meta_chat_reasoning_50_50_system / trainer_state.json
ryanmarten's picture
End of training
ac66006 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 1235,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004048582995951417,
"grad_norm": 11.347427834246414,
"learning_rate": 3.2258064516129035e-07,
"loss": 1.1793,
"step": 1
},
{
"epoch": 0.008097165991902834,
"grad_norm": 8.666632003495156,
"learning_rate": 6.451612903225807e-07,
"loss": 1.0933,
"step": 2
},
{
"epoch": 0.012145748987854251,
"grad_norm": 7.706634495860031,
"learning_rate": 9.67741935483871e-07,
"loss": 0.9687,
"step": 3
},
{
"epoch": 0.016194331983805668,
"grad_norm": 15.264459374946204,
"learning_rate": 1.2903225806451614e-06,
"loss": 1.1074,
"step": 4
},
{
"epoch": 0.020242914979757085,
"grad_norm": 9.960670273784594,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.1599,
"step": 5
},
{
"epoch": 0.024291497975708502,
"grad_norm": 6.315477446910755,
"learning_rate": 1.935483870967742e-06,
"loss": 1.0018,
"step": 6
},
{
"epoch": 0.02834008097165992,
"grad_norm": 5.357548596478053,
"learning_rate": 2.2580645161290324e-06,
"loss": 0.9219,
"step": 7
},
{
"epoch": 0.032388663967611336,
"grad_norm": 11.904450616978572,
"learning_rate": 2.580645161290323e-06,
"loss": 1.0542,
"step": 8
},
{
"epoch": 0.03643724696356275,
"grad_norm": 4.037182002636295,
"learning_rate": 2.903225806451613e-06,
"loss": 0.9005,
"step": 9
},
{
"epoch": 0.04048582995951417,
"grad_norm": 5.0419816658351,
"learning_rate": 3.225806451612903e-06,
"loss": 0.9271,
"step": 10
},
{
"epoch": 0.044534412955465584,
"grad_norm": 4.239857816967519,
"learning_rate": 3.548387096774194e-06,
"loss": 0.8423,
"step": 11
},
{
"epoch": 0.048582995951417005,
"grad_norm": 3.8510913945027374,
"learning_rate": 3.870967741935484e-06,
"loss": 0.8408,
"step": 12
},
{
"epoch": 0.05263157894736842,
"grad_norm": 3.329748616762171,
"learning_rate": 4.193548387096774e-06,
"loss": 0.8196,
"step": 13
},
{
"epoch": 0.05668016194331984,
"grad_norm": 2.821562431366395,
"learning_rate": 4.516129032258065e-06,
"loss": 0.8217,
"step": 14
},
{
"epoch": 0.06072874493927125,
"grad_norm": 2.0866742394573206,
"learning_rate": 4.838709677419355e-06,
"loss": 0.8009,
"step": 15
},
{
"epoch": 0.06477732793522267,
"grad_norm": 2.4262108112350584,
"learning_rate": 5.161290322580646e-06,
"loss": 0.8334,
"step": 16
},
{
"epoch": 0.06882591093117409,
"grad_norm": 2.4819170757764644,
"learning_rate": 5.483870967741935e-06,
"loss": 0.7679,
"step": 17
},
{
"epoch": 0.0728744939271255,
"grad_norm": 3.26183501887231,
"learning_rate": 5.806451612903226e-06,
"loss": 0.7685,
"step": 18
},
{
"epoch": 0.07692307692307693,
"grad_norm": 2.6994801199080807,
"learning_rate": 6.129032258064517e-06,
"loss": 0.7249,
"step": 19
},
{
"epoch": 0.08097165991902834,
"grad_norm": 2.881481878293029,
"learning_rate": 6.451612903225806e-06,
"loss": 0.7181,
"step": 20
},
{
"epoch": 0.08502024291497975,
"grad_norm": 5.011272252973126,
"learning_rate": 6.774193548387097e-06,
"loss": 0.7956,
"step": 21
},
{
"epoch": 0.08906882591093117,
"grad_norm": 1.6716874431733777,
"learning_rate": 7.096774193548388e-06,
"loss": 0.6957,
"step": 22
},
{
"epoch": 0.0931174089068826,
"grad_norm": 1.6602577152142572,
"learning_rate": 7.4193548387096784e-06,
"loss": 0.7466,
"step": 23
},
{
"epoch": 0.09716599190283401,
"grad_norm": 1.6594595367719083,
"learning_rate": 7.741935483870968e-06,
"loss": 0.7312,
"step": 24
},
{
"epoch": 0.10121457489878542,
"grad_norm": 1.613534466122695,
"learning_rate": 8.064516129032258e-06,
"loss": 0.7539,
"step": 25
},
{
"epoch": 0.10526315789473684,
"grad_norm": 1.538667063216181,
"learning_rate": 8.387096774193549e-06,
"loss": 0.6726,
"step": 26
},
{
"epoch": 0.10931174089068826,
"grad_norm": 1.5148368019998546,
"learning_rate": 8.70967741935484e-06,
"loss": 0.688,
"step": 27
},
{
"epoch": 0.11336032388663968,
"grad_norm": 1.396269219115153,
"learning_rate": 9.03225806451613e-06,
"loss": 0.6896,
"step": 28
},
{
"epoch": 0.11740890688259109,
"grad_norm": 1.2079287791263986,
"learning_rate": 9.35483870967742e-06,
"loss": 0.66,
"step": 29
},
{
"epoch": 0.1214574898785425,
"grad_norm": 1.3083129076496907,
"learning_rate": 9.67741935483871e-06,
"loss": 0.6878,
"step": 30
},
{
"epoch": 0.12550607287449392,
"grad_norm": 1.5049014562129688,
"learning_rate": 1e-05,
"loss": 0.6683,
"step": 31
},
{
"epoch": 0.12955465587044535,
"grad_norm": 1.4870261409376142,
"learning_rate": 1.0322580645161291e-05,
"loss": 0.6367,
"step": 32
},
{
"epoch": 0.13360323886639677,
"grad_norm": 1.7405016727759932,
"learning_rate": 1.0645161290322582e-05,
"loss": 0.6623,
"step": 33
},
{
"epoch": 0.13765182186234817,
"grad_norm": 1.1025882823318527,
"learning_rate": 1.096774193548387e-05,
"loss": 0.6424,
"step": 34
},
{
"epoch": 0.1417004048582996,
"grad_norm": 1.711244261935097,
"learning_rate": 1.1290322580645164e-05,
"loss": 0.6673,
"step": 35
},
{
"epoch": 0.145748987854251,
"grad_norm": 1.243425817871486,
"learning_rate": 1.1612903225806453e-05,
"loss": 0.6879,
"step": 36
},
{
"epoch": 0.14979757085020243,
"grad_norm": 1.4413459431408993,
"learning_rate": 1.1935483870967743e-05,
"loss": 0.6677,
"step": 37
},
{
"epoch": 0.15384615384615385,
"grad_norm": 1.7490884500315658,
"learning_rate": 1.2258064516129034e-05,
"loss": 0.6112,
"step": 38
},
{
"epoch": 0.15789473684210525,
"grad_norm": 1.4014999337528382,
"learning_rate": 1.2580645161290324e-05,
"loss": 0.6434,
"step": 39
},
{
"epoch": 0.16194331983805668,
"grad_norm": 2.158009070312463,
"learning_rate": 1.2903225806451613e-05,
"loss": 0.6762,
"step": 40
},
{
"epoch": 0.1659919028340081,
"grad_norm": 1.3815309915299507,
"learning_rate": 1.3225806451612903e-05,
"loss": 0.6945,
"step": 41
},
{
"epoch": 0.1700404858299595,
"grad_norm": 1.1482862170995356,
"learning_rate": 1.3548387096774194e-05,
"loss": 0.6577,
"step": 42
},
{
"epoch": 0.17408906882591094,
"grad_norm": 1.197270637242798,
"learning_rate": 1.3870967741935486e-05,
"loss": 0.6244,
"step": 43
},
{
"epoch": 0.17813765182186234,
"grad_norm": 1.1504196605125483,
"learning_rate": 1.4193548387096776e-05,
"loss": 0.6254,
"step": 44
},
{
"epoch": 0.18218623481781376,
"grad_norm": 1.1743272289506776,
"learning_rate": 1.4516129032258066e-05,
"loss": 0.6015,
"step": 45
},
{
"epoch": 0.1862348178137652,
"grad_norm": 2.6265346726994525,
"learning_rate": 1.4838709677419357e-05,
"loss": 0.6712,
"step": 46
},
{
"epoch": 0.1902834008097166,
"grad_norm": 1.2571240975308322,
"learning_rate": 1.5161290322580646e-05,
"loss": 0.6483,
"step": 47
},
{
"epoch": 0.19433198380566802,
"grad_norm": 1.4999896825051986,
"learning_rate": 1.5483870967741936e-05,
"loss": 0.6152,
"step": 48
},
{
"epoch": 0.19838056680161945,
"grad_norm": 1.0873827043627597,
"learning_rate": 1.5806451612903226e-05,
"loss": 0.6578,
"step": 49
},
{
"epoch": 0.20242914979757085,
"grad_norm": 1.3491767951787395,
"learning_rate": 1.6129032258064517e-05,
"loss": 0.6594,
"step": 50
},
{
"epoch": 0.20647773279352227,
"grad_norm": 1.6340233002190434,
"learning_rate": 1.6451612903225807e-05,
"loss": 0.6006,
"step": 51
},
{
"epoch": 0.21052631578947367,
"grad_norm": 1.2919129306405928,
"learning_rate": 1.6774193548387098e-05,
"loss": 0.6888,
"step": 52
},
{
"epoch": 0.2145748987854251,
"grad_norm": 1.3269632250668109,
"learning_rate": 1.7096774193548388e-05,
"loss": 0.6278,
"step": 53
},
{
"epoch": 0.21862348178137653,
"grad_norm": 1.1373042669624553,
"learning_rate": 1.741935483870968e-05,
"loss": 0.5985,
"step": 54
},
{
"epoch": 0.22267206477732793,
"grad_norm": 1.2545462471921844,
"learning_rate": 1.774193548387097e-05,
"loss": 0.6648,
"step": 55
},
{
"epoch": 0.22672064777327935,
"grad_norm": 1.5374642516597983,
"learning_rate": 1.806451612903226e-05,
"loss": 0.6807,
"step": 56
},
{
"epoch": 0.23076923076923078,
"grad_norm": 1.2702119017314726,
"learning_rate": 1.838709677419355e-05,
"loss": 0.6495,
"step": 57
},
{
"epoch": 0.23481781376518218,
"grad_norm": 1.3530473559445024,
"learning_rate": 1.870967741935484e-05,
"loss": 0.6167,
"step": 58
},
{
"epoch": 0.2388663967611336,
"grad_norm": 1.5517965457389118,
"learning_rate": 1.903225806451613e-05,
"loss": 0.6439,
"step": 59
},
{
"epoch": 0.242914979757085,
"grad_norm": 1.2414001362295908,
"learning_rate": 1.935483870967742e-05,
"loss": 0.6368,
"step": 60
},
{
"epoch": 0.24696356275303644,
"grad_norm": 1.5676588986769422,
"learning_rate": 1.967741935483871e-05,
"loss": 0.6631,
"step": 61
},
{
"epoch": 0.25101214574898784,
"grad_norm": 3.749973329741446,
"learning_rate": 2e-05,
"loss": 0.6601,
"step": 62
},
{
"epoch": 0.2550607287449393,
"grad_norm": 1.5518276336036936,
"learning_rate": 2.0322580645161292e-05,
"loss": 0.6797,
"step": 63
},
{
"epoch": 0.2591093117408907,
"grad_norm": 1.2587569753274126,
"learning_rate": 2.0645161290322582e-05,
"loss": 0.6396,
"step": 64
},
{
"epoch": 0.2631578947368421,
"grad_norm": 1.324159010807197,
"learning_rate": 2.0967741935483873e-05,
"loss": 0.6141,
"step": 65
},
{
"epoch": 0.26720647773279355,
"grad_norm": 1.3470192440023554,
"learning_rate": 2.1290322580645163e-05,
"loss": 0.6826,
"step": 66
},
{
"epoch": 0.27125506072874495,
"grad_norm": 1.2512986599009488,
"learning_rate": 2.161290322580645e-05,
"loss": 0.6799,
"step": 67
},
{
"epoch": 0.27530364372469635,
"grad_norm": 1.3469332098237967,
"learning_rate": 2.193548387096774e-05,
"loss": 0.6472,
"step": 68
},
{
"epoch": 0.2793522267206478,
"grad_norm": 1.6191425334399308,
"learning_rate": 2.225806451612903e-05,
"loss": 0.6336,
"step": 69
},
{
"epoch": 0.2834008097165992,
"grad_norm": 1.3329825396222077,
"learning_rate": 2.2580645161290328e-05,
"loss": 0.6168,
"step": 70
},
{
"epoch": 0.2874493927125506,
"grad_norm": 1.379066816754817,
"learning_rate": 2.2903225806451615e-05,
"loss": 0.6818,
"step": 71
},
{
"epoch": 0.291497975708502,
"grad_norm": 1.2717497508615527,
"learning_rate": 2.3225806451612906e-05,
"loss": 0.6264,
"step": 72
},
{
"epoch": 0.29554655870445345,
"grad_norm": 1.105043781026222,
"learning_rate": 2.3548387096774196e-05,
"loss": 0.6303,
"step": 73
},
{
"epoch": 0.29959514170040485,
"grad_norm": 1.1834962464677936,
"learning_rate": 2.3870967741935486e-05,
"loss": 0.6525,
"step": 74
},
{
"epoch": 0.30364372469635625,
"grad_norm": 1.2864984833450814,
"learning_rate": 2.4193548387096777e-05,
"loss": 0.612,
"step": 75
},
{
"epoch": 0.3076923076923077,
"grad_norm": 1.0759838264401245,
"learning_rate": 2.4516129032258067e-05,
"loss": 0.5951,
"step": 76
},
{
"epoch": 0.3117408906882591,
"grad_norm": 1.4999449440027763,
"learning_rate": 2.4838709677419358e-05,
"loss": 0.6505,
"step": 77
},
{
"epoch": 0.3157894736842105,
"grad_norm": 1.2847306949750221,
"learning_rate": 2.5161290322580648e-05,
"loss": 0.614,
"step": 78
},
{
"epoch": 0.31983805668016196,
"grad_norm": 1.832398269799664,
"learning_rate": 2.548387096774194e-05,
"loss": 0.6377,
"step": 79
},
{
"epoch": 0.32388663967611336,
"grad_norm": 1.1478664270299517,
"learning_rate": 2.5806451612903226e-05,
"loss": 0.6507,
"step": 80
},
{
"epoch": 0.32793522267206476,
"grad_norm": 2.906015598218103,
"learning_rate": 2.6129032258064516e-05,
"loss": 0.6244,
"step": 81
},
{
"epoch": 0.3319838056680162,
"grad_norm": 1.2633467650066428,
"learning_rate": 2.6451612903225806e-05,
"loss": 0.5984,
"step": 82
},
{
"epoch": 0.3360323886639676,
"grad_norm": 0.9396758231188451,
"learning_rate": 2.6774193548387097e-05,
"loss": 0.6198,
"step": 83
},
{
"epoch": 0.340080971659919,
"grad_norm": 1.2216397056772976,
"learning_rate": 2.7096774193548387e-05,
"loss": 0.6235,
"step": 84
},
{
"epoch": 0.3441295546558704,
"grad_norm": 1.1927758771374821,
"learning_rate": 2.7419354838709678e-05,
"loss": 0.684,
"step": 85
},
{
"epoch": 0.3481781376518219,
"grad_norm": 1.9879194015828368,
"learning_rate": 2.774193548387097e-05,
"loss": 0.6574,
"step": 86
},
{
"epoch": 0.3522267206477733,
"grad_norm": 1.4053058680109733,
"learning_rate": 2.8064516129032262e-05,
"loss": 0.6128,
"step": 87
},
{
"epoch": 0.3562753036437247,
"grad_norm": 2.343948277847098,
"learning_rate": 2.8387096774193552e-05,
"loss": 0.6575,
"step": 88
},
{
"epoch": 0.3603238866396761,
"grad_norm": 1.1119799300115432,
"learning_rate": 2.8709677419354843e-05,
"loss": 0.6171,
"step": 89
},
{
"epoch": 0.3643724696356275,
"grad_norm": 1.116467831695336,
"learning_rate": 2.9032258064516133e-05,
"loss": 0.6813,
"step": 90
},
{
"epoch": 0.3684210526315789,
"grad_norm": 1.108395563244231,
"learning_rate": 2.9354838709677423e-05,
"loss": 0.6428,
"step": 91
},
{
"epoch": 0.3724696356275304,
"grad_norm": 1.4159577720255188,
"learning_rate": 2.9677419354838714e-05,
"loss": 0.6159,
"step": 92
},
{
"epoch": 0.3765182186234818,
"grad_norm": 1.727986891929083,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.6276,
"step": 93
},
{
"epoch": 0.3805668016194332,
"grad_norm": 1.1628825886960616,
"learning_rate": 3.032258064516129e-05,
"loss": 0.6232,
"step": 94
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.332553817923901,
"learning_rate": 3.0645161290322585e-05,
"loss": 0.6451,
"step": 95
},
{
"epoch": 0.38866396761133604,
"grad_norm": 1.082439720360465,
"learning_rate": 3.096774193548387e-05,
"loss": 0.6352,
"step": 96
},
{
"epoch": 0.39271255060728744,
"grad_norm": 0.9697488120290962,
"learning_rate": 3.1290322580645166e-05,
"loss": 0.6282,
"step": 97
},
{
"epoch": 0.3967611336032389,
"grad_norm": 1.2749367734118122,
"learning_rate": 3.161290322580645e-05,
"loss": 0.6473,
"step": 98
},
{
"epoch": 0.4008097165991903,
"grad_norm": 1.0815735234038646,
"learning_rate": 3.193548387096774e-05,
"loss": 0.625,
"step": 99
},
{
"epoch": 0.4048582995951417,
"grad_norm": 1.4969491670645156,
"learning_rate": 3.2258064516129034e-05,
"loss": 0.6323,
"step": 100
},
{
"epoch": 0.4089068825910931,
"grad_norm": 1.1473919182114656,
"learning_rate": 3.258064516129033e-05,
"loss": 0.6036,
"step": 101
},
{
"epoch": 0.41295546558704455,
"grad_norm": 1.079083010092156,
"learning_rate": 3.2903225806451614e-05,
"loss": 0.6372,
"step": 102
},
{
"epoch": 0.41700404858299595,
"grad_norm": 1.112214567731508,
"learning_rate": 3.322580645161291e-05,
"loss": 0.6353,
"step": 103
},
{
"epoch": 0.42105263157894735,
"grad_norm": 1.0421852600435355,
"learning_rate": 3.3548387096774195e-05,
"loss": 0.6867,
"step": 104
},
{
"epoch": 0.4251012145748988,
"grad_norm": 1.3208599818463809,
"learning_rate": 3.387096774193549e-05,
"loss": 0.6599,
"step": 105
},
{
"epoch": 0.4291497975708502,
"grad_norm": 1.0133860870682334,
"learning_rate": 3.4193548387096776e-05,
"loss": 0.6049,
"step": 106
},
{
"epoch": 0.4331983805668016,
"grad_norm": 1.4544255621004913,
"learning_rate": 3.451612903225807e-05,
"loss": 0.6402,
"step": 107
},
{
"epoch": 0.43724696356275305,
"grad_norm": 0.9592825578147338,
"learning_rate": 3.483870967741936e-05,
"loss": 0.6236,
"step": 108
},
{
"epoch": 0.44129554655870445,
"grad_norm": 1.141442161134121,
"learning_rate": 3.516129032258065e-05,
"loss": 0.6042,
"step": 109
},
{
"epoch": 0.44534412955465585,
"grad_norm": 1.2075163207704633,
"learning_rate": 3.548387096774194e-05,
"loss": 0.6482,
"step": 110
},
{
"epoch": 0.4493927125506073,
"grad_norm": 1.0211851292434546,
"learning_rate": 3.5806451612903225e-05,
"loss": 0.6228,
"step": 111
},
{
"epoch": 0.4534412955465587,
"grad_norm": 1.1455390964881076,
"learning_rate": 3.612903225806452e-05,
"loss": 0.6579,
"step": 112
},
{
"epoch": 0.4574898785425101,
"grad_norm": 0.9933456551669142,
"learning_rate": 3.6451612903225805e-05,
"loss": 0.6328,
"step": 113
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.117234647190974,
"learning_rate": 3.67741935483871e-05,
"loss": 0.6505,
"step": 114
},
{
"epoch": 0.46558704453441296,
"grad_norm": 1.032928866156577,
"learning_rate": 3.7096774193548386e-05,
"loss": 0.6801,
"step": 115
},
{
"epoch": 0.46963562753036436,
"grad_norm": 0.900071123371017,
"learning_rate": 3.741935483870968e-05,
"loss": 0.6531,
"step": 116
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.8417538260697496,
"learning_rate": 3.7741935483870974e-05,
"loss": 0.5569,
"step": 117
},
{
"epoch": 0.4777327935222672,
"grad_norm": 0.9522360710986643,
"learning_rate": 3.806451612903226e-05,
"loss": 0.6182,
"step": 118
},
{
"epoch": 0.4817813765182186,
"grad_norm": 0.9639297769962729,
"learning_rate": 3.8387096774193555e-05,
"loss": 0.612,
"step": 119
},
{
"epoch": 0.48582995951417,
"grad_norm": 0.9662212627182868,
"learning_rate": 3.870967741935484e-05,
"loss": 0.6096,
"step": 120
},
{
"epoch": 0.4898785425101215,
"grad_norm": 1.3390949752718082,
"learning_rate": 3.9032258064516135e-05,
"loss": 0.6547,
"step": 121
},
{
"epoch": 0.4939271255060729,
"grad_norm": 1.4667059167323184,
"learning_rate": 3.935483870967742e-05,
"loss": 0.7453,
"step": 122
},
{
"epoch": 0.4979757085020243,
"grad_norm": 1.0546385070882134,
"learning_rate": 3.9677419354838716e-05,
"loss": 0.7509,
"step": 123
},
{
"epoch": 0.5020242914979757,
"grad_norm": 1.3853865692032852,
"learning_rate": 4e-05,
"loss": 0.664,
"step": 124
},
{
"epoch": 0.5060728744939271,
"grad_norm": 1.1337079788944244,
"learning_rate": 3.9999920040266474e-05,
"loss": 0.6284,
"step": 125
},
{
"epoch": 0.5101214574898786,
"grad_norm": 1.0005988884268517,
"learning_rate": 3.999968016170525e-05,
"loss": 0.6582,
"step": 126
},
{
"epoch": 0.5141700404858299,
"grad_norm": 1.1010292118550422,
"learning_rate": 3.9999280366234394e-05,
"loss": 0.6212,
"step": 127
},
{
"epoch": 0.5182186234817814,
"grad_norm": 0.9390843330414733,
"learning_rate": 3.999872065705065e-05,
"loss": 0.6588,
"step": 128
},
{
"epoch": 0.5222672064777328,
"grad_norm": 0.9996089505319317,
"learning_rate": 3.999800103862945e-05,
"loss": 0.6577,
"step": 129
},
{
"epoch": 0.5263157894736842,
"grad_norm": 1.2443071245296287,
"learning_rate": 3.9997121516724834e-05,
"loss": 0.6762,
"step": 130
},
{
"epoch": 0.5303643724696356,
"grad_norm": 1.0539301474992129,
"learning_rate": 3.999608209836944e-05,
"loss": 0.6559,
"step": 131
},
{
"epoch": 0.5344129554655871,
"grad_norm": 1.1784566724771592,
"learning_rate": 3.999488279187443e-05,
"loss": 0.6273,
"step": 132
},
{
"epoch": 0.5384615384615384,
"grad_norm": 1.4368869205458423,
"learning_rate": 3.9993523606829425e-05,
"loss": 0.6874,
"step": 133
},
{
"epoch": 0.5425101214574899,
"grad_norm": 1.9406407173487628,
"learning_rate": 3.9992004554102435e-05,
"loss": 0.727,
"step": 134
},
{
"epoch": 0.5465587044534413,
"grad_norm": 1.2634266843659512,
"learning_rate": 3.999032564583976e-05,
"loss": 0.6368,
"step": 135
},
{
"epoch": 0.5506072874493927,
"grad_norm": 1.8583762587632695,
"learning_rate": 3.998848689546591e-05,
"loss": 0.6688,
"step": 136
},
{
"epoch": 0.5546558704453441,
"grad_norm": 1.0879887954123943,
"learning_rate": 3.9986488317683496e-05,
"loss": 0.6488,
"step": 137
},
{
"epoch": 0.5587044534412956,
"grad_norm": 1.0330366614614181,
"learning_rate": 3.9984329928473067e-05,
"loss": 0.6508,
"step": 138
},
{
"epoch": 0.562753036437247,
"grad_norm": 1.0397409773534914,
"learning_rate": 3.998201174509306e-05,
"loss": 0.6668,
"step": 139
},
{
"epoch": 0.5668016194331984,
"grad_norm": 0.8409760409191535,
"learning_rate": 3.997953378607961e-05,
"loss": 0.6139,
"step": 140
},
{
"epoch": 0.5708502024291497,
"grad_norm": 0.9573178019070454,
"learning_rate": 3.9976896071246403e-05,
"loss": 0.6455,
"step": 141
},
{
"epoch": 0.5748987854251012,
"grad_norm": 1.0110588858251772,
"learning_rate": 3.997409862168454e-05,
"loss": 0.6135,
"step": 142
},
{
"epoch": 0.5789473684210527,
"grad_norm": 1.2108706174324149,
"learning_rate": 3.997114145976237e-05,
"loss": 0.7342,
"step": 143
},
{
"epoch": 0.582995951417004,
"grad_norm": 1.2163757033398532,
"learning_rate": 3.9968024609125245e-05,
"loss": 0.6597,
"step": 144
},
{
"epoch": 0.5870445344129555,
"grad_norm": 0.985452536818492,
"learning_rate": 3.996474809469545e-05,
"loss": 0.657,
"step": 145
},
{
"epoch": 0.5910931174089069,
"grad_norm": 0.9058493709924511,
"learning_rate": 3.996131194267189e-05,
"loss": 0.69,
"step": 146
},
{
"epoch": 0.5951417004048583,
"grad_norm": 1.5405755083333639,
"learning_rate": 3.995771618052995e-05,
"loss": 0.6778,
"step": 147
},
{
"epoch": 0.5991902834008097,
"grad_norm": 1.5327231107061816,
"learning_rate": 3.9953960837021256e-05,
"loss": 0.6517,
"step": 148
},
{
"epoch": 0.6032388663967612,
"grad_norm": 1.012791099979681,
"learning_rate": 3.995004594217343e-05,
"loss": 0.6667,
"step": 149
},
{
"epoch": 0.6072874493927125,
"grad_norm": 1.143007932649428,
"learning_rate": 3.994597152728987e-05,
"loss": 0.6523,
"step": 150
},
{
"epoch": 0.611336032388664,
"grad_norm": 1.2816544086729924,
"learning_rate": 3.994173762494947e-05,
"loss": 0.6961,
"step": 151
},
{
"epoch": 0.6153846153846154,
"grad_norm": 1.0827692568087062,
"learning_rate": 3.9937344269006424e-05,
"loss": 0.6692,
"step": 152
},
{
"epoch": 0.6194331983805668,
"grad_norm": 1.098611388662187,
"learning_rate": 3.993279149458988e-05,
"loss": 0.6833,
"step": 153
},
{
"epoch": 0.6234817813765182,
"grad_norm": 0.9318538974971404,
"learning_rate": 3.9928079338103695e-05,
"loss": 0.6101,
"step": 154
},
{
"epoch": 0.6275303643724697,
