{ "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": 0.87, "step": 500 }, { "epoch": 0.04, "learning_rate": 1.983751629726206e-05, "loss": 0.7054, "step": 1000 }, { "epoch": 0.06, "learning_rate": 1.9756029986962192e-05, "loss": 0.661, "step": 1500 }, { "epoch": 0.08, "learning_rate": 1.9674543676662323e-05, "loss": 0.6353, "step": 2000 }, { "epoch": 0.1, "learning_rate": 1.9593057366362454e-05, "loss": 0.626, "step": 2500 }, { "epoch": 0.12, "learning_rate": 1.9511571056062585e-05, "loss": 0.6097, "step": 3000 }, { "epoch": 0.14, "learning_rate": 1.9430084745762713e-05, "loss": 0.5873, "step": 3500 }, { "epoch": 0.16, "learning_rate": 1.9348598435462844e-05, "loss": 0.6014, "step": 4000 }, { "epoch": 0.18, "learning_rate": 1.9267275097783573e-05, "loss": 0.5693, "step": 4500 }, { "epoch": 0.2, "learning_rate": 1.9185951760104305e-05, "loss": 0.5796, "step": 5000 }, { "epoch": 0.22, "learning_rate": 1.9104465449804433e-05, "loss": 0.5727, "step": 5500 }, { "epoch": 0.24, "learning_rate": 1.9022979139504564e-05, "loss": 0.5765, "step": 6000 }, { "epoch": 0.26, "learning_rate": 1.8941492829204696e-05, "loss": 0.5676, "step": 6500 }, { "epoch": 0.29, "learning_rate": 1.8860495436766623e-05, "loss": 0.561, "step": 7000 }, { "epoch": 0.31, "learning_rate": 1.8779009126466758e-05, "loss": 0.55, "step": 7500 }, { "epoch": 0.33, "learning_rate": 1.8697522816166886e-05, "loss": 0.5409, "step": 8000 }, { "epoch": 0.35, "learning_rate": 1.8616036505867017e-05, "loss": 0.5422, "step": 8500 }, { "epoch": 0.37, "learning_rate": 1.8534550195567148e-05, "loss": 0.5615, "step": 9000 }, { "epoch": 0.39, "learning_rate": 1.8453063885267276e-05, "loss": 0.5539, "step": 9500 }, { "epoch": 0.41, "learning_rate": 1.8371577574967407e-05, "loss": 0.5484, "step": 10000 }, { "epoch": 0.43, "learning_rate": 1.8290091264667538e-05, "loss": 0.5325, "step": 10500 }, { "epoch": 0.45, "learning_rate": 1.820860495436767e-05, "loss": 0.5254, "step": 11000 }, { "epoch": 0.47, "learning_rate": 1.8127118644067797e-05, "loss": 0.5269, "step": 11500 }, { "epoch": 0.49, "learning_rate": 1.804563233376793e-05, "loss": 0.5443, "step": 12000 }, { "epoch": 0.51, "learning_rate": 1.796414602346806e-05, "loss": 0.5331, "step": 12500 }, { "epoch": 0.53, "learning_rate": 1.7882659713168187e-05, "loss": 0.5199, "step": 13000 }, { "epoch": 0.55, "learning_rate": 1.780117340286832e-05, "loss": 0.5298, "step": 13500 }, { "epoch": 0.57, "learning_rate": 1.771968709256845e-05, "loss": 0.5314, "step": 14000 }, { "epoch": 0.59, "learning_rate": 1.763820078226858e-05, "loss": 0.5179, "step": 14500 }, { "epoch": 0.61, "learning_rate": 1.755687744458931e-05, "loss": 0.5288, "step": 15000 }, { "epoch": 0.63, "learning_rate": 1.747539113428944e-05, "loss": 0.525, "step": 15500 }, { "epoch": 0.65, "learning_rate": 1.7393904823989572e-05, "loss": 0.5161, "step": 16000 }, { "epoch": 0.67, "learning_rate": 1.73124185136897e-05, "loss": 0.5243, "step": 16500 }, { "epoch": 0.69, "learning_rate": 1.723093220338983e-05, "loss": 0.5145, "step": 17000 }, { "epoch": 0.71, "learning_rate": 1.7149445893089963e-05, "loss": 0.5353, "step": 17500 }, { "epoch": 0.73, "learning_rate": 1.7068285528031293e-05, "loss": 0.5261, "step": 