fengyao1909's picture
Upload folder using huggingface_hub
e8015d6 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.3333333333333333,
"eval_steps": 500,
"global_step": 157,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0021231422505307855,
"grad_norm": 6.233692311689662,
"learning_rate": 0.0,
"loss": 1.3677,
"step": 1
},
{
"epoch": 0.004246284501061571,
"grad_norm": 6.03613913611666,
"learning_rate": 2.0833333333333333e-07,
"loss": 1.4092,
"step": 2
},
{
"epoch": 0.006369426751592357,
"grad_norm": 5.843878249582451,
"learning_rate": 4.1666666666666667e-07,
"loss": 1.4005,
"step": 3
},
{
"epoch": 0.008492569002123142,
"grad_norm": 6.2100854445399305,
"learning_rate": 6.25e-07,
"loss": 1.4027,
"step": 4
},
{
"epoch": 0.010615711252653927,
"grad_norm": 5.700197750457959,
"learning_rate": 8.333333333333333e-07,
"loss": 1.3859,
"step": 5
},
{
"epoch": 0.012738853503184714,
"grad_norm": 5.708735743112818,
"learning_rate": 1.0416666666666667e-06,
"loss": 1.4135,
"step": 6
},
{
"epoch": 0.014861995753715499,
"grad_norm": 5.584947016091844,
"learning_rate": 1.25e-06,
"loss": 1.3083,
"step": 7
},
{
"epoch": 0.016985138004246284,
"grad_norm": 5.5017514099891125,
"learning_rate": 1.4583333333333335e-06,
"loss": 1.3761,
"step": 8
},
{
"epoch": 0.01910828025477707,
"grad_norm": 5.240707118749176,
"learning_rate": 1.6666666666666667e-06,
"loss": 1.374,
"step": 9
},
{
"epoch": 0.021231422505307854,
"grad_norm": 5.159923266283754,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.3837,
"step": 10
},
{
"epoch": 0.02335456475583864,
"grad_norm": 4.23439034643152,
"learning_rate": 2.0833333333333334e-06,
"loss": 1.3278,
"step": 11
},
{
"epoch": 0.025477707006369428,
"grad_norm": 4.23403012426509,
"learning_rate": 2.2916666666666666e-06,
"loss": 1.3095,
"step": 12
},
{
"epoch": 0.027600849256900213,
"grad_norm": 4.033469595815775,
"learning_rate": 2.5e-06,
"loss": 1.3736,
"step": 13
},
{
"epoch": 0.029723991507430998,
"grad_norm": 3.4140671410294092,
"learning_rate": 2.7083333333333334e-06,
"loss": 1.2966,
"step": 14
},
{
"epoch": 0.03184713375796178,
"grad_norm": 3.2252832240083946,
"learning_rate": 2.916666666666667e-06,
"loss": 1.3211,
"step": 15
},
{
"epoch": 0.03397027600849257,
"grad_norm": 2.958759736658132,
"learning_rate": 3.125e-06,
"loss": 1.284,
"step": 16
},
{
"epoch": 0.036093418259023353,
"grad_norm": 3.0082219342870764,
"learning_rate": 3.3333333333333333e-06,
"loss": 1.2574,
"step": 17
},
{
"epoch": 0.03821656050955414,
"grad_norm": 2.7607212501829186,
"learning_rate": 3.5416666666666673e-06,
"loss": 1.2222,
"step": 18
},
{
"epoch": 0.040339702760084924,
"grad_norm": 1.7794382970826705,
"learning_rate": 3.7500000000000005e-06,
"loss": 1.1952,
"step": 19
},
{
"epoch": 0.04246284501061571,
"grad_norm": 1.9000490101749923,
"learning_rate": 3.958333333333333e-06,
"loss": 1.0968,
"step": 20
},
{
"epoch": 0.044585987261146494,
"grad_norm": 2.184749467485857,
"learning_rate": 4.166666666666667e-06,
"loss": 1.2058,
"step": 21
},
{
"epoch": 0.04670912951167728,
"grad_norm": 2.076583497979509,
"learning_rate": 4.3750000000000005e-06,
