emrecan's picture
End of training
29c4bf7
raw
history blame
33.1 kB
{
"best_metric": 0.7648497198166072,
"best_model_checkpoint": "/home/user/emrecan/models/bert-base-multilingual-cased_allnli_tr/checkpoint-80000",
"epoch": 3.0,
"global_step": 88320,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03,
"learning_rate": 1.9773550724637682e-05,
"loss": 0.8623,
"step": 1000
},
{
"epoch": 0.03,
"eval_accuracy": 0.5917473255221599,
"eval_loss": 0.9075943827629089,
"eval_runtime": 49.3629,
"eval_samples_per_second": 198.833,
"eval_steps_per_second": 6.219,
"step": 1000
},
{
"epoch": 0.07,
"learning_rate": 1.9547101449275363e-05,
"loss": 0.7528,
"step": 2000
},
{
"epoch": 0.07,
"eval_accuracy": 0.6119205298013245,
"eval_loss": 0.8586783409118652,
"eval_runtime": 49.2395,
"eval_samples_per_second": 199.332,
"eval_steps_per_second": 6.235,
"step": 2000
},
{
"epoch": 0.1,
"learning_rate": 1.9320652173913047e-05,
"loss": 0.7074,
"step": 3000
},
{
"epoch": 0.1,
"eval_accuracy": 0.664696892511462,
"eval_loss": 0.7866753935813904,
"eval_runtime": 49.3359,
"eval_samples_per_second": 198.942,
"eval_steps_per_second": 6.223,
"step": 3000
},
{
"epoch": 0.14,
"learning_rate": 1.9094202898550727e-05,
"loss": 0.6949,
"step": 4000
},
{
"epoch": 0.14,
"eval_accuracy": 0.6772287315333673,
"eval_loss": 0.7473878264427185,
"eval_runtime": 49.2961,
"eval_samples_per_second": 199.103,
"eval_steps_per_second": 6.228,
"step": 4000
},
{
"epoch": 0.17,
"learning_rate": 1.8867753623188408e-05,
"loss": 0.6681,
"step": 5000
},
{
"epoch": 0.17,
"eval_accuracy": 0.6814060112073357,
"eval_loss": 0.7661426067352295,
"eval_runtime": 49.2761,
"eval_samples_per_second": 199.184,
"eval_steps_per_second": 6.23,
"step": 5000
},
{
"epoch": 0.2,
"learning_rate": 1.864130434782609e-05,
"loss": 0.6597,
"step": 6000
},
{
"epoch": 0.2,
"eval_accuracy": 0.6943453897096281,
"eval_loss": 0.7264170050621033,
"eval_runtime": 49.2718,
"eval_samples_per_second": 199.201,
"eval_steps_per_second": 6.231,
"step": 6000
},
{
"epoch": 0.24,
"learning_rate": 1.841485507246377e-05,
"loss": 0.6495,
"step": 7000
},
{
"epoch": 0.24,
"eval_accuracy": 0.6781456953642384,
"eval_loss": 0.7841250896453857,
"eval_runtime": 49.2905,
"eval_samples_per_second": 199.126,
"eval_steps_per_second": 6.228,
"step": 7000
},
{
"epoch": 0.27,
"learning_rate": 1.818840579710145e-05,
"loss": 0.6323,
"step": 8000
},
{
"epoch": 0.27,
"eval_accuracy": 0.6951604686704025,
"eval_loss": 0.7256377339363098,
"eval_runtime": 49.2057,
"eval_samples_per_second": 199.469,
"eval_steps_per_second": 6.239,
"step": 8000
},
{
"epoch": 0.31,
"learning_rate": 1.7961956521739134e-05,
"loss": 0.6308,
"step": 9000
},
{
"epoch": 0.31,
"eval_accuracy": 0.6957717778909832,
"eval_loss": 0.7319093346595764,
"eval_runtime": 49.2073,
"eval_samples_per_second": 199.462,
"eval_steps_per_second": 6.239,
"step": 9000
},
{
"epoch": 0.34,
"learning_rate": 1.7735507246376815e-05,
"loss": 0.6254,
"step": 10000
},
{
"epoch": 0.34,
"eval_accuracy": 0.7003565970453388,
"eval_loss": 0.7053707838058472,
"eval_runtime": 49.2797,
"eval_samples_per_second": 199.169,
"eval_steps_per_second": 6.23,
"step": 10000
},
{
"epoch": 0.37,
"learning_rate": 1.7509057971014495e-05,
"loss": 0.6233,
"step": 11000