"grad_norm": 0.8516242435843555,
"learning_rate": 3.992320783722616e-05,
"loss": 0.6269,
"step": 155
},
{
"epoch": 0.631578947368421,
"grad_norm": 1.1394908845411607,
"learning_rate": 3.991817703090965e-05,
"loss": 0.6358,
"step": 156
},
{
"epoch": 0.6356275303643725,
"grad_norm": 0.9236485706928472,
"learning_rate": 3.991298695938038e-05,
"loss": 0.6212,
"step": 157
},
{
"epoch": 0.6396761133603239,
"grad_norm": 0.758007386761704,
"learning_rate": 3.9907637664138e-05,
"loss": 0.6121,
"step": 158
},
{
"epoch": 0.6437246963562753,
"grad_norm": 0.9628921810735493,
"learning_rate": 3.990212918795534e-05,
"loss": 0.6354,
"step": 159
},
{
"epoch": 0.6477732793522267,
"grad_norm": 0.9297861030562014,
"learning_rate": 3.9896461574878036e-05,
"loss": 0.6577,
"step": 160
},
{
"epoch": 0.6518218623481782,
"grad_norm": 0.8385560550577852,
"learning_rate": 3.989063487022416e-05,
"loss": 0.6693,
"step": 161
},
{
"epoch": 0.6558704453441295,
"grad_norm": 1.8095061305380937,
"learning_rate": 3.9884649120583903e-05,
"loss": 0.6595,
"step": 162
},
{
"epoch": 0.659919028340081,
"grad_norm": 1.0137045194784642,
"learning_rate": 3.987850437381915e-05,
"loss": 0.6671,
"step": 163
},
{
"epoch": 0.6639676113360324,
"grad_norm": 1.0012591103742574,
"learning_rate": 3.987220067906313e-05,
"loss": 0.6912,
"step": 164
},
{
"epoch": 0.6680161943319838,
"grad_norm": 1.0375997373179913,
"learning_rate": 3.986573808672002e-05,
"loss": 0.5927,
"step": 165
},
{
"epoch": 0.6720647773279352,
"grad_norm": 0.8233934808921543,
"learning_rate": 3.9859116648464534e-05,
"loss": 0.6194,
"step": 166
},
{
"epoch": 0.6761133603238867,
"grad_norm": 1.0840053447467655,
"learning_rate": 3.985233641724152e-05,
"loss": 0.6542,
"step": 167
},
{
"epoch": 0.680161943319838,
"grad_norm": 1.0094375394511608,
"learning_rate": 3.984539744726554e-05,
"loss": 0.6834,
"step": 168
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.9219106063883598,
"learning_rate": 3.983829979402038e-05,
"loss": 0.6573,
"step": 169
},
{
"epoch": 0.6882591093117408,
"grad_norm": 1.152849124384071,
"learning_rate": 3.983104351425871e-05,
"loss": 0.659,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 1.040532632026705,
"learning_rate": 3.982362866600154e-05,
"loss": 0.6094,
"step": 171
},
{
"epoch": 0.6963562753036437,
"grad_norm": 0.8545786312295408,
"learning_rate": 3.9816055308537806e-05,
"loss": 0.6297,
"step": 172
},
{
"epoch": 0.7004048582995951,
"grad_norm": 0.9451937331956216,
"learning_rate": 3.980832350242388e-05,
"loss": 0.7097,
"step": 173
},
{
"epoch": 0.7044534412955465,
"grad_norm": 1.155646489904595,
"learning_rate": 3.980043330948306e-05,
"loss": 0.6715,
"step": 174
},
{
"epoch": 0.708502024291498,
"grad_norm": 0.955439497334856,
"learning_rate": 3.9792384792805124e-05,
"loss": 0.6477,
"step": 175
},
{
"epoch": 0.7125506072874493,
"grad_norm": 0.829624533203592,
"learning_rate": 3.978417801674579e-05,
"loss": 0.7032,
"step": 176
},
{
"epoch": 0.7165991902834008,
"grad_norm": 1.14440837353669,
"learning_rate": 3.9775813046926244e-05,
"loss": 0.6369,
"step": 177
},
{
"epoch": 0.7206477732793523,
"grad_norm": 0.8580574548133278,
"learning_rate": 3.976728995023254e-05,
"loss": 0.623,
"step": 178
},
{
"epoch": 0.7246963562753036,
"grad_norm": 0.8992113729029236,
"learning_rate": 3.975860879481514e-05,
"loss": 0.6465,
"step": 179
},
{
"epoch": 0.728744939271255,
"grad_norm": 0.9206742872484379,
"learning_rate": 3.974976965008833e-05,
"loss": 0.6169,
"step": 180
},
{
"epoch": 0.7327935222672065,
"grad_norm": 0.9598999852819109,
"learning_rate": 3.9740772586729674e-05,
"loss": 0.6289,
"step": 181
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.9008650832040583,
"learning_rate": 3.973161767667946e-05,
"loss": 0.6511,
"step": 182
},
{
"epoch": 0.7408906882591093,
"grad_norm": 0.983187646681521,
"learning_rate": 3.972230499314009e-05,
"loss": 0.6788,
"step": 183
},
{
"epoch": 0.7449392712550608,
"grad_norm": 0.8877729439846112,
"learning_rate": 3.9712834610575545e-05,
"loss": 0.6648,
"step": 184
},
{
"epoch": 0.7489878542510121,
"grad_norm": 1.0028990771717767,
"learning_rate": 3.9703206604710746e-05,
"loss": 0.7145,
"step": 185
},
{
"epoch": 0.7530364372469636,
"grad_norm": 2.4066720730740028,
"learning_rate": 3.969342105253098e-05,
"loss": 0.6785,
"step": 186
},
{
"epoch": 0.757085020242915,
"grad_norm": 1.1464843511376637,
"learning_rate": 3.9683478032281246e-05,
"loss": 0.6799,
"step": 187
},
{
"epoch": 0.7611336032388664,
"grad_norm": 0.9914990747512855,
"learning_rate": 3.967337762346568e-05,
"loss": 0.6734,
"step": 188
},
{
"epoch": 0.7651821862348178,
"grad_norm": 1.5029856000487658,
"learning_rate": 3.9663119906846885e-05,
"loss": 0.6343,
"step": 189
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.8207415754744064,
"learning_rate": 3.965270496444528e-05,
"loss": 0.7317,
"step": 190
},
{
"epoch": 0.7732793522267206,
"grad_norm": 1.1159363293221234,
"learning_rate": 3.9642132879538465e-05,
"loss": 0.652,
"step": 191
},
{
"epoch": 0.7773279352226721,
"grad_norm": 1.2776577136557608,
"learning_rate": 3.9631403736660566e-05,
"loss": 0.6173,
"step": 192
},
{
"epoch": 0.7813765182186235,
"grad_norm": 1.1154647951508423,
"learning_rate": 3.962051762160151e-05,
"loss": 0.6688,
"step": 193
},
{
"epoch": 0.7854251012145749,
"grad_norm": 1.0481822533900207,
"learning_rate": 3.960947462140638e-05,
"loss": 0.6174,
"step": 194
},
{
"epoch": 0.7894736842105263,
"grad_norm": 1.32255041939358,
"learning_rate": 3.959827482437472e-05,
"loss": 0.6567,
"step": 195
},
{
"epoch": 0.7935222672064778,
"grad_norm": 0.8307579190938454,
"learning_rate": 3.958691832005981e-05,
"loss": 0.6389,
"step": 196
},
{
"epoch": 0.7975708502024291,
"grad_norm": 1.2012858566160414,
"learning_rate": 3.9575405199267944e-05,
"loss": 0.7301,
"step": 197
},
{
"epoch": 0.8016194331983806,
"grad_norm": 1.0965603539072692,
"learning_rate": 3.956373555405774e-05,
"loss": 0.6541,
"step": 198
},
{
"epoch": 0.805668016194332,
"grad_norm": 0.9112220500224274,
"learning_rate": 3.955190947773937e-05,
"loss": 0.694,
"step": 199
},
{
"epoch": 0.8097165991902834,
"grad_norm": 1.6944427512925182,
"learning_rate": 3.9539927064873824e-05,
"loss": 0.6827,
"step": 200
},
{
"epoch": 0.8137651821862348,
"grad_norm": 1.3903833009267403,
"learning_rate": 3.952778841127214e-05,
"loss": 0.6944,
"step": 201
},
{
"epoch": 0.8178137651821862,
"grad_norm": 0.997868658099856,
"learning_rate": 3.95154936139947e-05,
"loss": 0.6474,
"step": 202
},
{
"epoch": 0.8218623481781376,
"grad_norm": 1.0195581560495972,
"learning_rate": 3.950304277135034e-05,
"loss": 0.6822,
"step": 203
},
{
"epoch": 0.8259109311740891,
"grad_norm": 1.3527238983462668,
"learning_rate": 3.94904359828957e-05,
"loss": 0.6338,
"step": 204
},
{
"epoch": 0.8299595141700404,
"grad_norm": 1.005468556762221,
"learning_rate": 3.94776733494343e-05,
"loss": 0.7323,
"step": 205
},
{
"epoch": 0.8340080971659919,
"grad_norm": 1.0115295923261716,
"learning_rate": 3.9464754973015826e-05,
"loss": 0.6771,
"step": 206
},
{
"epoch": 0.8380566801619433,
"grad_norm": 1.2183513240958506,
"learning_rate": 3.9451680956935275e-05,
"loss": 0.6799,
"step": 207
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.9025813211250548,
"learning_rate": 3.943845140573213e-05,
"loss": 0.6741,
"step": 208
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.9114353663513252,
"learning_rate": 3.942506642518952e-05,
"loss": 0.6685,
"step": 209
},
{
"epoch": 0.8502024291497976,
"grad_norm": 0.9685801648196057,
"learning_rate": 3.9411526122333396e-05,
"loss": 0.6587,
"step": 210
},
{
"epoch": 0.854251012145749,
"grad_norm": 1.177888659028034,
"learning_rate": 3.939783060543167e-05,
"loss": 0.6469,
"step": 211
},
{
"epoch": 0.8582995951417004,
"grad_norm": 1.131035423814624,
"learning_rate": 3.938397998399333e-05,
"loss": 0.6855,
"step": 212
},
{
"epoch": 0.8623481781376519,
"grad_norm": 0.9785545440866603,
"learning_rate": 3.936997436876756e-05,
"loss": 0.6039,
"step": 213
},
{
"epoch": 0.8663967611336032,
"grad_norm": 1.1037493320020864,
"learning_rate": 3.935581387174289e-05,
"loss": 0.6569,
"step": 214
},
{
"epoch": 0.8704453441295547,
"grad_norm": 1.0589254791674074,
"learning_rate": 3.9341498606146295e-05,
"loss": 0.6826,
"step": 215
},
{
"epoch": 0.8744939271255061,
"grad_norm": 0.8748783082046494,
"learning_rate": 3.932702868644224e-05,
"loss": 0.6656,
"step": 216
},
{
"epoch": 0.8785425101214575,
"grad_norm": 0.9457014257519218,
"learning_rate": 3.931240422833181e-05,
"loss": 0.6693,
"step": 217
},
{
"epoch": 0.8825910931174089,
"grad_norm": 0.9201483322258449,
"learning_rate": 3.9297625348751806e-05,
"loss": 0.6707,
"step": 218
},
{
"epoch": 0.8866396761133604,
"grad_norm": 0.8201356760414583,
"learning_rate": 3.9282692165873744e-05,
"loss": 0.6708,
"step": 219
},
{
"epoch": 0.8906882591093117,
"grad_norm": 0.9444883128245389,
"learning_rate": 3.9267604799102945e-05,
"loss": 0.6193,
"step": 220
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.9224355464681916,
"learning_rate": 3.9252363369077615e-05,
"loss": 0.6973,
"step": 221
},
{
"epoch": 0.8987854251012146,
"grad_norm": 1.0226645783742,
"learning_rate": 3.9236967997667796e-05,
"loss": 0.7194,
"step": 222
},
{
"epoch": 0.902834008097166,
"grad_norm": 0.9414289883515873,
"learning_rate": 3.9221418807974496e-05,
"loss": 0.6379,
"step": 223
},
{
"epoch": 0.9068825910931174,
"grad_norm": 2.1501872309217984,
"learning_rate": 3.92057159243286e-05,
"loss": 0.6867,
"step": 224
},
{
"epoch": 0.9109311740890689,
"grad_norm": 1.7839545596895272,
"learning_rate": 3.9189859472289956e-05,
"loss": 0.6939,
"step": 225
},
{
"epoch": 0.9149797570850202,
"grad_norm": 0.9675516682810018,
"learning_rate": 3.9173849578646324e-05,
"loss": 0.67,
"step": 226
},
{
"epoch": 0.9190283400809717,
"grad_norm": 0.923216925969234,
"learning_rate": 3.9157686371412395e-05,
"loss": 0.6703,
"step": 227
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.7203510110355501,
"learning_rate": 3.914136997982874e-05,
"loss": 0.5814,
"step": 228
},
{
"epoch": 0.9271255060728745,
"grad_norm": 2.5719618228814696,
"learning_rate": 3.912490053436079e-05,
"loss": 0.6921,
"step": 229
},
{
"epoch": 0.9311740890688259,
"grad_norm": 3.076590818372445,
"learning_rate": 3.9108278166697805e-05,
"loss": 0.6793,
"step": 230
},
{
"epoch": 0.9352226720647774,
"grad_norm": 0.9224426500009192,
"learning_rate": 3.909150300975178e-05,
"loss": 0.7124,
"step": 231
},
{
"epoch": 0.9392712550607287,
"grad_norm": 0.9416464315260672,
"learning_rate": 3.907457519765642e-05,
"loss": 0.652,
"step": 232
},
{
"epoch": 0.9433198380566802,
"grad_norm": 0.9509241186770211,
"learning_rate": 3.905749486576607e-05,
"loss": 0.6308,
"step": 233
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.8999515317578459,
"learning_rate": 3.90402621506546e-05,
"loss": 0.5913,
"step": 234
},
{
"epoch": 0.951417004048583,
"grad_norm": 0.8619785631995822,
"learning_rate": 3.902287719011435e-05,
"loss": 0.6618,
"step": 235
},
{
"epoch": 0.9554655870445344,
"grad_norm": 1.486662668757939,
"learning_rate": 3.900534012315499e-05,
"loss": 0.6871,
"step": 236
},
{
"epoch": 0.9595141700404858,
"grad_norm": 1.28452087046863,
"learning_rate": 3.8987651090002454e-05,
"loss": 0.6838,
"step": 237
},
{
"epoch": 0.9635627530364372,
"grad_norm": 1.3935140590821884,
"learning_rate": 3.896981023209777e-05,
"loss": 0.6623,
"step": 238
},
{
"epoch": 0.9676113360323887,
"grad_norm": 1.353056062554968,
"learning_rate": 3.8951817692095966e-05,
"loss": 0.6455,
"step": 239
},
{
"epoch": 0.97165991902834,
"grad_norm": 1.0477336216545576,
"learning_rate": 3.893367361386491e-05,
"loss": 0.6179,
"step": 240
},
{
"epoch": 0.9757085020242915,
"grad_norm": 1.0235964609525114,
"learning_rate": 3.891537814248417e-05,
"loss": 0.6397,
"step": 241
},
{
"epoch": 0.979757085020243,
"grad_norm": 0.9145622254775393,
"learning_rate": 3.889693142424384e-05,
"loss": 0.649,
"step": 242
},
{
"epoch": 0.9838056680161943,
"grad_norm": 1.1638189902507223,
"learning_rate": 3.8878333606643405e-05,
"loss": 0.6785,
"step": 243
},
{
"epoch": 0.9878542510121457,
"grad_norm": 1.1269082499055099,
"learning_rate": 3.885958483839049e-05,
"loss": 0.5904,
"step": 244
},
{
"epoch": 0.9919028340080972,
"grad_norm": 0.846845934657142,
"learning_rate": 3.8840685269399786e-05,
"loss": 0.6268,
"step": 245
},
{
"epoch": 0.9959514170040485,
"grad_norm": 0.8248498696171718,
"learning_rate": 3.882163505079171e-05,
"loss": 0.6726,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 2.3353882750894126,
"learning_rate": 3.880243433489132e-05,
"loss": 0.6771,
"step": 247
},
{
"epoch": 1.0040485829959513,
"grad_norm": 0.8646392403226383,
"learning_rate": 3.878308327522702e-05,
"loss": 0.4675,
"step": 248
},
{
"epoch": 1.008097165991903,
"grad_norm": 0.9834737535052662,
"learning_rate": 3.876358202652936e-05,
"loss": 0.4749,
"step": 249
},
{
"epoch": 1.0121457489878543,
"grad_norm": 0.822170695780851,
"learning_rate": 3.8743930744729835e-05,
"loss": 0.4362,
"step": 250
},
{
"epoch": 1.0161943319838056,
"grad_norm": 0.738436654854948,
"learning_rate": 3.8724129586959535e-05,
"loss": 0.4481,
"step": 251
},
{
"epoch": 1.0202429149797572,
"grad_norm": 0.7513315817807182,
"learning_rate": 3.870417871154801e-05,
"loss": 0.4199,
"step": 252
},
{
"epoch": 1.0242914979757085,
"grad_norm": 0.7053847097210041,
"learning_rate": 3.868407827802192e-05,
"loss": 0.4959,
"step": 253
},
{
"epoch": 1.0283400809716599,
"grad_norm": 0.7776376454699596,
"learning_rate": 3.8663828447103804e-05,
"loss": 0.437,
"step": 254
},
{
"epoch": 1.0323886639676114,
"grad_norm": 0.8274084858580887,
"learning_rate": 3.8643429380710765e-05,
"loss": 0.4646,
"step": 255
},
{
"epoch": 1.0364372469635628,
"grad_norm": 0.8691061437539319,
"learning_rate": 3.862288124195319e-05,
"loss": 0.4441,
"step": 256
},
{
"epoch": 1.040485829959514,
"grad_norm": 0.6684343944649395,
"learning_rate": 3.8602184195133464e-05,
"loss": 0.4472,
"step": 257
},
{
"epoch": 1.0445344129554657,
"grad_norm": 1.132570789583865,
"learning_rate": 3.8581338405744604e-05,
"loss": 0.4237,
"step": 258
},
{
"epoch": 1.048582995951417,
"grad_norm": 1.0272594443104808,
"learning_rate": 3.8560344040468996e-05,
"loss": 0.4669,
"step": 259
},
{
"epoch": 1.0526315789473684,
"grad_norm": 1.260207129059265,
"learning_rate": 3.853920126717703e-05,
"loss": 0.4307,
"step": 260
},
{
"epoch": 1.05668016194332,
"grad_norm": 4.0141852646437295,
"learning_rate": 3.851791025492574e-05,
"loss": 0.4902,
"step": 261
},
{
"epoch": 1.0607287449392713,
"grad_norm": 0.7479671648739642,
"learning_rate": 3.8496471173957515e-05,
"loss": 0.4364,
"step": 262
},
{
"epoch": 1.0647773279352226,
"grad_norm": 0.7436974018338265,
"learning_rate": 3.847488419569866e-05,
"loss": 0.4498,
"step": 263
},
{
"epoch": 1.0688259109311742,
"grad_norm": 0.7931109346411394,
"learning_rate": 3.8453149492758077e-05,
"loss": 0.4311,
"step": 264
},
{
"epoch": 1.0728744939271255,
"grad_norm": 0.7484626248591651,
"learning_rate": 3.8431267238925886e-05,
"loss": 0.4087,
"step": 265
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.8327248630127905,
"learning_rate": 3.840923760917199e-05,
"loss": 0.4477,
"step": 266
},
{
"epoch": 1.0809716599190284,
"grad_norm": 0.7634175559233908,
"learning_rate": 3.838706077964473e-05,
"loss": 0.4163,
"step": 267
},
{
"epoch": 1.0850202429149798,
"grad_norm": 0.9604102670373103,
"learning_rate": 3.836473692766944e-05,
"loss": 0.4551,
"step": 268
},
{
"epoch": 1.0890688259109311,
"grad_norm": 0.7320477877508798,
"learning_rate": 3.834226623174705e-05,
"loss": 0.4254,
"step": 269
},
{
"epoch": 1.0931174089068827,
"grad_norm": 0.7975900865072115,
"learning_rate": 3.831964887155264e-05,
"loss": 0.4257,
"step": 270
},
{
"epoch": 1.097165991902834,
"grad_norm": 0.7463566125136203,
"learning_rate": 3.829688502793403e-05,
"loss": 0.4498,
"step": 271
},
{
"epoch": 1.1012145748987854,
"grad_norm": 0.7772942006010995,
"learning_rate": 3.827397488291029e-05,
"loss": 0.517,
"step": 272
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.7320915178200773,
"learning_rate": 3.8250918619670343e-05,
"loss": 0.4269,
"step": 273
},
{
"epoch": 1.1093117408906883,
"grad_norm": 0.81233741062334,
"learning_rate": 3.822771642257145e-05,
"loss": 0.4166,
"step": 274
},
{
"epoch": 1.1133603238866396,
"grad_norm": 0.6714669770174236,
"learning_rate": 3.820436847713776e-05,
"loss": 0.3572,
"step": 275
},
{
"epoch": 1.117408906882591,
"grad_norm": 1.5619216856362823,
"learning_rate": 3.8180874970058826e-05,
"loss": 0.4145,
"step": 276
},
{
"epoch": 1.1214574898785425,
"grad_norm": 0.7458125430891894,
"learning_rate": 3.8157236089188103e-05,
"loss": 0.408,
"step": 277
},
{
"epoch": 1.125506072874494,
"grad_norm": 0.7054607431487282,
"learning_rate": 3.813345202354145e-05,
"loss": 0.398,
"step": 278
},
{
"epoch": 1.1295546558704452,
"grad_norm": 1.1246782631559533,
"learning_rate": 3.810952296329563e-05,
"loss": 0.4284,
"step": 279
},
{
"epoch": 1.1336032388663968,
"grad_norm": 0.9730864097755725,
"learning_rate": 3.808544909978676e-05,
"loss": 0.4252,
"step": 280
},
{
"epoch": 1.1376518218623481,
"grad_norm": 0.8323970902177198,
"learning_rate": 3.806123062550882e-05,
"loss": 0.461,
"step": 281
},
{
"epoch": 1.1417004048582995,
"grad_norm": 0.7963628022339492,
"learning_rate": 3.803686773411208e-05,
"loss": 0.4631,
"step": 282
},
{
"epoch": 1.145748987854251,
"grad_norm": 1.1338332340030868,
"learning_rate": 3.801236062040158e-05,
"loss": 0.4287,
"step": 283
},
{
"epoch": 1.1497975708502024,
"grad_norm": 1.1748740297593823,
"learning_rate": 3.798770948033553e-05,
"loss": 0.4759,
"step": 284
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.7851070493123286,
"learning_rate": 3.796291451102381e-05,
"loss": 0.4163,
"step": 285
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.8769467973448833,
"learning_rate": 3.793797591072631e-05,
"loss": 0.437,
"step": 286
},
{
"epoch": 1.1619433198380567,
"grad_norm": 0.95081744548635,
"learning_rate": 3.791289387885144e-05,
"loss": 0.4365,
"step": 287
},
{
"epoch": 1.165991902834008,
"grad_norm": 0.722582848933919,
"learning_rate": 3.788766861595443e-05,
"loss": 0.4121,
"step": 288
},
{
"epoch": 1.1700404858299596,
"grad_norm": 1.0008375210900828,
"learning_rate": 3.7862300323735835e-05,
"loss": 0.4437,
"step": 289
},
{
"epoch": 1.174089068825911,
"grad_norm": 1.1114122428932158,
"learning_rate": 3.783678920503983e-05,
"loss": 0.4624,
"step": 290
},
{
"epoch": 1.1781376518218623,
"grad_norm": 0.8106307189583402,
"learning_rate": 3.781113546385265e-05,
"loss": 0.4319,
"step": 291
},
{
"epoch": 1.1821862348178138,
"grad_norm": 1.0990500373013536,
"learning_rate": 3.7785339305300916e-05,
"loss": 0.4273,
"step": 292
},
{
"epoch": 1.1862348178137652,
"grad_norm": 0.8748398071748422,
"learning_rate": 3.7759400935650025e-05,
"loss": 0.5236,
"step": 293
},
{
"epoch": 1.1902834008097165,
"grad_norm": 0.8736721167388509,
"learning_rate": 3.77333205623025e-05,
"loss": 0.4474,
"step": 294
},
{
"epoch": 1.194331983805668,
"grad_norm": 0.710647869261301,
"learning_rate": 3.7707098393796295e-05,
"loss": 0.4093,
"step": 295
},
{
"epoch": 1.1983805668016194,
"grad_norm": 0.8086675138829419,
"learning_rate": 3.7680734639803185e-05,
"loss": 0.4226,
"step": 296
},
{
"epoch": 1.2024291497975708,
"grad_norm": 1.7177983634025258,
"learning_rate": 3.765422951112704e-05,
"loss": 0.4391,
"step": 297
},
{
"epoch": 1.2064777327935223,
"grad_norm": 0.8153702442007229,
"learning_rate": 3.762758321970216e-05,
"loss": 0.4131,
"step": 298
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.710136612509509,
"learning_rate": 3.7600795978591584e-05,
"loss": 0.43,
"step": 299
},
{
"epoch": 1.214574898785425,
"grad_norm": 0.9500138630672614,
"learning_rate": 3.757386800198538e-05,
"loss": 0.4637,
"step": 300
},
{
"epoch": 1.2186234817813766,
"grad_norm": 1.006350915778543,
"learning_rate": 3.7546799505198925e-05,
"loss": 0.4361,
"step": 301
},
{
"epoch": 1.222672064777328,
"grad_norm": 0.8823583743212201,
"learning_rate": 3.7519590704671197e-05,
"loss": 0.4154,
"step": 302
},
{
"epoch": 1.2267206477732793,
"grad_norm": 0.7278042477811579,
"learning_rate": 3.749224181796305e-05,
"loss": 0.4379,
"step": 303
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.7477276566060855,
"learning_rate": 3.746475306375545e-05,
"loss": 0.4339,
"step": 304
},
{
"epoch": 1.2348178137651822,
"grad_norm": 1.0237377133696974,
"learning_rate": 3.743712466184774e-05,
"loss": 0.4041,
"step": 305
},
{
"epoch": 1.2388663967611335,
"grad_norm": 1.2106729252814101,
"learning_rate": 3.7409356833155885e-05,
"loss": 0.4076,
"step": 306
},
{
"epoch": 1.242914979757085,
"grad_norm": 0.7028345432986266,
"learning_rate": 3.73814497997107e-05,
"loss": 0.4175,
"step": 307
},
{
"epoch": 1.2469635627530364,
"grad_norm": 1.165248474809656,
"learning_rate": 3.7353403784656095e-05,
"loss": 0.4417,
"step": 308
},
{
"epoch": 1.2510121457489878,
"grad_norm": 0.6676535341564642,
"learning_rate": 3.732521901224724e-05,
"loss": 0.4568,
"step": 309
},
{
"epoch": 1.2550607287449393,
