gpt2_xl_discovery / trainer_state.json
prajjwal1
v1
738a15a
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 105231,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.9762427421577294e-05,
"loss": 0.9294,
"step": 500
},
{
"epoch": 0.03,
"learning_rate": 4.9524854843154585e-05,
"loss": 0.8906,
"step": 1000
},
{
"epoch": 0.04,
"learning_rate": 4.9287282264731876e-05,
"loss": 0.8884,
"step": 1500
},
{
"epoch": 0.06,
"learning_rate": 4.904970968630917e-05,
"loss": 0.8958,
"step": 2000
},
{
"epoch": 0.07,
"learning_rate": 4.8812137107886466e-05,
"loss": 0.8973,
"step": 2500
},
{
"epoch": 0.09,
"learning_rate": 4.857456452946375e-05,
"loss": 0.8955,
"step": 3000
},
{
"epoch": 0.1,
"learning_rate": 4.833699195104104e-05,
"loss": 0.8949,
"step": 3500
},
{
"epoch": 0.11,
"learning_rate": 4.809941937261834e-05,
"loss": 0.897,
"step": 4000
},
{
"epoch": 0.13,
"learning_rate": 4.786184679419563e-05,
"loss": 0.9011,
"step": 4500
},
{
"epoch": 0.14,
"learning_rate": 4.7624274215772915e-05,
"loss": 0.902,
"step": 5000
},
{
"epoch": 0.16,
"learning_rate": 4.738670163735021e-05,
"loss": 0.8925,
"step": 5500
},
{
"epoch": 0.17,
"learning_rate": 4.7149129058927504e-05,
"loss": 0.8982,
"step": 6000
},
{
"epoch": 0.19,
"learning_rate": 4.6911556480504796e-05,
"loss": 0.8934,
"step": 6500
},
{
"epoch": 0.2,
"learning_rate": 4.667398390208209e-05,
"loss": 0.8977,
"step": 7000
},
{
"epoch": 0.21,
"learning_rate": 4.643641132365938e-05,
"loss": 0.8981,
"step": 7500
},
{
"epoch": 0.23,
"learning_rate": 4.619883874523667e-05,
"loss": 0.8979,
"step": 8000
},
{
"epoch": 0.24,
"learning_rate": 4.596126616681397e-05,
"loss": 0.9024,
"step": 8500
},
{
"epoch": 0.26,
"learning_rate": 4.572369358839125e-05,
"loss": 0.8902,
"step": 9000
},
{
"epoch": 0.27,
"learning_rate": 4.548612100996854e-05,
"loss": 0.8972,
"step": 9500
},
{
"epoch": 0.29,
"learning_rate": 4.524854843154584e-05,
"loss": 0.8991,
"step": 10000
},
{
"epoch": 0.3,
"learning_rate": 4.501097585312313e-05,
"loss": 0.8913,
"step": 10500
},
{
"epoch": 0.31,
"learning_rate": 4.4773403274700424e-05,
"loss": 0.8995,
"step": 11000
},
{
"epoch": 0.33,
"learning_rate": 4.4535830696277715e-05,
"loss": 0.8954,
"step": 11500
},
{
"epoch": 0.34,
"learning_rate": 4.4298258117855006e-05,
"loss": 0.8964,
"step": 12000
},
{
"epoch": 0.36,
"learning_rate": 4.40606855394323e-05,
"loss": 0.9008,
"step": 12500
},
{
"epoch": 0.37,
"learning_rate": 4.382311296100959e-05,
"loss": 0.8956,
"step": 13000
},
{
"epoch": 0.38,
"learning_rate": 4.358554038258688e-05,
"loss": 0.8993,
"step": 13500
},
{
"epoch": 0.4,
"learning_rate": 4.334796780416417e-05,
"loss": 0.8918,
"step": 14000
},
{
"epoch": 0.41,
"learning_rate": 4.311039522574147e-05,
"loss": 0.8973,
"step": 14500
},
{
"epoch": 0.43,
