rclip / trainer_state.json
kaveh's picture
init
12cb5cd
raw
history blame
29 kB
{
"best_metric": 0.3388192057609558,
"best_model_checkpoint": "outputs/output_8_clip14_cxrbert/checkpoint-22500",
"epoch": 8.0,
"global_step": 43608,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.09,
"learning_rate": 4.999999755266707e-05,
"loss": 0.7951,
"step": 500
},
{
"epoch": 0.09,
"eval_loss": 1.1912389993667603,
"eval_runtime": 139.1386,
"eval_samples_per_second": 58.747,
"eval_steps_per_second": 2.451,
"step": 500
},
{
"epoch": 0.18,
"learning_rate": 4.993123185382302e-05,
"loss": 0.5887,
"step": 1000
},
{
"epoch": 0.18,
"eval_loss": 0.9833270907402039,
"eval_runtime": 139.0379,
"eval_samples_per_second": 58.79,
"eval_steps_per_second": 2.453,
"step": 1000
},
{
"epoch": 0.28,
"learning_rate": 4.972693864808811e-05,
"loss": 0.5023,
"step": 1500
},
{
"epoch": 0.28,
"eval_loss": 0.8458877205848694,
"eval_runtime": 139.1851,
"eval_samples_per_second": 58.728,
"eval_steps_per_second": 2.45,
"step": 1500
},
{
"epoch": 0.37,
"learning_rate": 4.938822848423147e-05,
"loss": 0.4709,
"step": 2000
},
{
"epoch": 0.37,
"eval_loss": 0.8479061126708984,
"eval_runtime": 138.6519,
"eval_samples_per_second": 58.953,
"eval_steps_per_second": 2.459,
"step": 2000
},
{
"epoch": 0.46,
"learning_rate": 4.891694260878015e-05,
"loss": 0.4484,
"step": 2500
},
{
"epoch": 0.46,
"eval_loss": 0.766708493232727,
"eval_runtime": 138.9466,
"eval_samples_per_second": 58.828,
"eval_steps_per_second": 2.454,
"step": 2500
},
{
"epoch": 0.55,
"learning_rate": 4.831564295690475e-05,
"loss": 0.4319,
"step": 3000
},
{
"epoch": 0.55,
"eval_loss": 0.8092461228370667,
"eval_runtime": 138.9741,
"eval_samples_per_second": 58.817,
"eval_steps_per_second": 2.454,
"step": 3000
},
{
"epoch": 0.64,
"learning_rate": 4.7587598225603125e-05,
"loss": 0.4181,
"step": 3500
},
{
"epoch": 0.64,
"eval_loss": 0.6963649392127991,
"eval_runtime": 138.9254,
"eval_samples_per_second": 58.837,
"eval_steps_per_second": 2.455,
"step": 3500
},
{
"epoch": 0.73,
"learning_rate": 4.673676610488902e-05,
"loss": 0.4107,
"step": 4000
},
{
"epoch": 0.73,
"eval_loss": 0.6463401913642883,
"eval_runtime": 138.8758,
"eval_samples_per_second": 58.858,
"eval_steps_per_second": 2.455,
"step": 4000
},
{
"epoch": 0.83,
"learning_rate": 4.576777176357795e-05,
"loss": 0.3723,
"step": 4500
},
{
"epoch": 0.83,
"eval_loss": 0.7892907857894897,
"eval_runtime": 138.7167,
"eval_samples_per_second": 58.926,
"eval_steps_per_second": 2.458,
"step": 4500
},
{
"epoch": 0.92,
"learning_rate": 4.468588270662272e-05,
"loss": 0.3746,
"step": 5000
},
{
"epoch": 0.92,
"eval_loss": 0.686305582523346,
"eval_runtime": 139.0126,
"eval_samples_per_second": 58.8,
"eval_steps_per_second": 2.453,
"step": 5000
},
{
"epoch": 1.01,
"learning_rate": 4.349698014067534e-05,