"grad_norm": 0.7344321580322212,
"learning_rate": 3.7296895707848844e-05,
"loss": 0.4225,
"step": 310
},
{
"epoch": 1.2591093117408907,
"grad_norm": 0.6765582540934203,
"learning_rate": 3.7268434097933275e-05,
"loss": 0.4055,
"step": 311
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.8247096467102359,
"learning_rate": 3.723983441007882e-05,
"loss": 0.4175,
"step": 312
},
{
"epoch": 1.2672064777327936,
"grad_norm": 0.8421998642694128,
"learning_rate": 3.721109687296781e-05,
"loss": 0.4357,
"step": 313
},
{
"epoch": 1.271255060728745,
"grad_norm": 0.8173514155130153,
"learning_rate": 3.718222171638484e-05,
"loss": 0.4299,
"step": 314
},
{
"epoch": 1.2753036437246963,
"grad_norm": 0.688760432250666,
"learning_rate": 3.715320917121488e-05,
"loss": 0.4297,
"step": 315
},
{
"epoch": 1.2793522267206479,
"grad_norm": 0.7962152254510303,
"learning_rate": 3.712405946944149e-05,
"loss": 0.4428,
"step": 316
},
{
"epoch": 1.2834008097165992,
"grad_norm": 0.7721419135594909,
"learning_rate": 3.709477284414488e-05,
"loss": 0.4404,
"step": 317
},
{
"epoch": 1.2874493927125505,
"grad_norm": 2.887313050800867,
"learning_rate": 3.7065349529500136e-05,
"loss": 0.4867,
"step": 318
},
{
"epoch": 1.291497975708502,
"grad_norm": 1.2627786597355672,
"learning_rate": 3.7035789760775305e-05,
"loss": 0.4835,
"step": 319
},
{
"epoch": 1.2955465587044535,
"grad_norm": 0.6647775432444061,
"learning_rate": 3.70060937743295e-05,
"loss": 0.4244,
"step": 320
},
{
"epoch": 1.2995951417004048,
"grad_norm": 0.7857300403673259,
"learning_rate": 3.6976261807611045e-05,
"loss": 0.4572,
"step": 321
},
{
"epoch": 1.3036437246963564,
"grad_norm": 1.5512422809412043,
"learning_rate": 3.694629409915555e-05,
"loss": 0.447,
"step": 322
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.6434576931290944,
"learning_rate": 3.691619088858401e-05,
"loss": 0.4221,
"step": 323
},
{
"epoch": 1.311740890688259,
"grad_norm": 0.7116944061038822,
"learning_rate": 3.68859524166009e-05,
"loss": 0.425,
"step": 324
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.8589171238609007,
"learning_rate": 3.6855578924992226e-05,
"loss": 0.4082,
"step": 325
},
{
"epoch": 1.319838056680162,
"grad_norm": 0.6907774918289232,
"learning_rate": 3.6825070656623626e-05,
"loss": 0.4003,
"step": 326
},
{
"epoch": 1.3238866396761133,
"grad_norm": 1.2168075292424558,
"learning_rate": 3.67944278554384e-05,
"loss": 0.4238,
"step": 327
},
{
"epoch": 1.3279352226720649,
"grad_norm": 2.3910410468391987,
"learning_rate": 3.676365076645557e-05,
"loss": 0.4121,
"step": 328
},
{
"epoch": 1.3319838056680162,
"grad_norm": 0.914000809930577,
"learning_rate": 3.673273963576791e-05,
"loss": 0.3962,
"step": 329
},
{
"epoch": 1.3360323886639676,
"grad_norm": 1.100732713877438,
"learning_rate": 3.670169471054001e-05,
"loss": 0.4432,
"step": 330
},
{
"epoch": 1.3400809716599191,
"grad_norm": 0.7816843175829155,
"learning_rate": 3.6670516239006254e-05,
"loss": 0.4268,
"step": 331
},
{
"epoch": 1.3441295546558705,
"grad_norm": 0.9081763439316024,
"learning_rate": 3.663920447046888e-05,
"loss": 0.4211,
"step": 332
},
{
"epoch": 1.3481781376518218,
"grad_norm": 0.8531133269646999,
"learning_rate": 3.6607759655295954e-05,
"loss": 0.4296,
"step": 333
},
{
"epoch": 1.3522267206477734,
"grad_norm": 0.7044516871299804,
"learning_rate": 3.657618204491937e-05,
"loss": 0.4453,
"step": 334
},
{
"epoch": 1.3562753036437247,
"grad_norm": 0.7653900973293428,
"learning_rate": 3.6544471891832875e-05,
"loss": 0.4039,
"step": 335
},
{
"epoch": 1.360323886639676,
"grad_norm": 0.8182614172834239,
"learning_rate": 3.6512629449589983e-05,
"loss": 0.4518,
"step": 336
},
{
"epoch": 1.3643724696356276,
"grad_norm": 0.906764874327522,
"learning_rate": 3.6480654972802045e-05,
"loss": 0.4977,
"step": 337
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.8569385604743642,
"learning_rate": 3.644854871713611e-05,
"loss": 0.4398,
"step": 338
},
{
"epoch": 1.3724696356275303,
"grad_norm": 0.6406596202208564,
"learning_rate": 3.6416310939312924e-05,
"loss": 0.4011,
"step": 339
},
{
"epoch": 1.376518218623482,
"grad_norm": 2.176640277904198,
"learning_rate": 3.638394189710493e-05,
"loss": 0.4729,
"step": 340
},
{
"epoch": 1.3805668016194332,
"grad_norm": 0.8440677738535373,
"learning_rate": 3.635144184933412e-05,
"loss": 0.4144,
"step": 341
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.7268199513509292,
"learning_rate": 3.6318811055869994e-05,
"loss": 0.4676,
"step": 342
},
{
"epoch": 1.3886639676113361,
"grad_norm": 0.639644463990493,
"learning_rate": 3.628604977762752e-05,
"loss": 0.4565,
"step": 343
},
{
"epoch": 1.3927125506072875,
"grad_norm": 0.9994795971831902,
"learning_rate": 3.6253158276565004e-05,
"loss": 0.4083,
"step": 344
},
{
"epoch": 1.3967611336032388,
"grad_norm": 0.8015408280711819,
"learning_rate": 3.6220136815682016e-05,
"loss": 0.4568,
"step": 345
},
{
"epoch": 1.4008097165991904,
"grad_norm": 0.846366307382393,
"learning_rate": 3.618698565901727e-05,
"loss": 0.4321,
"step": 346
},
{
"epoch": 1.4048582995951417,
"grad_norm": 0.7342907132704195,
"learning_rate": 3.615370507164654e-05,
"loss": 0.4428,
"step": 347
},
{
"epoch": 1.408906882591093,
"grad_norm": 0.9726587589149706,
"learning_rate": 3.61202953196805e-05,
"loss": 0.4273,
"step": 348
},
{
"epoch": 1.4129554655870447,
"grad_norm": 2.1873826285062754,
"learning_rate": 3.6086756670262654e-05,
"loss": 0.4731,
"step": 349
},
{
"epoch": 1.417004048582996,
"grad_norm": 0.8015335714056537,
"learning_rate": 3.6053089391567135e-05,
"loss": 0.4375,
"step": 350
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.6227402733919594,
"learning_rate": 3.601929375279662e-05,
"loss": 0.4267,
"step": 351
},
{
"epoch": 1.425101214574899,
"grad_norm": 0.7952408106043131,
"learning_rate": 3.598537002418012e-05,
"loss": 0.4728,
"step": 352
},
{
"epoch": 1.4291497975708503,
"grad_norm": 0.7992024767836873,
"learning_rate": 3.5951318476970885e-05,
"loss": 0.4327,
"step": 353
},
{
"epoch": 1.4331983805668016,
"grad_norm": 1.0111561103371407,
"learning_rate": 3.591713938344416e-05,
"loss": 0.4641,
"step": 354
},
{
"epoch": 1.4372469635627532,
"grad_norm": 0.6596539297364921,
"learning_rate": 3.588283301689507e-05,
"loss": 0.4273,
"step": 355
},
{
"epoch": 1.4412955465587045,
"grad_norm": 0.6900582387357528,
"learning_rate": 3.5848399651636424e-05,
"loss": 0.3822,
"step": 356
},
{
"epoch": 1.4453441295546559,
"grad_norm": 0.8988065965128225,
"learning_rate": 3.5813839562996473e-05,
"loss": 0.4112,
"step": 357
},
{
"epoch": 1.4493927125506074,
"grad_norm": 0.6931751458740782,
"learning_rate": 3.5779153027316775e-05,
"loss": 0.3999,
"step": 358
},
{
"epoch": 1.4534412955465588,
"grad_norm": 0.7319342221162741,
"learning_rate": 3.574434032194994e-05,
"loss": 0.4131,
"step": 359
},
{
"epoch": 1.45748987854251,
"grad_norm": 0.8581700627783422,
"learning_rate": 3.5709401725257436e-05,
"loss": 0.4814,
"step": 360
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.7775363418095905,
"learning_rate": 3.5674337516607346e-05,
"loss": 0.4176,
"step": 361
},
{
"epoch": 1.465587044534413,
"grad_norm": 0.8636208483451866,
"learning_rate": 3.563914797637216e-05,
"loss": 0.5025,
"step": 362
},
{
"epoch": 1.4696356275303644,
"grad_norm": 0.7651114026401926,
"learning_rate": 3.560383338592649e-05,
"loss": 0.3993,
"step": 363
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.7488479782925682,
"learning_rate": 3.556839402764487e-05,
"loss": 0.4557,
"step": 364
},
{
"epoch": 1.4777327935222673,
"grad_norm": 1.0720594684348435,
"learning_rate": 3.553283018489946e-05,
"loss": 0.4577,
"step": 365
},
{
"epoch": 1.4817813765182186,
"grad_norm": 0.9488342182218016,
"learning_rate": 3.54971421420578e-05,
"loss": 0.4388,
"step": 366
},
{
"epoch": 1.48582995951417,
"grad_norm": 0.7343661687706139,
"learning_rate": 3.5461330184480524e-05,
"loss": 0.4081,
"step": 367
},
{
"epoch": 1.4898785425101215,
"grad_norm": 1.12668381277185,
"learning_rate": 3.54253945985191e-05,
"loss": 0.4242,
"step": 368
},
{
"epoch": 1.4939271255060729,
"grad_norm": 0.7389116679603945,
"learning_rate": 3.5389335671513524e-05,
"loss": 0.4177,
"step": 369
},
{
"epoch": 1.4979757085020242,
"grad_norm": 1.035272476638392,
"learning_rate": 3.5353153691789986e-05,
"loss": 0.481,
"step": 370
},
{
"epoch": 1.5020242914979756,
"grad_norm": 0.8624385373889963,
"learning_rate": 3.531684894865866e-05,
"loss": 0.4628,
"step": 371
},
{
"epoch": 1.5060728744939271,
"grad_norm": 0.7906471874439008,
"learning_rate": 3.528042173241128e-05,
"loss": 0.4509,
"step": 372
},
{
"epoch": 1.5101214574898787,
"grad_norm": 0.7421535319408779,
"learning_rate": 3.524387233431892e-05,
"loss": 0.4163,
"step": 373
},
{
"epoch": 1.5141700404858298,
"grad_norm": 0.7751536829375621,
"learning_rate": 3.520720104662958e-05,
"loss": 0.4305,
"step": 374
},
{
"epoch": 1.5182186234817814,
"grad_norm": 0.7442127284038789,
"learning_rate": 3.5170408162565904e-05,
"loss": 0.4489,
"step": 375
},
{
"epoch": 1.522267206477733,
"grad_norm": 1.0958933785723604,
"learning_rate": 3.513349397632281e-05,
"loss": 0.4657,
"step": 376
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.8479302638631345,
"learning_rate": 3.5096458783065145e-05,
"loss": 0.4247,
"step": 377
},
{
"epoch": 1.5303643724696356,
"grad_norm": 0.8496443985597879,
"learning_rate": 3.505930287892533e-05,
"loss": 0.4068,
"step": 378
},
{
"epoch": 1.5344129554655872,
"grad_norm": 0.9095740516298543,
"learning_rate": 3.502202656100099e-05,
"loss": 0.4867,
"step": 379
},
{
"epoch": 1.5384615384615383,
"grad_norm": 1.609398765548226,
"learning_rate": 3.498463012735256e-05,
"loss": 0.4517,
"step": 380
},
{
"epoch": 1.54251012145749,
"grad_norm": 0.7960372793064289,
"learning_rate": 3.494711387700094e-05,
"loss": 0.4262,
"step": 381
},
{
"epoch": 1.5465587044534415,
"grad_norm": 0.7434911414311022,
"learning_rate": 3.490947810992505e-05,
"loss": 0.4262,
"step": 382
},
{
"epoch": 1.5506072874493926,
"grad_norm": 1.7555996266520217,
"learning_rate": 3.48717231270595e-05,
"loss": 0.4612,
"step": 383
},
{
"epoch": 1.5546558704453441,
"grad_norm": 0.8555761141623766,
"learning_rate": 3.483384923029211e-05,
"loss": 0.437,
"step": 384
},
{
"epoch": 1.5587044534412957,
"grad_norm": 0.8445044263819196,
"learning_rate": 3.479585672246156e-05,
"loss": 0.4535,
"step": 385
},
{
"epoch": 1.5627530364372468,
"grad_norm": 1.0097020936968264,
"learning_rate": 3.4757745907354924e-05,
"loss": 0.4168,
"step": 386
},
{
"epoch": 1.5668016194331984,
"grad_norm": 0.8783015254106388,
"learning_rate": 3.471951708970527e-05,
"loss": 0.4118,
"step": 387
},
{
"epoch": 1.5708502024291497,
"grad_norm": 0.7212064822525112,
"learning_rate": 3.468117057518921e-05,
"loss": 0.455,
"step": 388
},
{
"epoch": 1.574898785425101,
"grad_norm": 1.132305209505929,
"learning_rate": 3.4642706670424426e-05,
"loss": 0.4917,
"step": 389
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.8174770423070618,
"learning_rate": 3.460412568296731e-05,
"loss": 0.4481,
"step": 390
},
{
"epoch": 1.582995951417004,
"grad_norm": 0.8857058091892888,
"learning_rate": 3.456542792131039e-05,
"loss": 0.4299,
"step": 391
},
{
"epoch": 1.5870445344129553,
"grad_norm": 0.9625957300602553,
"learning_rate": 3.4526613694879936e-05,
"loss": 0.4585,
"step": 392
},
{
"epoch": 1.591093117408907,
"grad_norm": 0.7992746119326002,
"learning_rate": 3.448768331403348e-05,
"loss": 0.4077,
"step": 393
},
{
"epoch": 1.5951417004048583,
"grad_norm": 1.0931847765555716,
"learning_rate": 3.44486370900573e-05,
"loss": 0.4581,
"step": 394
},
{
"epoch": 1.5991902834008096,
"grad_norm": 0.7505087445950795,
"learning_rate": 3.440947533516396e-05,
"loss": 0.4015,
"step": 395
},
{
"epoch": 1.6032388663967612,
"grad_norm": 0.744656334933216,
"learning_rate": 3.4370198362489816e-05,
"loss": 0.4535,
"step": 396
},
{
"epoch": 1.6072874493927125,
"grad_norm": 1.5688111288219428,
"learning_rate": 3.43308064860925e-05,
"loss": 0.4397,
"step": 397
},
{
"epoch": 1.6113360323886639,
"grad_norm": 0.9938235700871144,
"learning_rate": 3.4291300020948393e-05,
"loss": 0.4531,
"step": 398
},
{
"epoch": 1.6153846153846154,
"grad_norm": 1.0750829995513678,
"learning_rate": 3.425167928295015e-05,
"loss": 0.4662,
"step": 399
},
{
"epoch": 1.6194331983805668,
"grad_norm": 0.6390835265795476,
"learning_rate": 3.421194458890411e-05,
"loss": 0.48,
"step": 400
},
{
"epoch": 1.623481781376518,
"grad_norm": 2.3808679750624218,
"learning_rate": 3.4172096256527865e-05,
"loss": 0.4211,
"step": 401
},
{
"epoch": 1.6275303643724697,
"grad_norm": 0.7296682118384447,
"learning_rate": 3.4132134604447595e-05,
"loss": 0.4431,
"step": 402
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.8506495420311206,
"learning_rate": 3.409205995219561e-05,
"loss": 0.4414,
"step": 403
},
{
"epoch": 1.6356275303643724,
"grad_norm": 1.101768509927368,
"learning_rate": 3.4051872620207765e-05,
"loss": 0.3989,
"step": 404
},
{
"epoch": 1.639676113360324,
"grad_norm": 1.6202057632885984,
"learning_rate": 3.40115729298209e-05,
"loss": 0.4524,
"step": 405
},
{
"epoch": 1.6437246963562753,
"grad_norm": 1.1462579140421085,
"learning_rate": 3.397116120327025e-05,
"loss": 0.4503,
"step": 406
},
{
"epoch": 1.6477732793522266,
"grad_norm": 0.7850179451984087,
"learning_rate": 3.3930637763686916e-05,
"loss": 0.4351,
"step": 407
},
{
"epoch": 1.6518218623481782,
"grad_norm": 1.5947847436938563,
"learning_rate": 3.389000293509524e-05,
"loss": 0.4721,
"step": 408
},
{
"epoch": 1.6558704453441295,
"grad_norm": 1.0599622832897608,
"learning_rate": 3.3849257042410225e-05,
"loss": 0.4719,
"step": 409
},
{
"epoch": 1.6599190283400809,
"grad_norm": 0.9810086769609722,
"learning_rate": 3.380840041143494e-05,
"loss": 0.4402,
"step": 410
},
{
"epoch": 1.6639676113360324,
"grad_norm": 0.8086445498323963,
"learning_rate": 3.376743336885792e-05,
"loss": 0.4467,
"step": 411
},
{
"epoch": 1.6680161943319838,
"grad_norm": 0.8117605063140771,
"learning_rate": 3.372635624225055e-05,
"loss": 0.4286,
"step": 412
},
{
"epoch": 1.6720647773279351,
"grad_norm": 0.7819254067524829,
"learning_rate": 3.3685169360064445e-05,
"loss": 0.419,
"step": 413
},
{
"epoch": 1.6761133603238867,
"grad_norm": 0.7662172709118422,
"learning_rate": 3.3643873051628804e-05,
"loss": 0.432,
"step": 414
},
{
"epoch": 1.680161943319838,
"grad_norm": 0.8485862583889224,
"learning_rate": 3.3602467647147814e-05,
"loss": 0.4815,
"step": 415
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.7413156383836643,
"learning_rate": 3.356095347769799e-05,
"loss": 0.5279,
"step": 416
},
{
"epoch": 1.688259109311741,
"grad_norm": 0.9528390904054924,
"learning_rate": 3.351933087522552e-05,
"loss": 0.4122,
"step": 417
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.7751208944639435,
"learning_rate": 3.3477600172543626e-05,
"loss": 0.4947,
"step": 418
},
{
"epoch": 1.6963562753036436,
"grad_norm": 0.9544347809720647,
"learning_rate": 3.3435761703329894e-05,
"loss": 0.4691,
"step": 419
},
{
"epoch": 1.7004048582995952,
"grad_norm": 0.7774648418299829,
"learning_rate": 3.3393815802123606e-05,
"loss": 0.4313,
"step": 420
},
{
"epoch": 1.7044534412955465,
"grad_norm": 1.0153840354846804,
"learning_rate": 3.335176280432308e-05,
"loss": 0.4089,
"step": 421
},
{
"epoch": 1.708502024291498,
"grad_norm": 0.7355128286653033,
"learning_rate": 3.330960304618295e-05,
"loss": 0.4099,
"step": 422
},
{
"epoch": 1.7125506072874495,
"grad_norm": 0.8740367348861144,
"learning_rate": 3.326733686481153e-05,
"loss": 0.3947,
"step": 423
},
{
"epoch": 1.7165991902834008,
"grad_norm": 0.7298365141486302,
"learning_rate": 3.3224964598168084e-05,
"loss": 0.4212,
"step": 424
},
{
"epoch": 1.7206477732793521,
"grad_norm": 0.7252064669958557,
"learning_rate": 3.318248658506012e-05,
"loss": 0.4112,
"step": 425
},
{
"epoch": 1.7246963562753037,
"grad_norm": 0.8259047106363374,
"learning_rate": 3.313990316514069e-05,
"loss": 0.4817,
"step": 426
},
{
"epoch": 1.728744939271255,
"grad_norm": 0.7354669929476801,
"learning_rate": 3.309721467890571e-05,
"loss": 0.4306,
"step": 427
},
{
"epoch": 1.7327935222672064,
"grad_norm": 0.7233702997318384,
"learning_rate": 3.305442146769115e-05,
"loss": 0.4543,
"step": 428
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.7109015826374887,
"learning_rate": 3.30115238736704e-05,
"loss": 0.4154,
"step": 429
},
{
"epoch": 1.7408906882591093,
"grad_norm": 0.7139342903495831,
"learning_rate": 3.296852223985148e-05,
"loss": 0.4674,
"step": 430
},
{
"epoch": 1.7449392712550607,
"grad_norm": 1.5750178703504834,
"learning_rate": 3.2925416910074315e-05,
"loss": 0.4593,
"step": 431
},
{
"epoch": 1.7489878542510122,
"grad_norm": 0.7074778602110503,
"learning_rate": 3.288220822900796e-05,
"loss": 0.4319,
"step": 432
},
{
"epoch": 1.7530364372469636,
"grad_norm": 0.6048926595955478,
"learning_rate": 3.283889654214788e-05,
"loss": 0.4169,
"step": 433
},
{
"epoch": 1.757085020242915,
"grad_norm": 0.7418080507524061,
"learning_rate": 3.279548219581317e-05,
"loss": 0.4158,
"step": 434
},
{
"epoch": 1.7611336032388665,
"grad_norm": 0.7364191284826843,
"learning_rate": 3.275196553714379e-05,
"loss": 0.438,
"step": 435
},
{
"epoch": 1.7651821862348178,
"grad_norm": 0.892600609887849,
"learning_rate": 3.270834691409778e-05,
"loss": 0.4284,
"step": 436
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.7078061769468528,
"learning_rate": 3.266462667544849e-05,
"loss": 0.4611,
"step": 437
},
{
"epoch": 1.7732793522267207,
"grad_norm": 0.6918750472457025,
"learning_rate": 3.262080517078178e-05,
"loss": 0.4241,
"step": 438
},
{
"epoch": 1.777327935222672,
"grad_norm": 0.7278349205145843,
"learning_rate": 3.257688275049323e-05,
"loss": 0.4487,
"step": 439
},
{
"epoch": 1.7813765182186234,
"grad_norm": 1.0617738432661417,
"learning_rate": 3.253285976578535e-05,
"loss": 0.4531,
"step": 440
},
{
"epoch": 1.785425101214575,
"grad_norm": 0.8958466802903048,
"learning_rate": 3.2488736568664756e-05,
"loss": 0.4269,
"step": 441
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.9346673476414686,
"learning_rate": 3.244451351193935e-05,
"loss": 0.481,
"step": 442
},
{
"epoch": 1.7935222672064777,
"grad_norm": 0.9554052984930238,
"learning_rate": 3.240019094921551e-05,
"loss": 0.4443,
"step": 443
},
{
"epoch": 1.7975708502024292,
"grad_norm": 0.7490113506578192,
"learning_rate": 3.235576923489527e-05,
"loss": 0.4278,
"step": 444
},
{
"epoch": 1.8016194331983806,
"grad_norm": 0.8105955486918456,
"learning_rate": 3.2311248724173476e-05,
"loss": 0.4254,
"step": 445
},
{
"epoch": 1.805668016194332,
"grad_norm": 0.9018708065343257,
"learning_rate": 3.226662977303494e-05,
"loss": 0.4553,
"step": 446
},
{
"epoch": 1.8097165991902835,
"grad_norm": 1.061363908890884,
"learning_rate": 3.2221912738251626e-05,
"loss": 0.4525,
"step": 447
},
{
"epoch": 1.8137651821862348,
"grad_norm": 0.7514141824085137,
"learning_rate": 3.2177097977379736e-05,
"loss": 0.4372,
"step": 448
},
{
"epoch": 1.8178137651821862,
"grad_norm": 1.1844664421472848,
"learning_rate": 3.21321858487569e-05,
"loss": 0.4516,
"step": 449
},
{
"epoch": 1.8218623481781377,
"grad_norm": 0.8564640570844934,
"learning_rate": 3.208717671149931e-05,
"loss": 0.4174,
"step": 450
},
{
"epoch": 1.825910931174089,
"grad_norm": 0.7533209692717147,
"learning_rate": 3.204207092549882e-05,
"loss": 0.4342,
"step": 451
},
{
"epoch": 1.8299595141700404,
"grad_norm": 1.1278937690511064,
"learning_rate": 3.19968688514201e-05,
"loss": 0.4801,
"step": 452
},
{
"epoch": 1.834008097165992,
"grad_norm": 1.1056319551163027,
"learning_rate": 3.195157085069774e-05,
"loss": 0.4351,
"step": 453
},
{
"epoch": 1.8380566801619433,
"grad_norm": 0.923909409467763,
"learning_rate": 3.190617728553332e-05,
"loss": 0.4354,
"step": 454
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.8744770411709978,
"learning_rate": 3.1860688518892606e-05,
"loss": 0.4234,
"step": 455
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.6512757271404485,
"learning_rate": 3.181510491450254e-05,
"loss": 0.4148,
"step": 456
},
{
"epoch": 1.8502024291497976,
"grad_norm": 0.8748212783167156,
"learning_rate": 3.176942683684842e-05,
"loss": 0.4266,
"step": 457
},
{
"epoch": 1.854251012145749,
"grad_norm": 0.7589889060753323,
"learning_rate": 3.1723654651170934e-05,
"loss": 0.4347,
"step": 458
},
{
"epoch": 1.8582995951417005,
"grad_norm": 0.7617932997848421,
"learning_rate": 3.167778872346327e-05,
"loss": 0.3825,
"step": 459
},
{
"epoch": 1.8623481781376519,
"grad_norm": 0.665605645341861,
"learning_rate": 3.163182942046815e-05,
"loss": 0.403,
"step": 460
},
{
"epoch": 1.8663967611336032,
"grad_norm": 0.6708314493391023,
"learning_rate": 3.158577710967494e-05,
"loss": 0.4209,
"step": 461
},
{
"epoch": 1.8704453441295548,
"grad_norm": 0.9277455462322933,
"learning_rate": 3.153963215931669e-05,
"loss": 0.3961,
"step": 462
},
{
"epoch": 1.874493927125506,
"grad_norm": 0.9039144092558545,
"learning_rate": 3.14933949383672e-05,
"loss": 0.4148,
"step": 463
},
{
"epoch": 1.8785425101214575,
"grad_norm": 0.7523991281084527,
"learning_rate": 3.144706581653805e-05,
"loss": 0.4154,
"step": 464
