adaption_chartsense_sft / trainer_state.json
Avinaash's picture
Add 11 files
775c308 verified
Raw
History Blame Contribute Delete
50.6 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.0,
"eval_steps": 54,
"global_step": 272,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.014705882352941176,
"grad_norm": 0.7467471361160278,
"learning_rate": 0.0,
"loss": 1.44091796875,
"step": 1
},
{
"epoch": 0.029411764705882353,
"grad_norm": 1.1072983741760254,
"learning_rate": 1.1111111111111112e-05,
"loss": 1.223876953125,
"step": 2
},
{
"epoch": 0.04411764705882353,
"grad_norm": 0.744576096534729,
"learning_rate": 2.2222222222222223e-05,
"loss": 1.4482421875,
"step": 3
},
{
"epoch": 0.058823529411764705,
"grad_norm": 0.7001456618309021,
"learning_rate": 3.3333333333333335e-05,
"loss": 1.45068359375,
"step": 4
},
{
"epoch": 0.07352941176470588,
"grad_norm": 0.5546188354492188,
"learning_rate": 4.4444444444444447e-05,
"loss": 1.118896484375,
"step": 5
},
{
"epoch": 0.08823529411764706,
"grad_norm": 0.4852159023284912,
"learning_rate": 5.555555555555556e-05,
"loss": 1.24560546875,
"step": 6
},
{
"epoch": 0.10294117647058823,
"grad_norm": 0.27731776237487793,
"learning_rate": 6.666666666666667e-05,
"loss": 1.15478515625,
"step": 7
},
{
"epoch": 0.11764705882352941,
"grad_norm": 0.22583669424057007,
"learning_rate": 7.777777777777778e-05,
"loss": 1.15869140625,
"step": 8
},
{
"epoch": 0.1323529411764706,
"grad_norm": 0.3496478796005249,
"learning_rate": 8.888888888888889e-05,
"loss": 1.19384765625,
"step": 9
},
{
"epoch": 0.14705882352941177,
"grad_norm": 0.4567904472351074,
"learning_rate": 0.0001,
"loss": 1.18798828125,
"step": 10
},
{
"epoch": 0.16176470588235295,
"grad_norm": 0.3518531322479248,
"learning_rate": 9.999678955227749e-05,
"loss": 0.973876953125,
"step": 11
},
{
"epoch": 0.17647058823529413,
"grad_norm": 0.3234044313430786,
"learning_rate": 9.998715866719772e-05,
"loss": 1.203125,
"step": 12
},
{
"epoch": 0.19117647058823528,
"grad_norm": 0.2109292596578598,
"learning_rate": 9.997110871895861e-05,
"loss": 1.1884765625,
"step": 13
},
{
"epoch": 0.20588235294117646,
"grad_norm": 0.18434831500053406,
"learning_rate": 9.994864199767213e-05,
"loss": 1.228515625,
"step": 14
},
{
"epoch": 0.22058823529411764,
"grad_norm": 0.17678029835224152,
"learning_rate": 9.99197617090376e-05,
"loss": 1.0146484375,
"step": 15
},
{
"epoch": 0.23529411764705882,
"grad_norm": 0.16223645210266113,
"learning_rate": 9.988447197388422e-05,
"loss": 1.1533203125,
"step": 16
},
{
"epoch": 0.25,
"grad_norm": 2.8353583812713623,
"learning_rate": 9.984277782758306e-05,
"loss": 1.1015625,
"step": 17
},
{
"epoch": 0.2647058823529412,
"grad_norm": 0.1727093905210495,
"learning_rate": 9.979468521932868e-05,
"loss": 1.1015625,
"step": 18
},
{
"epoch": 0.27941176470588236,
"grad_norm": 0.15514397621154785,
"learning_rate": 9.974020101129016e-05,
"loss": 1.1240234375,
"step": 19
},
{
"epoch": 0.29411764705882354,
"grad_norm": 0.15237128734588623,
"learning_rate": 9.967933297763205e-05,
"loss": 1.10302734375,
"step": 20
},
{
"epoch": 0.3088235294117647,
"grad_norm": 0.12800024449825287,
"learning_rate": 9.961208980340498e-05,
"loss": 1.2314453125,
"step": 21
},
{
"epoch": 0.3235294117647059,
"grad_norm": 0.09313075989484787,
"learning_rate": 9.953848108330654e-05,
"loss": 1.06005859375,
"step": 22
},
{
"epoch": 0.3382352941176471,
"grad_norm": 0.11129424721002579,
"learning_rate": 9.945851732031221e-05,
"loss": 1.02490234375,
"step": 23
},
{
"epoch": 0.35294117647058826,
"grad_norm": 0.10968948900699615,
"learning_rate": 9.937220992417666e-05,
"loss": 1.111083984375,
"step": 24
},
{
"epoch": 0.36764705882352944,
"grad_norm": 0.11814726144075394,
"learning_rate": 9.927957120980584e-05,
"loss": 1.05908203125,
"step": 25
},
{
"epoch": 0.38235294117647056,
"grad_norm": 0.11258970201015472,
"learning_rate": 9.918061439549971e-05,
"loss": 1.16650390625,
"step": 26
},
{
"epoch": 0.39705882352941174,
"grad_norm": 0.1053420752286911,
"learning_rate": 9.907535360106624e-05,
"loss": 1.10693359375,
"step": 27
},
{
"epoch": 0.4117647058823529,
"grad_norm": 0.09652009606361389,
"learning_rate": 9.896380384580667e-05,
"loss": 0.9814453125,
"step": 28
},
{
"epoch": 0.4264705882352941,
"grad_norm": 0.08952522277832031,
"learning_rate": 9.884598104637246e-05,
"loss": 0.90625,
"step": 29
},
{
"epoch": 0.4411764705882353,
"grad_norm": 0.0945228561758995,
"learning_rate": 9.872190201449416e-05,
"loss": 1.114501953125,
"step": 30
},
{
"epoch": 0.45588235294117646,
"grad_norm": 0.09587328881025314,
"learning_rate": 9.859158445458272e-05,
"loss": 1.0986328125,
"step": 31
},
{
"epoch": 0.47058823529411764,
"grad_norm": 0.0860571637749672,
"learning_rate": 9.845504696120315e-05,
"loss": 1.15380859375,
"step": 32
},
{
"epoch": 0.4852941176470588,
"grad_norm": 0.08797360956668854,
"learning_rate": 9.831230901642144e-05,
"loss": 1.01141357421875,
"step": 33
},
{
"epoch": 0.5,
"grad_norm": 0.08988713473081589,
"learning_rate": 9.816339098702467e-05,
