{
"best_metric": 0.8164658634538152,
"best_model_checkpoint": "./results/checkpoint-62000",
"epoch": 2.526075619295958,
"global_step": 62000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 2.5e-05,
"loss": 1.0225,
"step": 250
},
{
"epoch": 0.02,
"learning_rate": 5e-05,
"loss": 0.8271,
"step": 500
},
{
"epoch": 0.02,
"eval_accuracy": 0.661847389558233,
"eval_loss": 0.8168992400169373,
"eval_runtime": 9.3368,
"eval_samples_per_second": 266.686,
"step": 500
},
{
"epoch": 0.03,
"learning_rate": 4.98290761909971e-05,
"loss": 0.7821,
"step": 750
},
{
"epoch": 0.04,
"learning_rate": 4.965815238199421e-05,
"loss": 0.7463,
"step": 1000
},
{
"epoch": 0.04,
"eval_accuracy": 0.708433734939759,
"eval_loss": 0.7267000675201416,
"eval_runtime": 9.0373,
"eval_samples_per_second": 275.525,
"step": 1000
},
{
"epoch": 0.05,
"learning_rate": 4.9487228572991307e-05,
"loss": 0.7266,
"step": 1250
},
{
"epoch": 0.06,
"learning_rate": 4.9316304763988406e-05,
"loss": 0.7156,
"step": 1500
},
{
"epoch": 0.06,
"eval_accuracy": 0.7180722891566265,
"eval_loss": 0.6909603476524353,
"eval_runtime": 9.3259,
"eval_samples_per_second": 266.998,
"step": 1500
},
{
"epoch": 0.07,
"learning_rate": 4.914538095498551e-05,
"loss": 0.7193,
"step": 1750
},
{
"epoch": 0.08,
"learning_rate": 4.897445714598261e-05,
"loss": 0.7073,
"step": 2000
},
{
"epoch": 0.08,
"eval_accuracy": 0.7176706827309237,
"eval_loss": 0.6720558404922485,
"eval_runtime": 9.0667,
"eval_samples_per_second": 274.632,
"step": 2000
},
{
"epoch": 0.09,
"learning_rate": 4.880353333697971e-05,
"loss": 0.6817,
"step": 2250
},
{
"epoch": 0.1,
"learning_rate": 4.8632609527976816e-05,
"loss": 0.6825,
"step": 2500
},
{
"epoch": 0.1,
"eval_accuracy": 0.7373493975903614,
"eval_loss": 0.6640393137931824,
"eval_runtime": 9.298,
"eval_samples_per_second": 267.799,
"step": 2500
},
{
"epoch": 0.11,
"learning_rate": 4.8461685718973915e-05,
"loss": 0.6759,
"step": 2750
},
{
"epoch": 0.12,
"learning_rate": 4.8290761909971014e-05,
"loss": 0.6886,
"step": 3000
},
{
"epoch": 0.12,
"eval_accuracy": 0.7578313253012048,
"eval_loss": 0.6005652546882629,
"eval_runtime": 9.0815,
"eval_samples_per_second": 274.184,
"step": 3000
},
{
"epoch": 0.13,
"learning_rate": 4.811983810096811e-05,
"loss": 0.6971,
"step": 3250
},
{
"epoch": 0.14,
"learning_rate": 4.794891429196521e-05,
"loss": 0.6642,
"step": 3500
},
{
"epoch": 0.14,
"eval_accuracy": 0.7582329317269076,
"eval_loss": 0.5993230938911438,
"eval_runtime": 8.992,
"eval_samples_per_second": 276.912,
"step": 3500
},
{
"epoch": 0.15,
"learning_rate": 4.777799048296232e-05,
"loss": 0.664,
"step": 3750
},
{
"epoch": 0.16,
"learning_rate": 4.760706667395942e-05,
"loss": 0.671,
"step": 4000
},
{
"epoch": 0.16,
"eval_accuracy": 0.7610441767068273,
"eval_loss": 0.5954618453979492,
"eval_runtime": 8.999,
"eval_samples_per_second": 276.698,
"step": 4000
},
{
"epoch": 0.17,
"learning_rate": 4.7436142864956516e-05,
"loss": 0.6421,
"step": 4250
},
{
"epoch": 0.18,
"learning_rate": 4.726521905595362e-05,
"loss": 0.6768,
"step": 4500
},
{
"epoch": 0.18,
"eval_accuracy": 0.7200803212851405,
"eval_loss": 0.6515903472900391,
"eval_runtime": 9.1435,
"eval_samples_per_second": 272.325,
"step": 4500
},
{
"epoch": 0.19,
"learning_rate": 4.709429524695072e-05,
"loss": 0.6606,
"step": 4750
},
{
"epoch": 0.2,
"learning_rate": 4.692337143794782e-05,
"loss": 0.6511,
"step": 5000
},
{
"epoch": 0.2,
"eval_accuracy": 0.7518072289156627,
"eval_loss": 0.6293097734451294,
"eval_runtime": 9.1699,
"eval_samples_per_second": 271.54,
"step": 5000
},
{
"epoch": 0.21,
"learning_rate": 4.6752447628944926e-05,
"loss": 0.651,
"step": 5250
},
{
"epoch": 0.22,
"learning_rate": 4.6581523819942026e-05,
"loss": 0.6729,
"step": 5500
},
{
"epoch": 0.22,
"eval_accuracy": 0.748995983935743,
"eval_loss": 0.6285788416862488,
"eval_runtime": 9.1397,
"eval_samples_per_second": 272.439,
"step": 5500
},
{
"epoch": 0.23,
"learning_rate": 4.6410600010939125e-05,
"loss": 0.6573,
"step": 5750
},
{
"epoch": 0.24,
"learning_rate": 4.623967620193623e-05,
"loss": 0.6335,
"step": 6000
},
{
"epoch": 0.24,
"eval_accuracy": 0.7485943775100402,
"eval_loss": 0.6102361679077148,
"eval_runtime": 9.2187,
"eval_samples_per_second": 270.105,
"step": 6000
},
{
"epoch": 0.25,
"learning_rate": 4.606875239293333e-05,
"loss": 0.6171,
"step": 6250
},
{
"epoch": 0.26,
"learning_rate": 4.589782858393043e-05,
"loss": 0.6445,
"step": 6500
},
{
"epoch": 0.26,
"eval_accuracy": 0.7578313253012048,
"eval_loss": 0.5952242612838745,
"eval_runtime": 9.1091,
"eval_samples_per_second": 273.352,
"step": 6500
},
{
"epoch": 0.28,
"learning_rate": 4.5726904774927535e-05,
"loss": 0.6371,
"step": 6750
},
{
"epoch": 0.29,
"learning_rate": 4.5555980965924634e-05,
"loss": 0.6285,
"step": 7000
},
{
"epoch": 0.29,
"eval_accuracy": 0.7550200803212851,
"eval_loss": 0.5844168663024902,
"eval_runtime": 9.3454,
"eval_samples_per_second": 266.442,
"step": 7000
},
{
"epoch": 0.3,
"learning_rate": 4.538505715692173e-05,
"loss": 0.645,
"step": 7250
},
{
"epoch": 0.31,
"learning_rate": 4.521413334791884e-05,
"loss": 0.6614,
"step": 7500
},
{
"epoch": 0.31,
"eval_accuracy": 0.7706827309236948,
"eval_loss": 0.5955267548561096,
"eval_runtime": 9.1185,
"eval_samples_per_second": 273.072,
"step": 7500
},
{
"epoch": 0.32,
"learning_rate": 4.504320953891594e-05,
"loss": 0.6042,
"step": 7750
},
{
"epoch": 0.33,
