c1_science_10d_16s_3k / trainer_state.json
ryanmarten's picture
Upload model
c751bdd verified
raw
history blame
39.5 kB
{
"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": 6.571710244916427,
"learning_rate": 8.695652173913044e-07,
"loss": 1.0468,
"step": 1
},
{
"epoch": 0.060759493670886074,
"grad_norm": 6.584888261901075,
"learning_rate": 1.7391304347826088e-06,
"loss": 1.0646,
"step": 2
},
{
"epoch": 0.09113924050632911,
"grad_norm": 6.5829190846891965,
"learning_rate": 2.6086956521739132e-06,
"loss": 1.0622,
"step": 3
},
{
"epoch": 0.12151898734177215,
"grad_norm": 5.986274057023929,
"learning_rate": 3.4782608695652175e-06,
"loss": 1.0444,
"step": 4
},
{
"epoch": 0.1518987341772152,
"grad_norm": 4.916943092294921,
"learning_rate": 4.347826086956522e-06,
"loss": 1.0046,
"step": 5
},
{
"epoch": 0.18227848101265823,
"grad_norm": 2.5806009746099323,
"learning_rate": 5.2173913043478265e-06,
"loss": 0.9958,
"step": 6
},
{
"epoch": 0.21265822784810126,
"grad_norm": 2.1381745868721245,
"learning_rate": 6.086956521739132e-06,
"loss": 0.9489,
"step": 7
},
{
"epoch": 0.2430379746835443,
"grad_norm": 4.409782277140252,
"learning_rate": 6.956521739130435e-06,
"loss": 0.931,
"step": 8
},
{
"epoch": 0.27341772151898736,
"grad_norm": 4.700716606778443,
"learning_rate": 7.82608695652174e-06,
"loss": 0.9391,
"step": 9
},
{
"epoch": 0.3037974683544304,
"grad_norm": 4.563103880723426,
"learning_rate": 8.695652173913044e-06,
"loss": 0.9154,
"step": 10
},
{
"epoch": 0.3341772151898734,
"grad_norm": 4.356293158183602,
"learning_rate": 9.565217391304349e-06,
"loss": 0.8921,
"step": 11
},
{
"epoch": 0.36455696202531646,
"grad_norm": 3.8837074690270024,
"learning_rate": 1.0434782608695653e-05,
"loss": 0.8708,
"step": 12
},
{
"epoch": 0.3949367088607595,
"grad_norm": 3.131081058420418,
"learning_rate": 1.1304347826086957e-05,
"loss": 0.8306,
"step": 13
},
{
"epoch": 0.4253164556962025,
"grad_norm": 2.692877292452746,
"learning_rate": 1.2173913043478263e-05,
"loss": 0.8636,
"step": 14
},
{
"epoch": 0.45569620253164556,
"grad_norm": 2.660046552948952,
"learning_rate": 1.3043478260869566e-05,
"loss": 0.8264,
"step": 15
},
{
"epoch": 0.4860759493670886,
"grad_norm": 2.6247949320839616,
"learning_rate": 1.391304347826087e-05,
"loss": 0.8247,
"step": 16
},
{
"epoch": 0.5164556962025316,
"grad_norm": 2.6097971785450556,
"learning_rate": 1.4782608695652174e-05,
"loss": 0.8329,
"step": 17
},
{
"epoch": 0.5468354430379747,
"grad_norm": 1.986529333432244,
"learning_rate": 1.565217391304348e-05,
"loss": 0.8031,
"step": 18
},
{
"epoch": 0.5772151898734177,
"grad_norm": 1.6970915565643376,
"learning_rate": 1.6521739130434785e-05,
"loss": 0.8077,
"step": 19
},
{
"epoch": 0.6075949367088608,
"grad_norm": 1.47944010935288,
"learning_rate": 1.739130434782609e-05,
"loss": 0.7619,
"step": 20
},
{
"epoch": 0.6379746835443038,
"grad_norm": 1.600785511063542,
"learning_rate": 1.8260869565217393e-05,
"loss": 0.7921,
"step": 21
},
{
"epoch": 0.6683544303797468,
"grad_norm": 1.2463876261791966,
"learning_rate": 1.9130434782608697e-05,
"loss": 0.7752,
"step": 22
},
{
"epoch": 0.6987341772151898,
"grad_norm": 1.2765022654214535,
"learning_rate": 2e-05,
"loss": 0.7474,
"step": 23
},
{
"epoch": 0.7291139240506329,
"grad_norm": 1.0820472673303314,
"learning_rate": 1.999877856940653e-05,
"loss": 0.7321,
"step": 24
},
{
"epoch": 0.759493670886076,
"grad_norm": 1.033830390932466,
"learning_rate": 1.999511457600466e-05,
"loss": 0.7557,
"step": 25
},
{
"epoch": 0.789873417721519,
"grad_norm": 1.15093605459584,
"learning_rate": 1.9989008914857115e-05,
"loss": 0.7621,
"step": 26
},
{
"epoch": 0.8202531645569621,
"grad_norm": 0.90808619471091,
"learning_rate": 1.998046307749216e-05,
"loss": 0.7026,
"step": 27
},
{
"epoch": 0.850632911392405,
"grad_norm": 1.1592780803576461,
"learning_rate": 1.9969479151539238e-05,