"loss": 1.102783203125,
"step": 34
},
{
"epoch": 0.5147058823529411,
"grad_norm": 0.0863901823759079,
"learning_rate": 9.800831412161502e-05,
"loss": 1.111572265625,
"step": 35
},
{
"epoch": 0.5294117647058824,
"grad_norm": 0.10267789661884308,
"learning_rate": 9.784710054757779e-05,
"loss": 1.02783203125,
"step": 36
},
{
"epoch": 0.5441176470588235,
"grad_norm": 0.08884095400571823,
"learning_rate": 9.767977326792415e-05,
"loss": 0.891845703125,
"step": 37
},
{
"epoch": 0.5588235294117647,
"grad_norm": 0.08703789114952087,
"learning_rate": 9.750635615800895e-05,
"loss": 1.1357421875,
"step": 38
},
{
"epoch": 0.5735294117647058,
"grad_norm": 0.07673384249210358,
"learning_rate": 9.7326873962124e-05,
"loss": 1.16943359375,
"step": 39
},
{
"epoch": 0.5882352941176471,
"grad_norm": 0.08323892205953598,
"learning_rate": 9.714135228996734e-05,
"loss": 1.08056640625,
"step": 40
},
{
"epoch": 0.6029411764705882,
"grad_norm": 0.10157938301563263,
"learning_rate": 9.694981761298924e-05,
"loss": 1.0703125,
"step": 41
},
{
"epoch": 0.6176470588235294,
"grad_norm": 0.08656392991542816,
"learning_rate": 9.675229726061491e-05,
"loss": 1.16162109375,
"step": 42
},
{
"epoch": 0.6323529411764706,
"grad_norm": 0.0794377475976944,
"learning_rate": 9.654881941634502e-05,
"loss": 1.04736328125,
"step": 43
},
{
"epoch": 0.6470588235294118,
"grad_norm": 0.0774090439081192,
"learning_rate": 9.633941311373433e-05,
"loss": 0.98876953125,
"step": 44
},
{
"epoch": 0.6617647058823529,
"grad_norm": 0.0841091051697731,
"learning_rate": 9.612410823224893e-05,
"loss": 1.05810546875,
"step": 45
},
{
"epoch": 0.6764705882352942,
"grad_norm": 0.07540322840213776,
"learning_rate": 9.59029354930029e-05,
"loss": 1.07861328125,
"step": 46
},
{
"epoch": 0.6911764705882353,
"grad_norm": 0.07488111406564713,
"learning_rate": 9.567592645437475e-05,
"loss": 0.994140625,
"step": 47
},
{
"epoch": 0.7058823529411765,
"grad_norm": 0.07723530381917953,
"learning_rate": 9.544311350750454e-05,
"loss": 0.98095703125,
"step": 48
},
{
"epoch": 0.7205882352941176,
"grad_norm": 0.08420762419700623,
"learning_rate": 9.520452987167204e-05,
"loss": 1.0869140625,
"step": 49
},
{
"epoch": 0.7352941176470589,
"grad_norm": 0.09217358380556107,
"learning_rate": 9.496020958955682e-05,
"loss": 1.117431640625,
"step": 50
},
{
"epoch": 0.75,
"grad_norm": 0.07303526997566223,
"learning_rate": 9.471018752238085e-05,
"loss": 0.791015625,
"step": 51
},
{
"epoch": 0.7647058823529411,
"grad_norm": 0.08568691462278366,
"learning_rate": 9.44544993449341e-05,
"loss": 0.94921875,
"step": 52
},
{
"epoch": 0.7794117647058824,
"grad_norm": 0.08517300337553024,
"learning_rate": 9.419318154048448e-05,
"loss": 1.0791015625,
"step": 53
},
{
"epoch": 0.7941176470588235,
"grad_norm": 0.08076217770576477,
"learning_rate": 9.3926271395572e-05,
"loss": 1.13037109375,
"step": 54
},
{
"epoch": 0.8088235294117647,
"grad_norm": 0.10536608099937439,
"learning_rate": 9.365380699468851e-05,
"loss": 1.0185546875,
"step": 55
},
{
"epoch": 0.8235294117647058,
"grad_norm": 0.08304548263549805,
"learning_rate": 9.337582721484357e-05,
"loss": 1.013916015625,
"step": 56
},
{
"epoch": 0.8235294117647058,
"eval_loss": 0.990234375,
"eval_runtime": 5.5244,
"eval_samples_per_second": 0.724,
"eval_steps_per_second": 0.181,
"step": 56
},
{
"epoch": 0.8382352941176471,
"grad_norm": 0.08484519273042679,
"learning_rate": 9.309237172001724e-05,
"loss": 0.95849609375,
"step": 57
},
{
"epoch": 0.8529411764705882,
"grad_norm": 0.08013865351676941,
"learning_rate": 9.280348095550055e-05,
"loss": 0.970947265625,
"step": 58
},
{
"epoch": 0.8676470588235294,
"grad_norm": 0.08895693719387054,
"learning_rate": 9.250919614212444e-05,
"loss": 0.9736328125,
"step": 59
},
{
"epoch": 0.8823529411764706,
"grad_norm": 0.08965151011943817,
"learning_rate": 9.220955927037823e-05,
"loss": 1.06201171875,
"step": 60
},
{
"epoch": 0.8970588235294118,
"grad_norm": 0.09408161789178848,
"learning_rate": 9.1904613094418e-05,
"loss": 0.890869140625,
"step": 61
},
{
"epoch": 0.9117647058823529,
"grad_norm": 0.08738987892866135,
"learning_rate": 9.159440112596626e-05,
"loss": 0.969970703125,
"step": 62
},
{
"epoch": 0.9264705882352942,
"grad_norm": 0.08372481167316437,
"learning_rate": 9.127896762810332e-05,
"loss": 1.00732421875,
"step": 63
},
{
"epoch": 0.9411764705882353,
"grad_norm": 0.08490575104951859,
"learning_rate": 9.095835760895167e-05,
"loss": 0.8583984375,
"step": 64
},
{
"epoch": 0.9558823529411765,
"grad_norm": 0.08689001947641373,
"learning_rate": 9.063261681525376e-05,
"loss": 0.917724609375,
"step": 65
},
{
"epoch": 0.9705882352941176,
"grad_norm": 0.086466945707798,
"learning_rate": 9.030179172584465e-05,
"loss": 0.830322265625,
"step": 66
},
{
"epoch": 0.9852941176470589,
"grad_norm": 0.08034230023622513,
"learning_rate": 8.996592954502015e-05,
"loss": 1.06298828125,
"step": 67
},
{
"epoch": 1.0,
"grad_norm": 0.07923819869756699,
"learning_rate": 8.962507819580127e-05,
"loss": 1.10595703125,