},
{
"epoch": 0.37,
"eval_accuracy": 0.708507386653082,
"eval_loss": 0.7068630456924438,
"eval_runtime": 55.8024,
"eval_samples_per_second": 175.889,
"eval_steps_per_second": 5.502,
"step": 11000
},
{
"epoch": 0.41,
"learning_rate": 1.7282608695652176e-05,
"loss": 0.6165,
"step": 12000
},
{
"epoch": 0.41,
"eval_accuracy": 0.7180845644421804,
"eval_loss": 0.687969982624054,
"eval_runtime": 50.9115,
"eval_samples_per_second": 192.785,
"eval_steps_per_second": 6.03,
"step": 12000
},
{
"epoch": 0.44,
"learning_rate": 1.7056159420289856e-05,
"loss": 0.6033,
"step": 13000
},
{
"epoch": 0.44,
"eval_accuracy": 0.719714722363729,
"eval_loss": 0.6844114065170288,
"eval_runtime": 49.3739,
"eval_samples_per_second": 198.789,
"eval_steps_per_second": 6.218,
"step": 13000
},
{
"epoch": 0.48,
"learning_rate": 1.6829710144927537e-05,
"loss": 0.6014,
"step": 14000
},
{
"epoch": 0.48,
"eval_accuracy": 0.7128884360672441,
"eval_loss": 0.6752753853797913,
"eval_runtime": 49.3697,
"eval_samples_per_second": 198.806,
"eval_steps_per_second": 6.218,
"step": 14000
},
{
"epoch": 0.51,
"learning_rate": 1.6603260869565218e-05,
"loss": 0.5947,
"step": 15000
},
{
"epoch": 0.51,
"eval_accuracy": 0.7039225674987264,
"eval_loss": 0.7000291347503662,
"eval_runtime": 49.3889,
"eval_samples_per_second": 198.729,
"eval_steps_per_second": 6.216,
"step": 15000
},
{
"epoch": 0.54,
"learning_rate": 1.6376811594202898e-05,
"loss": 0.5965,
"step": 16000
},
{
"epoch": 0.54,
"eval_accuracy": 0.7263372389200203,
"eval_loss": 0.670754075050354,
"eval_runtime": 49.311,
"eval_samples_per_second": 199.043,
"eval_steps_per_second": 6.226,
"step": 16000
},
{
"epoch": 0.58,
"learning_rate": 1.615036231884058e-05,
"loss": 0.5979,
"step": 17000
},
{
"epoch": 0.58,
"eval_accuracy": 0.7284768211920529,
"eval_loss": 0.656209409236908,
"eval_runtime": 49.4043,
"eval_samples_per_second": 198.667,
"eval_steps_per_second": 6.214,
"step": 17000
},
{
"epoch": 0.61,
"learning_rate": 1.5923913043478263e-05,
"loss": 0.5787,
"step": 18000
},
{
"epoch": 0.61,
"eval_accuracy": 0.7296994396332145,
"eval_loss": 0.6554355621337891,
"eval_runtime": 50.2795,
"eval_samples_per_second": 195.209,
"eval_steps_per_second": 6.106,
"step": 18000
},
{
"epoch": 0.65,
"learning_rate": 1.5697463768115943e-05,
"loss": 0.58,
"step": 19000
},
{
"epoch": 0.65,
"eval_accuracy": 0.7315333672949567,
"eval_loss": 0.6544056534767151,
"eval_runtime": 49.3813,
"eval_samples_per_second": 198.76,
"eval_steps_per_second": 6.217,
"step": 19000
},
{
"epoch": 0.68,
"learning_rate": 1.5471014492753624e-05,
"loss": 0.574,
"step": 20000
},
{
"epoch": 0.68,
"eval_accuracy": 0.7338767193071829,
"eval_loss": 0.6549123525619507,
"eval_runtime": 49.3189,
"eval_samples_per_second": 199.011,
"eval_steps_per_second": 6.225,
"step": 20000
},
{
"epoch": 0.71,
"learning_rate": 1.5244565217391305e-05,
"loss": 0.5751,
"step": 21000
},
{
"epoch": 0.71,
"eval_accuracy": 0.7288843606724401,
"eval_loss": 0.6545295715332031,
"eval_runtime": 52.3727,
"eval_samples_per_second": 187.407,
"eval_steps_per_second": 5.862,
"step": 21000
},
{
"epoch": 0.75,
"learning_rate": 1.5018115942028985e-05,
"loss": 0.5659,
"step": 22000
},
{
"epoch": 0.75,
"eval_accuracy": 0.7371370351502802,
