adaption_final_final / trainer_state.json
huyxdang's picture
Add 11 files
d1b2ed1 verified
Raw
History Blame Contribute Delete
76.9 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 83,
"global_step": 418,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004784688995215311,
"grad_norm": 0.48090460896492004,
"learning_rate": 0.0,
"loss": 1.5244140625,
"step": 1
},
{
"epoch": 0.009569377990430622,
"grad_norm": 0.5485076904296875,
"learning_rate": 4.7619047619047615e-06,
"loss": 1.5478515625,
"step": 2
},
{
"epoch": 0.014354066985645933,
"grad_norm": 0.6669546365737915,
"learning_rate": 9.523809523809523e-06,
"loss": 1.62890625,
"step": 3
},
{
"epoch": 0.019138755980861243,
"grad_norm": 0.4594496190547943,
"learning_rate": 1.4285714285714285e-05,
"loss": 1.4521484375,
"step": 4
},
{
"epoch": 0.023923444976076555,
"grad_norm": 0.48580074310302734,
"learning_rate": 1.9047619047619046e-05,
"loss": 1.4169921875,
"step": 5
},
{
"epoch": 0.028708133971291867,
"grad_norm": 0.43042290210723877,
"learning_rate": 2.380952380952381e-05,
"loss": 1.4130859375,
"step": 6
},
{
"epoch": 0.03349282296650718,
"grad_norm": 0.44704583287239075,
"learning_rate": 2.857142857142857e-05,
"loss": 1.3681640625,
"step": 7
},
{
"epoch": 0.03827751196172249,
"grad_norm": 0.4594089388847351,
"learning_rate": 3.3333333333333335e-05,
"loss": 1.431640625,
"step": 8
},
{
"epoch": 0.0430622009569378,
"grad_norm": 0.34071800112724304,
"learning_rate": 3.809523809523809e-05,
"loss": 1.34375,
"step": 9
},
{
"epoch": 0.04784688995215311,
"grad_norm": 0.404588907957077,
"learning_rate": 4.2857142857142856e-05,
"loss": 1.3408203125,
"step": 10
},
{
"epoch": 0.05263157894736842,
"grad_norm": 0.3034615218639374,
"learning_rate": 4.761904761904762e-05,
"loss": 1.1591796875,
"step": 11
},
{
"epoch": 0.05741626794258373,
"grad_norm": 0.32653138041496277,
"learning_rate": 5.2380952380952384e-05,
"loss": 1.34765625,
"step": 12
},
{
"epoch": 0.06220095693779904,
"grad_norm": 0.22854778170585632,
"learning_rate": 5.714285714285714e-05,
"loss": 1.244140625,
"step": 13
},
{
"epoch": 0.06698564593301436,
"grad_norm": 0.18688103556632996,
"learning_rate": 6.19047619047619e-05,
"loss": 1.25439453125,
"step": 14
},
{
"epoch": 0.07177033492822966,
"grad_norm": 0.20101380348205566,
"learning_rate": 6.666666666666667e-05,
"loss": 1.283203125,
"step": 15
},
{
"epoch": 0.07655502392344497,
"grad_norm": 0.25939851999282837,
"learning_rate": 7.142857142857143e-05,
"loss": 1.29443359375,
"step": 16
},
{
"epoch": 0.08133971291866028,
"grad_norm": 0.3485074043273926,
"learning_rate": 7.619047619047618e-05,
"loss": 1.2841796875,
"step": 17
},
{
"epoch": 0.0861244019138756,
"grad_norm": 0.3645748198032379,
"learning_rate": 8.095238095238096e-05,
"loss": 1.2626953125,
"step": 18
},
{
"epoch": 0.09090909090909091,
"grad_norm": 0.2573246657848358,
"learning_rate": 8.571428571428571e-05,
"loss": 1.04833984375,
"step": 19
},
{
"epoch": 0.09569377990430622,
"grad_norm": 0.18779775500297546,
"learning_rate": 9.047619047619048e-05,
"loss": 1.1328125,
"step": 20
},
{
"epoch": 0.10047846889952153,
"grad_norm": 0.24399079382419586,
"learning_rate": 9.523809523809524e-05,
"loss": 1.1494140625,
"step": 21
},
{
"epoch": 0.10526315789473684,
"grad_norm": 0.2191498726606369,
"learning_rate": 0.0001,
"loss": 1.11572265625,
"step": 22
},
{
"epoch": 0.11004784688995216,
"grad_norm": 0.19746914505958557,
"learning_rate": 9.999859103896227e-05,
"loss": 1.13623046875,
"step": 23
},
{
"epoch": 0.11483253588516747,
"grad_norm": 0.168879434466362,
"learning_rate": 9.999436424407889e-05,
"loss": 1.17529296875,
"step": 24
},
{
"epoch": 0.11961722488038277,
"grad_norm": 0.16985730826854706,
"learning_rate": 9.998731988003383e-05,
"loss": 1.07470703125,
"step": 25
},
{
"epoch": 0.12440191387559808,
"grad_norm": 0.18533837795257568,
"learning_rate": 9.997745838794862e-05,
"loss": 1.09423828125,
"step": 26
},
{
"epoch": 0.1291866028708134,
"grad_norm": 0.16401302814483643,
"learning_rate": 9.996478038535475e-05,
"loss": 1.0947265625,
"step": 27
},
{
"epoch": 0.1339712918660287,
"grad_norm": 0.14787301421165466,
"learning_rate": 9.994928666615494e-05,
"loss": 1.11181640625,
"step": 28
},
{
"epoch": 0.13875598086124402,
"grad_norm": 0.12298636883497238,
"learning_rate": 9.993097820057351e-05,
"loss": 1.03515625,
"step": 29
},
{
"epoch": 0.14354066985645933,
"grad_norm": 0.13817098736763,
"learning_rate": 9.990985613509554e-05,
"loss": 1.0498046875,
"step": 30
},
{
"epoch": 0.14832535885167464,
"grad_norm": 0.10902783274650574,
"learning_rate": 9.988592179239512e-05,
"loss": 1.0576171875,
"step": 31
},
{
"epoch": 0.15311004784688995,
"grad_norm": 0.1320575773715973,
"learning_rate": 9.985917667125256e-05,
"loss": 1.04150390625,
"step": 32
},
{
"epoch": 0.15789473684210525,
"grad_norm": 0.1314796805381775,
"learning_rate": 9.982962244646046e-05,
"loss": 1.02587890625,
"step": 33
},
{
"epoch": 0.16267942583732056,
"grad_norm": 0.11765120178461075,
"learning_rate": 9.979726096871885e-05,
"loss": 1.0537109375,
"step": 34
},
{
"epoch": 0.1674641148325359,
"grad_norm": 0.11312679201364517,
"learning_rate": 9.976209426451938e-05,
"loss": 1.0400390625,
"step": 35
},
{
"epoch": 0.1722488038277512,
"grad_norm": 0.10504739731550217,
"learning_rate": 9.972412453601828e-05,
"loss": 1.02490234375,
"step": 36
},
{
"epoch": 0.17703349282296652,
"grad_norm": 0.11522939801216125,
"learning_rate": 9.968335416089861e-05,
"loss": 1.029296875,
"step": 37
},
{
"epoch": 0.18181818181818182,
"grad_norm": 0.09458598494529724,
"learning_rate": 9.963978569222122e-05,
"loss": 1.0068359375,
"step": 38
},
{
"epoch": 0.18660287081339713,
"grad_norm": 0.10005984455347061,
"learning_rate": 9.959342185826503e-05,
"loss": 1.02978515625,
"step": 39
},
{
"epoch": 0.19138755980861244,
"grad_norm": 0.11210939288139343,
"learning_rate": 9.954426556235606e-05,
"loss": 1.01513671875,
"step": 40
},
{
"epoch": 0.19617224880382775,
"grad_norm": 0.09823155403137207,
"learning_rate": 9.949231988268566e-05,
"loss": 0.9716796875,
"step": 41
},
{
"epoch": 0.20095693779904306,
"grad_norm": 0.09227780252695084,
"learning_rate": 9.943758807211777e-05,
"loss": 0.790771484375,
"step": 42
},
{
"epoch": 0.20574162679425836,
"grad_norm": 0.12116223573684692,
"learning_rate": 9.938007355798524e-05,
"loss": 1.001953125,
"step": 43
},
{
"epoch": 0.21052631578947367,
"grad_norm": 0.10298804938793182,
"learning_rate": 9.931977994187515e-05,
"loss": 1.02099609375,
"step": 44
},
{
"epoch": 0.215311004784689,
"grad_norm": 0.09136996418237686,
"learning_rate": 9.925671099940333e-05,
"loss": 0.78271484375,
"step": 45
},
{
"epoch": 0.22009569377990432,
"grad_norm": 0.08057890832424164,
"learning_rate": 9.919087067997785e-05,
"loss": 1.00537109375,
"step": 46
},
{
"epoch": 0.22488038277511962,
"grad_norm": 0.0971514880657196,
"learning_rate": 9.912226310655185e-05,
"loss": 1.02294921875,
"step": 47
},
{
"epoch": 0.22966507177033493,
"grad_norm": 0.0978764072060585,
"learning_rate": 9.905089257536524e-05,
"loss": 0.96142578125,
"step": 48
},
{
"epoch": 0.23444976076555024,
"grad_norm": 0.10639040172100067,
"learning_rate": 9.897676355567569e-05,
"loss": 0.95703125,
"step": 49
},
{
"epoch": 0.23923444976076555,
"grad_norm": 0.09742165356874466,
"learning_rate": 9.889988068947877e-05,
"loss": 0.9775390625,
"step": 50
},
{
"epoch": 0.24401913875598086,
"grad_norm": 0.10995205491781235,
"learning_rate": 9.882024879121728e-05,
"loss": 1.00244140625,
"step": 51
},
{
"epoch": 0.24880382775119617,
"grad_norm": 0.1241854727268219,
"learning_rate": 9.873787284747975e-05,
"loss": 0.90478515625,
