{ "best_metric": null, "best_model_checkpoint": null, "epoch": 5.0, "global_step": 122720, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.02, "learning_rate": 1.991900260756193e-05, "loss": 0.8126, "step": 500 }, { "epoch": 0.04, "learning_rate": 1.983767926988266e-05, "loss": 0.6298, "step": 1000 }, { "epoch": 0.06, "learning_rate": 1.9756192959582793e-05, "loss": 0.5816, "step": 1500 }, { "epoch": 0.08, "learning_rate": 1.967470664928292e-05, "loss": 0.5448, "step": 2000 }, { "epoch": 0.1, "learning_rate": 1.9593220338983052e-05, "loss": 0.5418, "step": 2500 }, { "epoch": 0.12, "learning_rate": 1.9511734028683183e-05, "loss": 0.5363, "step": 3000 }, { "epoch": 0.14, "learning_rate": 1.9430410691003915e-05, "loss": 0.5124, "step": 3500 }, { "epoch": 0.16, "learning_rate": 1.9348924380704043e-05, "loss": 0.5199, "step": 4000 }, { "epoch": 0.18, "learning_rate": 1.9267438070404174e-05, "loss": 0.4912, "step": 4500 }, { "epoch": 0.2, "learning_rate": 1.9185951760104305e-05, "loss": 0.4827, "step": 5000 }, { "epoch": 0.22, "learning_rate": 1.9104465449804433e-05, "loss": 0.4819, "step": 5500 }, { "epoch": 0.24, "learning_rate": 1.9023305084745763e-05, "loss": 0.4897, "step": 6000 }, { "epoch": 0.26, "learning_rate": 1.8941818774445894e-05, "loss": 0.4875, "step": 6500 }, { "epoch": 0.29, "learning_rate": 1.8860332464146026e-05, "loss": 0.4813, "step": 7000 }, { "epoch": 0.31, "learning_rate": 1.8778846153846157e-05, "loss": 0.4721, "step": 7500 }, { "epoch": 0.33, "learning_rate": 1.8697359843546285e-05, "loss": 0.4629, "step": 8000 }, { "epoch": 0.35, "learning_rate": 1.8615873533246416e-05, "loss": 0.4671, "step": 8500 }, { "epoch": 0.37, "learning_rate": 1.8534387222946547e-05, "loss": 0.4841, "step": 9000 }, { "epoch": 0.39, "learning_rate": 1.8452900912646675e-05, "loss": 0.4595, "step": 9500 }, { "epoch": 0.41, "learning_rate": 1.8371577574967407e-05, "loss": 0.4584, "step": 10000 }, { "epoch": 0.43, "learning_rate": 1.8290091264667538e-05, "loss": 0.4544, "step": 10500 }, { "epoch": 0.45, "learning_rate": 1.820860495436767e-05, "loss": 0.4578, "step": 11000 }, { "epoch": 0.47, "learning_rate": 1.8127118644067797e-05, "loss": 0.4599, "step": 11500 }, { "epoch": 0.49, "learning_rate": 1.804563233376793e-05, "loss": 0.4514, "step": 12000 }, { "epoch": 0.51, "learning_rate": 1.796447196870926e-05, "loss": 0.4578, "step": 12500 }, { "epoch": 0.53, "learning_rate": 1.788298565840939e-05, "loss": 0.4533, "step": 13000 }, { "epoch": 0.55, "learning_rate": 1.7801499348109517e-05, "loss": 0.4528, "step": 13500 }, { "epoch": 0.57, "learning_rate": 1.772017601043025e-05, "loss": 0.4496, "step": 14000 }, { "epoch": 0.59, "learning_rate": 1.763868970013038e-05, "loss": 0.4443, "step": 14500 }, { "epoch": 0.61, "learning_rate": 1.7557203389830512e-05, "loss": 0.4603, "step": 15000 }, { "epoch": 0.63, "learning_rate": 1.747571707953064e-05, "loss": 0.4541, "step": 15500 }, { "epoch": 0.65, "learning_rate": 1.739423076923077e-05, "loss": 0.4447, "step": 16000 }, { "epoch": 0.67, "learning_rate": 1.7312744458930902e-05, "loss": 0.4501, "step": 16500 }, { "epoch": 0.69, "learning_rate": 1.723125814863103e-05, "loss": 0.4342, "step": 17000 }, { "epoch": 0.71, "learning_rate": 1.714977183833116e-05, "loss": 0.4501, "step": 17500 }, { "epoch": 0.73, "learning_rate": 1.7068285528031293e-05, "loss": 0.4532, "step": 18000 }, { "epoch": 0.75, "learning_rate": 1.6986799217731424e-05, "loss": 0.4313, "step": 18500 }, { "epoch": 0.77, "learning_rate": 1.690531290743155e-05, "loss": 0.4339, "step": 19000 }, { "epoch": 0.79, "learning_rate": 1.6823989569752284e-05, "loss": 0.4355, "step": 19500 }, { "epoch": 0.81, "learning_rate": 1.6742503259452415e-05, "loss": 0.4155, "step": 20000 }, { "epoch": 0.84, "learning_rate": 1.6661016949152543e-05, "loss": 0.4201, "step": 20500 }, { "epoch": 0.86, "learning_rate": 1.6579530638852674e-05, "loss": 0.4278, "step": 21000 }, { "epoch": 0.88, "learning_rate": 1.6498044328552805e-05, "loss": 0.4436, "step": 21500 }, { "epoch": 0.9, "learning_rate": 1.6416558018252936e-05, "loss": 0.4228, "step": 22000 }, { "epoch": 0.92, "learning_rate": 1.6335071707953064e-05, "loss": 0.4225, "step": 22500 }, { "epoch": 0.94, "learning_rate": 1.6253585397653195e-05, "loss": 0.4264, "step": 23000 }, { "epoch": 0.96, "learning_rate": 1.6172099087353327e-05, "loss": 0.4079, "step": 23500 }, { "epoch": 0.98, "learning_rate": 1.6090612777053454e-05, "loss": 0.4117, "step": 24000 }, { "epoch": 1.0, "learning_rate": 1.6009126466753586e-05, "loss": 0.418, "step": 24500 }, { "epoch": 1.02, "learning_rate": 1.5927640156453717e-05, "loss": 0.3114, "step": 25000 }, { "epoch": 1.04, "learning_rate": 1.584631681877445e-05, "loss": 0.3058, "step": 25500 }, { "epoch": 1.06, "learning_rate": 1.576515645371578e-05, "loss": 0.3029, "step": 26000 }, { "epoch": 1.08, "learning_rate": 1.5683833116036508e-05, "loss": 0.2873, "step": 26500 }, { "epoch": 1.1, "learning_rate": 1.5602346805736636e-05, "loss": 0.2856, "step": 27000 }, { "epoch": 1.12, "learning_rate": 1.5520860495436767e-05, "loss": 0.2867, "step": 27500 }, { "epoch": 1.14, "learning_rate": 1.5439374185136898e-05, "loss": 0.2954, "step": 28000 }, { "epoch": 1.16, "learning_rate": 1.535788787483703e-05, "loss": 0.2994, "step": 28500 }, { "epoch": 1.18, "learning_rate": 1.527640156453716e-05, "loss": 0.2925, "step": 29000 }, { "epoch": 1.2, "learning_rate": 1.519491525423729e-05, "loss": 0.2886, "step": 29500 }, { "epoch": 1.22, "learning_rate": 1.5113428943937421e-05, "loss": 0.3094, "step": 30000 }, { "epoch": 1.24, "learning_rate": 1.503194263363755e-05, "loss": 0.3007, "step": 30500 }, { "epoch": 