CashWise / trainer_state.json
RayNene's picture
Upload 9 files
ec4ca0f verified
Raw History Blame Contribute Delete
60.7 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 65,
"global_step": 327,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.009174311926605505,
"grad_norm": 0.22605681419372559,
"learning_rate": 0.0,
"loss": 1.728515625,
"step": 1
},
{
"epoch": 0.01834862385321101,
"grad_norm": 0.19375473260879517,
"learning_rate": 1.1764705882352942e-05,
"loss": 1.6748046875,
"step": 2
},
{
"epoch": 0.027522935779816515,
"grad_norm": 0.19125375151634216,
"learning_rate": 2.3529411764705884e-05,
"loss": 1.685546875,
"step": 3
},
{
"epoch": 0.03669724770642202,
"grad_norm": 0.17866532504558563,
"learning_rate": 3.529411764705883e-05,
"loss": 1.6982421875,
"step": 4
},
{
"epoch": 0.045871559633027525,
"grad_norm": 0.1125837117433548,
"learning_rate": 4.705882352941177e-05,
"loss": 1.5732421875,
"step": 5
},
{
"epoch": 0.05504587155963303,
"grad_norm": 0.09169502556324005,
"learning_rate": 5.882352941176471e-05,
"loss": 1.49609375,
"step": 6
},
{
"epoch": 0.06422018348623854,
"grad_norm": 0.1863684207201004,
"learning_rate": 7.058823529411765e-05,
"loss": 1.5703125,
"step": 7
},
{
"epoch": 0.07339449541284404,
"grad_norm": 0.13687698543071747,
"learning_rate": 8.23529411764706e-05,
"loss": 1.5126953125,
"step": 8
},
{
"epoch": 0.08256880733944955,
"grad_norm": 0.10407889634370804,
"learning_rate": 9.411764705882353e-05,
"loss": 1.4423828125,
"step": 9
},
{
"epoch": 0.09174311926605505,
"grad_norm": 0.07455075532197952,
"learning_rate": 0.00010588235294117647,
"loss": 1.4189453125,
"step": 10
},
{
"epoch": 0.10091743119266056,
"grad_norm": 0.06540790945291519,
"learning_rate": 0.00011764705882352942,
"loss": 1.326171875,
"step": 11
},
{
"epoch": 0.11009174311926606,
"grad_norm": 0.06495493650436401,
"learning_rate": 0.00012941176470588237,
"loss": 1.41015625,
"step": 12
},
{
"epoch": 0.11926605504587157,
"grad_norm": 0.05407547950744629,
"learning_rate": 0.0001411764705882353,
"loss": 1.3798828125,
"step": 13
},
{
"epoch": 0.12844036697247707,
"grad_norm": 0.057004112750291824,
"learning_rate": 0.00015294117647058822,
"loss": 1.3544921875,
"step": 14
},
{
"epoch": 0.13761467889908258,
"grad_norm": 0.06124591454863548,
"learning_rate": 0.0001647058823529412,
"loss": 1.3291015625,
"step": 15
},
{
"epoch": 0.14678899082568808,
"grad_norm": 0.06303200125694275,
"learning_rate": 0.00017647058823529413,
"loss": 1.17919921875,
"step": 16
},
{
"epoch": 0.1559633027522936,
"grad_norm": 0.06791990995407104,
"learning_rate": 0.00018823529411764707,
"loss": 1.3544921875,
"step": 17
},
{
"epoch": 0.1651376146788991,
"grad_norm": 0.061412930488586426,
"learning_rate": 0.0002,
"loss": 1.337890625,
"step": 18
},
{
"epoch": 0.1743119266055046,
"grad_norm": 0.062398556619882584,
"learning_rate": 0.00019999537847661833,
"loss": 1.3291015625,
"step": 19
},
{
"epoch": 0.1834862385321101,
"grad_norm": 0.05166026949882507,
"learning_rate": 0.00019998151438110616,
"loss": 1.2568359375,
"step": 20
},
{
"epoch": 0.1926605504587156,
"grad_norm": 0.05899413302540779,
"learning_rate": 0.00019995840913731333,
"loss": 1.259765625,
"step": 21
},
{
"epoch": 0.2018348623853211,
"grad_norm": 0.05736570805311203,
"learning_rate": 0.0001999260651181603,
"loss": 1.267578125,
"step": 22
},
{
"epoch": 0.21100917431192662,
"grad_norm": 0.05624833330512047,
"learning_rate": 0.00019988448564539475,
"loss": 1.2783203125,
"step": 23
},
{
"epoch": 0.22018348623853212,
"grad_norm": 0.05654808506369591,
"learning_rate": 0.0001998336749892501,
"loss": 1.26171875,
"step": 24
},
{
"epoch": 0.22935779816513763,
"grad_norm": 0.04877206310629845,
"learning_rate": 0.0001997736383680071,
"loss": 1.26953125,
"step": 25
},
{
"epoch": 0.23853211009174313,
"grad_norm": 0.046762432903051376,
"learning_rate": 0.00019970438194745797,
"loss": 1.2158203125,
"step": 26
},
{
"epoch": 0.24770642201834864,
"grad_norm": 0.0464463084936142,
"learning_rate": 0.0001996259128402731,
"loss": 1.2373046875,
"step": 27
},
{
"epoch": 0.25688073394495414,
"grad_norm": 0.04192713275551796,
"learning_rate": 0.00019953823910527058,
"loss": 1.1982421875,
"step": 28
},
{
"epoch": 0.26605504587155965,
"grad_norm": 0.05698929354548454,
"learning_rate": 0.00019944136974658848,
"loss": 1.1953125,
"step": 29
},
{
"epoch": 0.27522935779816515,
"grad_norm": 0.04239416494965553,
"learning_rate": 0.00019933531471276036,
"loss": 1.201171875,
"step": 30
},
{
"epoch": 0.28440366972477066,
"grad_norm": 0.04659745469689369,
"learning_rate": 0.00019922008489569327,
"loss": 1.15625,
"step": 31
},
{
"epoch": 0.29357798165137616,
"grad_norm": 0.04682194069027901,
"learning_rate": 0.00019909569212954925,
"loss": 1.17626953125,
"step": 32
},
{
"epoch": 0.30275229357798167,
"grad_norm": 0.03945130109786987,
"learning_rate": 0.00019896214918953002,
"loss": 1.181640625,
"step": 33
},
{
"epoch": 0.3119266055045872,
"grad_norm": 0.05257967486977577,
"learning_rate": 0.000198819469790565,
"loss": 1.169921875,
"step": 34
},
{
"epoch": 0.3211009174311927,
"grad_norm": 0.041180651634931564,
"learning_rate": 0.0001986676685859025,
"loss": 1.1591796875,
"step": 35
},
{
"epoch": 0.3302752293577982,
"grad_norm": 0.042695317417383194,
"learning_rate": 0.00019850676116560517,
"loss": 1.16357421875,
"step": 36
},
{
"epoch": 0.3394495412844037,
"grad_norm": 0.04641354829072952,
"learning_rate": 0.00019833676405494862,
"loss": 1.1279296875,
"step": 37
},
{
"epoch": 0.3486238532110092,
"grad_norm": 0.044326573610305786,
"learning_rate": 0.00019815769471272452,
"loss": 1.1513671875,
"step": 38
},
{
"epoch": 0.3577981651376147,
"grad_norm": 0.052880674600601196,
"learning_rate": 0.00019796957152944732,
"loss": 1.1435546875,
"step": 39
},
{
"epoch": 0.3669724770642202,
"grad_norm": 0.040993958711624146,
"learning_rate": 0.00019777241382546568,
"loss": 1.1416015625,
"step": 40
},
{
"epoch": 0.3761467889908257,
"grad_norm": 0.042091961950063705,
"learning_rate": 0.00019756624184897825,