"step": 52
},
{
"epoch": 0.2535885167464115,
"grad_norm": 0.11184431612491608,
"learning_rate": 9.865275801668819e-05,
"loss": 0.9521484375,
"step": 53
},
{
"epoch": 0.2583732057416268,
"grad_norm": 0.1032576933503151,
"learning_rate": 9.856490962877505e-05,
"loss": 0.8046875,
"step": 54
},
{
"epoch": 0.2631578947368421,
"grad_norm": 0.10926849395036697,
"learning_rate": 9.847433318484948e-05,
"loss": 0.95849609375,
"step": 55
},
{
"epoch": 0.2679425837320574,
"grad_norm": 0.09815143048763275,
"learning_rate": 9.838103435685287e-05,
"loss": 0.88134765625,
"step": 56
},
{
"epoch": 0.2727272727272727,
"grad_norm": 0.11383997648954391,
"learning_rate": 9.828501898720354e-05,
"loss": 0.96533203125,
"step": 57
},
{
"epoch": 0.27751196172248804,
"grad_norm": 0.10146185010671616,
"learning_rate": 9.818629308843109e-05,
"loss": 0.88330078125,
"step": 58
},
{
"epoch": 0.2822966507177033,
"grad_norm": 0.13061702251434326,
"learning_rate": 9.80848628427997e-05,
"loss": 0.9423828125,
"step": 59
},
{
"epoch": 0.28708133971291866,
"grad_norm": 0.13290119171142578,
"learning_rate": 9.798073460192112e-05,
"loss": 0.94384765625,
"step": 60
},
{
"epoch": 0.291866028708134,
"grad_norm": 0.10585255175828934,
"learning_rate": 9.787391488635688e-05,
"loss": 0.94775390625,
"step": 61
},
{
"epoch": 0.2966507177033493,
"grad_norm": 0.11498890817165375,
"learning_rate": 9.776441038520995e-05,
"loss": 0.9736328125,
"step": 62
},
{
"epoch": 0.3014354066985646,
"grad_norm": 0.12990789115428925,
"learning_rate": 9.765222795570594e-05,
"loss": 0.94921875,
"step": 63
},
{
"epoch": 0.3062200956937799,
"grad_norm": 0.11351815611124039,
"learning_rate": 9.75373746227636e-05,
"loss": 0.94189453125,
"step": 64
},
{
"epoch": 0.31100478468899523,
"grad_norm": 0.12847229838371277,
"learning_rate": 9.741985757855499e-05,
"loss": 0.9375,
"step": 65
},
{
"epoch": 0.3157894736842105,
"grad_norm": 0.12766319513320923,
"learning_rate": 9.729968418205505e-05,
"loss": 0.93310546875,
"step": 66
},
{
"epoch": 0.32057416267942584,
"grad_norm": 0.11785779893398285,
"learning_rate": 9.717686195858084e-05,
"loss": 0.9189453125,
"step": 67
},
{
"epoch": 0.3253588516746411,
"grad_norm": 0.12866587936878204,
"learning_rate": 9.705139859932026e-05,
"loss": 0.96875,
"step": 68
},
{
"epoch": 0.33014354066985646,
"grad_norm": 0.13676884770393372,
"learning_rate": 9.692330196085039e-05,
"loss": 0.923828125,
"step": 69
},
{
"epoch": 0.3349282296650718,
"grad_norm": 0.27054962515830994,
"learning_rate": 9.679258006464557e-05,
"loss": 0.91650390625,
"step": 70
},
{
"epoch": 0.3397129186602871,
"grad_norm": 0.10657820850610733,
"learning_rate": 9.665924109657509e-05,
"loss": 0.9130859375,
"step": 71
},
{
"epoch": 0.3444976076555024,
"grad_norm": 0.12274233996868134,
"learning_rate": 9.652329340639053e-05,
"loss": 0.93359375,
"step": 72
},
{
"epoch": 0.3492822966507177,
"grad_norm": 0.13598042726516724,
"learning_rate": 9.638474550720292e-05,
"loss": 0.916015625,
"step": 73
},
{
"epoch": 0.35406698564593303,
"grad_norm": 0.14431379735469818,
"learning_rate": 9.624360607494971e-05,
"loss": 0.87548828125,
"step": 74
},
{
"epoch": 0.3588516746411483,
"grad_norm": 0.10417573153972626,
"learning_rate": 9.609988394785139e-05,
"loss": 0.939453125,
"step": 75
},
{
"epoch": 0.36363636363636365,
"grad_norm": 0.13331657648086548,
"learning_rate": 9.595358812585801e-05,
"loss": 0.90771484375,
"step": 76
},
{
"epoch": 0.3684210526315789,
"grad_norm": 0.13184547424316406,
"learning_rate": 9.580472777008574e-05,
"loss": 0.90576171875,
"step": 77
},
{
"epoch": 0.37320574162679426,
"grad_norm": 0.14161042869091034,
"learning_rate": 9.565331220224312e-05,
"loss": 0.876953125,
"step": 78
},
{
"epoch": 0.37799043062200954,
"grad_norm": 0.14439073204994202,
"learning_rate": 9.549935090404726e-05,
"loss": 0.90771484375,
"step": 79
},
{
"epoch": 0.3827751196172249,
"grad_norm": 0.11294569820165634,
"learning_rate": 9.534285351663014e-05,
"loss": 0.8955078125,
"step": 80
},
{
"epoch": 0.3875598086124402,
"grad_norm": 0.11165916174650192,
"learning_rate": 9.518382983993501e-05,
"loss": 0.9697265625,
"step": 81
},
{
"epoch": 0.3923444976076555,
"grad_norm": 0.12481875717639923,
"learning_rate": 9.502228983210244e-05,
"loss": 0.896484375,
"step": 82
},
{
"epoch": 0.39712918660287083,
"grad_norm": 0.10466470569372177,
"learning_rate": 9.485824360884704e-05,
"loss": 0.9091796875,
"step": 83
},
{
"epoch": 0.4019138755980861,
"grad_norm": 0.11642592400312424,
"learning_rate": 9.469170144282373e-05,
"loss": 0.88427734375,
"step": 84
},
{
"epoch": 0.40669856459330145,
"grad_norm": 0.16490517556667328,
"learning_rate": 9.452267376298467e-05,
"loss": 0.9072265625,
"step": 85
},
{
"epoch": 0.41148325358851673,
"grad_norm": 0.1355501413345337,
"learning_rate": 9.435117115392608e-05,
"loss": 0.9560546875,
"step": 86
},
{
"epoch": 0.41148325358851673,
"eval_loss": 0.9066162109375,
"eval_runtime": 6.328,
"eval_samples_per_second": 1.106,
"eval_steps_per_second": 0.158,
"step": 86
},
{
"epoch": 0.41626794258373206,
"grad_norm": 0.18226318061351776,
"learning_rate": 9.41772043552255e-05,
"loss": 0.880859375,
"step": 87
},
{
"epoch": 0.42105263157894735,
"grad_norm": 0.13384920358657837,
"learning_rate": 9.400078426076915e-05,
"loss": 0.9052734375,
"step": 88
},
{
"epoch": 0.4258373205741627,
"grad_norm": 0.154852494597435,
"learning_rate": 9.382192191806994e-05,
"loss": 0.96142578125,
"step": 89
},
{
"epoch": 0.430622009569378,
"grad_norm": 0.10967659950256348,
"learning_rate": 9.364062852757552e-05,
"loss": 0.79150390625,
"step": 90
},
{
"epoch": 0.4354066985645933,
"grad_norm": 0.1447352170944214,
"learning_rate": 9.345691544196688e-05,
"loss": 0.89111328125,
"step": 91
},
{
"epoch": 0.44019138755980863,
"grad_norm": 0.1410771757364273,
"learning_rate": 9.327079416544764e-05,
"loss": 0.9208984375,
"step": 92
},
{
"epoch": 0.4449760765550239,
"grad_norm": 0.12851056456565857,
"learning_rate": 9.308227635302339e-05,
"loss": 0.89306640625,
"step": 93
},
{
"epoch": 0.44976076555023925,
"grad_norm": 0.1392790973186493,
"learning_rate": 9.289137380977205e-05,
"loss": 0.86962890625,
"step": 94
},
{
"epoch": 0.45454545454545453,
"grad_norm": 0.13085263967514038,
"learning_rate": 9.26980984901045e-05,
"loss": 0.8564453125,
"step": 95
},
{
"epoch": 0.45933014354066987,
"grad_norm": 0.15431171655654907,
"learning_rate": 9.25024624970161e-05,
"loss": 0.89501953125,
"step": 96
},
{
"epoch": 0.46411483253588515,
"grad_norm": 0.14034897089004517,
"learning_rate": 9.230447808132869e-05,
"loss": 0.9677734375,
"step": 97
},
{
"epoch": 0.4688995215311005,
"grad_norm": 0.14690035581588745,
"learning_rate": 9.210415764092353e-05,
"loss": 0.86376953125,
"step": 98
},
{
"epoch": 0.47368421052631576,
"grad_norm": 0.1203850582242012,
"learning_rate": 9.190151371996483e-05,
"loss": 0.802734375,
"step": 99
},
{
"epoch": 0.4784688995215311,
"grad_norm": 0.18799160420894623,
"learning_rate": 9.169655900811437e-05,
"loss": 0.85009765625,
"step": 100
},
{
"epoch": 0.48325358851674644,
"grad_norm": 0.14786072075366974,
"learning_rate": 9.148930633973673e-05,
"loss": 0.880859375,
"step": 101
},
{
"epoch": 0.4880382775119617,
"grad_norm": 0.16505296528339386,
"learning_rate": 9.127976869309566e-05,
"loss": 0.88232421875,
"step": 102
},
{
"epoch": 0.49282296650717705,
"grad_norm": 0.1454196274280548,
"learning_rate": 9.106795918954142e-05,
"loss": 0.93212890625,
"step": 103
},
{
"epoch": 0.49760765550239233,
"grad_norm": 0.13809478282928467,
"learning_rate": 9.085389109268901e-05,
"loss": 0.86279296875,
"step": 104
},
{
"epoch": 0.5023923444976076,