"loss": 0.3667,
"step": 5500
},
{
"epoch": 1.01,
"eval_loss": 0.6910073161125183,
"eval_runtime": 138.8546,
"eval_samples_per_second": 58.867,
"eval_steps_per_second": 2.456,
"step": 5500
},
{
"epoch": 1.1,
"learning_rate": 4.220752700353382e-05,
"loss": 0.3253,
"step": 6000
},
{
"epoch": 1.1,
"eval_loss": 0.6863256096839905,
"eval_runtime": 138.8547,
"eval_samples_per_second": 58.867,
"eval_steps_per_second": 2.456,
"step": 6000
},
{
"epoch": 1.19,
"learning_rate": 4.082453283126738e-05,
"loss": 0.3274,
"step": 6500
},
{
"epoch": 1.19,
"eval_loss": 0.6445034146308899,
"eval_runtime": 138.846,
"eval_samples_per_second": 58.871,
"eval_steps_per_second": 2.456,
"step": 6500
},
{
"epoch": 1.28,
"learning_rate": 3.935551565400428e-05,
"loss": 0.3065,
"step": 7000
},
{
"epoch": 1.28,
"eval_loss": 0.5908203125,
"eval_runtime": 138.7486,
"eval_samples_per_second": 58.912,
"eval_steps_per_second": 2.458,
"step": 7000
},
{
"epoch": 1.38,
"learning_rate": 3.7808461127518854e-05,
"loss": 0.2834,
"step": 7500
},
{
"epoch": 1.38,
"eval_loss": 0.6137728691101074,
"eval_runtime": 139.0095,
"eval_samples_per_second": 58.802,
"eval_steps_per_second": 2.453,
"step": 7500
},
{
"epoch": 1.47,
"learning_rate": 3.6191779122780486e-05,
"loss": 0.293,
"step": 8000
},
{
"epoch": 1.47,
"eval_loss": 0.6515378355979919,
"eval_runtime": 139.1162,
"eval_samples_per_second": 58.757,
"eval_steps_per_second": 2.451,
"step": 8000
},
{
"epoch": 1.56,
"learning_rate": 3.4514258009446234e-05,
"loss": 0.303,
"step": 8500
},
{
"epoch": 1.56,
"eval_loss": 0.5806155800819397,
"eval_runtime": 138.924,
"eval_samples_per_second": 58.838,
"eval_steps_per_second": 2.455,
"step": 8500
},
{
"epoch": 1.65,
"learning_rate": 3.278501688181439e-05,
"loss": 0.2638,
"step": 9000
},
{
"epoch": 1.65,
"eval_loss": 0.5586961507797241,
"eval_runtime": 139.1301,
"eval_samples_per_second": 58.751,
"eval_steps_per_second": 2.451,
"step": 9000
},
{
"epoch": 1.74,
"learning_rate": 3.101345598694112e-05,
"loss": 0.2593,
"step": 9500
},
{
"epoch": 1.74,
"eval_loss": 0.5215563178062439,
"eval_runtime": 138.9957,
"eval_samples_per_second": 58.808,
"eval_steps_per_second": 2.453,
"step": 9500
},
{
"epoch": 1.83,
"learning_rate": 2.9209205624395885e-05,
"loss": 0.2451,
"step": 10000
},
{
"epoch": 1.83,
"eval_loss": 0.5282728672027588,
"eval_runtime": 138.8608,
"eval_samples_per_second": 58.865,
"eval_steps_per_second": 2.456,
"step": 10000
},
{
"epoch": 1.93,
"learning_rate": 2.7382073795438957e-05,
"loss": 0.2468,
"step": 10500
},
{
"epoch": 1.93,
"eval_loss": 0.5001487135887146,
"eval_runtime": 138.9207,
"eval_samples_per_second": 58.839,
"eval_steps_per_second": 2.455,
"step": 10500
},
{
"epoch": 2.02,
"learning_rate": 2.5541992886203175e-05,
"loss": 0.2295,
"step": 11000
},
{
"epoch": 2.02,