"loss": 0.7753,
"step": 28
},
{
"epoch": 0.8810126582278481,
"grad_norm": 1.0881166841141052,
"learning_rate": 1.9956059820218982e-05,
"loss": 0.7075,
"step": 29
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.8885254561219318,
"learning_rate": 1.9940208361687762e-05,
"loss": 0.727,
"step": 30
},
{
"epoch": 0.9417721518987342,
"grad_norm": 1.2824703578093604,
"learning_rate": 1.9921928648236855e-05,
"loss": 0.7181,
"step": 31
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.9825576866386416,
"learning_rate": 1.990122514534651e-05,
"loss": 0.6768,
"step": 32
},
{
"epoch": 1.010126582278481,
"grad_norm": 1.2158455805572395,
"learning_rate": 1.9878102910595097e-05,
"loss": 0.7094,
"step": 33
},
{
"epoch": 1.040506329113924,
"grad_norm": 0.9923158660662204,
"learning_rate": 1.985256759242359e-05,
"loss": 0.6611,
"step": 34
},
{
"epoch": 1.070886075949367,
"grad_norm": 0.9089019186870102,
"learning_rate": 1.982462542875576e-05,
"loss": 0.67,
"step": 35
},
{
"epoch": 1.1012658227848102,
"grad_norm": 0.7627833742076737,
"learning_rate": 1.979428324547432e-05,
"loss": 0.6189,
"step": 36
},
{
"epoch": 1.1316455696202532,
"grad_norm": 0.6939967503990835,
"learning_rate": 1.9761548454753455e-05,
"loss": 0.5971,
"step": 37
},
{
"epoch": 1.1620253164556962,
"grad_norm": 0.8217387241250471,
"learning_rate": 1.972642905324813e-05,
"loss": 0.6242,
"step": 38
},
{
"epoch": 1.1924050632911392,
"grad_norm": 0.6853136969708522,
"learning_rate": 1.9688933620140638e-05,
"loss": 0.6202,
"step": 39
},
{
"epoch": 1.2227848101265824,
"grad_norm": 0.8162936323626641,
"learning_rate": 1.96490713150448e-05,
"loss": 0.6071,
"step": 40
},
{
"epoch": 1.2531645569620253,
"grad_norm": 0.8521213485064775,
"learning_rate": 1.9606851875768404e-05,
"loss": 0.6357,
"step": 41
},
{
"epoch": 1.2835443037974683,
"grad_norm": 0.8708493902686254,
"learning_rate": 1.956228561593441e-05,
"loss": 0.6197,
"step": 42
},
{
"epoch": 1.3139240506329113,
"grad_norm": 0.7759754507613245,
"learning_rate": 1.9515383422461457e-05,
"loss": 0.6031,
"step": 43
},
{
"epoch": 1.3443037974683545,
"grad_norm": 0.8598689433498667,
"learning_rate": 1.9466156752904344e-05,
"loss": 0.6392,
"step": 44
},
{
"epoch": 1.3746835443037975,
"grad_norm": 0.7842680485348008,
"learning_rate": 1.9414617632655114e-05,
"loss": 0.6182,
"step": 45
},
{
"epoch": 1.4050632911392404,
"grad_norm": 0.8512495674609817,
"learning_rate": 1.9360778652005416e-05,
"loss": 0.6064,
"step": 46
},
{
"epoch": 1.4354430379746836,
"grad_norm": 1.0399579537207244,
"learning_rate": 1.9304652963070868e-05,
"loss": 0.6135,
"step": 47
},
{
"epoch": 1.4658227848101266,
"grad_norm": 0.7785410527633699,
"learning_rate": 1.9246254276578175e-05,
"loss": 0.5892,
"step": 48
},
{
"epoch": 1.4962025316455696,
"grad_norm": 0.781085354056731,
"learning_rate": 1.9185596858515797e-05,
"loss": 0.5693,
"step": 49
},
{
"epoch": 1.5265822784810128,
"grad_norm": 0.9638349374413304,
"learning_rate": 1.9122695526648968e-05,
"loss": 0.6014,
"step": 50
},
{
"epoch": 1.5569620253164556,
"grad_norm": 0.7248911543227423,
"learning_rate": 1.905756564689991e-05,
"loss": 0.612,
"step": 51
},
{
"epoch": 1.5873417721518988,
"grad_norm": 0.7185899614179934,
"learning_rate": 1.8990223129594146e-05,
"loss": 0.5901,
"step": 52
},
{
"epoch": 1.6177215189873417,
"grad_norm": 0.9403225382334981,
"learning_rate": 1.8920684425573865e-05,
"loss": 0.5776,
"step": 53
},
{
"epoch": 1.6481012658227847,
"grad_norm": 0.7277422223823456,
"learning_rate": 1.884896652217917e-05,
"loss": 0.5726,
"step": 54
},
{
"epoch": 1.678481012658228,
"grad_norm": 0.8484567601014534,
"learning_rate": 1.877508693909831e-05,
"loss": 0.583,
"step": 55
},
{
"epoch": 1.7088607594936709,
"grad_norm": 0.7273805263769564,
"learning_rate": 1.8699063724087905e-05,
"loss": 0.6398,
"step": 56