"grad_norm": 0.16683712601661682,
"learning_rate": 9.063757780758766e-05,
"loss": 0.888671875,
"step": 105
},
{
"epoch": 0.507177033492823,
"grad_norm": 0.12409838289022446,
"learning_rate": 9.041903287988142e-05,
"loss": 0.9150390625,
"step": 106
},
{
"epoch": 0.5119617224880383,
"grad_norm": 0.2158932238817215,
"learning_rate": 9.019826999496086e-05,
"loss": 0.87890625,
"step": 107
},
{
"epoch": 0.5167464114832536,
"grad_norm": 0.14570724964141846,
"learning_rate": 8.997530297710614e-05,
"loss": 0.9169921875,
"step": 108
},
{
"epoch": 0.5215311004784688,
"grad_norm": 0.14855045080184937,
"learning_rate": 8.975014578862129e-05,
"loss": 0.8984375,
"step": 109
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.14390894770622253,
"learning_rate": 8.95228125289599e-05,
"loss": 0.8857421875,
"step": 110
},
{
"epoch": 0.5311004784688995,
"grad_norm": 0.15073491632938385,
"learning_rate": 8.929331743384225e-05,
"loss": 0.85791015625,
"step": 111
},
{
"epoch": 0.5358851674641149,
"grad_norm": 0.13765165209770203,
"learning_rate": 8.906167487436373e-05,
"loss": 0.88916015625,
"step": 112
},
{
"epoch": 0.5406698564593302,
"grad_norm": 0.12538696825504303,
"learning_rate": 8.882789935609508e-05,
"loss": 0.9267578125,
"step": 113
},
{
"epoch": 0.5454545454545454,
"grad_norm": 0.18623046576976776,
"learning_rate": 8.859200551817393e-05,
"loss": 0.85205078125,
"step": 114
},
{
"epoch": 0.5502392344497608,
"grad_norm": 0.13965821266174316,
"learning_rate": 8.835400813238813e-05,
"loss": 0.83544921875,
"step": 115
},
{
"epoch": 0.5550239234449761,
"grad_norm": 0.14598041772842407,
"learning_rate": 8.811392210225074e-05,
"loss": 0.81884765625,
"step": 116
},
{
"epoch": 0.5598086124401914,
"grad_norm": 0.14970417320728302,
"learning_rate": 8.787176246206672e-05,
"loss": 0.9267578125,
"step": 117
},
{
"epoch": 0.5645933014354066,
"grad_norm": 0.1508273482322693,
"learning_rate": 8.762754437599157e-05,
"loss": 0.8349609375,
"step": 118
},
{
"epoch": 0.569377990430622,
"grad_norm": 0.15280179679393768,
"learning_rate": 8.738128313708162e-05,
"loss": 0.84423828125,
"step": 119
},
{
"epoch": 0.5741626794258373,
"grad_norm": 0.12097357213497162,
"learning_rate": 8.713299416633647e-05,
"loss": 0.83984375,
"step": 120
},
{
"epoch": 0.5789473684210527,
"grad_norm": 0.1372656673192978,
"learning_rate": 8.688269301173326e-05,
"loss": 0.91015625,
"step": 121
},
{
"epoch": 0.583732057416268,
"grad_norm": 0.14168693125247955,
"learning_rate": 8.663039534725309e-05,
"loss": 0.8994140625,
"step": 122
},
{
"epoch": 0.5885167464114832,
"grad_norm": 0.15086841583251953,
"learning_rate": 8.637611697189947e-05,
"loss": 0.861328125,
"step": 123
},
{
"epoch": 0.5933014354066986,
"grad_norm": 0.1354799121618271,
"learning_rate": 8.6119873808709e-05,
"loss": 0.78466796875,
"step": 124
},
{
"epoch": 0.5980861244019139,
"grad_norm": 0.13367100059986115,
"learning_rate": 8.586168190375428e-05,
"loss": 0.8994140625,
"step": 125
},
{
"epoch": 0.6028708133971292,
"grad_norm": 0.14267395436763763,
"learning_rate": 8.560155742513907e-05,
"loss": 0.89453125,
"step": 126
},
{
"epoch": 0.6076555023923444,
"grad_norm": 0.15500813722610474,
"learning_rate": 8.53395166619858e-05,
"loss": 0.81396484375,
"step": 127
},
{
"epoch": 0.6124401913875598,
"grad_norm": 0.1287531852722168,
"learning_rate": 8.507557602341559e-05,
"loss": 0.8720703125,
"step": 128
},
{
"epoch": 0.6172248803827751,
"grad_norm": 0.1518527865409851,
"learning_rate": 8.480975203752077e-05,
"loss": 0.84423828125,
"step": 129
},
{
"epoch": 0.6220095693779905,
"grad_norm": 0.14997772872447968,
"learning_rate": 8.454206135032967e-05,
"loss": 0.8974609375,
"step": 130
},
{
"epoch": 0.6267942583732058,
"grad_norm": 0.14649060368537903,
"learning_rate": 8.427252072476452e-05,
"loss": 0.89697265625,
"step": 131
},
{
"epoch": 0.631578947368421,
"grad_norm": 0.18086740374565125,
"learning_rate": 8.400114703959144e-05,
"loss": 0.83154296875,
"step": 132
},
{
"epoch": 0.6363636363636364,
"grad_norm": 0.1518823206424713,
"learning_rate": 8.372795728836377e-05,
"loss": 0.89208984375,
"step": 133
},
{
"epoch": 0.6411483253588517,
"grad_norm": 0.1383829265832901,
"learning_rate": 8.345296857835775e-05,
"loss": 0.87939453125,
"step": 134
},
{
"epoch": 0.645933014354067,
"grad_norm": 0.18899908661842346,
"learning_rate": 8.317619812950128e-05,
"loss": 0.84130859375,
"step": 135
},
{
"epoch": 0.6507177033492823,
"grad_norm": 0.12442408502101898,
"learning_rate": 8.289766327329565e-05,
"loss": 0.830810546875,
"step": 136
},
{
"epoch": 0.6555023923444976,
"grad_norm": 0.13153305649757385,
"learning_rate": 8.26173814517302e-05,
"loss": 0.8759765625,
"step": 137
},
{
"epoch": 0.6602870813397129,
"grad_norm": 0.16612817347049713,
"learning_rate": 8.23353702161901e-05,
"loss": 0.8876953125,
"step": 138
},
{
"epoch": 0.6650717703349283,
"grad_norm": 0.12367686629295349,
"learning_rate": 8.205164722635726e-05,
"loss": 0.8505859375,
"step": 139
},
{
"epoch": 0.6698564593301436,
"grad_norm": 0.11612129211425781,
"learning_rate": 8.176623024910447e-05,
"loss": 0.6962890625,
"step": 140
},
{
"epoch": 0.6746411483253588,
"grad_norm": 0.21645186841487885,
"learning_rate": 8.147913715738287e-05,
"loss": 0.85302734375,
"step": 141
},
{
"epoch": 0.6794258373205742,
"grad_norm": 0.1662965565919876,
"learning_rate": 8.119038592910271e-05,
"loss": 0.8515625,
"step": 142
},
{
"epoch": 0.6842105263157895,
"grad_norm": 0.1909109205007553,
"learning_rate": 8.089999464600756e-05,
"loss": 0.89404296875,
"step": 143
},
{
"epoch": 0.6889952153110048,
"grad_norm": 0.14662204682826996,
"learning_rate": 8.060798149254201e-05,
"loss": 0.86181640625,
"step": 144
},
{
"epoch": 0.69377990430622,
"grad_norm": 0.18106287717819214,
"learning_rate": 8.0314364754713e-05,
"loss": 0.833984375,
"step": 145
},
{
"epoch": 0.6985645933014354,
"grad_norm": 0.2151636928319931,
"learning_rate": 8.001916281894467e-05,
"loss": 0.8388671875,
"step": 146
},
{
"epoch": 0.7033492822966507,
"grad_norm": 0.17297010123729706,
"learning_rate": 7.972239417092707e-05,
"loss": 0.845703125,
"step": 147
},
{
"epoch": 0.7081339712918661,
"grad_norm": 0.18964099884033203,
"learning_rate": 7.94240773944585e-05,
"loss": 0.72802734375,
"step": 148
},
{
"epoch": 0.7129186602870813,
"grad_norm": 0.1460205465555191,
"learning_rate": 7.912423117028187e-05,
"loss": 0.85205078125,
"step": 149
},
{
"epoch": 0.7177033492822966,
"grad_norm": 0.14216914772987366,
"learning_rate": 7.882287427491483e-05,
"loss": 0.8818359375,
"step": 150
},
{
"epoch": 0.722488038277512,
"grad_norm": 0.1378447413444519,
"learning_rate": 7.8520025579474e-05,
"loss": 0.856689453125,
"step": 151
},
{
"epoch": 0.7272727272727273,
"grad_norm": 0.18658150732517242,
"learning_rate": 7.821570404849326e-05,
"loss": 0.8603515625,
"step": 152
},
{
"epoch": 0.7320574162679426,
"grad_norm": 0.14973101019859314,
"learning_rate": 7.790992873873615e-05,
"loss": 0.859375,
"step": 153
},
{
"epoch": 0.7368421052631579,
"grad_norm": 0.17674702405929565,
"learning_rate": 7.76027187980026e-05,
"loss": 0.8662109375,
"step": 154
},
{
"epoch": 0.7416267942583732,
"grad_norm": 0.17365242540836334,
"learning_rate": 7.729409346392979e-05,
"loss": 0.87158203125,
"step": 155
},
{
"epoch": 0.7464114832535885,
"grad_norm": 0.15215370059013367,
"learning_rate": 7.698407206278752e-05,
"loss": 0.8603515625,
"step": 156
},
{
"epoch": 0.7511961722488039,
"grad_norm": 0.17037777602672577,
"learning_rate": 7.667267400826808e-05,
"loss": 0.8818359375,
"step": 157
},
{
"epoch": 0.7559808612440191,
"grad_norm": 0.1910388469696045,
"learning_rate": 7.635991880027031e-05,
