d1_code_python_3k / trainer_state.json
ryanmarten's picture
End of training
3242adf verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 6.850632911392405,
"eval_steps": 500,
"global_step": 224,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.030379746835443037,
"grad_norm": 6.900967765530235,
"learning_rate": 8.695652173913044e-07,
"loss": 1.0701,
"step": 1
},
{
"epoch": 0.060759493670886074,
"grad_norm": 6.575531951311463,
"learning_rate": 1.7391304347826088e-06,
"loss": 1.0428,
"step": 2
},
{
"epoch": 0.09113924050632911,
"grad_norm": 6.4215988628473974,
"learning_rate": 2.6086956521739132e-06,
"loss": 1.0333,
"step": 3
},
{
"epoch": 0.12151898734177215,
"grad_norm": 5.97286810463995,
"learning_rate": 3.4782608695652175e-06,
"loss": 1.0287,
"step": 4
},
{
"epoch": 0.1518987341772152,
"grad_norm": 4.539445458567603,
"learning_rate": 4.347826086956522e-06,
"loss": 0.986,
"step": 5
},
{
"epoch": 0.18227848101265823,
"grad_norm": 2.7976966757232544,
"learning_rate": 5.2173913043478265e-06,
"loss": 0.9635,
"step": 6
},
{
"epoch": 0.21265822784810126,
"grad_norm": 2.5961464405887655,
"learning_rate": 6.086956521739132e-06,
"loss": 0.9403,
"step": 7
},
{
"epoch": 0.2430379746835443,
"grad_norm": 4.191157758536918,
"learning_rate": 6.956521739130435e-06,
"loss": 0.9413,
"step": 8
},
{
"epoch": 0.27341772151898736,
"grad_norm": 4.442211994169548,
"learning_rate": 7.82608695652174e-06,
"loss": 0.9618,
"step": 9
},
{
"epoch": 0.3037974683544304,
"grad_norm": 3.797392864859067,
"learning_rate": 8.695652173913044e-06,
"loss": 0.9109,
"step": 10
},
{
"epoch": 0.3341772151898734,
"grad_norm": 3.6149539350857385,
"learning_rate": 9.565217391304349e-06,
"loss": 0.8934,
"step": 11
},
{
"epoch": 0.36455696202531646,
"grad_norm": 3.089514596657137,
"learning_rate": 1.0434782608695653e-05,
"loss": 0.8838,
"step": 12
},
{
"epoch": 0.3949367088607595,
"grad_norm": 2.130812155192881,
"learning_rate": 1.1304347826086957e-05,
"loss": 0.8901,
"step": 13
},
{
"epoch": 0.4253164556962025,
"grad_norm": 1.5156598615847707,
"learning_rate": 1.2173913043478263e-05,
"loss": 0.8292,
"step": 14
},
{
"epoch": 0.45569620253164556,
"grad_norm": 1.5151462799510833,
"learning_rate": 1.3043478260869566e-05,
"loss": 0.8106,
"step": 15
},
{
"epoch": 0.4860759493670886,
"grad_norm": 1.315000402699009,
"learning_rate": 1.391304347826087e-05,
"loss": 0.7835,
"step": 16
},
{
"epoch": 0.5164556962025316,
"grad_norm": 1.0686914721404177,
"learning_rate": 1.4782608695652174e-05,
"loss": 0.7897,
"step": 17
},
{
"epoch": 0.5468354430379747,
"grad_norm": 1.2232559006839683,
"learning_rate": 1.565217391304348e-05,
"loss": 0.7656,
"step": 18
},
{
"epoch": 0.5772151898734177,
"grad_norm": 1.1524094785535905,
"learning_rate": 1.6521739130434785e-05,
"loss": 0.7679,
"step": 19
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.9356142584753656,
"learning_rate": 1.739130434782609e-05,
"loss": 0.7518,
"step": 20
},
{
"epoch": 0.6379746835443038,
"grad_norm": 1.0872139080338754,
"learning_rate": 1.8260869565217393e-05,
"loss": 0.761,
"step": 21
},
{
"epoch": 0.6683544303797468,
"grad_norm": 1.0048304807301232,
"learning_rate": 1.9130434782608697e-05,
"loss": 0.7405,
"step": 22
},
{
"epoch": 0.6987341772151898,
"grad_norm": 0.7871080094445352,
"learning_rate": 2e-05,
"loss": 0.7559,
"step": 23
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.7580947914049513,
"learning_rate": 1.999877856940653e-05,
"loss": 0.7176,
"step": 24
},
{
"epoch": 0.759493670886076,
"grad_norm": 0.736793411198655,
"learning_rate": 1.999511457600466e-05,
"loss": 0.7253,
"step": 25
},
{
"epoch": 0.789873417721519,
"grad_norm": 0.7367998443576975,
"learning_rate": 1.9989008914857115e-05,
"loss": 0.7197,
"step": 26
},
{
"epoch": 0.8202531645569621,
"grad_norm": 0.794945194059278,
"learning_rate": 1.998046307749216e-05,
"loss": 0.7206,
"step": 27