},
{
"epoch": 1.7392405063291139,
"grad_norm": 0.8112017793403115,
"learning_rate": 1.862091544856407e-05,
"loss": 0.6048,
"step": 57
},
{
"epoch": 1.769620253164557,
"grad_norm": 0.7447221693680904,
"learning_rate": 1.854066120306571e-05,
"loss": 0.6045,
"step": 58
},
{
"epoch": 1.8,
"grad_norm": 0.7800809256719868,
"learning_rate": 1.8458320592590976e-05,
"loss": 0.5815,
"step": 59
},
{
"epoch": 1.830379746835443,
"grad_norm": 0.7070384208924317,
"learning_rate": 1.837391373180801e-05,
"loss": 0.5844,
"step": 60
},
{
"epoch": 1.8607594936708862,
"grad_norm": 0.8534386134810288,
"learning_rate": 1.8287461240141217e-05,
"loss": 0.5928,
"step": 61
},
{
"epoch": 1.891139240506329,
"grad_norm": 0.7991713227608602,
"learning_rate": 1.8198984236734246e-05,
"loss": 0.5987,
"step": 62
},
{
"epoch": 1.9215189873417722,
"grad_norm": 0.7835569930011058,
"learning_rate": 1.8108504335290852e-05,
"loss": 0.5853,
"step": 63
},
{
"epoch": 1.9518987341772152,
"grad_norm": 0.8306022048153217,
"learning_rate": 1.8016043638794975e-05,
"loss": 0.5889,
"step": 64
},
{
"epoch": 1.9822784810126581,
"grad_norm": 0.6860569907741538,
"learning_rate": 1.7921624734111292e-05,
"loss": 0.5823,
"step": 65
},
{
"epoch": 2.020253164556962,
"grad_norm": 0.7989377323443457,
"learning_rate": 1.7825270686467567e-05,
"loss": 0.5004,
"step": 66
},
{
"epoch": 2.050632911392405,
"grad_norm": 0.6932589763968676,
"learning_rate": 1.7727005033820117e-05,
"loss": 0.4442,
"step": 67
},
{
"epoch": 2.081012658227848,
"grad_norm": 1.261951551373665,
"learning_rate": 1.762685178110382e-05,
"loss": 0.4529,
"step": 68
},
{
"epoch": 2.1113924050632913,
"grad_norm": 0.7827963460304641,
"learning_rate": 1.752483539436807e-05,
"loss": 0.4597,
"step": 69
},
{
"epoch": 2.141772151898734,
"grad_norm": 1.1352669012355001,
"learning_rate": 1.7420980794800013e-05,
"loss": 0.4567,
"step": 70
},
{
"epoch": 2.1721518987341772,
"grad_norm": 0.9247863155500291,
"learning_rate": 1.731531335263669e-05,
"loss": 0.4381,
"step": 71
},
{
"epoch": 2.2025316455696204,
"grad_norm": 0.6924916493077918,
"learning_rate": 1.720785888096743e-05,
"loss": 0.4547,
"step": 72
},
{
"epoch": 2.232911392405063,
"grad_norm": 0.8120055923792734,
"learning_rate": 1.7098643629428035e-05,
"loss": 0.4592,
"step": 73
},
{
"epoch": 2.2632911392405064,
"grad_norm": 0.6823279393839417,
"learning_rate": 1.698769427778842e-05,
"loss": 0.4403,
"step": 74
},
{
"epoch": 2.293670886075949,
"grad_norm": 1.2690156570062068,
"learning_rate": 1.687503792943506e-05,
"loss": 0.4211,
"step": 75
},
{
"epoch": 2.3240506329113924,
"grad_norm": 0.6298587598480608,
"learning_rate": 1.6760702104750046e-05,
"loss": 0.4218,
"step": 76
},
{
"epoch": 2.3544303797468356,
"grad_norm": 0.6815787062968733,
"learning_rate": 1.664471473438822e-05,
"loss": 0.4352,
"step": 77
},
{
"epoch": 2.3848101265822783,
"grad_norm": 0.6009356136430085,
"learning_rate": 1.6527104152454096e-05,
"loss": 0.4165,
"step": 78
},
{
"epoch": 2.4151898734177215,
"grad_norm": 0.6637178455182509,
"learning_rate": 1.6407899089580263e-05,
"loss": 0.436,
"step": 79
},
{
"epoch": 2.4455696202531647,
"grad_norm": 0.6317392617042757,
"learning_rate": 1.628712866590885e-05,
"loss": 0.4277,
"step": 80
},
{
"epoch": 2.4759493670886075,
"grad_norm": 0.7088088197088229,
"learning_rate": 1.6164822383977912e-05,
"loss": 0.4376,
"step": 81
},
{
"epoch": 2.5063291139240507,
"grad_norm": 0.5919345378484626,
"learning_rate": 1.604101012151436e-05,
"loss": 0.4398,
"step": 82
},
{
"epoch": 2.536708860759494,
"grad_norm": 0.5849589547319439,
"learning_rate": 1.5915722124135227e-05,
"loss": 0.432,
"step": 83
},
{
"epoch": 2.5670886075949366,
"grad_norm": 0.820570546030797,
"learning_rate": 1.5788988997959115e-05,
"loss": 0.4718,
"step": 84
},
{
"epoch": 2.59746835443038,
"grad_norm": 0.5942270512761931,