},
{
"epoch": 1.882591093117409,
"grad_norm": 0.7744901876544801,
"learning_rate": 3.140064516427566e-05,
"loss": 0.4236,
"step": 465
},
{
"epoch": 1.8866396761133604,
"grad_norm": 0.821665343703843,
"learning_rate": 3.135413335275833e-05,
"loss": 0.4461,
"step": 466
},
{
"epoch": 1.8906882591093117,
"grad_norm": 1.1003077623912663,
"learning_rate": 3.130753075389327e-05,
"loss": 0.419,
"step": 467
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.7378985296300666,
"learning_rate": 3.1260837740313625e-05,
"loss": 0.4894,
"step": 468
},
{
"epoch": 1.8987854251012146,
"grad_norm": 0.6357281853129614,
"learning_rate": 3.121405468537547e-05,
"loss": 0.4207,
"step": 469
},
{
"epoch": 1.902834008097166,
"grad_norm": 0.8668507994318915,
"learning_rate": 3.116718196315488e-05,
"loss": 0.4318,
"step": 470
},
{
"epoch": 1.9068825910931175,
"grad_norm": 0.815037495864932,
"learning_rate": 3.1120219948444894e-05,
"loss": 0.4242,
"step": 471
},
{
"epoch": 1.9109311740890689,
"grad_norm": 1.0223546312345944,
"learning_rate": 3.107316901675252e-05,
"loss": 0.449,
"step": 472
},
{
"epoch": 1.9149797570850202,
"grad_norm": 0.7575359342496305,
"learning_rate": 3.1026029544295745e-05,
"loss": 0.4129,
"step": 473
},
{
"epoch": 1.9190283400809718,
"grad_norm": 0.7781206679154609,
"learning_rate": 3.097880190800057e-05,
"loss": 0.4216,
"step": 474
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.8980732833339586,
"learning_rate": 3.093148648549788e-05,
"loss": 0.4468,
"step": 475
},
{
"epoch": 1.9271255060728745,
"grad_norm": 0.7128827568580286,
"learning_rate": 3.088408365512055e-05,
"loss": 0.4434,
"step": 476
},
{
"epoch": 1.931174089068826,
"grad_norm": 0.7988748617051469,
"learning_rate": 3.083659379590034e-05,
"loss": 0.454,
"step": 477
},
{
"epoch": 1.9352226720647774,
"grad_norm": 0.6251671389784828,
"learning_rate": 3.0789017287564913e-05,
"loss": 0.4101,
"step": 478
},
{
"epoch": 1.9392712550607287,
"grad_norm": 0.7117291695878791,
"learning_rate": 3.074135451053475e-05,
"loss": 0.4033,
"step": 479
},
{
"epoch": 1.9433198380566803,
"grad_norm": 0.8784164120980272,
"learning_rate": 3.069360584592015e-05,
"loss": 0.4472,
"step": 480
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.6548145140057734,
"learning_rate": 3.064577167551816e-05,
"loss": 0.4374,
"step": 481
},
{
"epoch": 1.951417004048583,
"grad_norm": 1.279004106660506,
"learning_rate": 3.059785238180953e-05,
"loss": 0.4335,
"step": 482
},
{
"epoch": 1.9554655870445345,
"grad_norm": 0.7529746393575133,
"learning_rate": 3.054984834795566e-05,
"loss": 0.4073,
"step": 483
},
{
"epoch": 1.9595141700404857,
"grad_norm": 0.6818439100894281,
"learning_rate": 3.0501759957795537e-05,
"loss": 0.4135,
"step": 484
},
{
"epoch": 1.9635627530364372,
"grad_norm": 0.7764179430535826,
"learning_rate": 3.0453587595842627e-05,
"loss": 0.4427,
"step": 485
},
{
"epoch": 1.9676113360323888,
"grad_norm": 0.64050740818794,
"learning_rate": 3.0405331647281856e-05,
"loss": 0.4101,
"step": 486
},
{
"epoch": 1.97165991902834,
"grad_norm": 0.7215229352949164,
"learning_rate": 3.0356992497966508e-05,
"loss": 0.4782,
"step": 487
},
{
"epoch": 1.9757085020242915,
"grad_norm": 0.8367993685667381,
"learning_rate": 3.0308570534415137e-05,
"loss": 0.411,
"step": 488
},
{
"epoch": 1.979757085020243,
"grad_norm": 0.7746546957730409,
"learning_rate": 3.0260066143808467e-05,
"loss": 0.4351,
"step": 489
},
{
"epoch": 1.9838056680161942,
"grad_norm": 0.9226709655649625,
"learning_rate": 3.021147971398632e-05,
"loss": 0.4263,
"step": 490
},
{
"epoch": 1.9878542510121457,
"grad_norm": 0.8348490519992745,
"learning_rate": 3.0162811633444478e-05,
"loss": 0.4241,
"step": 491
},
{
"epoch": 1.9919028340080973,
"grad_norm": 0.7198739027490211,
"learning_rate": 3.0114062291331635e-05,
"loss": 0.4406,
"step": 492
},
{
"epoch": 1.9959514170040484,
"grad_norm": 0.7438974188523678,
"learning_rate": 3.0065232077446217e-05,
"loss": 0.4461,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.8431675341144987,
"learning_rate": 3.001632138223332e-05,
"loss": 0.4376,
"step": 494
},
{
"epoch": 2.0040485829959516,
"grad_norm": 0.7201048536540752,
"learning_rate": 2.9967330596781565e-05,
"loss": 0.3399,
"step": 495
},
{
"epoch": 2.0080971659919027,
"grad_norm": 0.5877545118720884,
"learning_rate": 2.9918260112819952e-05,
"loss": 0.2788,
"step": 496
},
{
"epoch": 2.0121457489878543,
"grad_norm": 0.6073617905808089,
"learning_rate": 2.9869110322714778e-05,
"loss": 0.2782,
"step": 497
},
{
"epoch": 2.016194331983806,
"grad_norm": 0.8928263374618776,
"learning_rate": 2.9819881619466447e-05,
"loss": 0.3539,
"step": 498
},
{
"epoch": 2.020242914979757,
"grad_norm": 0.7138852256406425,
"learning_rate": 2.9770574396706364e-05,
"loss": 0.2807,
"step": 499
},
{
"epoch": 2.0242914979757085,
"grad_norm": 0.6680098979511734,
"learning_rate": 2.972118904869376e-05,
"loss": 0.2883,
"step": 500
},
{
"epoch": 2.02834008097166,
"grad_norm": 0.6507116058718492,
"learning_rate": 2.9671725970312563e-05,
"loss": 0.278,
"step": 501
},
{
"epoch": 2.032388663967611,
"grad_norm": 0.7906196481654697,
"learning_rate": 2.9622185557068238e-05,
"loss": 0.2877,
"step": 502
},
{
"epoch": 2.0364372469635628,
"grad_norm": 0.6124766474235545,
"learning_rate": 2.95725682050846e-05,
"loss": 0.2696,
"step": 503
},
{
"epoch": 2.0404858299595143,
"grad_norm": 0.5197626000958363,
"learning_rate": 2.9522874311100684e-05,
"loss": 0.28,
"step": 504
},
{
"epoch": 2.0445344129554655,
"grad_norm": 0.8581941208601401,
"learning_rate": 2.9473104272467526e-05,
"loss": 0.3001,
"step": 505
},
{
"epoch": 2.048582995951417,
"grad_norm": 0.6605342359007983,
"learning_rate": 2.942325848714505e-05,
"loss": 0.3051,
"step": 506
},
{
"epoch": 2.0526315789473686,
"grad_norm": 0.5647248318366876,
"learning_rate": 2.9373337353698804e-05,
"loss": 0.2746,
"step": 507
},
{
"epoch": 2.0566801619433197,
"grad_norm": 0.8990682776120723,
"learning_rate": 2.932334127129686e-05,
"loss": 0.2914,
"step": 508
},
{
"epoch": 2.0607287449392713,
"grad_norm": 0.53797063751215,
"learning_rate": 2.9273270639706547e-05,
"loss": 0.2671,
"step": 509
},
{
"epoch": 2.064777327935223,
"grad_norm": 0.6054236521746998,
"learning_rate": 2.922312585929131e-05,
"loss": 0.2737,
"step": 510
},
{
"epoch": 2.068825910931174,
"grad_norm": 0.687484973479785,
"learning_rate": 2.9172907331007478e-05,
"loss": 0.277,
"step": 511
},
{
"epoch": 2.0728744939271255,
"grad_norm": 0.5103051236105406,
"learning_rate": 2.9122615456401053e-05,
"loss": 0.2952,
"step": 512
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.7344712681626743,
"learning_rate": 2.9072250637604535e-05,
"loss": 0.3157,
"step": 513
},
{
"epoch": 2.080971659919028,
"grad_norm": 0.729220622554582,
"learning_rate": 2.9021813277333667e-05,
"loss": 0.2494,
"step": 514
},
{
"epoch": 2.08502024291498,
"grad_norm": 0.8168253431373407,
"learning_rate": 2.8971303778884244e-05,
"loss": 0.2717,
"step": 515
},
{
"epoch": 2.0890688259109313,
"grad_norm": 1.0120758057842474,
"learning_rate": 2.8920722546128863e-05,
"loss": 0.2823,
"step": 516
},
{
"epoch": 2.0931174089068825,
"grad_norm": 0.5449987018586017,
"learning_rate": 2.887006998351372e-05,
"loss": 0.2939,
"step": 517
},
{
"epoch": 2.097165991902834,
"grad_norm": 0.5137328009884246,
"learning_rate": 2.881934649605535e-05,
"loss": 0.2566,
"step": 518
},
{
"epoch": 2.1012145748987856,
"grad_norm": 0.6918618825082945,
"learning_rate": 2.8768552489337408e-05,
"loss": 0.3164,
"step": 519
},
{
"epoch": 2.1052631578947367,
"grad_norm": 0.5821829525142211,
"learning_rate": 2.8717688369507425e-05,
"loss": 0.283,
"step": 520
},
{
"epoch": 2.1093117408906883,
"grad_norm": 0.5571799024134502,
"learning_rate": 2.8666754543273535e-05,
"loss": 0.2729,
"step": 521
},
{
"epoch": 2.11336032388664,
"grad_norm": 0.5266104113829136,
"learning_rate": 2.8615751417901273e-05,
"loss": 0.2809,
"step": 522
},
{
"epoch": 2.117408906882591,
"grad_norm": 0.6458947195954288,
"learning_rate": 2.856467940121025e-05,
"loss": 0.27,
"step": 523
},
{
"epoch": 2.1214574898785425,
"grad_norm": 0.6286568928746501,
"learning_rate": 2.8513538901570964e-05,
"loss": 0.2552,
"step": 524
},
{
"epoch": 2.125506072874494,
"grad_norm": 0.5687773148210501,
"learning_rate": 2.846233032790149e-05,
"loss": 0.2723,
"step": 525
},
{
"epoch": 2.1295546558704452,
"grad_norm": 0.5773403665492544,
"learning_rate": 2.84110540896642e-05,
"loss": 0.2791,
"step": 526
},
{
"epoch": 2.133603238866397,
"grad_norm": 0.6008136310153109,
"learning_rate": 2.8359710596862555e-05,
"loss": 0.3125,
"step": 527
},
{
"epoch": 2.1376518218623484,
"grad_norm": 0.5854683712454358,
"learning_rate": 2.8308300260037734e-05,
"loss": 0.2512,
"step": 528
},
{
"epoch": 2.1417004048582995,
"grad_norm": 0.8686319732463594,
"learning_rate": 2.8256823490265424e-05,
"loss": 0.2765,
"step": 529
},
{
"epoch": 2.145748987854251,
"grad_norm": 0.5621216674902523,
"learning_rate": 2.8205280699152515e-05,
"loss": 0.2587,
"step": 530
},
{
"epoch": 2.1497975708502026,
"grad_norm": 0.6642016119075943,
"learning_rate": 2.8153672298833777e-05,
"loss": 0.2606,
"step": 531
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.5847723637415193,
"learning_rate": 2.8101998701968615e-05,
"loss": 0.2938,
"step": 532
},
{
"epoch": 2.1578947368421053,
"grad_norm": 1.4253324768462141,
"learning_rate": 2.805026032173773e-05,
"loss": 0.2453,
"step": 533
},
{
"epoch": 2.161943319838057,
"grad_norm": 0.4813856768486233,
"learning_rate": 2.7998457571839832e-05,
"loss": 0.2705,
"step": 534
},
{
"epoch": 2.165991902834008,
"grad_norm": 0.5679109302150801,
"learning_rate": 2.7946590866488322e-05,
"loss": 0.2996,
"step": 535
},
{
"epoch": 2.1700404858299596,
"grad_norm": 0.986201981585674,
"learning_rate": 2.7894660620408e-05,
"loss": 0.2916,
"step": 536
},
{
"epoch": 2.174089068825911,
"grad_norm": 1.054357378581762,
"learning_rate": 2.784266724883173e-05,
"loss": 0.2676,
"step": 537
},
{
"epoch": 2.1781376518218623,
"grad_norm": 0.5903137702574576,
"learning_rate": 2.7790611167497128e-05,
"loss": 0.2533,
"step": 538
},
{
"epoch": 2.182186234817814,
"grad_norm": 0.49312627054357666,
"learning_rate": 2.7738492792643224e-05,
"loss": 0.2731,
"step": 539
},
{
"epoch": 2.1862348178137654,
"grad_norm": 0.5838516079261789,
"learning_rate": 2.768631254100716e-05,
"loss": 0.2842,
"step": 540
},
{
"epoch": 2.1902834008097165,
"grad_norm": 0.502205973301598,
"learning_rate": 2.7634070829820836e-05,
"loss": 0.2581,
"step": 541
},
{
"epoch": 2.194331983805668,
"grad_norm": 0.6344674441769595,
"learning_rate": 2.7581768076807592e-05,
"loss": 0.2972,
"step": 542
},
{
"epoch": 2.1983805668016196,
"grad_norm": 0.5153053244435004,
"learning_rate": 2.7529404700178832e-05,
"loss": 0.2962,
"step": 543
},
{
"epoch": 2.2024291497975708,
"grad_norm": 0.5569687599553506,
"learning_rate": 2.7476981118630735e-05,
"loss": 0.2926,
"step": 544
},
{
"epoch": 2.2064777327935223,
"grad_norm": 0.6942269030413322,
"learning_rate": 2.742449775134085e-05,
"loss": 0.2756,
"step": 545
},
{
"epoch": 2.2105263157894735,
"grad_norm": 0.5547890129990676,
"learning_rate": 2.7371955017964788e-05,
"loss": 0.3012,
"step": 546
},
{
"epoch": 2.214574898785425,
"grad_norm": 0.45532589946392144,
"learning_rate": 2.7319353338632845e-05,
"loss": 0.2813,
"step": 547
},
{
"epoch": 2.2186234817813766,
"grad_norm": 0.6454738620170909,
"learning_rate": 2.7266693133946645e-05,
"loss": 0.2898,
"step": 548
},
{
"epoch": 2.2226720647773277,
"grad_norm": 0.5526720376652118,
"learning_rate": 2.7213974824975785e-05,
"loss": 0.2742,
"step": 549
},
{
"epoch": 2.2267206477732793,
"grad_norm": 0.6189784294712907,
"learning_rate": 2.7161198833254464e-05,
"loss": 0.2284,
"step": 550
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.48178920290813204,
"learning_rate": 2.7108365580778095e-05,
"loss": 0.3083,
"step": 551
},
{
"epoch": 2.234817813765182,
"grad_norm": 0.504089620156236,
"learning_rate": 2.705547548999996e-05,
"loss": 0.2942,
"step": 552
},
{
"epoch": 2.2388663967611335,
"grad_norm": 0.5328991349183593,
"learning_rate": 2.7002528983827817e-05,
"loss": 0.2832,
"step": 553
},
{
"epoch": 2.242914979757085,
"grad_norm": 0.6328309150650496,
"learning_rate": 2.6949526485620517e-05,
"loss": 0.3052,
"step": 554
},
{
"epoch": 2.246963562753036,
"grad_norm": 0.5481808880002431,
"learning_rate": 2.6896468419184623e-05,
"loss": 0.2879,
"step": 555
},
{
"epoch": 2.251012145748988,
"grad_norm": 0.6342144374937204,
"learning_rate": 2.6843355208771024e-05,
"loss": 0.2556,
"step": 556
},
{
"epoch": 2.2550607287449393,
"grad_norm": 0.8835867515447996,
"learning_rate": 2.6790187279071536e-05,
"loss": 0.2822,
"step": 557
},
{
"epoch": 2.2591093117408905,
"grad_norm": 2.5752586376002893,
"learning_rate": 2.67369650552155e-05,
"loss": 0.2659,
"step": 558
},
{
"epoch": 2.263157894736842,
"grad_norm": 0.6129150179590822,
"learning_rate": 2.6683688962766407e-05,
"loss": 0.2469,
"step": 559
},
{
"epoch": 2.2672064777327936,
"grad_norm": 0.7131633505354905,
"learning_rate": 2.6630359427718468e-05,
"loss": 0.2807,
"step": 560
},
{
"epoch": 2.2712550607287447,
"grad_norm": 0.6077310002139935,
"learning_rate": 2.6576976876493225e-05,
"loss": 0.2871,
"step": 561
},
{
"epoch": 2.2753036437246963,
"grad_norm": 0.5433210008901572,
"learning_rate": 2.652354173593614e-05,
"loss": 0.2829,
"step": 562
},
{
"epoch": 2.279352226720648,
"grad_norm": 0.5067877136359286,
"learning_rate": 2.647005443331317e-05,
"loss": 0.2763,
"step": 563
},
{
"epoch": 2.283400809716599,
"grad_norm": 0.5888452334550435,
"learning_rate": 2.6416515396307354e-05,
"loss": 0.3572,
"step": 564
},
{
"epoch": 2.2874493927125505,
"grad_norm": 0.5456481370872392,
"learning_rate": 2.636292505301542e-05,
"loss": 0.2755,
"step": 565
},
{
"epoch": 2.291497975708502,
"grad_norm": 0.6515323180835961,
"learning_rate": 2.630928383194431e-05,
"loss": 0.2815,
"step": 566
},
{
"epoch": 2.2955465587044532,
"grad_norm": 0.6212781071595338,
"learning_rate": 2.625559216200781e-05,
"loss": 0.2838,
"step": 567
},
{
"epoch": 2.299595141700405,
"grad_norm": 0.5682334686328586,
"learning_rate": 2.620185047252308e-05,
"loss": 0.2579,
"step": 568
},
{
"epoch": 2.3036437246963564,
"grad_norm": 0.5349077234993206,
"learning_rate": 2.614805919320723e-05,
"loss": 0.2464,
"step": 569
},
{
"epoch": 2.3076923076923075,
"grad_norm": 1.1125561838829638,
"learning_rate": 2.6094218754173895e-05,
"loss": 0.3044,
"step": 570
},
{
"epoch": 2.311740890688259,
"grad_norm": 0.607709489069194,
"learning_rate": 2.60403295859298e-05,
"loss": 0.3456,
"step": 571
},
{
"epoch": 2.3157894736842106,
"grad_norm": 0.5277425380360949,
"learning_rate": 2.5986392119371297e-05,
"loss": 0.2711,
"step": 572
},
{
"epoch": 2.3198380566801617,
"grad_norm": 0.5263717946757053,
"learning_rate": 2.5932406785780928e-05,
"loss": 0.2873,
"step": 573
},
{
"epoch": 2.3238866396761133,
"grad_norm": 0.5208369358932562,
"learning_rate": 2.5878374016823987e-05,
"loss": 0.2729,
"step": 574
},
{
"epoch": 2.327935222672065,
"grad_norm": 0.735575111501246,
"learning_rate": 2.5824294244545045e-05,
"loss": 0.2559,
"step": 575
},
{
"epoch": 2.331983805668016,
"grad_norm": 0.5007509450590277,
"learning_rate": 2.577016790136453e-05,
"loss": 0.2605,
"step": 576
},
{
"epoch": 2.3360323886639676,
"grad_norm": 0.5264849017126916,
"learning_rate": 2.5715995420075225e-05,
"loss": 0.3151,
"step": 577
},
{
"epoch": 2.340080971659919,
"grad_norm": 0.6605249314760724,
"learning_rate": 2.5661777233838864e-05,
"loss": 0.2364,
"step": 578
},
{
"epoch": 2.3441295546558703,
"grad_norm": 0.6561831944792644,
"learning_rate": 2.5607513776182616e-05,
"loss": 0.2521,
"step": 579
},
{
"epoch": 2.348178137651822,
"grad_norm": 0.6481589933290606,
"learning_rate": 2.5553205480995633e-05,
"loss": 0.2705,
"step": 580
},
{
"epoch": 2.3522267206477734,
"grad_norm": 0.5433940045802638,
"learning_rate": 2.5498852782525604e-05,
"loss": 0.258,
"step": 581
},
{
"epoch": 2.3562753036437245,
"grad_norm": 0.7079576823023057,
"learning_rate": 2.5444456115375252e-05,
"loss": 0.2872,
"step": 582
},
{
"epoch": 2.360323886639676,
"grad_norm": 0.631793538217695,
"learning_rate": 2.539001591449889e-05,
"loss": 0.277,
"step": 583
},
{
"epoch": 2.3643724696356276,
"grad_norm": 1.1380409048149267,
"learning_rate": 2.53355326151989e-05,
"loss": 0.2909,
"step": 584
},
{
"epoch": 2.3684210526315788,
"grad_norm": 0.697751331204932,
"learning_rate": 2.5281006653122288e-05,
"loss": 0.2967,
"step": 585
},
{
"epoch": 2.3724696356275303,
"grad_norm": 0.6156496169900497,
"learning_rate": 2.5226438464257217e-05,
"loss": 0.3137,
"step": 586
},
{
"epoch": 2.376518218623482,
"grad_norm": 0.566340224937903,
"learning_rate": 2.5171828484929447e-05,
"loss": 0.2761,
"step": 587
},
{
"epoch": 2.380566801619433,
"grad_norm": 0.6550495404194205,
"learning_rate": 2.5117177151798934e-05,
"loss": 0.2857,
"step": 588
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.5214054962982035,
"learning_rate": 2.506248490185627e-05,
"loss": 0.2669,
"step": 589
},
{
"epoch": 2.388663967611336,
"grad_norm": 0.5109326245480089,
"learning_rate": 2.500775217241923e-05,
"loss": 0.3049,
"step": 590
},
{
"epoch": 2.3927125506072873,
"grad_norm": 0.604551544123906,
"learning_rate": 2.4952979401129266e-05,
"loss": 0.2623,
"step": 591
},
{
"epoch": 2.396761133603239,
"grad_norm": 0.5138217730313275,
"learning_rate": 2.4898167025948e-05,
"loss": 0.2846,
"step": 592
},
{
"epoch": 2.4008097165991904,
"grad_norm": 0.6131881765843618,
"learning_rate": 2.4843315485153703e-05,
"loss": 0.2914,
"step": 593
},
{
"epoch": 2.4048582995951415,
"grad_norm": 0.5753998409785155,
"learning_rate": 2.4788425217337856e-05,
"loss": 0.2735,
"step": 594
},
{
"epoch": 2.408906882591093,
"grad_norm": 0.7784350772770747,
"learning_rate": 2.4733496661401572e-05,
"loss": 0.2584,
"step": 595
},
{
"epoch": 2.4129554655870447,
"grad_norm": 0.55164846400088,
"learning_rate": 2.4678530256552114e-05,
"loss": 0.2672,
"step": 596
},
{
"epoch": 2.417004048582996,
"grad_norm": 1.2262165301293713,
"learning_rate": 2.46235264422994e-05,
"loss": 0.2952,
"step": 597
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.5908437180586928,
"learning_rate": 2.4568485658452452e-05,
"loss": 0.246,
"step": 598
},
{
"epoch": 2.425101214574899,
"grad_norm": 0.5186008382373357,
"learning_rate": 2.451340834511592e-05,
"loss": 0.2755,
"step": 599
},
{
"epoch": 2.42914979757085,
"grad_norm": 0.49141475066533713,
"learning_rate": 2.4458294942686528e-05,
"loss": 0.2673,
"step": 600
},
{
"epoch": 2.4331983805668016,
"grad_norm": 0.5389468688736758,
"learning_rate": 2.440314589184958e-05,
"loss": 0.2846,
"step": 601
},
{
"epoch": 2.437246963562753,
"grad_norm": 0.5569145913074681,
"learning_rate": 2.434796163357541e-05,
"loss": 0.2949,
"step": 602
},
{
"epoch": 2.4412955465587043,
"grad_norm": 0.5651125763057591,
"learning_rate": 2.429274260911588e-05,
"loss": 0.2944,
"step": 603
},
{
"epoch": 2.445344129554656,
"grad_norm": 0.5479989048037583,
"learning_rate": 2.423748926000084e-05,
"loss": 0.2689,
"step": 604
},
{
"epoch": 2.4493927125506074,
"grad_norm": 0.5156145061855272,
"learning_rate": 2.418220202803459e-05,
"loss": 0.2629,
"step": 605
},
{
"epoch": 2.4534412955465585,
"grad_norm": 0.5538722534568596,
"learning_rate": 2.412688135529237e-05,
"loss": 0.2789,
"step": 606
},
{
"epoch": 2.45748987854251,
"grad_norm": 0.7820524734161053,
"learning_rate": 2.4071527684116807e-05,
"loss": 0.2594,
"step": 607
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.6123989420955539,
"learning_rate": 2.4016141457114372e-05,
"loss": 0.2666,
"step": 608
},
{
"epoch": 2.465587044534413,
"grad_norm": 0.5118197782084453,
"learning_rate": 2.396072311715187e-05,
"loss": 0.2753,
"step": 609
},
{
"epoch": 2.4696356275303644,
"grad_norm": 0.6971616091612334,
"learning_rate": 2.3905273107352863e-05,
"loss": 0.2863,
"step": 610
},
{
"epoch": 2.473684210526316,
"grad_norm": 0.5123626138412264,
"learning_rate": 2.3849791871094157e-05,
"loss": 0.2682,
"step": 611
},
{
"epoch": 2.477732793522267,
"grad_norm": 0.5339257545887731,
"learning_rate": 2.379427985200223e-05,
"loss": 0.257,
"step": 612
},
{
"epoch": 2.4817813765182186,
"grad_norm": 0.608370645716737,
"learning_rate": 2.373873749394972e-05,
"loss": 0.2898,
"step": 613
},
{
"epoch": 2.48582995951417,
"grad_norm": 0.5148213856029835,
"learning_rate": 2.3683165241051836e-05,
"loss": 0.2779,
"step": 614
},
{
"epoch": 2.4898785425101213,
"grad_norm": 0.8686765753851864,
"learning_rate": 2.3627563537662823e-05,
"loss": 0.2933,
"step": 615
},
{
"epoch": 2.493927125506073,
"grad_norm": 0.6012170916509771,
"learning_rate": 2.3571932828372433e-05,
"loss": 0.2601,
"step": 616
},
{
"epoch": 2.4979757085020244,
"grad_norm": 0.6298335585687407,
"learning_rate": 2.351627355800233e-05,
"loss": 0.2749,
"step": 617
},
{
"epoch": 2.5020242914979756,
"grad_norm": 0.5756541289086935,
"learning_rate": 2.346058617160256e-05,
"loss": 0.249,
"step": 618
},
{
"epoch": 2.506072874493927,
"grad_norm": 0.8880138403806707,
"learning_rate": 2.3404871114447976e-05,
"loss": 0.2951,
"step": 619
},
{