18000 }, { "epoch": 0.75, "learning_rate": 1.6986799217731424e-05, "loss": 0.5178, "step": 18500 }, { "epoch": 0.77, "learning_rate": 1.690531290743155e-05, "loss": 0.5108, "step": 19000 }, { "epoch": 0.79, "learning_rate": 1.6823826597131683e-05, "loss": 0.5169, "step": 19500 }, { "epoch": 0.81, "learning_rate": 1.6742340286831814e-05, "loss": 0.4949, "step": 20000 }, { "epoch": 0.84, "learning_rate": 1.6660853976531942e-05, "loss": 0.5012, "step": 20500 }, { "epoch": 0.86, "learning_rate": 1.6579367666232073e-05, "loss": 0.5169, "step": 21000 }, { "epoch": 0.88, "learning_rate": 1.6497881355932204e-05, "loss": 0.5201, "step": 21500 }, { "epoch": 0.9, "learning_rate": 1.6416395045632335e-05, "loss": 0.4954, "step": 22000 }, { "epoch": 0.92, "learning_rate": 1.6334908735332466e-05, "loss": 0.5005, "step": 22500 }, { "epoch": 0.94, "learning_rate": 1.6253422425032598e-05, "loss": 0.5063, "step": 23000 }, { "epoch": 0.96, "learning_rate": 1.6172099087353327e-05, "loss": 0.4957, "step": 23500 }, { "epoch": 0.98, "learning_rate": 1.6090612777053454e-05, "loss": 0.4903, "step": 24000 }, { "epoch": 1.0, "learning_rate": 1.6009126466753586e-05, "loss": 0.4883, "step": 24500 }, { "epoch": 1.02, "learning_rate": 1.5927640156453717e-05, "loss": 0.3715, "step": 25000 }, { "epoch": 1.04, "learning_rate": 1.5846153846153848e-05, "loss": 0.3628, "step": 25500 }, { "epoch": 1.06, "learning_rate": 1.5764830508474577e-05, "loss": 0.3589, "step": 26000 }, { "epoch": 1.08, "learning_rate": 1.5683344198174708e-05, "loss": 0.3532, "step": 26500 }, { "epoch": 1.1, "learning_rate": 1.560185788787484e-05, "loss": 0.3578, "step": 27000 }, { "epoch": 1.12, "learning_rate": 1.5520371577574967e-05, "loss": 0.3897, "step": 27500 }, { "epoch": 1.14, "learning_rate": 1.54390482398957e-05, "loss": 0.3671, "step": 28000 }, { "epoch": 1.16, "learning_rate": 1.535756192959583e-05, "loss": 0.3505, "step": 28500 }, { "epoch": 1.18, "learning_rate": 1.5276075619295962e-05, "loss": 0.3782, "step": 29000 }, { "epoch": 1.2, "learning_rate": 1.5194589308996091e-05, "loss": 0.3765, "step": 29500 }, { "epoch": 1.22, "learning_rate": 1.511310299869622e-05, "loss": 0.3658, "step": 30000 }, { "epoch": 1.24, "learning_rate": 1.503161668839635e-05, "loss": 0.3752, "step": 30500 }, { "epoch": 1.26, "learning_rate": 1.4950130378096481e-05, "loss": 0.3762, "step": 31000 }, { "epoch": 1.28, "learning_rate": 1.4868644067796611e-05, "loss": 0.3742, "step": 31500 }, { "epoch": 1.3, "learning_rate": 1.4787157757496742e-05, "loss": 0.3748, "step": 32000 }, { "epoch": 1.32, "learning_rate": 1.4705834419817473e-05, "loss": 0.3628, "step": 32500 }, { "epoch": 1.34, "learning_rate": 1.4624348109517604e-05, "loss": 0.3635, "step": 33000 }, { "epoch": 1.36, "learning_rate": 1.4542861799217733e-05, "loss": 0.3824, "step": 33500 }, { "epoch": 1.39, "learning_rate": 1.4461375488917863e-05, "loss": 0.3639, "step": 34000 }, { "epoch": 1.41, "learning_rate": 1.4380215123859193e-05, "loss": 0.3632, "step": 34500 }, { "epoch": 1.43, "learning_rate": 1.4298728813559324e-05, "loss": 0.3739, "step": 35000 }, { "epoch": 1.45, "learning_rate": 1.4217242503259454e-05, "loss": 0.3871, "step": 35500 }, { "epoch": 