"learning_rate": 4.487228572991304e-05,
"loss": 0.6497,
"step": 8000
},
{
"epoch": 0.33,
"eval_accuracy": 0.7803212851405622,
"eval_loss": 0.5732316970825195,
"eval_runtime": 9.5547,
"eval_samples_per_second": 260.605,
"step": 8000
},
{
"epoch": 0.34,
"learning_rate": 4.470136192091014e-05,
"loss": 0.6394,
"step": 8250
},
{
"epoch": 0.35,
"learning_rate": 4.453043811190724e-05,
"loss": 0.6109,
"step": 8500
},
{
"epoch": 0.35,
"eval_accuracy": 0.7594377510040161,
"eval_loss": 0.5940960049629211,
"eval_runtime": 9.1183,
"eval_samples_per_second": 273.077,
"step": 8500
},
{
"epoch": 0.36,
"learning_rate": 4.4359514302904335e-05,
"loss": 0.6324,
"step": 8750
},
{
"epoch": 0.37,
"learning_rate": 4.418859049390144e-05,
"loss": 0.6266,
"step": 9000
},
{
"epoch": 0.37,
"eval_accuracy": 0.7730923694779116,
"eval_loss": 0.5806799530982971,
"eval_runtime": 9.6937,
"eval_samples_per_second": 256.869,
"step": 9000
},
{
"epoch": 0.38,
"learning_rate": 4.401766668489854e-05,
"loss": 0.622,
"step": 9250
},
{
"epoch": 0.39,
"learning_rate": 4.384674287589564e-05,
"loss": 0.629,
"step": 9500
},
{
"epoch": 0.39,
"eval_accuracy": 0.770281124497992,
"eval_loss": 0.5808362364768982,
"eval_runtime": 9.089,
"eval_samples_per_second": 273.959,
"step": 9500
},
{
"epoch": 0.4,
"learning_rate": 4.3675819066892745e-05,
"loss": 0.6107,
"step": 9750
},
{
"epoch": 0.41,
"learning_rate": 4.3504895257889844e-05,
"loss": 0.6347,
"step": 10000
},
{
"epoch": 0.41,
"eval_accuracy": 0.7650602409638554,
"eval_loss": 0.5932815074920654,
"eval_runtime": 9.5702,
"eval_samples_per_second": 260.184,
"step": 10000
},
{
"epoch": 0.42,
"learning_rate": 4.333397144888694e-05,
"loss": 0.6441,
"step": 10250
},
{
"epoch": 0.43,
"learning_rate": 4.316304763988405e-05,
"loss": 0.6181,
"step": 10500
},
{
"epoch": 0.43,
"eval_accuracy": 0.7815261044176707,
"eval_loss": 0.5563480257987976,
"eval_runtime": 9.0503,
"eval_samples_per_second": 275.128,
"step": 10500
},
{
"epoch": 0.44,
"learning_rate": 4.299212383088115e-05,
"loss": 0.6133,
"step": 10750
},
{
"epoch": 0.45,
"learning_rate": 4.282120002187825e-05,
"loss": 0.6173,
"step": 11000
},
{
"epoch": 0.45,
"eval_accuracy": 0.7775100401606426,
"eval_loss": 0.5630077719688416,
"eval_runtime": 9.4887,
"eval_samples_per_second": 262.417,
"step": 11000
},
{
"epoch": 0.46,
"learning_rate": 4.265027621287535e-05,
"loss": 0.5979,
"step": 11250
},
{
"epoch": 0.47,
"learning_rate": 4.247935240387245e-05,
"loss": 0.6114,
"step": 11500
},
{
"epoch": 0.47,
"eval_accuracy": 0.7666666666666667,
"eval_loss": 0.5745715498924255,
"eval_runtime": 9.0458,
"eval_samples_per_second": 275.267,
"step": 11500
},
{
"epoch": 0.48,
"learning_rate": 4.230842859486955e-05,
"loss": 0.6168,
"step": 11750
},
{
"epoch": 0.49,
"learning_rate": 4.213750478586666e-05,
"loss": 0.602,
"step": 12000
},
{
"epoch": 0.49,
"eval_accuracy": 0.7614457831325301,
"eval_loss": 0.5948830246925354,
"eval_runtime": 9.3904,
"eval_samples_per_second": 265.165,
"step": 12000
},
{
"epoch": 0.5,
"learning_rate": 4.1966580976863756e-05,
"loss": 0.6268,
"step": 12250
},
{
"epoch": 0.51,
"learning_rate": 4.1795657167860855e-05,
"loss": 0.6041,
"step": 12500
},
{
"epoch": 0.51,
"eval_accuracy": 0.7755020080321285,
"eval_loss": 0.5688386559486389,
"eval_runtime": 9.0467,
"eval_samples_per_second": 275.238,
"step": 12500
},
{
"epoch": 0.52,
"learning_rate": 4.162473335885796e-05,
"loss": 0.6284,
"step": 12750
},
{
"epoch": 0.53,
"learning_rate": 4.145380954985506e-05,
"loss": 0.6193,
"step": 13000
},
{
"epoch": 0.53,
"eval_accuracy": 0.7590361445783133,
"eval_loss": 0.583430290222168,
"eval_runtime": 9.3111,
"eval_samples_per_second": 267.423,
"step": 13000
},
{
"epoch": 0.54,
"learning_rate": 4.128288574085216e-05,
"loss": 0.5877,
"step": 13250
},
{
"epoch": 0.55,
"learning_rate": 4.1111961931849265e-05,
"loss": 0.5842,
"step": 13500
},
{
"epoch": 0.55,
"eval_accuracy": 0.7614457831325301,
"eval_loss": 0.5992956161499023,
"eval_runtime": 9.0426,
"eval_samples_per_second": 275.365,
"step": 13500
},
{
"epoch": 0.56,
"learning_rate": 4.0941038122846365e-05,
"loss": 0.592,
"step": 13750
},
{
"epoch": 0.57,
"learning_rate": 4.0770114313843464e-05,
"loss": 0.6068,
"step": 14000
},
{
"epoch": 0.57,
"eval_accuracy": 0.7666666666666667,
"eval_loss": 0.582902193069458,
"eval_runtime": 9.2607,
"eval_samples_per_second": 268.878,
"step": 14000
},
{
"epoch": 0.58,
"learning_rate": 4.059919050484056e-05,
"loss": 0.6125,
"step": 14250
},
{
"epoch": 0.59,
"learning_rate": 4.042826669583766e-05,
"loss": 0.6012,
"step": 14500
},
{
"epoch": 0.59,
"eval_accuracy": 0.7799196787148595,
"eval_loss": 0.565805196762085,
"eval_runtime": 9.0222,
"eval_samples_per_second": 275.985,
"step": 14500
},
{
"epoch": 0.6,
"learning_rate": 4.025734288683477e-05,
"loss": 0.5845,
"step": 14750
},
{
"epoch": 0.61,
"learning_rate": 4.008641907783187e-05,
"loss": 0.6159,
"step": 15000
},
{
"epoch": 0.61,
"eval_accuracy": 0.7538152610441767,
"eval_loss": 0.6279062628746033,
"eval_runtime": 9.2223,
"eval_samples_per_second": 269.997,
"step": 15000
},
{
"epoch": 0.62,
"learning_rate": 3.9915495268828966e-05,
"loss": 0.5897,
"step": 15250
},
{
"epoch": 0.63,
"learning_rate": 3.974457145982607e-05,
"loss": 0.5996,
"step": 15500
},
{
"epoch": 0.63,
"eval_accuracy": 0.7726907630522089,
"eval_loss": 0.5615552067756653,
"eval_runtime": 9.0848,
"eval_samples_per_second": 274.085,