"step": 68
},
{
"epoch": 1.0147058823529411,
"grad_norm": 0.08265892416238785,
"learning_rate": 8.927928631309638e-05,
"loss": 1.02001953125,
"step": 69
},
{
"epoch": 1.0294117647058822,
"grad_norm": 0.07541780173778534,
"learning_rate": 8.892860323676158e-05,
"loss": 0.82562255859375,
"step": 70
},
{
"epoch": 1.0441176470588236,
"grad_norm": 0.09345775842666626,
"learning_rate": 8.857307900456055e-05,
"loss": 1.025634765625,
"step": 71
},
{
"epoch": 1.0588235294117647,
"grad_norm": 0.0891224667429924,
"learning_rate": 8.821276434502499e-05,
"loss": 1.06494140625,
"step": 72
},
{
"epoch": 1.0735294117647058,
"grad_norm": 0.08286001533269882,
"learning_rate": 8.784771067021609e-05,
"loss": 0.81640625,
"step": 73
},
{
"epoch": 1.088235294117647,
"grad_norm": 0.08618306368589401,
"learning_rate": 8.747797006838894e-05,
"loss": 0.93310546875,
"step": 74
},
{
"epoch": 1.1029411764705883,
"grad_norm": 0.0845174565911293,
"learning_rate": 8.710359529656009e-05,
"loss": 0.9404296875,
"step": 75
},
{
"epoch": 1.1176470588235294,
"grad_norm": 0.09461177885532379,
"learning_rate": 8.672463977297996e-05,
"loss": 0.94873046875,
"step": 76
},
{
"epoch": 1.1323529411764706,
"grad_norm": 0.08769670128822327,
"learning_rate": 8.634115756951063e-05,
"loss": 1.02392578125,
"step": 77
},
{
"epoch": 1.1470588235294117,
"grad_norm": 0.09109525382518768,
"learning_rate": 8.595320340391067e-05,
"loss": 0.990966796875,
"step": 78
},
{
"epoch": 1.161764705882353,
"grad_norm": 0.08751225471496582,
"learning_rate": 8.556083263202745e-05,
"loss": 0.810791015625,
"step": 79
},
{
"epoch": 1.1764705882352942,
"grad_norm": 0.0901452898979187,
"learning_rate": 8.516410123989887e-05,
"loss": 1.02490234375,
"step": 80
},
{
"epoch": 1.1911764705882353,
"grad_norm": 0.08694286644458771,
"learning_rate": 8.476306583576462e-05,
"loss": 1.009765625,
"step": 81
},
{
"epoch": 1.2058823529411764,
"grad_norm": 0.09788144379854202,
"learning_rate": 8.435778364198915e-05,
"loss": 1.060546875,
"step": 82
},
{
"epoch": 1.2205882352941178,
"grad_norm": 0.08373791724443436,
"learning_rate": 8.394831248689674e-05,
"loss": 0.87158203125,
"step": 83
},
{
"epoch": 1.2352941176470589,
"grad_norm": 0.09068770706653595,
"learning_rate": 8.353471079652014e-05,
"loss": 1.00732421875,
"step": 84
},
{
"epoch": 1.25,
"grad_norm": 0.09484093636274338,
"learning_rate": 8.311703758626404e-05,
"loss": 0.96533203125,
"step": 85
},
{
"epoch": 1.2647058823529411,
"grad_norm": 0.0946110263466835,
"learning_rate": 8.269535245248426e-05,
"loss": 0.966064453125,
"step": 86
},
{
"epoch": 1.2794117647058822,
"grad_norm": 0.08420296758413315,
"learning_rate": 8.226971556398425e-05,
"loss": 0.986572265625,
"step": 87
},
{
"epoch": 1.2941176470588236,
"grad_norm": 0.08452770859003067,
"learning_rate": 8.184018765342976e-05,
"loss": 0.97021484375,
"step": 88
},
{
"epoch": 1.3088235294117647,
"grad_norm": 0.09456127136945724,
"learning_rate": 8.14068300086831e-05,
"loss": 1.10791015625,
"step": 89
},
{
"epoch": 1.3235294117647058,
"grad_norm": 0.12618407607078552,
"learning_rate": 8.096970446405822e-05,
"loss": 0.943359375,
"step": 90
},
{
"epoch": 1.3382352941176472,
"grad_norm": 0.07619062066078186,
"learning_rate": 8.052887339149774e-05,
"loss": 0.9208984375,
"step": 91
},
{
"epoch": 1.3529411764705883,
"grad_norm": 0.10043489187955856,
"learning_rate": 8.008439969167335e-05,
"loss": 0.99658203125,
"step": 92
},
{
"epoch": 1.3676470588235294,
"grad_norm": 0.10213746130466461,
"learning_rate": 7.963634678501072e-05,
"loss": 0.95263671875,
"step": 93
},
{
"epoch": 1.3823529411764706,
"grad_norm": 0.10206730663776398,
"learning_rate": 7.918477860264023e-05,
"loss": 1.05126953125,
"step": 94
},
{
"epoch": 1.3970588235294117,
"grad_norm": 0.09070366621017456,
"learning_rate": 7.872975957727487e-05,
"loss": 0.995361328125,
"step": 95
},
{
"epoch": 1.4117647058823528,
"grad_norm": 0.08802887052297592,
"learning_rate": 7.82713546340166e-05,
"loss": 0.875,
"step": 96
},
{
"epoch": 1.4264705882352942,
"grad_norm": 0.086219422519207,
"learning_rate": 7.780962918109236e-05,
"loss": 0.813720703125,
"step": 97
},
{
"epoch": 1.4411764705882353,
"grad_norm": 0.0911615714430809,
"learning_rate": 7.73446491005212e-05,
"loss": 1.0087890625,
"step": 98
},
{
"epoch": 1.4558823529411764,
"grad_norm": 0.09299807995557785,
"learning_rate": 7.68764807387138e-05,
"loss": 0.98583984375,
"step": 99
},
{
"epoch": 1.4705882352941178,
"grad_norm": 0.08247499167919159,
"learning_rate": 7.640519089700579e-05,
"loss": 1.05078125,
"step": 100
},
{
"epoch": 1.4852941176470589,
"grad_norm": 0.07894891500473022,
"learning_rate": 7.593084682212598e-05,
"loss": 0.9261474609375,
"step": 101
},
{
"epoch": 1.5,
"grad_norm": 0.09833148121833801,
"learning_rate": 7.545351619660128e-05,
"loss": 1.007568359375,
"step": 102
},
{
"epoch": 1.5147058823529411,
"grad_norm": 0.08686123788356781,
"learning_rate": 7.49732671290992e-05,
"loss": 1.013916015625,
"step": 103
},
{
"epoch": 1.5294117647058822,