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.6610336013779932,
"learning_rate": 1.9969479151539238e-05,
"loss": 0.7224,
"step": 28
},
{
"epoch": 0.8810126582278481,
"grad_norm": 0.6919942538851096,
"learning_rate": 1.9956059820218982e-05,
"loss": 0.7189,
"step": 29
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.6140354305025851,
"learning_rate": 1.9940208361687762e-05,
"loss": 0.7296,
"step": 30
},
{
"epoch": 0.9417721518987342,
"grad_norm": 0.6709056568535675,
"learning_rate": 1.9921928648236855e-05,
"loss": 0.7047,
"step": 31
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.6484001484687504,
"learning_rate": 1.990122514534651e-05,
"loss": 0.713,
"step": 32
},
{
"epoch": 1.010126582278481,
"grad_norm": 0.6141911850605088,
"learning_rate": 1.9878102910595097e-05,
"loss": 0.707,
"step": 33
},
{
"epoch": 1.040506329113924,
"grad_norm": 0.5928634363916581,
"learning_rate": 1.985256759242359e-05,
"loss": 0.679,
"step": 34
},
{
"epoch": 1.070886075949367,
"grad_norm": 0.5279347538697585,
"learning_rate": 1.982462542875576e-05,
"loss": 0.6817,
"step": 35
},
{
"epoch": 1.1012658227848102,
"grad_norm": 0.49207519591692517,
"learning_rate": 1.979428324547432e-05,
"loss": 0.6638,
"step": 36
},
{
"epoch": 1.1316455696202532,
"grad_norm": 0.5383685675662327,
"learning_rate": 1.9761548454753455e-05,
"loss": 0.6612,
"step": 37
},
{
"epoch": 1.1620253164556962,
"grad_norm": 0.7010865119107363,
"learning_rate": 1.972642905324813e-05,
"loss": 0.6698,
"step": 38
},
{
"epoch": 1.1924050632911392,
"grad_norm": 0.4834295687560638,
"learning_rate": 1.9688933620140638e-05,
"loss": 0.7059,
"step": 39
},
{
"epoch": 1.2227848101265824,
"grad_norm": 0.6001307078259079,
"learning_rate": 1.96490713150448e-05,
"loss": 0.6519,
"step": 40
},
{
"epoch": 1.2531645569620253,
"grad_norm": 0.488509885492811,
"learning_rate": 1.9606851875768404e-05,
"loss": 0.6575,
"step": 41
},
{
"epoch": 1.2835443037974683,
"grad_norm": 0.6125726898942023,
"learning_rate": 1.956228561593441e-05,
"loss": 0.6557,
"step": 42
},
{
"epoch": 1.3139240506329113,
"grad_norm": 0.5071218136363533,
"learning_rate": 1.9515383422461457e-05,
"loss": 0.6426,
"step": 43
},
{
"epoch": 1.3443037974683545,
"grad_norm": 0.58281223183722,
"learning_rate": 1.9466156752904344e-05,
"loss": 0.631,
"step": 44
},
{
"epoch": 1.3746835443037975,
"grad_norm": 0.5327420125731059,
"learning_rate": 1.9414617632655114e-05,
"loss": 0.648,
"step": 45
},
{
"epoch": 1.4050632911392404,
"grad_norm": 0.5393301453944022,
"learning_rate": 1.9360778652005416e-05,
"loss": 0.6361,
"step": 46
},
{
"epoch": 1.4354430379746836,
"grad_norm": 0.7325255868613124,
"learning_rate": 1.9304652963070868e-05,
"loss": 0.6379,
"step": 47
},
{
"epoch": 1.4658227848101266,
"grad_norm": 0.5808540035045816,
"learning_rate": 1.9246254276578175e-05,
"loss": 0.6443,
"step": 48
},
{
"epoch": 1.4962025316455696,
"grad_norm": 0.5895396622256339,
"learning_rate": 1.9185596858515797e-05,
"loss": 0.6603,
"step": 49
},
{
"epoch": 1.5265822784810128,
"grad_norm": 0.6558711003499368,
"learning_rate": 1.9122695526648968e-05,
"loss": 0.6473,
"step": 50
},
{
"epoch": 1.5569620253164556,
"grad_norm": 0.518033203454323,
"learning_rate": 1.905756564689991e-05,
"loss": 0.6371,
"step": 51
},
{
"epoch": 1.5873417721518988,
"grad_norm": 0.6379670916019468,
"learning_rate": 1.8990223129594146e-05,
"loss": 0.6316,
"step": 52
},
{
"epoch": 1.6177215189873417,
"grad_norm": 0.5280296632581212,
"learning_rate": 1.8920684425573865e-05,
"loss": 0.6399,
"step": 53
},
{
"epoch": 1.6481012658227847,
"grad_norm": 0.6726534436175223,
"learning_rate": 1.884896652217917e-05,
"loss": 0.6423,
"step": 54
},
{
"epoch": 1.678481012658228,
"grad_norm": 0.44139421556384173,
"learning_rate": 1.877508693909831e-05,
"loss": 0.6354,
"step": 55
},
{
"epoch": 1.7088607594936709,
"grad_norm": 0.6331583286657771,