"learning_rate": 1.5660841702129533e-05,
"loss": 0.444,
"step": 85
},
{
"epoch": 2.6278481012658226,
"grad_norm": 0.5877349622542494,
"learning_rate": 1.5531311541251995e-05,
"loss": 0.4156,
"step": 86
},
{
"epoch": 2.6582278481012658,
"grad_norm": 0.7074455538437789,
"learning_rate": 1.540043015774676e-05,
"loss": 0.4559,
"step": 87
},
{
"epoch": 2.688607594936709,
"grad_norm": 0.6009228668998818,
"learning_rate": 1.5268229524119007e-05,
"loss": 0.4472,
"step": 88
},
{
"epoch": 2.7189873417721517,
"grad_norm": 0.6237321719245432,
"learning_rate": 1.513474193514842e-05,
"loss": 0.4424,
"step": 89
},
{
"epoch": 2.749367088607595,
"grad_norm": 0.6038850128760376,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.3886,
"step": 90
},
{
"epoch": 2.779746835443038,
"grad_norm": 0.6094845356241455,
"learning_rate": 1.4864036634258112e-05,
"loss": 0.434,
"step": 91
},
{
"epoch": 2.810126582278481,
"grad_norm": 0.664655172703334,
"learning_rate": 1.4726885051885654e-05,
"loss": 0.4204,
"step": 92
},
{
"epoch": 2.840506329113924,
"grad_norm": 0.64365610583345,
"learning_rate": 1.4588578757110359e-05,
"loss": 0.4498,
"step": 93
},
{
"epoch": 2.8708860759493673,
"grad_norm": 0.558508872697568,
"learning_rate": 1.4449151536240167e-05,
"loss": 0.4357,
"step": 94
},
{
"epoch": 2.90126582278481,
"grad_norm": 0.6412349883744713,
"learning_rate": 1.4308637449409705e-05,
"loss": 0.4515,
"step": 95
},
{
"epoch": 2.9316455696202532,
"grad_norm": 0.6052921897084557,
"learning_rate": 1.4167070822259868e-05,
"loss": 0.4249,
"step": 96
},
{
"epoch": 2.962025316455696,
"grad_norm": 0.6475401662226178,
"learning_rate": 1.402448623755254e-05,
"loss": 0.415,
"step": 97
},
{
"epoch": 2.992405063291139,
"grad_norm": 0.6213067810425908,
"learning_rate": 1.3880918526722497e-05,
"loss": 0.4299,
"step": 98
},
{
"epoch": 3.030379746835443,
"grad_norm": 0.9293603871494556,
"learning_rate": 1.3736402761368597e-05,
"loss": 0.3223,
"step": 99
},
{
"epoch": 3.060759493670886,
"grad_norm": 0.5899335803856,
"learning_rate": 1.3590974244686248e-05,
"loss": 0.3039,
"step": 100
},
{
"epoch": 3.091139240506329,
"grad_norm": 1.2809540789102956,
"learning_rate": 1.344466850284333e-05,
"loss": 0.387,
"step": 101
},
{
"epoch": 3.1215189873417724,
"grad_norm": 0.8986933927044485,
"learning_rate": 1.3297521276301666e-05,
"loss": 0.3116,
"step": 102
},
{
"epoch": 3.151898734177215,
"grad_norm": 0.7102894310919661,
"learning_rate": 1.3149568511086104e-05,
"loss": 0.3444,
"step": 103
},
{
"epoch": 3.1822784810126583,
"grad_norm": 0.8953531613474207,
"learning_rate": 1.300084635000341e-05,
"loss": 0.3119,
"step": 104
},
{
"epoch": 3.212658227848101,
"grad_norm": 0.7036215868342826,
"learning_rate": 1.2851391123813075e-05,
"loss": 0.3073,
"step": 105
},
{
"epoch": 3.2430379746835443,
"grad_norm": 0.6781709170943677,
"learning_rate": 1.2701239342352223e-05,
"loss": 0.3276,
"step": 106
},
{
"epoch": 3.2734177215189875,
"grad_norm": 0.6929725261564401,
"learning_rate": 1.2550427685616767e-05,
"loss": 0.3462,
"step": 107
},
{
"epoch": 3.3037974683544302,
"grad_norm": 0.7112868072324559,
"learning_rate": 1.239899299480098e-05,
"loss": 0.3457,
"step": 108
},
{
"epoch": 3.3341772151898734,
"grad_norm": 0.6257867773500337,
"learning_rate": 1.2246972263297718e-05,
"loss": 0.3107,
"step": 109
},
{
"epoch": 3.3645569620253166,
"grad_norm": 0.5957828232725987,
"learning_rate": 1.2094402627661447e-05,
"loss": 0.3437,
"step": 110
},
{
"epoch": 3.3949367088607594,
"grad_norm": 0.578906684628111,
"learning_rate": 1.1941321358536278e-05,
"loss": 0.3213,
"step": 111
},
{
"epoch": 3.4253164556962026,
"grad_norm": 0.6323367600589211,
"learning_rate": 1.1787765851551296e-05,
"loss": 0.3114,
"step": 112
},
{
"epoch": 3.4556962025316453,
"grad_norm": 0.6765034084565158,
"learning_rate": 1.1633773618185302e-05,
"loss": 0.3228,
