808e-06,
"loss": 0.6186,
"step": 2382
},
{
"epoch": 1.9732192159028161,
"grad_norm": 0.9011021256446838,
"learning_rate": 3.841831751693466e-06,
"loss": 0.6177,
"step": 2383
},
{
"epoch": 1.9740474875759249,
"grad_norm": 0.8824598789215088,
"learning_rate": 3.8409037564121895e-06,
"loss": 0.6392,
"step": 2384
},
{
"epoch": 1.9748757592490338,
"grad_norm": 0.8936701416969299,
"learning_rate": 3.839975501678536e-06,
"loss": 0.6508,
"step": 2385
},
{
"epoch": 1.9757040309221425,
"grad_norm": 0.9009436368942261,
"learning_rate": 3.839046987672113e-06,
"loss": 0.6169,
"step": 2386
},
{
"epoch": 1.9765323025952513,
"grad_norm": 0.8839378356933594,
"learning_rate": 3.8381182145725815e-06,
"loss": 0.6107,
"step": 2387
},
{
"epoch": 1.97736057426836,
"grad_norm": 0.9240637421607971,
"learning_rate": 3.837189182559649e-06,
"loss": 0.6083,
"step": 2388
},
{
"epoch": 1.9781888459414687,
"grad_norm": 0.881908118724823,
"learning_rate": 3.836259891813075e-06,
"loss": 0.5995,
"step": 2389
},
{
"epoch": 1.9790171176145774,
"grad_norm": 0.8754116296768188,
"learning_rate": 3.835330342512667e-06,
"loss": 0.6111,
"step": 2390
},
{
"epoch": 1.9798453892876864,
"grad_norm": 0.9713384509086609,
"learning_rate": 3.834400534838287e-06,
"loss": 0.641,
"step": 2391
},
{
"epoch": 1.9806736609607951,
"grad_norm": 0.9263403415679932,
"learning_rate": 3.833470468969842e-06,
"loss": 0.6254,
"step": 2392
},
{
"epoch": 1.981501932633904,
"grad_norm": 0.9033388495445251,
"learning_rate": 3.832540145087291e-06,
"loss": 0.6049,
"step": 2393
},
{
"epoch": 1.9823302043070128,
"grad_norm": 0.8710628151893616,
"learning_rate": 3.8316095633706455e-06,
"loss": 0.5999,
"step": 2394
},
{
"epoch": 1.9831584759801215,
"grad_norm": 0.9053885340690613,
"learning_rate": 3.830678723999961e-06,
"loss": 0.6393,
"step": 2395
},
{
"epoch": 1.9839867476532302,
"grad_norm": 0.9075314998626709,
"learning_rate": 3.829747627155348e-06,
"loss": 0.6188,
"step": 2396
},
{
"epoch": 1.984815019326339,
"grad_norm": 0.9151870012283325,
"learning_rate": 3.828816273016966e-06,
"loss": 0.625,
"step": 2397
},
{
"epoch": 1.9856432909994477,
"grad_norm": 0.903843104839325,
"learning_rate": 3.827884661765022e-06,
"loss": 0.618,
"step": 2398
},
{
"epoch": 1.9864715626725566,
"grad_norm": 0.875468373298645,
"learning_rate": 3.826952793579774e-06,
"loss": 0.6121,
"step": 2399
},
{
"epoch": 1.9872998343456654,
"grad_norm": 0.8908163905143738,
"learning_rate": 3.8260206686415295e-06,
"loss": 0.6292,
"step": 2400
},
{
"epoch": 1.9881281060187743,
"grad_norm": 0.8985518217086792,
"learning_rate": 3.825088287130647e-06,
"loss": 0.6356,
"step": 2401
},
{
"epoch": 1.988956377691883,
"grad_norm": 0.9277573227882385,
"learning_rate": 3.8241556492275335e-06,
"loss": 0.6424,
"step": 2402
},
{
"epoch": 1.9897846493649918,
"grad_norm": 0.9075362682342529,
"learning_rate": 3.823222755112646e-06,
"loss": 0.623,
"step": 2403
},
{
"epoch": 1.9906129210381005,
"grad_norm": 0.9038227796554565,
"learning_rate": 3.822289604966489e-06,
"loss": 0.6195,
"step": 2404
},
{
"epoch": 1.9914411927112092,
"grad_norm": 0.8869004845619202,
"learning_rate": 3.82135619896962e-06,
"loss": 0.6034,
"step": 2405
},
{
"epoch": 1.992269464384318,
"grad_norm": 0.9017077684402466,
"learning_rate": 3.820422537302645e-06,
"loss": 0.6469,
"step": 2406
},
{
"epoch": 1.9930977360574267,
"grad_norm": 0.8785225749015808,
"learning_rate": 3.819488620146218e-06,
"loss": 0.6327,
"step": 2407
},
{
"epoch": 1.9939260077305356,
"grad_norm": 1.0367780923843384,
"learning_rate": 3.818554447681041e-06,
"loss": 0.6259,
"step": 2408
},
{
"epoch": 1.9947542794036444,
"grad_norm": 0.9392277002334595,
"learning_rate": 3.817620020087872e-06,
"loss": 0.6393,
"step": 2409
},
{
"epoch": 1.9955825510767533,
"grad_norm": 0.8935515880584717,
"learning_rate": 3.816685337547511e-06,
"loss": 0.6153,
"step": 2410
},
{
"epoch": 1.996410822749862,
"grad_norm": 0.8895331025123596,
"learning_rate": 3.815750400240811e-06,
"loss": 0.5978,
"step": 2411
},
{
"epoch": 1.9972390944229708,
"grad_norm": 0.8733546733856201,
"learning_rate": 3.814815208348673e-06,
"loss": 0.6254,
"step": 2412
},
{
"epoch": 1.9980673660960795,
"grad_norm": 0.9177356958389282,
"learning_rate": 3.8138797620520507e-06,
"loss": 0.631,
"step": 2413
},
{
"epoch": 1.9988956377691882,
"grad_norm": 0.9259604811668396,
"learning_rate": 3.8129440615319414e-06,
"loss": 0.6138,
"step": 2414
}
],
"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": 6.096168561514906e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}