"eval_loss": 0.49750879406929016,
"eval_runtime": 138.9502,
"eval_samples_per_second": 58.827,
"eval_steps_per_second": 2.454,
"step": 11000
},
{
"epoch": 2.11,
"learning_rate": 2.3698965674712838e-05,
"loss": 0.1953,
"step": 11500
},
{
"epoch": 2.11,
"eval_loss": 0.4750489890575409,
"eval_runtime": 138.8668,
"eval_samples_per_second": 58.862,
"eval_steps_per_second": 2.456,
"step": 11500
},
{
"epoch": 2.2,
"learning_rate": 2.1863010955248543e-05,
"loss": 0.1954,
"step": 12000
},
{
"epoch": 2.2,
"eval_loss": 0.45723679661750793,
"eval_runtime": 139.0817,
"eval_samples_per_second": 58.771,
"eval_steps_per_second": 2.452,
"step": 12000
},
{
"epoch": 2.29,
"learning_rate": 2.0044109075646793e-05,
"loss": 0.1737,
"step": 12500
},
{
"epoch": 2.29,
"eval_loss": 0.4731180667877197,
"eval_runtime": 139.0822,
"eval_samples_per_second": 58.771,
"eval_steps_per_second": 2.452,
"step": 12500
},
{
"epoch": 2.38,
"learning_rate": 1.8252147683596503e-05,
"loss": 0.175,
"step": 13000
},
{
"epoch": 2.38,
"eval_loss": 0.4526049494743347,
"eval_runtime": 139.059,
"eval_samples_per_second": 58.781,
"eval_steps_per_second": 2.452,
"step": 13000
},
{
"epoch": 2.48,
"learning_rate": 1.6496867976858525e-05,
"loss": 0.1873,
"step": 13500
},
{
"epoch": 2.48,
"eval_loss": 0.4890150725841522,
"eval_runtime": 138.99,
"eval_samples_per_second": 58.81,
"eval_steps_per_second": 2.453,
"step": 13500
},
{
"epoch": 2.57,
"learning_rate": 1.4787811749594674e-05,
"loss": 0.1809,
"step": 14000
},
{
"epoch": 2.57,
"eval_loss": 0.4210197627544403,
"eval_runtime": 139.0599,
"eval_samples_per_second": 58.78,
"eval_steps_per_second": 2.452,
"step": 14000
},
{
"epoch": 2.66,
"learning_rate": 1.3134269522665521e-05,
"loss": 0.1711,
"step": 14500
},
{
"epoch": 2.66,
"eval_loss": 0.4197298586368561,
"eval_runtime": 139.0776,
"eval_samples_per_second": 58.773,
"eval_steps_per_second": 2.452,
"step": 14500
},
{
"epoch": 2.75,
"learning_rate": 1.1545230039863117e-05,
"loss": 0.1457,
"step": 15000
},
{
"epoch": 2.75,
"eval_loss": 0.3998343348503113,
"eval_runtime": 138.7787,
"eval_samples_per_second": 58.9,
"eval_steps_per_second": 2.457,
"step": 15000
},
{
"epoch": 2.84,
"learning_rate": 1.0029331404620077e-05,
"loss": 0.1583,
"step": 15500
},
{
"epoch": 2.84,
"eval_loss": 0.392282098531723,
"eval_runtime": 139.0688,
"eval_samples_per_second": 58.777,
"eval_steps_per_second": 2.452,
"step": 15500
},
{
"epoch": 2.94,
"learning_rate": 8.59481412281825e-06,
"loss": 0.1579,
"step": 16000
},
{
"epoch": 2.94,
"eval_loss": 0.3823428750038147,
"eval_runtime": 139.0069,
"eval_samples_per_second": 58.803,
"eval_steps_per_second": 2.453,
"step": 16000
},
{
"epoch": 3.03,
"learning_rate": 7.249476306959052e-06,
"loss": 0.1339,
"step": 16500
},
{
"epoch": 3.03,
"eval_loss": 0.3654000163078308,