1.26, "learning_rate": 1.495045632333768e-05, "loss": 0.2992, "step": 31000 }, { "epoch": 1.28, "learning_rate": 1.4868970013037811e-05, "loss": 0.2977, "step": 31500 }, { "epoch": 1.3, "learning_rate": 1.4787483702737941e-05, "loss": 0.3049, "step": 32000 }, { "epoch": 1.32, "learning_rate": 1.4705997392438072e-05, "loss": 0.3042, "step": 32500 }, { "epoch": 1.34, "learning_rate": 1.4624511082138202e-05, "loss": 0.2967, "step": 33000 }, { "epoch": 1.36, "learning_rate": 1.4543187744458934e-05, "loss": 0.3003, "step": 33500 }, { "epoch": 1.39, "learning_rate": 1.4461701434159063e-05, "loss": 0.3043, "step": 34000 }, { "epoch": 1.41, "learning_rate": 1.4380215123859193e-05, "loss": 0.2887, "step": 34500 }, { "epoch": 1.43, "learning_rate": 1.4298728813559324e-05, "loss": 0.2969, "step": 35000 }, { "epoch": 1.45, "learning_rate": 1.4217405475880053e-05, "loss": 0.2943, "step": 35500 }, { "epoch": 1.47, "learning_rate": 1.4135919165580182e-05, "loss": 0.302, "step": 36000 }, { "epoch": 1.49, "learning_rate": 1.4054432855280314e-05, "loss": 0.2866, "step": 36500 }, { "epoch": 1.51, "learning_rate": 1.3972946544980443e-05, "loss": 0.3001, "step": 37000 }, { "epoch": 1.53, "learning_rate": 1.3891460234680574e-05, "loss": 0.3074, "step": 37500 }, { "epoch": 1.55, "learning_rate": 1.3809973924380706e-05, "loss": 0.3015, "step": 38000 }, { "epoch": 1.57, "learning_rate": 1.3728487614080837e-05, "loss": 0.292, "step": 38500 }, { "epoch": 1.59, "learning_rate": 1.3647001303780966e-05, "loss": 0.3036, "step": 39000 }, { "epoch": 1.61, "learning_rate": 1.3565677966101695e-05, "loss": 0.2968, "step": 39500 }, { "epoch": 1.63, "learning_rate": 1.3484354628422427e-05, "loss": 0.2942, "step": 40000 }, { "epoch": 1.65, "learning_rate": 1.3402868318122557e-05, "loss": 0.29, "step": 40500 }, { "epoch": 1.67, "learning_rate": 1.3321382007822688e-05, "loss": 0.2974, "step": 41000 }, { "epoch": 1.69, "learning_rate": 1.3239895697522818e-05, "loss": 0.2956, "step": 41500 }, { "epoch": 1.71, "learning_rate": 1.3158409387222949e-05, "loss": 0.2975, "step": 42000 }, { "epoch": 1.73, "learning_rate": 1.3076923076923078e-05, "loss": 0.2974, "step": 42500 }, { "epoch": 1.75, "learning_rate": 1.2995436766623208e-05, "loss": 0.2985, "step": 43000 }, { "epoch": 1.77, "learning_rate": 1.2913950456323339e-05, "loss": 0.2932, "step": 43500 }, { "epoch": 1.79, "learning_rate": 1.2832464146023468e-05, "loss": 0.2953, "step": 44000 }, { "epoch": 1.81, "learning_rate": 1.27511408083442e-05, "loss": 0.299, "step": 44500 }, { "epoch": 1.83, "learning_rate": 1.266965449804433e-05, "loss": 0.284, "step": 45000 }, { "epoch": 1.85, "learning_rate": 1.2588168187744461e-05, "loss": 0.2934, "step": 45500 }, { "epoch": 1.87, "learning_rate": 1.2506681877444591e-05, "loss": 0.2926, "step": 46000 }, { "epoch": 1.89, "learning_rate": 