"loss": 1.1726,
"step": 22
},
{
"epoch": 0.04883227176220807,
"grad_norm": 1.6762408968746727,
"learning_rate": 4.583333333333333e-06,
"loss": 1.0816,
"step": 23
},
{
"epoch": 0.050955414012738856,
"grad_norm": 1.655763485736625,
"learning_rate": 4.791666666666668e-06,
"loss": 1.0881,
"step": 24
},
{
"epoch": 0.05307855626326964,
"grad_norm": 1.404756822747019,
"learning_rate": 5e-06,
"loss": 1.111,
"step": 25
},
{
"epoch": 0.055201698513800426,
"grad_norm": 1.7450925608231072,
"learning_rate": 4.999938256261134e-06,
"loss": 1.0968,
"step": 26
},
{
"epoch": 0.05732484076433121,
"grad_norm": 1.7455723981125268,
"learning_rate": 4.999753028094368e-06,
"loss": 1.135,
"step": 27
},
{
"epoch": 0.059447983014861996,
"grad_norm": 1.765180118972071,
"learning_rate": 4.999444324649045e-06,
"loss": 1.1332,
"step": 28
},
{
"epoch": 0.06157112526539278,
"grad_norm": 1.7025986108773807,
"learning_rate": 4.99901216117357e-06,
"loss": 1.1037,
"step": 29
},
{
"epoch": 0.06369426751592357,
"grad_norm": 1.6843781660517996,
"learning_rate": 4.998456559014653e-06,
"loss": 1.1036,
"step": 30
},
{
"epoch": 0.06581740976645435,
"grad_norm": 1.4971568544476521,
"learning_rate": 4.997777545616258e-06,
"loss": 1.0896,
"step": 31
},
{
"epoch": 0.06794055201698514,
"grad_norm": 1.2471663073080574,
"learning_rate": 4.996975154518245e-06,
"loss": 1.0816,
"step": 32
},
{
"epoch": 0.07006369426751592,
"grad_norm": 1.1099200791707107,
"learning_rate": 4.996049425354717e-06,
"loss": 1.0408,
"step": 33
},
{
"epoch": 0.07218683651804671,
"grad_norm": 1.1596577803643031,
"learning_rate": 4.995000403852057e-06,
"loss": 1.0348,
"step": 34
},
{
"epoch": 0.07430997876857749,
"grad_norm": 0.9836703479205239,
"learning_rate": 4.993828141826672e-06,
"loss": 1.0397,
"step": 35
},
{
"epoch": 0.07643312101910828,
"grad_norm": 4.973840678082251,
"learning_rate": 4.992532697182434e-06,
"loss": 1.0692,
"step": 36
},
{
"epoch": 0.07855626326963906,
"grad_norm": 0.9322137357537782,
"learning_rate": 4.991114133907822e-06,
"loss": 1.0056,
"step": 37
},
{
"epoch": 0.08067940552016985,
"grad_norm": 1.0203537152829516,
"learning_rate": 4.989572522072753e-06,
"loss": 0.9956,
"step": 38
},
{
"epoch": 0.08280254777070063,
"grad_norm": 1.0385035368583873,
"learning_rate": 4.9879079378251325e-06,
"loss": 0.9772,
"step": 39
},
{
"epoch": 0.08492569002123142,
"grad_norm": 0.9097103872410344,
"learning_rate": 4.986120463387084e-06,
"loss": 1.0256,
"step": 40
},
{
"epoch": 0.0870488322717622,
"grad_norm": 1.0611463152500384,
"learning_rate": 4.984210187050891e-06,
"loss": 0.9883,
"step": 41
},
{
"epoch": 0.08917197452229299,
"grad_norm": 0.779249244150645,
"learning_rate": 4.982177203174636e-06,
"loss": 0.9847,
"step": 42
},
{
"epoch": 0.09129511677282377,
"grad_norm": 0.7658008659883724,
"learning_rate": 4.9800216121775404e-06,
"loss": 0.9829,
"step": 43
},
{
"epoch": 0.09341825902335456,
"grad_norm": 0.7598772913633557,
"learning_rate": 4.977743520535001e-06,
"loss": 1.0496,
"step": 44
},
{
"epoch": 0.09554140127388536,
"grad_norm": 0.8141928977386432,
"learning_rate": 4.975343040773335e-06,