"loss": 1.15478515625,
"step": 41
},
{
"epoch": 0.3853211009174312,
"grad_norm": 0.042426034808158875,
"learning_rate": 0.00019735107677395415,
"loss": 1.142578125,
"step": 42
},
{
"epoch": 0.3944954128440367,
"grad_norm": 0.04127749428153038,
"learning_rate": 0.00019712694069795838,
"loss": 1.138671875,
"step": 43
},
{
"epoch": 0.4036697247706422,
"grad_norm": 0.04473963752388954,
"learning_rate": 0.00019689385663988242,
"loss": 1.130859375,
"step": 44
},
{
"epoch": 0.41284403669724773,
"grad_norm": 0.042432017624378204,
"learning_rate": 0.00019665184853758012,
"loss": 1.1298828125,
"step": 45
},
{
"epoch": 0.42201834862385323,
"grad_norm": 0.04143335670232773,
"learning_rate": 0.00019640094124540934,
"loss": 1.10546875,
"step": 46
},
{
"epoch": 0.43119266055045874,
"grad_norm": 0.041434649378061295,
"learning_rate": 0.00019614116053167936,
"loss": 1.14111328125,
"step": 47
},
{
"epoch": 0.44036697247706424,
"grad_norm": 0.04477505385875702,
"learning_rate": 0.00019587253307600438,
"loss": 1.08447265625,
"step": 48
},
{
"epoch": 0.44954128440366975,
"grad_norm": 0.04183594509959221,
"learning_rate": 0.00019559508646656385,
"loss": 1.0810546875,
"step": 49
},
{
"epoch": 0.45871559633027525,
"grad_norm": 0.045960474759340286,
"learning_rate": 0.00019530884919726854,
"loss": 1.10205078125,
"step": 50
},
{
"epoch": 0.46788990825688076,
"grad_norm": 0.042445652186870575,
"learning_rate": 0.00019501385066483484,
"loss": 1.078125,
"step": 51
},
{
"epoch": 0.47706422018348627,
"grad_norm": 0.04624992609024048,
"learning_rate": 0.0001947101211657653,
"loss": 1.060546875,
"step": 52
},
{
"epoch": 0.48623853211009177,
"grad_norm": 0.044562436640262604,
"learning_rate": 0.00019439769189323727,
"loss": 1.10400390625,
"step": 53
},
{
"epoch": 0.4954128440366973,
"grad_norm": 0.04860757291316986,
"learning_rate": 0.00019407659493389935,
"loss": 1.08203125,
"step": 54
},
{
"epoch": 0.5045871559633027,
"grad_norm": 0.04623546451330185,
"learning_rate": 0.00019374686326457613,
"loss": 1.08319091796875,
"step": 55
},
{
"epoch": 0.5137614678899083,
"grad_norm": 0.04296131432056427,
"learning_rate": 0.0001934085307488814,
"loss": 1.0693359375,
"step": 56
},
{
"epoch": 0.5229357798165137,
"grad_norm": 0.04321106895804405,
"learning_rate": 0.00019306163213374028,
"loss": 1.05126953125,
"step": 57
},
{
"epoch": 0.5321100917431193,
"grad_norm": 0.04590374231338501,
"learning_rate": 0.00019270620304582078,
"loss": 1.0947265625,
"step": 58
},
{
"epoch": 0.5412844036697247,
"grad_norm": 0.046930886805057526,
"learning_rate": 0.00019234227998787492,
"loss": 1.0908203125,
"step": 59
},
{
"epoch": 0.5504587155963303,
"grad_norm": 0.04378589987754822,
"learning_rate": 0.00019196990033498986,
"loss": 1.07373046875,
"step": 60
},
{
"epoch": 0.5596330275229358,
"grad_norm": 0.040084123611450195,
"learning_rate": 0.00019158910233074946,
"loss": 1.06103515625,
"step": 61
},
{
"epoch": 0.5688073394495413,
"grad_norm": 0.04478531703352928,
"learning_rate": 0.00019119992508330659,
"loss": 1.09375,
"step": 62
},
{
"epoch": 0.5779816513761468,
"grad_norm": 0.04241171479225159,
"learning_rate": 0.00019080240856136677,
"loss": 1.0390625,
"step": 63
},
{
"epoch": 0.5871559633027523,
"grad_norm": 0.0481603778898716,
"learning_rate": 0.00019039659359008332,
"loss": 1.09033203125,
"step": 64
},
{
"epoch": 0.5963302752293578,
"grad_norm": 0.04408343508839607,
"learning_rate": 0.0001899825218468647,
"loss": 1.06640625,
"step": 65
},
{
"epoch": 0.6055045871559633,
"grad_norm": 0.04367062449455261,
"learning_rate": 0.00018956023585709405,
"loss": 1.046875,
"step": 66
},
{
"epoch": 0.6146788990825688,
"grad_norm": 0.04499538615345955,
"learning_rate": 0.00018912977898976198,
"loss": 1.06201171875,
"step": 67
},
{
"epoch": 0.6146788990825688,
"eval_loss": 0.9586181640625,
"eval_runtime": 14.2411,
"eval_samples_per_second": 0.702,
"eval_steps_per_second": 0.14,
"step": 67
},
{
"epoch": 0.6238532110091743,
"grad_norm": 0.04286055266857147,
"learning_rate": 0.0001886911954530124,
"loss": 1.0322265625,
"step": 68
},
{
"epoch": 0.6330275229357798,
"grad_norm": 0.04268469288945198,
"learning_rate": 0.0001882445302896024,
"loss": 1.05126953125,
"step": 69
},
{
"epoch": 0.6422018348623854,
"grad_norm": 0.04346257448196411,
"learning_rate": 0.0001877898293722764,
"loss": 1.09521484375,
"step": 70
},
{
"epoch": 0.6513761467889908,
"grad_norm": 0.04236588254570961,
"learning_rate": 0.00018732713939905482,
"loss": 1.0234375,
"step": 71
},
{
"epoch": 0.6605504587155964,
"grad_norm": 0.040440477430820465,
"learning_rate": 0.0001868565078884383,
"loss": 1.041015625,
"step": 72
},
{
"epoch": 0.6697247706422018,
"grad_norm": 0.04488544538617134,
"learning_rate": 0.0001863779831745276,
"loss": 1.0458984375,
"step": 73
},
{
"epoch": 0.6788990825688074,
"grad_norm": 0.04207679256796837,
"learning_rate": 0.00018589161440205943,
"loss": 1.02734375,
"step": 74
},
{
"epoch": 0.6880733944954128,
"grad_norm": 0.04115382209420204,
"learning_rate": 0.0001853974515213594,
"loss": 0.951416015625,
"step": 75
},
{
"epoch": 0.6972477064220184,
"grad_norm": 0.04315188154578209,
"learning_rate": 0.00018489554528321216,
"loss": 1.05078125,
"step": 76
},
{
"epoch": 0.7064220183486238,
"grad_norm": 0.045279622077941895,
"learning_rate": 0.00018438594723364917,
"loss": 1.0283203125,
"step": 77
},
{
"epoch": 0.7155963302752294,
"grad_norm": 0.044494811445474625,
"learning_rate": 0.0001838687097086549,
"loss": 1.06005859375,
"step": 78
},
{
"epoch": 0.7247706422018348,
"grad_norm": 0.04344740882515907,
"learning_rate": 0.00018334388582879187,
"loss": 1.00537109375,
"step": 79
},
{
"epoch": 0.7339449541284404,
"grad_norm": 0.04212263971567154,
"learning_rate": 0.00018281152949374527,
"loss": 1.04345703125,
"step": 80
},
{
"epoch": 0.7431192660550459,
"grad_norm": 0.04220547899603844,
"learning_rate": 0.00018227169537678734,
"loss": 1.05126953125,
"step": 81
},
{
"epoch": 0.7522935779816514,
"grad_norm": 0.04242934286594391,