"step": 15500
},
{
"epoch": 0.64,
"learning_rate": 3.957364765082317e-05,
"loss": 0.5899,
"step": 15750
},
{
"epoch": 0.65,
"learning_rate": 3.940272384182027e-05,
"loss": 0.6014,
"step": 16000
},
{
"epoch": 0.65,
"eval_accuracy": 0.7883534136546185,
"eval_loss": 0.5467997193336487,
"eval_runtime": 9.0085,
"eval_samples_per_second": 276.404,
"step": 16000
},
{
"epoch": 0.66,
"learning_rate": 3.9231800032817376e-05,
"loss": 0.6057,
"step": 16250
},
{
"epoch": 0.67,
"learning_rate": 3.9060876223814475e-05,
"loss": 0.6118,
"step": 16500
},
{
"epoch": 0.67,
"eval_accuracy": 0.7626506024096386,
"eval_loss": 0.572376012802124,
"eval_runtime": 8.9948,
"eval_samples_per_second": 276.827,
"step": 16500
},
{
"epoch": 0.68,
"learning_rate": 3.8889952414811574e-05,
"loss": 0.6071,
"step": 16750
},
{
"epoch": 0.69,
"learning_rate": 3.871902860580868e-05,
"loss": 0.5986,
"step": 17000
},
{
"epoch": 0.69,
"eval_accuracy": 0.7718875502008032,
"eval_loss": 0.5707191228866577,
"eval_runtime": 9.1769,
"eval_samples_per_second": 271.333,
"step": 17000
},
{
"epoch": 0.7,
"learning_rate": 3.854810479680578e-05,
"loss": 0.5837,
"step": 17250
},
{
"epoch": 0.71,
"learning_rate": 3.837718098780288e-05,
"loss": 0.5946,
"step": 17500
},
{
"epoch": 0.71,
"eval_accuracy": 0.7726907630522089,
"eval_loss": 0.5725845098495483,
"eval_runtime": 9.1347,
"eval_samples_per_second": 272.587,
"step": 17500
},
{
"epoch": 0.72,
"learning_rate": 3.8206257178799984e-05,
"loss": 0.5858,
"step": 17750
},
{
"epoch": 0.73,
"learning_rate": 3.8035333369797084e-05,
"loss": 0.5768,
"step": 18000
},
{
"epoch": 0.73,
"eval_accuracy": 0.763855421686747,
"eval_loss": 0.5683776140213013,
"eval_runtime": 9.114,
"eval_samples_per_second": 273.207,
"step": 18000
},
{
"epoch": 0.74,
"learning_rate": 3.786440956079418e-05,
"loss": 0.6339,
"step": 18250
},
{
"epoch": 0.75,
"learning_rate": 3.769348575179129e-05,
"loss": 0.577,
"step": 18500
},
{
"epoch": 0.75,
"eval_accuracy": 0.7803212851405622,
"eval_loss": 0.5440804958343506,
"eval_runtime": 9.2459,
"eval_samples_per_second": 269.31,
"step": 18500
},
{
"epoch": 0.76,
"learning_rate": 3.752256194278839e-05,
"loss": 0.6009,
"step": 18750
},
{
"epoch": 0.77,
"learning_rate": 3.735163813378549e-05,
"loss": 0.5862,
"step": 19000
},
{
"epoch": 0.77,
"eval_accuracy": 0.7863453815261044,
"eval_loss": 0.5475257635116577,
"eval_runtime": 9.1052,
"eval_samples_per_second": 273.471,
"step": 19000
},
{
"epoch": 0.78,
"learning_rate": 3.7180714324782586e-05,
"loss": 0.591,
"step": 19250
},
{
"epoch": 0.79,
"learning_rate": 3.7009790515779685e-05,
"loss": 0.6009,
"step": 19500
},
{
"epoch": 0.79,
"eval_accuracy": 0.7730923694779116,
"eval_loss": 0.5383695960044861,
"eval_runtime": 9.4316,
"eval_samples_per_second": 264.007,
"step": 19500
},
{
"epoch": 0.8,
"learning_rate": 3.6838866706776784e-05,
"loss": 0.5783,
"step": 19750
},
{
"epoch": 0.81,
"learning_rate": 3.666794289777389e-05,
"loss": 0.5803,
"step": 20000
},
{
"epoch": 0.81,
"eval_accuracy": 0.7799196787148595,
"eval_loss": 0.5557690858840942,
"eval_runtime": 9.0883,
"eval_samples_per_second": 273.978,
"step": 20000
},
{
"epoch": 0.83,
"learning_rate": 3.649701908877099e-05,
"loss": 0.5826,
"step": 20250
},
{
"epoch": 0.84,
"learning_rate": 3.632609527976809e-05,
"loss": 0.5811,
"step": 20500
},
{
"epoch": 0.84,
"eval_accuracy": 0.7650602409638554,
"eval_loss": 0.5788130760192871,
"eval_runtime": 9.68,
"eval_samples_per_second": 257.231,
"step": 20500
},
{
"epoch": 0.85,
"learning_rate": 3.6155171470765194e-05,
"loss": 0.5937,
"step": 20750
},
{
"epoch": 0.86,
"learning_rate": 3.598424766176229e-05,
"loss": 0.597,
"step": 21000
},
{
"epoch": 0.86,
"eval_accuracy": 0.791566265060241,
"eval_loss": 0.528762698173523,
"eval_runtime": 9.0827,
"eval_samples_per_second": 274.147,
"step": 21000
},
{
"epoch": 0.87,
"learning_rate": 3.581332385275939e-05,
"loss": 0.5732,
"step": 21250
},
{
"epoch": 0.88,
"learning_rate": 3.56424000437565e-05,
"loss": 0.6004,
"step": 21500
},
{
"epoch": 0.88,
"eval_accuracy": 0.7819277108433735,
"eval_loss": 0.5339077711105347,
"eval_runtime": 9.6053,
"eval_samples_per_second": 259.233,
"step": 21500
},
{
"epoch": 0.89,
"learning_rate": 3.54714762347536e-05,
"loss": 0.5548,
"step": 21750
},
{
"epoch": 0.9,
"learning_rate": 3.53005524257507e-05,
"loss": 0.5718,
"step": 22000
},
{
"epoch": 0.9,
"eval_accuracy": 0.7759036144578313,
"eval_loss": 0.5448992848396301,
"eval_runtime": 8.9793,
"eval_samples_per_second": 277.305,
"step": 22000
},
{
"epoch": 0.91,
"learning_rate": 3.51296286167478e-05,
"loss": 0.5747,
"step": 22250
},
{
"epoch": 0.92,
"learning_rate": 3.49587048077449e-05,
"loss": 0.5922,
"step": 22500
},
{
"epoch": 0.92,
"eval_accuracy": 0.7827309236947791,
"eval_loss": 0.534243643283844,
"eval_runtime": 9.0048,
"eval_samples_per_second": 276.52,
"step": 22500
},
{
"epoch": 0.93,
"learning_rate": 3.4787780998742e-05,
"loss": 0.5665,
"step": 22750
},
{
"epoch": 0.94,
"learning_rate": 3.461685718973911e-05,
"loss": 0.5834,
"step": 23000
},
{
"epoch": 0.94,
"eval_accuracy": 0.8012048192771084,
"eval_loss": 0.517599880695343,
"eval_runtime": 9.1146,
"eval_samples_per_second": 273.189,
"step": 23000
},
{
"epoch": 0.95,
"learning_rate": 3.4445933380736206e-05,
"loss": 0.5759,
"step": 23250
},
{
"epoch": 0.96,
"learning_rate": 3.4275009571733305e-05,
"loss": 0.5791,
"step": 23500
},
{
"epoch": 0.96,