"learning_rate": 4.2872822647318754e-05,
"loss": 0.8952,
"step": 15000
},
{
"epoch": 0.44,
"learning_rate": 4.263525006889605e-05,
"loss": 0.902,
"step": 15500
},
{
"epoch": 0.46,
"learning_rate": 4.239767749047334e-05,
"loss": 0.898,
"step": 16000
},
{
"epoch": 0.47,
"learning_rate": 4.2160104912050635e-05,
"loss": 0.8949,
"step": 16500
},
{
"epoch": 0.48,
"learning_rate": 4.1922532333627926e-05,
"loss": 0.892,
"step": 17000
},
{
"epoch": 0.5,
"learning_rate": 4.168495975520522e-05,
"loss": 0.8938,
"step": 17500
},
{
"epoch": 0.51,
"learning_rate": 4.144738717678251e-05,
"loss": 0.8949,
"step": 18000
},
{
"epoch": 0.53,
"learning_rate": 4.12098145983598e-05,
"loss": 0.8914,
"step": 18500
},
{
"epoch": 0.54,
"learning_rate": 4.097224201993709e-05,
"loss": 0.891,
"step": 19000
},
{
"epoch": 0.56,
"learning_rate": 4.073466944151438e-05,
"loss": 0.8916,
"step": 19500
},
{
"epoch": 0.57,
"learning_rate": 4.049709686309168e-05,
"loss": 0.8938,
"step": 20000
},
{
"epoch": 0.58,
"learning_rate": 4.025952428466897e-05,
"loss": 0.8906,
"step": 20500
},
{
"epoch": 0.6,
"learning_rate": 4.0021951706246256e-05,
"loss": 0.8925,
"step": 21000
},
{
"epoch": 0.61,
"learning_rate": 3.9784379127823554e-05,
"loss": 0.8888,
"step": 21500
},
{
"epoch": 0.63,
"learning_rate": 3.9546806549400845e-05,
"loss": 0.8944,
"step": 22000
},
{
"epoch": 0.64,
"learning_rate": 3.9309233970978137e-05,
"loss": 0.889,
"step": 22500
},
{
"epoch": 0.66,
"learning_rate": 3.907166139255543e-05,
"loss": 0.8919,
"step": 23000
},
{
"epoch": 0.67,
"learning_rate": 3.883408881413272e-05,
"loss": 0.8929,
"step": 23500
},
{
"epoch": 0.68,
"learning_rate": 3.859651623571001e-05,
"loss": 0.8967,
"step": 24000
},
{
"epoch": 0.7,
"learning_rate": 3.835894365728731e-05,
"loss": 0.8951,
"step": 24500
},
{
"epoch": 0.71,
"learning_rate": 3.812137107886459e-05,
"loss": 0.8923,
"step": 25000
},
{
"epoch": 0.73,
"learning_rate": 3.7883798500441884e-05,
"loss": 0.8911,
"step": 25500
},
{
"epoch": 0.74,
"learning_rate": 3.764622592201918e-05,
"loss": 0.8929,
"step": 26000
},
{
"epoch": 0.76,
"learning_rate": 3.7408653343596473e-05,
"loss": 0.8902,
"step": 26500
},
{
"epoch": 0.77,
"learning_rate": 3.717108076517376e-05,
"loss": 0.8951,
"step": 27000
},
{
"epoch": 0.78,
"learning_rate": 3.6933508186751056e-05,
"loss": 0.8925,
"step": 27500
},
{
"epoch": 0.8,
"learning_rate": 3.669593560832835e-05,
"loss": 0.8939,
"step": 28000
},
{
"epoch": 0.81,
"learning_rate": 3.645836302990564e-05,
"loss": 0.8926,
"step": 28500
},
{
"epoch": 0.83,
"learning_rate": 3.622079045148293e-05,
"loss": 0.8907,
"step": 29000
},
{
"epoch": 0.84,
"learning_rate": 3.598321787306022e-05,
"loss": 0.8933,
"step": 29500
},
{
"epoch": 0.86,
"learning_rate": 3.574564529463751e-05,
"loss": 0.8867,
"step": 30000