"eval_runtime": 139.085,
"eval_samples_per_second": 58.77,
"eval_steps_per_second": 2.452,
"step": 16500
},
{
"epoch": 3.12,
"learning_rate": 6.00063128520765e-06,
"loss": 0.1164,
"step": 17000
},
{
"epoch": 3.12,
"eval_loss": 0.3591544032096863,
"eval_runtime": 139.2185,
"eval_samples_per_second": 58.713,
"eval_steps_per_second": 2.449,
"step": 17000
},
{
"epoch": 3.21,
"learning_rate": 4.855067845750841e-06,
"loss": 0.1217,
"step": 17500
},
{
"epoch": 3.21,
"eval_loss": 0.3641490936279297,
"eval_runtime": 139.1903,
"eval_samples_per_second": 58.725,
"eval_steps_per_second": 2.45,
"step": 17500
},
{
"epoch": 3.3,
"learning_rate": 3.8190133325820834e-06,
"loss": 0.119,
"step": 18000
},
{
"epoch": 3.3,
"eval_loss": 0.3553272783756256,
"eval_runtime": 139.0893,
"eval_samples_per_second": 58.768,
"eval_steps_per_second": 2.452,
"step": 18000
},
{
"epoch": 3.39,
"learning_rate": 2.8980997933272802e-06,
"loss": 0.1151,
"step": 18500
},
{
"epoch": 3.39,
"eval_loss": 0.35238373279571533,
"eval_runtime": 139.0702,
"eval_samples_per_second": 58.776,
"eval_steps_per_second": 2.452,
"step": 18500
},
{
"epoch": 3.49,
"learning_rate": 2.0973333631332525e-06,
"loss": 0.119,
"step": 19000
},
{
"epoch": 3.49,
"eval_loss": 0.3452140688896179,
"eval_runtime": 138.934,
"eval_samples_per_second": 58.834,
"eval_steps_per_second": 2.454,
"step": 19000
},
{
"epoch": 3.58,
"learning_rate": 1.4210670510499595e-06,
"loss": 0.102,
"step": 19500
},
{
"epoch": 3.58,
"eval_loss": 0.34390997886657715,
"eval_runtime": 139.1805,
"eval_samples_per_second": 58.729,
"eval_steps_per_second": 2.45,
"step": 19500
},
{
"epoch": 3.67,
"learning_rate": 8.729770768409501e-07,
"loss": 0.1085,
"step": 20000
},
{
"epoch": 3.67,
"eval_loss": 0.3422289192676544,
"eval_runtime": 139.0942,
"eval_samples_per_second": 58.766,
"eval_steps_per_second": 2.452,
"step": 20000
},
{
"epoch": 3.76,
"learning_rate": 4.5604288685657804e-07,
"loss": 0.1142,
"step": 20500
},
{
"epoch": 3.76,
"eval_loss": 0.33955371379852295,
"eval_runtime": 138.9826,
"eval_samples_per_second": 58.813,
"eval_steps_per_second": 2.454,
"step": 20500
},
{
"epoch": 3.85,
"learning_rate": 1.7253095760459415e-07,
"loss": 0.1038,
"step": 21000
},
{
"epoch": 3.85,
"eval_loss": 0.33917009830474854,
"eval_runtime": 139.1121,
"eval_samples_per_second": 58.758,
"eval_steps_per_second": 2.451,
"step": 21000
},
{
"epoch": 3.94,
"learning_rate": 2.3982475062916954e-08,
"loss": 0.1143,
"step": 21500
},
{
"epoch": 3.94,
"eval_loss": 0.33897778391838074,
"eval_runtime": 139.1906,
"eval_samples_per_second": 58.725,
"eval_steps_per_second": 2.45,
"step": 21500
},
{
"epoch": 4.04,
"learning_rate": 1.1204956710403336e-08,
"loss": 0.0983,
"step": 22000
},
{
"epoch": 4.04,
"eval_loss": 0.3389684855937958,
"eval_runtime": 139.2217,