"eval_accuracy": 0.7879518072289157,
"eval_loss": 0.5282337665557861,
"eval_runtime": 9.3392,
"eval_samples_per_second": 266.619,
"step": 23500
},
{
"epoch": 0.97,
"learning_rate": 3.410408576273041e-05,
"loss": 0.5892,
"step": 23750
},
{
"epoch": 0.98,
"learning_rate": 3.393316195372751e-05,
"loss": 0.5598,
"step": 24000
},
{
"epoch": 0.98,
"eval_accuracy": 0.7855421686746988,
"eval_loss": 0.5417413711547852,
"eval_runtime": 9.0517,
"eval_samples_per_second": 275.086,
"step": 24000
},
{
"epoch": 0.99,
"learning_rate": 3.376223814472461e-05,
"loss": 0.5695,
"step": 24250
},
{
"epoch": 1.0,
"learning_rate": 3.3591314335721715e-05,
"loss": 0.5663,
"step": 24500
},
{
"epoch": 1.0,
"eval_accuracy": 0.7811244979919679,
"eval_loss": 0.5498207807540894,
"eval_runtime": 9.2845,
"eval_samples_per_second": 268.188,
"step": 24500
},
{
"epoch": 1.01,
"learning_rate": 3.342039052671881e-05,
"loss": 0.4729,
"step": 24750
},
{
"epoch": 1.02,
"learning_rate": 3.324946671771591e-05,
"loss": 0.4511,
"step": 25000
},
{
"epoch": 1.02,
"eval_accuracy": 0.7779116465863454,
"eval_loss": 0.5648518204689026,
"eval_runtime": 9.0553,
"eval_samples_per_second": 274.977,
"step": 25000
},
{
"epoch": 1.03,
"learning_rate": 3.307854290871301e-05,
"loss": 0.4426,
"step": 25250
},
{
"epoch": 1.04,
"learning_rate": 3.290761909971011e-05,
"loss": 0.4361,
"step": 25500
},
{
"epoch": 1.04,
"eval_accuracy": 0.7959839357429719,
"eval_loss": 0.5434826016426086,
"eval_runtime": 9.2435,
"eval_samples_per_second": 269.378,
"step": 25500
},
{
"epoch": 1.05,
"learning_rate": 3.273669529070722e-05,
"loss": 0.4643,
"step": 25750
},
{
"epoch": 1.06,
"learning_rate": 3.2565771481704317e-05,
"loss": 0.4686,
"step": 26000
},
{
"epoch": 1.06,
"eval_accuracy": 0.7947791164658634,
"eval_loss": 0.5344994068145752,
"eval_runtime": 9.0719,
"eval_samples_per_second": 274.473,
"step": 26000
},
{
"epoch": 1.07,
"learning_rate": 3.2394847672701416e-05,
"loss": 0.4505,
"step": 26250
},
{
"epoch": 1.08,
"learning_rate": 3.222392386369852e-05,
"loss": 0.4483,
"step": 26500
},
{
"epoch": 1.08,
"eval_accuracy": 0.7983935742971887,
"eval_loss": 0.5487410426139832,
"eval_runtime": 9.2215,
"eval_samples_per_second": 270.021,
"step": 26500
},
{
"epoch": 1.09,
"learning_rate": 3.205300005469562e-05,
"loss": 0.4497,
"step": 26750
},
{
"epoch": 1.1,
"learning_rate": 3.188207624569272e-05,
"loss": 0.4608,
"step": 27000
},
{
"epoch": 1.1,
"eval_accuracy": 0.7879518072289157,
"eval_loss": 0.5529894232749939,
"eval_runtime": 9.098,
"eval_samples_per_second": 273.687,
"step": 27000
},
{
"epoch": 1.11,
"learning_rate": 3.1711152436689826e-05,
"loss": 0.4549,
"step": 27250
},
{
"epoch": 1.12,
"learning_rate": 3.1540228627686925e-05,
"loss": 0.4636,
"step": 27500
},
{
"epoch": 1.12,
"eval_accuracy": 0.7859437751004016,
"eval_loss": 0.5583757162094116,
"eval_runtime": 9.1565,
"eval_samples_per_second": 271.938,
"step": 27500
},
{
"epoch": 1.13,
"learning_rate": 3.1369304818684024e-05,
"loss": 0.4537,
"step": 27750
},
{
"epoch": 1.14,
"learning_rate": 3.119838100968113e-05,
"loss": 0.4628,
"step": 28000
},
{
"epoch": 1.14,
"eval_accuracy": 0.795582329317269,
"eval_loss": 0.5474898815155029,
"eval_runtime": 9.1471,
"eval_samples_per_second": 272.217,
"step": 28000
},
{
"epoch": 1.15,
"learning_rate": 3.102745720067823e-05,
"loss": 0.4648,
"step": 28250
},
{
"epoch": 1.16,
"learning_rate": 3.085653339167533e-05,
"loss": 0.476,
"step": 28500
},
{
"epoch": 1.16,
"eval_accuracy": 0.7911646586345381,
"eval_loss": 0.5516586899757385,
"eval_runtime": 8.9921,
"eval_samples_per_second": 276.91,
"step": 28500
},
{
"epoch": 1.17,
"learning_rate": 3.0685609582672434e-05,
"loss": 0.4794,
"step": 28750
},
{
"epoch": 1.18,
"learning_rate": 3.0514685773669533e-05,
"loss": 0.4537,
"step": 29000
},
{
"epoch": 1.18,
"eval_accuracy": 0.804417670682731,
"eval_loss": 0.5304700136184692,
"eval_runtime": 9.0138,
"eval_samples_per_second": 276.242,
"step": 29000
},
{
"epoch": 1.19,
"learning_rate": 3.0343761964666632e-05,
"loss": 0.481,
"step": 29250
},
{
"epoch": 1.2,
"learning_rate": 3.0172838155663735e-05,
"loss": 0.4715,
"step": 29500
},
{
"epoch": 1.2,
"eval_accuracy": 0.7939759036144578,
"eval_loss": 0.5461502075195312,
"eval_runtime": 9.0986,
"eval_samples_per_second": 273.667,
"step": 29500
},
{
"epoch": 1.21,
"learning_rate": 3.0001914346660837e-05,
"loss": 0.4385,
"step": 29750
},
{
"epoch": 1.22,
"learning_rate": 2.9830990537657933e-05,
"loss": 0.4633,
"step": 30000
},
{
"epoch": 1.22,
"eval_accuracy": 0.793574297188755,
"eval_loss": 0.5552789568901062,
"eval_runtime": 9.3994,
"eval_samples_per_second": 264.911,
"step": 30000
},
{
"epoch": 1.23,
"learning_rate": 2.9660066728655032e-05,
"loss": 0.4748,
"step": 30250
},
{
"epoch": 1.24,
"learning_rate": 2.9489142919652135e-05,
"loss": 0.466,
"step": 30500
},
{
"epoch": 1.24,
"eval_accuracy": 0.7907630522088354,
"eval_loss": 0.5751305222511292,
"eval_runtime": 9.0876,
"eval_samples_per_second": 274.0,
"step": 30500
},
{
"epoch": 1.25,
"learning_rate": 2.9318219110649237e-05,
"loss": 0.4705,
"step": 30750
},
{
"epoch": 1.26,
"learning_rate": 2.9147295301646336e-05,
"loss": 0.4815,
"step": 31000
},
{
"epoch": 1.26,
"eval_accuracy": 0.8008032128514057,
"eval_loss": 0.5165457725524902,
"eval_runtime": 9.6998,
"eval_samples_per_second": 256.707,
"step": 31000
},
{
"epoch": 1.27,
"learning_rate": 2.897637149264344e-05,