"step": 113
},
{
"epoch": 3.4860759493670885,
"grad_norm": 0.5928815732690462,
"learning_rate": 1.14793822766033e-05,
"loss": 0.3081,
"step": 114
},
{
"epoch": 3.5164556962025317,
"grad_norm": 0.553573258993955,
"learning_rate": 1.132462954246688e-05,
"loss": 0.3042,
"step": 115
},
{
"epoch": 3.546835443037975,
"grad_norm": 0.5848194295079728,
"learning_rate": 1.1169553219720828e-05,
"loss": 0.3053,
"step": 116
},
{
"epoch": 3.5772151898734177,
"grad_norm": 0.5772365456657795,
"learning_rate": 1.1014191191358118e-05,
"loss": 0.3375,
"step": 117
},
{
"epoch": 3.607594936708861,
"grad_norm": 0.5381449788350662,
"learning_rate": 1.085858141016566e-05,
"loss": 0.3163,
"step": 118
},
{
"epoch": 3.6379746835443036,
"grad_norm": 0.5216562838814953,
"learning_rate": 1.070276188945293e-05,
"loss": 0.2938,
"step": 119
},
{
"epoch": 3.668354430379747,
"grad_norm": 0.536579118947148,
"learning_rate": 1.0546770693765859e-05,
"loss": 0.3269,
"step": 120
},
{
"epoch": 3.69873417721519,
"grad_norm": 0.5341080183973637,
"learning_rate": 1.0390645929588197e-05,
"loss": 0.2987,
"step": 121
},
{
"epoch": 3.729113924050633,
"grad_norm": 0.5248291012635133,
"learning_rate": 1.0234425736032607e-05,
"loss": 0.3066,
"step": 122
},
{
"epoch": 3.759493670886076,
"grad_norm": 0.4993469436246074,
"learning_rate": 1.007814827552384e-05,
"loss": 0.357,
"step": 123
},
{
"epoch": 3.7898734177215188,
"grad_norm": 0.5318884508564896,
"learning_rate": 9.92185172447616e-06,
"loss": 0.3113,
"step": 124
},
{
"epoch": 3.820253164556962,
"grad_norm": 0.5643495474864679,
"learning_rate": 9.765574263967397e-06,
"loss": 0.3377,
"step": 125
},
{
"epoch": 3.850632911392405,
"grad_norm": 0.5261524951831621,
"learning_rate": 9.609354070411807e-06,
"loss": 0.3295,
"step": 126
},
{
"epoch": 3.8810126582278484,
"grad_norm": 0.5347883974392824,
"learning_rate": 9.453229306234143e-06,
"loss": 0.3361,
"step": 127
},
{
"epoch": 3.911392405063291,
"grad_norm": 0.515343390787002,
"learning_rate": 9.297238110547075e-06,
"loss": 0.3372,
"step": 128
},
{
"epoch": 3.9417721518987343,
"grad_norm": 0.5596963864172609,
"learning_rate": 9.14141858983434e-06,
"loss": 0.3051,
"step": 129
},
{
"epoch": 3.972151898734177,
"grad_norm": 0.4734596843335074,
"learning_rate": 8.985808808641883e-06,
"loss": 0.3027,
"step": 130
},
{
"epoch": 4.010126582278481,
"grad_norm": 0.49200411877665867,
"learning_rate": 8.830446780279175e-06,
"loss": 0.3019,
"step": 131
},
{
"epoch": 4.040506329113924,
"grad_norm": 0.5695776585815239,
"learning_rate": 8.675370457533122e-06,
"loss": 0.2514,
"step": 132
},
{
"epoch": 4.0708860759493675,
"grad_norm": 0.4357481693196861,
"learning_rate": 8.520617723396702e-06,
"loss": 0.2748,
"step": 133
},
{
"epoch": 4.10126582278481,
"grad_norm": 0.5509843753267881,
"learning_rate": 8.366226381814698e-06,
"loss": 0.2538,
"step": 134
},
{
"epoch": 4.131645569620253,
"grad_norm": 0.6348670712782296,
"learning_rate": 8.212234148448708e-06,
"loss": 0.2544,
"step": 135
},
{
"epoch": 4.162025316455696,
"grad_norm": 0.4524695786647717,
"learning_rate": 8.058678641463724e-06,
"loss": 0.2558,
"step": 136
},
{
"epoch": 4.192405063291139,
"grad_norm": 0.47506577387319127,
"learning_rate": 7.905597372338558e-06,
"loss": 0.2386,
"step": 137
},
{
"epoch": 4.222784810126583,
"grad_norm": 0.5155716168871031,
"learning_rate": 7.753027736702283e-06,
"loss": 0.2229,
"step": 138
},
{
"epoch": 4.253164556962025,
"grad_norm": 0.4585368837376427,
"learning_rate": 7.601007005199022e-06,
"loss": 0.2217,
"step": 139
},
{
"epoch": 4.283544303797468,
"grad_norm": 0.41330785361461114,
"learning_rate": 7.449572314383237e-06,
"loss": 0.219,
"step": 140
},
{
"epoch": 4.313924050632911,
"grad_norm": 0.4830241066548344,
"learning_rate": 7.298760657647779e-06,
"loss": 0.2694,
"step": 141
},
{
"epoch": 4.3443037974683545,