1.242519556714472e-05, "loss": 0.3106, "step": 46500 }, { "epoch": 1.91, "learning_rate": 1.2343709256844852e-05, "loss": 0.3061, "step": 47000 }, { "epoch": 1.94, "learning_rate": 1.2262222946544981e-05, "loss": 0.2924, "step": 47500 }, { "epoch": 1.96, "learning_rate": 1.2180736636245112e-05, "loss": 0.3064, "step": 48000 }, { "epoch": 1.98, "learning_rate": 1.2099250325945242e-05, "loss": 0.293, "step": 48500 }, { "epoch": 2.0, "learning_rate": 1.2017926988265974e-05, "loss": 0.2838, "step": 49000 }, { "epoch": 2.02, "learning_rate": 1.1936440677966104e-05, "loss": 0.2064, "step": 49500 }, { "epoch": 2.04, "learning_rate": 1.1855117340286832e-05, "loss": 0.2013, "step": 50000 }, { "epoch": 2.06, "learning_rate": 1.1773631029986962e-05, "loss": 0.1839, "step": 50500 }, { "epoch": 2.08, "learning_rate": 1.1692144719687093e-05, "loss": 0.2017, "step": 51000 }, { "epoch": 2.1, "learning_rate": 1.1610658409387223e-05, "loss": 0.1932, "step": 51500 }, { "epoch": 2.12, "learning_rate": 1.1529335071707955e-05, "loss": 0.1967, "step": 52000 }, { "epoch": 2.14, "learning_rate": 1.1447848761408084e-05, "loss": 0.1894, "step": 52500 }, { "epoch": 2.16, "learning_rate": 1.1366362451108216e-05, "loss": 0.1929, "step": 53000 }, { "epoch": 2.18, "learning_rate": 1.1284876140808345e-05, "loss": 0.194, "step": 53500 }, { "epoch": 2.2, "learning_rate": 1.1203552803129076e-05, "loss": 0.2006, "step": 54000 }, { "epoch": 2.22, "learning_rate": 1.1122066492829207e-05, "loss": 0.2096, "step": 54500 }, { "epoch": 2.24, "learning_rate": 1.1040580182529336e-05, "loss": 0.2047, "step": 55000 }, { "epoch": 2.26, "learning_rate": 1.0959093872229468e-05, "loss": 0.2006, "step": 55500 }, { "epoch": 2.28, "learning_rate": 1.0877607561929597e-05, "loss": 0.209, "step": 56000 }, { "epoch": 2.3, "learning_rate": 1.0796121251629728e-05, "loss": 0.201, "step": 56500 }, { "epoch": 2.32, "learning_rate": 1.0714797913950457e-05, "loss": 0.1952, "step": 57000 }, { "epoch": 2.34, "learning_rate": 1.0633474576271188e-05, "loss": 0.1966, "step": 57500 }, { "epoch": 2.36, "learning_rate": 1.0551988265971317e-05, "loss": 0.1941, "step": 58000 }, { "epoch": 2.38, "learning_rate": 1.0470501955671448e-05, "loss": 0.1978, "step": 58500 }, { "epoch": 2.4, "learning_rate": 1.0389015645371578e-05, "loss": 0.1953, "step": 59000 }, { "epoch": 2.42, "learning_rate": 1.030752933507171e-05, "loss": 0.1943, "step": 59500 }, { "epoch": 2.44, "learning_rate": 1.0226043024771839e-05, "loss": 0.207, "step": 60000 }, { "epoch": 2.46, "learning_rate": 1.014455671447197e-05, "loss": 0.2075, "step": 60500 }, { "epoch": 2.49, "learning_rate": 1.00630704041721e-05, "loss": 0.197, "step": 61000 }, { "epoch": 2.51, "learning_rate": 9.98158409387223e-06, "loss": 0.2191, "step": 61500 }, { "epoch": 2.53, "learning_rate": 9.900260756192961e-06, "loss": 