"loss": 0.4638,
"step": 31250
},
{
"epoch": 1.28,
"learning_rate": 2.880544768364054e-05,
"loss": 0.4544,
"step": 31500
},
{
"epoch": 1.28,
"eval_accuracy": 0.7847389558232932,
"eval_loss": 0.5818995833396912,
"eval_runtime": 9.0916,
"eval_samples_per_second": 273.879,
"step": 31500
},
{
"epoch": 1.29,
"learning_rate": 2.863452387463764e-05,
"loss": 0.4639,
"step": 31750
},
{
"epoch": 1.3,
"learning_rate": 2.8463600065634743e-05,
"loss": 0.4626,
"step": 32000
},
{
"epoch": 1.3,
"eval_accuracy": 0.8012048192771084,
"eval_loss": 0.5314044952392578,
"eval_runtime": 9.5865,
"eval_samples_per_second": 259.741,
"step": 32000
},
{
"epoch": 1.31,
"learning_rate": 2.8292676256631846e-05,
"loss": 0.4502,
"step": 32250
},
{
"epoch": 1.32,
"learning_rate": 2.8121752447628945e-05,
"loss": 0.4742,
"step": 32500
},
{
"epoch": 1.32,
"eval_accuracy": 0.8036144578313253,
"eval_loss": 0.5086196064949036,
"eval_runtime": 9.0534,
"eval_samples_per_second": 275.036,
"step": 32500
},
{
"epoch": 1.33,
"learning_rate": 2.7950828638626047e-05,
"loss": 0.4555,
"step": 32750
},
{
"epoch": 1.34,
"learning_rate": 2.777990482962315e-05,
"loss": 0.4643,
"step": 33000
},
{
"epoch": 1.34,
"eval_accuracy": 0.8012048192771084,
"eval_loss": 0.5454714894294739,
"eval_runtime": 9.453,
"eval_samples_per_second": 263.409,
"step": 33000
},
{
"epoch": 1.35,
"learning_rate": 2.760898102062025e-05,
"loss": 0.4597,
"step": 33250
},
{
"epoch": 1.36,
"learning_rate": 2.743805721161735e-05,
"loss": 0.4727,
"step": 33500
},
{
"epoch": 1.36,
"eval_accuracy": 0.7943775100401607,
"eval_loss": 0.5575410723686218,
"eval_runtime": 8.9953,
"eval_samples_per_second": 276.812,
"step": 33500
},
{
"epoch": 1.38,
"learning_rate": 2.7267133402614454e-05,
"loss": 0.4612,
"step": 33750
},
{
"epoch": 1.39,
"learning_rate": 2.7096209593611553e-05,
"loss": 0.4646,
"step": 34000
},
{
"epoch": 1.39,
"eval_accuracy": 0.7987951807228916,
"eval_loss": 0.5220562219619751,
"eval_runtime": 9.4179,
"eval_samples_per_second": 264.391,
"step": 34000
},
{
"epoch": 1.4,
"learning_rate": 2.6925285784608655e-05,
"loss": 0.46,
"step": 34250
},
{
"epoch": 1.41,
"learning_rate": 2.6754361975605758e-05,
"loss": 0.4573,
"step": 34500
},
{
"epoch": 1.41,
"eval_accuracy": 0.7939759036144578,
"eval_loss": 0.5481248497962952,
"eval_runtime": 9.0056,
"eval_samples_per_second": 276.493,
"step": 34500
},
{
"epoch": 1.42,
"learning_rate": 2.6583438166602857e-05,
"loss": 0.4633,
"step": 34750
},
{
"epoch": 1.43,
"learning_rate": 2.641251435759996e-05,
"loss": 0.4532,
"step": 35000
},
{
"epoch": 1.43,
"eval_accuracy": 0.793574297188755,
"eval_loss": 0.5569635629653931,
"eval_runtime": 9.3588,
"eval_samples_per_second": 266.061,
"step": 35000
},
{
"epoch": 1.44,
"learning_rate": 2.6241590548597055e-05,
"loss": 0.4431,
"step": 35250
},
{
"epoch": 1.45,
"learning_rate": 2.6070666739594158e-05,
"loss": 0.4637,
"step": 35500
},
{
"epoch": 1.45,
"eval_accuracy": 0.7971887550200804,
"eval_loss": 0.5141083002090454,
"eval_runtime": 9.013,
"eval_samples_per_second": 276.266,
"step": 35500
},
{
"epoch": 1.46,
"learning_rate": 2.5899742930591257e-05,
"loss": 0.4337,
"step": 35750
},
{
"epoch": 1.47,
"learning_rate": 2.572881912158836e-05,
"loss": 0.4572,
"step": 36000
},
{
"epoch": 1.47,
"eval_accuracy": 0.7887550200803213,
"eval_loss": 0.5640882253646851,
"eval_runtime": 9.316,
"eval_samples_per_second": 267.283,
"step": 36000
},
{
"epoch": 1.48,
"learning_rate": 2.5557895312585462e-05,
"loss": 0.4485,
"step": 36250
},
{
"epoch": 1.49,
"learning_rate": 2.538697150358256e-05,
"loss": 0.4613,
"step": 36500
},
{
"epoch": 1.49,
"eval_accuracy": 0.8004016064257028,
"eval_loss": 0.5442628860473633,
"eval_runtime": 9.0407,
"eval_samples_per_second": 275.422,
"step": 36500
},
{
"epoch": 1.5,
"learning_rate": 2.5216047694579664e-05,
"loss": 0.4547,
"step": 36750
},
{
"epoch": 1.51,
"learning_rate": 2.5045123885576766e-05,
"loss": 0.4614,
"step": 37000
},
{
"epoch": 1.51,
"eval_accuracy": 0.7987951807228916,
"eval_loss": 0.5075089931488037,
"eval_runtime": 9.22,
"eval_samples_per_second": 270.065,
"step": 37000
},
{
"epoch": 1.52,
"learning_rate": 2.4874200076573865e-05,
"loss": 0.4545,
"step": 37250
},
{
"epoch": 1.53,
"learning_rate": 2.4703276267570968e-05,
"loss": 0.4724,
"step": 37500
},
{
"epoch": 1.53,
"eval_accuracy": 0.8092369477911646,
"eval_loss": 0.5035138726234436,
"eval_runtime": 9.0797,
"eval_samples_per_second": 274.237,
"step": 37500
},
{
"epoch": 1.54,
"learning_rate": 2.453235245856807e-05,
"loss": 0.4451,
"step": 37750
},
{
"epoch": 1.55,
"learning_rate": 2.436142864956517e-05,
"loss": 0.4396,
"step": 38000
},
{
"epoch": 1.55,
"eval_accuracy": 0.7971887550200804,
"eval_loss": 0.5381002426147461,
"eval_runtime": 9.1884,
"eval_samples_per_second": 270.995,
"step": 38000
},
{
"epoch": 1.56,
"learning_rate": 2.4190504840562272e-05,
"loss": 0.459,
"step": 38250
},
{
"epoch": 1.57,
"learning_rate": 2.4019581031559374e-05,
"loss": 0.4561,
"step": 38500
},
{
"epoch": 1.57,
"eval_accuracy": 0.8100401606425702,
"eval_loss": 0.4913768470287323,
"eval_runtime": 9.0787,
"eval_samples_per_second": 274.268,
"step": 38500
},
{
"epoch": 1.58,
"learning_rate": 2.3848657222556474e-05,
"loss": 0.4503,
"step": 38750
},
{
"epoch": 1.59,
"learning_rate": 2.3677733413553576e-05,
"loss": 0.4643,
"step": 39000
},
{
"epoch": 1.59,
"eval_accuracy": 0.8092369477911646,