"loss": 0.8876953125,
"step": 158
},
{
"epoch": 0.7607655502392344,
"grad_norm": 0.13110846281051636,
"learning_rate": 7.604582602367885e-05,
"loss": 0.88427734375,
"step": 159
},
{
"epoch": 0.7655502392344498,
"grad_norm": 0.16954734921455383,
"learning_rate": 7.57304153471374e-05,
"loss": 0.74609375,
"step": 160
},
{
"epoch": 0.7703349282296651,
"grad_norm": 0.15228325128555298,
"learning_rate": 7.541370652181729e-05,
"loss": 0.83447265625,
"step": 161
},
{
"epoch": 0.7751196172248804,
"grad_norm": 0.1494646817445755,
"learning_rate": 7.509571938018049e-05,
"loss": 0.88525390625,
"step": 162
},
{
"epoch": 0.7799043062200957,
"grad_norm": 0.16694070398807526,
"learning_rate": 7.477647383473785e-05,
"loss": 0.8857421875,
"step": 163
},
{
"epoch": 0.784688995215311,
"grad_norm": 0.1934370994567871,
"learning_rate": 7.445598987680199e-05,
"loss": 0.8115234375,
"step": 164
},
{
"epoch": 0.7894736842105263,
"grad_norm": 0.15983052551746368,
"learning_rate": 7.413428757523559e-05,
"loss": 0.86181640625,
"step": 165
},
{
"epoch": 0.7942583732057417,
"grad_norm": 0.13332803547382355,
"learning_rate": 7.381138707519461e-05,
"loss": 0.92333984375,
"step": 166
},
{
"epoch": 0.7990430622009569,
"grad_norm": 0.12580889463424683,
"learning_rate": 7.348730859686674e-05,
"loss": 0.8427734375,
"step": 167
},
{
"epoch": 0.8038277511961722,
"grad_norm": 0.15740549564361572,
"learning_rate": 7.31620724342053e-05,
"loss": 0.6341552734375,
"step": 168
},
{
"epoch": 0.8086124401913876,
"grad_norm": 0.1947418600320816,
"learning_rate": 7.283569895365833e-05,
"loss": 0.81982421875,
"step": 169
},
{
"epoch": 0.8086124401913876,
"eval_loss": 0.8587646484375,
"eval_runtime": 6.3234,
"eval_samples_per_second": 1.107,
"eval_steps_per_second": 0.158,
"step": 169
},
{
"epoch": 0.8133971291866029,
"grad_norm": 0.16423191130161285,
"learning_rate": 7.250820859289327e-05,
"loss": 0.82861328125,
"step": 170
},
{
"epoch": 0.8181818181818182,
"grad_norm": 0.1542576253414154,
"learning_rate": 7.217962185951722e-05,
"loss": 0.8154296875,
"step": 171
},
{
"epoch": 0.8229665071770335,
"grad_norm": 0.15081188082695007,
"learning_rate": 7.184995932979258e-05,
"loss": 0.79541015625,
"step": 172
},
{
"epoch": 0.8277511961722488,
"grad_norm": 0.15851037204265594,
"learning_rate": 7.151924164734867e-05,
"loss": 0.86865234375,
"step": 173
},
{
"epoch": 0.8325358851674641,
"grad_norm": 0.15119867026805878,
"learning_rate": 7.118748952188904e-05,
"loss": 0.84130859375,
"step": 174
},
{
"epoch": 0.8373205741626795,
"grad_norm": 0.16937977075576782,
"learning_rate": 7.08547237278945e-05,
"loss": 0.84375,
"step": 175
},
{
"epoch": 0.8421052631578947,
"grad_norm": 0.16020779311656952,
"learning_rate": 7.052096510332233e-05,
"loss": 0.8779296875,
"step": 176
},
{
"epoch": 0.84688995215311,
"grad_norm": 0.1509438157081604,
"learning_rate": 7.018623454830135e-05,
"loss": 0.82275390625,
"step": 177
},
{
"epoch": 0.8516746411483254,
"grad_norm": 0.14054879546165466,
"learning_rate": 6.98505530238231e-05,
"loss": 0.86962890625,
"step": 178
},
{
"epoch": 0.8564593301435407,
"grad_norm": 0.1427825540304184,
"learning_rate": 6.95139415504293e-05,
"loss": 0.91845703125,
"step": 179
},
{
"epoch": 0.861244019138756,
"grad_norm": 0.167701855301857,
"learning_rate": 6.917642120689556e-05,
"loss": 0.81787109375,
"step": 180
},
{
"epoch": 0.8660287081339713,
"grad_norm": 0.13314808905124664,
"learning_rate": 6.883801312891139e-05,
"loss": 0.89453125,
"step": 181
},
{
"epoch": 0.8708133971291866,
"grad_norm": 0.15560881793498993,
"learning_rate": 6.849873850775662e-05,
"loss": 0.85595703125,
"step": 182
},
{
"epoch": 0.8755980861244019,
"grad_norm": 0.1258866786956787,
"learning_rate": 6.815861858897447e-05,
"loss": 0.71337890625,
"step": 183
},
{
"epoch": 0.8803827751196173,
"grad_norm": 0.17535783350467682,
"learning_rate": 6.781767467104111e-05,
"loss": 0.8671875,
"step": 184
},
{
"epoch": 0.8851674641148325,
"grad_norm": 0.1488444209098816,
"learning_rate": 6.747592810403193e-05,
"loss": 0.818359375,
"step": 185
},
{
"epoch": 0.8899521531100478,
"grad_norm": 0.14873863756656647,
"learning_rate": 6.713340028828461e-05,
"loss": 0.84521484375,
"step": 186
},
{
"epoch": 0.8947368421052632,
"grad_norm": 0.15342408418655396,
"learning_rate": 6.6790112673059e-05,
"loss": 0.828125,
"step": 187
},
{
"epoch": 0.8995215311004785,
"grad_norm": 0.1610623598098755,
"learning_rate": 6.644608675519398e-05,
"loss": 0.806640625,
"step": 188
},
{
"epoch": 0.9043062200956937,
"grad_norm": 0.1683485358953476,
"learning_rate": 6.610134407776127e-05,
"loss": 0.84521484375,
"step": 189
},
{
"epoch": 0.9090909090909091,
"grad_norm": 0.15010549128055573,
"learning_rate": 6.57559062287165e-05,
"loss": 0.845703125,
"step": 190
},
{
"epoch": 0.9138755980861244,
"grad_norm": 0.16008473932743073,
"learning_rate": 6.540979483954715e-05,
"loss": 0.71630859375,
"step": 191
},
{
"epoch": 0.9186602870813397,
"grad_norm": 0.14229531586170197,
"learning_rate": 6.506303158391828e-05,
"loss": 0.651611328125,
"step": 192
},
{
"epoch": 0.9234449760765551,
"grad_norm": 0.18040095269680023,
"learning_rate": 6.471563817631506e-05,
"loss": 0.85888671875,
"step": 193
},
{
"epoch": 0.9282296650717703,
"grad_norm": 0.14827679097652435,
"learning_rate": 6.436763637068307e-05,
"loss": 0.841796875,
"step": 194
},
{
"epoch": 0.9330143540669856,
"grad_norm": 0.1827199012041092,
"learning_rate": 6.401904795906609e-05,
"loss": 0.90673828125,
"step": 195
},
{
"epoch": 0.937799043062201,
"grad_norm": 0.18707910180091858,
"learning_rate": 6.366989477024148e-05,
"loss": 0.8154296875,
"step": 196
},
{
"epoch": 0.9425837320574163,
"grad_norm": 0.15842516720294952,
"learning_rate": 6.332019866835321e-05,
"loss": 0.828125,
"step": 197
},
{
"epoch": 0.9473684210526315,
"grad_norm": 0.18741275370121002,
"learning_rate": 6.29699815515427e-05,
"loss": 0.833984375,
"step": 198
},
{
"epoch": 0.9521531100478469,
"grad_norm": 0.14310209453105927,
"learning_rate": 6.261926535057764e-05,
"loss": 0.89599609375,
"step": 199
},
{
"epoch": 0.9569377990430622,
"grad_norm": 0.15200170874595642,
"learning_rate": 6.226807202747858e-05,
"loss": 0.85546875,
"step": 200
},
{
"epoch": 0.9617224880382775,
"grad_norm": 0.1716766655445099,
"learning_rate": 6.191642357414369e-05,
"loss": 0.7978515625,
"step": 201
},
{
"epoch": 0.9665071770334929,
"grad_norm": 0.17156939208507538,
"learning_rate": 6.156434201097162e-05,
"loss": 0.880615234375,
"step": 202
},
{
"epoch": 0.9712918660287081,
"grad_norm": 0.16476194560527802,
"learning_rate": 6.121184938548258e-05,
"loss": 0.79638671875,
"step": 203
},
{
"epoch": 0.9760765550239234,
"grad_norm": 0.16853675246238708,
"learning_rate": 6.0858967770937694e-05,
"loss": 0.8388671875,
"step": 204
},
{
"epoch": 0.9808612440191388,
"grad_norm": 0.14625944197177887,
"learning_rate": 6.0505719264956776e-05,
"loss": 0.88134765625,
"step": 205
},
{
"epoch": 0.9856459330143541,
"grad_norm": 0.15870210528373718,
"learning_rate": 6.015212598813457e-05,
"loss": 0.81005859375,
"step": 206
},
{
"epoch": 0.9904306220095693,
"grad_norm": 0.1566341668367386,
"learning_rate": 5.979821008265553e-05,
"loss": 0.90771484375,
"step": 207
},
{
"epoch": 0.9952153110047847,
"grad_norm": 0.15179800987243652,
"learning_rate": 5.944399371090726e-05,
"loss": 0.86181640625,
"step": 208
},
{
"epoch": 1.0,
"grad_norm": 0.13958945870399475,
"learning_rate": 5.9089499054092765e-05,
"loss": 0.86376953125,
"step": 209
},
{
"epoch": 1.0047846889952152,
"grad_norm": 0.12973764538764954,
"learning_rate": 5.8734748310841315e-05,
"loss": 0.884765625,
"step": 210