0.195, "step": 62000 }, { "epoch": 2.55, "learning_rate": 9.81877444589309e-06, "loss": 0.2089, "step": 62500 }, { "epoch": 2.57, "learning_rate": 9.737288135593222e-06, "loss": 0.2116, "step": 63000 }, { "epoch": 2.59, "learning_rate": 9.655801825293351e-06, "loss": 0.2021, "step": 63500 }, { "epoch": 2.61, "learning_rate": 9.574315514993483e-06, "loss": 0.2064, "step": 64000 }, { "epoch": 2.63, "learning_rate": 9.492829204693612e-06, "loss": 0.201, "step": 64500 }, { "epoch": 2.65, "learning_rate": 9.411342894393743e-06, "loss": 0.217, "step": 65000 }, { "epoch": 2.67, "learning_rate": 9.329856584093873e-06, "loss": 0.2103, "step": 65500 }, { "epoch": 2.69, "learning_rate": 9.248533246414603e-06, "loss": 0.2061, "step": 66000 }, { "epoch": 2.71, "learning_rate": 9.167046936114733e-06, "loss": 0.2004, "step": 66500 }, { "epoch": 2.73, "learning_rate": 9.085560625814864e-06, "loss": 0.1984, "step": 67000 }, { "epoch": 2.75, "learning_rate": 9.004074315514995e-06, "loss": 0.2145, "step": 67500 }, { "epoch": 2.77, "learning_rate": 8.922588005215125e-06, "loss": 0.2029, "step": 68000 }, { "epoch": 2.79, "learning_rate": 8.841264667535855e-06, "loss": 0.2132, "step": 68500 }, { "epoch": 2.81, "learning_rate": 8.759941329856584e-06, "loss": 0.2046, "step": 69000 }, { "epoch": 2.83, "learning_rate": 8.678455019556715e-06, "loss": 0.2029, "step": 69500 }, { "epoch": 2.85, "learning_rate": 8.596968709256845e-06, "loss": 0.2072, "step": 70000 }, { "epoch": 2.87, "learning_rate": 8.515482398956976e-06, "loss": 0.2068, "step": 70500 }, { "epoch": 2.89, "learning_rate": 8.434159061277705e-06, "loss": 0.2001, "step": 71000 }, { "epoch": 2.91, "learning_rate": 8.352672750977836e-06, "loss": 0.213, "step": 71500 }, { "epoch": 2.93, "learning_rate": 8.271186440677966e-06, "loss": 0.2027, "step": 72000 }, { "epoch": 2.95, "learning_rate": 8.189700130378097e-06, "loss": 0.2109, "step": 72500 }, { "epoch": 2.97, "learning_rate": 8.108213820078228e-06, "loss": 0.1989, "step": 73000 }, { "epoch": 2.99, "learning_rate": 8.026727509778358e-06, "loss": 0.2039, "step": 73500 }, { "epoch": 3.01, "learning_rate": 7.945241199478489e-06, "loss": 0.1444, "step": 74000 }, { "epoch": 3.04, "learning_rate": 7.863917861799218e-06, "loss": 0.1375, "step": 74500 }, { "epoch": 3.06, "learning_rate": 7.782431551499349e-06, "loss": 0.1237, "step": 75000 }, { "epoch": 3.08, "learning_rate": 7.700945241199478e-06, "loss": 0.1369, "step": 75500 }, { "epoch": 3.1, "learning_rate": 7.6194589308996095e-06, "loss": 0.1432, "step": 76000 }, { "epoch": 3.12, "learning_rate": 7.53797262059974e-06, "loss": 0.1283, "step": 76500 }, { "epoch": 3.14, "learning_rate": 7.456486310299869e-06, "loss": 0.1381, "step": 77000 }, { "epoch": 3.16, "learning_rate": 7.375000000000001e-06, "loss": 0.13, "step": 77500 }, { "epoch": 