"grad_norm": 0.09453439712524414,
"learning_rate": 7.449016814470976e-05,
"loss": 0.93017578125,
"step": 104
},
{
"epoch": 1.5441176470588234,
"grad_norm": 0.09585236757993698,
"learning_rate": 7.400428817516783e-05,
"loss": 0.807373046875,
"step": 105
},
{
"epoch": 1.5588235294117647,
"grad_norm": 0.09875720739364624,
"learning_rate": 7.351569654901744e-05,
"loss": 1.0361328125,
"step": 106
},
{
"epoch": 1.5735294117647058,
"grad_norm": 0.09748269617557526,
"learning_rate": 7.302446298171966e-05,
"loss": 1.0791015625,
"step": 107
},
{
"epoch": 1.5882352941176472,
"grad_norm": 0.10695629566907883,
"learning_rate": 7.253065756570502e-05,
"loss": 0.98779296875,
"step": 108
},
{
"epoch": 1.6029411764705883,
"grad_norm": 0.09422429651021957,
"learning_rate": 7.203435076037236e-05,
"loss": 0.9853515625,
"step": 109
},
{
"epoch": 1.6176470588235294,
"grad_norm": 0.10585072636604309,
"learning_rate": 7.15356133820351e-05,
"loss": 1.07958984375,
"step": 110
},
{
"epoch": 1.6176470588235294,
"eval_loss": 0.9423828125,
"eval_runtime": 5.5116,
"eval_samples_per_second": 0.726,
"eval_steps_per_second": 0.181,
"step": 110
},
{
"epoch": 1.6323529411764706,
"grad_norm": 0.14425884187221527,
"learning_rate": 7.103451659381685e-05,
"loss": 0.95947265625,
"step": 111
},
{
"epoch": 1.6470588235294117,
"grad_norm": 0.08809656649827957,
"learning_rate": 7.053113189549724e-05,
"loss": 0.9169921875,
"step": 112
},
{
"epoch": 1.6617647058823528,
"grad_norm": 0.1172780841588974,
"learning_rate": 7.002553111331001e-05,
"loss": 0.97265625,
"step": 113
},
{
"epoch": 1.6764705882352942,
"grad_norm": 0.10190092772245407,
"learning_rate": 6.951778638969421e-05,
"loss": 0.997802734375,
"step": 114
},
{
"epoch": 1.6911764705882353,
"grad_norm": 0.09937534481287003,
"learning_rate": 6.900797017300062e-05,
"loss": 0.91943359375,
"step": 115
},
{
"epoch": 1.7058823529411766,
"grad_norm": 0.10354841500520706,
"learning_rate": 6.84961552071542e-05,
"loss": 0.90087890625,
"step": 116
},
{
"epoch": 1.7205882352941178,
"grad_norm": 0.11600671708583832,
"learning_rate": 6.798241452127454e-05,
"loss": 1.00439453125,
"step": 117
},
{
"epoch": 1.7352941176470589,
"grad_norm": 0.09904322773218155,
"learning_rate": 6.746682141925566e-05,
"loss": 1.043701171875,
"step": 118
},
{
"epoch": 1.75,
"grad_norm": 0.08371817320585251,
"learning_rate": 6.694944946930647e-05,
"loss": 0.7327880859375,
"step": 119
},
{
"epoch": 1.7647058823529411,
"grad_norm": 0.0922434851527214,
"learning_rate": 6.643037249345353e-05,
"loss": 0.88134765625,
"step": 120
},
{
"epoch": 1.7794117647058822,
"grad_norm": 0.0983903557062149,
"learning_rate": 6.590966455700775e-05,
"loss": 1.00146484375,
"step": 121
},
{
"epoch": 1.7941176470588234,
"grad_norm": 0.09228742867708206,
"learning_rate": 6.53873999579962e-05,
"loss": 1.0556640625,
"step": 122
},
{
"epoch": 1.8088235294117647,
"grad_norm": 0.09647046029567719,
"learning_rate": 6.486365321656071e-05,
"loss": 0.94482421875,
"step": 123
},
{
"epoch": 1.8235294117647058,
"grad_norm": 0.09415315091609955,
"learning_rate": 6.433849906432501e-05,
"loss": 0.947021484375,
"step": 124
},
{
"epoch": 1.8382352941176472,
"grad_norm": 0.10217582434415817,
"learning_rate": 6.381201243373142e-05,
"loss": 0.8857421875,
"step": 125
},
{
"epoch": 1.8529411764705883,
"grad_norm": 0.10179689526557922,
"learning_rate": 6.3284268447349e-05,
"loss": 0.900634765625,
"step": 126
},
{
"epoch": 1.8676470588235294,
"grad_norm": 0.09513360261917114,
"learning_rate": 6.275534240715461e-05,
"loss": 0.9052734375,
"step": 127
},
{
"epoch": 1.8823529411764706,
"grad_norm": 0.09685131907463074,
"learning_rate": 6.22253097837883e-05,
"loss": 0.990234375,
"step": 128
},
{
"epoch": 1.8970588235294117,
"grad_norm": 0.10882526636123657,
"learning_rate": 6.169424620578465e-05,
"loss": 0.82373046875,
"step": 129
},
{
"epoch": 1.9117647058823528,
"grad_norm": 0.10201963037252426,
"learning_rate": 6.116222744878165e-05,
"loss": 0.904296875,
"step": 130
},
{
"epoch": 1.9264705882352942,
"grad_norm": 0.09556466341018677,
"learning_rate": 6.062932942470851e-05,
"loss": 0.937744140625,
"step": 131
},
{
"epoch": 1.9411764705882353,
"grad_norm": 0.0984668955206871,
"learning_rate": 6.009562817095399e-05,
"loss": 0.796142578125,
"step": 132
},
{
"epoch": 1.9558823529411766,
"grad_norm": 0.12080808728933334,
"learning_rate": 5.956119983951698e-05,
"loss": 0.857666015625,
"step": 133
},
{
"epoch": 1.9705882352941178,
"grad_norm": 0.10191366821527481,
"learning_rate": 5.902612068614057e-05,
"loss": 0.769287109375,
"step": 134
},
{
"epoch": 1.9852941176470589,
"grad_norm": 0.09555588662624359,
"learning_rate": 5.8490467059431364e-05,
"loss": 1.00244140625,
"step": 135
},
{
"epoch": 2.0,
"grad_norm": 0.10878229141235352,
"learning_rate": 5.795431538996568e-05,
"loss": 1.03173828125,
"step": 136
},
{
"epoch": 2.014705882352941,
"grad_norm": 0.11508411169052124,
"learning_rate": 5.741774217938373e-05,
"loss": 0.94580078125,
"step": 137
},
{