"learning_rate": 0.00018172443891916235,
"loss": 1.0078125,
"step": 82
},
{
"epoch": 0.7614678899082569,
"grad_norm": 0.04215303063392639,
"learning_rate": 0.00018116981632439286,
"loss": 1.0390625,
"step": 83
},
{
"epoch": 0.7706422018348624,
"grad_norm": 0.041623715311288834,
"learning_rate": 0.00018060788455250758,
"loss": 0.990234375,
"step": 84
},
{
"epoch": 0.7798165137614679,
"grad_norm": 0.04116380587220192,
"learning_rate": 0.00018003870131419136,
"loss": 1.01708984375,
"step": 85
},
{
"epoch": 0.7889908256880734,
"grad_norm": 0.040912870317697525,
"learning_rate": 0.00017946232506485857,
"loss": 1.02587890625,
"step": 86
},
{
"epoch": 0.7981651376146789,
"grad_norm": 0.041738320142030716,
"learning_rate": 0.00017887881499864953,
"loss": 1.00390625,
"step": 87
},
{
"epoch": 0.8073394495412844,
"grad_norm": 0.04347852244973183,
"learning_rate": 0.00017828823104235118,
"loss": 1.05615234375,
"step": 88
},
{
"epoch": 0.8165137614678899,
"grad_norm": 0.04572346433997154,
"learning_rate": 0.00017769063384924272,
"loss": 1.03564453125,
"step": 89
},
{
"epoch": 0.8256880733944955,
"grad_norm": 0.0426209457218647,
"learning_rate": 0.0001770860847928664,
"loss": 1.0302734375,
"step": 90
},
{
"epoch": 0.8348623853211009,
"grad_norm": 0.06725878268480301,
"learning_rate": 0.00017647464596072437,
"loss": 1.03369140625,
"step": 91
},
{
"epoch": 0.8440366972477065,
"grad_norm": 0.04196714237332344,
"learning_rate": 0.00017585638014790242,
"loss": 1.03369140625,
"step": 92
},
{
"epoch": 0.8532110091743119,
"grad_norm": 0.04758419468998909,
"learning_rate": 0.0001752313508506208,
"loss": 1.0244140625,
"step": 93
},
{
"epoch": 0.8623853211009175,
"grad_norm": 0.042311377823352814,
"learning_rate": 0.0001745996222597131,
"loss": 1.03564453125,
"step": 94
},
{
"epoch": 0.8715596330275229,
"grad_norm": 0.0418376661837101,
"learning_rate": 0.0001739612592540338,
"loss": 1.0185546875,
"step": 95
},
{
"epoch": 0.8807339449541285,
"grad_norm": 0.04389483109116554,
"learning_rate": 0.00017331632739379533,
"loss": 1.0380859375,
"step": 96
},
{
"epoch": 0.8899082568807339,
"grad_norm": 0.043066758662462234,
"learning_rate": 0.0001726648929138348,
"loss": 1.01611328125,
"step": 97
},
{
"epoch": 0.8990825688073395,
"grad_norm": 0.04372232407331467,
"learning_rate": 0.00017200702271681178,
"loss": 1.017578125,
"step": 98
},
{
"epoch": 0.908256880733945,
"grad_norm": 0.041941553354263306,
"learning_rate": 0.0001713427843663374,
"loss": 0.99365234375,
"step": 99
},
{
"epoch": 0.9174311926605505,
"grad_norm": 0.042411062866449356,
"learning_rate": 0.00017067224608003528,
"loss": 0.98828125,
"step": 100
},
{
"epoch": 0.926605504587156,
"grad_norm": 0.04098072275519371,
"learning_rate": 0.0001699954767225359,
"loss": 1.00341796875,
"step": 101
},
{
"epoch": 0.9357798165137615,
"grad_norm": 0.047187983989715576,
"learning_rate": 0.0001693125457984039,
"loss": 0.994140625,
"step": 102
},
{
"epoch": 0.944954128440367,
"grad_norm": 0.049710918217897415,
"learning_rate": 0.00016862352344500003,
"loss": 1.0400390625,
"step": 103
},
{
"epoch": 0.9541284403669725,
"grad_norm": 0.04946531727910042,
"learning_rate": 0.00016792848042527798,
"loss": 1.0224609375,
"step": 104
},
{
"epoch": 0.963302752293578,
"grad_norm": 0.043604061007499695,
"learning_rate": 0.00016722748812051706,
"loss": 1.01220703125,
"step": 105
},
{
"epoch": 0.9724770642201835,
"grad_norm": 0.05599333345890045,
"learning_rate": 0.00016652061852299118,
"loss": 0.9208984375,
"step": 106
},
{
"epoch": 0.981651376146789,
"grad_norm": 0.04117029532790184,
"learning_rate": 0.00016580794422857523,
"loss": 0.98828125,
"step": 107
},
{
"epoch": 0.9908256880733946,
"grad_norm": 0.04220104217529297,
"learning_rate": 0.00016508953842928967,
"loss": 1.01513671875,
"step": 108
},
{
"epoch": 1.0,
"grad_norm": 0.04640773683786392,
"learning_rate": 0.00016436547490578327,
"loss": 0.9697265625,
"step": 109
},
{
"epoch": 1.0091743119266054,
"grad_norm": 0.04251675307750702,
"learning_rate": 0.00016363582801975614,
"loss": 0.95849609375,
"step": 110
},
{
"epoch": 1.018348623853211,
"grad_norm": 0.043099816888570786,
"learning_rate": 0.0001629006727063226,
"loss": 0.9697265625,
"step": 111
},
{
"epoch": 1.0275229357798166,
"grad_norm": 0.046582065522670746,
"learning_rate": 0.0001621600844663153,
"loss": 0.978515625,
"step": 112
},
{
"epoch": 1.036697247706422,
"grad_norm": 0.04335556551814079,
"learning_rate": 0.0001614141393585313,
"loss": 0.99755859375,
"step": 113
},
{
"epoch": 1.0458715596330275,
"grad_norm": 0.04112216457724571,
"learning_rate": 0.0001606629139919208,
"loss": 0.97607421875,
"step": 114
},
{
"epoch": 1.0550458715596331,
"grad_norm": 0.047019608318805695,
"learning_rate": 0.00015990648551771923,
"loss": 0.94482421875,
"step": 115
},
{
"epoch": 1.0642201834862386,
"grad_norm": 0.04267769679427147,
"learning_rate": 0.00015914493162152396,
"loss": 0.97509765625,
"step": 116
},
{
"epoch": 1.073394495412844,
"grad_norm": 0.044427670538425446,
"learning_rate": 0.00015837833051531575,
"loss": 0.97412109375,
"step": 117
},
{
"epoch": 1.0825688073394495,
"grad_norm": 0.04493756592273712,
"learning_rate": 0.00015760676092942661,
"loss": 0.94970703125,
"step": 118
},
{
"epoch": 1.091743119266055,
"grad_norm": 0.04097391292452812,
"learning_rate": 0.00015683030210445392,
"loss": 0.953125,
"step": 119
},
{
"epoch": 1.1009174311926606,
"grad_norm": 0.04028734564781189,
"learning_rate": 0.00015604903378312244,
"loss": 0.9287109375,
"step": 120
},
{
"epoch": 1.110091743119266,
"grad_norm": 0.043582089245319366,
"learning_rate": 0.00015526303620209462,
"loss": 0.9580078125,
"step": 121
},
{
"epoch": 1.1192660550458715,
"grad_norm": 0.04319828003644943,
"learning_rate": 0.00015447239008373052,
"loss": 0.96630859375,
"step": 122
},
{
"epoch": 1.1284403669724772,
"grad_norm": 0.04270276054739952,
"learning_rate": 0.00015367717662779733,
"loss": 0.953125,
"step": 123
},
{
"epoch": 1.1376146788990826,
"grad_norm": 0.042849309742450714,