3.18, "learning_rate": 7.293513689700131e-06, "loss": 0.1329, "step": 78000 }, { "epoch": 3.2, "learning_rate": 7.212027379400261e-06, "loss": 0.1315, "step": 78500 }, { "epoch": 3.22, "learning_rate": 7.130541069100392e-06, "loss": 0.1214, "step": 79000 }, { "epoch": 3.24, "learning_rate": 7.0493807040417216e-06, "loss": 0.141, "step": 79500 }, { "epoch": 3.26, "learning_rate": 6.967894393741852e-06, "loss": 0.1485, "step": 80000 }, { "epoch": 3.28, "learning_rate": 6.886408083441982e-06, "loss": 0.1301, "step": 80500 }, { "epoch": 3.3, "learning_rate": 6.804921773142113e-06, "loss": 0.1519, "step": 81000 }, { "epoch": 3.32, "learning_rate": 6.723435462842243e-06, "loss": 0.1536, "step": 81500 }, { "epoch": 3.34, "learning_rate": 6.641949152542373e-06, "loss": 0.1207, "step": 82000 }, { "epoch": 3.36, "learning_rate": 6.5604628422425045e-06, "loss": 0.1455, "step": 82500 }, { "epoch": 3.38, "learning_rate": 6.478976531942635e-06, "loss": 0.1414, "step": 83000 }, { "epoch": 3.4, "learning_rate": 6.397816166883965e-06, "loss": 0.1283, "step": 83500 }, { "epoch": 3.42, "learning_rate": 6.316329856584094e-06, "loss": 0.1487, "step": 84000 }, { "epoch": 3.44, "learning_rate": 6.234843546284225e-06, "loss": 0.1312, "step": 84500 }, { "epoch": 3.46, "learning_rate": 6.153357235984355e-06, "loss": 0.1397, "step": 85000 }, { "epoch": 3.48, "learning_rate": 6.071870925684485e-06, "loss": 0.1173, "step": 85500 }, { "epoch": 3.5, "learning_rate": 5.990547588005216e-06, "loss": 0.1505, "step": 86000 }, { "epoch": 3.52, "learning_rate": 5.9090612777053454e-06, "loss": 0.1342, "step": 86500 }, { "epoch": 3.54, "learning_rate": 5.827574967405476e-06, "loss": 0.1423, "step": 87000 }, { "epoch": 3.57, "learning_rate": 5.746251629726206e-06, "loss": 0.1337, "step": 87500 }, { "epoch": 3.59, "learning_rate": 5.664765319426337e-06, "loss": 0.1253, "step": 88000 }, { "epoch": 3.61, "learning_rate": 5.583279009126468e-06, "loss": 0.124, "step": 88500 }, { "epoch": 3.63, "learning_rate": 5.501792698826598e-06, "loss": 0.1445, "step": 89000 }, { "epoch": 3.65, "learning_rate": 5.420306388526729e-06, "loss": 0.128, "step": 89500 }, { "epoch": 3.67, "learning_rate": 5.338820078226858e-06, "loss": 0.1359, "step": 90000 }, { "epoch": 3.69, "learning_rate": 5.257496740547589e-06, "loss": 0.1333, "step": 90500 }, { "epoch": 3.71, "learning_rate": 5.176010430247719e-06, "loss": 0.1304, "step": 91000 }, { "epoch": 3.73, "learning_rate": 5.094524119947849e-06, "loss": 0.1459, "step": 91500 }, { "epoch": 3.75, "learning_rate": 5.013037809647979e-06, "loss": 0.1265, "step": 92000 }, { "epoch": 3.77, "learning_rate": 4.93155149934811e-06, "loss": 0.1316, "step": 92500 }, { "epoch": 3.79, "learning_rate": 4.85006518904824e-06, "loss": 0.1418, "step": 93000 }, { "epoch": 3.81, "learning_rate": 