"grad_norm": 0.5111579283891808,
"learning_rate": 7.148608876186931e-06,
"loss": 0.237,
"step": 142
},
{
"epoch": 4.374683544303798,
"grad_norm": 0.4501333280070922,
"learning_rate": 6.999153649996595e-06,
"loss": 0.219,
"step": 143
},
{
"epoch": 4.405063291139241,
"grad_norm": 0.4677430262051855,
"learning_rate": 6.8504314889138956e-06,
"loss": 0.2681,
"step": 144
},
{
"epoch": 4.435443037974683,
"grad_norm": 0.453358547318383,
"learning_rate": 6.702478723698336e-06,
"loss": 0.2234,
"step": 145
},
{
"epoch": 4.465822784810126,
"grad_norm": 0.5155045393434438,
"learning_rate": 6.555331497156671e-06,
"loss": 0.2311,
"step": 146
},
{
"epoch": 4.49620253164557,
"grad_norm": 0.46182713500382666,
"learning_rate": 6.4090257553137566e-06,
"loss": 0.219,
"step": 147
},
{
"epoch": 4.526582278481013,
"grad_norm": 0.42303706778202704,
"learning_rate": 6.263597238631405e-06,
"loss": 0.2596,
"step": 148
},
{
"epoch": 4.556962025316456,
"grad_norm": 0.432662674573151,
"learning_rate": 6.119081473277502e-06,
"loss": 0.2196,
"step": 149
},
{
"epoch": 4.587341772151898,
"grad_norm": 0.4234332930723555,
"learning_rate": 5.975513762447465e-06,
"loss": 0.2243,
"step": 150
},
{
"epoch": 4.6177215189873415,
"grad_norm": 0.42460055403171637,
"learning_rate": 5.832929177740134e-06,
"loss": 0.27,
"step": 151
},
{
"epoch": 4.648101265822785,
"grad_norm": 0.42506826817752974,
"learning_rate": 5.6913625505902966e-06,
"loss": 0.2326,
"step": 152
},
{
"epoch": 4.678481012658228,
"grad_norm": 0.39979191016820853,
"learning_rate": 5.550848463759835e-06,
"loss": 0.2515,
"step": 153
},
{
"epoch": 4.708860759493671,
"grad_norm": 0.4008425062621252,
"learning_rate": 5.411421242889643e-06,
"loss": 0.2568,
"step": 154
},
{
"epoch": 4.739240506329114,
"grad_norm": 0.42723429229383986,
"learning_rate": 5.273114948114346e-06,
"loss": 0.2254,
"step": 155
},
{
"epoch": 4.769620253164557,
"grad_norm": 0.42260474316931795,
"learning_rate": 5.135963365741892e-06,
"loss": 0.2251,
"step": 156
},
{
"epoch": 4.8,
"grad_norm": 0.4067240585220915,
"learning_rate": 5.000000000000003e-06,
"loss": 0.2606,
"step": 157
},
{
"epoch": 4.830379746835443,
"grad_norm": 0.38823072250163293,
"learning_rate": 4.865258064851579e-06,
"loss": 0.2372,
"step": 158
},
{
"epoch": 4.860759493670886,
"grad_norm": 0.4191822779374117,
"learning_rate": 4.731770475880995e-06,
"loss": 0.2486,
"step": 159
},
{
"epoch": 4.891139240506329,
"grad_norm": 0.4461708890820075,
"learning_rate": 4.599569842253244e-06,
"loss": 0.2237,
"step": 160
},
{
"epoch": 4.921518987341772,
"grad_norm": 0.4174090508507243,
"learning_rate": 4.468688458748006e-06,
"loss": 0.2456,
"step": 161
},
{
"epoch": 4.951898734177215,
"grad_norm": 0.3874314184390399,
"learning_rate": 4.339158297870469e-06,
"loss": 0.2122,
"step": 162
},
{
"epoch": 4.982278481012658,
"grad_norm": 0.42666677238256934,
"learning_rate": 4.211011002040885e-06,
"loss": 0.2381,
"step": 163
},
{
"epoch": 5.020253164556962,
"grad_norm": 0.46547502767623333,
"learning_rate": 4.084277875864776e-06,
"loss": 0.2435,
"step": 164
},
{
"epoch": 5.050632911392405,
"grad_norm": 0.5162266310966511,
"learning_rate": 3.958989878485644e-06,
"loss": 0.211,
"step": 165
},
{
"epoch": 5.0810126582278485,
"grad_norm": 0.4134007391032785,
"learning_rate": 3.83517761602209e-06,
"loss": 0.2125,
"step": 166
},
{
"epoch": 5.111392405063291,
"grad_norm": 0.3473501092708669,
"learning_rate": 3.712871334091154e-06,
"loss": 0.2237,
"step": 167
},
{
"epoch": 5.141772151898734,
"grad_norm": 0.3696619109171167,
"learning_rate": 3.592100910419738e-06,
"loss": 0.1855,
"step": 168
},
{
"epoch": 5.172151898734177,
"grad_norm": 0.5069681270754923,
"learning_rate": 3.4728958475459052e-06,
"loss": 0.2202,
"step": 169
},
{
"epoch": 5.2025316455696204,
"grad_norm": 0.4762521157436079,
"learning_rate": 3.355285265611784e-06,