1.47, "learning_rate": 1.4135756192959585e-05, "loss": 0.3843, "step": 36000 }, { "epoch": 1.49, "learning_rate": 1.4054269882659714e-05, "loss": 0.3855, "step": 36500 }, { "epoch": 1.51, "learning_rate": 1.3972783572359845e-05, "loss": 0.3836, "step": 37000 }, { "epoch": 1.53, "learning_rate": 1.3891297262059975e-05, "loss": 0.3762, "step": 37500 }, { "epoch": 1.55, "learning_rate": 1.3809973924380706e-05, "loss": 0.3816, "step": 38000 }, { "epoch": 1.57, "learning_rate": 1.3728487614080837e-05, "loss": 0.3808, "step": 38500 }, { "epoch": 1.59, "learning_rate": 1.3647001303780966e-05, "loss": 0.3672, "step": 39000 }, { "epoch": 1.61, "learning_rate": 1.3565514993481097e-05, "loss": 0.3776, "step": 39500 }, { "epoch": 1.63, "learning_rate": 1.3484191655801826e-05, "loss": 0.3735, "step": 40000 }, { "epoch": 1.65, "learning_rate": 1.3402868318122557e-05, "loss": 0.3684, "step": 40500 }, { "epoch": 1.67, "learning_rate": 1.3321382007822688e-05, "loss": 0.3794, "step": 41000 }, { "epoch": 1.69, "learning_rate": 1.3239895697522818e-05, "loss": 0.3877, "step": 41500 }, { "epoch": 1.71, "learning_rate": 1.3158409387222949e-05, "loss": 0.369, "step": 42000 }, { "epoch": 1.73, "learning_rate": 1.3077249022164277e-05, "loss": 0.3797, "step": 42500 }, { "epoch": 1.75, "learning_rate": 1.2995762711864408e-05, "loss": 0.3814, "step": 43000 }, { "epoch": 1.77, "learning_rate": 1.2914276401564538e-05, "loss": 0.3805, "step": 43500 }, { "epoch": 1.79, "learning_rate": 1.2832790091264669e-05, "loss": 0.3756, "step": 44000 }, { "epoch": 1.81, "learning_rate": 1.2751303780964798e-05, "loss": 0.3884, "step": 44500 }, { "epoch": 1.83, "learning_rate": 1.266981747066493e-05, "loss": 0.3815, "step": 45000 }, { "epoch": 1.85, "learning_rate": 1.2588331160365059e-05, "loss": 0.3754, "step": 45500 }, { "epoch": 1.87, "learning_rate": 1.250684485006519e-05, "loss": 0.3738, "step": 46000 }, { "epoch": 1.89, "learning_rate": 1.242535853976532e-05, "loss": 0.3824, "step": 46500 }, { "epoch": 1.91, "learning_rate": 1.234387222946545e-05, "loss": 0.3752, "step": 47000 }, { "epoch": 1.94, "learning_rate": 1.226238591916558e-05, "loss": 0.3719, "step": 47500 }, { "epoch": 1.96, "learning_rate": 1.218089960886571e-05, "loss": 0.3749, "step": 48000 }, { "epoch": 1.98, "learning_rate": 1.2099413298565843e-05, "loss": 0.3674, "step": 48500 }, { "epoch": 2.0, "learning_rate": 1.2017926988265974e-05, "loss": 0.3789, "step": 49000 }, { "epoch": 2.02, "learning_rate": 1.1936440677966104e-05, "loss": 0.2588, "step": 49500 }, { "epoch": 2.04, "learning_rate": 1.1855117340286832e-05, "loss": 0.2309, "step": 50000 }, { "epoch": 2.06, "learning_rate": 1.1773794002607563e-05, "loss": 0.2401, "step": 50500 }, { "epoch": 2.08, "learning_rate": 1.1692470664928292e-05, "loss": 0.2278, "step": 51000 }, { "epoch": 2.1, "learning_rate": 1.1610984354628423e-05, "loss": 0.234, "step": 51500 }, { "epoch": 2.12, "learning_rate": 1.1529498044328553e-05, "loss": 0.2319, "step": 52000 }, { "epoch": 2.14, "learning_rate": 1.1448011734028684e-05, "loss": 0.2405, "step": 52500 }, { "epoch": 2.16, "learning_rate": 1.1366525423728813e-05, "loss": 0.2346, "step": 53000 }, { "epoch": 2.18, "learning_rate": 1.1285039113428945e-05, "loss": 