},
{
"epoch": 1.0095693779904307,
"grad_norm": 0.15666553378105164,
"learning_rate": 5.837976369581852e-05,
"loss": 0.8603515625,
"step": 211
},
{
"epoch": 1.014354066985646,
"grad_norm": 0.1602666825056076,
"learning_rate": 5.8024567438335074e-05,
"loss": 0.84521484375,
"step": 212
},
{
"epoch": 1.0191387559808613,
"grad_norm": 0.13860997557640076,
"learning_rate": 5.7669181780954906e-05,
"loss": 0.828125,
"step": 213
},
{
"epoch": 1.0239234449760766,
"grad_norm": 0.1436486691236496,
"learning_rate": 5.73136289781022e-05,
"loss": 0.80126953125,
"step": 214
},
{
"epoch": 1.0287081339712918,
"grad_norm": 0.1419706642627716,
"learning_rate": 5.6957931294667885e-05,
"loss": 0.84228515625,
"step": 215
},
{
"epoch": 1.0334928229665072,
"grad_norm": 0.14055973291397095,
"learning_rate": 5.660211100461542e-05,
"loss": 0.8125,
"step": 216
},
{
"epoch": 1.0382775119617225,
"grad_norm": 0.15234734117984772,
"learning_rate": 5.6246190389585895e-05,
"loss": 0.841796875,
"step": 217
},
{
"epoch": 1.0430622009569377,
"grad_norm": 0.1302803009748459,
"learning_rate": 5.589019173750283e-05,
"loss": 0.84912109375,
"step": 218
},
{
"epoch": 1.0478468899521531,
"grad_norm": 0.16531650722026825,
"learning_rate": 5.553413734117647e-05,
"loss": 0.791015625,
"step": 219
},
{
"epoch": 1.0526315789473684,
"grad_norm": 0.13514459133148193,
"learning_rate": 5.517804949690777e-05,
"loss": 0.73291015625,
"step": 220
},
{
"epoch": 1.0574162679425838,
"grad_norm": 0.16611863672733307,
"learning_rate": 5.482195050309223e-05,
"loss": 0.8291015625,
"step": 221
},
{
"epoch": 1.062200956937799,
"grad_norm": 0.1442006230354309,
"learning_rate": 5.446586265882354e-05,
"loss": 0.85009765625,
"step": 222
},
{
"epoch": 1.0669856459330143,
"grad_norm": 0.21945621073246002,
"learning_rate": 5.410980826249719e-05,
"loss": 0.8466796875,
"step": 223
},
{
"epoch": 1.0717703349282297,
"grad_norm": 0.15590190887451172,
"learning_rate": 5.375380961041411e-05,
"loss": 0.876953125,
"step": 224
},
{
"epoch": 1.076555023923445,
"grad_norm": 0.1533394306898117,
"learning_rate": 5.3397888995384594e-05,
"loss": 0.89697265625,
"step": 225
},
{
"epoch": 1.0813397129186604,
"grad_norm": 0.156230628490448,
"learning_rate": 5.3042068705332125e-05,
"loss": 0.85595703125,
"step": 226
},
{
"epoch": 1.0861244019138756,
"grad_norm": 0.16652673482894897,
"learning_rate": 5.2686371021897826e-05,
"loss": 0.84228515625,
"step": 227
},
{
"epoch": 1.0909090909090908,
"grad_norm": 0.1318468302488327,
"learning_rate": 5.2330818219045105e-05,
"loss": 0.73486328125,
"step": 228
},
{
"epoch": 1.0956937799043063,
"grad_norm": 0.15515613555908203,
"learning_rate": 5.197543256166493e-05,
"loss": 0.85888671875,
"step": 229
},
{
"epoch": 1.1004784688995215,
"grad_norm": 0.16461950540542603,
"learning_rate": 5.16202363041815e-05,
"loss": 0.79833984375,
"step": 230
},
{
"epoch": 1.1052631578947367,
"grad_norm": 0.17022602260112762,
"learning_rate": 5.1265251689158675e-05,
"loss": 0.78076171875,
"step": 231
},
{
"epoch": 1.1100478468899522,
"grad_norm": 0.15896818041801453,
"learning_rate": 5.0910500945907245e-05,
"loss": 0.8203125,
"step": 232
},
{
"epoch": 1.1148325358851674,
"grad_norm": 0.1486581265926361,
"learning_rate": 5.055600628909275e-05,
"loss": 0.90087890625,
"step": 233
},
{
"epoch": 1.1196172248803828,
"grad_norm": 0.16293200850486755,
"learning_rate": 5.020178991734449e-05,
"loss": 0.8046875,
"step": 234
},
{
"epoch": 1.124401913875598,
"grad_norm": 0.16731753945350647,
"learning_rate": 4.984787401186545e-05,
"loss": 0.80517578125,
"step": 235
},
{
"epoch": 1.1291866028708135,
"grad_norm": 0.1648280918598175,
"learning_rate": 4.949428073504324e-05,
"loss": 0.8291015625,
"step": 236
},
{
"epoch": 1.1339712918660287,
"grad_norm": 0.15662537515163422,
"learning_rate": 4.914103222906233e-05,
"loss": 0.83935546875,
"step": 237
},
{
"epoch": 1.138755980861244,
"grad_norm": 0.13991908729076385,
"learning_rate": 4.8788150614517434e-05,
"loss": 0.8056640625,
"step": 238
},
{
"epoch": 1.1435406698564594,
"grad_norm": 0.1668047159910202,
"learning_rate": 4.8435657989028395e-05,
"loss": 0.81005859375,
"step": 239
},
{
"epoch": 1.1483253588516746,
"grad_norm": 0.15042896568775177,
"learning_rate": 4.808357642585633e-05,
"loss": 0.83447265625,
"step": 240
},
{
"epoch": 1.1531100478468899,
"grad_norm": 0.15066254138946533,
"learning_rate": 4.773192797252144e-05,
"loss": 0.81787109375,
"step": 241
},
{
"epoch": 1.1578947368421053,
"grad_norm": 0.16723085939884186,
"learning_rate": 4.738073464942237e-05,
"loss": 0.80859375,
"step": 242
},
{
"epoch": 1.1626794258373205,
"grad_norm": 0.14816319942474365,
"learning_rate": 4.70300184484573e-05,
"loss": 0.8466796875,
"step": 243
},
{
"epoch": 1.167464114832536,
"grad_norm": 0.1575804203748703,
"learning_rate": 4.667980133164681e-05,
"loss": 0.826171875,
"step": 244
},
{
"epoch": 1.1722488038277512,
"grad_norm": 0.13895685970783234,
"learning_rate": 4.6330105229758526e-05,
"loss": 0.830078125,
"step": 245
},
{
"epoch": 1.1770334928229664,
"grad_norm": 0.1610182672739029,
"learning_rate": 4.5980952040933925e-05,
"loss": 0.8212890625,
"step": 246
},
{
"epoch": 1.1818181818181819,
"grad_norm": 0.1358628123998642,
"learning_rate": 4.5632363629316944e-05,
"loss": 0.8349609375,
"step": 247
},
{
"epoch": 1.186602870813397,
"grad_norm": 0.16048452258110046,
"learning_rate": 4.528436182368495e-05,
"loss": 0.845703125,
"step": 248
},
{
"epoch": 1.1913875598086126,
"grad_norm": 0.16002970933914185,
"learning_rate": 4.493696841608171e-05,
"loss": 0.81787109375,
"step": 249
},
{
"epoch": 1.1961722488038278,
"grad_norm": 0.15855439007282257,
"learning_rate": 4.459020516045285e-05,
"loss": 0.78271484375,
"step": 250
},
{
"epoch": 1.200956937799043,
"grad_norm": 0.14956708252429962,
"learning_rate": 4.4244093771283536e-05,
"loss": 0.6328125,
"step": 251
},
{
"epoch": 1.2057416267942584,
"grad_norm": 0.1500825136899948,
"learning_rate": 4.389865592223873e-05,
"loss": 0.82080078125,
"step": 252
},
{
"epoch": 1.2057416267942584,
"eval_loss": 0.8333740234375,
"eval_runtime": 6.313,
"eval_samples_per_second": 1.109,
"eval_steps_per_second": 0.158,
"step": 252
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.15775145590305328,
"learning_rate": 4.3553913244806046e-05,
"loss": 0.85693359375,
"step": 253
},
{
"epoch": 1.215311004784689,
"grad_norm": 0.14435818791389465,
"learning_rate": 4.3209887326941015e-05,
"loss": 0.64599609375,
"step": 254
},
{
"epoch": 1.2200956937799043,
"grad_norm": 0.13371528685092926,
"learning_rate": 4.286659971171541e-05,
"loss": 0.85302734375,
"step": 255
},
{
"epoch": 1.2248803827751196,
"grad_norm": 0.15325836837291718,
"learning_rate": 4.2524071895968076e-05,
"loss": 0.8505859375,
"step": 256
},
{
"epoch": 1.229665071770335,
"grad_norm": 0.1479969322681427,
"learning_rate": 4.2182325328958905e-05,
"loss": 0.80029296875,
"step": 257
},
{
"epoch": 1.2344497607655502,
"grad_norm": 0.17559714615345,
"learning_rate": 4.184138141102555e-05,
"loss": 0.783203125,
"step": 258
},
{
"epoch": 1.2392344497607655,
"grad_norm": 0.19007180631160736,
"learning_rate": 4.150126149224338e-05,
"loss": 0.81396484375,
"step": 259
},
{
"epoch": 1.244019138755981,
"grad_norm": 0.17040984332561493,
"learning_rate": 4.116198687108862e-05,
"loss": 0.83447265625,
"step": 260
},
{
"epoch": 1.2488038277511961,
"grad_norm": 0.1566859483718872,
"learning_rate": 4.0823578793104446e-05,
"loss": 0.744140625,
"step": 261
},
{
"epoch": 1.2535885167464116,
"grad_norm": 0.1724807471036911,
"learning_rate": 4.0486058449570716e-05,
"loss": 0.798828125,
"step": 262
},
{
"epoch": 1.2583732057416268,