"loss": 0.1671,
"step": 170
},
{
"epoch": 5.232911392405064,
"grad_norm": 0.47960208578525826,
"learning_rate": 3.2392978952499553e-06,
"loss": 0.1951,
"step": 171
},
{
"epoch": 5.263291139240506,
"grad_norm": 0.35924733837405015,
"learning_rate": 3.1249620705649417e-06,
"loss": 0.2149,
"step": 172
},
{
"epoch": 5.293670886075949,
"grad_norm": 0.348448947623233,
"learning_rate": 3.0123057222115835e-06,
"loss": 0.1779,
"step": 173
},
{
"epoch": 5.324050632911392,
"grad_norm": 0.404257135957341,
"learning_rate": 2.9013563705719673e-06,
"loss": 0.1839,
"step": 174
},
{
"epoch": 5.3544303797468356,
"grad_norm": 0.40593422037271537,
"learning_rate": 2.7921411190325753e-06,
"loss": 0.2051,
"step": 175
},
{
"epoch": 5.384810126582279,
"grad_norm": 0.3521219402602394,
"learning_rate": 2.6846866473633126e-06,
"loss": 0.185,
"step": 176
},
{
"epoch": 5.415189873417722,
"grad_norm": 0.32273379765564403,
"learning_rate": 2.579019205199992e-06,
"loss": 0.2011,
"step": 177
},
{
"epoch": 5.445569620253164,
"grad_norm": 0.32490098676443463,
"learning_rate": 2.4751646056319334e-06,
"loss": 0.1991,
"step": 178
},
{
"epoch": 5.4759493670886075,
"grad_norm": 0.3544739337877738,
"learning_rate": 2.373148218896182e-06,
"loss": 0.1772,
"step": 179
},
{
"epoch": 5.506329113924051,
"grad_norm": 0.3577236527157084,
"learning_rate": 2.2729949661798876e-06,
"loss": 0.2011,
"step": 180
},
{
"epoch": 5.536708860759494,
"grad_norm": 0.3506390551925719,
"learning_rate": 2.174729313532433e-06,
"loss": 0.2106,
"step": 181
},
{
"epoch": 5.567088607594937,
"grad_norm": 0.34715119704574837,
"learning_rate": 2.078375265888707e-06,
"loss": 0.1798,
"step": 182
},
{
"epoch": 5.597468354430379,
"grad_norm": 0.3651223054017652,
"learning_rate": 1.9839563612050273e-06,
"loss": 0.215,
"step": 183
},
{
"epoch": 5.627848101265823,
"grad_norm": 0.33377854019749137,
"learning_rate": 1.8914956647091497e-06,
"loss": 0.1856,
"step": 184
},
{
"epoch": 5.658227848101266,
"grad_norm": 0.351922253411728,
"learning_rate": 1.8010157632657544e-06,
"loss": 0.2213,
"step": 185
},
{
"epoch": 5.688607594936709,
"grad_norm": 0.35104562168350345,
"learning_rate": 1.7125387598587862e-06,
"loss": 0.2019,
"step": 186
},
{
"epoch": 5.718987341772152,
"grad_norm": 0.31701739152330494,
"learning_rate": 1.6260862681919965e-06,
"loss": 0.1541,
"step": 187
},
{
"epoch": 5.749367088607595,
"grad_norm": 0.34192159832257357,
"learning_rate": 1.5416794074090258e-06,
"loss": 0.1714,
"step": 188
},
{
"epoch": 5.779746835443038,
"grad_norm": 0.3200613805323288,
"learning_rate": 1.459338796934293e-06,
"loss": 0.1575,
"step": 189
},
{
"epoch": 5.810126582278481,
"grad_norm": 0.3553716020021964,
"learning_rate": 1.3790845514359363e-06,
"loss": 0.1565,
"step": 190
},
{
"epoch": 5.840506329113924,
"grad_norm": 0.3176563059889475,
"learning_rate": 1.300936275912098e-06,
"loss": 0.1609,
"step": 191
},
{
"epoch": 5.870886075949367,
"grad_norm": 0.3236672089672163,
"learning_rate": 1.224913060901688e-06,
"loss": 0.2001,
"step": 192
},
{
"epoch": 5.9012658227848105,
"grad_norm": 0.3540755256144387,
"learning_rate": 1.1510334778208332e-06,
"loss": 0.2026,
"step": 193
},
{
"epoch": 5.931645569620253,
"grad_norm": 0.33591079192847045,
"learning_rate": 1.0793155744261352e-06,
"loss": 0.2137,
"step": 194
},
{
"epoch": 5.962025316455696,
"grad_norm": 0.3268029265342323,
"learning_rate": 1.0097768704058542e-06,
"loss": 0.1824,
"step": 195
},
{
"epoch": 5.992405063291139,
"grad_norm": 0.3343098338067192,
"learning_rate": 9.424343531000968e-07,
"loss": 0.2031,
"step": 196
},
{
"epoch": 6.030379746835443,
"grad_norm": 0.3588586366271071,
"learning_rate": 8.773044733510338e-07,
"loss": 0.1966,
"step": 197
},
{
"epoch": 6.060759493670886,
"grad_norm": 0.2713228166680348,
"learning_rate": 8.144031414842012e-07,
"loss": 0.1619,
"step": 198
},
{