"eval_loss": 0.6466848850250244,
"eval_runtime": 51.7409,
"eval_samples_per_second": 189.695,
"eval_steps_per_second": 5.933,
"step": 22000
},
{
"epoch": 0.78,
"learning_rate": 1.479166666666667e-05,
"loss": 0.5732,
"step": 23000
},
{
"epoch": 0.78,
"eval_accuracy": 0.736220071319409,
"eval_loss": 0.6447662711143494,
"eval_runtime": 51.1755,
"eval_samples_per_second": 191.791,
"eval_steps_per_second": 5.999,
"step": 23000
},
{
"epoch": 0.82,
"learning_rate": 1.456521739130435e-05,
"loss": 0.5637,
"step": 24000
},
{
"epoch": 0.82,
"eval_accuracy": 0.7355068772287315,
"eval_loss": 0.6520141959190369,
"eval_runtime": 51.0797,
"eval_samples_per_second": 192.151,
"eval_steps_per_second": 6.01,
"step": 24000
},
{
"epoch": 0.85,
"learning_rate": 1.433876811594203e-05,
"loss": 0.5648,
"step": 25000
},
{
"epoch": 0.85,
"eval_accuracy": 0.7344880285277636,
"eval_loss": 0.6411919593811035,
"eval_runtime": 51.1242,
"eval_samples_per_second": 191.983,
"eval_steps_per_second": 6.005,
"step": 25000
},
{
"epoch": 0.88,
"learning_rate": 1.4112318840579711e-05,
"loss": 0.5622,
"step": 26000
},
{
"epoch": 0.88,
"eval_accuracy": 0.735812531839022,
"eval_loss": 0.6350020170211792,
"eval_runtime": 51.0256,
"eval_samples_per_second": 192.355,
"eval_steps_per_second": 6.017,
"step": 26000
},
{
"epoch": 0.92,
"learning_rate": 1.3885869565217392e-05,
"loss": 0.5579,
"step": 27000
},
{
"epoch": 0.92,
"eval_accuracy": 0.7392766174223128,
"eval_loss": 0.6346594095230103,
"eval_runtime": 51.1443,
"eval_samples_per_second": 191.908,
"eval_steps_per_second": 6.003,
"step": 27000
},
{
"epoch": 0.95,
"learning_rate": 1.3659420289855074e-05,
"loss": 0.5518,
"step": 28000
},
{
"epoch": 0.95,
"eval_accuracy": 0.739174732552216,
"eval_loss": 0.6417071223258972,
"eval_runtime": 52.294,
"eval_samples_per_second": 187.689,
"eval_steps_per_second": 5.871,
"step": 28000
},
{
"epoch": 0.99,
"learning_rate": 1.3432971014492755e-05,
"loss": 0.5547,
"step": 29000
},
{
"epoch": 0.99,
"eval_accuracy": 0.7436576668364748,
"eval_loss": 0.6321312785148621,
"eval_runtime": 51.1442,
"eval_samples_per_second": 191.908,
"eval_steps_per_second": 6.003,
"step": 29000
},
{
"epoch": 1.02,
"learning_rate": 1.3206521739130435e-05,
"loss": 0.524,
"step": 30000
},
{
"epoch": 1.02,
"eval_accuracy": 0.7412124299541518,
"eval_loss": 0.6430493593215942,
"eval_runtime": 51.2108,
"eval_samples_per_second": 191.659,
"eval_steps_per_second": 5.995,
"step": 30000
},
{
"epoch": 1.05,
"learning_rate": 1.2980072463768116e-05,
"loss": 0.4982,
"step": 31000
},
{
"epoch": 1.05,
"eval_accuracy": 0.7457972491085074,
"eval_loss": 0.6252649426460266,
"eval_runtime": 57.7511,
"eval_samples_per_second": 169.954,
"eval_steps_per_second": 5.316,
"step": 31000
},
{
"epoch": 1.09,
"learning_rate": 1.2753623188405797e-05,
"loss": 0.5002,
"step": 32000
},
{
"epoch": 1.09,
"eval_accuracy": 0.7418237391747325,
"eval_loss": 0.6316350698471069,
"eval_runtime": 52.7966,
"eval_samples_per_second": 185.902,
"eval_steps_per_second": 5.815,
"step": 32000
},
{
"epoch": 1.12,
"learning_rate": 1.252717391304348e-05,
"loss": 0.4993,
"step": 33000
},
{
"epoch": 1.12,
"eval_accuracy": 0.7486500254712175,
"eval_loss": 0.6196975111961365,
"eval_runtime": 51.1876,
"eval_samples_per_second": 191.746,