0.2382, "step": 53500 }, { "epoch": 2.2, "learning_rate": 1.1203552803129076e-05, "loss": 0.2455, "step": 54000 }, { "epoch": 2.22, "learning_rate": 1.1122066492829207e-05, "loss": 0.2385, "step": 54500 }, { "epoch": 2.24, "learning_rate": 1.1040580182529336e-05, "loss": 0.2447, "step": 55000 }, { "epoch": 2.26, "learning_rate": 1.0959093872229468e-05, "loss": 0.244, "step": 55500 }, { "epoch": 2.28, "learning_rate": 1.0877607561929597e-05, "loss": 0.2324, "step": 56000 }, { "epoch": 2.3, "learning_rate": 1.0796121251629728e-05, "loss": 0.2533, "step": 56500 }, { "epoch": 2.32, "learning_rate": 1.0714634941329858e-05, "loss": 0.2406, "step": 57000 }, { "epoch": 2.34, "learning_rate": 1.0633311603650587e-05, "loss": 0.2474, "step": 57500 }, { "epoch": 2.36, "learning_rate": 1.0551825293350716e-05, "loss": 0.2397, "step": 58000 }, { "epoch": 2.38, "learning_rate": 1.0470338983050847e-05, "loss": 0.2433, "step": 58500 }, { "epoch": 2.4, "learning_rate": 1.038885267275098e-05, "loss": 0.2568, "step": 59000 }, { "epoch": 2.42, "learning_rate": 1.030752933507171e-05, "loss": 0.227, "step": 59500 }, { "epoch": 2.44, "learning_rate": 1.0226043024771839e-05, "loss": 0.2339, "step": 60000 }, { "epoch": 2.46, "learning_rate": 1.014455671447197e-05, "loss": 0.2414, "step": 60500 }, { "epoch": 2.49, "learning_rate": 1.00630704041721e-05, "loss": 0.2448, "step": 61000 }, { "epoch": 2.51, "learning_rate": 9.98158409387223e-06, "loss": 0.255, "step": 61500 }, { "epoch": 2.53, "learning_rate": 9.900097783572362e-06, "loss": 0.2332, "step": 62000 }, { "epoch": 2.55, "learning_rate": 9.81877444589309e-06, "loss": 0.2453, "step": 62500 }, { "epoch": 2.57, "learning_rate": 9.737288135593222e-06, "loss": 0.2319, "step": 63000 }, { "epoch": 2.59, "learning_rate": 9.655801825293351e-06, "loss": 0.2517, "step": 63500 }, { "epoch": 2.61, "learning_rate": 9.574315514993483e-06, "loss": 0.2449, "step": 64000 }, { "epoch": 2.63, "learning_rate": 9.492992177314211e-06, "loss": 0.247, "step": 64500 }, { "epoch": 2.65, "learning_rate": 9.411505867014343e-06, "loss": 0.2441, "step": 65000 }, { "epoch": 2.67, "learning_rate": 9.330019556714472e-06, "loss": 0.241, "step": 65500 }, { "epoch": 2.69, "learning_rate": 9.248533246414603e-06, "loss": 0.2583, "step": 66000 }, { "epoch": 2.71, "learning_rate": 9.167209908735332e-06, "loss": 0.2497, "step": 66500 }, { "epoch": 2.73, "learning_rate": 9.085723598435463e-06, "loss": 0.2426, "step": 67000 }, { "epoch": 2.75, "learning_rate": 9.004237288135595e-06, "loss": 0.2529, "step": 67500 }, { "epoch": 2.77, "learning_rate": 8.922750977835724e-06, "loss": 0.2416, "step": 68000 }, { "epoch": 2.79, "learning_rate": 8.841427640156455e-06, "loss": 0.2468, "step": 68500 }, { "epoch": 2.81, "learning_rate": 8.759941329856584e-06, "loss": 0.2612, "step": 69000 }, { "epoch": 2.83, "learning_rate": 8.678455019556715e-06, "loss": 0.2496, "step": 69500 }, { "epoch": 2.85, "learning_rate": 8.596968709256845e-06, "loss": 0.2644, "step": 70000 }, { "epoch": 2.87, "learning_rate": 8.515482398956976e-06, "loss": 0.2314, "step": 70500 }, { "epoch": 2.89, "learning_rate": 8.433996088657106e-06, "loss": 0.2427, "step": 71000 }, { "epoch": 2.91, "learning_rate": 