"epoch": 6.091139240506329,
"grad_norm": 0.3500882262386705,
"learning_rate": 7.537457234218271e-07,
"loss": 0.1907,
"step": 199
},
{
"epoch": 6.121518987341772,
"grad_norm": 0.335448550594209,
"learning_rate": 6.953470369291349e-07,
"loss": 0.1711,
"step": 200
},
{
"epoch": 6.151898734177215,
"grad_norm": 0.31632384454067425,
"learning_rate": 6.392213479945852e-07,
"loss": 0.2338,
"step": 201
},
{
"epoch": 6.182278481012658,
"grad_norm": 0.30598375246696663,
"learning_rate": 5.853823673448877e-07,
"loss": 0.1467,
"step": 202
},
{
"epoch": 6.2126582278481015,
"grad_norm": 0.3001671556277259,
"learning_rate": 5.33843247095659e-07,
"loss": 0.1907,
"step": 203
},
{
"epoch": 6.243037974683545,
"grad_norm": 0.28456378192655113,
"learning_rate": 4.846165775385459e-07,
"loss": 0.1429,
"step": 204
},
{
"epoch": 6.273417721518987,
"grad_norm": 0.2908463467436283,
"learning_rate": 4.3771438406559173e-07,
"loss": 0.1648,
"step": 205
},
{
"epoch": 6.30379746835443,
"grad_norm": 0.3047163062035505,
"learning_rate": 3.931481242315993e-07,
"loss": 0.169,
"step": 206
},
{
"epoch": 6.334177215189873,
"grad_norm": 0.2925779697765088,
"learning_rate": 3.5092868495520294e-07,
"loss": 0.1797,
"step": 207
},
{
"epoch": 6.364556962025317,
"grad_norm": 0.2874622808103571,
"learning_rate": 3.110663798593616e-07,
"loss": 0.1497,
"step": 208
},
{
"epoch": 6.39493670886076,
"grad_norm": 0.2903038716340111,
"learning_rate": 2.735709467518699e-07,
"loss": 0.1658,
"step": 209
},
{
"epoch": 6.425316455696202,
"grad_norm": 0.29311481243284,
"learning_rate": 2.384515452465475e-07,
"loss": 0.1859,
"step": 210
},
{
"epoch": 6.455696202531645,
"grad_norm": 0.29807460647131795,
"learning_rate": 2.0571675452567997e-07,
"loss": 0.1688,
"step": 211
},
{
"epoch": 6.4860759493670885,
"grad_norm": 0.31717937043968736,
"learning_rate": 1.7537457124423896e-07,
"loss": 0.1891,
"step": 212
},
{
"epoch": 6.516455696202532,
"grad_norm": 0.29155618807746464,
"learning_rate": 1.474324075764111e-07,
"loss": 0.1711,
"step": 213
},
{
"epoch": 6.546835443037975,
"grad_norm": 0.2934592403237966,
"learning_rate": 1.2189708940490653e-07,
"loss": 0.1777,
"step": 214
},
{
"epoch": 6.577215189873417,
"grad_norm": 0.3033189007155214,
"learning_rate": 9.877485465349057e-08,
"loss": 0.1825,
"step": 215
},
{
"epoch": 6.6075949367088604,
"grad_norm": 0.28465506705550975,
"learning_rate": 7.807135176314707e-08,
"loss": 0.1685,
"step": 216
},
{
"epoch": 6.637974683544304,
"grad_norm": 0.28494872266415744,
"learning_rate": 5.979163831223988e-08,
"loss": 0.1835,
"step": 217
},
{
"epoch": 6.668354430379747,
"grad_norm": 0.28501659601138096,
"learning_rate": 4.394017978101905e-08,
"loss": 0.1633,
"step": 218
},
{
"epoch": 6.69873417721519,
"grad_norm": 0.2913187231660811,
"learning_rate": 3.0520848460765525e-08,
"loss": 0.18,
"step": 219
},
{
"epoch": 6.729113924050633,
"grad_norm": 0.2898243561952896,
"learning_rate": 1.9536922507841227e-08,
"loss": 0.1688,
"step": 220
},
{
"epoch": 6.759493670886076,
"grad_norm": 0.30141955235920126,
"learning_rate": 1.099108514288627e-08,
"loss": 0.2004,
"step": 221
},
{
"epoch": 6.789873417721519,
"grad_norm": 0.2967842814028097,
"learning_rate": 4.885423995341088e-09,
"loss": 0.1789,
"step": 222
},
{
"epoch": 6.820253164556962,
"grad_norm": 0.28235982181228925,
"learning_rate": 1.2214305934699078e-09,
"loss": 0.1866,
"step": 223
},
{
"epoch": 6.850632911392405,
"grad_norm": 0.30226486624869175,
"learning_rate": 0.0,
"loss": 0.2146,
"step": 224
},
{
"epoch": 6.850632911392405,
"step": 224,
"total_flos": 3.788104761815859e+17,
"train_loss": 0.40834243855039987,
"train_runtime": 25597.751,
"train_samples_per_second": 0.864,
"train_steps_per_second": 0.009
}
],
"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": 3.788104761815859e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}