"learning_rate": 0.00015287747750313007,
"loss": 0.9189453125,
"step": 124
},
{
"epoch": 1.146788990825688,
"grad_norm": 0.03792624920606613,
"learning_rate": 0.0001520733748392444,
"loss": 0.8414306640625,
"step": 125
},
{
"epoch": 1.1559633027522935,
"grad_norm": 0.044329360127449036,
"learning_rate": 0.00015126495121790178,
"loss": 0.984375,
"step": 126
},
{
"epoch": 1.165137614678899,
"grad_norm": 0.04070119559764862,
"learning_rate": 0.0001504522896646281,
"loss": 0.9658203125,
"step": 127
},
{
"epoch": 1.1743119266055047,
"grad_norm": 0.05330684036016464,
"learning_rate": 0.00014963547364018712,
"loss": 0.96240234375,
"step": 128
},
{
"epoch": 1.18348623853211,
"grad_norm": 0.044252488762140274,
"learning_rate": 0.00014881458703200896,
"loss": 0.90771484375,
"step": 129
},
{
"epoch": 1.1926605504587156,
"grad_norm": 0.04215725511312485,
"learning_rate": 0.00014798971414557488,
"loss": 0.93212890625,
"step": 130
},
{
"epoch": 1.2018348623853212,
"grad_norm": 0.050028230994939804,
"learning_rate": 0.00014716093969575886,
"loss": 0.947265625,
"step": 131
},
{
"epoch": 1.2110091743119267,
"grad_norm": 0.0427655465900898,
"learning_rate": 0.00014632834879812755,
"loss": 0.94482421875,
"step": 132
},
{
"epoch": 1.2110091743119267,
"eval_loss": 0.8802490234375,
"eval_runtime": 14.2508,
"eval_samples_per_second": 0.702,
"eval_steps_per_second": 0.14,
"step": 132
},
{
"epoch": 1.2201834862385321,
"grad_norm": 0.043557606637477875,
"learning_rate": 0.0001454920269601987,
"loss": 0.92626953125,
"step": 133
},
{
"epoch": 1.2293577981651376,
"grad_norm": 0.042206183075904846,
"learning_rate": 0.00014465206007265963,
"loss": 0.98095703125,
"step": 134
},
{
"epoch": 1.238532110091743,
"grad_norm": 0.04260166734457016,
"learning_rate": 0.0001438085344005461,
"loss": 0.92529296875,
"step": 135
},
{
"epoch": 1.2477064220183487,
"grad_norm": 0.04281189292669296,
"learning_rate": 0.00014296153657438306,
"loss": 0.93603515625,
"step": 136
},
{
"epoch": 1.2568807339449541,
"grad_norm": 0.04335340857505798,
"learning_rate": 0.00014211115358128722,
"loss": 0.9169921875,
"step": 137
},
{
"epoch": 1.2660550458715596,
"grad_norm": 0.04184405878186226,
"learning_rate": 0.00014125747275603383,
"loss": 0.90283203125,
"step": 138
},
{
"epoch": 1.2752293577981653,
"grad_norm": 0.042529623955488205,
"learning_rate": 0.00014040058177208726,
"loss": 0.91845703125,
"step": 139
},
{
"epoch": 1.2844036697247707,
"grad_norm": 0.043326206505298615,
"learning_rate": 0.00013954056863259677,
"loss": 0.8837890625,
"step": 140
},
{
"epoch": 1.2935779816513762,
"grad_norm": 0.044169191271066666,
"learning_rate": 0.00013867752166135867,
"loss": 0.906005859375,
"step": 141
},
{
"epoch": 1.3027522935779816,
"grad_norm": 0.04223386570811272,
"learning_rate": 0.00013781152949374526,
"loss": 0.9306640625,
"step": 142
},
{
"epoch": 1.311926605504587,
"grad_norm": 0.04900440201163292,
"learning_rate": 0.00013694268106760225,
"loss": 0.89404296875,
"step": 143
},
{
"epoch": 1.3211009174311927,
"grad_norm": 0.04095913842320442,
"learning_rate": 0.0001360710656141143,
"loss": 0.896484375,
"step": 144
},
{
"epoch": 1.3302752293577982,
"grad_norm": 0.04405977204442024,
"learning_rate": 0.0001351967726486415,
"loss": 0.8916015625,
"step": 145
},
{
"epoch": 1.3394495412844036,
"grad_norm": 0.047476377338171005,
"learning_rate": 0.0001343198919615256,
"loss": 0.86669921875,
"step": 146
},
{
"epoch": 1.3486238532110093,
"grad_norm": 0.043907489627599716,
"learning_rate": 0.0001334405136088688,
"loss": 0.8984375,
"step": 147
},
{
"epoch": 1.3577981651376148,
"grad_norm": 0.04861440137028694,
"learning_rate": 0.00013255872790328486,
"loss": 0.88134765625,
"step": 148
},
{
"epoch": 1.3669724770642202,
"grad_norm": 0.04644455388188362,
"learning_rate": 0.0001316746254046239,
"loss": 0.89111328125,
"step": 149
},
{
"epoch": 1.3761467889908257,
"grad_norm": 0.046188563108444214,
"learning_rate": 0.00013078829691067192,
"loss": 0.90625,
"step": 150
},
{
"epoch": 1.385321100917431,
"grad_norm": 0.04571763426065445,
"learning_rate": 0.00012989983344782575,
"loss": 0.90185546875,
"step": 151
},
{
"epoch": 1.3944954128440368,
"grad_norm": 0.04731906205415726,
"learning_rate": 0.00012900932626174467,
"loss": 0.8935546875,
"step": 152
},
{
"epoch": 1.4036697247706422,
"grad_norm": 0.047201234847307205,
"learning_rate": 0.0001281168668079794,
"loss": 0.8798828125,
"step": 153
},
{
"epoch": 1.4128440366972477,
"grad_norm": 0.045363202691078186,
"learning_rate": 0.00012722254674257956,
"loss": 0.888671875,
"step": 154
},
{
"epoch": 1.4220183486238533,
"grad_norm": 0.043558184057474136,
"learning_rate": 0.00012632645791268058,
"loss": 0.87158203125,
"step": 155
},
{
"epoch": 1.4311926605504588,
"grad_norm": 0.04717644304037094,
"learning_rate": 0.00012542869234707083,
"loss": 0.90771484375,
"step": 156
},
{
"epoch": 1.4403669724770642,
"grad_norm": 0.04378174617886543,
"learning_rate": 0.00012452934224674043,
"loss": 0.84326171875,
"step": 157
},
{
"epoch": 1.4495412844036697,
"grad_norm": 0.04657544195652008,
"learning_rate": 0.0001236284999754119,
"loss": 0.8466796875,
"step": 158
},
{
"epoch": 1.4587155963302751,
"grad_norm": 0.05050445348024368,
"learning_rate": 0.00012272625805005464,
"loss": 0.86474609375,
"step": 159
},
{
"epoch": 1.4678899082568808,
"grad_norm": 0.048389460891485214,
"learning_rate": 0.00012182270913138319,
"loss": 0.85595703125,
"step": 160
},
{
"epoch": 1.4770642201834863,
"grad_norm": 0.051234446465969086,
"learning_rate": 0.00012091794601434094,
"loss": 0.828857421875,
"step": 161
},
{
"epoch": 1.4862385321100917,
"grad_norm": 0.047934044152498245,
"learning_rate": 0.00012001206161857013,
"loss": 0.87841796875,
"step": 162
},
{
"epoch": 1.4954128440366974,
"grad_norm": 0.05460440739989281,
"learning_rate": 0.0001191051489788689,
"loss": 0.85546875,
"step": 163
},
{
"epoch": 1.5045871559633026,
"grad_norm": 0.05092587694525719,
"learning_rate": 0.0001181973012356366,
"loss": 0.85552978515625,
"step": 164