"loss": 1.0519,
"step": 45
},
{
"epoch": 0.09766454352441614,
"grad_norm": 0.8911722172701033,
"learning_rate": 4.972820291464219e-06,
"loss": 1.0345,
"step": 46
},
{
"epoch": 0.09978768577494693,
"grad_norm": 0.8358128773624875,
"learning_rate": 4.970175397218832e-06,
"loss": 0.9987,
"step": 47
},
{
"epoch": 0.10191082802547771,
"grad_norm": 0.882046205562171,
"learning_rate": 4.967408488681702e-06,
"loss": 1.0004,
"step": 48
},
{
"epoch": 0.1040339702760085,
"grad_norm": 0.8308643665449907,
"learning_rate": 4.964519702524251e-06,
"loss": 0.9867,
"step": 49
},
{
"epoch": 0.10615711252653928,
"grad_norm": 0.8727814847927396,
"learning_rate": 4.9615091814380465e-06,
"loss": 1.0236,
"step": 50
},
{
"epoch": 0.10828025477707007,
"grad_norm": 0.8128090079858757,
"learning_rate": 4.958377074127751e-06,
"loss": 1.0096,
"step": 51
},
{
"epoch": 0.11040339702760085,
"grad_norm": 0.7041201796080805,
"learning_rate": 4.955123535303775e-06,
"loss": 1.0161,
"step": 52
},
{
"epoch": 0.11252653927813164,
"grad_norm": 0.7034976999278052,
"learning_rate": 4.951748725674643e-06,
"loss": 0.9444,
"step": 53
},
{
"epoch": 0.11464968152866242,
"grad_norm": 0.7111287475587301,
"learning_rate": 4.948252811939044e-06,
"loss": 1.0034,
"step": 54
},
{
"epoch": 0.11677282377919321,
"grad_norm": 0.7112458705577068,
"learning_rate": 4.944635966777607e-06,
"loss": 1.0413,
"step": 55
},
{
"epoch": 0.11889596602972399,
"grad_norm": 0.7351439597774859,
"learning_rate": 4.940898368844366e-06,
"loss": 1.0663,
"step": 56
},
{
"epoch": 0.12101910828025478,
"grad_norm": 0.7120598412888705,
"learning_rate": 4.937040202757937e-06,
"loss": 1.0058,
"step": 57
},
{
"epoch": 0.12314225053078556,
"grad_norm": 0.8009131327695405,
"learning_rate": 4.933061659092401e-06,
"loss": 1.0048,
"step": 58
},
{
"epoch": 0.12526539278131635,
"grad_norm": 0.8355353984744548,
"learning_rate": 4.928962934367887e-06,
"loss": 1.0005,
"step": 59
},
{
"epoch": 0.12738853503184713,
"grad_norm": 0.7292500219199575,
"learning_rate": 4.924744231040865e-06,
"loss": 0.9948,
"step": 60
},
{
"epoch": 0.12951167728237792,
"grad_norm": 0.7441720822069011,
"learning_rate": 4.920405757494147e-06,
"loss": 0.9203,
"step": 61
},
{
"epoch": 0.1316348195329087,
"grad_norm": 0.7164757410132886,
"learning_rate": 4.915947728026599e-06,
"loss": 0.9826,
"step": 62
},
{
"epoch": 0.1337579617834395,
"grad_norm": 0.6688811212861865,
"learning_rate": 4.911370362842544e-06,
"loss": 1.0041,
"step": 63
},
{
"epoch": 0.13588110403397027,
"grad_norm": 0.7276933998413905,
"learning_rate": 4.906673888040895e-06,
"loss": 1.0254,
"step": 64
},
{
"epoch": 0.13800424628450106,
"grad_norm": 0.8980292377631762,
"learning_rate": 4.901858535603982e-06,
"loss": 1.0341,
"step": 65
},
{
"epoch": 0.14012738853503184,
"grad_norm": 0.7213457650159387,
"learning_rate": 4.896924543386099e-06,
"loss": 0.9732,
"step": 66
},
{
"epoch": 0.14225053078556263,
"grad_norm": 0.6575608860005288,
"learning_rate": 4.891872155101746e-06,
"loss": 0.9758,
"step": 67
},
{
"epoch": 0.14437367303609341,
"grad_norm": 0.7227626661588851,