"epoch": 2.5101214574898787,
"grad_norm": 0.6096175159569388,
"learning_rate": 2.334912883203469e-05,
"loss": 0.2993,
"step": 620
},
{
"epoch": 2.51417004048583,
"grad_norm": 0.5331872802501344,
"learning_rate": 2.3293359770076505e-05,
"loss": 0.2572,
"step": 621
},
{
"epoch": 2.5182186234817814,
"grad_norm": 0.5758087471776424,
"learning_rate": 2.323756437450136e-05,
"loss": 0.2779,
"step": 622
},
{
"epoch": 2.522267206477733,
"grad_norm": 0.5359082453581675,
"learning_rate": 2.3181743091447754e-05,
"loss": 0.3179,
"step": 623
},
{
"epoch": 2.526315789473684,
"grad_norm": 0.49258941364923203,
"learning_rate": 2.3125896367261164e-05,
"loss": 0.2723,
"step": 624
},
{
"epoch": 2.5303643724696356,
"grad_norm": 0.5825283142179529,
"learning_rate": 2.307002464849052e-05,
"loss": 0.2597,
"step": 625
},
{
"epoch": 2.534412955465587,
"grad_norm": 0.5026793346013811,
"learning_rate": 2.30141283818846e-05,
"loss": 0.2625,
"step": 626
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.6926729180523932,
"learning_rate": 2.2958208014388462e-05,
"loss": 0.2464,
"step": 627
},
{
"epoch": 2.54251012145749,
"grad_norm": 1.6206707367426703,
"learning_rate": 2.2902263993139857e-05,
"loss": 0.2976,
"step": 628
},
{
"epoch": 2.5465587044534415,
"grad_norm": 0.5341501689223896,
"learning_rate": 2.2846296765465708e-05,
"loss": 0.2516,
"step": 629
},
{
"epoch": 2.5506072874493926,
"grad_norm": 0.8073125451352318,
"learning_rate": 2.2790306778878465e-05,
"loss": 0.2654,
"step": 630
},
{
"epoch": 2.554655870445344,
"grad_norm": 0.5895372956681044,
"learning_rate": 2.273429448107257e-05,
"loss": 0.2633,
"step": 631
},
{
"epoch": 2.5587044534412957,
"grad_norm": 0.5508576950018121,
"learning_rate": 2.2678260319920868e-05,
"loss": 0.2809,
"step": 632
},
{
"epoch": 2.562753036437247,
"grad_norm": 0.5793401350226242,
"learning_rate": 2.262220474347102e-05,
"loss": 0.2452,
"step": 633
},
{
"epoch": 2.5668016194331984,
"grad_norm": 0.5035278643439726,
"learning_rate": 2.2566128199941918e-05,
"loss": 0.2739,
"step": 634
},
{
"epoch": 2.57085020242915,
"grad_norm": 0.5027064032811989,
"learning_rate": 2.2510031137720104e-05,
"loss": 0.2727,
"step": 635
},
{
"epoch": 2.574898785425101,
"grad_norm": 0.5508239593439942,
"learning_rate": 2.2453914005356206e-05,
"loss": 0.2846,
"step": 636
},
{
"epoch": 2.5789473684210527,
"grad_norm": 0.5691191100897592,
"learning_rate": 2.2397777251561302e-05,
"loss": 0.2894,
"step": 637
},
{
"epoch": 2.582995951417004,
"grad_norm": 0.5168974691836294,
"learning_rate": 2.2341621325203393e-05,
"loss": 0.2835,
"step": 638
},
{
"epoch": 2.5870445344129553,
"grad_norm": 0.6330555370167597,
"learning_rate": 2.2285446675303764e-05,
"loss": 0.2867,
"step": 639
},
{
"epoch": 2.591093117408907,
"grad_norm": 0.5179047442599396,
"learning_rate": 2.2229253751033423e-05,
"loss": 0.2675,
"step": 640
},
{
"epoch": 2.5951417004048585,
"grad_norm": 0.5785890796120687,
"learning_rate": 2.2173043001709492e-05,
"loss": 0.3081,
"step": 641
},
{
"epoch": 2.5991902834008096,
"grad_norm": 0.5293372026104409,
"learning_rate": 2.2116814876791627e-05,
"loss": 0.2336,
"step": 642
},
{
"epoch": 2.603238866396761,
"grad_norm": 0.5320190053268438,
"learning_rate": 2.2060569825878413e-05,
"loss": 0.2765,
"step": 643
},
{
"epoch": 2.6072874493927127,
"grad_norm": 0.4427153922131416,
"learning_rate": 2.2004308298703777e-05,
"loss": 0.2681,
"step": 644
},
{
"epoch": 2.611336032388664,
"grad_norm": 0.5343124742730944,
"learning_rate": 2.1948030745133396e-05,
"loss": 0.238,
"step": 645
},
{
"epoch": 2.6153846153846154,
"grad_norm": 1.2841338349528637,
"learning_rate": 2.1891737615161083e-05,
"loss": 0.2558,
"step": 646
},
{
"epoch": 2.619433198380567,
"grad_norm": 0.6188295584327519,
"learning_rate": 2.1835429358905218e-05,
"loss": 0.2266,
"step": 647
},
{
"epoch": 2.623481781376518,
"grad_norm": 0.5214096087382686,
"learning_rate": 2.17791064266051e-05,
"loss": 0.2673,
"step": 648
},
{
"epoch": 2.6275303643724697,
"grad_norm": 1.4030482440130942,
"learning_rate": 2.172276926861741e-05,
"loss": 0.299,
"step": 649
},
{
"epoch": 2.6315789473684212,
"grad_norm": 0.5622085964713419,
"learning_rate": 2.166641833541255e-05,
"loss": 0.254,
"step": 650
},
{
"epoch": 2.6356275303643724,
"grad_norm": 0.6343515745475755,
"learning_rate": 2.161005407757109e-05,
"loss": 0.2595,
"step": 651
},
{
"epoch": 2.639676113360324,
"grad_norm": 0.5233217601287751,
"learning_rate": 2.155367694578013e-05,
"loss": 0.2711,
"step": 652
},
{
"epoch": 2.6437246963562755,
"grad_norm": 0.5535649061290727,
"learning_rate": 2.1497287390829712e-05,
"loss": 0.2552,
"step": 653
},
{
"epoch": 2.6477732793522266,
"grad_norm": 0.8866119222189809,
"learning_rate": 2.1440885863609208e-05,
"loss": 0.3007,
"step": 654
},
{
"epoch": 2.651821862348178,
"grad_norm": 0.5072616487086111,
"learning_rate": 2.1384472815103737e-05,
"loss": 0.2679,
"step": 655
},
{
"epoch": 2.6558704453441297,
"grad_norm": 0.5643490166552982,
"learning_rate": 2.1328048696390534e-05,
"loss": 0.2867,
"step": 656
},
{
"epoch": 2.659919028340081,
"grad_norm": 0.5481154805477989,
"learning_rate": 2.1271613958635342e-05,
"loss": 0.2896,
"step": 657
},
{
"epoch": 2.6639676113360324,
"grad_norm": 0.5151860184628099,
"learning_rate": 2.1215169053088825e-05,
"loss": 0.2647,
"step": 658
},
{
"epoch": 2.668016194331984,
"grad_norm": 0.5102505408009816,
"learning_rate": 2.1158714431082937e-05,
"loss": 0.2792,
"step": 659
},
{
"epoch": 2.672064777327935,
"grad_norm": 0.5387306796979847,
"learning_rate": 2.110225054402733e-05,
"loss": 0.2811,
"step": 660
},
{
"epoch": 2.6761133603238867,
"grad_norm": 0.5377765064683591,
"learning_rate": 2.1045777843405747e-05,
"loss": 0.2926,
"step": 661
},
{
"epoch": 2.6801619433198383,
"grad_norm": 0.4804486852342911,
"learning_rate": 2.09892967807724e-05,
"loss": 0.2576,
"step": 662
},
{
"epoch": 2.6842105263157894,
"grad_norm": 0.5037675345251194,
"learning_rate": 2.093280780774835e-05,
"loss": 0.2984,
"step": 663
},
{
"epoch": 2.688259109311741,
"grad_norm": 0.4889312154165904,
"learning_rate": 2.087631137601793e-05,
"loss": 0.282,
"step": 664
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.8237397055644391,
"learning_rate": 2.0819807937325104e-05,
"loss": 0.3394,
"step": 665
},
{
"epoch": 2.6963562753036436,
"grad_norm": 0.6286022753051812,
"learning_rate": 2.0763297943469847e-05,
"loss": 0.2558,
"step": 666
},
{
"epoch": 2.700404858299595,
"grad_norm": 0.5291907319344537,
"learning_rate": 2.070678184630458e-05,
"loss": 0.2594,
"step": 667
},
{
"epoch": 2.7044534412955468,
"grad_norm": 0.49978324375368893,
"learning_rate": 2.06502600977305e-05,
"loss": 0.2564,
"step": 668
},
{
"epoch": 2.708502024291498,
"grad_norm": 0.562498638953671,
"learning_rate": 2.0593733149694008e-05,
"loss": 0.2761,
"step": 669
},
{
"epoch": 2.7125506072874495,
"grad_norm": 0.5118794485358056,
"learning_rate": 2.0537201454183074e-05,
"loss": 0.253,
"step": 670
},
{
"epoch": 2.716599190283401,
"grad_norm": 0.9064630133366244,
"learning_rate": 2.0480665463223626e-05,
"loss": 0.255,
"step": 671
},
{
"epoch": 2.720647773279352,
"grad_norm": 0.7882580947726364,
"learning_rate": 2.042412562887594e-05,
"loss": 0.289,
"step": 672
},
{
"epoch": 2.7246963562753037,
"grad_norm": 0.5625107969662394,
"learning_rate": 2.0367582403231033e-05,
"loss": 0.304,
"step": 673
},
{
"epoch": 2.7287449392712553,
"grad_norm": 0.5006310947715809,
"learning_rate": 2.0311036238407023e-05,
"loss": 0.2925,
"step": 674
},
{
"epoch": 2.7327935222672064,
"grad_norm": 0.5343443102465788,
"learning_rate": 2.0254487586545534e-05,
"loss": 0.2815,
"step": 675
},
{
"epoch": 2.736842105263158,
"grad_norm": 0.557408656089659,
"learning_rate": 2.0197936899808088e-05,
"loss": 0.2586,
"step": 676
},
{
"epoch": 2.7408906882591095,
"grad_norm": 0.5062066174903348,
"learning_rate": 2.0141384630372466e-05,
"loss": 0.3025,
"step": 677
},
{
"epoch": 2.7449392712550607,
"grad_norm": 0.5117169031005852,
"learning_rate": 2.0084831230429106e-05,
"loss": 0.2869,
"step": 678
},
{
"epoch": 2.748987854251012,
"grad_norm": 0.5908589753200898,
"learning_rate": 2.0028277152177492e-05,
"loss": 0.2584,
"step": 679
},
{
"epoch": 2.753036437246964,
"grad_norm": 1.2590585792893534,
"learning_rate": 1.9971722847822518e-05,
"loss": 0.2809,
"step": 680
},
{
"epoch": 2.757085020242915,
"grad_norm": 0.5521660978938729,
"learning_rate": 1.99151687695709e-05,
"loss": 0.2583,
"step": 681
},
{
"epoch": 2.7611336032388665,
"grad_norm": 0.5301746022159416,
"learning_rate": 1.9858615369627537e-05,
"loss": 0.2629,
"step": 682
},
{
"epoch": 2.765182186234818,
"grad_norm": 0.5638454704965183,
"learning_rate": 1.9802063100191916e-05,
"loss": 0.2309,
"step": 683
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.49123813291767854,
"learning_rate": 1.974551241345447e-05,
"loss": 0.2537,
"step": 684
},
{
"epoch": 2.7732793522267207,
"grad_norm": 0.6210788550224456,
"learning_rate": 1.9688963761592984e-05,
"loss": 0.2558,
"step": 685
},
{
"epoch": 2.7773279352226723,
"grad_norm": 0.5241588924948405,
"learning_rate": 1.963241759676897e-05,
"loss": 0.2564,
"step": 686
},
{
"epoch": 2.7813765182186234,
"grad_norm": 0.5217491267495588,
"learning_rate": 1.9575874371124066e-05,
"loss": 0.2634,
"step": 687
},
{
"epoch": 2.785425101214575,
"grad_norm": 0.4991164130029741,
"learning_rate": 1.951933453677638e-05,
"loss": 0.2741,
"step": 688
},
{
"epoch": 2.7894736842105265,
"grad_norm": 0.5078421017614728,
"learning_rate": 1.9462798545816932e-05,
"loss": 0.251,
"step": 689
},
{
"epoch": 2.7935222672064777,
"grad_norm": 0.6102705733315366,
"learning_rate": 1.9406266850305996e-05,
"loss": 0.255,
"step": 690
},
{
"epoch": 2.7975708502024292,
"grad_norm": 0.5000765956076475,
"learning_rate": 1.9349739902269503e-05,
"loss": 0.2912,
"step": 691
},
{
"epoch": 2.801619433198381,
"grad_norm": 0.650901609406627,
"learning_rate": 1.9293218153695425e-05,
"loss": 0.2714,
"step": 692
},
{
"epoch": 2.805668016194332,
"grad_norm": 0.5159765651627471,
"learning_rate": 1.9236702056530153e-05,
"loss": 0.2723,
"step": 693
},
{
"epoch": 2.8097165991902835,
"grad_norm": 0.5760640827173734,
"learning_rate": 1.9180192062674907e-05,
"loss": 0.2594,
"step": 694
},
{
"epoch": 2.813765182186235,
"grad_norm": 0.4799676144013165,
"learning_rate": 1.9123688623982076e-05,
"loss": 0.2734,
"step": 695
},
{
"epoch": 2.817813765182186,
"grad_norm": 0.5376352816851193,
"learning_rate": 1.906719219225165e-05,
"loss": 0.3103,
"step": 696
},
{
"epoch": 2.8218623481781377,
"grad_norm": 0.5410342394373145,
"learning_rate": 1.9010703219227603e-05,
"loss": 0.2701,
"step": 697
},
{
"epoch": 2.8259109311740893,
"grad_norm": 0.5709749258234796,
"learning_rate": 1.8954222156594256e-05,
"loss": 0.2647,
"step": 698
},
{
"epoch": 2.8299595141700404,
"grad_norm": 0.48955296931387543,
"learning_rate": 1.8897749455972673e-05,
"loss": 0.2862,
"step": 699
},
{
"epoch": 2.834008097165992,
"grad_norm": 0.7972257053018811,
"learning_rate": 1.884128556891707e-05,
"loss": 0.3081,
"step": 700
},
{
"epoch": 2.8380566801619436,
"grad_norm": 0.5101376269372113,
"learning_rate": 1.878483094691119e-05,
"loss": 0.2962,
"step": 701
},
{
"epoch": 2.8421052631578947,
"grad_norm": 0.5809942211088953,
"learning_rate": 1.8728386041364664e-05,
"loss": 0.2704,
"step": 702
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.8412234254397132,
"learning_rate": 1.867195130360947e-05,
"loss": 0.2617,
"step": 703
},
{
"epoch": 2.850202429149798,
"grad_norm": 0.8823734597796186,
"learning_rate": 1.8615527184896263e-05,
"loss": 0.2711,
"step": 704
},
{
"epoch": 2.854251012145749,
"grad_norm": 0.72464571599413,
"learning_rate": 1.8559114136390795e-05,
"loss": 0.2842,
"step": 705
},
{
"epoch": 2.8582995951417005,
"grad_norm": 0.5141410666587896,
"learning_rate": 1.8502712609170298e-05,
"loss": 0.2578,
"step": 706
},
{
"epoch": 2.862348178137652,
"grad_norm": 0.6182581876128471,
"learning_rate": 1.8446323054219876e-05,
"loss": 0.2657,
"step": 707
},
{
"epoch": 2.866396761133603,
"grad_norm": 0.5574853604755492,
"learning_rate": 1.838994592242891e-05,
"loss": 0.2498,
"step": 708
},
{
"epoch": 2.8704453441295548,
"grad_norm": 0.4964650661744877,
"learning_rate": 1.8333581664587453e-05,
"loss": 0.3069,
"step": 709
},
{
"epoch": 2.8744939271255063,
"grad_norm": 0.4975127906112016,
"learning_rate": 1.8277230731382593e-05,
"loss": 0.2774,
"step": 710
},
{
"epoch": 2.8785425101214575,
"grad_norm": 0.5105802448082029,
"learning_rate": 1.82208935733949e-05,
"loss": 0.287,
"step": 711
},
{
"epoch": 2.882591093117409,
"grad_norm": 0.6028892606803838,
"learning_rate": 1.8164570641094793e-05,
"loss": 0.2836,
"step": 712
},
{
"epoch": 2.8866396761133606,
"grad_norm": 0.5109323265758638,
"learning_rate": 1.810826238483892e-05,
"loss": 0.2591,
"step": 713
},
{
"epoch": 2.8906882591093117,
"grad_norm": 0.7671157039442934,
"learning_rate": 1.8051969254866608e-05,
"loss": 0.2728,
"step": 714
},
{
"epoch": 2.8947368421052633,
"grad_norm": 0.48158879440956637,
"learning_rate": 1.7995691701296226e-05,
"loss": 0.2636,
"step": 715
},
{
"epoch": 2.898785425101215,
"grad_norm": 0.5943412117396312,
"learning_rate": 1.7939430174121593e-05,
"loss": 0.2711,
"step": 716
},
{
"epoch": 2.902834008097166,
"grad_norm": 0.4994411850713464,
"learning_rate": 1.788318512320838e-05,
"loss": 0.2695,
"step": 717
},
{
"epoch": 2.9068825910931175,
"grad_norm": 0.5131332107488845,
"learning_rate": 1.782695699829051e-05,
"loss": 0.3134,
"step": 718
},
{
"epoch": 2.910931174089069,
"grad_norm": 0.5861180843300459,
"learning_rate": 1.7770746248966587e-05,
"loss": 0.2668,
"step": 719
},
{
"epoch": 2.91497975708502,
"grad_norm": 0.49709045880826275,
"learning_rate": 1.7714553324696243e-05,
"loss": 0.2735,
"step": 720
},
{
"epoch": 2.919028340080972,
"grad_norm": 0.5949163056643956,
"learning_rate": 1.7658378674796614e-05,
"loss": 0.2533,
"step": 721
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.6524262124038026,
"learning_rate": 1.76022227484387e-05,
"loss": 0.2987,
"step": 722
},
{
"epoch": 2.9271255060728745,
"grad_norm": 0.6082162428602788,
"learning_rate": 1.7546085994643807e-05,
"loss": 0.2608,
"step": 723
},
{
"epoch": 2.931174089068826,
"grad_norm": 0.47943963795713695,
"learning_rate": 1.7489968862279902e-05,
"loss": 0.2798,
"step": 724
},
{
"epoch": 2.9352226720647776,
"grad_norm": 0.4991524993098724,
"learning_rate": 1.743387180005809e-05,
"loss": 0.2603,
"step": 725
},
{
"epoch": 2.9392712550607287,
"grad_norm": 0.4944489353855595,
"learning_rate": 1.737779525652899e-05,
"loss": 0.2548,
"step": 726
},
{
"epoch": 2.9433198380566803,
"grad_norm": 0.506770704411323,
"learning_rate": 1.732173968007914e-05,
"loss": 0.2849,
"step": 727
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.6835625174773496,
"learning_rate": 1.7265705518927437e-05,
"loss": 0.3061,
"step": 728
},
{
"epoch": 2.951417004048583,
"grad_norm": 0.599163249329214,
"learning_rate": 1.7209693221121542e-05,
"loss": 0.2916,
"step": 729
},
{
"epoch": 2.9554655870445345,
"grad_norm": 0.7835324770342864,
"learning_rate": 1.7153703234534302e-05,
"loss": 0.2692,
"step": 730
},
{
"epoch": 2.9595141700404857,
"grad_norm": 0.7064855163855814,
"learning_rate": 1.709773600686015e-05,
"loss": 0.2605,
"step": 731
},
{
"epoch": 2.9635627530364372,
"grad_norm": 0.7362782177276238,
"learning_rate": 1.7041791985611548e-05,
"loss": 0.3023,
"step": 732
},
{
"epoch": 2.967611336032389,
"grad_norm": 0.5219177949372421,
"learning_rate": 1.69858716181154e-05,
"loss": 0.2815,
"step": 733
},
{
"epoch": 2.97165991902834,
"grad_norm": 0.715369905545974,
"learning_rate": 1.692997535150948e-05,
"loss": 0.2648,
"step": 734
},
{
"epoch": 2.9757085020242915,
"grad_norm": 0.5599757386355709,
"learning_rate": 1.687410363273884e-05,
"loss": 0.2689,
"step": 735
},
{
"epoch": 2.979757085020243,
"grad_norm": 0.5969717878391205,
"learning_rate": 1.6818256908552257e-05,
"loss": 0.2539,
"step": 736
},
{
"epoch": 2.983805668016194,
"grad_norm": 0.42315970998382635,
"learning_rate": 1.6762435625498646e-05,
"loss": 0.2696,
"step": 737
},
{
"epoch": 2.9878542510121457,
"grad_norm": 0.5840806815885051,
"learning_rate": 1.67066402299235e-05,
"loss": 0.2409,
"step": 738
},
{
"epoch": 2.9919028340080973,
"grad_norm": 0.484596615336308,
"learning_rate": 1.6650871167965313e-05,
"loss": 0.246,
"step": 739
},
{
"epoch": 2.9959514170040484,
"grad_norm": 0.5913211515996577,
"learning_rate": 1.6595128885552028e-05,
"loss": 0.2781,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 1.4060702405851195,
"learning_rate": 1.653941382839745e-05,
"loss": 0.2695,
"step": 741
},
{
"epoch": 3.0040485829959516,
"grad_norm": 0.36729582056830073,
"learning_rate": 1.6483726441997673e-05,
"loss": 0.1663,
"step": 742
},
{
"epoch": 3.0080971659919027,
"grad_norm": 0.39898427327209457,
"learning_rate": 1.642806717162757e-05,
"loss": 0.2114,
"step": 743
},
{
"epoch": 3.0121457489878543,
"grad_norm": 0.3267766795137925,
"learning_rate": 1.637243646233718e-05,
"loss": 0.1925,
"step": 744
},
{
"epoch": 3.016194331983806,
"grad_norm": 0.3343423426210911,
"learning_rate": 1.6316834758948174e-05,
"loss": 0.2323,
"step": 745
},
{
"epoch": 3.020242914979757,
"grad_norm": 0.38286055068930724,
"learning_rate": 1.6261262506050282e-05,
"loss": 0.2104,
"step": 746
},
{
"epoch": 3.0242914979757085,
"grad_norm": 0.4610402600321375,
"learning_rate": 1.620572014799777e-05,
"loss": 0.2086,
"step": 747
},
{
"epoch": 3.02834008097166,
"grad_norm": 0.35581149720747096,
"learning_rate": 1.6150208128905857e-05,
"loss": 0.1998,
"step": 748
},
{
"epoch": 3.032388663967611,
"grad_norm": 0.37892912090231057,
"learning_rate": 1.6094726892647147e-05,
"loss": 0.183,
"step": 749
},
{
"epoch": 3.0364372469635628,
"grad_norm": 0.3738103876762375,
"learning_rate": 1.6039276882848135e-05,
"loss": 0.1775,
"step": 750
},
{
"epoch": 3.0404858299595143,
"grad_norm": 0.4128209232870934,
"learning_rate": 1.5983858542885635e-05,
"loss": 0.208,
"step": 751
},
{
"epoch": 3.0445344129554655,
"grad_norm": 0.3701781488723076,
"learning_rate": 1.5928472315883203e-05,
"loss": 0.1935,
"step": 752
},
{
"epoch": 3.048582995951417,
"grad_norm": 0.3715346204109034,
"learning_rate": 1.5873118644707633e-05,
"loss": 0.1983,
"step": 753
},
{
"epoch": 3.0526315789473686,
"grad_norm": 0.34494019530287495,
"learning_rate": 1.5817797971965413e-05,
"loss": 0.1668,
"step": 754
},
{
"epoch": 3.0566801619433197,
"grad_norm": 0.3985147460962563,
"learning_rate": 1.576251073999917e-05,
"loss": 0.2176,
"step": 755
},
{
"epoch": 3.0607287449392713,
"grad_norm": 0.40133157067454245,
"learning_rate": 1.5707257390884126e-05,
"loss": 0.1916,
"step": 756
},
{
"epoch": 3.064777327935223,
"grad_norm": 0.39425212411994615,
"learning_rate": 1.5652038366424595e-05,
"loss": 0.1761,
"step": 757
},
{
"epoch": 3.068825910931174,
"grad_norm": 0.6333466695704362,
"learning_rate": 1.5596854108150424e-05,
"loss": 0.2261,
"step": 758
},
{
"epoch": 3.0728744939271255,
"grad_norm": 1.3771388721289768,
"learning_rate": 1.5541705057313476e-05,
"loss": 0.2244,
"step": 759
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.37091475281272707,
"learning_rate": 1.5486591654884086e-05,
"loss": 0.1869,
"step": 760
},
{
"epoch": 3.080971659919028,
"grad_norm": 0.3681078194331674,
"learning_rate": 1.543151434154755e-05,
"loss": 0.1681,
"step": 761
},
{
"epoch": 3.08502024291498,
"grad_norm": 0.4064629355984237,
"learning_rate": 1.5376473557700615e-05,
"loss": 0.1709,
"step": 762
},
{
"epoch": 3.0890688259109313,
"grad_norm": 0.3751121491703414,
"learning_rate": 1.5321469743447893e-05,
"loss": 0.1731,
"step": 763
},
{
"epoch": 3.0931174089068825,
"grad_norm": 0.3796566429727685,
"learning_rate": 1.5266503338598434e-05,
"loss": 0.184,
"step": 764
},
{
"epoch": 3.097165991902834,
"grad_norm": 0.312707664724226,
"learning_rate": 1.5211574782662149e-05,
"loss": 0.1777,
"step": 765
},
{
"epoch": 3.1012145748987856,
"grad_norm": 0.7359465951366814,
"learning_rate": 1.5156684514846307e-05,
"loss": 0.2017,
"step": 766
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.39485978105305053,
"learning_rate": 1.5101832974052012e-05,
"loss": 0.2009,
"step": 767
},
{
"epoch": 3.1093117408906883,
"grad_norm": 0.4155937983698579,
"learning_rate": 1.5047020598870737e-05,
"loss": 0.1888,
"step": 768
},
{
"epoch": 3.11336032388664,
"grad_norm": 0.3554676741635712,
"learning_rate": 1.4992247827580778e-05,
"loss": 0.1949,
"step": 769
},
{
"epoch": 3.117408906882591,
"grad_norm": 0.3613818838654057,
"learning_rate": 1.493751509814374e-05,
"loss": 0.2087,
"step": 770
},
{
"epoch": 3.1214574898785425,
"grad_norm": 0.34725289516714314,
"learning_rate": 1.4882822848201073e-05,
"loss": 0.1681,
"step": 771
},
{
"epoch": 3.125506072874494,
"grad_norm": 0.3377706239711726,
"learning_rate": 1.4828171515070553e-05,
"loss": 0.1881,
"step": 772
},
{
"epoch": 3.1295546558704452,
"grad_norm": 0.39201145247113467,
"learning_rate": 1.4773561535742793e-05,
"loss": 0.1777,
"step": 773
},
{
"epoch": 3.133603238866397,
"grad_norm": 2.613998573516557,
"learning_rate": 1.4718993346877715e-05,
"loss": 0.1921,