},
{
"epoch": 0.87,
"learning_rate": 3.550807271621481e-05,
"loss": 0.8881,
"step": 30500
},
{
"epoch": 0.88,
"learning_rate": 3.5270500137792095e-05,
"loss": 0.8887,
"step": 31000
},
{
"epoch": 0.9,
"learning_rate": 3.5032927559369386e-05,
"loss": 0.8934,
"step": 31500
},
{
"epoch": 0.91,
"learning_rate": 3.4795354980946684e-05,
"loss": 0.888,
"step": 32000
},
{
"epoch": 0.93,
"learning_rate": 3.4557782402523975e-05,
"loss": 0.8844,
"step": 32500
},
{
"epoch": 0.94,
"learning_rate": 3.432020982410126e-05,
"loss": 0.8877,
"step": 33000
},
{
"epoch": 0.96,
"learning_rate": 3.408263724567856e-05,
"loss": 0.887,
"step": 33500
},
{
"epoch": 0.97,
"learning_rate": 3.384506466725585e-05,
"loss": 0.8879,
"step": 34000
},
{
"epoch": 0.98,
"learning_rate": 3.360749208883314e-05,
"loss": 0.888,
"step": 34500
},
{
"epoch": 1.0,
"learning_rate": 3.336991951041043e-05,
"loss": 0.885,
"step": 35000
},
{
"epoch": 1.01,
"learning_rate": 3.313234693198772e-05,
"loss": 0.7613,
"step": 35500
},
{
"epoch": 1.03,
"learning_rate": 3.2894774353565014e-05,
"loss": 0.7311,
"step": 36000
},
{
"epoch": 1.04,
"learning_rate": 3.265720177514231e-05,
"loss": 0.7264,
"step": 36500
},
{
"epoch": 1.05,
"learning_rate": 3.24196291967196e-05,
"loss": 0.7286,
"step": 37000
},
{
"epoch": 1.07,
"learning_rate": 3.218205661829689e-05,
"loss": 0.7323,
"step": 37500
},
{
"epoch": 1.08,
"learning_rate": 3.1944484039874186e-05,
"loss": 0.7263,
"step": 38000
},
{
"epoch": 1.1,
"learning_rate": 3.170691146145148e-05,
"loss": 0.7283,
"step": 38500
},
{
"epoch": 1.11,
"learning_rate": 3.146933888302876e-05,
"loss": 0.7314,
"step": 39000
},
{
"epoch": 1.13,
"learning_rate": 3.123176630460606e-05,
"loss": 0.7359,
"step": 39500
},
{
"epoch": 1.14,
"learning_rate": 3.099419372618335e-05,
"loss": 0.7309,
"step": 40000
},
{
"epoch": 1.15,
"learning_rate": 3.075662114776064e-05,
"loss": 0.7298,
"step": 40500
},
{
"epoch": 1.17,
"learning_rate": 3.0519048569337934e-05,
"loss": 0.7363,
"step": 41000
},
{
"epoch": 1.18,
"learning_rate": 3.0281475990915225e-05,
"loss": 0.7333,
"step": 41500
},
{
"epoch": 1.2,
"learning_rate": 3.004390341249252e-05,
"loss": 0.7342,
"step": 42000
},
{
"epoch": 1.21,
"learning_rate": 2.980633083406981e-05,
"loss": 0.735,
"step": 42500
},
{
"epoch": 1.23,
"learning_rate": 2.95687582556471e-05,
"loss": 0.7365,
"step": 43000
},
{
"epoch": 1.24,
"learning_rate": 2.9331185677224394e-05,
"loss": 0.7335,
"step": 43500
},
{
"epoch": 1.25,
"learning_rate": 2.9093613098801685e-05,
"loss": 0.7339,
"step": 44000
},
{
"epoch": 1.27,
"learning_rate": 2.885604052037898e-05,
"loss": 0.7369,
"step": 44500
},
{
"epoch": 1.28,
"learning_rate": 2.8618467941956267e-05,
"loss": 0.7362,
"step": 45000
},
{
"epoch": 1.3,
"learning_rate": 2.8380895363533562e-05,