8.352672750977836e-06, "loss": 0.2529, "step": 71500 }, { "epoch": 2.93, "learning_rate": 8.271186440677966e-06, "loss": 0.2463, "step": 72000 }, { "epoch": 2.95, "learning_rate": 8.189700130378097e-06, "loss": 0.2446, "step": 72500 }, { "epoch": 2.97, "learning_rate": 8.108213820078228e-06, "loss": 0.2457, "step": 73000 }, { "epoch": 2.99, "learning_rate": 8.026727509778358e-06, "loss": 0.2435, "step": 73500 }, { "epoch": 3.01, "learning_rate": 7.945241199478489e-06, "loss": 0.1715, "step": 74000 }, { "epoch": 3.04, "learning_rate": 7.863917861799218e-06, "loss": 0.1444, "step": 74500 }, { "epoch": 3.06, "learning_rate": 7.782431551499349e-06, "loss": 0.1493, "step": 75000 }, { "epoch": 3.08, "learning_rate": 7.700945241199478e-06, "loss": 0.1546, "step": 75500 }, { "epoch": 3.1, "learning_rate": 7.6194589308996095e-06, "loss": 0.1532, "step": 76000 }, { "epoch": 3.12, "learning_rate": 7.53797262059974e-06, "loss": 0.1717, "step": 76500 }, { "epoch": 3.14, "learning_rate": 7.45664928292047e-06, "loss": 0.1646, "step": 77000 }, { "epoch": 3.16, "learning_rate": 7.375162972620601e-06, "loss": 0.1613, "step": 77500 }, { "epoch": 3.18, "learning_rate": 7.293676662320731e-06, "loss": 0.1695, "step": 78000 }, { "epoch": 3.2, "learning_rate": 7.2121903520208615e-06, "loss": 0.1655, "step": 78500 }, { "epoch": 3.22, "learning_rate": 7.130867014341591e-06, "loss": 0.1509, "step": 79000 }, { "epoch": 3.24, "learning_rate": 7.0493807040417216e-06, "loss": 0.1621, "step": 79500 }, { "epoch": 3.26, "learning_rate": 6.967894393741852e-06, "loss": 0.1589, "step": 80000 }, { "epoch": 3.28, "learning_rate": 6.886408083441982e-06, "loss": 0.1693, "step": 80500 }, { "epoch": 3.3, "learning_rate": 6.804921773142113e-06, "loss": 0.1702, "step": 81000 }, { "epoch": 3.32, "learning_rate": 6.723598435462842e-06, "loss": 0.1749, "step": 81500 }, { "epoch": 3.34, "learning_rate": 6.642112125162973e-06, "loss": 0.1603, "step": 82000 }, { "epoch": 3.36, "learning_rate": 6.560625814863103e-06, "loss": 0.1671, "step": 82500 }, { "epoch": 3.38, "learning_rate": 6.479139504563234e-06, "loss": 0.1714, "step": 83000 }, { "epoch": 3.4, "learning_rate": 6.397816166883965e-06, "loss": 0.1722, "step": 83500 }, { "epoch": 3.42, "learning_rate": 6.316329856584094e-06, "loss": 0.1686, "step": 84000 }, { "epoch": 3.44, "learning_rate": 6.234843546284225e-06, "loss": 0.1553, "step": 84500 }, { "epoch": 3.46, "learning_rate": 6.153357235984355e-06, "loss": 0.1584, "step": 85000 }, { "epoch": 3.48, "learning_rate": 6.072033898305086e-06, "loss": 0.1665, "step": 85500 }, { "epoch": 3.5, "learning_rate": 5.990547588005216e-06, "loss": 0.1607, "step": 86000 }, { "epoch": 3.52, "learning_rate": 5.9090612777053454e-06, "loss": 0.1748, "step": 86500 }, { "epoch": 3.54, "learning_rate": 5.827737940026076e-06, "loss": 0.1688, "step": 87000 }, { "epoch": 3.57, "learning_rate": 5.746251629726206e-06, "loss": 0.1681, "step": 87500 }, { "epoch": 3.59, "learning_rate": 5.664765319426337e-06, "loss": 0.171, "step": 88000 }, { "epoch": 3.61, "learning_rate": 5.583279009126468e-06, "loss": 0.1744, "step": 88500 }, { "epoch": 3.63, "learning_rate": 5.501792698826598e-06, "loss": 0.1648, "step": 89000 }, { "epoch": 3.65, "learning_rate": 5.420306388526729e-06, "loss": 0.1645, "step": 89500 }, { "epoch": 3.67, "learning_rate": 5.338820078226858e-06, "loss": 0.1518, "step": 90000 }, { "epoch": 3.69, "learning_rate": 5.2573337679269884e-06, "loss": 0.1715, "step": 90500 }, { "epoch": 3.71, "learning_rate": 5.175847457627119e-06, "loss": 0.1768, "step": 91000 }, { "epoch": 3.73, "learning_rate": 5.094524119947849e-06, "loss": 0.1655, "step": 91500 }, { "epoch": 3.75, "learning_rate": 5.013037809647979e-06, "loss": 0.1783, "step": 92000 }, { "epoch": 3.77, "learning_rate": 4.93155149934811e-06, "loss": 0.1721, "step": 92500 }, { "epoch": 3.79, "learning_rate": 4.85006518904824e-06, "loss": 0.1683, "step": 93000 }, { "epoch": 3.81, "learning_rate": 4.76874185136897e-06, "loss": 0.1725, "step": 93500 }, { "epoch": 3.83, "learning_rate": 4.6872555410691005e-06, "loss": 0.1744, "step": 94000 }, { "epoch": 3.85, "learning_rate": 4.605769230769231e-06, "loss": 0.1729, "step": 94500 }, { "epoch": 3.87, "learning_rate": 4.524282920469362e-06, "loss": 0.1642, "step": 95000 }, { "epoch": 3.89, "learning_rate": 4.442959582790092e-06, "loss": 0.1653, "step": 95500 }, { "epoch": 3.91, "learning_rate": 4.361473272490222e-06, "loss": 0.1914, "step": 96000 }, { "epoch": 3.93, "learning_rate": 4.2799869621903525e-06, "loss": 0.1626, "step": 96500 }, { "epoch": 3.95, "learning_rate": 4.198500651890483e-06, "loss": 0.1808, "step": 97000 }, { "epoch": 3.97, "learning_rate": 4.117014341590613e-06, "loss": 0.1617, "step": 97500 }, { "epoch": 3.99, "learning_rate": 4.035691003911344e-06, "loss": 0.1755, "step": 98000 }, { "epoch": 4.01, "learning_rate": 3.954204693611473e-06, "loss": 0.1162, "step": 98500 }, { "epoch": 4.03, "learning_rate": 3.872881355932204e-06, "loss": 0.0986, "step": 99000 }, { "epoch": 4.05, "learning_rate": 3.7913950456323338e-06, "loss": 0.0932, "step": 99500 }, { "epoch": 4.07, "learning_rate": 3.709908735332464e-06, "loss": 0.0952, "step": 100000 }, { "epoch": 4.09, "learning_rate": 3.628422425032595e-06, "loss": 0.0933, "step": 100500 }, { "epoch": 4.12, "learning_rate": 3.5469361147327252e-06, "loss": 0.1057, "step": 101000 }, { "epoch": 4.14, "learning_rate": 3.4654498044328556e-06, "loss": 0.0988, "step": 101500 }, { "epoch": 4.16, "learning_rate": 3.383963494132986e-06, "loss": 0.0894, "step": 102000 }, { "epoch": 4.18, "learning_rate": 3.3024771838331167e-06, "loss": 0.0959, "step": 102500 }, { "epoch": 4.2, "learning_rate": 3.221316818774446e-06, "loss": 0.1033, "step": 103000 }, { "epoch": 4.22, "learning_rate": 3.1398305084745766e-06, "loss": 0.1019, "step": 103500 }, { "epoch": 4.24, "learning_rate": 3.058344198174707e-06, "loss": 0.0942, "step": 104000 }, { "epoch": 4.26, "learning_rate": 2.9768578878748373e-06, "loss": 0.1, "step": 104500 }, { "epoch": 4.28, "learning_rate": 2.8955345501955674e-06, "loss": 0.1174, "step": 105000 }, { "epoch": 4.3, "learning_rate": 2.8140482398956974e-06, "loss": 0.0919, "step": 105500 }, { "epoch": 4.32, "learning_rate": 2.732561929595828e-06, "loss": 0.1201, "step": 106000 }, { "epoch": 4.34, "learning_rate": 2.6510756192959585e-06, "loss": 0.1025, "step": 106500 }, { "epoch": 4.36, "learning_rate": 