"step": 774
},
{
"epoch": 3.1376518218623484,
"grad_norm": 0.3803700158433991,
"learning_rate": 1.466446738480111e-05,
"loss": 0.1705,
"step": 775
},
{
"epoch": 3.1417004048582995,
"grad_norm": 0.5091898620697394,
"learning_rate": 1.460998408550112e-05,
"loss": 0.1935,
"step": 776
},
{
"epoch": 3.145748987854251,
"grad_norm": 1.2569024266879065,
"learning_rate": 1.4555543884624751e-05,
"loss": 0.2549,
"step": 777
},
{
"epoch": 3.1497975708502026,
"grad_norm": 0.3109508574155528,
"learning_rate": 1.4501147217474402e-05,
"loss": 0.185,
"step": 778
},
{
"epoch": 3.1538461538461537,
"grad_norm": 0.3549887573708317,
"learning_rate": 1.444679451900437e-05,
"loss": 0.1938,
"step": 779
},
{
"epoch": 3.1578947368421053,
"grad_norm": 0.35687069615458666,
"learning_rate": 1.4392486223817397e-05,
"loss": 0.1721,
"step": 780
},
{
"epoch": 3.161943319838057,
"grad_norm": 0.3059176151500034,
"learning_rate": 1.4338222766161143e-05,
"loss": 0.1729,
"step": 781
},
{
"epoch": 3.165991902834008,
"grad_norm": 0.34382501443562963,
"learning_rate": 1.4284004579924777e-05,
"loss": 0.1878,
"step": 782
},
{
"epoch": 3.1700404858299596,
"grad_norm": 0.5610137865178165,
"learning_rate": 1.4229832098635479e-05,
"loss": 0.202,
"step": 783
},
{
"epoch": 3.174089068825911,
"grad_norm": 0.34477181608608076,
"learning_rate": 1.4175705755454963e-05,
"loss": 0.209,
"step": 784
},
{
"epoch": 3.1781376518218623,
"grad_norm": 0.35659358170020816,
"learning_rate": 1.412162598317602e-05,
"loss": 0.1933,
"step": 785
},
{
"epoch": 3.182186234817814,
"grad_norm": 0.35615364467487304,
"learning_rate": 1.406759321421907e-05,
"loss": 0.174,
"step": 786
},
{
"epoch": 3.1862348178137654,
"grad_norm": 0.3465008537259742,
"learning_rate": 1.401360788062871e-05,
"loss": 0.1974,
"step": 787
},
{
"epoch": 3.1902834008097165,
"grad_norm": 0.3980919579953644,
"learning_rate": 1.3959670414070204e-05,
"loss": 0.1757,
"step": 788
},
{
"epoch": 3.194331983805668,
"grad_norm": 0.34810139262362977,
"learning_rate": 1.3905781245826108e-05,
"loss": 0.2132,
"step": 789
},
{
"epoch": 3.1983805668016196,
"grad_norm": 0.3139460447438319,
"learning_rate": 1.3851940806792778e-05,
"loss": 0.1622,
"step": 790
},
{
"epoch": 3.2024291497975708,
"grad_norm": 0.3916773716857556,
"learning_rate": 1.379814952747693e-05,
"loss": 0.2093,
"step": 791
},
{
"epoch": 3.2064777327935223,
"grad_norm": 0.34422709562580756,
"learning_rate": 1.3744407837992193e-05,
"loss": 0.1965,
"step": 792
},
{
"epoch": 3.2105263157894735,
"grad_norm": 0.3465958117476038,
"learning_rate": 1.3690716168055692e-05,
"loss": 0.1735,
"step": 793
},
{
"epoch": 3.214574898785425,
"grad_norm": 0.3693457076822894,
"learning_rate": 1.363707494698459e-05,
"loss": 0.2009,
"step": 794
},
{
"epoch": 3.2186234817813766,
"grad_norm": 0.3288744881488745,
"learning_rate": 1.358348460369265e-05,
"loss": 0.1734,
"step": 795
},
{
"epoch": 3.2226720647773277,
"grad_norm": 0.501148884121069,
"learning_rate": 1.352994556668684e-05,
"loss": 0.1771,
"step": 796
},
{
"epoch": 3.2267206477732793,
"grad_norm": 0.41902671193602165,
"learning_rate": 1.3476458264063865e-05,
"loss": 0.2177,
"step": 797
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.41951392173133084,
"learning_rate": 1.342302312350678e-05,
"loss": 0.1961,
"step": 798
},
{
"epoch": 3.234817813765182,
"grad_norm": 0.4035471507755397,
"learning_rate": 1.3369640572281537e-05,
"loss": 0.2014,
"step": 799
},
{
"epoch": 3.2388663967611335,
"grad_norm": 0.3392102157989639,
"learning_rate": 1.3316311037233596e-05,
"loss": 0.1656,
"step": 800
},
{
"epoch": 3.242914979757085,
"grad_norm": 0.4329699854598184,
"learning_rate": 1.326303494478451e-05,
"loss": 0.2011,
"step": 801
},
{
"epoch": 3.246963562753036,
"grad_norm": 0.3847381486928026,
"learning_rate": 1.3209812720928472e-05,
"loss": 0.1906,
"step": 802
},
{
"epoch": 3.251012145748988,
"grad_norm": 0.3016918919683117,
"learning_rate": 1.315664479122898e-05,
"loss": 0.2006,
"step": 803
},
{
"epoch": 3.2550607287449393,
"grad_norm": 0.3623070120936601,
"learning_rate": 1.3103531580815378e-05,
"loss": 0.1875,
"step": 804
},
{
"epoch": 3.2591093117408905,
"grad_norm": 0.292321528168288,
"learning_rate": 1.3050473514379488e-05,
"loss": 0.23,
"step": 805
},
{
"epoch": 3.263157894736842,
"grad_norm": 0.30509904186799114,
"learning_rate": 1.2997471016172188e-05,
"loss": 0.173,
"step": 806
},
{
"epoch": 3.2672064777327936,
"grad_norm": 0.46354633538476103,
"learning_rate": 1.2944524510000042e-05,
"loss": 0.2011,
"step": 807
},
{
"epoch": 3.2712550607287447,
"grad_norm": 0.37759280019360963,
"learning_rate": 1.2891634419221908e-05,
"loss": 0.275,
"step": 808
},
{
"epoch": 3.2753036437246963,
"grad_norm": 0.3202595638557271,
"learning_rate": 1.2838801166745545e-05,
"loss": 0.2141,
"step": 809
},
{
"epoch": 3.279352226720648,
"grad_norm": 0.9305796703090534,
"learning_rate": 1.2786025175024216e-05,
"loss": 0.1675,
"step": 810
},
{
"epoch": 3.283400809716599,
"grad_norm": 0.38642266099025707,
"learning_rate": 1.2733306866053357e-05,
"loss": 0.1528,
"step": 811
},
{
"epoch": 3.2874493927125505,
"grad_norm": 0.3474572690707698,
"learning_rate": 1.2680646661367163e-05,
"loss": 0.1741,
"step": 812
},
{
"epoch": 3.291497975708502,
"grad_norm": 0.6198635944019383,
"learning_rate": 1.2628044982035219e-05,
"loss": 0.2308,
"step": 813
},
{
"epoch": 3.2955465587044532,
"grad_norm": 0.3653813232480011,
"learning_rate": 1.2575502248659155e-05,
"loss": 0.2954,
"step": 814
},
{
"epoch": 3.299595141700405,
"grad_norm": 0.38751519012088936,
"learning_rate": 1.2523018881369269e-05,
"loss": 0.1941,
"step": 815
},
{
"epoch": 3.3036437246963564,
"grad_norm": 0.45900274081472026,
"learning_rate": 1.2470595299821176e-05,
"loss": 0.1828,
"step": 816
},
{
"epoch": 3.3076923076923075,
"grad_norm": 0.33834439266851873,
"learning_rate": 1.2418231923192416e-05,
"loss": 0.1749,
"step": 817
},
{
"epoch": 3.311740890688259,
"grad_norm": 0.3339837574883418,
"learning_rate": 1.2365929170179165e-05,
"loss": 0.1844,
"step": 818
},
{
"epoch": 3.3157894736842106,
"grad_norm": 0.4085401554288196,
"learning_rate": 1.2313687458992848e-05,
"loss": 0.1869,
"step": 819
},
{
"epoch": 3.3198380566801617,
"grad_norm": 0.33471818626171623,
"learning_rate": 1.2261507207356785e-05,
"loss": 0.139,
"step": 820
},
{
"epoch": 3.3238866396761133,
"grad_norm": 0.4001249486180455,
"learning_rate": 1.220938883250288e-05,
"loss": 0.1833,
"step": 821
},
{
"epoch": 3.327935222672065,
"grad_norm": 0.3094315035505681,
"learning_rate": 1.215733275116827e-05,
"loss": 0.2017,
"step": 822
},
{
"epoch": 3.331983805668016,
"grad_norm": 0.37476567876557776,
"learning_rate": 1.2105339379592005e-05,
"loss": 0.2119,
"step": 823
},
{
"epoch": 3.3360323886639676,
"grad_norm": 0.3111739121193045,
"learning_rate": 1.2053409133511681e-05,
"loss": 0.1972,
"step": 824
},
{
"epoch": 3.340080971659919,
"grad_norm": 0.35785427680467147,
"learning_rate": 1.2001542428160173e-05,
"loss": 0.2021,
"step": 825
},
{
"epoch": 3.3441295546558703,
"grad_norm": 0.36569088299957603,
"learning_rate": 1.1949739678262278e-05,
"loss": 0.1951,
"step": 826
},
{
"epoch": 3.348178137651822,
"grad_norm": 0.4079152359457104,
"learning_rate": 1.189800129803139e-05,
"loss": 0.1889,
"step": 827
},
{
"epoch": 3.3522267206477734,
"grad_norm": 0.383862206064576,
"learning_rate": 1.1846327701166228e-05,
"loss": 0.1836,
"step": 828
},
{
"epoch": 3.3562753036437245,
"grad_norm": 0.3521819763025046,
"learning_rate": 1.1794719300847493e-05,
"loss": 0.2283,
"step": 829
},
{
"epoch": 3.360323886639676,
"grad_norm": 0.43549425444047335,
"learning_rate": 1.1743176509734582e-05,
"loss": 0.1942,
"step": 830
},
{
"epoch": 3.3643724696356276,
"grad_norm": 0.3263417523725222,
"learning_rate": 1.1691699739962275e-05,
"loss": 0.1847,
"step": 831
},
{
"epoch": 3.3684210526315788,
"grad_norm": 0.38897906734781984,
"learning_rate": 1.1640289403137456e-05,
"loss": 0.1956,
"step": 832
},
{
"epoch": 3.3724696356275303,
"grad_norm": 0.49885169637682164,
"learning_rate": 1.1588945910335802e-05,
"loss": 0.2006,
"step": 833
},
{
"epoch": 3.376518218623482,
"grad_norm": 0.36523775179770673,
"learning_rate": 1.1537669672098517e-05,
"loss": 0.2166,
"step": 834
},
{
"epoch": 3.380566801619433,
"grad_norm": 0.3192405221766178,
"learning_rate": 1.1486461098429036e-05,
"loss": 0.1968,
"step": 835
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.36720526398462194,
"learning_rate": 1.1435320598789747e-05,
"loss": 0.1906,
"step": 836
},
{
"epoch": 3.388663967611336,
"grad_norm": 0.4290049228584132,
"learning_rate": 1.1384248582098736e-05,
"loss": 0.1762,
"step": 837
},
{
"epoch": 3.3927125506072873,
"grad_norm": 1.7003613768778485,
"learning_rate": 1.1333245456726467e-05,
"loss": 0.2225,
"step": 838
},
{
"epoch": 3.396761133603239,
"grad_norm": 0.3915694740556354,
"learning_rate": 1.128231163049258e-05,
"loss": 0.1944,
"step": 839
},
{
"epoch": 3.4008097165991904,
"grad_norm": 0.4254700286049747,
"learning_rate": 1.123144751066259e-05,
"loss": 0.1905,
"step": 840
},
{
"epoch": 3.4048582995951415,
"grad_norm": 0.32535508143080777,
"learning_rate": 1.1180653503944657e-05,
"loss": 0.1588,
"step": 841
},
{
"epoch": 3.408906882591093,
"grad_norm": 0.3714527244494471,
"learning_rate": 1.1129930016486287e-05,
"loss": 0.2119,
"step": 842
},
{
"epoch": 3.4129554655870447,
"grad_norm": 0.41579106616485856,
"learning_rate": 1.107927745387114e-05,
"loss": 0.183,
"step": 843
},
{
"epoch": 3.417004048582996,
"grad_norm": 0.3462928774776382,
"learning_rate": 1.1028696221115766e-05,
"loss": 0.2158,
"step": 844
},
{
"epoch": 3.4210526315789473,
"grad_norm": 0.3009720414045167,
"learning_rate": 1.0978186722666341e-05,
"loss": 0.1919,
"step": 845
},
{
"epoch": 3.425101214574899,
"grad_norm": 0.3322221317008958,
"learning_rate": 1.0927749362395475e-05,
"loss": 0.1801,
"step": 846
},
{
"epoch": 3.42914979757085,
"grad_norm": 0.38169462065921794,
"learning_rate": 1.0877384543598954e-05,
"loss": 0.209,
"step": 847
},
{
"epoch": 3.4331983805668016,
"grad_norm": 0.3880313818305254,
"learning_rate": 1.0827092668992529e-05,
"loss": 0.1975,
"step": 848
},
{
"epoch": 3.437246963562753,
"grad_norm": 0.3405415773677355,
"learning_rate": 1.077687414070869e-05,
"loss": 0.2003,
"step": 849
},
{
"epoch": 3.4412955465587043,
"grad_norm": 0.3206987068265626,
"learning_rate": 1.0726729360293451e-05,
"loss": 0.179,
"step": 850
},
{
"epoch": 3.445344129554656,
"grad_norm": 0.3437912210045598,
"learning_rate": 1.0676658728703147e-05,
"loss": 0.1814,
"step": 851
},
{
"epoch": 3.4493927125506074,
"grad_norm": 0.31698755082717495,
"learning_rate": 1.0626662646301198e-05,
"loss": 0.179,
"step": 852
},
{
"epoch": 3.4534412955465585,
"grad_norm": 0.32840964320680477,
"learning_rate": 1.0576741512854958e-05,
"loss": 0.1751,
"step": 853
},
{
"epoch": 3.45748987854251,
"grad_norm": 0.3487436201465418,
"learning_rate": 1.052689572753247e-05,
"loss": 0.1739,
"step": 854
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.31421656348399274,
"learning_rate": 1.0477125688899328e-05,
"loss": 0.1666,
"step": 855
},
{
"epoch": 3.465587044534413,
"grad_norm": 0.36794361039906287,
"learning_rate": 1.0427431794915406e-05,
"loss": 0.19,
"step": 856
},
{
"epoch": 3.4696356275303644,
"grad_norm": 0.42446584655437325,
"learning_rate": 1.0377814442931769e-05,
"loss": 0.1721,
"step": 857
},
{
"epoch": 3.473684210526316,
"grad_norm": 0.31881804662535085,
"learning_rate": 1.0328274029687439e-05,
"loss": 0.2169,
"step": 858
},
{
"epoch": 3.477732793522267,
"grad_norm": 0.33473363081803,
"learning_rate": 1.0278810951306253e-05,
"loss": 0.1994,
"step": 859
},
{
"epoch": 3.4817813765182186,
"grad_norm": 0.30963788162018513,
"learning_rate": 1.0229425603293647e-05,
"loss": 0.2054,
"step": 860
},
{
"epoch": 3.48582995951417,
"grad_norm": 0.36188363087896175,
"learning_rate": 1.0180118380533558e-05,
"loss": 0.156,
"step": 861
},
{
"epoch": 3.4898785425101213,
"grad_norm": 0.3121144213972495,
"learning_rate": 1.0130889677285225e-05,
"loss": 0.1756,
"step": 862
},
{
"epoch": 3.493927125506073,
"grad_norm": 0.4008046974094533,
"learning_rate": 1.008173988718005e-05,
"loss": 0.2041,
"step": 863
},
{
"epoch": 3.4979757085020244,
"grad_norm": 0.3412752439852175,
"learning_rate": 1.0032669403218439e-05,
"loss": 0.2021,
"step": 864
},
{
"epoch": 3.5020242914979756,
"grad_norm": 0.3822255568304466,
"learning_rate": 9.983678617766677e-06,
"loss": 0.1898,
"step": 865
},
{
"epoch": 3.506072874493927,
"grad_norm": 0.5739422670968678,
"learning_rate": 9.934767922553788e-06,
"loss": 0.1978,
"step": 866
},
{
"epoch": 3.5101214574898787,
"grad_norm": 0.3478922221351245,
"learning_rate": 9.885937708668371e-06,
"loss": 0.1518,
"step": 867
},
{
"epoch": 3.51417004048583,
"grad_norm": 0.3197137789818536,
"learning_rate": 9.837188366555524e-06,
"loss": 0.187,
"step": 868
},
{
"epoch": 3.5182186234817814,
"grad_norm": 0.33251087367501386,
"learning_rate": 9.788520286013691e-06,
"loss": 0.1915,
"step": 869
},
{
"epoch": 3.522267206477733,
"grad_norm": 0.3020359174737125,
"learning_rate": 9.739933856191537e-06,
"loss": 0.19,
"step": 870
},
{
"epoch": 3.526315789473684,
"grad_norm": 0.4143584524694363,
"learning_rate": 9.691429465584866e-06,
"loss": 0.2363,
"step": 871
},
{
"epoch": 3.5303643724696356,
"grad_norm": 0.43201647009724836,
"learning_rate": 9.643007502033492e-06,
"loss": 0.2285,
"step": 872
},
{
"epoch": 3.534412955465587,
"grad_norm": 0.3481410506029055,
"learning_rate": 9.594668352718152e-06,
"loss": 0.1675,
"step": 873
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.4082817199860991,
"learning_rate": 9.546412404157383e-06,
"loss": 0.1909,
"step": 874
},
{
"epoch": 3.54251012145749,
"grad_norm": 0.3837904843129899,
"learning_rate": 9.498240042204471e-06,
"loss": 0.2107,
"step": 875
},
{
"epoch": 3.5465587044534415,
"grad_norm": 0.4096829353397311,
"learning_rate": 9.450151652044339e-06,
"loss": 0.2239,
"step": 876
},
{
"epoch": 3.5506072874493926,
"grad_norm": 0.3367191617608033,
"learning_rate": 9.402147618190472e-06,
"loss": 0.1605,
"step": 877
},
{
"epoch": 3.554655870445344,
"grad_norm": 0.46263797350765123,
"learning_rate": 9.354228324481841e-06,
"loss": 0.1742,
"step": 878
},
{
"epoch": 3.5587044534412957,
"grad_norm": 0.32752038799796007,
"learning_rate": 9.306394154079852e-06,
"loss": 0.2637,
"step": 879
},
{
"epoch": 3.562753036437247,
"grad_norm": 0.30353519609728236,
"learning_rate": 9.258645489465254e-06,
"loss": 0.2365,
"step": 880
},
{
"epoch": 3.5668016194331984,
"grad_norm": 0.34226133244300616,
"learning_rate": 9.21098271243509e-06,
"loss": 0.2143,
"step": 881
},
{
"epoch": 3.57085020242915,
"grad_norm": 0.355066311048994,
"learning_rate": 9.163406204099659e-06,
"loss": 0.1805,
"step": 882
},
{
"epoch": 3.574898785425101,
"grad_norm": 0.33204466844327185,
"learning_rate": 9.115916344879452e-06,
"loss": 0.1939,
"step": 883
},
{
"epoch": 3.5789473684210527,
"grad_norm": 0.30917533432801503,
"learning_rate": 9.068513514502126e-06,
"loss": 0.2061,
"step": 884
},
{
"epoch": 3.582995951417004,
"grad_norm": 0.3902383788239698,
"learning_rate": 9.021198091999439e-06,
"loss": 0.2214,
"step": 885
},
{
"epoch": 3.5870445344129553,
"grad_norm": 0.3576818788748052,
"learning_rate": 8.973970455704251e-06,
"loss": 0.1918,
"step": 886
},
{
"epoch": 3.591093117408907,
"grad_norm": 0.3526462240766607,
"learning_rate": 8.926830983247495e-06,
"loss": 0.193,
"step": 887
},
{
"epoch": 3.5951417004048585,
"grad_norm": 0.32256472556977,
"learning_rate": 8.879780051555116e-06,
"loss": 0.1698,
"step": 888
},
{
"epoch": 3.5991902834008096,
"grad_norm": 0.35877636001643803,
"learning_rate": 8.83281803684512e-06,
"loss": 0.2145,
"step": 889
},
{
"epoch": 3.603238866396761,
"grad_norm": 0.38593002901565876,
"learning_rate": 8.785945314624532e-06,
"loss": 0.1737,
"step": 890
},
{
"epoch": 3.6072874493927127,
"grad_norm": 0.32310202493496654,
"learning_rate": 8.739162259686385e-06,
"loss": 0.2008,
"step": 891
},
{
"epoch": 3.611336032388664,
"grad_norm": 0.36334691603682523,
"learning_rate": 8.692469246106734e-06,
"loss": 0.1943,
"step": 892
},
{
"epoch": 3.6153846153846154,
"grad_norm": 0.36737759183034935,
"learning_rate": 8.645866647241677e-06,
"loss": 0.1744,
"step": 893
},
{
"epoch": 3.619433198380567,
"grad_norm": 0.5279172318771499,
"learning_rate": 8.599354835724348e-06,
"loss": 0.2588,
"step": 894
},
{
"epoch": 3.623481781376518,
"grad_norm": 0.6619856480032166,
"learning_rate": 8.552934183461958e-06,
"loss": 0.2063,
"step": 895
},
{
"epoch": 3.6275303643724697,
"grad_norm": 0.4564335124768016,
"learning_rate": 8.506605061632805e-06,
"loss": 0.1925,
"step": 896
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.41588536531770937,
"learning_rate": 8.460367840683308e-06,
"loss": 0.2715,
"step": 897
},
{
"epoch": 3.6356275303643724,
"grad_norm": 0.45679426899036013,
"learning_rate": 8.414222890325065e-06,
"loss": 0.1979,
"step": 898
},
{
"epoch": 3.639676113360324,
"grad_norm": 0.489178930411812,
"learning_rate": 8.368170579531855e-06,
"loss": 0.2116,
"step": 899
},
{
"epoch": 3.6437246963562755,
"grad_norm": 0.3377746492610638,
"learning_rate": 8.322211276536733e-06,
"loss": 0.2135,
"step": 900
},
{
"epoch": 3.6477732793522266,
"grad_norm": 0.33822225101076725,
"learning_rate": 8.27634534882907e-06,
"loss": 0.1965,
"step": 901
},
{
"epoch": 3.651821862348178,
"grad_norm": 0.3505371040916298,
"learning_rate": 8.230573163151585e-06,
"loss": 0.1951,
"step": 902
},
{
"epoch": 3.6558704453441297,
"grad_norm": 0.2951664075828084,
"learning_rate": 8.184895085497466e-06,
"loss": 0.1926,
"step": 903
},
{
"epoch": 3.659919028340081,
"grad_norm": 0.46089752018946656,
"learning_rate": 8.139311481107401e-06,
"loss": 0.2146,
"step": 904
},
{
"epoch": 3.6639676113360324,
"grad_norm": 0.34406319099780924,
"learning_rate": 8.093822714466686e-06,
"loss": 0.2471,
"step": 905
},
{
"epoch": 3.668016194331984,
"grad_norm": 0.378556263640653,
"learning_rate": 8.048429149302271e-06,
"loss": 0.209,
"step": 906
},
{
"epoch": 3.672064777327935,
"grad_norm": 0.3239395521010079,
"learning_rate": 8.003131148579902e-06,
"loss": 0.1514,
"step": 907
},
{
"epoch": 3.6761133603238867,
"grad_norm": 0.4654398150284495,
"learning_rate": 7.957929074501182e-06,
"loss": 0.1745,
"step": 908
},
{
"epoch": 3.6801619433198383,
"grad_norm": 0.4560343489848329,
"learning_rate": 7.912823288500695e-06,
"loss": 0.1942,
"step": 909
},
{
"epoch": 3.6842105263157894,
"grad_norm": 0.3679826970581447,
"learning_rate": 7.867814151243102e-06,
"loss": 0.2167,
"step": 910
},
{
"epoch": 3.688259109311741,
"grad_norm": 2.147157260579936,
"learning_rate": 7.822902022620267e-06,
"loss": 0.2009,
"step": 911
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.37831384062299395,
"learning_rate": 7.778087261748377e-06,
"loss": 0.1979,
"step": 912
},
{
"epoch": 3.6963562753036436,
"grad_norm": 0.35006673918853104,
"learning_rate": 7.733370226965056e-06,
"loss": 0.1578,
"step": 913
},
{
"epoch": 3.700404858299595,
"grad_norm": 0.43984048475488574,
"learning_rate": 7.688751275826527e-06,
"loss": 0.1905,
"step": 914
},
{
"epoch": 3.7044534412955468,
"grad_norm": 0.32889792531246786,
"learning_rate": 7.644230765104735e-06,
"loss": 0.1723,
"step": 915
},
{
"epoch": 3.708502024291498,
"grad_norm": 0.42182862791010456,
"learning_rate": 7.599809050784502e-06,
"loss": 0.1902,
"step": 916
},
{
"epoch": 3.7125506072874495,
"grad_norm": 0.43153466435651,
"learning_rate": 7.555486488060661e-06,
"loss": 0.1856,
"step": 917
},
{
"epoch": 3.716599190283401,
"grad_norm": 0.5054653318602943,
"learning_rate": 7.511263431335249e-06,
"loss": 0.1859,
"step": 918
},
{
"epoch": 3.720647773279352,
"grad_norm": 0.35595840125731004,
"learning_rate": 7.467140234214656e-06,
"loss": 0.1687,
"step": 919
},
{
"epoch": 3.7246963562753037,
"grad_norm": 0.3544079534329402,
"learning_rate": 7.423117249506777e-06,
"loss": 0.1817,
"step": 920
},
{
"epoch": 3.7287449392712553,
"grad_norm": 0.411189565196018,
"learning_rate": 7.379194829218231e-06,
"loss": 0.1939,
"step": 921
},
{
"epoch": 3.7327935222672064,
"grad_norm": 0.33436277661710506,
"learning_rate": 7.335373324551518e-06,
"loss": 0.2027,
"step": 922
},
{
"epoch": 3.736842105263158,
"grad_norm": 0.3082265638770454,
"learning_rate": 7.291653085902224e-06,
"loss": 0.1837,
"step": 923
},
{
"epoch": 3.7408906882591095,
"grad_norm": 0.37049194942030683,
"learning_rate": 7.248034462856213e-06,
"loss": 0.1665,
"step": 924
},
{
"epoch": 3.7449392712550607,
"grad_norm": 0.34566668712473825,
"learning_rate": 7.20451780418683e-06,
"loss": 0.1995,
"step": 925
},
{
"epoch": 3.748987854251012,
"grad_norm": 0.35782245169207916,
"learning_rate": 7.161103457852128e-06,
"loss": 0.1965,
"step": 926
},
{
"epoch": 3.753036437246964,
"grad_norm": 0.3194860273537156,
"learning_rate": 7.117791770992046e-06,
"loss": 0.1792,
"step": 927
},
{
"epoch": 3.757085020242915,
"grad_norm": 0.3120208187892621,
"learning_rate": 7.074583089925688e-06,
"loss": 0.1659,
"step": 928
},
{
"epoch": 3.7611336032388665,
"grad_norm": 0.3548625545635257,
"learning_rate": 7.0314777601485155e-06,