"grad_norm": 0.1529332548379898,
"learning_rate": 4.014944697617692e-05,
"loss": 0.6728515625,
"step": 263
},
{
"epoch": 1.263157894736842,
"grad_norm": 0.1585606038570404,
"learning_rate": 3.981376545169867e-05,
"loss": 0.81298828125,
"step": 264
},
{
"epoch": 1.2679425837320575,
"grad_norm": 0.167527973651886,
"learning_rate": 3.947903489667768e-05,
"loss": 0.740234375,
"step": 265
},
{
"epoch": 1.2727272727272727,
"grad_norm": 0.1705494225025177,
"learning_rate": 3.91452762721055e-05,
"loss": 0.8134765625,
"step": 266
},
{
"epoch": 1.277511961722488,
"grad_norm": 0.12672428786754608,
"learning_rate": 3.8812510478110975e-05,
"loss": 0.75439453125,
"step": 267
},
{
"epoch": 1.2822966507177034,
"grad_norm": 0.19668768346309662,
"learning_rate": 3.848075835265135e-05,
"loss": 0.7919921875,
"step": 268
},
{
"epoch": 1.2870813397129186,
"grad_norm": 0.17751078307628632,
"learning_rate": 3.8150040670207435e-05,
"loss": 0.79296875,
"step": 269
},
{
"epoch": 1.291866028708134,
"grad_norm": 0.1517217457294464,
"learning_rate": 3.7820378140482794e-05,
"loss": 0.81591796875,
"step": 270
},
{
"epoch": 1.2966507177033493,
"grad_norm": 0.17382237315177917,
"learning_rate": 3.7491791407106736e-05,
"loss": 0.83349609375,
"step": 271
},
{
"epoch": 1.3014354066985647,
"grad_norm": 0.17997704446315765,
"learning_rate": 3.7164301046341696e-05,
"loss": 0.80126953125,
"step": 272
},
{
"epoch": 1.30622009569378,
"grad_norm": 0.16072958707809448,
"learning_rate": 3.6837927565794714e-05,
"loss": 0.80810546875,
"step": 273
},
{
"epoch": 1.3110047846889952,
"grad_norm": 0.18286632001399994,
"learning_rate": 3.651269140313326e-05,
"loss": 0.79248046875,
"step": 274
},
{
"epoch": 1.3157894736842106,
"grad_norm": 0.17933203279972076,
"learning_rate": 3.61886129248054e-05,
"loss": 0.79443359375,
"step": 275
},
{
"epoch": 1.3205741626794258,
"grad_norm": 0.15538223087787628,
"learning_rate": 3.586571242476441e-05,
"loss": 0.794921875,
"step": 276
},
{
"epoch": 1.325358851674641,
"grad_norm": 0.1968618482351303,
"learning_rate": 3.554401012319801e-05,
"loss": 0.82763671875,
"step": 277
},
{
"epoch": 1.3301435406698565,
"grad_norm": 0.17610980570316315,
"learning_rate": 3.5223526165262176e-05,
"loss": 0.7880859375,
"step": 278
},
{
"epoch": 1.3349282296650717,
"grad_norm": 0.19299069046974182,
"learning_rate": 3.490428061981952e-05,
"loss": 0.77734375,
"step": 279
},
{
"epoch": 1.339712918660287,
"grad_norm": 0.15427900850772858,
"learning_rate": 3.4586293478182725e-05,
"loss": 0.7939453125,
"step": 280
},
{
"epoch": 1.3444976076555024,
"grad_norm": 0.1793992519378662,
"learning_rate": 3.4269584652862595e-05,
"loss": 0.80859375,
"step": 281
},
{
"epoch": 1.3492822966507176,
"grad_norm": 0.17524075508117676,
"learning_rate": 3.3954173976321156e-05,
"loss": 0.78759765625,
"step": 282
},
{
"epoch": 1.354066985645933,
"grad_norm": 0.19400925934314728,
"learning_rate": 3.3640081199729676e-05,
"loss": 0.7431640625,
"step": 283
},
{
"epoch": 1.3588516746411483,
"grad_norm": 0.1983603835105896,
"learning_rate": 3.3327325991731935e-05,
"loss": 0.82666015625,
"step": 284
},
{
"epoch": 1.3636363636363638,
"grad_norm": 0.19066844880580902,
"learning_rate": 3.301592793721249e-05,
"loss": 0.79150390625,
"step": 285
},
{
"epoch": 1.368421052631579,
"grad_norm": 0.18725162744522095,
"learning_rate": 3.270590653607023e-05,
"loss": 0.7890625,
"step": 286
},
{
"epoch": 1.3732057416267942,
"grad_norm": 0.1936885416507721,
"learning_rate": 3.239728120199741e-05,
"loss": 0.7568359375,
"step": 287
},
{
"epoch": 1.3779904306220097,
"grad_norm": 0.17952080070972443,
"learning_rate": 3.209007126126385e-05,
"loss": 0.79541015625,
"step": 288
},
{
"epoch": 1.3827751196172249,
"grad_norm": 0.1761673539876938,
"learning_rate": 3.178429595150676e-05,
"loss": 0.78564453125,
"step": 289
},
{
"epoch": 1.38755980861244,
"grad_norm": 0.16253837943077087,
"learning_rate": 3.1479974420526006e-05,
"loss": 0.85595703125,
"step": 290
},
{
"epoch": 1.3923444976076556,
"grad_norm": 0.16691721975803375,
"learning_rate": 3.1177125725085185e-05,
"loss": 0.78369140625,
"step": 291
},
{
"epoch": 1.3971291866028708,
"grad_norm": 0.1457083523273468,
"learning_rate": 3.087576882971814e-05,
"loss": 0.805908203125,
"step": 292
},
{
"epoch": 1.401913875598086,
"grad_norm": 0.16206428408622742,
"learning_rate": 3.057592260554151e-05,
"loss": 0.77783203125,
"step": 293
},
{
"epoch": 1.4066985645933014,
"grad_norm": 0.18252208828926086,
"learning_rate": 3.0277605829072952e-05,
"loss": 0.79296875,
"step": 294
},
{
"epoch": 1.4114832535885167,
"grad_norm": 0.15930020809173584,
"learning_rate": 2.998083718105536e-05,
"loss": 0.8466796875,
"step": 295
},
{
"epoch": 1.4162679425837321,
"grad_norm": 0.1903473138809204,
"learning_rate": 2.968563524528703e-05,
"loss": 0.76171875,
"step": 296
},
{
"epoch": 1.4210526315789473,
"grad_norm": 0.18638069927692413,
"learning_rate": 2.9392018507458006e-05,
"loss": 0.7958984375,
"step": 297
},
{
"epoch": 1.4258373205741628,
"grad_norm": 0.17060671746730804,
"learning_rate": 2.9100005353992456e-05,
"loss": 0.849609375,
"step": 298
},
{
"epoch": 1.430622009569378,
"grad_norm": 0.14795249700546265,
"learning_rate": 2.8809614070897296e-05,
"loss": 0.70068359375,
"step": 299
},
{
"epoch": 1.4354066985645932,
"grad_norm": 0.1723978966474533,
"learning_rate": 2.8520862842617137e-05,
"loss": 0.79150390625,
"step": 300
},
{
"epoch": 1.4401913875598087,
"grad_norm": 0.17499007284641266,
"learning_rate": 2.823376975089554e-05,
"loss": 0.81494140625,
"step": 301
},
{
"epoch": 1.444976076555024,
"grad_norm": 0.1675194799900055,
"learning_rate": 2.7948352773642767e-05,
"loss": 0.7890625,
"step": 302
},
{
"epoch": 1.4497607655502391,
"grad_norm": 0.17450399696826935,
"learning_rate": 2.7664629783809915e-05,
"loss": 0.767578125,
"step": 303
},
{
"epoch": 1.4545454545454546,
"grad_norm": 0.15929684042930603,
"learning_rate": 2.73826185482698e-05,
"loss": 0.7548828125,
"step": 304
},
{
"epoch": 1.4593301435406698,
"grad_norm": 0.18379977345466614,
"learning_rate": 2.7102336726704346e-05,
"loss": 0.7890625,
"step": 305
},
{
"epoch": 1.464114832535885,
"grad_norm": 0.16011302173137665,
"learning_rate": 2.682380187049874e-05,
"loss": 0.863037109375,
"step": 306
},
{
"epoch": 1.4688995215311005,
"grad_norm": 0.1894303858280182,
"learning_rate": 2.6547031421642254e-05,
"loss": 0.76171875,
"step": 307
},
{
"epoch": 1.4736842105263157,
"grad_norm": 0.1478191614151001,
"learning_rate": 2.6272042711636242e-05,
"loss": 0.720458984375,
"step": 308
},
{
"epoch": 1.4784688995215312,
"grad_norm": 0.20156490802764893,
"learning_rate": 2.599885296040857e-05,
"loss": 0.7412109375,
"step": 309
},
{
"epoch": 1.4832535885167464,
"grad_norm": 0.17944565415382385,
"learning_rate": 2.5727479275235506e-05,
"loss": 0.77685546875,
"step": 310
},
{
"epoch": 1.4880382775119618,
"grad_norm": 0.19786421954631805,
"learning_rate": 2.5457938649670333e-05,
"loss": 0.77978515625,
"step": 311
},
{
"epoch": 1.492822966507177,
"grad_norm": 0.18372617661952972,
"learning_rate": 2.5190247962479242e-05,
"loss": 0.83203125,
"step": 312
},
{
"epoch": 1.4976076555023923,
"grad_norm": 0.17823582887649536,
"learning_rate": 2.4924423976584422e-05,
"loss": 0.76171875,
"step": 313
},
{
"epoch": 1.5023923444976077,
"grad_norm": 0.23215219378471375,
"learning_rate": 2.4660483338014212e-05,
"loss": 0.7861328125,
"step": 314
},
{
"epoch": 1.507177033492823,
"grad_norm": 0.15614977478981018,
"learning_rate": 2.4398442574860948e-05,
"loss": 0.826171875,
"step": 315
},
{
"epoch": 1.5119617224880382,
"grad_norm": 0.17831632494926453,