"eval_steps_per_second": 5.998,
"step": 33000
},
{
"epoch": 1.15,
"learning_rate": 1.2300724637681161e-05,
"loss": 0.4963,
"step": 34000
},
{
"epoch": 1.15,
"eval_accuracy": 0.7462047885888945,
"eval_loss": 0.6307246685028076,
"eval_runtime": 51.0522,
"eval_samples_per_second": 192.254,
"eval_steps_per_second": 6.013,
"step": 34000
},
{
"epoch": 1.19,
"learning_rate": 1.2074275362318842e-05,
"loss": 0.504,
"step": 35000
},
{
"epoch": 1.19,
"eval_accuracy": 0.7480387162506368,
"eval_loss": 0.627221405506134,
"eval_runtime": 51.0856,
"eval_samples_per_second": 192.129,
"eval_steps_per_second": 6.01,
"step": 35000
},
{
"epoch": 1.22,
"learning_rate": 1.1847826086956522e-05,
"loss": 0.4922,
"step": 36000
},
{
"epoch": 1.22,
"eval_accuracy": 0.7432501273560876,
"eval_loss": 0.6410390138626099,
"eval_runtime": 51.1349,
"eval_samples_per_second": 191.943,
"eval_steps_per_second": 6.004,
"step": 36000
},
{
"epoch": 1.26,
"learning_rate": 1.1621376811594205e-05,
"loss": 0.5016,
"step": 37000
},
{
"epoch": 1.26,
"eval_accuracy": 0.7461029037187977,
"eval_loss": 0.6295490860939026,
"eval_runtime": 51.0732,
"eval_samples_per_second": 192.175,
"eval_steps_per_second": 6.011,
"step": 37000
},
{
"epoch": 1.29,
"learning_rate": 1.1394927536231885e-05,
"loss": 0.4957,
"step": 38000
},
{
"epoch": 1.29,
"eval_accuracy": 0.7505858380030566,
"eval_loss": 0.6182788014411926,
"eval_runtime": 51.1969,
"eval_samples_per_second": 191.711,
"eval_steps_per_second": 5.996,
"step": 38000
},
{
"epoch": 1.32,
"learning_rate": 1.1168478260869566e-05,
"loss": 0.4883,
"step": 39000
},
{
"epoch": 1.32,
"eval_accuracy": 0.7501782985226694,
"eval_loss": 0.6260754466056824,
"eval_runtime": 51.1903,
"eval_samples_per_second": 191.735,
"eval_steps_per_second": 5.997,
"step": 39000
},
{
"epoch": 1.36,
"learning_rate": 1.0942028985507247e-05,
"loss": 0.4985,
"step": 40000
},
{
"epoch": 1.36,
"eval_accuracy": 0.7495669893020886,
"eval_loss": 0.6315430402755737,
"eval_runtime": 51.1506,
"eval_samples_per_second": 191.884,
"eval_steps_per_second": 6.002,
"step": 40000
},
{
"epoch": 1.39,
"learning_rate": 1.0715579710144927e-05,
"loss": 0.4885,
"step": 41000
},
{
"epoch": 1.39,
"eval_accuracy": 0.7529291900152827,
"eval_loss": 0.6188690066337585,
"eval_runtime": 52.5846,
"eval_samples_per_second": 186.651,
"eval_steps_per_second": 5.838,
"step": 41000
},
{
"epoch": 1.43,
"learning_rate": 1.0489130434782611e-05,
"loss": 0.4909,
"step": 42000
},
{
"epoch": 1.43,
"eval_accuracy": 0.7473255221599593,
"eval_loss": 0.6188654899597168,
"eval_runtime": 51.6594,
"eval_samples_per_second": 189.994,
"eval_steps_per_second": 5.943,
"step": 42000
},
{
"epoch": 1.46,
"learning_rate": 1.0262681159420292e-05,
"loss": 0.4894,
"step": 43000
},
{
"epoch": 1.46,
"eval_accuracy": 0.7432501273560876,
"eval_loss": 0.631429135799408,
"eval_runtime": 51.1492,
"eval_samples_per_second": 191.889,
"eval_steps_per_second": 6.002,
"step": 43000
},
{
"epoch": 1.49,
"learning_rate": 1.0036231884057972e-05,
"loss": 0.4912,
"step": 44000
},
{
"epoch": 1.49,
"eval_accuracy": 0.7445746306673459,
"eval_loss": 0.6183902025222778,
"eval_runtime": 51.0914,
"eval_samples_per_second": 192.107,
"eval_steps_per_second": 6.009,
"step": 44000
},
{