"eval_loss": 0.5171190500259399,
"eval_runtime": 8.9997,
"eval_samples_per_second": 276.677,
"step": 39000
},
{
"epoch": 1.6,
"learning_rate": 2.3506809604550675e-05,
"loss": 0.4624,
"step": 39250
},
{
"epoch": 1.61,
"learning_rate": 2.3335885795547778e-05,
"loss": 0.443,
"step": 39500
},
{
"epoch": 1.61,
"eval_accuracy": 0.7951807228915663,
"eval_loss": 0.5365191102027893,
"eval_runtime": 9.0082,
"eval_samples_per_second": 276.415,
"step": 39500
},
{
"epoch": 1.62,
"learning_rate": 2.3164961986544877e-05,
"loss": 0.4467,
"step": 39750
},
{
"epoch": 1.63,
"learning_rate": 2.299403817754198e-05,
"loss": 0.4525,
"step": 40000
},
{
"epoch": 1.63,
"eval_accuracy": 0.8036144578313253,
"eval_loss": 0.5161953568458557,
"eval_runtime": 9.0761,
"eval_samples_per_second": 274.346,
"step": 40000
},
{
"epoch": 1.64,
"learning_rate": 2.2823114368539082e-05,
"loss": 0.4591,
"step": 40250
},
{
"epoch": 1.65,
"learning_rate": 2.265219055953618e-05,
"loss": 0.4618,
"step": 40500
},
{
"epoch": 1.65,
"eval_accuracy": 0.8080321285140563,
"eval_loss": 0.49769964814186096,
"eval_runtime": 9.2939,
"eval_samples_per_second": 267.916,
"step": 40500
},
{
"epoch": 1.66,
"learning_rate": 2.2481266750533284e-05,
"loss": 0.4434,
"step": 40750
},
{
"epoch": 1.67,
"learning_rate": 2.2310342941530386e-05,
"loss": 0.4561,
"step": 41000
},
{
"epoch": 1.67,
"eval_accuracy": 0.8092369477911646,
"eval_loss": 0.4977361559867859,
"eval_runtime": 8.9616,
"eval_samples_per_second": 277.852,
"step": 41000
},
{
"epoch": 1.68,
"learning_rate": 2.2139419132527485e-05,
"loss": 0.4483,
"step": 41250
},
{
"epoch": 1.69,
"learning_rate": 2.1968495323524588e-05,
"loss": 0.4512,
"step": 41500
},
{
"epoch": 1.69,
"eval_accuracy": 0.802008032128514,
"eval_loss": 0.4957820475101471,
"eval_runtime": 9.0164,
"eval_samples_per_second": 276.164,
"step": 41500
},
{
"epoch": 1.7,
"learning_rate": 2.179757151452169e-05,
"loss": 0.4473,
"step": 41750
},
{
"epoch": 1.71,
"learning_rate": 2.162664770551879e-05,
"loss": 0.4564,
"step": 42000
},
{
"epoch": 1.71,
"eval_accuracy": 0.7987951807228916,
"eval_loss": 0.5506803393363953,
"eval_runtime": 9.1173,
"eval_samples_per_second": 273.106,
"step": 42000
},
{
"epoch": 1.72,
"learning_rate": 2.145572389651589e-05,
"loss": 0.4491,
"step": 42250
},
{
"epoch": 1.73,
"learning_rate": 2.128480008751299e-05,
"loss": 0.4331,
"step": 42500
},
{
"epoch": 1.73,
"eval_accuracy": 0.8140562248995984,
"eval_loss": 0.5114361047744751,
"eval_runtime": 9.554,
"eval_samples_per_second": 260.625,
"step": 42500
},
{
"epoch": 1.74,
"learning_rate": 2.1113876278510094e-05,
"loss": 0.4631,
"step": 42750
},
{
"epoch": 1.75,
"learning_rate": 2.0942952469507193e-05,
"loss": 0.4598,
"step": 43000
},
{
"epoch": 1.75,
"eval_accuracy": 0.8032128514056225,
"eval_loss": 0.5167751312255859,
"eval_runtime": 9.065,
"eval_samples_per_second": 274.682,
"step": 43000
},
{
"epoch": 1.76,
"learning_rate": 2.0772028660504295e-05,
"loss": 0.433,
"step": 43250
},
{
"epoch": 1.77,
"learning_rate": 2.0601104851501398e-05,
"loss": 0.4711,
"step": 43500
},
{
"epoch": 1.77,
"eval_accuracy": 0.8100401606425702,
"eval_loss": 0.49471431970596313,
"eval_runtime": 9.4303,
"eval_samples_per_second": 264.043,
"step": 43500
},
{
"epoch": 1.78,
"learning_rate": 2.0430181042498497e-05,
"loss": 0.4638,
"step": 43750
},
{
"epoch": 1.79,
"learning_rate": 2.02592572334956e-05,
"loss": 0.4133,
"step": 44000
},
{
"epoch": 1.79,
"eval_accuracy": 0.8144578313253013,
"eval_loss": 0.5469810962677002,
"eval_runtime": 9.0719,
"eval_samples_per_second": 274.474,
"step": 44000
},
{
"epoch": 1.8,
"learning_rate": 2.0088333424492702e-05,
"loss": 0.4558,
"step": 44250
},
{
"epoch": 1.81,
"learning_rate": 1.99174096154898e-05,
"loss": 0.4442,
"step": 44500
},
{
"epoch": 1.81,
"eval_accuracy": 0.8024096385542169,
"eval_loss": 0.5169267654418945,
"eval_runtime": 9.3219,
"eval_samples_per_second": 267.114,
"step": 44500
},
{
"epoch": 1.82,
"learning_rate": 1.97464858064869e-05,
"loss": 0.4412,
"step": 44750
},
{
"epoch": 1.83,
"learning_rate": 1.9575561997484003e-05,
"loss": 0.4425,
"step": 45000
},
{
"epoch": 1.83,
"eval_accuracy": 0.8076305220883534,
"eval_loss": 0.5110898613929749,
"eval_runtime": 9.0459,
"eval_samples_per_second": 275.262,
"step": 45000
},
{
"epoch": 1.84,
"learning_rate": 1.9404638188481102e-05,
"loss": 0.4443,
"step": 45250
},
{
"epoch": 1.85,
"learning_rate": 1.9233714379478204e-05,
"loss": 0.4522,
"step": 45500
},
{
"epoch": 1.85,
"eval_accuracy": 0.8032128514056225,
"eval_loss": 0.527312159538269,
"eval_runtime": 8.9896,
"eval_samples_per_second": 276.986,
"step": 45500
},
{
"epoch": 1.86,
"learning_rate": 1.9062790570475307e-05,
"loss": 0.45,
"step": 45750
},
{
"epoch": 1.87,
"learning_rate": 1.8891866761472406e-05,
"loss": 0.4451,
"step": 46000
},
{
"epoch": 1.87,
"eval_accuracy": 0.8132530120481928,
"eval_loss": 0.4912641644477844,
"eval_runtime": 9.0608,
"eval_samples_per_second": 274.811,
"step": 46000
},
{
"epoch": 1.88,
"learning_rate": 1.872094295246951e-05,
"loss": 0.4321,
"step": 46250
},
{
"epoch": 1.89,
"learning_rate": 1.855001914346661e-05,
"loss": 0.463,
"step": 46500
},
{
"epoch": 1.89,
"eval_accuracy": 0.8040160642570281,
"eval_loss": 0.5067523717880249,
"eval_runtime": 8.9793,
"eval_samples_per_second": 277.305,
"step": 46500
},
{
"epoch": 1.9,
"learning_rate": 1.837909533446371e-05,
"loss": 0.44,