"learning_rate": 2.4138318096245716e-05,
"loss": 0.7841796875,
"step": 316
},
{
"epoch": 1.5167464114832536,
"grad_norm": 0.17954230308532715,
"learning_rate": 2.3880126191291014e-05,
"loss": 0.82275390625,
"step": 317
},
{
"epoch": 1.5215311004784688,
"grad_norm": 0.17791250348091125,
"learning_rate": 2.362388302810053e-05,
"loss": 0.8046875,
"step": 318
},
{
"epoch": 1.526315789473684,
"grad_norm": 0.16783450543880463,
"learning_rate": 2.3369604652746925e-05,
"loss": 0.79052734375,
"step": 319
},
{
"epoch": 1.5311004784688995,
"grad_norm": 0.18951308727264404,
"learning_rate": 2.3117306988266753e-05,
"loss": 0.763671875,
"step": 320
},
{
"epoch": 1.535885167464115,
"grad_norm": 0.17476274073123932,
"learning_rate": 2.2867005833663535e-05,
"loss": 0.7998046875,
"step": 321
},
{
"epoch": 1.5406698564593302,
"grad_norm": 0.15480978786945343,
"learning_rate": 2.261871686291838e-05,
"loss": 0.8359375,
"step": 322
},
{
"epoch": 1.5454545454545454,
"grad_norm": 0.18359163403511047,
"learning_rate": 2.237245562400843e-05,
"loss": 0.76025390625,
"step": 323
},
{
"epoch": 1.5502392344497609,
"grad_norm": 0.1858016848564148,
"learning_rate": 2.2128237537933288e-05,
"loss": 0.74853515625,
"step": 324
},
{
"epoch": 1.555023923444976,
"grad_norm": 0.18103976547718048,
"learning_rate": 2.1886077897749263e-05,
"loss": 0.73486328125,
"step": 325
},
{
"epoch": 1.5598086124401913,
"grad_norm": 0.1770430952310562,
"learning_rate": 2.1645991867611875e-05,
"loss": 0.83642578125,
"step": 326
},
{
"epoch": 1.5645933014354068,
"grad_norm": 0.1885480433702469,
"learning_rate": 2.1407994481826075e-05,
"loss": 0.74609375,
"step": 327
},
{
"epoch": 1.569377990430622,
"grad_norm": 0.2102678418159485,
"learning_rate": 2.1172100643904923e-05,
"loss": 0.75537109375,
"step": 328
},
{
"epoch": 1.5741626794258372,
"grad_norm": 0.15284086763858795,
"learning_rate": 2.0938325125636278e-05,
"loss": 0.75244140625,
"step": 329
},
{
"epoch": 1.5789473684210527,
"grad_norm": 0.17530620098114014,
"learning_rate": 2.070668256615777e-05,
"loss": 0.82080078125,
"step": 330
},
{
"epoch": 1.583732057416268,
"grad_norm": 0.19387437403202057,
"learning_rate": 2.047718747104009e-05,
"loss": 0.81298828125,
"step": 331
},
{
"epoch": 1.588516746411483,
"grad_norm": 0.18224063515663147,
"learning_rate": 2.0249854211378727e-05,
"loss": 0.7744140625,
"step": 332
},
{
"epoch": 1.5933014354066986,
"grad_norm": 0.16329993307590485,
"learning_rate": 2.0024697022893874e-05,
"loss": 0.712890625,
"step": 333
},
{
"epoch": 1.598086124401914,
"grad_norm": 0.17031338810920715,
"learning_rate": 1.980173000503916e-05,
"loss": 0.8125,
"step": 334
},
{
"epoch": 1.6028708133971292,
"grad_norm": 0.17068995535373688,
"learning_rate": 1.9580967120118587e-05,
"loss": 0.80908203125,
"step": 335
},
{
"epoch": 1.6028708133971292,
"eval_loss": 0.823486328125,
"eval_runtime": 6.6013,
"eval_samples_per_second": 1.06,
"eval_steps_per_second": 0.151,
"step": 335
},
{
"epoch": 1.6076555023923444,
"grad_norm": 0.17699700593948364,
"learning_rate": 1.9362422192412356e-05,
"loss": 0.73193359375,
"step": 336
},
{
"epoch": 1.61244019138756,
"grad_norm": 0.22851988673210144,
"learning_rate": 1.9146108907311005e-05,
"loss": 0.79150390625,
"step": 337
},
{
"epoch": 1.6172248803827751,
"grad_norm": 0.19925302267074585,
"learning_rate": 1.893204081045859e-05,
"loss": 0.7587890625,
"step": 338
},
{
"epoch": 1.6220095693779903,
"grad_norm": 0.18237553536891937,
"learning_rate": 1.8720231306904346e-05,
"loss": 0.81005859375,
"step": 339
},
{
"epoch": 1.6267942583732058,
"grad_norm": 0.16293460130691528,
"learning_rate": 1.8510693660263283e-05,
"loss": 0.8125,
"step": 340
},
{
"epoch": 1.631578947368421,
"grad_norm": 0.1894524097442627,
"learning_rate": 1.830344099188565e-05,
"loss": 0.74560546875,
"step": 341
},
{
"epoch": 1.6363636363636362,
"grad_norm": 0.1865224838256836,
"learning_rate": 1.8098486280035172e-05,
"loss": 0.80712890625,
"step": 342
},
{
"epoch": 1.6411483253588517,
"grad_norm": 0.16558879613876343,
"learning_rate": 1.7895842359076488e-05,
"loss": 0.80029296875,
"step": 343
},
{
"epoch": 1.6459330143540671,
"grad_norm": 0.17985814809799194,
"learning_rate": 1.7695521918671315e-05,
"loss": 0.75634765625,
"step": 344
},
{
"epoch": 1.6507177033492821,
"grad_norm": 0.15998484194278717,
"learning_rate": 1.7497537502983906e-05,
"loss": 0.758544921875,
"step": 345
},
{
"epoch": 1.6555023923444976,
"grad_norm": 0.16847531497478485,
"learning_rate": 1.73019015098955e-05,
"loss": 0.7998046875,
"step": 346
},
{
"epoch": 1.660287081339713,
"grad_norm": 0.17358282208442688,
"learning_rate": 1.7108626190227966e-05,
"loss": 0.80810546875,
"step": 347
},
{
"epoch": 1.6650717703349283,
"grad_norm": 0.1544700413942337,
"learning_rate": 1.6917723646976615e-05,
"loss": 0.779296875,
"step": 348
},
{
"epoch": 1.6698564593301435,
"grad_norm": 0.1284680813550949,
"learning_rate": 1.672920583455237e-05,
"loss": 0.637939453125,
"step": 349
},
{
"epoch": 1.674641148325359,
"grad_norm": 0.1909414529800415,
"learning_rate": 1.6543084558033112e-05,
"loss": 0.7666015625,
"step": 350
},
{
"epoch": 1.6794258373205742,
"grad_norm": 0.1910911351442337,
"learning_rate": 1.6359371472424493e-05,
"loss": 0.77099609375,
"step": 351
},
{
"epoch": 1.6842105263157894,
"grad_norm": 0.17626284062862396,
"learning_rate": 1.6178078081930058e-05,
"loss": 0.81201171875,
"step": 352
},
{
"epoch": 1.6889952153110048,
"grad_norm": 0.17833667993545532,
"learning_rate": 1.5999215739230856e-05,
"loss": 0.78173828125,
"step": 353
},
{
"epoch": 1.69377990430622,
"grad_norm": 0.17231453955173492,
"learning_rate": 1.582279564477453e-05,
"loss": 0.7607421875,
"step": 354
},
{
"epoch": 1.6985645933014353,
"grad_norm": 0.19059370458126068,
"learning_rate": 1.564882884607393e-05,
"loss": 0.759765625,
"step": 355
},
{
"epoch": 1.7033492822966507,
"grad_norm": 0.20854324102401733,
"learning_rate": 1.5477326237015343e-05,
"loss": 0.76708984375,
"step": 356
},
{
"epoch": 1.7081339712918662,
"grad_norm": 0.1557876467704773,
"learning_rate": 1.530829855717628e-05,
"loss": 0.6611328125,
"step": 357
},
{
"epoch": 1.7129186602870812,
"grad_norm": 0.1856634020805359,
"learning_rate": 1.514175639115298e-05,
"loss": 0.77587890625,
"step": 358
},
{
"epoch": 1.7177033492822966,
"grad_norm": 0.16173864901065826,
"learning_rate": 1.4977710167897552e-05,
"loss": 0.80810546875,
"step": 359
},
{
"epoch": 1.722488038277512,
"grad_norm": 0.15028072893619537,
"learning_rate": 1.481617016006501e-05,
"loss": 0.7861328125,
"step": 360
},
{
"epoch": 1.7272727272727273,
"grad_norm": 0.16341842710971832,
"learning_rate": 1.465714648336986e-05,
"loss": 0.7880859375,
"step": 361
},
{
"epoch": 1.7320574162679425,
"grad_norm": 0.177590012550354,
"learning_rate": 1.4500649095952762e-05,
"loss": 0.78515625,
"step": 362
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.1711808741092682,
"learning_rate": 1.4346687797756886e-05,
"loss": 0.791015625,
"step": 363
},
{
"epoch": 1.7416267942583732,
"grad_norm": 0.17956730723381042,
"learning_rate": 1.4195272229914247e-05,
"loss": 0.79931640625,
"step": 364
},
{
"epoch": 1.7464114832535884,
"grad_norm": 0.18436218798160553,
"learning_rate": 1.4046411874142001e-05,
"loss": 0.78955078125,
"step": 365
},
{
"epoch": 1.7511961722488039,
"grad_norm": 0.17020408809185028,
"learning_rate": 1.3900116052148632e-05,
"loss": 0.80908203125,
"step": 366
},
{
"epoch": 1.755980861244019,
"grad_norm": 0.18755203485488892,
"learning_rate": 1.3756393925050293e-05,
"loss": 0.81201171875,
"step": 367
},
{
"epoch": 1.7607655502392343,
"grad_norm": 0.17153725028038025,