"epoch": 2.0294117647058822,
"grad_norm": 0.09879229962825775,
"learning_rate": 5.6880823989474096e-05,
"loss": 0.777099609375,
"step": 138
},
{
"epoch": 2.0441176470588234,
"grad_norm": 0.11903475224971771,
"learning_rate": 5.634363743124919e-05,
"loss": 0.96875,
"step": 139
},
{
"epoch": 2.0588235294117645,
"grad_norm": 0.0968213602900505,
"learning_rate": 5.5806259154014004e-05,
"loss": 1.00732421875,
"step": 140
},
{
"epoch": 2.073529411764706,
"grad_norm": 0.10713297128677368,
"learning_rate": 5.526876583442928e-05,
"loss": 0.761962890625,
"step": 141
},
{
"epoch": 2.088235294117647,
"grad_norm": 0.11751079559326172,
"learning_rate": 5.4731234165570745e-05,
"loss": 0.877685546875,
"step": 142
},
{
"epoch": 2.1029411764705883,
"grad_norm": 0.09544238448143005,
"learning_rate": 5.4193740845986006e-05,
"loss": 0.890380859375,
"step": 143
},
{
"epoch": 2.1176470588235294,
"grad_norm": 0.10961336642503738,
"learning_rate": 5.365636256875084e-05,
"loss": 0.89404296875,
"step": 144
},
{
"epoch": 2.1323529411764706,
"grad_norm": 0.10094622522592545,
"learning_rate": 5.311917601052594e-05,
"loss": 0.97265625,
"step": 145
},
{
"epoch": 2.1470588235294117,
"grad_norm": 0.10295704007148743,
"learning_rate": 5.258225782061629e-05,
"loss": 0.935302734375,
"step": 146
},
{
"epoch": 2.161764705882353,
"grad_norm": 0.10666032880544662,
"learning_rate": 5.204568461003434e-05,
"loss": 0.762939453125,
"step": 147
},
{
"epoch": 2.176470588235294,
"grad_norm": 0.10679176449775696,
"learning_rate": 5.150953294056864e-05,
"loss": 0.96826171875,
"step": 148
},
{
"epoch": 2.1911764705882355,
"grad_norm": 0.1029231995344162,
"learning_rate": 5.097387931385945e-05,
"loss": 0.9541015625,
"step": 149
},
{
"epoch": 2.2058823529411766,
"grad_norm": 0.11718638241291046,
"learning_rate": 5.043880016048303e-05,
"loss": 1.007568359375,
"step": 150
},
{
"epoch": 2.2205882352941178,
"grad_norm": 0.1020374596118927,
"learning_rate": 4.9904371829046016e-05,
"loss": 0.8232421875,
"step": 151
},
{
"epoch": 2.235294117647059,
"grad_norm": 0.10561268031597137,
"learning_rate": 4.937067057529151e-05,
"loss": 0.957275390625,
"step": 152
},
{
"epoch": 2.25,
"grad_norm": 0.11097069084644318,
"learning_rate": 4.883777255121835e-05,
"loss": 0.91162109375,
"step": 153
},
{
"epoch": 2.264705882352941,
"grad_norm": 0.11146210134029388,
"learning_rate": 4.8305753794215367e-05,
"loss": 0.91552734375,
"step": 154
},
{
"epoch": 2.2794117647058822,
"grad_norm": 0.10153888165950775,
"learning_rate": 4.7774690216211714e-05,
"loss": 0.93505859375,
"step": 155
},
{
"epoch": 2.2941176470588234,
"grad_norm": 0.10157613456249237,
"learning_rate": 4.724465759284541e-05,
"loss": 0.92138671875,
"step": 156
},
{
"epoch": 2.3088235294117645,
"grad_norm": 0.10186035186052322,
"learning_rate": 4.671573155265101e-05,
"loss": 1.05810546875,
"step": 157
},
{
"epoch": 2.323529411764706,
"grad_norm": 0.10635138303041458,
"learning_rate": 4.6187987566268586e-05,
"loss": 0.89501953125,
"step": 158
},
{
"epoch": 2.338235294117647,
"grad_norm": 0.09509985893964767,
"learning_rate": 4.5661500935675e-05,
"loss": 0.87939453125,
"step": 159
},
{
"epoch": 2.3529411764705883,
"grad_norm": 0.1142621859908104,
"learning_rate": 4.513634678343932e-05,
"loss": 0.949462890625,
"step": 160
},
{
"epoch": 2.3676470588235294,
"grad_norm": 0.09869854897260666,
"learning_rate": 4.4612600042003815e-05,
"loss": 0.90625,
"step": 161
},
{
"epoch": 2.3823529411764706,
"grad_norm": 0.10569340735673904,
"learning_rate": 4.409033544299226e-05,
"loss": 1.00634765625,
"step": 162
},
{
"epoch": 2.3970588235294117,
"grad_norm": 0.17637315392494202,
"learning_rate": 4.3569627506546476e-05,
"loss": 0.951416015625,
"step": 163
},
{
"epoch": 2.411764705882353,
"grad_norm": 0.1038566380739212,
"learning_rate": 4.3050550530693545e-05,
"loss": 0.8310546875,
"step": 164
},
{
"epoch": 2.411764705882353,
"eval_loss": 0.9248046875,
"eval_runtime": 5.5331,
"eval_samples_per_second": 0.723,
"eval_steps_per_second": 0.181,
"step": 164
},
{
"epoch": 2.426470588235294,
"grad_norm": 0.09599288552999496,
"learning_rate": 4.253317858074432e-05,
"loss": 0.775146484375,
"step": 165
},
{
"epoch": 2.4411764705882355,
"grad_norm": 0.11044510453939438,
"learning_rate": 4.2017585478725466e-05,
"loss": 0.965087890625,
"step": 166
},
{
"epoch": 2.4558823529411766,
"grad_norm": 0.1129738911986351,
"learning_rate": 4.150384479284581e-05,
"loss": 0.939453125,
"step": 167
},
{
"epoch": 2.4705882352941178,
"grad_norm": 0.10469809174537659,
"learning_rate": 4.099202982699938e-05,
"loss": 1.0068359375,
"step": 168
},
{
"epoch": 2.485294117647059,
"grad_norm": 0.09376489371061325,
"learning_rate": 4.04822136103058e-05,
"loss": 0.8895263671875,
"step": 169
},
{
"epoch": 2.5,
"grad_norm": 0.10994348675012589,
"learning_rate": 3.997446888669001e-05,
"loss": 0.966064453125,
"step": 170
},
{
"epoch": 2.514705882352941,
"grad_norm": 0.10015475749969482,
"learning_rate": 3.9468868104502775e-05,
"loss": 0.97119140625,