"epoch": 1.53,
"learning_rate": 9.809782608695653e-06,
"loss": 0.4851,
"step": 45000
},
{
"epoch": 1.53,
"eval_accuracy": 0.7461029037187977,
"eval_loss": 0.6257576942443848,
"eval_runtime": 50.9837,
"eval_samples_per_second": 192.513,
"eval_steps_per_second": 6.022,
"step": 45000
},
{
"epoch": 1.56,
"learning_rate": 9.583333333333335e-06,
"loss": 0.4879,
"step": 46000
},
{
"epoch": 1.56,
"eval_accuracy": 0.7480387162506368,
"eval_loss": 0.6286013126373291,
"eval_runtime": 51.1172,
"eval_samples_per_second": 192.01,
"eval_steps_per_second": 6.006,
"step": 46000
},
{
"epoch": 1.6,
"learning_rate": 9.356884057971016e-06,
"loss": 0.4907,
"step": 47000
},
{
"epoch": 1.6,
"eval_accuracy": 0.7511971472236373,
"eval_loss": 0.6196326613426208,
"eval_runtime": 51.072,
"eval_samples_per_second": 192.18,
"eval_steps_per_second": 6.011,
"step": 47000
},
{
"epoch": 1.63,
"learning_rate": 9.130434782608697e-06,
"loss": 0.4884,
"step": 48000
},
{
"epoch": 1.63,
"eval_accuracy": 0.7526235354049924,
"eval_loss": 0.6156549453735352,
"eval_runtime": 51.1757,
"eval_samples_per_second": 191.79,
"eval_steps_per_second": 5.999,
"step": 48000
},
{
"epoch": 1.66,
"learning_rate": 8.903985507246377e-06,
"loss": 0.4755,
"step": 49000
},
{
"epoch": 1.66,
"eval_accuracy": 0.7591441670911869,
"eval_loss": 0.6055976152420044,
"eval_runtime": 50.9202,
"eval_samples_per_second": 192.752,
"eval_steps_per_second": 6.029,
"step": 49000
},
{
"epoch": 1.7,
"learning_rate": 8.677536231884058e-06,
"loss": 0.4811,
"step": 50000
},
{
"epoch": 1.7,
"eval_accuracy": 0.7582272032603159,
"eval_loss": 0.5976621508598328,
"eval_runtime": 51.1974,
"eval_samples_per_second": 191.709,
"eval_steps_per_second": 5.996,
"step": 50000
},
{
"epoch": 1.73,
"learning_rate": 8.45108695652174e-06,
"loss": 0.4787,
"step": 51000
},
{
"epoch": 1.73,
"eval_accuracy": 0.7620988283239939,
"eval_loss": 0.5914710164070129,
"eval_runtime": 52.4733,
"eval_samples_per_second": 187.047,
"eval_steps_per_second": 5.851,
"step": 51000
},
{
"epoch": 1.77,
"learning_rate": 8.22463768115942e-06,
"loss": 0.4779,
"step": 52000
},
{
"epoch": 1.77,
"eval_accuracy": 0.7583290881304127,
"eval_loss": 0.6014041304588318,
"eval_runtime": 51.6886,
"eval_samples_per_second": 189.887,
"eval_steps_per_second": 5.939,
"step": 52000
},
{
"epoch": 1.8,
"learning_rate": 7.998188405797103e-06,
"loss": 0.4767,
"step": 53000
},
{
"epoch": 1.8,
"eval_accuracy": 0.7623025980641874,
"eval_loss": 0.6041266918182373,
"eval_runtime": 51.0745,
"eval_samples_per_second": 192.17,
"eval_steps_per_second": 6.011,
"step": 53000
},
{
"epoch": 1.83,
"learning_rate": 7.771739130434784e-06,
"loss": 0.4737,
"step": 54000
},
{
"epoch": 1.83,
"eval_accuracy": 0.7562913907284768,
"eval_loss": 0.6093412637710571,
"eval_runtime": 51.0594,
"eval_samples_per_second": 192.227,
"eval_steps_per_second": 6.013,
"step": 54000
},
{
"epoch": 1.87,
"learning_rate": 7.545289855072464e-06,
"loss": 0.4836,
"step": 55000
},
{
"epoch": 1.87,
"eval_accuracy": 0.7568008150789608,
"eval_loss": 0.6001136302947998,
"eval_runtime": 51.1468,
"eval_samples_per_second": 191.899,
"eval_steps_per_second": 6.002,
"step": 55000
},
{
"epoch": 1.9,
"learning_rate": 7.318840579710146e-06,
"loss": 0.4765,
"step": 56000
},
{