"learning_rate": 1.361525449279708e-05,
"loss": 0.814697265625,
"step": 368
},
{
"epoch": 1.7655502392344498,
"grad_norm": 0.19658948481082916,
"learning_rate": 1.3476706593609479e-05,
"loss": 0.678466796875,
"step": 369
},
{
"epoch": 1.7703349282296652,
"grad_norm": 0.1711190640926361,
"learning_rate": 1.3340758903424912e-05,
"loss": 0.7666015625,
"step": 370
},
{
"epoch": 1.7751196172248804,
"grad_norm": 0.18621385097503662,
"learning_rate": 1.3207419935354437e-05,
"loss": 0.8125,
"step": 371
},
{
"epoch": 1.7799043062200957,
"grad_norm": 0.18893972039222717,
"learning_rate": 1.3076698039149621e-05,
"loss": 0.81201171875,
"step": 372
},
{
"epoch": 1.784688995215311,
"grad_norm": 0.20575562119483948,
"learning_rate": 1.294860140067975e-05,
"loss": 0.7373046875,
"step": 373
},
{
"epoch": 1.7894736842105263,
"grad_norm": 0.18274357914924622,
"learning_rate": 1.282313804141916e-05,
"loss": 0.79052734375,
"step": 374
},
{
"epoch": 1.7942583732057416,
"grad_norm": 0.17074786126613617,
"learning_rate": 1.2700315817944964e-05,
"loss": 0.85595703125,
"step": 375
},
{
"epoch": 1.799043062200957,
"grad_norm": 0.15272174775600433,
"learning_rate": 1.2580142421445024e-05,
"loss": 0.78369140625,
"step": 376
},
{
"epoch": 1.8038277511961722,
"grad_norm": 0.16635139286518097,
"learning_rate": 1.2462625377236414e-05,
"loss": 0.579345703125,
"step": 377
},
{
"epoch": 1.8086124401913874,
"grad_norm": 0.19089093804359436,
"learning_rate": 1.2347772044294074e-05,
"loss": 0.751953125,
"step": 378
},
{
"epoch": 1.813397129186603,
"grad_norm": 0.1820306032896042,
"learning_rate": 1.2235589614790056e-05,
"loss": 0.76318359375,
"step": 379
},
{
"epoch": 1.8181818181818183,
"grad_norm": 0.17462246119976044,
"learning_rate": 1.2126085113643134e-05,
"loss": 0.75,
"step": 380
},
{
"epoch": 1.8229665071770333,
"grad_norm": 0.19027724862098694,
"learning_rate": 1.2019265398078883e-05,
"loss": 0.7275390625,
"step": 381
},
{
"epoch": 1.8277511961722488,
"grad_norm": 0.1650601029396057,
"learning_rate": 1.1915137157200304e-05,
"loss": 0.8056640625,
"step": 382
},
{
"epoch": 1.8325358851674642,
"grad_norm": 0.18004824221134186,
"learning_rate": 1.181370691156891e-05,
"loss": 0.7744140625,
"step": 383
},
{
"epoch": 1.8373205741626795,
"grad_norm": 0.19831469655036926,
"learning_rate": 1.171498101279645e-05,
"loss": 0.7734375,
"step": 384
},
{
"epoch": 1.8421052631578947,
"grad_norm": 0.18922831118106842,
"learning_rate": 1.1618965643147134e-05,
"loss": 0.8095703125,
"step": 385
},
{
"epoch": 1.8468899521531101,
"grad_norm": 0.17712976038455963,
"learning_rate": 1.1525666815150516e-05,
"loss": 0.759765625,
"step": 386
},
{
"epoch": 1.8516746411483254,
"grad_norm": 0.17649629712104797,
"learning_rate": 1.1435090371224953e-05,
"loss": 0.8046875,
"step": 387
},
{
"epoch": 1.8564593301435406,
"grad_norm": 0.17713221907615662,
"learning_rate": 1.134724198331183e-05,
"loss": 0.85107421875,
"step": 388
},
{
"epoch": 1.861244019138756,
"grad_norm": 0.2169307917356491,
"learning_rate": 1.1262127152520268e-05,
"loss": 0.75146484375,
"step": 389
},
{
"epoch": 1.8660287081339713,
"grad_norm": 0.16737069189548492,
"learning_rate": 1.1179751208782736e-05,
"loss": 0.83251953125,
"step": 390
},
{
"epoch": 1.8708133971291865,
"grad_norm": 0.17925404012203217,
"learning_rate": 1.1100119310521241e-05,
"loss": 0.79345703125,
"step": 391
},
{
"epoch": 1.875598086124402,
"grad_norm": 0.15294474363327026,
"learning_rate": 1.1023236444324319e-05,
"loss": 0.662841796875,
"step": 392
},
{
"epoch": 1.8803827751196174,
"grad_norm": 0.1851387619972229,
"learning_rate": 1.0949107424634761e-05,
"loss": 0.801513671875,
"step": 393
},
{
"epoch": 1.8851674641148324,
"grad_norm": 0.22087079286575317,
"learning_rate": 1.0877736893448146e-05,
"loss": 0.75634765625,
"step": 394
},
{
"epoch": 1.8899521531100478,
"grad_norm": 0.19715087115764618,
"learning_rate": 1.0809129320022162e-05,
"loss": 0.783203125,
"step": 395
},
{
"epoch": 1.8947368421052633,
"grad_norm": 0.1615232527256012,
"learning_rate": 1.0743289000596695e-05,
"loss": 0.7705078125,
"step": 396
},
{
"epoch": 1.8995215311004785,
"grad_norm": 0.18735022842884064,
"learning_rate": 1.0680220058124857e-05,
"loss": 0.7421875,
"step": 397
},
{
"epoch": 1.9043062200956937,
"grad_norm": 0.18647435307502747,
"learning_rate": 1.0619926442014758e-05,
"loss": 0.78515625,
"step": 398
},
{
"epoch": 1.9090909090909092,
"grad_norm": 0.1795354187488556,
"learning_rate": 1.0562411927882231e-05,
"loss": 0.7861328125,
"step": 399
},
{
"epoch": 1.9138755980861244,
"grad_norm": 0.15958954393863678,
"learning_rate": 1.0507680117314347e-05,
"loss": 0.66650390625,
"step": 400
},
{
"epoch": 1.9186602870813396,
"grad_norm": 0.1826360523700714,
"learning_rate": 1.045573443764395e-05,
"loss": 0.6036376953125,
"step": 401
},
{
"epoch": 1.923444976076555,
"grad_norm": 0.19836081564426422,
"learning_rate": 1.0406578141734968e-05,
"loss": 0.79296875,
"step": 402
},
{
"epoch": 1.9282296650717703,
"grad_norm": 0.18211688101291656,
"learning_rate": 1.0360214307778779e-05,
"loss": 0.78466796875,
"step": 403
},
{
"epoch": 1.9330143540669855,
"grad_norm": 0.1684381514787674,
"learning_rate": 1.0316645839101402e-05,
"loss": 0.84619140625,
"step": 404
},
{
"epoch": 1.937799043062201,
"grad_norm": 0.19366608560085297,
"learning_rate": 1.0275875463981722e-05,
"loss": 0.7548828125,
"step": 405
},
{
"epoch": 1.9425837320574164,
"grad_norm": 0.19184112548828125,
"learning_rate": 1.023790573548063e-05,
"loss": 0.7685546875,
"step": 406
},
{
"epoch": 1.9473684210526314,
"grad_norm": 0.20356374979019165,
"learning_rate": 1.0202739031281154e-05,
"loss": 0.77392578125,
"step": 407
},
{
"epoch": 1.9521531100478469,
"grad_norm": 0.16546234488487244,
"learning_rate": 1.0170377553539548e-05,
"loss": 0.84423828125,
"step": 408
},
{
"epoch": 1.9569377990430623,
"grad_norm": 0.16989609599113464,
"learning_rate": 1.0140823328747442e-05,
"loss": 0.798828125,
"step": 409
},
{
"epoch": 1.9617224880382775,
"grad_norm": 0.1848805546760559,
"learning_rate": 1.011407820760488e-05,
"loss": 0.74072265625,
"step": 410
},
{
"epoch": 1.9665071770334928,
"grad_norm": 0.1985882818698883,
"learning_rate": 1.0090143864904471e-05,
"loss": 0.81787109375,
"step": 411
},
{
"epoch": 1.9712918660287082,
"grad_norm": 0.2124672383069992,
"learning_rate": 1.0069021799426498e-05,
"loss": 0.7373046875,
"step": 412
},
{
"epoch": 1.9760765550239234,
"grad_norm": 0.18534432351589203,
"learning_rate": 1.0050713333845062e-05,
"loss": 0.7841796875,
"step": 413
},
{
"epoch": 1.9808612440191387,
"grad_norm": 0.18089644610881805,
"learning_rate": 1.0035219614645259e-05,
"loss": 0.822265625,
"step": 414
},
{
"epoch": 1.985645933014354,
"grad_norm": 0.25077196955680847,
"learning_rate": 1.0022541612051384e-05,
"loss": 0.75390625,
"step": 415
},
{
"epoch": 1.9904306220095693,
"grad_norm": 0.18322627246379852,
"learning_rate": 1.0012680119966181e-05,
"loss": 0.8486328125,
"step": 416
},
{
"epoch": 1.9952153110047846,
"grad_norm": 0.19022634625434875,
"learning_rate": 1.0005635755921122e-05,
"loss": 0.8056640625,
"step": 417
},
{
"epoch": 2.0,
"grad_norm": 0.20409126579761505,
"learning_rate": 1.0001408961037736e-05,
"loss": 0.79931640625,
"step": 418
},
{
"epoch": 2.0,
"eval_loss": 0.8175048828125,
"eval_runtime": 6.3173,
"eval_samples_per_second": 1.108,
"eval_steps_per_second": 0.158,
"step": 418
}
],
"logging_steps": 1.0,
"max_steps": 418,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"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": 3.435107961085205e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}