"learning_rate": 4.886701620313595e-06,
"loss": 0.9736,
"step": 68
},
{
"epoch": 0.1464968152866242,
"grad_norm": 0.7820534177881011,
"learning_rate": 4.88141319442017e-06,
"loss": 0.9519,
"step": 69
},
{
"epoch": 0.14861995753715498,
"grad_norm": 0.7464763101124039,
"learning_rate": 4.876007138643216e-06,
"loss": 1.0623,
"step": 70
},
{
"epoch": 0.15074309978768577,
"grad_norm": 0.8031125528787094,
"learning_rate": 4.870483720014814e-06,
"loss": 1.0023,
"step": 71
},
{
"epoch": 0.15286624203821655,
"grad_norm": 0.7252756266115659,
"learning_rate": 4.8648432113641765e-06,
"loss": 1.025,
"step": 72
},
{
"epoch": 0.15498938428874734,
"grad_norm": 0.7685738411441779,
"learning_rate": 4.8590858913041775e-06,
"loss": 0.9701,
"step": 73
},
{
"epoch": 0.15711252653927812,
"grad_norm": 0.7119059019537681,
"learning_rate": 4.853212044217591e-06,
"loss": 0.9471,
"step": 74
},
{
"epoch": 0.1592356687898089,
"grad_norm": 0.670302973142834,
"learning_rate": 4.847221960243042e-06,
"loss": 0.9918,
"step": 75
},
{
"epoch": 0.1613588110403397,
"grad_norm": 0.7274878719799441,
"learning_rate": 4.8411159352606735e-06,
"loss": 0.9114,
"step": 76
},
{
"epoch": 0.16348195329087048,
"grad_norm": 0.6706478978040786,
"learning_rate": 4.834894270877536e-06,
"loss": 0.9628,
"step": 77
},
{
"epoch": 0.16560509554140126,
"grad_norm": 0.7197495557043765,
"learning_rate": 4.828557274412686e-06,
"loss": 0.9792,
"step": 78
},
{
"epoch": 0.16772823779193205,
"grad_norm": 0.6740897649139851,
"learning_rate": 4.822105258882007e-06,
"loss": 0.978,
"step": 79
},
{
"epoch": 0.16985138004246284,
"grad_norm": 0.7298858435711825,
"learning_rate": 4.815538542982751e-06,
"loss": 1.0749,
"step": 80
},
{
"epoch": 0.17197452229299362,
"grad_norm": 0.7437812096942298,
"learning_rate": 4.808857451077788e-06,
"loss": 1.0027,
"step": 81
},
{
"epoch": 0.1740976645435244,
"grad_norm": 0.7793372891907893,
"learning_rate": 4.802062313179595e-06,
"loss": 1.056,
"step": 82
},
{
"epoch": 0.1762208067940552,
"grad_norm": 0.7200522013491976,
"learning_rate": 4.795153464933948e-06,
"loss": 0.9804,
"step": 83
},
{
"epoch": 0.17834394904458598,
"grad_norm": 0.6800845586517711,
"learning_rate": 4.7881312476033444e-06,
"loss": 1.0096,
"step": 84
},
{
"epoch": 0.18046709129511676,
"grad_norm": 0.6876028206701021,
"learning_rate": 4.7809960080501464e-06,
"loss": 0.9886,
"step": 85
},
{
"epoch": 0.18259023354564755,
"grad_norm": 0.7119685223963281,
"learning_rate": 4.773748098719448e-06,
"loss": 1.019,
"step": 86
},
{
"epoch": 0.18471337579617833,
"grad_norm": 0.6948951583741163,
"learning_rate": 4.766387877621667e-06,
"loss": 1.0177,
"step": 87
},
{
"epoch": 0.18683651804670912,
"grad_norm": 0.7252384478947084,
"learning_rate": 4.758915708314858e-06,
"loss": 0.952,
"step": 88
},
{
"epoch": 0.18895966029723993,
"grad_norm": 0.7132363877921689,
"learning_rate": 4.751331959886758e-06,
"loss": 0.9311,
"step": 89
},
{
"epoch": 0.1910828025477707,
"grad_norm": 0.6672381978657935,
"learning_rate": 4.743637006936552e-06,
"loss": 1.0019,
"step": 90
},
{
"epoch": 0.1932059447983015,