4.768578878748371e-06, "loss": 0.1419, "step": 93500 }, { "epoch": 3.83, "learning_rate": 4.687092568448501e-06, "loss": 0.137, "step": 94000 }, { "epoch": 3.85, "learning_rate": 4.6056062581486315e-06, "loss": 0.133, "step": 94500 }, { "epoch": 3.87, "learning_rate": 4.524282920469362e-06, "loss": 0.1423, "step": 95000 }, { "epoch": 3.89, "learning_rate": 4.4427966101694915e-06, "loss": 0.1395, "step": 95500 }, { "epoch": 3.91, "learning_rate": 4.361310299869622e-06, "loss": 0.1381, "step": 96000 }, { "epoch": 3.93, "learning_rate": 4.279823989569752e-06, "loss": 0.1403, "step": 96500 }, { "epoch": 3.95, "learning_rate": 4.198500651890483e-06, "loss": 0.119, "step": 97000 }, { "epoch": 3.97, "learning_rate": 4.117014341590613e-06, "loss": 0.1339, "step": 97500 }, { "epoch": 3.99, "learning_rate": 4.0355280312907435e-06, "loss": 0.1317, "step": 98000 }, { "epoch": 4.01, "learning_rate": 3.954041720990874e-06, "loss": 0.0799, "step": 98500 }, { "epoch": 4.03, "learning_rate": 3.872555410691004e-06, "loss": 0.0724, "step": 99000 }, { "epoch": 4.05, "learning_rate": 3.791069100391135e-06, "loss": 0.088, "step": 99500 }, { "epoch": 4.07, "learning_rate": 3.709582790091265e-06, "loss": 0.0731, "step": 100000 }, { "epoch": 4.09, "learning_rate": 3.6280964797913952e-06, "loss": 0.084, "step": 100500 }, { "epoch": 4.12, "learning_rate": 3.5467731421121254e-06, "loss": 0.0797, "step": 101000 }, { "epoch": 4.14, "learning_rate": 3.4654498044328556e-06, "loss": 0.0703, "step": 101500 }, { "epoch": 4.16, "learning_rate": 3.383963494132986e-06, "loss": 0.0748, "step": 102000 }, { "epoch": 4.18, "learning_rate": 3.3024771838331167e-06, "loss": 0.0757, "step": 102500 }, { "epoch": 4.2, "learning_rate": 3.2209908735332466e-06, "loss": 0.0696, "step": 103000 }, { "epoch": 4.22, "learning_rate": 3.139504563233377e-06, "loss": 0.0763, "step": 103500 }, { "epoch": 4.24, "learning_rate": 3.0580182529335073e-06, "loss": 0.0723, "step": 104000 }, { "epoch": 4.26, "learning_rate": 2.9765319426336376e-06, "loss": 0.085, "step": 104500 }, { "epoch": 4.28, "learning_rate": 2.8950456323337684e-06, "loss": 0.073, "step": 105000 }, { "epoch": 4.3, "learning_rate": 2.8135593220338988e-06, "loss": 0.0858, "step": 105500 }, { "epoch": 4.32, "learning_rate": 2.7320730117340287e-06, "loss": 0.0784, "step": 106000 }, { "epoch": 4.34, "learning_rate": 2.650749674054759e-06, "loss": 0.0853, "step": 106500 }, { "epoch": 4.36, "learning_rate": 2.569263363754889e-06, "loss": 0.0741, "step": 107000 }, { "epoch": 4.38, "learning_rate": 2.4877770534550195e-06, "loss": 0.0772, "step": 107500 }, { "epoch": 4.4, "learning_rate": 2.4062907431551503e-06, "loss": 0.0737, "step": 108000 }, { "epoch": 4.42, "learning_rate": 2.3248044328552807e-06, "loss": 0.0636, "step": 108500 }, { "epoch": 4.44, "learning_rate": 2.2434810951760104e-06, "loss": 