2.5697522816166886e-06, "loss": 0.0948, "step": 107000 }, { "epoch": 4.38, "learning_rate": 2.488265971316819e-06, "loss": 0.092, "step": 107500 }, { "epoch": 4.4, "learning_rate": 2.4067796610169493e-06, "loss": 0.1014, "step": 108000 }, { "epoch": 4.42, "learning_rate": 2.3252933507170797e-06, "loss": 0.0994, "step": 108500 }, { "epoch": 4.44, "learning_rate": 2.24380704041721e-06, "loss": 0.0776, "step": 109000 }, { "epoch": 4.46, "learning_rate": 2.1623207301173404e-06, "loss": 0.0908, "step": 109500 }, { "epoch": 4.48, "learning_rate": 2.0808344198174707e-06, "loss": 0.0979, "step": 110000 }, { "epoch": 4.5, "learning_rate": 1.999511082138201e-06, "loss": 0.0908, "step": 110500 }, { "epoch": 4.52, "learning_rate": 1.9180247718383312e-06, "loss": 0.1044, "step": 111000 }, { "epoch": 4.54, "learning_rate": 1.8365384615384618e-06, "loss": 0.099, "step": 111500 }, { "epoch": 4.56, "learning_rate": 1.7550521512385921e-06, "loss": 0.0967, "step": 112000 }, { "epoch": 4.58, "learning_rate": 1.6735658409387223e-06, "loss": 0.0871, "step": 112500 }, { "epoch": 4.6, "learning_rate": 1.5922425032594526e-06, "loss": 0.101, "step": 113000 }, { "epoch": 4.62, "learning_rate": 1.510756192959583e-06, "loss": 0.0913, "step": 113500 }, { "epoch": 4.64, "learning_rate": 1.4292698826597131e-06, "loss": 0.0887, "step": 114000 }, { "epoch": 4.67, "learning_rate": 1.3477835723598437e-06, "loss": 0.089, "step": 114500 }, { "epoch": 4.69, "learning_rate": 1.266297262059974e-06, "loss": 0.0973, "step": 115000 }, { "epoch": 4.71, "learning_rate": 1.1848109517601044e-06, "loss": 0.0938, "step": 115500 }, { "epoch": 4.73, "learning_rate": 1.1033246414602347e-06, "loss": 0.1079, "step": 116000 }, { "epoch": 4.75, "learning_rate": 1.0220013037809649e-06, "loss": 0.1027, "step": 116500 }, { "epoch": 4.77, "learning_rate": 9.405149934810953e-07, "loss": 0.0991, "step": 117000 }, { "epoch": 4.79, "learning_rate": 8.590286831812256e-07, "loss": 0.1028, "step": 117500 }, { "epoch": 4.81, "learning_rate": 7.775423728813559e-07, "loss": 0.0997, "step": 118000 }, { "epoch": 4.83, "learning_rate": 6.962190352020862e-07, "loss": 0.0964, "step": 118500 }, { "epoch": 4.85, "learning_rate": 6.147327249022164e-07, "loss": 0.0954, "step": 119000 }, { "epoch": 4.87, "learning_rate": 5.332464146023469e-07, "loss": 0.0963, "step": 119500 }, { "epoch": 4.89, "learning_rate": 4.517601043024772e-07, "loss": 0.0816, "step": 120000 }, { "epoch": 4.91, "learning_rate": 3.702737940026076e-07, "loss": 0.1004, "step": 120500 }, { "epoch": 4.93, "learning_rate": 2.889504563233377e-07, "loss": 0.1038, "step": 121000 }, { "epoch": 4.95, "learning_rate": 2.0746414602346806e-07, "loss": 0.0843, "step": 121500 }, { "epoch": 4.97, "learning_rate": 1.2597783572359844e-07, "loss": 0.1104, "step": 122000 }, { "epoch": 4.99, "learning_rate": 4.449152542372882e-08, "loss": 0.0853, "step": 122500 }, { "epoch": 5.0, "step": 122720, "total_flos": 1.2826585381404672e+17, "train_loss": 0.2862651685536918, "train_runtime": 5523.0542, "train_samples_per_second": 355.512, "train_steps_per_second": 22.22 } ], "max_steps": 122720, "num_train_epochs": 5, "total_flos": 1.2826585381404672e+17, "trial_name": null, "trial_params": null }