"step": 171
},
{
"epoch": 2.5294117647058822,
"grad_norm": 0.10487864911556244,
"learning_rate": 3.896548340618317e-05,
"loss": 0.88720703125,
"step": 172
},
{
"epoch": 2.5441176470588234,
"grad_norm": 0.110616035759449,
"learning_rate": 3.846438661796492e-05,
"loss": 0.77001953125,
"step": 173
},
{
"epoch": 2.5588235294117645,
"grad_norm": 0.11588707566261292,
"learning_rate": 3.7965649239627656e-05,
"loss": 0.990234375,
"step": 174
},
{
"epoch": 2.5735294117647056,
"grad_norm": 0.11195044219493866,
"learning_rate": 3.746934243429499e-05,
"loss": 1.037109375,
"step": 175
},
{
"epoch": 2.588235294117647,
"grad_norm": 0.12539899349212646,
"learning_rate": 3.697553701828036e-05,
"loss": 0.94677734375,
"step": 176
},
{
"epoch": 2.6029411764705883,
"grad_norm": 0.11014209687709808,
"learning_rate": 3.648430345098257e-05,
"loss": 0.94482421875,
"step": 177
},
{
"epoch": 2.6176470588235294,
"grad_norm": 0.12514397501945496,
"learning_rate": 3.599571182483218e-05,
"loss": 1.0400390625,
"step": 178
},
{
"epoch": 2.6323529411764706,
"grad_norm": 0.13499069213867188,
"learning_rate": 3.550983185529023e-05,
"loss": 0.9189453125,
"step": 179
},
{
"epoch": 2.6470588235294117,
"grad_norm": 0.11633610725402832,
"learning_rate": 3.502673287090083e-05,
"loss": 0.88525390625,
"step": 180
},
{
"epoch": 2.661764705882353,
"grad_norm": 0.11220976710319519,
"learning_rate": 3.4546483803398735e-05,
"loss": 0.93115234375,
"step": 181
},
{
"epoch": 2.6764705882352944,
"grad_norm": 0.10854318737983704,
"learning_rate": 3.4069153177874026e-05,
"loss": 0.9609375,
"step": 182
},
{
"epoch": 2.6911764705882355,
"grad_norm": 0.1204732209444046,
"learning_rate": 3.359480910299424e-05,
"loss": 0.8818359375,
"step": 183
},
{
"epoch": 2.7058823529411766,
"grad_norm": 0.11446163803339005,
"learning_rate": 3.312351926128622e-05,
"loss": 0.86328125,
"step": 184
},
{
"epoch": 2.7205882352941178,
"grad_norm": 0.11880532652139664,
"learning_rate": 3.265535089947881e-05,
"loss": 0.96630859375,
"step": 185
},
{
"epoch": 2.735294117647059,
"grad_norm": 0.12591275572776794,
"learning_rate": 3.219037081890766e-05,
"loss": 1.00830078125,
"step": 186
},
{
"epoch": 2.75,
"grad_norm": 0.10148783028125763,
"learning_rate": 3.172864536598341e-05,
"loss": 0.70513916015625,
"step": 187
},
{
"epoch": 2.764705882352941,
"grad_norm": 0.13037607073783875,
"learning_rate": 3.127024042272515e-05,
"loss": 0.8466796875,
"step": 188
},
{
"epoch": 2.7794117647058822,
"grad_norm": 0.12317381799221039,
"learning_rate": 3.081522139735977e-05,
"loss": 0.9638671875,
"step": 189
},
{
"epoch": 2.7941176470588234,
"grad_norm": 0.11696892976760864,
"learning_rate": 3.0363653214989295e-05,
"loss": 1.0205078125,
"step": 190
},
{
"epoch": 2.8088235294117645,
"grad_norm": 0.11366445571184158,
"learning_rate": 2.991560030832666e-05,
"loss": 0.90771484375,
"step": 191
},
{
"epoch": 2.8235294117647056,
"grad_norm": 0.10979506373405457,
"learning_rate": 2.947112660850227e-05,
"loss": 0.914306640625,
"step": 192
},
{
"epoch": 2.838235294117647,
"grad_norm": 0.13060979545116425,
"learning_rate": 2.90302955359418e-05,
"loss": 0.849609375,
"step": 193
},
{
"epoch": 2.8529411764705883,
"grad_norm": 0.11327344179153442,
"learning_rate": 2.859316999131691e-05,
"loss": 0.866455078125,
"step": 194
},
{
"epoch": 2.8676470588235294,
"grad_norm": 0.10893069952726364,
"learning_rate": 2.8159812346570253e-05,
"loss": 0.8720703125,
"step": 195
},
{
"epoch": 2.8823529411764706,
"grad_norm": 0.1116403266787529,
"learning_rate": 2.7730284436015763e-05,
"loss": 0.956787109375,
"step": 196
},
{
"epoch": 2.8970588235294117,
"grad_norm": 0.11271966993808746,
"learning_rate": 2.7304647547515756e-05,
"loss": 0.79345703125,
"step": 197
},
{
"epoch": 2.911764705882353,
"grad_norm": 0.1136358305811882,
"learning_rate": 2.6882962413735973e-05,
"loss": 0.86865234375,
"step": 198
},
{
"epoch": 2.9264705882352944,
"grad_norm": 0.10731173306703568,
"learning_rate": 2.646528920347986e-05,
"loss": 0.90576171875,
"step": 199
},
{
"epoch": 2.9411764705882355,
"grad_norm": 0.11048389226198196,
"learning_rate": 2.6051687513103284e-05,
"loss": 0.76611328125,
"step": 200
},
{
"epoch": 2.9558823529411766,
"grad_norm": 0.11182138323783875,
"learning_rate": 2.5642216358010856e-05,
"loss": 0.82470703125,
"step": 201
},
{
"epoch": 2.9705882352941178,
"grad_norm": 0.11982689052820206,
"learning_rate": 2.5236934164235387e-05,
"loss": 0.739013671875,
"step": 202
},
{
"epoch": 2.985294117647059,
"grad_norm": 0.11637600511312485,
"learning_rate": 2.4835898760101155e-05,
"loss": 0.973388671875,
"step": 203
},
{
"epoch": 3.0,
"grad_norm": 0.12302377820014954,
"learning_rate": 2.4439167367972557e-05,
"loss": 0.990234375,
"step": 204
},
{
"epoch": 3.014705882352941,
"grad_norm": 0.11300607025623322,
"learning_rate": 2.404679659608934e-05,
"loss": 0.908203125,
"step": 205
},
{
"epoch": 3.0294117647058822,
"grad_norm": 0.09651143103837967,
"learning_rate": 2.3658842430489367e-05,
"loss": 0.7515869140625,
"step": 206