},
{
"epoch": 1.5137614678899083,
"grad_norm": 0.04715915769338608,
"learning_rate": 0.0001172886116253081,
"loss": 0.85693359375,
"step": 165
},
{
"epoch": 1.5229357798165137,
"grad_norm": 0.04569736868143082,
"learning_rate": 0.00011637917347077859,
"loss": 0.83984375,
"step": 166
},
{
"epoch": 1.5321100917431192,
"grad_norm": 0.04889059439301491,
"learning_rate": 0.00011546908017181906,
"loss": 0.87158203125,
"step": 167
},
{
"epoch": 1.5412844036697249,
"grad_norm": 0.0470917671918869,
"learning_rate": 0.00011455842519548416,
"loss": 0.87255859375,
"step": 168
},
{
"epoch": 1.5504587155963303,
"grad_norm": 0.04809945449233055,
"learning_rate": 0.00011364730206651318,
"loss": 0.85400390625,
"step": 169
},
{
"epoch": 1.5596330275229358,
"grad_norm": 0.046025026589632034,
"learning_rate": 0.00011273580435772486,
"loss": 0.85546875,
"step": 170
},
{
"epoch": 1.5688073394495414,
"grad_norm": 0.05134977772831917,
"learning_rate": 0.00011182402568040751,
"loss": 0.876953125,
"step": 171
},
{
"epoch": 1.5779816513761467,
"grad_norm": 0.047295693308115005,
"learning_rate": 0.00011091205967470502,
"loss": 0.833984375,
"step": 172
},
{
"epoch": 1.5871559633027523,
"grad_norm": 0.04927787184715271,
"learning_rate": 0.00011000000000000002,
"loss": 0.873046875,
"step": 173
},
{
"epoch": 1.5963302752293578,
"grad_norm": 0.04760082811117172,
"learning_rate": 0.00010908794032529504,
"loss": 0.8583984375,
"step": 174
},
{
"epoch": 1.6055045871559632,
"grad_norm": 0.04950103536248207,
"learning_rate": 0.00010817597431959252,
"loss": 0.8310546875,
"step": 175
},
{
"epoch": 1.614678899082569,
"grad_norm": 0.05043569207191467,
"learning_rate": 0.00010726419564227519,
"loss": 0.85595703125,
"step": 176
},
{
"epoch": 1.6238532110091743,
"grad_norm": 0.06437493115663528,
"learning_rate": 0.00010635269793348683,
"loss": 0.8251953125,
"step": 177
},
{
"epoch": 1.6330275229357798,
"grad_norm": 0.048829928040504456,
"learning_rate": 0.00010544157480451587,
"loss": 0.84423828125,
"step": 178
},
{
"epoch": 1.6422018348623855,
"grad_norm": 0.04708709940314293,
"learning_rate": 0.00010453091982818098,
"loss": 0.89404296875,
"step": 179
},
{
"epoch": 1.6513761467889907,
"grad_norm": 0.04794275760650635,
"learning_rate": 0.00010362082652922145,
"loss": 0.82568359375,
"step": 180
},
{
"epoch": 1.6605504587155964,
"grad_norm": 0.052562177181243896,
"learning_rate": 0.00010271138837469194,
"loss": 0.8466796875,
"step": 181
},
{
"epoch": 1.6697247706422018,
"grad_norm": 0.14999158680438995,
"learning_rate": 0.00010180269876436344,
"loss": 0.8388671875,
"step": 182
},
{
"epoch": 1.6788990825688073,
"grad_norm": 0.04777876287698746,
"learning_rate": 0.00010089485102113112,
"loss": 0.82861328125,
"step": 183
},
{
"epoch": 1.688073394495413,
"grad_norm": 0.0467647947371006,
"learning_rate": 9.998793838142988e-05,
"loss": 0.7744140625,
"step": 184
},
{
"epoch": 1.6972477064220184,
"grad_norm": 0.050435010343790054,
"learning_rate": 9.908205398565907e-05,
"loss": 0.85693359375,
"step": 185
},
{
"epoch": 1.7064220183486238,
"grad_norm": 0.04828406870365143,
"learning_rate": 9.817729086861682e-05,
"loss": 0.83447265625,
"step": 186
},
{
"epoch": 1.7155963302752295,
"grad_norm": 0.04797462373971939,
"learning_rate": 9.727374194994536e-05,
"loss": 0.85986328125,
"step": 187
},
{
"epoch": 1.7247706422018347,
"grad_norm": 0.09186355024576187,
"learning_rate": 9.637150002458813e-05,
"loss": 0.810546875,
"step": 188
},
{
"epoch": 1.7339449541284404,
"grad_norm": 0.05213315784931183,
"learning_rate": 9.54706577532596e-05,
"loss": 0.845703125,
"step": 189
},
{
"epoch": 1.7431192660550459,
"grad_norm": 0.050209175795316696,
"learning_rate": 9.457130765292921e-05,
"loss": 0.85498046875,
"step": 190
},
{
"epoch": 1.7522935779816513,
"grad_norm": 0.052893489599227905,
"learning_rate": 9.367354208731945e-05,
"loss": 0.81884765625,
"step": 191
},
{
"epoch": 1.761467889908257,
"grad_norm": 0.050216108560562134,
"learning_rate": 9.277745325742044e-05,
"loss": 0.84619140625,
"step": 192
},
{
"epoch": 1.7706422018348624,
"grad_norm": 0.050006598234176636,
"learning_rate": 9.188313319202058e-05,
"loss": 0.8017578125,
"step": 193
},
{
"epoch": 1.7798165137614679,
"grad_norm": 0.04936378449201584,
"learning_rate": 9.099067373825534e-05,
"loss": 0.83056640625,
"step": 194
},
{
"epoch": 1.7889908256880735,
"grad_norm": 0.05004863813519478,
"learning_rate": 9.010016655217431e-05,
"loss": 0.83447265625,
"step": 195
},
{
"epoch": 1.7981651376146788,
"grad_norm": 0.04911724850535393,
"learning_rate": 8.921170308932811e-05,
"loss": 0.8203125,
"step": 196
},
{
"epoch": 1.8073394495412844,
"grad_norm": 0.05456596240401268,
"learning_rate": 8.832537459537613e-05,
"loss": 0.86474609375,
"step": 197
},
{
"epoch": 1.8073394495412844,
"eval_loss": 0.861328125,
"eval_runtime": 14.241,
"eval_samples_per_second": 0.702,
"eval_steps_per_second": 0.14,
"step": 197
},
{
"epoch": 1.81651376146789,
"grad_norm": 0.04900031164288521,
"learning_rate": 8.744127209671516e-05,
"loss": 0.84716796875,
"step": 198
},
{
"epoch": 1.8256880733944953,
"grad_norm": 0.04944784939289093,
"learning_rate": 8.655948639113122e-05,
"loss": 0.84375,
"step": 199
},
{
"epoch": 1.834862385321101,
"grad_norm": 0.04886430874466896,
"learning_rate": 8.568010803847445e-05,
"loss": 0.84765625,
"step": 200
},
{
"epoch": 1.8440366972477065,
"grad_norm": 0.05156690254807472,
"learning_rate": 8.480322735135853e-05,
"loss": 0.841796875,
"step": 201
},
{
"epoch": 1.853211009174312,
"grad_norm": 0.05328977480530739,
"learning_rate": 8.392893438588572e-05,
"loss": 0.8167724609375,
"step": 202
},
{
"epoch": 1.8623853211009176,
"grad_norm": 0.053657256066799164,
"learning_rate": 8.305731893239779e-05,
"loss": 0.85302734375,
"step": 203
},
{
"epoch": 1.8715596330275228,
"grad_norm": 0.09988856315612793,
"learning_rate": 8.218847050625476e-05,
"loss": 0.8369140625,
"step": 204
},
{
"epoch": 1.8807339449541285,
"grad_norm": 0.05004033446311951,
"learning_rate": 8.132247833864137e-05,