"loss": 0.1877,
"step": 929
},
{
"epoch": 3.765182186234818,
"grad_norm": 0.3198869428387834,
"learning_rate": 6.988476126329604e-06,
"loss": 0.1869,
"step": 930
},
{
"epoch": 3.769230769230769,
"grad_norm": 0.3098316588030814,
"learning_rate": 6.945578532308857e-06,
"loss": 0.1824,
"step": 931
},
{
"epoch": 3.7732793522267207,
"grad_norm": 0.48933499005529835,
"learning_rate": 6.902785321094301e-06,
"loss": 0.2055,
"step": 932
},
{
"epoch": 3.7773279352226723,
"grad_norm": 0.46476217170660333,
"learning_rate": 6.8600968348593115e-06,
"loss": 0.1797,
"step": 933
},
{
"epoch": 3.7813765182186234,
"grad_norm": 0.3253070783870269,
"learning_rate": 6.817513414939892e-06,
"loss": 0.2295,
"step": 934
},
{
"epoch": 3.785425101214575,
"grad_norm": 0.32763757091653367,
"learning_rate": 6.775035401831924e-06,
"loss": 0.2066,
"step": 935
},
{
"epoch": 3.7894736842105265,
"grad_norm": 0.3223878305440969,
"learning_rate": 6.732663135188473e-06,
"loss": 0.2071,
"step": 936
},
{
"epoch": 3.7935222672064777,
"grad_norm": 0.4008244022810568,
"learning_rate": 6.690396953817055e-06,
"loss": 0.2478,
"step": 937
},
{
"epoch": 3.7975708502024292,
"grad_norm": 0.34257306190021997,
"learning_rate": 6.648237195676927e-06,
"loss": 0.1738,
"step": 938
},
{
"epoch": 3.801619433198381,
"grad_norm": 0.3273838011164468,
"learning_rate": 6.606184197876395e-06,
"loss": 0.1969,
"step": 939
},
{
"epoch": 3.805668016194332,
"grad_norm": 0.3453945199145201,
"learning_rate": 6.564238296670109e-06,
"loss": 0.179,
"step": 940
},
{
"epoch": 3.8097165991902835,
"grad_norm": 0.31167868086264816,
"learning_rate": 6.52239982745638e-06,
"loss": 0.1691,
"step": 941
},
{
"epoch": 3.813765182186235,
"grad_norm": 0.5576551988554141,
"learning_rate": 6.480669124774484e-06,
"loss": 0.1915,
"step": 942
},
{
"epoch": 3.817813765182186,
"grad_norm": 0.31032305663830706,
"learning_rate": 6.439046522302013e-06,
"loss": 0.222,
"step": 943
},
{
"epoch": 3.8218623481781377,
"grad_norm": 0.33610215000665733,
"learning_rate": 6.397532352852191e-06,
"loss": 0.1884,
"step": 944
},
{
"epoch": 3.8259109311740893,
"grad_norm": 0.3144247386603838,
"learning_rate": 6.356126948371202e-06,
"loss": 0.1966,
"step": 945
},
{
"epoch": 3.8299595141700404,
"grad_norm": 0.38394747994198175,
"learning_rate": 6.314830639935559e-06,
"loss": 0.1944,
"step": 946
},
{
"epoch": 3.834008097165992,
"grad_norm": 0.5900051987136615,
"learning_rate": 6.273643757749448e-06,
"loss": 0.2198,
"step": 947
},
{
"epoch": 3.8380566801619436,
"grad_norm": 0.3065595717269623,
"learning_rate": 6.2325666311420856e-06,
"loss": 0.1877,
"step": 948
},
{
"epoch": 3.8421052631578947,
"grad_norm": 0.38863847512361255,
"learning_rate": 6.1915995885650675e-06,
"loss": 0.1774,
"step": 949
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.3110993552802247,
"learning_rate": 6.150742957589784e-06,
"loss": 0.2333,
"step": 950
},
{
"epoch": 3.850202429149798,
"grad_norm": 0.30643745160482655,
"learning_rate": 6.109997064904774e-06,
"loss": 0.188,
"step": 951
},
{
"epoch": 3.854251012145749,
"grad_norm": 0.3148192507158416,
"learning_rate": 6.069362236313092e-06,
"loss": 0.1859,
"step": 952
},
{
"epoch": 3.8582995951417005,
"grad_norm": 0.3271728436420623,
"learning_rate": 6.0288387967297565e-06,
"loss": 0.1963,
"step": 953
},
{
"epoch": 3.862348178137652,
"grad_norm": 0.29927646000242525,
"learning_rate": 5.988427070179108e-06,
"loss": 0.2003,
"step": 954
},
{
"epoch": 3.866396761133603,
"grad_norm": 0.3273520969877742,
"learning_rate": 5.948127379792234e-06,
"loss": 0.1824,
"step": 955
},
{
"epoch": 3.8704453441295548,
"grad_norm": 0.2995358888216968,
"learning_rate": 5.90794004780439e-06,
"loss": 0.169,
"step": 956
},
{
"epoch": 3.8744939271255063,
"grad_norm": 0.32667372623910734,
"learning_rate": 5.867865395552405e-06,
"loss": 0.1616,
"step": 957
},
{
"epoch": 3.8785425101214575,
"grad_norm": 0.3183033831937247,
"learning_rate": 5.8279037434721364e-06,
"loss": 0.2153,
"step": 958
},
{
"epoch": 3.882591093117409,
"grad_norm": 0.33044062022547,
"learning_rate": 5.788055411095892e-06,
"loss": 0.1909,
"step": 959
},
{
"epoch": 3.8866396761133606,
"grad_norm": 0.31110542284923676,
"learning_rate": 5.7483207170498625e-06,
"loss": 0.2111,
"step": 960
},
{
"epoch": 3.8906882591093117,
"grad_norm": 0.32471745932237517,
"learning_rate": 5.708699979051613e-06,
"loss": 0.1533,
"step": 961
},
{
"epoch": 3.8947368421052633,
"grad_norm": 0.32436492353907886,
"learning_rate": 5.66919351390751e-06,
"loss": 0.1916,
"step": 962
},
{
"epoch": 3.898785425101215,
"grad_norm": 0.4312349559144098,
"learning_rate": 5.629801637510188e-06,
"loss": 0.1781,
"step": 963
},
{
"epoch": 3.902834008097166,
"grad_norm": 0.4641798392833268,
"learning_rate": 5.590524664836043e-06,
"loss": 0.2019,
"step": 964
},
{
"epoch": 3.9068825910931175,
"grad_norm": 0.3530397321356778,
"learning_rate": 5.551362909942708e-06,
"loss": 0.1966,
"step": 965
},
{
"epoch": 3.910931174089069,
"grad_norm": 0.41832205870375205,
"learning_rate": 5.51231668596653e-06,
"loss": 0.2049,
"step": 966
},
{
"epoch": 3.91497975708502,
"grad_norm": 0.37491683211549987,
"learning_rate": 5.47338630512007e-06,
"loss": 0.1932,
"step": 967
},
{
"epoch": 3.919028340080972,
"grad_norm": 0.3298884809019581,
"learning_rate": 5.434572078689617e-06,
"loss": 0.188,
"step": 968
},
{
"epoch": 3.9230769230769234,
"grad_norm": 0.327106005542329,
"learning_rate": 5.395874317032694e-06,
"loss": 0.1787,
"step": 969
},
{
"epoch": 3.9271255060728745,
"grad_norm": 1.0722569383780576,
"learning_rate": 5.357293329575573e-06,
"loss": 0.2425,
"step": 970
},
{
"epoch": 3.931174089068826,
"grad_norm": 0.32854450513561706,
"learning_rate": 5.318829424810797e-06,
"loss": 0.2396,
"step": 971
},
{
"epoch": 3.9352226720647776,
"grad_norm": 0.5776727210748016,
"learning_rate": 5.280482910294728e-06,
"loss": 0.2058,
"step": 972
},
{
"epoch": 3.9392712550607287,
"grad_norm": 0.4922472782720792,
"learning_rate": 5.242254092645078e-06,
"loss": 0.2167,
"step": 973
},
{
"epoch": 3.9433198380566803,
"grad_norm": 0.32663609961783235,
"learning_rate": 5.2041432775384445e-06,
"loss": 0.166,
"step": 974
},
{
"epoch": 3.9473684210526314,
"grad_norm": 0.3808307497105545,
"learning_rate": 5.166150769707892e-06,
"loss": 0.1875,
"step": 975
},
{
"epoch": 3.951417004048583,
"grad_norm": 0.3291002683513231,
"learning_rate": 5.128276872940503e-06,
"loss": 0.2113,
"step": 976
},
{
"epoch": 3.9554655870445345,
"grad_norm": 0.7166189122843486,
"learning_rate": 5.090521890074953e-06,
"loss": 0.1929,
"step": 977
},
{
"epoch": 3.9595141700404857,
"grad_norm": 0.3674856844317609,
"learning_rate": 5.0528861229990675e-06,
"loss": 0.2073,
"step": 978
},
{
"epoch": 3.9635627530364372,
"grad_norm": 0.3273292077524206,
"learning_rate": 5.015369872647442e-06,
"loss": 0.1697,
"step": 979
},
{
"epoch": 3.967611336032389,
"grad_norm": 0.3349648127454134,
"learning_rate": 4.9779734389990195e-06,
"loss": 0.1462,
"step": 980
},
{
"epoch": 3.97165991902834,
"grad_norm": 0.29416612948735005,
"learning_rate": 4.940697121074675e-06,
"loss": 0.1663,
"step": 981
},
{
"epoch": 3.9757085020242915,
"grad_norm": 0.33418669452361977,
"learning_rate": 4.903541216934864e-06,
"loss": 0.1903,
"step": 982
},
{
"epoch": 3.979757085020243,
"grad_norm": 0.4090023859746256,
"learning_rate": 4.8665060236772e-06,
"loss": 0.1704,
"step": 983
},
{
"epoch": 3.983805668016194,
"grad_norm": 0.3081879478624426,
"learning_rate": 4.829591837434102e-06,
"loss": 0.2019,
"step": 984
},
{
"epoch": 3.9878542510121457,
"grad_norm": 0.3013454227413771,
"learning_rate": 4.792798953370424e-06,
"loss": 0.1898,
"step": 985
},
{
"epoch": 3.9919028340080973,
"grad_norm": 0.43334515093609627,
"learning_rate": 4.756127665681083e-06,
"loss": 0.2021,
"step": 986
},
{
"epoch": 3.9959514170040484,
"grad_norm": 0.2931221154846842,
"learning_rate": 4.719578267588725e-06,
"loss": 0.1835,
"step": 987
},
{
"epoch": 4.0,
"grad_norm": 0.32098191941211907,
"learning_rate": 4.6831510513413525e-06,
"loss": 0.1582,
"step": 988
},
{
"epoch": 4.004048582995951,
"grad_norm": 0.30637568587985525,
"learning_rate": 4.6468463082100176e-06,
"loss": 0.1673,
"step": 989
},
{
"epoch": 4.008097165991903,
"grad_norm": 0.25043789731935634,
"learning_rate": 4.610664328486485e-06,
"loss": 0.1727,
"step": 990
},
{
"epoch": 4.012145748987854,
"grad_norm": 0.2384251079627586,
"learning_rate": 4.5746054014809e-06,
"loss": 0.1372,
"step": 991
},
{
"epoch": 4.016194331983805,
"grad_norm": 0.21336972941446272,
"learning_rate": 4.5386698155194766e-06,
"loss": 0.1346,
"step": 992
},
{
"epoch": 4.020242914979757,
"grad_norm": 0.21465739259891164,
"learning_rate": 4.502857857942204e-06,
"loss": 0.1521,
"step": 993
},
{
"epoch": 4.0242914979757085,
"grad_norm": 0.22359638061597334,
"learning_rate": 4.467169815100549e-06,
"loss": 0.1642,
"step": 994
},
{
"epoch": 4.02834008097166,
"grad_norm": 0.2535976427667772,
"learning_rate": 4.431605972355137e-06,
"loss": 0.1416,
"step": 995
},
{
"epoch": 4.032388663967612,
"grad_norm": 0.2590334674560289,
"learning_rate": 4.396166614073516e-06,
"loss": 0.1643,
"step": 996
},
{
"epoch": 4.036437246963563,
"grad_norm": 0.2428758965344956,
"learning_rate": 4.360852023627846e-06,
"loss": 0.1422,
"step": 997
},
{
"epoch": 4.040485829959514,
"grad_norm": 0.22911960061056788,
"learning_rate": 4.325662483392656e-06,
"loss": 0.1495,
"step": 998
},
{
"epoch": 4.044534412955466,
"grad_norm": 0.24315754872154474,
"learning_rate": 4.29059827474257e-06,
"loss": 0.1556,
"step": 999
},
{
"epoch": 4.048582995951417,
"grad_norm": 0.22250895414873575,
"learning_rate": 4.255659678050063e-06,
"loss": 0.1404,
"step": 1000
},
{
"epoch": 4.052631578947368,
"grad_norm": 0.22737985573603037,
"learning_rate": 4.22084697268323e-06,
"loss": 0.1275,
"step": 1001
},
{
"epoch": 4.05668016194332,
"grad_norm": 0.23396571916342662,
"learning_rate": 4.186160437003535e-06,
"loss": 0.127,
"step": 1002
},
{
"epoch": 4.060728744939271,
"grad_norm": 0.21435595523151432,
"learning_rate": 4.151600348363582e-06,
"loss": 0.1449,
"step": 1003
},
{
"epoch": 4.064777327935222,
"grad_norm": 0.22697465296643657,
"learning_rate": 4.11716698310493e-06,
"loss": 0.1722,
"step": 1004
},
{
"epoch": 4.068825910931174,
"grad_norm": 0.21265434379853587,
"learning_rate": 4.082860616555848e-06,
"loss": 0.1397,
"step": 1005
},
{
"epoch": 4.0728744939271255,
"grad_norm": 0.20875251162101438,
"learning_rate": 4.048681523029125e-06,
"loss": 0.1766,
"step": 1006
},
{
"epoch": 4.076923076923077,
"grad_norm": 0.19533912932956926,
"learning_rate": 4.014629975819882e-06,
"loss": 0.1762,
"step": 1007
},
{
"epoch": 4.080971659919029,
"grad_norm": 0.22132254650955624,
"learning_rate": 3.9807062472033855e-06,
"loss": 0.1257,
"step": 1008
},
{
"epoch": 4.08502024291498,
"grad_norm": 0.22704200182882345,
"learning_rate": 3.94691060843287e-06,
"loss": 0.1752,
"step": 1009
},
{
"epoch": 4.089068825910931,
"grad_norm": 0.31484557829373877,
"learning_rate": 3.913243329737353e-06,
"loss": 0.1734,
"step": 1010
},
{
"epoch": 4.093117408906883,
"grad_norm": 0.22586613174314385,
"learning_rate": 3.879704680319503e-06,
"loss": 0.163,
"step": 1011
},
{
"epoch": 4.097165991902834,
"grad_norm": 0.22391656183934527,
"learning_rate": 3.846294928353471e-06,
"loss": 0.1603,
"step": 1012
},
{
"epoch": 4.101214574898785,
"grad_norm": 0.22210320569999367,
"learning_rate": 3.8130143409827345e-06,
"loss": 0.1632,
"step": 1013
},
{
"epoch": 4.105263157894737,
"grad_norm": 0.21070872638441018,
"learning_rate": 3.7798631843179887e-06,
"loss": 0.1634,
"step": 1014
},
{
"epoch": 4.109311740890688,
"grad_norm": 0.2121590351915006,
"learning_rate": 3.7468417234349973e-06,
"loss": 0.1589,
"step": 1015
},
{
"epoch": 4.113360323886639,
"grad_norm": 0.2544137433520658,
"learning_rate": 3.7139502223724823e-06,
"loss": 0.1589,
"step": 1016
},
{
"epoch": 4.117408906882591,
"grad_norm": 0.19161667684772057,
"learning_rate": 3.6811889441300095e-06,
"loss": 0.1459,
"step": 1017
},
{
"epoch": 4.1214574898785425,
"grad_norm": 0.22603776060324327,
"learning_rate": 3.648558150665886e-06,
"loss": 0.1595,
"step": 1018
},
{
"epoch": 4.125506072874494,
"grad_norm": 0.20109293814342907,
"learning_rate": 3.616058102895075e-06,
"loss": 0.1647,
"step": 1019
},
{
"epoch": 4.129554655870446,
"grad_norm": 0.2162057157662541,
"learning_rate": 3.583689060687079e-06,
"loss": 0.1706,
"step": 1020
},
{
"epoch": 4.133603238866397,
"grad_norm": 0.20288984593346035,
"learning_rate": 3.5514512828639023e-06,
"loss": 0.1432,
"step": 1021
},
{
"epoch": 4.137651821862348,
"grad_norm": 0.2161088067095598,
"learning_rate": 3.51934502719796e-06,
"loss": 0.1575,
"step": 1022
},
{
"epoch": 4.1417004048583,
"grad_norm": 0.22609941033866293,
"learning_rate": 3.4873705504100185e-06,
"loss": 0.2097,
"step": 1023
},
{
"epoch": 4.145748987854251,
"grad_norm": 0.20825339905166154,
"learning_rate": 3.4555281081671345e-06,
"loss": 0.1467,
"step": 1024
},
{
"epoch": 4.149797570850202,
"grad_norm": 0.1964847366947185,
"learning_rate": 3.4238179550806326e-06,
"loss": 0.1531,
"step": 1025
},
{
"epoch": 4.153846153846154,
"grad_norm": 0.21869161949555094,
"learning_rate": 3.3922403447040497e-06,
"loss": 0.1486,
"step": 1026
},
{
"epoch": 4.157894736842105,
"grad_norm": 0.22654984784271182,
"learning_rate": 3.3607955295311247e-06,
"loss": 0.1976,
"step": 1027
},
{
"epoch": 4.161943319838056,
"grad_norm": 0.2701814629607038,
"learning_rate": 3.32948376099375e-06,
"loss": 0.1907,
"step": 1028
},
{
"epoch": 4.165991902834008,
"grad_norm": 0.2208564137107608,
"learning_rate": 3.2983052894599976e-06,
"loss": 0.1515,
"step": 1029
},
{
"epoch": 4.17004048582996,
"grad_norm": 0.23360541032444251,
"learning_rate": 3.2672603642320945e-06,
"loss": 0.1663,
"step": 1030
},
{
"epoch": 4.174089068825911,
"grad_norm": 0.20590765430048533,
"learning_rate": 3.236349233544438e-06,
"loss": 0.1432,
"step": 1031
},
{
"epoch": 4.178137651821863,
"grad_norm": 0.21170863655903652,
"learning_rate": 3.205572144561604e-06,
"loss": 0.1774,
"step": 1032
},
{
"epoch": 4.182186234817814,
"grad_norm": 0.22750586919812252,
"learning_rate": 3.174929343376374e-06,
"loss": 0.1802,
"step": 1033
},
{
"epoch": 4.186234817813765,
"grad_norm": 0.21808714678779093,
"learning_rate": 3.1444210750077777e-06,
"loss": 0.133,
"step": 1034
},
{
"epoch": 4.190283400809717,
"grad_norm": 0.21992636793152326,
"learning_rate": 3.1140475833991046e-06,
"loss": 0.1839,
"step": 1035
},
{
"epoch": 4.194331983805668,
"grad_norm": 0.23446566950917616,
"learning_rate": 3.0838091114159895e-06,
"loss": 0.2026,
"step": 1036
},
{
"epoch": 4.198380566801619,
"grad_norm": 0.21311652085785393,
"learning_rate": 3.0537059008444524e-06,
"loss": 0.1851,
"step": 1037
},
{
"epoch": 4.202429149797571,
"grad_norm": 0.2169200118497063,
"learning_rate": 3.0237381923889563e-06,
"loss": 0.1805,
"step": 1038
},
{
"epoch": 4.206477732793522,
"grad_norm": 0.2185244117372126,
"learning_rate": 2.9939062256704997e-06,
"loss": 0.1748,
"step": 1039
},
{
"epoch": 4.2105263157894735,
"grad_norm": 0.223738900819331,
"learning_rate": 2.964210239224696e-06,
"loss": 0.1611,
"step": 1040
},
{
"epoch": 4.2145748987854255,
"grad_norm": 0.248110207679201,
"learning_rate": 2.934650470499869e-06,
"loss": 0.1623,
"step": 1041
},
{
"epoch": 4.218623481781377,
"grad_norm": 0.23894950246920824,
"learning_rate": 2.905227155855128e-06,
"loss": 0.1536,
"step": 1042
},
{
"epoch": 4.222672064777328,
"grad_norm": 0.2354142507419388,
"learning_rate": 2.875940530558519e-06,
"loss": 0.1901,
"step": 1043
},
{
"epoch": 4.22672064777328,
"grad_norm": 0.24540383927082926,
"learning_rate": 2.8467908287851154e-06,
"loss": 0.1994,
"step": 1044
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.20984905557980657,
"learning_rate": 2.8177782836151603e-06,
"loss": 0.1537,
"step": 1045
},
{
"epoch": 4.234817813765182,
"grad_norm": 0.2293644427259079,
"learning_rate": 2.7889031270321874e-06,
"loss": 0.1982,
"step": 1046
},
{
"epoch": 4.238866396761134,
"grad_norm": 0.2078565952440226,
"learning_rate": 2.760165589921184e-06,
"loss": 0.1951,
"step": 1047
},
{
"epoch": 4.242914979757085,
"grad_norm": 0.2216992464184189,
"learning_rate": 2.7315659020667305e-06,
"loss": 0.1662,
"step": 1048
},
{
"epoch": 4.246963562753036,
"grad_norm": 0.35767984474570874,
"learning_rate": 2.703104292151164e-06,
"loss": 0.1614,
"step": 1049
},
{
"epoch": 4.251012145748988,
"grad_norm": 0.2088975713843415,
"learning_rate": 2.6747809877527607e-06,
"loss": 0.1405,
"step": 1050
},
{
"epoch": 4.255060728744939,
"grad_norm": 0.22488920466535145,
"learning_rate": 2.6465962153439082e-06,
"loss": 0.142,
"step": 1051
},
{
"epoch": 4.2591093117408905,
"grad_norm": 0.189155779231829,
"learning_rate": 2.618550200289303e-06,
"loss": 0.1664,
"step": 1052
},
{
"epoch": 4.2631578947368425,
"grad_norm": 0.22067924022113414,
"learning_rate": 2.59064316684412e-06,
"loss": 0.1503,
"step": 1053
},
{
"epoch": 4.267206477732794,
"grad_norm": 0.19917968784611217,
"learning_rate": 2.562875338152264e-06,
"loss": 0.1328,
"step": 1054
},
{
"epoch": 4.271255060728745,
"grad_norm": 0.41636791014030744,
"learning_rate": 2.5352469362445552e-06,
"loss": 0.1557,
"step": 1055
},
{
"epoch": 4.275303643724697,
"grad_norm": 0.21022255517985752,
"learning_rate": 2.5077581820369523e-06,
"loss": 0.1677,
"step": 1056
},
{
"epoch": 4.279352226720648,
"grad_norm": 0.25483745843510586,
"learning_rate": 2.480409295328803e-06,
"loss": 0.1356,
"step": 1057
},
{
"epoch": 4.283400809716599,
"grad_norm": 0.22879870490077578,
"learning_rate": 2.45320049480108e-06,
"loss": 0.1994,
"step": 1058
},
{
"epoch": 4.287449392712551,
"grad_norm": 0.20960694369703398,
"learning_rate": 2.4261319980146293e-06,
"loss": 0.1478,
"step": 1059
},
{
"epoch": 4.291497975708502,
"grad_norm": 0.21064578983604818,
"learning_rate": 2.399204021408421e-06,
"loss": 0.1567,
"step": 1060
},
{
"epoch": 4.295546558704453,
"grad_norm": 0.20183319382915318,
"learning_rate": 2.3724167802978395e-06,
"loss": 0.1691,
"step": 1061
},
{
"epoch": 4.299595141700405,
"grad_norm": 0.1910145664645432,
"learning_rate": 2.3457704888729625e-06,
"loss": 0.1412,
"step": 1062
},
{
"epoch": 4.303643724696356,
"grad_norm": 0.22435010789455495,
"learning_rate": 2.3192653601968162e-06,
"loss": 0.1211,
"step": 1063
},
{
"epoch": 4.3076923076923075,
"grad_norm": 0.3442858133149616,
"learning_rate": 2.2929016062037057e-06,
"loss": 0.1519,
"step": 1064
},
{
"epoch": 4.3117408906882595,
"grad_norm": 0.2218029672624173,
"learning_rate": 2.2666794376975055e-06,
"loss": 0.1716,
"step": 1065
},
{
"epoch": 4.315789473684211,
"grad_norm": 0.9479688705776099,
"learning_rate": 2.24059906434998e-06,
"loss": 0.1928,
"step": 1066
},
{
"epoch": 4.319838056680162,
"grad_norm": 0.218204963474193,
"learning_rate": 2.214660694699091e-06,
"loss": 0.2091,
"step": 1067
},
{
"epoch": 4.323886639676114,
"grad_norm": 0.23112847069294906,
"learning_rate": 2.188864536147355e-06,
"loss": 0.1691,
"step": 1068
},
{
"epoch": 4.327935222672065,
"grad_norm": 0.2145066494818169,
"learning_rate": 2.1632107949601757e-06,
"loss": 0.1631,
"step": 1069
},
{
"epoch": 4.331983805668016,
"grad_norm": 0.19601762087329422,
"learning_rate": 2.137699676264171e-06,
"loss": 0.1596,
"step": 1070
},
{
"epoch": 4.336032388663968,
"grad_norm": 0.21858295211203707,
"learning_rate": 2.1123313840455738e-06,
"loss": 0.1496,
"step": 1071
},
{
"epoch": 4.340080971659919,
"grad_norm": 0.23372312680114538,
"learning_rate": 2.08710612114857e-06,
"loss": 0.1769,
"step": 1072
},
{
"epoch": 4.34412955465587,
"grad_norm": 0.22600061431008292,
"learning_rate": 2.0620240892736953e-06,
"loss": 0.1551,
"step": 1073
},
{
"epoch": 4.348178137651822,
"grad_norm": 0.20597593135817527,
"learning_rate": 2.0370854889762e-06,
"loss": 0.1527,
"step": 1074
},
{
"epoch": 4.352226720647773,
"grad_norm": 0.28583151325535894,
"learning_rate": 2.0122905196644726e-06,
"loss": 0.1327,
"step": 1075
},
{
"epoch": 4.3562753036437245,
"grad_norm": 0.2180042981751587,
"learning_rate": 1.987639379598427e-06,
"loss": 0.1416,
"step": 1076
},
{
"epoch": 4.3603238866396765,
"grad_norm": 0.2058051780962486,
"learning_rate": 1.963132265887919e-06,
"loss": 0.166,
"step": 1077
},
{
"epoch": 4.364372469635628,
"grad_norm": 0.19955929868244662,
"learning_rate": 1.9387693744911784e-06,
"loss": 0.1419,
"step": 1078
},
{
"epoch": 4.368421052631579,
"grad_norm": 0.2093203039685621,
"learning_rate": 1.9145509002132368e-06,
"loss": 0.1569,
"step": 1079
},
{
"epoch": 4.372469635627531,
"grad_norm": 0.2163528557633722,
"learning_rate": 1.8904770367043723e-06,
"loss": 0.1585,
"step": 1080
},
{
"epoch": 4.376518218623482,
"grad_norm": 0.236412526750367,
"learning_rate": 1.8665479764585504e-06,
"loss": 0.1394,
"step": 1081
},
{
"epoch": 4.380566801619433,
"grad_norm": 0.2276058891355952,
"learning_rate": 1.8427639108118977e-06,
"loss": 0.1591,
"step": 1082
},
{
"epoch": 4.384615384615385,
"grad_norm": 0.2364781898913223,
"learning_rate": 1.8191250299411755e-06,
"loss": 0.1546,
"step": 1083
},
{
"epoch": 4.388663967611336,
"grad_norm": 0.20527873619674436,