"learning_rate": 1.8699063724087905e-05,
"loss": 0.6495,
"step": 56
},
{
"epoch": 1.7392405063291139,
"grad_norm": 0.5584456420892069,
"learning_rate": 1.862091544856407e-05,
"loss": 0.6539,
"step": 57
},
{
"epoch": 1.769620253164557,
"grad_norm": 0.47468281272491225,
"learning_rate": 1.854066120306571e-05,
"loss": 0.6376,
"step": 58
},
{
"epoch": 1.8,
"grad_norm": 0.621199560085674,
"learning_rate": 1.8458320592590976e-05,
"loss": 0.6212,
"step": 59
},
{
"epoch": 1.830379746835443,
"grad_norm": 0.5043287978462778,
"learning_rate": 1.837391373180801e-05,
"loss": 0.6368,
"step": 60
},
{
"epoch": 1.8607594936708862,
"grad_norm": 0.6919342397174377,
"learning_rate": 1.8287461240141217e-05,
"loss": 0.6265,
"step": 61
},
{
"epoch": 1.891139240506329,
"grad_norm": 0.4613310213113171,
"learning_rate": 1.8198984236734246e-05,
"loss": 0.631,
"step": 62
},
{
"epoch": 1.9215189873417722,
"grad_norm": 0.5200305742641289,
"learning_rate": 1.8108504335290852e-05,
"loss": 0.6156,
"step": 63
},
{
"epoch": 1.9518987341772152,
"grad_norm": 0.44777041092973485,
"learning_rate": 1.8016043638794975e-05,
"loss": 0.6216,
"step": 64
},
{
"epoch": 1.9822784810126581,
"grad_norm": 0.6321033144749344,
"learning_rate": 1.7921624734111292e-05,
"loss": 0.612,
"step": 65
},
{
"epoch": 2.020253164556962,
"grad_norm": 0.4423975149728961,
"learning_rate": 1.7825270686467567e-05,
"loss": 0.6075,
"step": 66
},
{
"epoch": 2.050632911392405,
"grad_norm": 0.5610954653102551,
"learning_rate": 1.7727005033820117e-05,
"loss": 0.569,
"step": 67
},
{
"epoch": 2.081012658227848,
"grad_norm": 0.48099169316251267,
"learning_rate": 1.762685178110382e-05,
"loss": 0.5695,
"step": 68
},
{
"epoch": 2.1113924050632913,
"grad_norm": 0.45798848320831864,
"learning_rate": 1.752483539436807e-05,
"loss": 0.5858,
"step": 69
},
{
"epoch": 2.141772151898734,
"grad_norm": 0.6067995595017763,
"learning_rate": 1.7420980794800013e-05,
"loss": 0.5893,
"step": 70
},
{
"epoch": 2.1721518987341772,
"grad_norm": 0.4455538598795637,
"learning_rate": 1.731531335263669e-05,
"loss": 0.5665,
"step": 71
},
{
"epoch": 2.2025316455696204,
"grad_norm": 0.5139225500993996,
"learning_rate": 1.720785888096743e-05,
"loss": 0.5758,
"step": 72
},
{
"epoch": 2.232911392405063,
"grad_norm": 0.5263467271351159,
"learning_rate": 1.7098643629428035e-05,
"loss": 0.5643,
"step": 73
},
{
"epoch": 2.2632911392405064,
"grad_norm": 0.589787310312181,
"learning_rate": 1.698769427778842e-05,
"loss": 0.5674,
"step": 74
},
{
"epoch": 2.293670886075949,
"grad_norm": 0.6550139817828793,
"learning_rate": 1.687503792943506e-05,
"loss": 0.5521,
"step": 75
},
{
"epoch": 2.3240506329113924,
"grad_norm": 0.5394217131127085,
"learning_rate": 1.6760702104750046e-05,
"loss": 0.5721,
"step": 76
},
{
"epoch": 2.3544303797468356,
"grad_norm": 0.6672433677365825,
"learning_rate": 1.664471473438822e-05,
"loss": 0.5785,
"step": 77
},
{
"epoch": 2.3848101265822783,
"grad_norm": 0.6174561016714003,
"learning_rate": 1.6527104152454096e-05,
"loss": 0.5636,
"step": 78
},
{
"epoch": 2.4151898734177215,
"grad_norm": 0.5427839185996065,
"learning_rate": 1.6407899089580263e-05,
"loss": 0.5741,
"step": 79
},
{
"epoch": 2.4455696202531647,
"grad_norm": 0.5547114175762482,
"learning_rate": 1.628712866590885e-05,
"loss": 0.5549,
"step": 80
},
{
"epoch": 2.4759493670886075,
"grad_norm": 0.6520468209899203,
"learning_rate": 1.6164822383977912e-05,
"loss": 0.5628,
"step": 81
},
{
"epoch": 2.5063291139240507,
"grad_norm": 0.484126618255132,
"learning_rate": 1.604101012151436e-05,
"loss": 0.5653,
"step": 82
},
{
"epoch": 2.536708860759494,
"grad_norm": 0.5420479074718673,
"learning_rate": 1.5915722124135227e-05,
"loss": 0.5788,
"step": 83
},
{
"epoch": 2.5670886075949366,
"grad_norm": 0.7113403279539883,
"learning_rate": 1.5788988997959115e-05,
"loss": 0.5442,
"step": 84
},
{
"epoch": 2.59746835443038,