"eval_samples_per_second": 58.712,
"eval_steps_per_second": 2.449,
"step": 22000
},
{
"epoch": 4.13,
"learning_rate": 1.3426786181872375e-07,
"loss": 0.0974,
"step": 22500
},
{
"epoch": 4.13,
"eval_loss": 0.3388192057609558,
"eval_runtime": 139.1103,
"eval_samples_per_second": 58.759,
"eval_steps_per_second": 2.451,
"step": 22500
},
{
"epoch": 4.22,
"learning_rate": 3.925022138680762e-07,
"loss": 0.1007,
"step": 23000
},
{
"epoch": 4.22,
"eval_loss": 0.33886849880218506,
"eval_runtime": 139.2186,
"eval_samples_per_second": 58.713,
"eval_steps_per_second": 2.449,
"step": 23000
},
{
"epoch": 4.31,
"learning_rate": 7.845042371392303e-07,
"loss": 0.0903,
"step": 23500
},
{
"epoch": 4.31,
"eval_loss": 0.33964774012565613,
"eval_runtime": 139.1917,
"eval_samples_per_second": 58.725,
"eval_steps_per_second": 2.45,
"step": 23500
},
{
"epoch": 4.4,
"learning_rate": 1.308142987713265e-06,
"loss": 0.095,
"step": 24000
},
{
"epoch": 4.4,
"eval_loss": 0.3394069969654083,
"eval_runtime": 139.0407,
"eval_samples_per_second": 58.789,
"eval_steps_per_second": 2.453,
"step": 24000
},
{
"epoch": 4.49,
"learning_rate": 1.960571937396438e-06,
"loss": 0.0955,
"step": 24500
},
{
"epoch": 4.49,
"eval_loss": 0.3435823619365692,
"eval_runtime": 138.8656,
"eval_samples_per_second": 58.863,
"eval_steps_per_second": 2.456,
"step": 24500
},
{
"epoch": 4.59,
"learning_rate": 2.7382444475993473e-06,
"loss": 0.1032,
"step": 25000
},
{
"epoch": 4.59,
"eval_loss": 0.3425971269607544,
"eval_runtime": 139.1498,
"eval_samples_per_second": 58.742,
"eval_steps_per_second": 2.451,
"step": 25000
},
{
"epoch": 4.68,
"learning_rate": 3.636933049053598e-06,
"loss": 0.1037,
"step": 25500
},
{
"epoch": 4.68,
"eval_loss": 0.3484514653682709,
"eval_runtime": 139.2414,
"eval_samples_per_second": 58.704,
"eval_steps_per_second": 2.449,
"step": 25500
},
{
"epoch": 4.77,
"learning_rate": 4.651752422560337e-06,
"loss": 0.103,
"step": 26000
},
{
"epoch": 4.77,
"eval_loss": 0.35472801327705383,
"eval_runtime": 139.171,
"eval_samples_per_second": 58.733,
"eval_steps_per_second": 2.45,
"step": 26000
},
{
"epoch": 4.86,
"learning_rate": 5.777185955846176e-06,
"loss": 0.0987,
"step": 26500
},
{
"epoch": 4.86,
"eval_loss": 0.355197936296463,
"eval_runtime": 139.255,
"eval_samples_per_second": 58.698,
"eval_steps_per_second": 2.449,
"step": 26500
},
{
"epoch": 4.95,
"learning_rate": 7.007115732161859e-06,
"loss": 0.1076,
"step": 27000
},
{
"epoch": 4.95,
"eval_loss": 0.35372602939605713,
"eval_runtime": 139.256,
"eval_samples_per_second": 58.698,
"eval_steps_per_second": 2.449,
"step": 27000
},
{
"epoch": 5.04,
"learning_rate": 8.334855787604286e-06,
"loss": 0.1134,
"step": 27500
},
{
"epoch": 5.04,
"eval_loss": 0.35491758584976196,
"eval_runtime": 139.157,
"eval_samples_per_second": 58.739,