"loss": 0.7375,
"step": 45500
},
{
"epoch": 1.31,
"learning_rate": 2.8143322785110853e-05,
"loss": 0.7317,
"step": 46000
},
{
"epoch": 1.33,
"learning_rate": 2.7905750206688148e-05,
"loss": 0.7348,
"step": 46500
},
{
"epoch": 1.34,
"learning_rate": 2.7668177628265436e-05,
"loss": 0.741,
"step": 47000
},
{
"epoch": 1.35,
"learning_rate": 2.7430605049842727e-05,
"loss": 0.7332,
"step": 47500
},
{
"epoch": 1.37,
"learning_rate": 2.7193032471420022e-05,
"loss": 0.7389,
"step": 48000
},
{
"epoch": 1.38,
"learning_rate": 2.6955459892997313e-05,
"loss": 0.738,
"step": 48500
},
{
"epoch": 1.4,
"learning_rate": 2.67178873145746e-05,
"loss": 0.7434,
"step": 49000
},
{
"epoch": 1.41,
"learning_rate": 2.6480314736151896e-05,
"loss": 0.74,
"step": 49500
},
{
"epoch": 1.43,
"learning_rate": 2.624274215772919e-05,
"loss": 0.738,
"step": 50000
},
{
"epoch": 1.44,
"learning_rate": 2.600516957930648e-05,
"loss": 0.7352,
"step": 50500
},
{
"epoch": 1.45,
"learning_rate": 2.576759700088377e-05,
"loss": 0.7366,
"step": 51000
},
{
"epoch": 1.47,
"learning_rate": 2.5530024422461064e-05,
"loss": 0.7391,
"step": 51500
},
{
"epoch": 1.48,
"learning_rate": 2.5292451844038355e-05,
"loss": 0.7374,
"step": 52000
},
{
"epoch": 1.5,
"learning_rate": 2.505487926561565e-05,
"loss": 0.7424,
"step": 52500
},
{
"epoch": 1.51,
"learning_rate": 2.481730668719294e-05,
"loss": 0.7404,
"step": 53000
},
{
"epoch": 1.53,
"learning_rate": 2.457973410877023e-05,
"loss": 0.7409,
"step": 53500
},
{
"epoch": 1.54,
"learning_rate": 2.4342161530347524e-05,
"loss": 0.7383,
"step": 54000
},
{
"epoch": 1.55,
"learning_rate": 2.4104588951924815e-05,
"loss": 0.7349,
"step": 54500
},
{
"epoch": 1.57,
"learning_rate": 2.3867016373502106e-05,
"loss": 0.7383,
"step": 55000
},
{
"epoch": 1.58,
"learning_rate": 2.3629443795079398e-05,
"loss": 0.741,
"step": 55500
},
{
"epoch": 1.6,
"learning_rate": 2.3391871216656692e-05,
"loss": 0.7409,
"step": 56000
},
{
"epoch": 1.61,
"learning_rate": 2.315429863823398e-05,
"loss": 0.7345,
"step": 56500
},
{
"epoch": 1.62,
"learning_rate": 2.291672605981127e-05,
"loss": 0.7389,
"step": 57000
},
{
"epoch": 1.64,
"learning_rate": 2.2679153481388566e-05,
"loss": 0.7383,
"step": 57500
},
{
"epoch": 1.65,
"learning_rate": 2.2441580902965857e-05,
"loss": 0.7364,
"step": 58000
},
{
"epoch": 1.67,
"learning_rate": 2.220400832454315e-05,
"loss": 0.7388,
"step": 58500
},
{
"epoch": 1.68,
"learning_rate": 2.196643574612044e-05,
"loss": 0.741,
"step": 59000
},
{
"epoch": 1.7,
"learning_rate": 2.1728863167697734e-05,
"loss": 0.7387,
"step": 59500
},
{
"epoch": 1.71,
"learning_rate": 2.1491290589275022e-05,
"loss": 0.741,
"step": 60000
},
{
"epoch": 1.72,
"learning_rate": 2.1253718010852317e-05,
"loss": 0.7397,
"step": 60500
},
{
"epoch": 1.74,