"step": 46750
},
{
"epoch": 1.91,
"learning_rate": 1.8208171525460813e-05,
"loss": 0.4173,
"step": 47000
},
{
"epoch": 1.91,
"eval_accuracy": 0.8008032128514057,
"eval_loss": 0.5309551954269409,
"eval_runtime": 9.0236,
"eval_samples_per_second": 275.944,
"step": 47000
},
{
"epoch": 1.93,
"learning_rate": 1.803724771645791e-05,
"loss": 0.4377,
"step": 47250
},
{
"epoch": 1.94,
"learning_rate": 1.7866323907455014e-05,
"loss": 0.4336,
"step": 47500
},
{
"epoch": 1.94,
"eval_accuracy": 0.8036144578313253,
"eval_loss": 0.5289146900177002,
"eval_runtime": 9.1489,
"eval_samples_per_second": 272.164,
"step": 47500
},
{
"epoch": 1.95,
"learning_rate": 1.7695400098452113e-05,
"loss": 0.4424,
"step": 47750
},
{
"epoch": 1.96,
"learning_rate": 1.7524476289449216e-05,
"loss": 0.4266,
"step": 48000
},
{
"epoch": 1.96,
"eval_accuracy": 0.8048192771084337,
"eval_loss": 0.516535758972168,
"eval_runtime": 9.1038,
"eval_samples_per_second": 273.513,
"step": 48000
},
{
"epoch": 1.97,
"learning_rate": 1.735355248044632e-05,
"loss": 0.4276,
"step": 48250
},
{
"epoch": 1.98,
"learning_rate": 1.7182628671443417e-05,
"loss": 0.4336,
"step": 48500
},
{
"epoch": 1.98,
"eval_accuracy": 0.8104417670682731,
"eval_loss": 0.5314404368400574,
"eval_runtime": 9.1174,
"eval_samples_per_second": 273.103,
"step": 48500
},
{
"epoch": 1.99,
"learning_rate": 1.701170486244052e-05,
"loss": 0.4286,
"step": 48750
},
{
"epoch": 2.0,
"learning_rate": 1.6840781053437622e-05,
"loss": 0.4342,
"step": 49000
},
{
"epoch": 2.0,
"eval_accuracy": 0.814859437751004,
"eval_loss": 0.4977148473262787,
"eval_runtime": 9.12,
"eval_samples_per_second": 273.026,
"step": 49000
},
{
"epoch": 2.01,
"learning_rate": 1.666985724443472e-05,
"loss": 0.3449,
"step": 49250
},
{
"epoch": 2.02,
"learning_rate": 1.6498933435431824e-05,
"loss": 0.3156,
"step": 49500
},
{
"epoch": 2.02,
"eval_accuracy": 0.8128514056224899,
"eval_loss": 0.5999274849891663,
"eval_runtime": 9.1186,
"eval_samples_per_second": 273.069,
"step": 49500
},
{
"epoch": 2.03,
"learning_rate": 1.6328009626428927e-05,
"loss": 0.2989,
"step": 49750
},
{
"epoch": 2.04,
"learning_rate": 1.6157085817426022e-05,
"loss": 0.3013,
"step": 50000
},
{
"epoch": 2.04,
"eval_accuracy": 0.8028112449799196,
"eval_loss": 0.6256367564201355,
"eval_runtime": 9.2774,
"eval_samples_per_second": 268.395,
"step": 50000
},
{
"epoch": 2.05,
"learning_rate": 1.5986162008423125e-05,
"loss": 0.2947,
"step": 50250
},
{
"epoch": 2.06,
"learning_rate": 1.5815238199420227e-05,
"loss": 0.2849,
"step": 50500
},
{
"epoch": 2.06,
"eval_accuracy": 0.8116465863453816,
"eval_loss": 0.6139395236968994,
"eval_runtime": 9.009,
"eval_samples_per_second": 276.389,
"step": 50500
},
{
"epoch": 2.07,
"learning_rate": 1.5644314390417327e-05,
"loss": 0.2929,
"step": 50750
},
{
"epoch": 2.08,
"learning_rate": 1.547339058141443e-05,
"loss": 0.2927,
"step": 51000
},
{
"epoch": 2.08,
"eval_accuracy": 0.808433734939759,
"eval_loss": 0.5880476236343384,
"eval_runtime": 8.9456,
"eval_samples_per_second": 278.35,
"step": 51000
},
{
"epoch": 2.09,
"learning_rate": 1.530246677241153e-05,
"loss": 0.2725,
"step": 51250
},
{
"epoch": 2.1,
"learning_rate": 1.5131542963408632e-05,
"loss": 0.2944,
"step": 51500
},
{
"epoch": 2.1,
"eval_accuracy": 0.8064257028112449,
"eval_loss": 0.6175798177719116,
"eval_runtime": 9.0878,
"eval_samples_per_second": 273.995,
"step": 51500
},
{
"epoch": 2.11,
"learning_rate": 1.4960619154405733e-05,
"loss": 0.2812,
"step": 51750
},
{
"epoch": 2.12,
"learning_rate": 1.4789695345402834e-05,
"loss": 0.2869,
"step": 52000
},
{
"epoch": 2.12,
"eval_accuracy": 0.8100401606425702,
"eval_loss": 0.5959600806236267,
"eval_runtime": 9.5023,
"eval_samples_per_second": 262.042,
"step": 52000
},
{
"epoch": 2.13,
"learning_rate": 1.4618771536399937e-05,
"loss": 0.2726,
"step": 52250
},
{
"epoch": 2.14,
"learning_rate": 1.4447847727397037e-05,
"loss": 0.3034,
"step": 52500
},
{
"epoch": 2.14,
"eval_accuracy": 0.7907630522088354,
"eval_loss": 0.6271839141845703,
"eval_runtime": 9.0834,
"eval_samples_per_second": 274.126,
"step": 52500
},
{
"epoch": 2.15,
"learning_rate": 1.4276923918394136e-05,
"loss": 0.3019,
"step": 52750
},
{
"epoch": 2.16,
"learning_rate": 1.4106000109391237e-05,
"loss": 0.279,
"step": 53000
},
{
"epoch": 2.16,
"eval_accuracy": 0.8096385542168675,
"eval_loss": 0.6030941009521484,
"eval_runtime": 9.4494,
"eval_samples_per_second": 263.51,
"step": 53000
},
{
"epoch": 2.17,
"learning_rate": 1.3935076300388338e-05,
"loss": 0.3079,
"step": 53250
},
{
"epoch": 2.18,
"learning_rate": 1.376415249138544e-05,
"loss": 0.2896,
"step": 53500
},
{
"epoch": 2.18,
"eval_accuracy": 0.8068273092369478,
"eval_loss": 0.6132158637046814,
"eval_runtime": 9.0141,
"eval_samples_per_second": 276.233,
"step": 53500
},
{
"epoch": 2.19,
"learning_rate": 1.3593228682382541e-05,
"loss": 0.2823,
"step": 53750
},
{
"epoch": 2.2,
"learning_rate": 1.3422304873379642e-05,
"loss": 0.2952,
"step": 54000
},
{
"epoch": 2.2,
"eval_accuracy": 0.8064257028112449,
"eval_loss": 0.6195886135101318,
"eval_runtime": 8.9618,
"eval_samples_per_second": 277.847,
"step": 54000
},
{
"epoch": 2.21,
"learning_rate": 1.3251381064376745e-05,
"loss": 0.3049,
"step": 54250
},
{
"epoch": 2.22,
"learning_rate": 1.3080457255373846e-05,
"loss": 0.2921,
"step": 54500
},
{
"epoch": 2.22,
"eval_accuracy": 0.8076305220883534,
"eval_loss": 0.6113378405570984,