"eval_steps_per_second": 2.45,
"step": 27500
},
{
"epoch": 5.14,
"learning_rate": 9.753188456373041e-06,
"loss": 0.1044,
"step": 28000
},
{
"epoch": 5.14,
"eval_loss": 0.362209677696228,
"eval_runtime": 139.1058,
"eval_samples_per_second": 58.761,
"eval_steps_per_second": 2.451,
"step": 28000
},
{
"epoch": 5.23,
"learning_rate": 1.1254403606386926e-05,
"loss": 0.1099,
"step": 28500
},
{
"epoch": 5.23,
"eval_loss": 0.37740227580070496,
"eval_runtime": 139.2551,
"eval_samples_per_second": 58.698,
"eval_steps_per_second": 2.449,
"step": 28500
},
{
"epoch": 5.32,
"learning_rate": 1.2830340551973424e-05,
"loss": 0.1129,
"step": 29000
},
{
"epoch": 5.32,
"eval_loss": 0.387184202671051,
"eval_runtime": 138.9995,
"eval_samples_per_second": 58.806,
"eval_steps_per_second": 2.453,
"step": 29000
},
{
"epoch": 5.41,
"learning_rate": 1.4472432415791445e-05,
"loss": 0.1235,
"step": 29500
},
{
"epoch": 5.41,
"eval_loss": 0.3766579329967499,
"eval_runtime": 139.1988,
"eval_samples_per_second": 58.722,
"eval_steps_per_second": 2.45,
"step": 29500
},
{
"epoch": 5.5,
"learning_rate": 1.6171752698833968e-05,
"loss": 0.1099,
"step": 30000
},
{
"epoch": 5.5,
"eval_loss": 0.3879966139793396,
"eval_runtime": 139.1987,
"eval_samples_per_second": 58.722,
"eval_steps_per_second": 2.45,
"step": 30000
},
{
"epoch": 5.6,
"learning_rate": 1.7919063805352744e-05,
"loss": 0.1331,
"step": 30500
},
{
"epoch": 5.6,
"eval_loss": 0.41808027029037476,
"eval_runtime": 139.3796,
"eval_samples_per_second": 58.646,
"eval_steps_per_second": 2.447,
"step": 30500
},
{
"epoch": 5.69,
"learning_rate": 1.9704867258922042e-05,
"loss": 0.134,
"step": 31000
},
{
"epoch": 5.69,
"eval_loss": 0.4090297818183899,
"eval_runtime": 139.2797,
"eval_samples_per_second": 58.688,
"eval_steps_per_second": 2.448,
"step": 31000
},
{
"epoch": 5.78,
"learning_rate": 2.1519455336663182e-05,
"loss": 0.142,
"step": 31500
},
{
"epoch": 5.78,
"eval_loss": 0.4044671654701233,
"eval_runtime": 139.2242,
"eval_samples_per_second": 58.711,
"eval_steps_per_second": 2.449,
"step": 31500
},
{
"epoch": 5.87,
"learning_rate": 2.335296384094446e-05,
"loss": 0.1441,
"step": 32000
},
{
"epoch": 5.87,
"eval_loss": 0.41757142543792725,
"eval_runtime": 139.1671,
"eval_samples_per_second": 58.735,
"eval_steps_per_second": 2.45,
"step": 32000
},
{
"epoch": 5.96,
"learning_rate": 2.51954257216856e-05,
"loss": 0.1577,
"step": 32500
},
{
"epoch": 5.96,
"eval_loss": 0.43774479627609253,
"eval_runtime": 139.219,
"eval_samples_per_second": 58.713,
"eval_steps_per_second": 2.449,
"step": 32500
},
{
"epoch": 6.05,
"learning_rate": 2.703682525777417e-05,
"loss": 0.1539,
"step": 33000
},
{
"epoch": 6.05,
"eval_loss": 0.43269890546798706,
"eval_runtime": 139.3068,
"eval_samples_per_second": 58.676,
"eval_steps_per_second": 2.448,
"step": 33000