"grad_norm": 0.429383282793752,
"learning_rate": 1.5660841702129533e-05,
"loss": 0.5619,
"step": 85
},
{
"epoch": 2.6278481012658226,
"grad_norm": 0.6318811293709081,
"learning_rate": 1.5531311541251995e-05,
"loss": 0.5593,
"step": 86
},
{
"epoch": 2.6582278481012658,
"grad_norm": 0.6928814959645925,
"learning_rate": 1.540043015774676e-05,
"loss": 0.5578,
"step": 87
},
{
"epoch": 2.688607594936709,
"grad_norm": 0.4648419968624839,
"learning_rate": 1.5268229524119007e-05,
"loss": 0.5634,
"step": 88
},
{
"epoch": 2.7189873417721517,
"grad_norm": 0.8082968039049384,
"learning_rate": 1.513474193514842e-05,
"loss": 0.5688,
"step": 89
},
{
"epoch": 2.749367088607595,
"grad_norm": 0.5498108509329553,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.5511,
"step": 90
},
{
"epoch": 2.779746835443038,
"grad_norm": 0.6122855179978333,
"learning_rate": 1.4864036634258112e-05,
"loss": 0.5742,
"step": 91
},
{
"epoch": 2.810126582278481,
"grad_norm": 0.6439369733260818,
"learning_rate": 1.4726885051885654e-05,
"loss": 0.5882,
"step": 92
},
{
"epoch": 2.840506329113924,
"grad_norm": 0.5763840390625248,
"learning_rate": 1.4588578757110359e-05,
"loss": 0.5708,
"step": 93
},
{
"epoch": 2.8708860759493673,
"grad_norm": 0.7986689835904099,
"learning_rate": 1.4449151536240167e-05,
"loss": 0.5482,
"step": 94
},
{
"epoch": 2.90126582278481,
"grad_norm": 0.5075982184175216,
"learning_rate": 1.4308637449409705e-05,
"loss": 0.5509,
"step": 95
},
{
"epoch": 2.9316455696202532,
"grad_norm": 0.6691445005481782,
"learning_rate": 1.4167070822259868e-05,
"loss": 0.5636,
"step": 96
},
{
"epoch": 2.962025316455696,
"grad_norm": 0.5531032668698979,
"learning_rate": 1.402448623755254e-05,
"loss": 0.5731,
"step": 97
},
{
"epoch": 2.992405063291139,
"grad_norm": 0.5916337115795276,
"learning_rate": 1.3880918526722497e-05,
"loss": 0.5755,
"step": 98
},
{
"epoch": 3.030379746835443,
"grad_norm": 0.7155757273866694,
"learning_rate": 1.3736402761368597e-05,
"loss": 0.5171,
"step": 99
},
{
"epoch": 3.060759493670886,
"grad_norm": 0.454608440541494,
"learning_rate": 1.3590974244686248e-05,
"loss": 0.5226,
"step": 100
},
{
"epoch": 3.091139240506329,
"grad_norm": 0.7489476692267287,
"learning_rate": 1.344466850284333e-05,
"loss": 0.519,
"step": 101
},
{
"epoch": 3.1215189873417724,
"grad_norm": 0.45509443137692646,
"learning_rate": 1.3297521276301666e-05,
"loss": 0.5242,
"step": 102
},
{
"epoch": 3.151898734177215,
"grad_norm": 0.6831099475936916,
"learning_rate": 1.3149568511086104e-05,
"loss": 0.5241,
"step": 103
},
{
"epoch": 3.1822784810126583,
"grad_norm": 0.540474264374005,
"learning_rate": 1.300084635000341e-05,
"loss": 0.529,
"step": 104
},
{
"epoch": 3.212658227848101,
"grad_norm": 0.5108682889881786,
"learning_rate": 1.2851391123813075e-05,
"loss": 0.5078,
"step": 105
},
{
"epoch": 3.2430379746835443,
"grad_norm": 0.624358953369949,
"learning_rate": 1.2701239342352223e-05,
"loss": 0.4978,
"step": 106
},
{
"epoch": 3.2734177215189875,
"grad_norm": 0.4757734539892761,
"learning_rate": 1.2550427685616767e-05,
"loss": 0.5011,
"step": 107
},
{
"epoch": 3.3037974683544302,
"grad_norm": 0.5122672512349897,
"learning_rate": 1.239899299480098e-05,
"loss": 0.4819,
"step": 108
},
{
"epoch": 3.3341772151898734,
"grad_norm": 0.42555093602885546,
"learning_rate": 1.2246972263297718e-05,
"loss": 0.5137,
"step": 109
},
{
"epoch": 3.3645569620253166,
"grad_norm": 0.5113060564073688,
"learning_rate": 1.2094402627661447e-05,
"loss": 0.4969,
"step": 110
},
{
"epoch": 3.3949367088607594,
"grad_norm": 0.4519268860422576,
"learning_rate": 1.1941321358536278e-05,
"loss": 0.5074,
"step": 111
},
{
"epoch": 3.4253164556962026,
"grad_norm": 0.44136682471859834,
"learning_rate": 1.1787765851551296e-05,
"loss": 0.4899,
"step": 112
},
{
"epoch": 3.4556962025316453,
"grad_norm": 0.45052812191914127,
"learning_rate": 1.1633773618185302e-05,