"grad_norm": 0.7074654491248366,
"learning_rate": 4.735831229556374e-06,
"loss": 0.9249,
"step": 91
},
{
"epoch": 0.19532908704883228,
"grad_norm": 0.8263034494745758,
"learning_rate": 4.727915013312527e-06,
"loss": 0.9627,
"step": 92
},
{
"epoch": 0.19745222929936307,
"grad_norm": 0.7564665006141713,
"learning_rate": 4.719888749226442e-06,
"loss": 0.9788,
"step": 93
},
{
"epoch": 0.19957537154989385,
"grad_norm": 0.7401712745437141,
"learning_rate": 4.711752833755362e-06,
"loss": 0.9595,
"step": 94
},
{
"epoch": 0.20169851380042464,
"grad_norm": 0.7177547311512174,
"learning_rate": 4.70350766877276e-06,
"loss": 1.0125,
"step": 95
},
{
"epoch": 0.20382165605095542,
"grad_norm": 0.8365049197183004,
"learning_rate": 4.695153661548486e-06,
"loss": 0.9617,
"step": 96
},
{
"epoch": 0.2059447983014862,
"grad_norm": 0.8275524843143772,
"learning_rate": 4.686691224728652e-06,
"loss": 0.9583,
"step": 97
},
{
"epoch": 0.208067940552017,
"grad_norm": 0.7849897218743559,
"learning_rate": 4.678120776315251e-06,
"loss": 0.9432,
"step": 98
},
{
"epoch": 0.21019108280254778,
"grad_norm": 0.7266338220692218,
"learning_rate": 4.669442739645506e-06,
"loss": 1.0112,
"step": 99
},
{
"epoch": 0.21231422505307856,
"grad_norm": 0.6786209555674868,
"learning_rate": 4.660657543370958e-06,
"loss": 1.0122,
"step": 100
},
{
"epoch": 0.21443736730360935,
"grad_norm": 0.7396143819036078,
"learning_rate": 4.651765621436303e-06,
"loss": 1.0491,
"step": 101
},
{
"epoch": 0.21656050955414013,
"grad_norm": 0.745708791916274,
"learning_rate": 4.642767413057942e-06,
"loss": 0.966,
"step": 102
},
{
"epoch": 0.21868365180467092,
"grad_norm": 0.7477607378275345,
"learning_rate": 4.6336633627023e-06,
"loss": 0.9129,
"step": 103
},
{
"epoch": 0.2208067940552017,
"grad_norm": 0.6930283267824098,
"learning_rate": 4.624453920063863e-06,
"loss": 0.9769,
"step": 104
},
{
"epoch": 0.2229299363057325,
"grad_norm": 0.6762283490594463,
"learning_rate": 4.6151395400429665e-06,
"loss": 0.9383,
"step": 105
},
{
"epoch": 0.22505307855626328,
"grad_norm": 0.7313509583041526,
"learning_rate": 4.605720682723331e-06,
"loss": 0.9428,
"step": 106
},
{
"epoch": 0.22717622080679406,
"grad_norm": 0.7156367191749474,
"learning_rate": 4.596197813349328e-06,
"loss": 0.9834,
"step": 107
},
{
"epoch": 0.22929936305732485,
"grad_norm": 0.6828464400232375,
"learning_rate": 4.586571402303006e-06,
"loss": 0.9504,
"step": 108
},
{
"epoch": 0.23142250530785563,
"grad_norm": 0.74716164026175,
"learning_rate": 4.576841925080853e-06,
"loss": 0.9998,
"step": 109
},
{
"epoch": 0.23354564755838642,
"grad_norm": 0.6856088912154495,
"learning_rate": 4.56700986227031e-06,
"loss": 1.0149,
"step": 110
},
{
"epoch": 0.2356687898089172,
"grad_norm": 0.7033606189520333,
"learning_rate": 4.557075699526032e-06,
"loss": 0.9584,
"step": 111
},
{
"epoch": 0.23779193205944799,
"grad_norm": 0.7043436392355145,
"learning_rate": 4.547039927545899e-06,
"loss": 0.9779,
"step": 112
},
{
"epoch": 0.23991507430997877,
"grad_norm": 0.8202453116913221,
"learning_rate": 4.536903042046778e-06,
"loss": 0.9505,
"step": 113
},
{
"epoch": 0.24203821656050956,