},
{
"epoch": 3.0441176470588234,
"grad_norm": 0.1241968423128128,
"learning_rate": 2.3275360227020064e-05,
"loss": 0.939453125,
"step": 207
},
{
"epoch": 3.0588235294117645,
"grad_norm": 0.1058558002114296,
"learning_rate": 2.2896404703439925e-05,
"loss": 0.978759765625,
"step": 208
},
{
"epoch": 3.073529411764706,
"grad_norm": 0.11273318529129028,
"learning_rate": 2.2522029931611072e-05,
"loss": 0.734619140625,
"step": 209
},
{
"epoch": 3.088235294117647,
"grad_norm": 0.11540959775447845,
"learning_rate": 2.2152289329783928e-05,
"loss": 0.849609375,
"step": 210
},
{
"epoch": 3.1029411764705883,
"grad_norm": 0.14421139657497406,
"learning_rate": 2.178723565497503e-05,
"loss": 0.863525390625,
"step": 211
},
{
"epoch": 3.1176470588235294,
"grad_norm": 0.12274222820997238,
"learning_rate": 2.1426920995439453e-05,
"loss": 0.8642578125,
"step": 212
},
{
"epoch": 3.1323529411764706,
"grad_norm": 0.09908393770456314,
"learning_rate": 2.1071396763238433e-05,
"loss": 0.94775390625,
"step": 213
},
{
"epoch": 3.1470588235294117,
"grad_norm": 0.11080903559923172,
"learning_rate": 2.072071368690363e-05,
"loss": 0.908203125,
"step": 214
},
{
"epoch": 3.161764705882353,
"grad_norm": 0.10474691540002823,
"learning_rate": 2.037492180419874e-05,
"loss": 0.7373046875,
"step": 215
},
{
"epoch": 3.176470588235294,
"grad_norm": 0.1099315881729126,
"learning_rate": 2.0034070454979865e-05,
"loss": 0.9423828125,
"step": 216
},
{
"epoch": 3.1911764705882355,
"grad_norm": 0.11670861393213272,
"learning_rate": 1.9698208274155355e-05,
"loss": 0.92236328125,
"step": 217
},
{
"epoch": 3.2058823529411766,
"grad_norm": 0.10826999694108963,
"learning_rate": 1.9367383184746255e-05,
"loss": 0.978515625,
"step": 218
},
{
"epoch": 3.2058823529411766,
"eval_loss": 0.9169921875,
"eval_runtime": 5.5373,
"eval_samples_per_second": 0.722,
"eval_steps_per_second": 0.181,
"step": 218
},
{
"epoch": 3.2205882352941178,
"grad_norm": 0.10669431835412979,
"learning_rate": 1.9041642391048334e-05,
"loss": 0.798828125,
"step": 219
},
{
"epoch": 3.235294117647059,
"grad_norm": 0.10440842062234879,
"learning_rate": 1.8721032371896673e-05,
"loss": 0.9296875,
"step": 220
},
{
"epoch": 3.25,
"grad_norm": 0.1103634312748909,
"learning_rate": 1.840559887403376e-05,
"loss": 0.8857421875,
"step": 221
},
{
"epoch": 3.264705882352941,
"grad_norm": 0.11505292356014252,
"learning_rate": 1.8095386905582014e-05,
"loss": 0.890869140625,
"step": 222
},
{
"epoch": 3.2794117647058822,
"grad_norm": 0.11045239120721817,
"learning_rate": 1.779044072962178e-05,
"loss": 0.908935546875,
"step": 223
},
{
"epoch": 3.2941176470588234,
"grad_norm": 0.107695572078228,
"learning_rate": 1.7490803857875565e-05,
"loss": 0.89599609375,
"step": 224
},
{
"epoch": 3.3088235294117645,
"grad_norm": 0.10176144540309906,
"learning_rate": 1.719651904449946e-05,
"loss": 1.03271484375,
"step": 225
},
{
"epoch": 3.323529411764706,
"grad_norm": 0.11387171596288681,
"learning_rate": 1.6907628279982757e-05,
"loss": 0.870849609375,
"step": 226
},
{
"epoch": 3.338235294117647,
"grad_norm": 0.10046130418777466,
"learning_rate": 1.6624172785156438e-05,
"loss": 0.858154296875,
"step": 227
},
{
"epoch": 3.3529411764705883,
"grad_norm": 0.11487563699483871,
"learning_rate": 1.6346193005311497e-05,
"loss": 0.922607421875,
"step": 228
},
{
"epoch": 3.3676470588235294,
"grad_norm": 0.1084771454334259,
"learning_rate": 1.6073728604428005e-05,
"loss": 0.88525390625,
"step": 229
},
{
"epoch": 3.3823529411764706,
"grad_norm": 0.11252636462450027,
"learning_rate": 1.580681845951552e-05,
"loss": 0.981689453125,
"step": 230
},
{
"epoch": 3.3970588235294117,
"grad_norm": 0.11360497027635574,
"learning_rate": 1.554550065506592e-05,
"loss": 0.927490234375,
"step": 231
},
{
"epoch": 3.411764705882353,
"grad_norm": 0.1092953234910965,
"learning_rate": 1.528981247761917e-05,
"loss": 0.80810546875,
"step": 232
},
{
"epoch": 3.426470588235294,
"grad_norm": 0.1018129214644432,
"learning_rate": 1.5039790410443166e-05,
"loss": 0.75537109375,
"step": 233
},
{
"epoch": 3.4411764705882355,
"grad_norm": 0.1259973645210266,
"learning_rate": 1.4795470128327967e-05,
"loss": 0.94287109375,
"step": 234
},
{
"epoch": 3.4558823529411766,
"grad_norm": 0.12532128393650055,
"learning_rate": 1.4556886492495475e-05,
"loss": 0.9169921875,
"step": 235
},
{
"epoch": 3.4705882352941178,
"grad_norm": 0.12567968666553497,
"learning_rate": 1.4324073545625252e-05,
"loss": 0.984375,
"step": 236
},
{
"epoch": 3.485294117647059,
"grad_norm": 0.10256142914295197,
"learning_rate": 1.4097064506997108e-05,
"loss": 0.87225341796875,
"step": 237
},
{
"epoch": 3.5,
"grad_norm": 0.11032252013683319,
"learning_rate": 1.3875891767751076e-05,
"loss": 0.943115234375,
"step": 238
},
{
"epoch": 3.514705882352941,
"grad_norm": 0.11118175089359283,
"learning_rate": 1.3660586886265678e-05,
"loss": 0.948974609375,
"step": 239
},
{
"epoch": 3.5294117647058822,
"grad_norm": 0.18247246742248535,
"learning_rate": 1.3451180583654977e-05,