},
{
"epoch": 6.15,
"learning_rate": 2.8867152503059856e-05,
"loss": 0.1475,
"step": 33500
},
{
"epoch": 6.15,
"eval_loss": 0.4586590826511383,
"eval_runtime": 139.2759,
"eval_samples_per_second": 58.689,
"eval_steps_per_second": 2.448,
"step": 33500
},
{
"epoch": 6.24,
"learning_rate": 3.0676457700956226e-05,
"loss": 0.1616,
"step": 34000
},
{
"epoch": 6.24,
"eval_loss": 0.47090479731559753,
"eval_runtime": 139.1928,
"eval_samples_per_second": 58.724,
"eval_steps_per_second": 2.45,
"step": 34000
},
{
"epoch": 6.33,
"learning_rate": 3.2454905371848176e-05,
"loss": 0.1671,
"step": 34500
},
{
"epoch": 6.33,
"eval_loss": 0.49197548627853394,
"eval_runtime": 139.1637,
"eval_samples_per_second": 58.737,
"eval_steps_per_second": 2.45,
"step": 34500
},
{
"epoch": 6.42,
"learning_rate": 3.4192827779284355e-05,
"loss": 0.1792,
"step": 35000
},
{
"epoch": 6.42,
"eval_loss": 0.48025813698768616,
"eval_runtime": 139.2895,
"eval_samples_per_second": 58.684,
"eval_steps_per_second": 2.448,
"step": 35000
},
{
"epoch": 6.51,
"learning_rate": 3.588077748430818e-05,
"loss": 0.2025,
"step": 35500
},
{
"epoch": 6.51,
"eval_loss": 0.5274905562400818,
"eval_runtime": 138.8747,
"eval_samples_per_second": 58.859,
"eval_steps_per_second": 2.455,
"step": 35500
},
{
"epoch": 6.6,
"learning_rate": 3.7509578702240475e-05,
"loss": 0.1823,
"step": 36000
},
{
"epoch": 6.6,
"eval_loss": 0.5114786028862,
"eval_runtime": 139.2682,
"eval_samples_per_second": 58.692,
"eval_steps_per_second": 2.449,
"step": 36000
},
{
"epoch": 6.7,
"learning_rate": 3.9070377182734444e-05,
"loss": 0.2123,
"step": 36500
},
{
"epoch": 6.7,
"eval_loss": 0.4975065290927887,
"eval_runtime": 138.9217,
"eval_samples_per_second": 58.839,
"eval_steps_per_second": 2.455,
"step": 36500
},
{
"epoch": 6.79,
"learning_rate": 4.0554688341953205e-05,
"loss": 0.2043,
"step": 37000
},
{
"epoch": 6.79,
"eval_loss": 0.48896968364715576,
"eval_runtime": 139.3258,
"eval_samples_per_second": 58.668,
"eval_steps_per_second": 2.448,
"step": 37000
},
{
"epoch": 6.88,
"learning_rate": 4.19544433852203e-05,
"loss": 0.2086,
"step": 37500
},
{
"epoch": 6.88,
"eval_loss": 0.5374048352241516,
"eval_runtime": 139.1786,
"eval_samples_per_second": 58.73,
"eval_steps_per_second": 2.45,
"step": 37500
},
{
"epoch": 6.97,
"learning_rate": 4.326203316941825e-05,
"loss": 0.2299,
"step": 38000
},
{
"epoch": 6.97,
"eval_loss": 0.5565398335456848,
"eval_runtime": 139.0129,
"eval_samples_per_second": 58.8,
"eval_steps_per_second": 2.453,
"step": 38000
},
{
"epoch": 7.06,
"learning_rate": 4.44703495666965e-05,
"loss": 0.2151,
"step": 38500
},
{
"epoch": 7.06,
"eval_loss": 0.6073034405708313,
"eval_runtime": 139.179,
"eval_samples_per_second": 58.73,
"eval_steps_per_second": 2.45,
"step": 38500
},
{
"epoch": 7.15,