"grad_norm": 0.7436505838525592,
"learning_rate": 4.526665543740038e-06,
"loss": 0.9561,
"step": 114
},
{
"epoch": 0.24416135881104034,
"grad_norm": 0.6478361281734111,
"learning_rate": 4.516327938306818e-06,
"loss": 1.0084,
"step": 115
},
{
"epoch": 0.24628450106157113,
"grad_norm": 0.7674555846640165,
"learning_rate": 4.505890736373045e-06,
"loss": 0.9767,
"step": 116
},
{
"epoch": 0.2484076433121019,
"grad_norm": 0.7184796384578326,
"learning_rate": 4.495354453484216e-06,
"loss": 0.9801,
"step": 117
},
{
"epoch": 0.2505307855626327,
"grad_norm": 0.677609956598202,
"learning_rate": 4.4847196100799305e-06,
"loss": 0.9825,
"step": 118
},
{
"epoch": 0.2526539278131635,
"grad_norm": 0.7123651122668029,
"learning_rate": 4.473986731468183e-06,
"loss": 1.0107,
"step": 119
},
{
"epoch": 0.25477707006369427,
"grad_norm": 0.7156231222559138,
"learning_rate": 4.463156347799419e-06,
"loss": 1.0779,
"step": 120
},
{
"epoch": 0.25690021231422505,
"grad_norm": 0.7717775840138449,
"learning_rate": 4.452228994040341e-06,
"loss": 0.9622,
"step": 121
},
{
"epoch": 0.25902335456475584,
"grad_norm": 0.7512072014948016,
"learning_rate": 4.4412052099474916e-06,
"loss": 0.9625,
"step": 122
},
{
"epoch": 0.2611464968152866,
"grad_norm": 0.7124270062225315,
"learning_rate": 4.430085540040587e-06,
"loss": 1.0169,
"step": 123
},
{
"epoch": 0.2632696390658174,
"grad_norm": 0.7058601891740411,
"learning_rate": 4.418870533575626e-06,
"loss": 0.9819,
"step": 124
},
{
"epoch": 0.2653927813163482,
"grad_norm": 0.6903185636770165,
"learning_rate": 4.40756074451775e-06,
"loss": 0.9398,
"step": 125
},
{
"epoch": 0.267515923566879,
"grad_norm": 0.7046682682406727,
"learning_rate": 4.396156731513888e-06,
"loss": 1.0341,
"step": 126
},
{
"epoch": 0.26963906581740976,
"grad_norm": 0.7862567299960644,
"learning_rate": 4.384659057865165e-06,
"loss": 0.947,
"step": 127
},
{
"epoch": 0.27176220806794055,
"grad_norm": 0.6195482427854749,
"learning_rate": 4.373068291499065e-06,
"loss": 0.8899,
"step": 128
},
{
"epoch": 0.27388535031847133,
"grad_norm": 0.6985398850512,
"learning_rate": 4.36138500494139e-06,
"loss": 0.9726,
"step": 129
},
{
"epoch": 0.2760084925690021,
"grad_norm": 0.668756553679393,
"learning_rate": 4.349609775287977e-06,
"loss": 0.9855,
"step": 130
},
{
"epoch": 0.2781316348195329,
"grad_norm": 0.6581795928554366,
"learning_rate": 4.337743184176188e-06,
"loss": 0.9596,
"step": 131
},
{
"epoch": 0.2802547770700637,
"grad_norm": 0.6337202167292365,
"learning_rate": 4.325785817756186e-06,
"loss": 0.9839,
"step": 132
},
{
"epoch": 0.2823779193205945,
"grad_norm": 0.6680059124711757,
"learning_rate": 4.313738266661979e-06,
"loss": 0.9324,
"step": 133
},
{
"epoch": 0.28450106157112526,
"grad_norm": 0.6598271036519225,
"learning_rate": 4.301601125982246e-06,
"loss": 1.0066,
"step": 134
},
{
"epoch": 0.28662420382165604,
"grad_norm": 0.7183648150650218,
"learning_rate": 4.289374995230942e-06,
"loss": 1.0136,
"step": 135
},
{
"epoch": 0.28874734607218683,
"grad_norm": 0.742221341756798,
"learning_rate": 4.277060478317687e-06,
"loss": 1.0008,