"loss": 0.85595703125,
"step": 205
},
{
"epoch": 1.889908256880734,
"grad_norm": 0.05268581211566925,
"learning_rate": 8.045943136740323e-05,
"loss": 0.830078125,
"step": 206
},
{
"epoch": 1.8990825688073394,
"grad_norm": 0.05199107900261879,
"learning_rate": 7.959941822791277e-05,
"loss": 0.8291015625,
"step": 207
},
{
"epoch": 1.908256880733945,
"grad_norm": 0.04928615316748619,
"learning_rate": 7.87425272439662e-05,
"loss": 0.814453125,
"step": 208
},
{
"epoch": 1.9174311926605505,
"grad_norm": 0.05711093917489052,
"learning_rate": 7.788884641871281e-05,
"loss": 0.80908203125,
"step": 209
},
{
"epoch": 1.926605504587156,
"grad_norm": 0.053956203162670135,
"learning_rate": 7.703846342561696e-05,
"loss": 0.8310546875,
"step": 210
},
{
"epoch": 1.9357798165137616,
"grad_norm": 0.05253409966826439,
"learning_rate": 7.619146559945389e-05,
"loss": 0.8154296875,
"step": 211
},
{
"epoch": 1.9449541284403669,
"grad_norm": 0.05143842473626137,
"learning_rate": 7.53479399273404e-05,
"loss": 0.86767578125,
"step": 212
},
{
"epoch": 1.9541284403669725,
"grad_norm": 0.05183498561382294,
"learning_rate": 7.450797303980134e-05,
"loss": 0.83984375,
"step": 213
},
{
"epoch": 1.963302752293578,
"grad_norm": 0.05124508589506149,
"learning_rate": 7.367165120187252e-05,
"loss": 0.83203125,
"step": 214
},
{
"epoch": 1.9724770642201834,
"grad_norm": 0.0484778992831707,
"learning_rate": 7.283906030424117e-05,
"loss": 0.760009765625,
"step": 215
},
{
"epoch": 1.981651376146789,
"grad_norm": 0.052525345236063004,
"learning_rate": 7.201028585442513e-05,
"loss": 0.8046875,
"step": 216
},
{
"epoch": 1.9908256880733946,
"grad_norm": 0.05047764629125595,
"learning_rate": 7.118541296799104e-05,
"loss": 0.8359375,
"step": 217
},
{
"epoch": 2.0,
"grad_norm": 0.05746862664818764,
"learning_rate": 7.036452635981292e-05,
"loss": 0.7060546875,
"step": 218
},
{
"epoch": 2.0091743119266057,
"grad_norm": 0.0637534111738205,
"learning_rate": 6.954771033537195e-05,
"loss": 0.70068359375,
"step": 219
},
{
"epoch": 2.018348623853211,
"grad_norm": 0.06224729120731354,
"learning_rate": 6.873504878209827e-05,
"loss": 0.79931640625,
"step": 220
},
{
"epoch": 2.0275229357798166,
"grad_norm": 0.06609112024307251,
"learning_rate": 6.792662516075561e-05,
"loss": 0.8037109375,
"step": 221
},
{
"epoch": 2.036697247706422,
"grad_norm": 0.05905923247337341,
"learning_rate": 6.712252249686996e-05,
"loss": 0.82373046875,
"step": 222
},
{
"epoch": 2.0458715596330275,
"grad_norm": 0.05165673419833183,
"learning_rate": 6.632282337220273e-05,
"loss": 0.80908203125,
"step": 223
},
{
"epoch": 2.055045871559633,
"grad_norm": 0.05654323101043701,
"learning_rate": 6.552760991626949e-05,
"loss": 0.7763671875,
"step": 224
},
{
"epoch": 2.0642201834862384,
"grad_norm": 0.052633751183748245,
"learning_rate": 6.473696379790541e-05,
"loss": 0.8056640625,
"step": 225
},
{
"epoch": 2.073394495412844,
"grad_norm": 0.05082765966653824,
"learning_rate": 6.395096621687761e-05,
"loss": 0.80615234375,
"step": 226
},
{
"epoch": 2.0825688073394497,
"grad_norm": 0.06019085273146629,
"learning_rate": 6.316969789554607e-05,
"loss": 0.78271484375,
"step": 227
},
{
"epoch": 2.091743119266055,
"grad_norm": 0.05313262343406677,
"learning_rate": 6.239323907057342e-05,
"loss": 0.7822265625,
"step": 228
},
{
"epoch": 2.1009174311926606,
"grad_norm": 0.05167046934366226,
"learning_rate": 6.162166948468425e-05,
"loss": 0.767578125,
"step": 229
},
{
"epoch": 2.1100917431192663,
"grad_norm": 0.05418609082698822,
"learning_rate": 6.085506837847611e-05,
"loss": 0.7939453125,
"step": 230
},
{
"epoch": 2.1192660550458715,
"grad_norm": 0.05151104927062988,
"learning_rate": 6.0093514482280775e-05,
"loss": 0.79736328125,
"step": 231
},
{
"epoch": 2.128440366972477,
"grad_norm": 0.05162815377116203,
"learning_rate": 5.933708600807924e-05,
"loss": 0.7890625,
"step": 232
},
{
"epoch": 2.1376146788990824,
"grad_norm": 0.05146823078393936,
"learning_rate": 5.858586064146867e-05,
"loss": 0.7587890625,
"step": 233
},
{
"epoch": 2.146788990825688,
"grad_norm": 0.046313222497701645,
"learning_rate": 5.783991553368473e-05,
"loss": 0.6990966796875,
"step": 234
},
{
"epoch": 2.1559633027522938,
"grad_norm": 0.051184896379709244,
"learning_rate": 5.709932729367744e-05,
"loss": 0.81982421875,
"step": 235
},
{
"epoch": 2.165137614678899,
"grad_norm": 0.05151189863681793,
"learning_rate": 5.6364171980243886e-05,
"loss": 0.80517578125,
"step": 236
},
{
"epoch": 2.1743119266055047,
"grad_norm": 0.051978498697280884,
"learning_rate": 5.563452509421674e-05,
"loss": 0.7998046875,
"step": 237
},
{
"epoch": 2.18348623853211,
"grad_norm": 0.05189194157719612,
"learning_rate": 5.491046157071034e-05,
"loss": 0.75146484375,
"step": 238
},
{
"epoch": 2.1926605504587156,
"grad_norm": 0.0518280528485775,
"learning_rate": 5.419205577142476e-05,
"loss": 0.77001953125,
"step": 239
},
{
"epoch": 2.2018348623853212,
"grad_norm": 0.05103209614753723,
"learning_rate": 5.347938147700885e-05,
"loss": 0.7841796875,
"step": 240
},
{
"epoch": 2.2110091743119265,
"grad_norm": 0.07511334121227264,
"learning_rate": 5.2772511879482976e-05,
"loss": 0.7802734375,
"step": 241
},
{
"epoch": 2.220183486238532,
"grad_norm": 0.049933455884456635,
"learning_rate": 5.207151957472205e-05,
"loss": 0.767578125,
"step": 242
},
{
"epoch": 2.229357798165138,
"grad_norm": 0.05338992550969124,
"learning_rate": 5.137647655500002e-05,
"loss": 0.81396484375,
"step": 243
},
{
"epoch": 2.238532110091743,
"grad_norm": 0.0521296001970768,
"learning_rate": 5.068745420159615e-05,
"loss": 0.76513671875,
"step": 244
},
{
"epoch": 2.2477064220183487,
"grad_norm": 0.0525696724653244,
"learning_rate": 5.000452327746413e-05,
"loss": 0.771484375,
"step": 245
},
{
"epoch": 2.2568807339449544,
"grad_norm": 0.05091803893446922,
"learning_rate": 4.9327753919964744e-05,
"loss": 0.75390625,
"step": 246
},
{
"epoch": 2.2660550458715596,
"grad_norm": 0.05207496136426926,
"learning_rate": 4.865721563366263e-05,
"loss": 0.7421875,