"loss": 0.4928,
"step": 113
},
{
"epoch": 3.4860759493670885,
"grad_norm": 0.5082695841633321,
"learning_rate": 1.14793822766033e-05,
"loss": 0.487,
"step": 114
},
{
"epoch": 3.5164556962025317,
"grad_norm": 0.4518831447437105,
"learning_rate": 1.132462954246688e-05,
"loss": 0.4886,
"step": 115
},
{
"epoch": 3.546835443037975,
"grad_norm": 0.4253791310421495,
"learning_rate": 1.1169553219720828e-05,
"loss": 0.5036,
"step": 116
},
{
"epoch": 3.5772151898734177,
"grad_norm": 0.40779849432123344,
"learning_rate": 1.1014191191358118e-05,
"loss": 0.5045,
"step": 117
},
{
"epoch": 3.607594936708861,
"grad_norm": 0.4106538409604475,
"learning_rate": 1.085858141016566e-05,
"loss": 0.4883,
"step": 118
},
{
"epoch": 3.6379746835443036,
"grad_norm": 0.43313847577907466,
"learning_rate": 1.070276188945293e-05,
"loss": 0.5054,
"step": 119
},
{
"epoch": 3.668354430379747,
"grad_norm": 0.34876315441846834,
"learning_rate": 1.0546770693765859e-05,
"loss": 0.5041,
"step": 120
},
{
"epoch": 3.69873417721519,
"grad_norm": 0.41865891160785923,
"learning_rate": 1.0390645929588197e-05,
"loss": 0.5086,
"step": 121
},
{
"epoch": 3.729113924050633,
"grad_norm": 0.3693619907415195,
"learning_rate": 1.0234425736032607e-05,
"loss": 0.5193,
"step": 122
},
{
"epoch": 3.759493670886076,
"grad_norm": 0.36211686116553843,
"learning_rate": 1.007814827552384e-05,
"loss": 0.4813,
"step": 123
},
{
"epoch": 3.7898734177215188,
"grad_norm": 0.3806993063290118,
"learning_rate": 9.92185172447616e-06,
"loss": 0.5085,
"step": 124
},
{
"epoch": 3.820253164556962,
"grad_norm": 0.4932836571388558,
"learning_rate": 9.765574263967397e-06,
"loss": 0.5159,
"step": 125
},
{
"epoch": 3.850632911392405,
"grad_norm": 0.4071054457640591,
"learning_rate": 9.609354070411807e-06,
"loss": 0.5173,
"step": 126
},
{
"epoch": 3.8810126582278484,
"grad_norm": 0.35848235745482016,
"learning_rate": 9.453229306234143e-06,
"loss": 0.4984,
"step": 127
},
{
"epoch": 3.911392405063291,
"grad_norm": 0.3919033582608312,
"learning_rate": 9.297238110547075e-06,
"loss": 0.4944,
"step": 128
},
{
"epoch": 3.9417721518987343,
"grad_norm": 0.38074628156321555,
"learning_rate": 9.14141858983434e-06,
"loss": 0.5004,
"step": 129
},
{
"epoch": 3.972151898734177,
"grad_norm": 0.5871821774469409,
"learning_rate": 8.985808808641883e-06,
"loss": 0.5368,
"step": 130
},
{
"epoch": 4.010126582278481,
"grad_norm": 0.39202627766042153,
"learning_rate": 8.830446780279175e-06,
"loss": 0.4772,
"step": 131
},
{
"epoch": 4.040506329113924,
"grad_norm": 0.46673082578063746,
"learning_rate": 8.675370457533122e-06,
"loss": 0.4651,
"step": 132
},
{
"epoch": 4.0708860759493675,
"grad_norm": 0.39818505604079746,
"learning_rate": 8.520617723396702e-06,
"loss": 0.4738,
"step": 133
},
{
"epoch": 4.10126582278481,
"grad_norm": 0.47016249893100936,
"learning_rate": 8.366226381814698e-06,
"loss": 0.4504,
"step": 134
},
{
"epoch": 4.131645569620253,
"grad_norm": 0.4392270738204623,
"learning_rate": 8.212234148448708e-06,
"loss": 0.4526,
"step": 135
},
{
"epoch": 4.162025316455696,
"grad_norm": 0.41637627336060623,
"learning_rate": 8.058678641463724e-06,
"loss": 0.4441,
"step": 136
},
{
"epoch": 4.192405063291139,
"grad_norm": 0.47209200948212127,
"learning_rate": 7.905597372338558e-06,
"loss": 0.4486,
"step": 137
},
{
"epoch": 4.222784810126583,
"grad_norm": 0.3623330697097205,
"learning_rate": 7.753027736702283e-06,
"loss": 0.4682,
"step": 138
},
{
"epoch": 4.253164556962025,
"grad_norm": 0.4050255588965073,
"learning_rate": 7.601007005199022e-06,
"loss": 0.4551,
"step": 139
},
{
"epoch": 4.283544303797468,
"grad_norm": 0.413750496464073,
"learning_rate": 7.449572314383237e-06,
"loss": 0.4556,
"step": 140
},
{
"epoch": 4.313924050632911,
"grad_norm": 0.3685313565905941,
"learning_rate": 7.298760657647779e-06,
"loss": 0.4495,
"step": 141
},
{
"epoch": 4.3443037974683545,