"eval_runtime": 9.0032,
"eval_samples_per_second": 276.569,
"step": 54500
},
{
"epoch": 2.23,
"learning_rate": 1.2909533446370946e-05,
"loss": 0.2765,
"step": 54750
},
{
"epoch": 2.24,
"learning_rate": 1.2738609637368049e-05,
"loss": 0.2958,
"step": 55000
},
{
"epoch": 2.24,
"eval_accuracy": 0.8064257028112449,
"eval_loss": 0.6207754611968994,
"eval_runtime": 9.0338,
"eval_samples_per_second": 275.632,
"step": 55000
},
{
"epoch": 2.25,
"learning_rate": 1.2567685828365146e-05,
"loss": 0.2888,
"step": 55250
},
{
"epoch": 2.26,
"learning_rate": 1.239676201936225e-05,
"loss": 0.2996,
"step": 55500
},
{
"epoch": 2.26,
"eval_accuracy": 0.8128514056224899,
"eval_loss": 0.5894312262535095,
"eval_runtime": 8.9904,
"eval_samples_per_second": 276.961,
"step": 55500
},
{
"epoch": 2.27,
"learning_rate": 1.2225838210359351e-05,
"loss": 0.2994,
"step": 55750
},
{
"epoch": 2.28,
"learning_rate": 1.205491440135645e-05,
"loss": 0.288,
"step": 56000
},
{
"epoch": 2.28,
"eval_accuracy": 0.8052208835341366,
"eval_loss": 0.6171417832374573,
"eval_runtime": 9.2622,
"eval_samples_per_second": 268.834,
"step": 56000
},
{
"epoch": 2.29,
"learning_rate": 1.1883990592353553e-05,
"loss": 0.2902,
"step": 56250
},
{
"epoch": 2.3,
"learning_rate": 1.1713066783350654e-05,
"loss": 0.3005,
"step": 56500
},
{
"epoch": 2.3,
"eval_accuracy": 0.8112449799196787,
"eval_loss": 0.5888203978538513,
"eval_runtime": 8.9926,
"eval_samples_per_second": 276.894,
"step": 56500
},
{
"epoch": 2.31,
"learning_rate": 1.1542142974347755e-05,
"loss": 0.3011,
"step": 56750
},
{
"epoch": 2.32,
"learning_rate": 1.1371219165344857e-05,
"loss": 0.3082,
"step": 57000
},
{
"epoch": 2.32,
"eval_accuracy": 0.8076305220883534,
"eval_loss": 0.6049151420593262,
"eval_runtime": 9.2353,
"eval_samples_per_second": 269.619,
"step": 57000
},
{
"epoch": 2.33,
"learning_rate": 1.1200295356341956e-05,
"loss": 0.273,
"step": 57250
},
{
"epoch": 2.34,
"learning_rate": 1.1029371547339059e-05,
"loss": 0.2773,
"step": 57500
},
{
"epoch": 2.34,
"eval_accuracy": 0.810843373493976,
"eval_loss": 0.6248819231987,
"eval_runtime": 9.0359,
"eval_samples_per_second": 275.569,
"step": 57500
},
{
"epoch": 2.35,
"learning_rate": 1.085844773833616e-05,
"loss": 0.3164,
"step": 57750
},
{
"epoch": 2.36,
"learning_rate": 1.068752392933326e-05,
"loss": 0.2824,
"step": 58000
},
{
"epoch": 2.36,
"eval_accuracy": 0.8136546184738955,
"eval_loss": 0.5784918069839478,
"eval_runtime": 9.2382,
"eval_samples_per_second": 269.533,
"step": 58000
},
{
"epoch": 2.37,
"learning_rate": 1.0516600120330363e-05,
"loss": 0.2992,
"step": 58250
},
{
"epoch": 2.38,
"learning_rate": 1.0345676311327464e-05,
"loss": 0.293,
"step": 58500
},
{
"epoch": 2.38,
"eval_accuracy": 0.8072289156626506,
"eval_loss": 0.611821711063385,
"eval_runtime": 9.0825,
"eval_samples_per_second": 274.153,
"step": 58500
},
{
"epoch": 2.39,
"learning_rate": 1.0174752502324563e-05,
"loss": 0.2742,
"step": 58750
},
{
"epoch": 2.4,
"learning_rate": 1.0003828693321665e-05,
"loss": 0.2927,
"step": 59000
},
{
"epoch": 2.4,
"eval_accuracy": 0.804417670682731,
"eval_loss": 0.6135737895965576,
"eval_runtime": 9.1843,
"eval_samples_per_second": 271.114,
"step": 59000
},
{
"epoch": 2.41,
"learning_rate": 9.832904884318766e-06,
"loss": 0.282,
"step": 59250
},
{
"epoch": 2.42,
"learning_rate": 9.661981075315867e-06,
"loss": 0.3021,
"step": 59500
},
{
"epoch": 2.42,
"eval_accuracy": 0.8088353413654619,
"eval_loss": 0.5996263027191162,
"eval_runtime": 9.1543,
"eval_samples_per_second": 272.003,
"step": 59500
},
{
"epoch": 2.43,
"learning_rate": 9.49105726631297e-06,
"loss": 0.2931,
"step": 59750
},
{
"epoch": 2.44,
"learning_rate": 9.320133457310069e-06,
"loss": 0.2745,
"step": 60000
},
{
"epoch": 2.44,
"eval_accuracy": 0.810843373493976,
"eval_loss": 0.5868379473686218,
"eval_runtime": 9.1158,
"eval_samples_per_second": 273.152,
"step": 60000
},
{
"epoch": 2.45,
"learning_rate": 9.149209648307171e-06,
"loss": 0.2897,
"step": 60250
},
{
"epoch": 2.46,
"learning_rate": 8.978285839304272e-06,
"loss": 0.2919,
"step": 60500
},
{
"epoch": 2.46,
"eval_accuracy": 0.8096385542168675,
"eval_loss": 0.5863232016563416,
"eval_runtime": 9.7927,
"eval_samples_per_second": 254.27,
"step": 60500
},
{
"epoch": 2.48,
"learning_rate": 8.807362030301373e-06,
"loss": 0.2903,
"step": 60750
},
{
"epoch": 2.49,
"learning_rate": 8.636438221298475e-06,
"loss": 0.2662,
"step": 61000
},
{
"epoch": 2.49,
"eval_accuracy": 0.804417670682731,
"eval_loss": 0.636022686958313,
"eval_runtime": 10.4967,
"eval_samples_per_second": 237.217,
"step": 61000
},
{
"epoch": 2.5,
"learning_rate": 8.465514412295575e-06,
"loss": 0.2908,
"step": 61250
},
{
"epoch": 2.51,
"learning_rate": 8.294590603292677e-06,
"loss": 0.2977,
"step": 61500
},
{
"epoch": 2.51,
"eval_accuracy": 0.8104417670682731,
"eval_loss": 0.596953272819519,
"eval_runtime": 9.4896,
"eval_samples_per_second": 262.393,
"step": 61500
},
{
"epoch": 2.52,
"learning_rate": 8.123666794289778e-06,
"loss": 0.2723,
"step": 61750
},
{
"epoch": 2.53,
"learning_rate": 7.952742985286879e-06,
"loss": 0.2785,
"step": 62000
},
{
"epoch": 2.53,
"eval_accuracy": 0.8164658634538152,
"eval_loss": 0.5756428241729736,
"eval_runtime": 9.105,
"eval_samples_per_second": 273.476,
"step": 62000
}
],
"max_steps": 73632,
"num_train_epochs": 3,
"total_flos": 55970352570961800,
"trial_name": null,
"trial_params": null
}