0.0795, "step": 109000 }, { "epoch": 4.46, "learning_rate": 2.161994784876141e-06, "loss": 0.0825, "step": 109500 }, { "epoch": 4.48, "learning_rate": 2.0805084745762715e-06, "loss": 0.0755, "step": 110000 }, { "epoch": 4.5, "learning_rate": 1.999022164276402e-06, "loss": 0.0691, "step": 110500 }, { "epoch": 4.52, "learning_rate": 1.917535853976532e-06, "loss": 0.0892, "step": 111000 }, { "epoch": 4.54, "learning_rate": 1.8360495436766626e-06, "loss": 0.0667, "step": 111500 }, { "epoch": 4.56, "learning_rate": 1.754563233376793e-06, "loss": 0.0884, "step": 112000 }, { "epoch": 4.58, "learning_rate": 1.673076923076923e-06, "loss": 0.0826, "step": 112500 }, { "epoch": 4.6, "learning_rate": 1.591916558018253e-06, "loss": 0.0731, "step": 113000 }, { "epoch": 4.62, "learning_rate": 1.5104302477183836e-06, "loss": 0.076, "step": 113500 }, { "epoch": 4.64, "learning_rate": 1.4289439374185137e-06, "loss": 0.0752, "step": 114000 }, { "epoch": 4.67, "learning_rate": 1.3474576271186443e-06, "loss": 0.0947, "step": 114500 }, { "epoch": 4.69, "learning_rate": 1.2661342894393744e-06, "loss": 0.0617, "step": 115000 }, { "epoch": 4.71, "learning_rate": 1.1846479791395046e-06, "loss": 0.0826, "step": 115500 }, { "epoch": 4.73, "learning_rate": 1.1033246414602347e-06, "loss": 0.0745, "step": 116000 }, { "epoch": 4.75, "learning_rate": 1.021838331160365e-06, "loss": 0.0867, "step": 116500 }, { "epoch": 4.77, "learning_rate": 9.403520208604955e-07, "loss": 0.0788, "step": 117000 }, { "epoch": 4.79, "learning_rate": 8.588657105606259e-07, "loss": 0.0732, "step": 117500 }, { "epoch": 4.81, "learning_rate": 7.773794002607562e-07, "loss": 0.0644, "step": 118000 }, { "epoch": 4.83, "learning_rate": 6.960560625814864e-07, "loss": 0.0745, "step": 118500 }, { "epoch": 4.85, "learning_rate": 6.145697522816167e-07, "loss": 0.0761, "step": 119000 }, { "epoch": 4.87, "learning_rate": 5.330834419817471e-07, "loss": 0.0864, "step": 119500 }, { "epoch": 4.89, "learning_rate": 4.5159713168187747e-07, "loss": 0.0647, "step": 120000 }, { "epoch": 4.91, "learning_rate": 3.7011082138200787e-07, "loss": 0.0679, "step": 120500 }, { "epoch": 4.93, "learning_rate": 2.886245110821382e-07, "loss": 0.08, "step": 121000 }, { "epoch": 4.95, "learning_rate": 2.071382007822686e-07, "loss": 0.085, "step": 121500 }, { "epoch": 4.97, "learning_rate": 1.258148631029987e-07, "loss": 0.0764, "step": 122000 }, { "epoch": 4.99, "learning_rate": 4.432855280312908e-08, "loss": 0.0787, "step": 122500 }, { "epoch": 5.0, "step": 122720, "total_flos": 4.574656834771123e+17, "train_loss": 0.23671778235317364, "train_runtime": 11417.9442, "train_samples_per_second": 171.967, "train_steps_per_second": 10.748 } ], "max_steps": 122720, "num_train_epochs": 5, "total_flos": 4.574656834771123e+17, "trial_name": null, "trial_params": null }