"grad_norm": 0.4044396771891203,
"learning_rate": 7.148608876186931e-06,
"loss": 0.4584,
"step": 142
},
{
"epoch": 4.374683544303798,
"grad_norm": 0.410443646965677,
"learning_rate": 6.999153649996595e-06,
"loss": 0.4556,
"step": 143
},
{
"epoch": 4.405063291139241,
"grad_norm": 0.3245734750867703,
"learning_rate": 6.8504314889138956e-06,
"loss": 0.4641,
"step": 144
},
{
"epoch": 4.435443037974683,
"grad_norm": 0.37261308661584225,
"learning_rate": 6.702478723698336e-06,
"loss": 0.4508,
"step": 145
},
{
"epoch": 4.465822784810126,
"grad_norm": 0.34093924046209567,
"learning_rate": 6.555331497156671e-06,
"loss": 0.4745,
"step": 146
},
{
"epoch": 4.49620253164557,
"grad_norm": 0.3428032870470108,
"learning_rate": 6.4090257553137566e-06,
"loss": 0.4502,
"step": 147
},
{
"epoch": 4.526582278481013,
"grad_norm": 0.3285802437545611,
"learning_rate": 6.263597238631405e-06,
"loss": 0.4674,
"step": 148
},
{
"epoch": 4.556962025316456,
"grad_norm": 0.361487655597335,
"learning_rate": 6.119081473277502e-06,
"loss": 0.4404,
"step": 149
},
{
"epoch": 4.587341772151898,
"grad_norm": 0.309193952586586,
"learning_rate": 5.975513762447465e-06,
"loss": 0.4528,
"step": 150
},
{
"epoch": 4.6177215189873415,
"grad_norm": 0.3256472467115911,
"learning_rate": 5.832929177740134e-06,
"loss": 0.4331,
"step": 151
},
{
"epoch": 4.648101265822785,
"grad_norm": 0.3424060964080675,
"learning_rate": 5.6913625505902966e-06,
"loss": 0.4422,
"step": 152
},
{
"epoch": 4.678481012658228,
"grad_norm": 0.3322337561374123,
"learning_rate": 5.550848463759835e-06,
"loss": 0.4437,
"step": 153
},
{
"epoch": 4.708860759493671,
"grad_norm": 0.31796302787060954,
"learning_rate": 5.411421242889643e-06,
"loss": 0.4515,
"step": 154
},
{
"epoch": 4.739240506329114,
"grad_norm": 0.31460303564559244,
"learning_rate": 5.273114948114346e-06,
"loss": 0.4569,
"step": 155
},
{
"epoch": 4.769620253164557,
"grad_norm": 0.32650502785455265,
"learning_rate": 5.135963365741892e-06,
"loss": 0.4618,
"step": 156
},
{
"epoch": 4.8,
"grad_norm": 0.29008971109013515,
"learning_rate": 5.000000000000003e-06,
"loss": 0.4549,
"step": 157
},
{
"epoch": 4.830379746835443,
"grad_norm": 0.3199562335154913,
"learning_rate": 4.865258064851579e-06,
"loss": 0.4389,
"step": 158
},
{
"epoch": 4.860759493670886,
"grad_norm": 0.3448626930084008,
"learning_rate": 4.731770475880995e-06,
"loss": 0.47,
"step": 159
},
{
"epoch": 4.891139240506329,
"grad_norm": 0.31654798009497775,
"learning_rate": 4.599569842253244e-06,
"loss": 0.4378,
"step": 160
},
{
"epoch": 4.921518987341772,
"grad_norm": 0.3263546165090644,
"learning_rate": 4.468688458748006e-06,
"loss": 0.4569,
"step": 161
},
{
"epoch": 4.951898734177215,
"grad_norm": 0.28764483105829874,
"learning_rate": 4.339158297870469e-06,
"loss": 0.4488,
"step": 162
},
{
"epoch": 4.982278481012658,
"grad_norm": 0.3033336320035795,
"learning_rate": 4.211011002040885e-06,
"loss": 0.4542,
"step": 163
},
{
"epoch": 5.020253164556962,
"grad_norm": 0.4133001335899334,
"learning_rate": 4.084277875864776e-06,
"loss": 0.4197,
"step": 164
},
{
"epoch": 5.050632911392405,
"grad_norm": 0.3894156451707503,
"learning_rate": 3.958989878485644e-06,
"loss": 0.4389,
"step": 165
},
{
"epoch": 5.0810126582278485,
"grad_norm": 0.3206114121682523,
"learning_rate": 3.83517761602209e-06,
"loss": 0.4193,
"step": 166
},
{
"epoch": 5.111392405063291,
"grad_norm": 0.4202820690299635,
"learning_rate": 3.712871334091154e-06,
"loss": 0.414,
"step": 167
},
{
"epoch": 5.141772151898734,
"grad_norm": 0.4419485939262147,
"learning_rate": 3.592100910419738e-06,
"loss": 0.4237,
"step": 168
},
{
"epoch": 5.172151898734177,
"grad_norm": 0.3660526714013728,
"learning_rate": 3.4728958475459052e-06,
"loss": 0.4356,
"step": 169
},
{
"epoch": 5.2025316455696204,
"grad_norm": 0.36926377394725557,
"learning_rate": 3.355285265611784e-06,
"loss": 0.4309,
"step": 170