"learning_rate": 2.1016145432429608e-05,
"loss": 0.7389,
"step": 61000
},
{
"epoch": 1.75,
"learning_rate": 2.07785728540069e-05,
"loss": 0.7403,
"step": 61500
},
{
"epoch": 1.77,
"learning_rate": 2.054100027558419e-05,
"loss": 0.7448,
"step": 62000
},
{
"epoch": 1.78,
"learning_rate": 2.0303427697161485e-05,
"loss": 0.7396,
"step": 62500
},
{
"epoch": 1.8,
"learning_rate": 2.0065855118738773e-05,
"loss": 0.7389,
"step": 63000
},
{
"epoch": 1.81,
"learning_rate": 1.9828282540316068e-05,
"loss": 0.739,
"step": 63500
},
{
"epoch": 1.82,
"learning_rate": 1.959070996189336e-05,
"loss": 0.7406,
"step": 64000
},
{
"epoch": 1.84,
"learning_rate": 1.935313738347065e-05,
"loss": 0.7407,
"step": 64500
},
{
"epoch": 1.85,
"learning_rate": 1.9115564805047942e-05,
"loss": 0.7405,
"step": 65000
},
{
"epoch": 1.87,
"learning_rate": 1.8877992226625236e-05,
"loss": 0.7336,
"step": 65500
},
{
"epoch": 1.88,
"learning_rate": 1.8640419648202524e-05,
"loss": 0.7349,
"step": 66000
},
{
"epoch": 1.9,
"learning_rate": 1.840284706977982e-05,
"loss": 0.7414,
"step": 66500
},
{
"epoch": 1.91,
"learning_rate": 1.816527449135711e-05,
"loss": 0.7467,
"step": 67000
},
{
"epoch": 1.92,
"learning_rate": 1.79277019129344e-05,
"loss": 0.7415,
"step": 67500
},
{
"epoch": 1.94,
"learning_rate": 1.7690129334511693e-05,
"loss": 0.7402,
"step": 68000
},
{
"epoch": 1.95,
"learning_rate": 1.7452556756088987e-05,
"loss": 0.7336,
"step": 68500
},
{
"epoch": 1.97,
"learning_rate": 1.7214984177666275e-05,
"loss": 0.7382,
"step": 69000
},
{
"epoch": 1.98,
"learning_rate": 1.697741159924357e-05,
"loss": 0.7398,
"step": 69500
},
{
"epoch": 2.0,
"learning_rate": 1.673983902082086e-05,
"loss": 0.7388,
"step": 70000
},
{
"epoch": 2.01,
"learning_rate": 1.6502266442398152e-05,
"loss": 0.6306,
"step": 70500
},
{
"epoch": 2.02,
"learning_rate": 1.6264693863975444e-05,
"loss": 0.5889,
"step": 71000
},
{
"epoch": 2.04,
"learning_rate": 1.602712128555274e-05,
"loss": 0.583,
"step": 71500
},
{
"epoch": 2.05,
"learning_rate": 1.578954870713003e-05,
"loss": 0.582,
"step": 72000
},
{
"epoch": 2.07,
"learning_rate": 1.555197612870732e-05,
"loss": 0.5843,
"step": 72500
},
{
"epoch": 2.08,
"learning_rate": 1.5314403550284612e-05,
"loss": 0.5826,
"step": 73000
},
{
"epoch": 2.1,
"learning_rate": 1.5076830971861905e-05,
"loss": 0.5884,
"step": 73500
},
{
"epoch": 2.11,
"learning_rate": 1.4839258393439195e-05,
"loss": 0.5861,
"step": 74000
},
{
"epoch": 2.12,
"learning_rate": 1.460168581501649e-05,
"loss": 0.5826,
"step": 74500
},
{
"epoch": 2.14,
"learning_rate": 1.4364113236593779e-05,
"loss": 0.5847,
"step": 75000
},
{
"epoch": 2.15,
"learning_rate": 1.4126540658171072e-05,
"loss": 0.5829,
"step": 75500
},
{
"epoch": 2.17,
"learning_rate": 1.3888968079748363e-05,
"loss": 0.584,