"step": 136
},
{
"epoch": 0.2908704883227176,
"grad_norm": 0.6824304036868553,
"learning_rate": 4.264658183517935e-06,
"loss": 0.9381,
"step": 137
},
{
"epoch": 0.2929936305732484,
"grad_norm": 0.6505573173225403,
"learning_rate": 4.252168723442927e-06,
"loss": 0.9489,
"step": 138
},
{
"epoch": 0.2951167728237792,
"grad_norm": 0.6950634835305272,
"learning_rate": 4.23959271500943e-06,
"loss": 0.9373,
"step": 139
},
{
"epoch": 0.29723991507430997,
"grad_norm": 0.6872675526003413,
"learning_rate": 4.226930779409271e-06,
"loss": 0.9984,
"step": 140
},
{
"epoch": 0.29936305732484075,
"grad_norm": 0.7693317009374261,
"learning_rate": 4.214183542078646e-06,
"loss": 0.9391,
"step": 141
},
{
"epoch": 0.30148619957537154,
"grad_norm": 0.6809609221970698,
"learning_rate": 4.201351632667227e-06,
"loss": 1.0107,
"step": 142
},
{
"epoch": 0.3036093418259023,
"grad_norm": 0.6776804857260985,
"learning_rate": 4.1884356850070695e-06,
"loss": 1.0316,
"step": 143
},
{
"epoch": 0.3057324840764331,
"grad_norm": 0.6858474793512133,
"learning_rate": 4.175436337081289e-06,
"loss": 0.9071,
"step": 144
},
{
"epoch": 0.3078556263269639,
"grad_norm": 0.7262200113022307,
"learning_rate": 4.162354230992562e-06,
"loss": 0.9751,
"step": 145
},
{
"epoch": 0.3099787685774947,
"grad_norm": 0.7048851306107318,
"learning_rate": 4.149190012931402e-06,
"loss": 0.9222,
"step": 146
},
{
"epoch": 0.31210191082802546,
"grad_norm": 0.7661302534267068,
"learning_rate": 4.135944333144244e-06,
"loss": 0.9493,
"step": 147
},
{
"epoch": 0.31422505307855625,
"grad_norm": 0.6647279751401796,
"learning_rate": 4.122617845901322e-06,
"loss": 1.0373,
"step": 148
},
{
"epoch": 0.31634819532908703,
"grad_norm": 0.677884233878186,
"learning_rate": 4.109211209464354e-06,
"loss": 1.0299,
"step": 149
},
{
"epoch": 0.3184713375796178,
"grad_norm": 0.6808985503614562,
"learning_rate": 4.095725086054029e-06,
"loss": 0.9564,
"step": 150
},
{
"epoch": 0.3205944798301486,
"grad_norm": 0.6844572260666089,
"learning_rate": 4.0821601418172926e-06,
"loss": 0.9942,
"step": 151
},
{
"epoch": 0.3227176220806794,
"grad_norm": 0.687636477644006,
"learning_rate": 4.068517046794443e-06,
"loss": 0.9925,
"step": 152
},
{
"epoch": 0.3248407643312102,
"grad_norm": 0.7121462034889847,
"learning_rate": 4.054796474886038e-06,
"loss": 1.0232,
"step": 153
},
{
"epoch": 0.32696390658174096,
"grad_norm": 0.6791139957154484,
"learning_rate": 4.040999103819606e-06,
"loss": 1.0147,
"step": 154
},
{
"epoch": 0.32908704883227174,
"grad_norm": 0.6317316423507299,
"learning_rate": 4.0271256151161666e-06,
"loss": 1.0073,
"step": 155
},
{
"epoch": 0.33121019108280253,
"grad_norm": 0.7021672955838567,
"learning_rate": 4.013176694056571e-06,
"loss": 0.9892,
"step": 156
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.6535581508427883,
"learning_rate": 3.999153029647651e-06,
"loss": 1.0406,
"step": 157
}
],
"logging_steps": 1,
"max_steps": 471,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 157,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 23254277554176.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}