"step": 247
},
{
"epoch": 2.2752293577981653,
"grad_norm": 0.05462362617254257,
"learning_rate": 4.79929772831882e-05,
"loss": 0.76416015625,
"step": 248
},
{
"epoch": 2.2844036697247705,
"grad_norm": 0.051704030483961105,
"learning_rate": 4.733510708616524e-05,
"loss": 0.73388671875,
"step": 249
},
{
"epoch": 2.293577981651376,
"grad_norm": 0.05338239669799805,
"learning_rate": 4.6683672606204674e-05,
"loss": 0.739990234375,
"step": 250
},
{
"epoch": 2.302752293577982,
"grad_norm": 0.05035558342933655,
"learning_rate": 4.603874074596622e-05,
"loss": 0.7724609375,
"step": 251
},
{
"epoch": 2.311926605504587,
"grad_norm": 0.06754462420940399,
"learning_rate": 4.540037774028691e-05,
"loss": 0.73876953125,
"step": 252
},
{
"epoch": 2.3211009174311927,
"grad_norm": 0.05083058029413223,
"learning_rate": 4.4768649149379225e-05,
"loss": 0.74609375,
"step": 253
},
{
"epoch": 2.330275229357798,
"grad_norm": 0.06331224739551544,
"learning_rate": 4.4143619852097595e-05,
"loss": 0.73486328125,
"step": 254
},
{
"epoch": 2.3394495412844036,
"grad_norm": 0.050961848348379135,
"learning_rate": 4.352535403927567e-05,
"loss": 0.712890625,
"step": 255
},
{
"epoch": 2.3486238532110093,
"grad_norm": 0.0516008697450161,
"learning_rate": 4.291391520713364e-05,
"loss": 0.73828125,
"step": 256
},
{
"epoch": 2.3577981651376145,
"grad_norm": 0.05565928295254707,
"learning_rate": 4.2309366150757294e-05,
"loss": 0.72900390625,
"step": 257
},
{
"epoch": 2.36697247706422,
"grad_norm": 0.054245349019765854,
"learning_rate": 4.1711768957648825e-05,
"loss": 0.73974609375,
"step": 258
},
{
"epoch": 2.376146788990826,
"grad_norm": 0.06098080426454544,
"learning_rate": 4.112118500135047e-05,
"loss": 0.75,
"step": 259
},
{
"epoch": 2.385321100917431,
"grad_norm": 0.05248161405324936,
"learning_rate": 4.053767493514142e-05,
"loss": 0.74951171875,
"step": 260
},
{
"epoch": 2.3944954128440368,
"grad_norm": 0.05365797132253647,
"learning_rate": 3.996129868580866e-05,
"loss": 0.7412109375,
"step": 261
},
{
"epoch": 2.4036697247706424,
"grad_norm": 0.05491862818598747,
"learning_rate": 3.939211544749246e-05,
"loss": 0.72119140625,
"step": 262
},
{
"epoch": 2.4036697247706424,
"eval_loss": 0.854248046875,
"eval_runtime": 14.2343,
"eval_samples_per_second": 0.703,
"eval_steps_per_second": 0.141,
"step": 262
},
{
"epoch": 2.4128440366972477,
"grad_norm": 0.054970841854810715,
"learning_rate": 3.8830183675607146e-05,
"loss": 0.73388671875,
"step": 263
},
{
"epoch": 2.4220183486238533,
"grad_norm": 0.055204931646585464,
"learning_rate": 3.827556108083766e-05,
"loss": 0.72314453125,
"step": 264
},
{
"epoch": 2.4311926605504586,
"grad_norm": 0.056462518870830536,
"learning_rate": 3.7728304623212684e-05,
"loss": 0.74951171875,
"step": 265
},
{
"epoch": 2.4403669724770642,
"grad_norm": 0.054016437381505966,
"learning_rate": 3.718847050625475e-05,
"loss": 0.6943359375,
"step": 266
},
{
"epoch": 2.44954128440367,
"grad_norm": 0.10177717357873917,
"learning_rate": 3.6656114171208155e-05,
"loss": 0.69580078125,
"step": 267
},
{
"epoch": 2.458715596330275,
"grad_norm": 0.05562096834182739,
"learning_rate": 3.613129029134516e-05,
"loss": 0.70849609375,
"step": 268
},
{
"epoch": 2.467889908256881,
"grad_norm": 0.05390914902091026,
"learning_rate": 3.561405276635085e-05,
"loss": 0.70703125,
"step": 269
},
{
"epoch": 2.477064220183486,
"grad_norm": 0.055356986820697784,
"learning_rate": 3.5104454716787866e-05,
"loss": 0.674072265625,
"step": 270
},
{
"epoch": 2.4862385321100917,
"grad_norm": 0.053944528102874756,
"learning_rate": 3.4602548478640616e-05,
"loss": 0.72314453125,
"step": 271
},
{
"epoch": 2.4954128440366974,
"grad_norm": 0.054691873490810394,
"learning_rate": 3.4108385597940614e-05,
"loss": 0.705078125,
"step": 272
},
{
"epoch": 2.5045871559633026,
"grad_norm": 0.05734097957611084,
"learning_rate": 3.3622016825472416e-05,
"loss": 0.693939208984375,
"step": 273
},
{
"epoch": 2.5137614678899083,
"grad_norm": 0.056674521416425705,
"learning_rate": 3.31434921115617e-05,
"loss": 0.70849609375,
"step": 274
},
{
"epoch": 2.522935779816514,
"grad_norm": 0.0582982674241066,
"learning_rate": 3.267286060094521e-05,
"loss": 0.6943359375,
"step": 275
},
{
"epoch": 2.532110091743119,
"grad_norm": 0.057676538825035095,
"learning_rate": 3.221017062772363e-05,
"loss": 0.72119140625,
"step": 276
},
{
"epoch": 2.541284403669725,
"grad_norm": 0.05714472383260727,
"learning_rate": 3.17554697103976e-05,
"loss": 0.7275390625,
"step": 277
},
{
"epoch": 2.5504587155963305,
"grad_norm": 0.05490856245160103,
"learning_rate": 3.130880454698761e-05,
"loss": 0.705322265625,
"step": 278
},
{
"epoch": 2.5596330275229358,
"grad_norm": 0.053883641958236694,
"learning_rate": 3.087022101023802e-05,
"loss": 0.7099609375,
"step": 279
},
{
"epoch": 2.5688073394495414,
"grad_norm": 0.056989215314388275,
"learning_rate": 3.0439764142905935e-05,
"loss": 0.72802734375,
"step": 280
},
{
"epoch": 2.5779816513761467,
"grad_norm": 0.05444977059960365,
"learning_rate": 3.00174781531353e-05,
"loss": 0.69140625,
"step": 281
},
{
"epoch": 2.5871559633027523,
"grad_norm": 0.06884799897670746,
"learning_rate": 2.9603406409916694e-05,
"loss": 0.72021484375,
"step": 282
},
{
"epoch": 2.5963302752293576,
"grad_norm": 0.05767287313938141,
"learning_rate": 2.919759143863326e-05,
"loss": 0.70849609375,
"step": 283
},
{
"epoch": 2.6055045871559632,
"grad_norm": 0.05668262764811516,
"learning_rate": 2.8800074916693433e-05,
"loss": 0.6875,
"step": 284
},
{
"epoch": 2.614678899082569,
"grad_norm": 0.05650593340396881,
"learning_rate": 2.8410897669250548e-05,
"loss": 0.712890625,
"step": 285
},
{
"epoch": 2.623853211009174,
"grad_norm": 0.056592393666505814,
"learning_rate": 2.8030099665010122e-05,
"loss": 0.6845703125,
"step": 286
},
{
"epoch": 2.63302752293578,
"grad_norm": 0.0557887926697731,
"learning_rate": 2.765772001212509e-05,
"loss": 0.69921875,
"step": 287
},
{
"epoch": 2.6422018348623855,
"grad_norm": 0.055601976811885834,
"learning_rate": 2.7293796954179253e-05,