"learning_rate": 1.7956315228622446e-06,
"loss": 0.1722,
"step": 1084
},
{
"epoch": 4.392712550607287,
"grad_norm": 0.20646962212051848,
"learning_rate": 1.7722835774285553e-06,
"loss": 0.1614,
"step": 1085
},
{
"epoch": 4.396761133603239,
"grad_norm": 0.2657816491094132,
"learning_rate": 1.7490813803296624e-06,
"loss": 0.1992,
"step": 1086
},
{
"epoch": 4.40080971659919,
"grad_norm": 0.20309421764869837,
"learning_rate": 1.726025117089718e-06,
"loss": 0.1348,
"step": 1087
},
{
"epoch": 4.4048582995951415,
"grad_norm": 0.21076026799590744,
"learning_rate": 1.70311497206598e-06,
"loss": 0.1561,
"step": 1088
},
{
"epoch": 4.4089068825910935,
"grad_norm": 0.2258957268024169,
"learning_rate": 1.6803511284473638e-06,
"loss": 0.1218,
"step": 1089
},
{
"epoch": 4.412955465587045,
"grad_norm": 0.20994047535138147,
"learning_rate": 1.6577337682529538e-06,
"loss": 0.1413,
"step": 1090
},
{
"epoch": 4.417004048582996,
"grad_norm": 0.21509776180160262,
"learning_rate": 1.6352630723305617e-06,
"loss": 0.1452,
"step": 1091
},
{
"epoch": 4.421052631578947,
"grad_norm": 0.20927009258402973,
"learning_rate": 1.612939220355272e-06,
"loss": 0.1479,
"step": 1092
},
{
"epoch": 4.425101214574899,
"grad_norm": 0.1946318080635211,
"learning_rate": 1.5907623908280112e-06,
"loss": 0.1353,
"step": 1093
},
{
"epoch": 4.42914979757085,
"grad_norm": 0.21052629194631087,
"learning_rate": 1.568732761074121e-06,
"loss": 0.154,
"step": 1094
},
{
"epoch": 4.433198380566802,
"grad_norm": 0.2332806850299172,
"learning_rate": 1.5468505072419237e-06,
"loss": 0.1618,
"step": 1095
},
{
"epoch": 4.437246963562753,
"grad_norm": 0.2349128761758036,
"learning_rate": 1.5251158043013448e-06,
"loss": 0.1377,
"step": 1096
},
{
"epoch": 4.441295546558704,
"grad_norm": 0.20295467868933417,
"learning_rate": 1.5035288260424885e-06,
"loss": 0.1573,
"step": 1097
},
{
"epoch": 4.445344129554655,
"grad_norm": 0.23645406603555275,
"learning_rate": 1.4820897450742623e-06,
"loss": 0.1619,
"step": 1098
},
{
"epoch": 4.449392712550607,
"grad_norm": 0.20474124140151304,
"learning_rate": 1.4607987328229767e-06,
"loss": 0.1785,
"step": 1099
},
{
"epoch": 4.4534412955465585,
"grad_norm": 0.22802907326248278,
"learning_rate": 1.4396559595310056e-06,
"loss": 0.1519,
"step": 1100
},
{
"epoch": 4.4574898785425106,
"grad_norm": 0.27411508474377916,
"learning_rate": 1.4186615942553973e-06,
"loss": 0.1844,
"step": 1101
},
{
"epoch": 4.461538461538462,
"grad_norm": 0.2259245261383601,
"learning_rate": 1.3978158048665424e-06,
"loss": 0.1661,
"step": 1102
},
{
"epoch": 4.465587044534413,
"grad_norm": 0.20548407166274904,
"learning_rate": 1.3771187580468115e-06,
"loss": 0.1367,
"step": 1103
},
{
"epoch": 4.469635627530364,
"grad_norm": 0.19372778686848557,
"learning_rate": 1.3565706192892392e-06,
"loss": 0.1727,
"step": 1104
},
{
"epoch": 4.473684210526316,
"grad_norm": 0.2133946634642161,
"learning_rate": 1.3361715528961993e-06,
"loss": 0.1486,
"step": 1105
},
{
"epoch": 4.477732793522267,
"grad_norm": 0.22380139922169895,
"learning_rate": 1.3159217219780817e-06,
"loss": 0.1539,
"step": 1106
},
{
"epoch": 4.481781376518219,
"grad_norm": 0.21406507348165438,
"learning_rate": 1.2958212884519949e-06,
"loss": 0.184,
"step": 1107
},
{
"epoch": 4.48582995951417,
"grad_norm": 0.32798668908681305,
"learning_rate": 1.2758704130404675e-06,
"loss": 0.1352,
"step": 1108
},
{
"epoch": 4.489878542510121,
"grad_norm": 0.19067760886435756,
"learning_rate": 1.2560692552701738e-06,
"loss": 0.1478,
"step": 1109
},
{
"epoch": 4.493927125506072,
"grad_norm": 0.21666187735278622,
"learning_rate": 1.2364179734706382e-06,
"loss": 0.1728,
"step": 1110
},
{
"epoch": 4.497975708502024,
"grad_norm": 0.19395068594991652,
"learning_rate": 1.2169167247729852e-06,
"loss": 0.1383,
"step": 1111
},
{
"epoch": 4.502024291497976,
"grad_norm": 0.23664424103239146,
"learning_rate": 1.1975656651086864e-06,
"loss": 0.1349,
"step": 1112
},
{
"epoch": 4.506072874493928,
"grad_norm": 0.21089814822115047,
"learning_rate": 1.1783649492082926e-06,
"loss": 0.154,
"step": 1113
},
{
"epoch": 4.510121457489879,
"grad_norm": 0.2598822262903524,
"learning_rate": 1.1593147306002183e-06,
"loss": 0.1653,
"step": 1114
},
{
"epoch": 4.51417004048583,
"grad_norm": 0.21789865205914039,
"learning_rate": 1.140415161609505e-06,
"loss": 0.1575,
"step": 1115
},
{
"epoch": 4.518218623481781,
"grad_norm": 0.22760141471912446,
"learning_rate": 1.121666393356604e-06,
"loss": 0.1862,
"step": 1116
},
{
"epoch": 4.522267206477733,
"grad_norm": 0.21551327374552762,
"learning_rate": 1.1030685757561632e-06,
"loss": 0.2173,
"step": 1117
},
{
"epoch": 4.526315789473684,
"grad_norm": 0.23739214944129486,
"learning_rate": 1.0846218575158373e-06,
"loss": 0.1493,
"step": 1118
},
{
"epoch": 4.530364372469636,
"grad_norm": 0.2537118387752405,
"learning_rate": 1.066326386135097e-06,
"loss": 0.1383,
"step": 1119
},
{
"epoch": 4.534412955465587,
"grad_norm": 0.2150850492039092,
"learning_rate": 1.0481823079040398e-06,
"loss": 0.1476,
"step": 1120
},
{
"epoch": 4.538461538461538,
"grad_norm": 0.22975770923849428,
"learning_rate": 1.0301897679022344e-06,
"loss": 0.1741,
"step": 1121
},
{
"epoch": 4.5425101214574894,
"grad_norm": 0.35509310091007024,
"learning_rate": 1.0123489099975491e-06,
"loss": 0.1525,
"step": 1122
},
{
"epoch": 4.5465587044534415,
"grad_norm": 0.21536529676795763,
"learning_rate": 9.9465987684501e-07,
"loss": 0.1667,
"step": 1123
},
{
"epoch": 4.550607287449393,
"grad_norm": 0.22670644193337858,
"learning_rate": 9.771228098856534e-07,
"loss": 0.1666,
"step": 1124
},
{
"epoch": 4.554655870445345,
"grad_norm": 0.23059885005131137,
"learning_rate": 9.597378493454013e-07,
"loss": 0.1537,
"step": 1125
},
{
"epoch": 4.558704453441296,
"grad_norm": 0.46814752638489293,
"learning_rate": 9.425051342339353e-07,
"loss": 0.1487,
"step": 1126
},
{
"epoch": 4.562753036437247,
"grad_norm": 0.22111344479751077,
"learning_rate": 9.254248023435864e-07,
"loss": 0.1671,
"step": 1127
},
{
"epoch": 4.566801619433198,
"grad_norm": 0.22861040245148814,
"learning_rate": 9.084969902482288e-07,
"loss": 0.1304,
"step": 1128
},
{
"epoch": 4.57085020242915,
"grad_norm": 0.24549303438942224,
"learning_rate": 8.917218333021993e-07,
"loss": 0.2272,
"step": 1129
},
{
"epoch": 4.574898785425101,
"grad_norm": 0.21554104774341226,
"learning_rate": 8.750994656392109e-07,
"loss": 0.2043,
"step": 1130
},
{
"epoch": 4.578947368421053,
"grad_norm": 0.21321210608264854,
"learning_rate": 8.58630020171265e-07,
"loss": 0.1562,
"step": 1131
},
{
"epoch": 4.582995951417004,
"grad_norm": 0.24134363410716286,
"learning_rate": 8.423136285876099e-07,
"loss": 0.1686,
"step": 1132
},
{
"epoch": 4.587044534412955,
"grad_norm": 0.20601058680502568,
"learning_rate": 8.26150421353682e-07,
"loss": 0.1571,
"step": 1133
},
{
"epoch": 4.5910931174089065,
"grad_norm": 0.21915395082271497,
"learning_rate": 8.101405277100549e-07,
"loss": 0.1624,
"step": 1134
},
{
"epoch": 4.5951417004048585,
"grad_norm": 0.21875372139428342,
"learning_rate": 7.942840756714077e-07,
"loss": 0.1431,
"step": 1135
},
{
"epoch": 4.59919028340081,
"grad_norm": 0.22877657326791467,
"learning_rate": 7.785811920255093e-07,
"loss": 0.1602,
"step": 1136
},
{
"epoch": 4.603238866396762,
"grad_norm": 0.25250287161813395,
"learning_rate": 7.630320023322024e-07,
"loss": 0.1997,
"step": 1137
},
{
"epoch": 4.607287449392713,
"grad_norm": 0.2249969188049647,
"learning_rate": 7.476366309223903e-07,
"loss": 0.1656,
"step": 1138
},
{
"epoch": 4.611336032388664,
"grad_norm": 0.2361573330597956,
"learning_rate": 7.323952008970537e-07,
"loss": 0.1699,
"step": 1139
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.1970435782028268,
"learning_rate": 7.173078341262618e-07,
"loss": 0.1693,
"step": 1140
},
{
"epoch": 4.619433198380567,
"grad_norm": 0.29492205760762835,
"learning_rate": 7.023746512481989e-07,
"loss": 0.1903,
"step": 1141
},
{
"epoch": 4.623481781376518,
"grad_norm": 0.3224185283125639,
"learning_rate": 6.875957716681902e-07,
"loss": 0.1644,
"step": 1142
},
{
"epoch": 4.62753036437247,
"grad_norm": 0.19362926476738668,
"learning_rate": 6.729713135577687e-07,
"loss": 0.1315,
"step": 1143
},
{
"epoch": 4.631578947368421,
"grad_norm": 0.20722937875515698,
"learning_rate": 6.585013938537143e-07,
"loss": 0.1925,
"step": 1144
},
{
"epoch": 4.635627530364372,
"grad_norm": 0.23643484178291838,
"learning_rate": 6.441861282571115e-07,
"loss": 0.1378,
"step": 1145
},
{
"epoch": 4.6396761133603235,
"grad_norm": 0.22742836356198745,
"learning_rate": 6.300256312324471e-07,
"loss": 0.2276,
"step": 1146
},
{
"epoch": 4.6437246963562755,
"grad_norm": 0.20708623458651013,
"learning_rate": 6.160200160066776e-07,
"loss": 0.1797,
"step": 1147
},
{
"epoch": 4.647773279352227,
"grad_norm": 0.2200544080314877,
"learning_rate": 6.021693945683327e-07,
"loss": 0.1811,
"step": 1148
},
{
"epoch": 4.651821862348179,
"grad_norm": 0.31297179635512745,
"learning_rate": 5.884738776666044e-07,
"loss": 0.1687,
"step": 1149
},
{
"epoch": 4.65587044534413,
"grad_norm": 0.20792275920918452,
"learning_rate": 5.749335748104878e-07,
"loss": 0.147,
"step": 1150
},
{
"epoch": 4.659919028340081,
"grad_norm": 0.2201833807428425,
"learning_rate": 5.615485942678778e-07,
"loss": 0.1528,
"step": 1151
},
{
"epoch": 4.663967611336032,
"grad_norm": 0.2590638066960072,
"learning_rate": 5.483190430647289e-07,
"loss": 0.137,
"step": 1152
},
{
"epoch": 4.668016194331984,
"grad_norm": 0.20854384536228351,
"learning_rate": 5.352450269841747e-07,
"loss": 0.1634,
"step": 1153
},
{
"epoch": 4.672064777327935,
"grad_norm": 0.19614365435400488,
"learning_rate": 5.22326650565701e-07,
"loss": 0.1702,
"step": 1154
},
{
"epoch": 4.676113360323887,
"grad_norm": 0.24164363575942743,
"learning_rate": 5.095640171043048e-07,
"loss": 0.1659,
"step": 1155
},
{
"epoch": 4.680161943319838,
"grad_norm": 0.19555176414716516,
"learning_rate": 4.969572286496594e-07,
"loss": 0.1768,
"step": 1156
},
{
"epoch": 4.684210526315789,
"grad_norm": 0.21331343384150162,
"learning_rate": 4.845063860053079e-07,
"loss": 0.1512,
"step": 1157
},
{
"epoch": 4.6882591093117405,
"grad_norm": 0.22902569966778585,
"learning_rate": 4.7221158872785997e-07,
"loss": 0.1644,
"step": 1158
},
{
"epoch": 4.6923076923076925,
"grad_norm": 0.21271692682769244,
"learning_rate": 4.600729351261857e-07,
"loss": 0.153,
"step": 1159
},
{
"epoch": 4.696356275303644,
"grad_norm": 0.20285824759567977,
"learning_rate": 4.4809052226063355e-07,
"loss": 0.1622,
"step": 1160
},
{
"epoch": 4.700404858299595,
"grad_norm": 0.21170111366596434,
"learning_rate": 4.3626444594226046e-07,
"loss": 0.1414,
"step": 1161
},
{
"epoch": 4.704453441295547,
"grad_norm": 0.21273494622282368,
"learning_rate": 4.245948007320588e-07,
"loss": 0.1605,
"step": 1162
},
{
"epoch": 4.708502024291498,
"grad_norm": 0.21580236712426215,
"learning_rate": 4.130816799401971e-07,
"loss": 0.1697,
"step": 1163
},
{
"epoch": 4.712550607287449,
"grad_norm": 0.21870312064468417,
"learning_rate": 4.017251756252827e-07,
"loss": 0.1486,
"step": 1164
},
{
"epoch": 4.716599190283401,
"grad_norm": 0.2086620351430627,
"learning_rate": 3.9052537859362473e-07,
"loss": 0.1792,
"step": 1165
},
{
"epoch": 4.720647773279352,
"grad_norm": 0.4567164021809264,
"learning_rate": 3.7948237839849687e-07,
"loss": 0.1637,
"step": 1166
},
{
"epoch": 4.724696356275303,
"grad_norm": 0.2054042759870951,
"learning_rate": 3.6859626333943797e-07,
"loss": 0.1547,
"step": 1167
},
{
"epoch": 4.728744939271255,
"grad_norm": 0.22787366706709217,
"learning_rate": 3.578671204615347e-07,
"loss": 0.1603,
"step": 1168
},
{
"epoch": 4.732793522267206,
"grad_norm": 0.21723775852665975,
"learning_rate": 3.472950355547267e-07,
"loss": 0.1625,
"step": 1169
},
{
"epoch": 4.7368421052631575,
"grad_norm": 0.21284171162336726,
"learning_rate": 3.3688009315312244e-07,
"loss": 0.1617,
"step": 1170
},
{
"epoch": 4.7408906882591095,
"grad_norm": 0.24326548251914945,
"learning_rate": 3.266223765343224e-07,
"loss": 0.1459,
"step": 1171
},
{
"epoch": 4.744939271255061,
"grad_norm": 0.21207539894851096,
"learning_rate": 3.16521967718757e-07,
"loss": 0.1658,
"step": 1172
},
{
"epoch": 4.748987854251012,
"grad_norm": 0.21318810407360125,
"learning_rate": 3.0657894746902726e-07,
"loss": 0.1396,
"step": 1173
},
{
"epoch": 4.753036437246964,
"grad_norm": 0.20817448396242322,
"learning_rate": 2.9679339528925654e-07,
"loss": 0.1657,
"step": 1174
},
{
"epoch": 4.757085020242915,
"grad_norm": 0.21270363958177355,
"learning_rate": 2.871653894244597e-07,
"loss": 0.1608,
"step": 1175
},
{
"epoch": 4.761133603238866,
"grad_norm": 0.22777002845499997,
"learning_rate": 2.7769500685991266e-07,
"loss": 0.1481,
"step": 1176
},
{
"epoch": 4.765182186234818,
"grad_norm": 0.23029921558396857,
"learning_rate": 2.683823233205485e-07,
"loss": 0.1441,
"step": 1177
},
{
"epoch": 4.769230769230769,
"grad_norm": 0.29549946679303096,
"learning_rate": 2.592274132703287e-07,
"loss": 0.1437,
"step": 1178
},
{
"epoch": 4.77327935222672,
"grad_norm": 0.22559412304611295,
"learning_rate": 2.502303499116754e-07,
"loss": 0.1561,
"step": 1179
},
{
"epoch": 4.777327935222672,
"grad_norm": 0.19586512694682767,
"learning_rate": 2.413912051848666e-07,
"loss": 0.1628,
"step": 1180
},
{
"epoch": 4.781376518218623,
"grad_norm": 0.21129488110932507,
"learning_rate": 2.3271004976746613e-07,
"loss": 0.1541,
"step": 1181
},
{
"epoch": 4.7854251012145745,
"grad_norm": 0.22211289100549902,
"learning_rate": 2.2418695307376393e-07,
"loss": 0.1829,
"step": 1182
},
{
"epoch": 4.7894736842105265,
"grad_norm": 0.21331460474172007,
"learning_rate": 2.1582198325420968e-07,
"loss": 0.1582,
"step": 1183
},
{
"epoch": 4.793522267206478,
"grad_norm": 0.2635297510599588,
"learning_rate": 2.0761520719488447e-07,
"loss": 0.1633,
"step": 1184
},
{
"epoch": 4.797570850202429,
"grad_norm": 0.23317341525953963,
"learning_rate": 1.9956669051694798e-07,
"loss": 0.165,
"step": 1185
},
{
"epoch": 4.801619433198381,
"grad_norm": 0.20725802802308055,
"learning_rate": 1.916764975761254e-07,
"loss": 0.1411,
"step": 1186
},
{
"epoch": 4.805668016194332,
"grad_norm": 0.2096446173791536,
"learning_rate": 1.839446914621923e-07,
"loss": 0.1316,
"step": 1187
},
{
"epoch": 4.809716599190283,
"grad_norm": 0.22639226514703667,
"learning_rate": 1.7637133399845962e-07,
"loss": 0.1473,
"step": 1188
},
{
"epoch": 4.813765182186235,
"grad_norm": 0.27539747986913277,
"learning_rate": 1.6895648574129397e-07,
"loss": 0.1914,
"step": 1189
},
{
"epoch": 4.817813765182186,
"grad_norm": 0.2681109574931658,
"learning_rate": 1.6170020597962465e-07,
"loss": 0.1548,
"step": 1190
},
{
"epoch": 4.821862348178137,
"grad_norm": 0.27218213871115765,
"learning_rate": 1.5460255273447078e-07,
"loss": 0.2009,
"step": 1191
},
{
"epoch": 4.825910931174089,
"grad_norm": 0.19721315680412155,
"learning_rate": 1.4766358275847715e-07,
"loss": 0.1371,
"step": 1192
},
{
"epoch": 4.82995951417004,
"grad_norm": 0.21411480603149827,
"learning_rate": 1.4088335153546573e-07,
"loss": 0.149,
"step": 1193
},
{
"epoch": 4.834008097165992,
"grad_norm": 0.20943034870718674,
"learning_rate": 1.3426191327998496e-07,
"loss": 0.1946,
"step": 1194
},
{
"epoch": 4.838056680161944,
"grad_norm": 0.28001239308224035,
"learning_rate": 1.277993209368744e-07,
"loss": 0.1608,
"step": 1195
},
{
"epoch": 4.842105263157895,
"grad_norm": 0.20586202037364087,
"learning_rate": 1.214956261808542e-07,
"loss": 0.1791,
"step": 1196
},
{
"epoch": 4.846153846153846,
"grad_norm": 0.21344831105649956,
"learning_rate": 1.1535087941609846e-07,
"loss": 0.1503,
"step": 1197
},
{
"epoch": 4.850202429149798,
"grad_norm": 0.2069072014754558,
"learning_rate": 1.0936512977583802e-07,
"loss": 0.1767,
"step": 1198
},
{
"epoch": 4.854251012145749,
"grad_norm": 0.21898155696813743,
"learning_rate": 1.0353842512196732e-07,
"loss": 0.219,
"step": 1199
},
{
"epoch": 4.8582995951417,
"grad_norm": 0.4010526789488682,
"learning_rate": 9.787081204466253e-08,
"loss": 0.1982,
"step": 1200
},
{
"epoch": 4.862348178137652,
"grad_norm": 0.2847964930281046,
"learning_rate": 9.236233586200627e-08,
"loss": 0.1339,
"step": 1201
},
{
"epoch": 4.866396761133603,
"grad_norm": 0.21908206669982577,
"learning_rate": 8.70130406196279e-08,
"loss": 0.1966,
"step": 1202
},
{
"epoch": 4.870445344129554,
"grad_norm": 0.21007269127098252,
"learning_rate": 8.182296909035047e-08,
"loss": 0.1668,
"step": 1203
},
{
"epoch": 4.874493927125506,
"grad_norm": 0.24270813403182873,
"learning_rate": 7.679216277384439e-08,
"loss": 0.177,
"step": 1204
},
{
"epoch": 4.8785425101214575,
"grad_norm": 0.25096049029974016,
"learning_rate": 7.19206618963053e-08,
"loss": 0.1643,
"step": 1205
},
{
"epoch": 4.882591093117409,
"grad_norm": 0.22528468394925152,
"learning_rate": 6.720850541012347e-08,
"loss": 0.1449,
"step": 1206
},
{
"epoch": 4.886639676113361,
"grad_norm": 0.24843986475553187,
"learning_rate": 6.265573099357714e-08,
"loss": 0.1531,
"step": 1207
},
{
"epoch": 4.890688259109312,
"grad_norm": 0.20275803035443452,
"learning_rate": 5.826237505053067e-08,
"loss": 0.1513,
"step": 1208
},
{
"epoch": 4.894736842105263,
"grad_norm": 0.21277623162434223,
"learning_rate": 5.4028472710137e-08,
"loss": 0.1514,
"step": 1209
},
{
"epoch": 4.898785425101215,
"grad_norm": 0.2221050417915764,
"learning_rate": 4.9954057826571146e-08,
"loss": 0.1547,
"step": 1210
},
{
"epoch": 4.902834008097166,
"grad_norm": 0.24476576247063722,
"learning_rate": 4.6039162978743776e-08,
"loss": 0.1727,
"step": 1211
},
{
"epoch": 4.906882591093117,
"grad_norm": 0.21254208813835604,
"learning_rate": 4.228381947004812e-08,
"loss": 0.148,
"step": 1212
},
{
"epoch": 4.910931174089069,
"grad_norm": 0.20613271325739702,
"learning_rate": 3.868805732811343e-08,
"loss": 0.1576,
"step": 1213
},
{
"epoch": 4.91497975708502,
"grad_norm": 0.21503475844799994,
"learning_rate": 3.525190530455636e-08,
"loss": 0.1712,
"step": 1214
},
{
"epoch": 4.919028340080971,
"grad_norm": 0.2317600317657127,
"learning_rate": 3.197539087475887e-08,
"loss": 0.1648,
"step": 1215
},
{
"epoch": 4.923076923076923,
"grad_norm": 0.22356149051251492,
"learning_rate": 2.8858540237641786e-08,
"loss": 0.1699,
"step": 1216
},
{
"epoch": 4.9271255060728745,
"grad_norm": 0.2114657803083758,
"learning_rate": 2.5901378315458247e-08,
"loss": 0.1635,
"step": 1217
},
{
"epoch": 4.931174089068826,
"grad_norm": 0.29547060937874886,
"learning_rate": 2.3103928753598347e-08,
"loss": 0.163,
"step": 1218
},
{
"epoch": 4.935222672064778,
"grad_norm": 0.21502402670338336,
"learning_rate": 2.0466213920393718e-08,
"loss": 0.1943,
"step": 1219
},
{
"epoch": 4.939271255060729,
"grad_norm": 0.212501867966559,
"learning_rate": 1.7988254906942115e-08,
"loss": 0.1488,
"step": 1220
},
{
"epoch": 4.94331983805668,
"grad_norm": 0.24049856849181864,
"learning_rate": 1.5670071526936447e-08,
"loss": 0.1862,
"step": 1221
},
{
"epoch": 4.947368421052632,
"grad_norm": 0.19466176877316402,
"learning_rate": 1.3511682316509344e-08,
"loss": 0.1515,
"step": 1222
},
{
"epoch": 4.951417004048583,
"grad_norm": 0.20857241157075918,
"learning_rate": 1.1513104534086605e-08,
"loss": 0.1563,
"step": 1223
},
{
"epoch": 4.955465587044534,
"grad_norm": 0.20845637539801218,
"learning_rate": 9.67435416023843e-09,
"loss": 0.1305,
"step": 1224
},
{
"epoch": 4.959514170040486,
"grad_norm": 0.19368852381120746,
"learning_rate": 7.995445897566178e-09,
"loss": 0.1539,
"step": 1225
},
{
"epoch": 4.963562753036437,
"grad_norm": 0.2150471079271013,
"learning_rate": 6.476393170578021e-09,
"loss": 0.1612,
"step": 1226
},
{
"epoch": 4.967611336032388,
"grad_norm": 0.22828381255965383,
"learning_rate": 5.11720812557348e-09,
"loss": 0.1466,
"step": 1227
},
{
"epoch": 4.97165991902834,
"grad_norm": 0.2517354584251829,
"learning_rate": 3.917901630563492e-09,
"loss": 0.1411,
"step": 1228
},
{
"epoch": 4.9757085020242915,
"grad_norm": 0.20465023598823315,
"learning_rate": 2.878483275170485e-09,
"loss": 0.1496,
"step": 1229
},
{
"epoch": 4.979757085020243,
"grad_norm": 0.23945909047116662,
"learning_rate": 1.998961370555108e-09,
"loss": 0.1594,
"step": 1230
},
{
"epoch": 4.983805668016195,
"grad_norm": 0.21836736199283446,
"learning_rate": 1.2793429493518361e-09,
"loss": 0.1604,
"step": 1231
},
{
"epoch": 4.987854251012146,
"grad_norm": 0.2834052660566203,
"learning_rate": 7.196337656112384e-10,
"loss": 0.1682,
"step": 1232
},
{
"epoch": 4.991902834008097,
"grad_norm": 0.23391080203916695,
"learning_rate": 3.198382947511292e-10,
"loss": 0.1428,
"step": 1233
},
{
"epoch": 4.995951417004049,
"grad_norm": 0.22275147545931687,
"learning_rate": 7.995973352770137e-11,
"loss": 0.1624,
"step": 1234
},
{
"epoch": 5.0,
"grad_norm": 0.19572743950755267,
"learning_rate": 0.0,
"loss": 0.1551,
"step": 1235
},
{
"epoch": 5.0,
"step": 1235,
"total_flos": 1.4236951748269834e+18,
"train_loss": 0.3489899976835077,
"train_runtime": 22405.389,
"train_samples_per_second": 7.052,
"train_steps_per_second": 0.055
}
],
"logging_steps": 1,
"max_steps": 1235,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.4236951748269834e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}