},
{
"epoch": 5.232911392405064,
"grad_norm": 0.3770922165745436,
"learning_rate": 3.2392978952499553e-06,
"loss": 0.4017,
"step": 171
},
{
"epoch": 5.263291139240506,
"grad_norm": 0.36009582277341656,
"learning_rate": 3.1249620705649417e-06,
"loss": 0.4287,
"step": 172
},
{
"epoch": 5.293670886075949,
"grad_norm": 0.32402855594401253,
"learning_rate": 3.0123057222115835e-06,
"loss": 0.403,
"step": 173
},
{
"epoch": 5.324050632911392,
"grad_norm": 0.32856792046727396,
"learning_rate": 2.9013563705719673e-06,
"loss": 0.4128,
"step": 174
},
{
"epoch": 5.3544303797468356,
"grad_norm": 0.3415102656117101,
"learning_rate": 2.7921411190325753e-06,
"loss": 0.4244,
"step": 175
},
{
"epoch": 5.384810126582279,
"grad_norm": 0.33400049954552397,
"learning_rate": 2.6846866473633126e-06,
"loss": 0.4151,
"step": 176
},
{
"epoch": 5.415189873417722,
"grad_norm": 0.3198760407550781,
"learning_rate": 2.579019205199992e-06,
"loss": 0.4409,
"step": 177
},
{
"epoch": 5.445569620253164,
"grad_norm": 0.3100042926736425,
"learning_rate": 2.4751646056319334e-06,
"loss": 0.4128,
"step": 178
},
{
"epoch": 5.4759493670886075,
"grad_norm": 0.2893924306167645,
"learning_rate": 2.373148218896182e-06,
"loss": 0.4256,
"step": 179
},
{
"epoch": 5.506329113924051,
"grad_norm": 0.2999713636551937,
"learning_rate": 2.2729949661798876e-06,
"loss": 0.4072,
"step": 180
},
{
"epoch": 5.536708860759494,
"grad_norm": 0.3253580352285578,
"learning_rate": 2.174729313532433e-06,
"loss": 0.4081,
"step": 181
},
{
"epoch": 5.567088607594937,
"grad_norm": 0.3175492041018933,
"learning_rate": 2.078375265888707e-06,
"loss": 0.3984,
"step": 182
},
{
"epoch": 5.597468354430379,
"grad_norm": 0.2916187796414224,
"learning_rate": 1.9839563612050273e-06,
"loss": 0.4183,
"step": 183
},
{
"epoch": 5.627848101265823,
"grad_norm": 0.29422073920797104,
"learning_rate": 1.8914956647091497e-06,
"loss": 0.4129,
"step": 184
},
{
"epoch": 5.658227848101266,
"grad_norm": 0.2942344841501321,
"learning_rate": 1.8010157632657544e-06,
"loss": 0.414,
"step": 185
},
{
"epoch": 5.688607594936709,
"grad_norm": 0.30053743179793696,
"learning_rate": 1.7125387598587862e-06,
"loss": 0.4461,
"step": 186
},
{
"epoch": 5.718987341772152,
"grad_norm": 0.329795052074006,
"learning_rate": 1.6260862681919965e-06,
"loss": 0.4047,
"step": 187
},
{
"epoch": 5.749367088607595,
"grad_norm": 0.293102040583313,
"learning_rate": 1.5416794074090258e-06,
"loss": 0.4181,
"step": 188
},
{
"epoch": 5.779746835443038,
"grad_norm": 0.3050351458985814,
"learning_rate": 1.459338796934293e-06,
"loss": 0.3856,
"step": 189
},
{
"epoch": 5.810126582278481,
"grad_norm": 0.27984047689532165,
"learning_rate": 1.3790845514359363e-06,
"loss": 0.4271,
"step": 190
},
{
"epoch": 5.840506329113924,
"grad_norm": 0.2645085833333402,
"learning_rate": 1.300936275912098e-06,
"loss": 0.422,
"step": 191
},
{
"epoch": 5.870886075949367,
"grad_norm": 0.2911168330595136,
"learning_rate": 1.224913060901688e-06,
"loss": 0.4334,
"step": 192
},
{
"epoch": 5.9012658227848105,
"grad_norm": 0.3086615720282332,
"learning_rate": 1.1510334778208332e-06,
"loss": 0.4177,
"step": 193
},
{
"epoch": 5.931645569620253,
"grad_norm": 0.268883531310211,
"learning_rate": 1.0793155744261352e-06,
"loss": 0.425,
"step": 194
},
{
"epoch": 5.962025316455696,
"grad_norm": 0.2604288538855863,
"learning_rate": 1.0097768704058542e-06,
"loss": 0.4102,
"step": 195
},
{
"epoch": 5.992405063291139,
"grad_norm": 0.3595203385253698,
"learning_rate": 9.424343531000968e-07,
"loss": 0.4143,
"step": 196
},
{
"epoch": 6.030379746835443,
"grad_norm": 0.3094543245332767,
"learning_rate": 8.773044733510338e-07,
"loss": 0.4073,
"step": 197
},
{
"epoch": 6.060759493670886,
"grad_norm": 0.29413748585079713,
"learning_rate": 8.144031414842012e-07,
"loss": 0.3952,
"step": 198
},
{
"epoch": 6.091139240506329,
"grad_norm": 0.2785982509263817,