"loss": 0.75,
"step": 288
},
{
"epoch": 2.6513761467889907,
"grad_norm": 0.05699892342090607,
"learning_rate": 2.6938367866259774e-05,
"loss": 0.68994140625,
"step": 289
},
{
"epoch": 2.6605504587155964,
"grad_norm": 0.05384403467178345,
"learning_rate": 2.6591469251118628e-05,
"loss": 0.70947265625,
"step": 290
},
{
"epoch": 2.669724770642202,
"grad_norm": 0.05626530572772026,
"learning_rate": 2.6253136735423894e-05,
"loss": 0.70166015625,
"step": 291
},
{
"epoch": 2.6788990825688073,
"grad_norm": 0.055328190326690674,
"learning_rate": 2.592340506610067e-05,
"loss": 0.68896484375,
"step": 292
},
{
"epoch": 2.688073394495413,
"grad_norm": 0.052280981093645096,
"learning_rate": 2.5602308106762773e-05,
"loss": 0.64404296875,
"step": 293
},
{
"epoch": 2.6972477064220186,
"grad_norm": 0.0791119635105133,
"learning_rate": 2.5289878834234703e-05,
"loss": 0.7177734375,
"step": 294
},
{
"epoch": 2.706422018348624,
"grad_norm": 0.05575219914317131,
"learning_rate": 2.4986149335165166e-05,
"loss": 0.7001953125,
"step": 295
},
{
"epoch": 2.7155963302752295,
"grad_norm": 0.05544380843639374,
"learning_rate": 2.4691150802731462e-05,
"loss": 0.72265625,
"step": 296
},
{
"epoch": 2.7247706422018347,
"grad_norm": 0.05615422502160072,
"learning_rate": 2.4404913533436186e-05,
"loss": 0.6787109375,
"step": 297
},
{
"epoch": 2.7339449541284404,
"grad_norm": 0.05648210644721985,
"learning_rate": 2.412746692399561e-05,
"loss": 0.70703125,
"step": 298
},
{
"epoch": 2.7431192660550456,
"grad_norm": 0.057345759123563766,
"learning_rate": 2.3858839468320684e-05,
"loss": 0.71728515625,
"step": 299
},
{
"epoch": 2.7522935779816513,
"grad_norm": 0.05655364319682121,
"learning_rate": 2.359905875459068e-05,
"loss": 0.685546875,
"step": 300
},
{
"epoch": 2.761467889908257,
"grad_norm": 0.05608190968632698,
"learning_rate": 2.334815146241989e-05,
"loss": 0.71533203125,
"step": 301
},
{
"epoch": 2.770642201834862,
"grad_norm": 0.05728856101632118,
"learning_rate": 2.3106143360117582e-05,
"loss": 0.671875,
"step": 302
},
{
"epoch": 2.779816513761468,
"grad_norm": 0.056054387241601944,
"learning_rate": 2.2873059302041626e-05,
"loss": 0.7001953125,
"step": 303
},
{
"epoch": 2.7889908256880735,
"grad_norm": 0.05643802881240845,
"learning_rate": 2.2648923226045855e-05,
"loss": 0.69921875,
"step": 304
},
{
"epoch": 2.7981651376146788,
"grad_norm": 0.056719593703746796,
"learning_rate": 2.2433758151021757e-05,
"loss": 0.68994140625,
"step": 305
},
{
"epoch": 2.8073394495412844,
"grad_norm": 0.057057030498981476,
"learning_rate": 2.2227586174534346e-05,
"loss": 0.73291015625,
"step": 306
},
{
"epoch": 2.81651376146789,
"grad_norm": 0.05694695934653282,
"learning_rate": 2.203042847055271e-05,
"loss": 0.716796875,
"step": 307
},
{
"epoch": 2.8256880733944953,
"grad_norm": 0.055483054369688034,
"learning_rate": 2.18423052872755e-05,
"loss": 0.71728515625,
"step": 308
},
{
"epoch": 2.834862385321101,
"grad_norm": 0.05642176792025566,
"learning_rate": 2.1663235945051392e-05,
"loss": 0.72021484375,
"step": 309
},
{
"epoch": 2.8440366972477067,
"grad_norm": 0.05773378536105156,
"learning_rate": 2.149323883439487e-05,
"loss": 0.70947265625,
"step": 310
},
{
"epoch": 2.853211009174312,
"grad_norm": 0.05909159779548645,
"learning_rate": 2.133233141409752e-05,
"loss": 0.6746826171875,
"step": 311
},
{
"epoch": 2.8623853211009176,
"grad_norm": 0.057984307408332825,
"learning_rate": 2.1180530209435036e-05,
"loss": 0.720703125,
"step": 312
},
{
"epoch": 2.871559633027523,
"grad_norm": 0.05757308378815651,
"learning_rate": 2.1037850810469978e-05,
"loss": 0.70947265625,
"step": 313
},
{
"epoch": 2.8807339449541285,
"grad_norm": 0.0598280169069767,
"learning_rate": 2.090430787045078e-05,
"loss": 0.73046875,
"step": 314
},
{
"epoch": 2.8899082568807337,
"grad_norm": 0.059042032808065414,
"learning_rate": 2.0779915104306773e-05,
"loss": 0.70654296875,
"step": 315
},
{
"epoch": 2.8990825688073394,
"grad_norm": 0.05784536525607109,
"learning_rate": 2.0664685287239648e-05,
"loss": 0.7060546875,
"step": 316
},
{
"epoch": 2.908256880733945,
"grad_norm": 0.05720442906022072,
"learning_rate": 2.055863025341151e-05,
"loss": 0.6923828125,
"step": 317
},
{
"epoch": 2.9174311926605503,
"grad_norm": 0.05674874037504196,
"learning_rate": 2.046176089472944e-05,
"loss": 0.6845703125,
"step": 318
},
{
"epoch": 2.926605504587156,
"grad_norm": 0.05587884783744812,
"learning_rate": 2.03740871597269e-05,
"loss": 0.70751953125,
"step": 319
},
{
"epoch": 2.9357798165137616,
"grad_norm": 0.056729506701231,
"learning_rate": 2.029561805254203e-05,
"loss": 0.69482421875,
"step": 320
},
{
"epoch": 2.944954128440367,
"grad_norm": 0.06823433935642242,
"learning_rate": 2.0226361631992915e-05,
"loss": 0.74951171875,
"step": 321
},
{
"epoch": 2.9541284403669725,
"grad_norm": 0.061061933636665344,
"learning_rate": 2.0166325010749924e-05,
"loss": 0.71728515625,
"step": 322
},
{
"epoch": 2.963302752293578,
"grad_norm": 0.05851565673947334,
"learning_rate": 2.0115514354605253e-05,
"loss": 0.7138671875,
"step": 323
},
{
"epoch": 2.9724770642201834,
"grad_norm": 0.05435410141944885,
"learning_rate": 2.0073934881839693e-05,
"loss": 0.655517578125,
"step": 324
},
{
"epoch": 2.981651376146789,
"grad_norm": 0.05809434503316879,
"learning_rate": 2.00415908626867e-05,
"loss": 0.685546875,
"step": 325
},
{
"epoch": 2.9908256880733948,
"grad_norm": 0.05944964662194252,
"learning_rate": 2.001848561889385e-05,
"loss": 0.7177734375,
"step": 326
},
{
"epoch": 3.0,
"grad_norm": 0.0742282122373581,
"learning_rate": 2.0004621523381677e-05,
"loss": 0.5322265625,
"step": 327
},
{
"epoch": 3.0,
"eval_loss": 0.860107421875,
"eval_runtime": 14.2601,
"eval_samples_per_second": 0.701,
"eval_steps_per_second": 0.14,
"step": 327
}
],
"logging_steps": 1.0,
"max_steps": 327,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"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": 2.7076181610234642e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}