"step": 76000
},
{
"epoch": 2.18,
"learning_rate": 1.3651395501325656e-05,
"loss": 0.5895,
"step": 76500
},
{
"epoch": 2.2,
"learning_rate": 1.3413822922902947e-05,
"loss": 0.5891,
"step": 77000
},
{
"epoch": 2.21,
"learning_rate": 1.317625034448024e-05,
"loss": 0.5846,
"step": 77500
},
{
"epoch": 2.22,
"learning_rate": 1.293867776605753e-05,
"loss": 0.5892,
"step": 78000
},
{
"epoch": 2.24,
"learning_rate": 1.2701105187634823e-05,
"loss": 0.5871,
"step": 78500
},
{
"epoch": 2.25,
"learning_rate": 1.2463532609212116e-05,
"loss": 0.5849,
"step": 79000
},
{
"epoch": 2.27,
"learning_rate": 1.2225960030789407e-05,
"loss": 0.5842,
"step": 79500
},
{
"epoch": 2.28,
"learning_rate": 1.1988387452366698e-05,
"loss": 0.582,
"step": 80000
},
{
"epoch": 2.29,
"learning_rate": 1.1750814873943991e-05,
"loss": 0.5872,
"step": 80500
},
{
"epoch": 2.31,
"learning_rate": 1.1513242295521283e-05,
"loss": 0.5819,
"step": 81000
},
{
"epoch": 2.32,
"learning_rate": 1.1275669717098576e-05,
"loss": 0.5855,
"step": 81500
},
{
"epoch": 2.34,
"learning_rate": 1.1038097138675867e-05,
"loss": 0.5867,
"step": 82000
},
{
"epoch": 2.35,
"learning_rate": 1.0800524560253158e-05,
"loss": 0.5903,
"step": 82500
},
{
"epoch": 2.37,
"learning_rate": 1.0562951981830451e-05,
"loss": 0.5859,
"step": 83000
},
{
"epoch": 2.38,
"learning_rate": 1.0325379403407742e-05,
"loss": 0.5897,
"step": 83500
},
{
"epoch": 2.39,
"learning_rate": 1.0087806824985034e-05,
"loss": 0.5844,
"step": 84000
},
{
"epoch": 2.41,
"learning_rate": 9.850234246562325e-06,
"loss": 0.585,
"step": 84500
},
{
"epoch": 2.42,
"learning_rate": 9.612661668139616e-06,
"loss": 0.5846,
"step": 85000
},
{
"epoch": 2.44,
"learning_rate": 9.375089089716907e-06,
"loss": 0.5847,
"step": 85500
},
{
"epoch": 2.45,
"learning_rate": 9.1375165112942e-06,
"loss": 0.5844,
"step": 86000
},
{
"epoch": 2.47,
"learning_rate": 8.899943932871492e-06,
"loss": 0.5853,
"step": 86500
},
{
"epoch": 2.48,
"learning_rate": 8.662371354448785e-06,
"loss": 0.5918,
"step": 87000
},
{
"epoch": 2.49,
"learning_rate": 8.424798776026076e-06,
"loss": 0.5872,
"step": 87500
},
{
"epoch": 2.51,
"learning_rate": 8.187226197603367e-06,
"loss": 0.5845,
"step": 88000
},
{
"epoch": 2.52,
"learning_rate": 7.94965361918066e-06,
"loss": 0.5867,
"step": 88500
},
{
"epoch": 2.54,
"learning_rate": 7.712081040757951e-06,
"loss": 0.5855,
"step": 89000
},
{
"epoch": 2.55,
"learning_rate": 7.4745084623352435e-06,
"loss": 0.5846,
"step": 89500
},
{
"epoch": 2.57,
"learning_rate": 7.236935883912535e-06,
"loss": 0.5869,
"step": 90000
},
{
"epoch": 2.58,
"learning_rate": 6.999363305489827e-06,
"loss": 0.5836,
"step": 90500
},
{
"epoch": 2.59,
"learning_rate": 6.761790727067119e-06,
"loss": 0.5827,
"step": 91000
},
{
"epoch": 2.61,