"epoch": 1.9,
"eval_accuracy": 0.7600611309220581,
"eval_loss": 0.6109462380409241,
"eval_runtime": 51.0882,
"eval_samples_per_second": 192.119,
"eval_steps_per_second": 6.009,
"step": 56000
},
{
"epoch": 1.94,
"learning_rate": 7.092391304347826e-06,
"loss": 0.4776,
"step": 57000
},
{
"epoch": 1.94,
"eval_accuracy": 0.7598573611818645,
"eval_loss": 0.6045997142791748,
"eval_runtime": 51.0831,
"eval_samples_per_second": 192.138,
"eval_steps_per_second": 6.01,
"step": 57000
},
{
"epoch": 1.97,
"learning_rate": 6.865942028985509e-06,
"loss": 0.4769,
"step": 58000
},
{
"epoch": 1.97,
"eval_accuracy": 0.7568008150789608,
"eval_loss": 0.5969610214233398,
"eval_runtime": 51.092,
"eval_samples_per_second": 192.105,
"eval_steps_per_second": 6.009,
"step": 58000
},
{
"epoch": 2.0,
"learning_rate": 6.639492753623189e-06,
"loss": 0.4654,
"step": 59000
},
{
"epoch": 2.0,
"eval_accuracy": 0.7613856342333164,
"eval_loss": 0.6146702170372009,
"eval_runtime": 51.0794,
"eval_samples_per_second": 192.152,
"eval_steps_per_second": 6.01,
"step": 59000
},
{
"epoch": 2.04,
"learning_rate": 6.41304347826087e-06,
"loss": 0.4144,
"step": 60000
},
{
"epoch": 2.04,
"eval_accuracy": 0.7565970453387671,
"eval_loss": 0.6438983678817749,
"eval_runtime": 51.1073,
"eval_samples_per_second": 192.047,
"eval_steps_per_second": 6.007,
"step": 60000
},
{
"epoch": 2.07,
"learning_rate": 6.186594202898551e-06,
"loss": 0.4101,
"step": 61000
},
{
"epoch": 2.07,
"eval_accuracy": 0.7527254202750892,
"eval_loss": 0.637277364730835,
"eval_runtime": 52.7247,
"eval_samples_per_second": 186.156,
"eval_steps_per_second": 5.823,
"step": 61000
},
{
"epoch": 2.11,
"learning_rate": 5.960144927536232e-06,
"loss": 0.4192,
"step": 62000
},
{
"epoch": 2.11,
"eval_accuracy": 0.7575140091696383,
"eval_loss": 0.6135603189468384,
"eval_runtime": 52.6526,
"eval_samples_per_second": 186.411,
"eval_steps_per_second": 5.831,
"step": 62000
},
{
"epoch": 2.14,
"learning_rate": 5.733695652173914e-06,
"loss": 0.4128,
"step": 63000
},
{
"epoch": 2.14,
"eval_accuracy": 0.7559857361181864,
"eval_loss": 0.6282700896263123,
"eval_runtime": 51.1146,
"eval_samples_per_second": 192.02,
"eval_steps_per_second": 6.006,
"step": 63000
},
{
"epoch": 2.17,
"learning_rate": 5.507246376811595e-06,
"loss": 0.4204,
"step": 64000
},
{
"epoch": 2.17,
"eval_accuracy": 0.7625063678043811,
"eval_loss": 0.618690013885498,
"eval_runtime": 51.1974,
"eval_samples_per_second": 191.709,
"eval_steps_per_second": 5.996,
"step": 64000
},
{
"epoch": 2.21,
"learning_rate": 5.2807971014492755e-06,
"loss": 0.4114,
"step": 65000
},
{
"epoch": 2.21,
"eval_accuracy": 0.7620988283239939,
"eval_loss": 0.6127009987831116,
"eval_runtime": 51.1304,
"eval_samples_per_second": 191.96,
"eval_steps_per_second": 6.004,
"step": 65000
},
{
"epoch": 2.24,
"learning_rate": 5.054347826086957e-06,
"loss": 0.4097,
"step": 66000
},
{
"epoch": 2.24,
"eval_accuracy": 0.7626082526744778,
"eval_loss": 0.618818998336792,
"eval_runtime": 51.1752,
"eval_samples_per_second": 191.792,
"eval_steps_per_second": 5.999,
"step": 66000
},
{
"epoch": 2.28,
"learning_rate": 4.8278985507246375e-06,
"loss": 0.4129,
"step": 67000
},
{
"epoch": 2.28,
"eval_accuracy": 0.7639327559857361,
"eval_loss": 0.6156466603279114,