"learning_rate": 4.5572824104633835e-05,
"loss": 0.222,
"step": 39000
},
{
"epoch": 7.15,
"eval_loss": 0.5468436479568481,
"eval_runtime": 139.1916,
"eval_samples_per_second": 58.725,
"eval_steps_per_second": 2.45,
"step": 39000
},
{
"epoch": 7.25,
"learning_rate": 4.656346367280503e-05,
"loss": 0.236,
"step": 39500
},
{
"epoch": 7.25,
"eval_loss": 0.5504103899002075,
"eval_runtime": 139.2016,
"eval_samples_per_second": 58.721,
"eval_steps_per_second": 2.45,
"step": 39500
},
{
"epoch": 7.34,
"learning_rate": 4.743688310164889e-05,
"loss": 0.2031,
"step": 40000
},
{
"epoch": 7.34,
"eval_loss": 0.5548919439315796,
"eval_runtime": 139.2056,
"eval_samples_per_second": 58.719,
"eval_steps_per_second": 2.45,
"step": 40000
},
{
"epoch": 7.43,
"learning_rate": 4.818833443653748e-05,
"loss": 0.2251,
"step": 40500
},
{
"epoch": 7.43,
"eval_loss": 0.5905419588088989,
"eval_runtime": 139.1367,
"eval_samples_per_second": 58.748,
"eval_steps_per_second": 2.451,
"step": 40500
},
{
"epoch": 7.52,
"learning_rate": 4.881373274791077e-05,
"loss": 0.2251,
"step": 41000
},
{
"epoch": 7.52,
"eval_loss": 0.6011632680892944,
"eval_runtime": 139.1129,
"eval_samples_per_second": 58.758,
"eval_steps_per_second": 2.451,
"step": 41000
},
{
"epoch": 7.61,
"learning_rate": 4.9309678337171785e-05,
"loss": 0.2464,
"step": 41500
},
{
"epoch": 7.61,
"eval_loss": 0.5931146740913391,
"eval_runtime": 139.1046,
"eval_samples_per_second": 58.762,
"eval_steps_per_second": 2.451,
"step": 41500
},
{
"epoch": 7.71,
"learning_rate": 4.9673475217629615e-05,
"loss": 0.2451,
"step": 42000
},
{
"epoch": 7.71,
"eval_loss": 0.6498579978942871,
"eval_runtime": 139.0889,
"eval_samples_per_second": 58.768,
"eval_steps_per_second": 2.452,
"step": 42000
},
{
"epoch": 7.8,
"learning_rate": 4.990314577002693e-05,
"loss": 0.2463,
"step": 42500
},
{
"epoch": 7.8,
"eval_loss": 0.5696046948432922,
"eval_runtime": 139.0384,
"eval_samples_per_second": 58.79,
"eval_steps_per_second": 2.453,
"step": 42500
},
{
"epoch": 7.89,
"learning_rate": 4.999744149298381e-05,
"loss": 0.2385,
"step": 43000
},
{
"epoch": 7.89,
"eval_loss": 0.5360204577445984,
"eval_runtime": 139.0432,
"eval_samples_per_second": 58.787,
"eval_steps_per_second": 2.452,
"step": 43000
},
{
"epoch": 7.98,
"learning_rate": 4.995584978991786e-05,
"loss": 0.2353,
"step": 43500
},
{
"epoch": 7.98,
"eval_loss": 0.5489608645439148,
"eval_runtime": 139.1804,
"eval_samples_per_second": 58.73,
"eval_steps_per_second": 2.45,
"step": 43500
},
{
"epoch": 8.0,
"step": 43608,
"total_flos": 1.5655045448788992e+17,
"train_loss": 0.21080181559736064,
"train_runtime": 44575.4239,
"train_samples_per_second": 11.741,
"train_steps_per_second": 0.978
}
],
"max_steps": 43608,
"num_train_epochs": 8,
"total_flos": 1.5655045448788992e+17,
"trial_name": null,
"trial_params": null
}