"loss": 0.86572265625,
"step": 240
},
{
"epoch": 3.5441176470588234,
"grad_norm": 0.11330296099185944,
"learning_rate": 1.3247702739385104e-05,
"loss": 0.752197265625,
"step": 241
},
{
"epoch": 3.5588235294117645,
"grad_norm": 0.14029426872730255,
"learning_rate": 1.3050182387010768e-05,
"loss": 0.9677734375,
"step": 242
},
{
"epoch": 3.5735294117647056,
"grad_norm": 0.10321860760450363,
"learning_rate": 1.2858647710032667e-05,
"loss": 1.01513671875,
"step": 243
},
{
"epoch": 3.588235294117647,
"grad_norm": 0.11792850494384766,
"learning_rate": 1.2673126037876027e-05,
"loss": 0.92529296875,
"step": 244
},
{
"epoch": 3.6029411764705883,
"grad_norm": 0.11715029925107956,
"learning_rate": 1.2493643841991063e-05,
"loss": 0.923828125,
"step": 245
},
{
"epoch": 3.6176470588235294,
"grad_norm": 0.12409134954214096,
"learning_rate": 1.2320226732075857e-05,
"loss": 1.02099609375,
"step": 246
},
{
"epoch": 3.6323529411764706,
"grad_norm": 0.11123975366353989,
"learning_rate": 1.2152899452422219e-05,
"loss": 0.89697265625,
"step": 247
},
{
"epoch": 3.6470588235294117,
"grad_norm": 0.09144707024097443,
"learning_rate": 1.1991685878384982e-05,
"loss": 0.868408203125,
"step": 248
},
{
"epoch": 3.661764705882353,
"grad_norm": 0.12097735702991486,
"learning_rate": 1.1836609012975333e-05,
"loss": 0.91064453125,
"step": 249
},
{
"epoch": 3.6764705882352944,
"grad_norm": 0.11031545698642731,
"learning_rate": 1.1687690983578583e-05,
"loss": 0.942138671875,
"step": 250
},
{
"epoch": 3.6911764705882355,
"grad_norm": 0.10610030591487885,
"learning_rate": 1.154495303879687e-05,
"loss": 0.863525390625,
"step": 251
},
{
"epoch": 3.7058823529411766,
"grad_norm": 0.10821826756000519,
"learning_rate": 1.140841554541729e-05,
"loss": 0.843994140625,
"step": 252
},
{
"epoch": 3.7205882352941178,
"grad_norm": 0.1155221089720726,
"learning_rate": 1.1278097985505839e-05,
"loss": 0.945556640625,
"step": 253
},
{
"epoch": 3.735294117647059,
"grad_norm": 0.11489252001047134,
"learning_rate": 1.1154018953627554e-05,
"loss": 0.99169921875,
"step": 254
},
{
"epoch": 3.75,
"grad_norm": 0.0971110463142395,
"learning_rate": 1.1036196154193338e-05,
"loss": 0.692779541015625,
"step": 255
},
{
"epoch": 3.764705882352941,
"grad_norm": 0.11258389800786972,
"learning_rate": 1.0924646398933763e-05,
"loss": 0.832275390625,
"step": 256
},
{
"epoch": 3.7794117647058822,
"grad_norm": 0.1158316507935524,
"learning_rate": 1.0819385604500302e-05,
"loss": 0.94677734375,
"step": 257
},
{
"epoch": 3.7941176470588234,
"grad_norm": 0.1032714769244194,
"learning_rate": 1.0720428790194178e-05,
"loss": 1.0029296875,
"step": 258
},
{
"epoch": 3.8088235294117645,
"grad_norm": 0.1114787757396698,
"learning_rate": 1.0627790075823346e-05,
"loss": 0.8896484375,
"step": 259
},
{
"epoch": 3.8235294117647056,
"grad_norm": 0.11017472296953201,
"learning_rate": 1.0541482679687797e-05,
"loss": 0.89990234375,
"step": 260
},
{
"epoch": 3.838235294117647,
"grad_norm": 0.11716391891241074,
"learning_rate": 1.0461518916693454e-05,
"loss": 0.8349609375,
"step": 261
},
{
"epoch": 3.8529411764705883,
"grad_norm": 0.11990869045257568,
"learning_rate": 1.0387910196595022e-05,
"loss": 0.85009765625,
"step": 262
},
{
"epoch": 3.8676470588235294,
"grad_norm": 0.10665875673294067,
"learning_rate": 1.0320667022367954e-05,
"loss": 0.8564453125,
"step": 263
},
{
"epoch": 3.8823529411764706,
"grad_norm": 0.11088557541370392,
"learning_rate": 1.0259798988709827e-05,
"loss": 0.939208984375,
"step": 264
},
{
"epoch": 3.8970588235294117,
"grad_norm": 0.12513795495033264,
"learning_rate": 1.0205314780671325e-05,
"loss": 0.7763671875,
"step": 265
},
{
"epoch": 3.911764705882353,
"grad_norm": 0.1149444654583931,
"learning_rate": 1.0157222172416945e-05,
"loss": 0.854736328125,
"step": 266
},
{
"epoch": 3.9264705882352944,
"grad_norm": 0.11747992783784866,
"learning_rate": 1.0115528026115793e-05,
"loss": 0.89208984375,
"step": 267
},
{
"epoch": 3.9411764705882355,
"grad_norm": 0.11170803755521774,
"learning_rate": 1.0080238290962398e-05,
"loss": 0.7529296875,
"step": 268
},
{
"epoch": 3.9558823529411766,
"grad_norm": 0.13009798526763916,
"learning_rate": 1.0051358002327869e-05,
"loss": 0.809814453125,
"step": 269
},
{
"epoch": 3.9705882352941178,
"grad_norm": 0.13089829683303833,
"learning_rate": 1.0028891281041395e-05,
"loss": 0.72265625,
"step": 270
},
{
"epoch": 3.985294117647059,
"grad_norm": 0.11525954306125641,
"learning_rate": 1.0012841332802281e-05,
"loss": 0.960205078125,
"step": 271
},
{
"epoch": 4.0,
"grad_norm": 0.1164756715297699,
"learning_rate": 1.0003210447722509e-05,
"loss": 0.97216796875,
"step": 272
},
{
"epoch": 4.0,
"eval_loss": 0.9111328125,
"eval_runtime": 5.5409,
"eval_samples_per_second": 0.722,
"eval_steps_per_second": 0.18,
"step": 272
}
],
"logging_steps": 1.0,
"max_steps": 272,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 0,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.3666297122777465e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}