"eval_runtime": 51.1634,
"eval_samples_per_second": 191.836,
"eval_steps_per_second": 6.0,
"step": 67000
},
{
"epoch": 2.31,
"learning_rate": 4.601449275362319e-06,
"loss": 0.4085,
"step": 68000
},
{
"epoch": 2.31,
"eval_accuracy": 0.7615894039735099,
"eval_loss": 0.6232104301452637,
"eval_runtime": 51.1826,
"eval_samples_per_second": 191.764,
"eval_steps_per_second": 5.998,
"step": 68000
},
{
"epoch": 2.34,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.4074,
"step": 69000
},
{
"epoch": 2.34,
"eval_accuracy": 0.7604686704024453,
"eval_loss": 0.6239916682243347,
"eval_runtime": 51.0688,
"eval_samples_per_second": 192.192,
"eval_steps_per_second": 6.011,
"step": 69000
},
{
"epoch": 2.38,
"learning_rate": 4.148550724637682e-06,
"loss": 0.409,
"step": 70000
},
{
"epoch": 2.38,
"eval_accuracy": 0.7591441670911869,
"eval_loss": 0.6152721643447876,
"eval_runtime": 51.1662,
"eval_samples_per_second": 191.826,
"eval_steps_per_second": 6.0,
"step": 70000
},
{
"epoch": 2.41,
"learning_rate": 3.9221014492753625e-06,
"loss": 0.4046,
"step": 71000
},
{
"epoch": 2.41,
"eval_accuracy": 0.7587366276107997,
"eval_loss": 0.6375284194946289,
"eval_runtime": 52.7067,
"eval_samples_per_second": 186.219,
"eval_steps_per_second": 5.825,
"step": 71000
},
{
"epoch": 2.45,
"learning_rate": 3.6956521739130436e-06,
"loss": 0.4117,
"step": 72000
},
{
"epoch": 2.45,
"eval_accuracy": 0.7629139072847683,
"eval_loss": 0.6144647598266602,
"eval_runtime": 52.5798,
"eval_samples_per_second": 186.669,
"eval_steps_per_second": 5.839,
"step": 72000
},
{
"epoch": 2.48,
"learning_rate": 3.4692028985507246e-06,
"loss": 0.4002,
"step": 73000
},
{
"epoch": 2.48,
"eval_accuracy": 0.7609780947529292,
"eval_loss": 0.6278636455535889,
"eval_runtime": 51.1134,
"eval_samples_per_second": 192.024,
"eval_steps_per_second": 6.006,
"step": 73000
},
{
"epoch": 2.51,
"learning_rate": 3.242753623188406e-06,
"loss": 0.4042,
"step": 74000
},
{
"epoch": 2.51,
"eval_accuracy": 0.7646459500764137,
"eval_loss": 0.6176004409790039,
"eval_runtime": 51.1309,
"eval_samples_per_second": 191.958,
"eval_steps_per_second": 6.004,
"step": 74000
},
{
"epoch": 2.55,
"learning_rate": 3.016304347826087e-06,
"loss": 0.4055,
"step": 75000
},
{
"epoch": 2.55,
"eval_accuracy": 0.7643402954661233,
"eval_loss": 0.627702534198761,
"eval_runtime": 51.1017,
"eval_samples_per_second": 192.068,
"eval_steps_per_second": 6.008,
"step": 75000
},
{
"epoch": 2.58,
"learning_rate": 2.7898550724637686e-06,
"loss": 0.4021,
"step": 76000
},
{
"epoch": 2.58,
"eval_accuracy": 0.7642384105960265,
"eval_loss": 0.619607150554657,
"eval_runtime": 51.1091,
"eval_samples_per_second": 192.04,
"eval_steps_per_second": 6.007,
"step": 76000
},
{
"epoch": 2.62,
"learning_rate": 2.563405797101449e-06,
"loss": 0.4081,
"step": 77000
},
{
"epoch": 2.62,
"eval_accuracy": 0.7658685685175751,
"eval_loss": 0.6127172708511353,
"eval_runtime": 51.0673,
"eval_samples_per_second": 192.197,
"eval_steps_per_second": 6.012,
"step": 77000
},
{
"epoch": 2.65,
"learning_rate": 2.3369565217391307e-06,
"loss": 0.408,
"step": 78000
},
{
"epoch": 2.65,
"eval_accuracy": 0.7638308711156393,
"eval_loss": 0.6236501336097717,
"eval_runtime": 51.0905,
"eval_samples_per_second": 192.11,