"learning_rate": 7.537457234218271e-07,
"loss": 0.4011,
"step": 199
},
{
"epoch": 6.121518987341772,
"grad_norm": 0.25001601970971643,
"learning_rate": 6.953470369291349e-07,
"loss": 0.4168,
"step": 200
},
{
"epoch": 6.151898734177215,
"grad_norm": 0.26060007361890414,
"learning_rate": 6.392213479945852e-07,
"loss": 0.4128,
"step": 201
},
{
"epoch": 6.182278481012658,
"grad_norm": 0.2654458778682752,
"learning_rate": 5.853823673448877e-07,
"loss": 0.394,
"step": 202
},
{
"epoch": 6.2126582278481015,
"grad_norm": 0.26056785189425213,
"learning_rate": 5.33843247095659e-07,
"loss": 0.3923,
"step": 203
},
{
"epoch": 6.243037974683545,
"grad_norm": 0.2699015301179899,
"learning_rate": 4.846165775385459e-07,
"loss": 0.4219,
"step": 204
},
{
"epoch": 6.273417721518987,
"grad_norm": 0.266459366555451,
"learning_rate": 4.3771438406559173e-07,
"loss": 0.3992,
"step": 205
},
{
"epoch": 6.30379746835443,
"grad_norm": 0.26091523912608355,
"learning_rate": 3.931481242315993e-07,
"loss": 0.3968,
"step": 206
},
{
"epoch": 6.334177215189873,
"grad_norm": 0.2587381009672222,
"learning_rate": 3.5092868495520294e-07,
"loss": 0.4084,
"step": 207
},
{
"epoch": 6.364556962025317,
"grad_norm": 0.26538726519000366,
"learning_rate": 3.110663798593616e-07,
"loss": 0.4118,
"step": 208
},
{
"epoch": 6.39493670886076,
"grad_norm": 0.2515495107618738,
"learning_rate": 2.735709467518699e-07,
"loss": 0.3926,
"step": 209
},
{
"epoch": 6.425316455696202,
"grad_norm": 0.24861153616287043,
"learning_rate": 2.384515452465475e-07,
"loss": 0.4037,
"step": 210
},
{
"epoch": 6.455696202531645,
"grad_norm": 0.2757372443454538,
"learning_rate": 2.0571675452567997e-07,
"loss": 0.418,
"step": 211
},
{
"epoch": 6.4860759493670885,
"grad_norm": 0.2566478950086312,
"learning_rate": 1.7537457124423896e-07,
"loss": 0.3973,
"step": 212
},
{
"epoch": 6.516455696202532,
"grad_norm": 0.25440369563099446,
"learning_rate": 1.474324075764111e-07,
"loss": 0.406,
"step": 213
},
{
"epoch": 6.546835443037975,
"grad_norm": 0.256266722394824,
"learning_rate": 1.2189708940490653e-07,
"loss": 0.4105,
"step": 214
},
{
"epoch": 6.577215189873417,
"grad_norm": 0.24510177892488455,
"learning_rate": 9.877485465349057e-08,
"loss": 0.3835,
"step": 215
},
{
"epoch": 6.6075949367088604,
"grad_norm": 0.2554638538206936,
"learning_rate": 7.807135176314707e-08,
"loss": 0.3802,
"step": 216
},
{
"epoch": 6.637974683544304,
"grad_norm": 0.2576741131768959,
"learning_rate": 5.979163831223988e-08,
"loss": 0.4075,
"step": 217
},
{
"epoch": 6.668354430379747,
"grad_norm": 0.24135841990075899,
"learning_rate": 4.394017978101905e-08,
"loss": 0.4181,
"step": 218
},
{
"epoch": 6.69873417721519,
"grad_norm": 0.25415706473002286,
"learning_rate": 3.0520848460765525e-08,
"loss": 0.3994,
"step": 219
},
{
"epoch": 6.729113924050633,
"grad_norm": 0.2519962661393518,
"learning_rate": 1.9536922507841227e-08,
"loss": 0.3804,
"step": 220
},
{
"epoch": 6.759493670886076,
"grad_norm": 0.2519483781563465,
"learning_rate": 1.099108514288627e-08,
"loss": 0.407,
"step": 221
},
{
"epoch": 6.789873417721519,
"grad_norm": 0.25482488351515514,
"learning_rate": 4.885423995341088e-09,
"loss": 0.4187,
"step": 222
},
{
"epoch": 6.820253164556962,
"grad_norm": 0.27219942697806976,
"learning_rate": 1.2214305934699078e-09,
"loss": 0.3991,
"step": 223
},
{
"epoch": 6.850632911392405,
"grad_norm": 0.2486644947264924,
"learning_rate": 0.0,
"loss": 0.4068,
"step": 224
},
{
"epoch": 6.850632911392405,
"step": 224,
"total_flos": 9.144986299206205e+17,
"train_loss": 0.5499288342627031,
"train_runtime": 27786.7275,
"train_samples_per_second": 0.796,
"train_steps_per_second": 0.008
}
],
"logging_steps": 1,
"max_steps": 224,
"num_input_tokens_seen": 0,
"num_train_epochs": 7,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 9.144986299206205e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}