"learning_rate": 6.52421814864441e-06,
"loss": 0.5864,
"step": 91500
},
{
"epoch": 2.62,
"learning_rate": 6.286645570221702e-06,
"loss": 0.5881,
"step": 92000
},
{
"epoch": 2.64,
"learning_rate": 6.049072991798995e-06,
"loss": 0.5842,
"step": 92500
},
{
"epoch": 2.65,
"learning_rate": 5.811500413376287e-06,
"loss": 0.5874,
"step": 93000
},
{
"epoch": 2.67,
"learning_rate": 5.573927834953579e-06,
"loss": 0.5838,
"step": 93500
},
{
"epoch": 2.68,
"learning_rate": 5.336355256530871e-06,
"loss": 0.5863,
"step": 94000
},
{
"epoch": 2.69,
"learning_rate": 5.098782678108162e-06,
"loss": 0.5851,
"step": 94500
},
{
"epoch": 2.71,
"learning_rate": 4.861210099685454e-06,
"loss": 0.5861,
"step": 95000
},
{
"epoch": 2.72,
"learning_rate": 4.623637521262746e-06,
"loss": 0.5872,
"step": 95500
},
{
"epoch": 2.74,
"learning_rate": 4.3860649428400385e-06,
"loss": 0.5793,
"step": 96000
},
{
"epoch": 2.75,
"learning_rate": 4.14849236441733e-06,
"loss": 0.5828,
"step": 96500
},
{
"epoch": 2.77,
"learning_rate": 3.910919785994622e-06,
"loss": 0.5863,
"step": 97000
},
{
"epoch": 2.78,
"learning_rate": 3.6733472075719136e-06,
"loss": 0.5874,
"step": 97500
},
{
"epoch": 2.79,
"learning_rate": 3.4357746291492057e-06,
"loss": 0.5854,
"step": 98000
},
{
"epoch": 2.81,
"learning_rate": 3.198202050726497e-06,
"loss": 0.5843,
"step": 98500
},
{
"epoch": 2.82,
"learning_rate": 2.960629472303789e-06,
"loss": 0.5815,
"step": 99000
},
{
"epoch": 2.84,
"learning_rate": 2.7230568938810807e-06,
"loss": 0.5838,
"step": 99500
},
{
"epoch": 2.85,
"learning_rate": 2.485484315458373e-06,
"loss": 0.5837,
"step": 100000
},
{
"epoch": 2.87,
"learning_rate": 2.2479117370356646e-06,
"loss": 0.5805,
"step": 100500
},
{
"epoch": 2.88,
"learning_rate": 2.0103391586129562e-06,
"loss": 0.5829,
"step": 101000
},
{
"epoch": 2.89,
"learning_rate": 1.7727665801902484e-06,
"loss": 0.5848,
"step": 101500
},
{
"epoch": 2.91,
"learning_rate": 1.53519400176754e-06,
"loss": 0.5852,
"step": 102000
},
{
"epoch": 2.92,
"learning_rate": 1.2976214233448317e-06,
"loss": 0.5881,
"step": 102500
},
{
"epoch": 2.94,
"learning_rate": 1.0600488449221236e-06,
"loss": 0.583,
"step": 103000
},
{
"epoch": 2.95,
"learning_rate": 8.224762664994155e-07,
"loss": 0.5828,
"step": 103500
},
{
"epoch": 2.96,
"learning_rate": 5.849036880767075e-07,
"loss": 0.586,
"step": 104000
},
{
"epoch": 2.98,
"learning_rate": 3.473311096539993e-07,
"loss": 0.5817,
"step": 104500
},
{
"epoch": 2.99,
"learning_rate": 1.0975853123129116e-07,
"loss": 0.5815,
"step": 105000
},
{
"epoch": 3.0,
"step": 105231,
"total_flos": 4154092029960192000,
"train_runtime": 251423.0864,
"train_samples_per_second": 0.419
}
],
"max_steps": 105231,
"num_train_epochs": 3,
"total_flos": 4154092029960192000,
"trial_name": null,
"trial_params": null
}