"eval_steps_per_second": 6.009,
"step": 78000
},
{
"epoch": 2.68,
"learning_rate": 2.1105072463768117e-06,
"loss": 0.3997,
"step": 79000
},
{
"epoch": 2.68,
"eval_accuracy": 0.7636271013754458,
"eval_loss": 0.6190339922904968,
"eval_runtime": 51.097,
"eval_samples_per_second": 192.086,
"eval_steps_per_second": 6.008,
"step": 79000
},
{
"epoch": 2.72,
"learning_rate": 1.884057971014493e-06,
"loss": 0.4093,
"step": 80000
},
{
"epoch": 2.72,
"eval_accuracy": 0.7648497198166072,
"eval_loss": 0.615200936794281,
"eval_runtime": 51.0806,
"eval_samples_per_second": 192.147,
"eval_steps_per_second": 6.01,
"step": 80000
},
{
"epoch": 2.75,
"learning_rate": 1.657608695652174e-06,
"loss": 0.4095,
"step": 81000
},
{
"epoch": 2.75,
"eval_accuracy": 0.7627101375445746,
"eval_loss": 0.6154515743255615,
"eval_runtime": 52.4793,
"eval_samples_per_second": 187.026,
"eval_steps_per_second": 5.85,
"step": 81000
},
{
"epoch": 2.79,
"learning_rate": 1.4311594202898552e-06,
"loss": 0.4088,
"step": 82000
},
{
"epoch": 2.79,
"eval_accuracy": 0.7641365257259297,
"eval_loss": 0.6130374073982239,
"eval_runtime": 52.6234,
"eval_samples_per_second": 186.514,
"eval_steps_per_second": 5.834,
"step": 82000
},
{
"epoch": 2.82,
"learning_rate": 1.2047101449275363e-06,
"loss": 0.4063,
"step": 83000
},
{
"epoch": 2.82,
"eval_accuracy": 0.7646459500764137,
"eval_loss": 0.6072085499763489,
"eval_runtime": 51.0506,
"eval_samples_per_second": 192.26,
"eval_steps_per_second": 6.014,
"step": 83000
},
{
"epoch": 2.85,
"learning_rate": 9.782608695652175e-07,
"loss": 0.3978,
"step": 84000
},
{
"epoch": 2.85,
"eval_accuracy": 0.7661742231278655,
"eval_loss": 0.6128284931182861,
"eval_runtime": 51.2162,
"eval_samples_per_second": 191.639,
"eval_steps_per_second": 5.994,
"step": 84000
},
{
"epoch": 2.89,
"learning_rate": 7.518115942028987e-07,
"loss": 0.4034,
"step": 85000
},
{
"epoch": 2.89,
"eval_accuracy": 0.7627101375445746,
"eval_loss": 0.6156770586967468,
"eval_runtime": 51.1115,
"eval_samples_per_second": 192.031,
"eval_steps_per_second": 6.006,
"step": 85000
},
{
"epoch": 2.92,
"learning_rate": 5.253623188405797e-07,
"loss": 0.4044,
"step": 86000
},
{
"epoch": 2.92,
"eval_accuracy": 0.7660723382577688,
"eval_loss": 0.6127380728721619,
"eval_runtime": 51.0045,
"eval_samples_per_second": 192.434,
"eval_steps_per_second": 6.019,
"step": 86000
},
{
"epoch": 2.96,
"learning_rate": 2.989130434782609e-07,
"loss": 0.403,
"step": 87000
},
{
"epoch": 2.96,
"eval_accuracy": 0.7663779928680591,
"eval_loss": 0.612598717212677,
"eval_runtime": 51.1269,
"eval_samples_per_second": 191.973,
"eval_steps_per_second": 6.005,
"step": 87000
},
{
"epoch": 2.99,
"learning_rate": 7.246376811594204e-08,
"loss": 0.4033,
"step": 88000
},
{
"epoch": 2.99,
"eval_accuracy": 0.7661742231278655,
"eval_loss": 0.6143723726272583,
"eval_runtime": 50.9677,
"eval_samples_per_second": 192.573,
"eval_steps_per_second": 6.023,
"step": 88000
},
{
"epoch": 3.0,
"step": 88320,
"total_flos": 1.4721628163172653e+17,
"train_loss": 0.5041872973027437,
"train_runtime": 53589.3429,
"train_samples_per_second": 52.738,
"train_steps_per_second": 1.648
}
],
"max_steps": 88320,
"num_train_epochs": 3,
"total_flos": 1.4721628163172653e+17,
"trial_name": null,
"trial_params": null
}