{ "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": 1.991932855280313e-05, "loss": 0.886, "step": 500 }, { "epoch": 0.04, "learning_rate": 1.983784224250326e-05, "loss": 0.6985, "step": 1000 }, { "epoch": 0.06, "learning_rate": 1.975635593220339e-05, "loss": 0.6446, "step": 1500 }, { "epoch": 0.08, "learning_rate": 1.967486962190352e-05, "loss": 0.6309, "step": 2000 }, { "epoch": 0.1, "learning_rate": 1.9593383311603653e-05, "loss": 0.6153, "step": 2500 }, { "epoch": 0.12, "learning_rate": 1.9511897001303784e-05, "loss": 0.6165, "step": 3000 }, { "epoch": 0.14, "learning_rate": 1.9430410691003915e-05, "loss": 0.5867, "step": 3500 }, { "epoch": 0.16, "learning_rate": 1.9348924380704043e-05, "loss": 0.59, "step": 4000 }, { "epoch": 0.18, "learning_rate": 1.9267438070404174e-05, "loss": 0.5768, "step": 4500 }, { "epoch": 0.2, "learning_rate": 1.9185951760104305e-05, "loss": 0.5758, "step": 5000 }, { "epoch": 0.22, "learning_rate": 1.9104465449804433e-05, "loss": 0.5651, "step": 5500 }, { "epoch": 0.24, "learning_rate": 1.9022979139504564e-05, "loss": 0.5702, "step": 6000 }, { "epoch": 0.26, "learning_rate": 1.8941492829204696e-05, "loss": 0.5707, "step": 6500 }, { "epoch": 0.29, "learning_rate": 1.8860006518904827e-05, "loss": 0.5567, "step": 7000 }, { "epoch": 0.31, "learning_rate": 1.8778520208604955e-05, "loss": 0.5417, "step": 7500 }, { "epoch": 0.33, "learning_rate": 1.8697033898305086e-05, "loss": 0.5497, "step": 8000 }, { "epoch": 0.35, "learning_rate": 1.8615710560625818e-05, "loss": 0.5467, "step": 8500 }, { "epoch": 0.37, "learning_rate": 1.8534224250325946e-05, "loss": 0.5696, "step": 9000 }, { "epoch": 0.39, "learning_rate": 1.8452737940026077e-05, "loss": 0.5536, "step": 9500 }, { "epoch": 0.41, "learning_rate": 1.837125162972621e-05, "loss": 0.5521, "step": 10000 }, { "epoch": 0.43, "learning_rate": 1.828976531942634e-05, "loss": 0.535, "step": 10500 }, { "epoch": 0.45, "learning_rate": 1.8208279009126467e-05, "loss": 0.5277, "step": 11000 }, { "epoch": 0.47, "learning_rate": 1.81267926988266e-05, "loss": 0.5418, "step": 11500 }, { "epoch": 0.49, "learning_rate": 1.804530638852673e-05, "loss": 0.5358, "step": 12000 }, { "epoch": 0.51, "learning_rate": 1.7963820078226857e-05, "loss": 0.5349, "step": 12500 }, { "epoch": 0.53, "learning_rate": 1.788249674054759e-05, "loss": 0.5253, "step": 13000 }, { "epoch": 0.55, "learning_rate": 1.780101043024772e-05, "loss": 0.5261, "step": 13500 }, { "epoch": 0.57, "learning_rate": 1.771968709256845e-05, "loss": 0.5424, "step": 14000 }, { "epoch": 0.59, "learning_rate": 1.763820078226858e-05, "loss": 0.5237, "step": 14500 }, { "epoch": 0.61, "learning_rate": 1.755671447196871e-05, "loss": 0.5307, "step": 15000 }, { "epoch": 0.63, "learning_rate": 1.747522816166884e-05, "loss": 0.5365, "step": 15500 }, { "epoch": 0.65, "learning_rate": 1.739374185136897e-05, "loss": 0.5146, "step": 16000 }, { "epoch": 0.67, "learning_rate": 1.73124185136897e-05, "loss": 0.53, "step": 16500 }, { "epoch": 0.69, "learning_rate": 1.723093220338983e-05, "loss": 0.522, "step": 17000 }, { "epoch": 0.71, "learning_rate": 1.7149445893089963e-05, "loss": 0.5336, "step": 17500 }, { "epoch": 0.73, "learning_rate": 1.7067959582790094e-05, "loss": 0.5299, "step": 18000 }, { "epoch": 0.75, "learning_rate": 1.698647327249022e-05, "loss": 0.511, "step": 18500 }, { "epoch": 0.77, "learning_rate": 1.6904986962190353e-05, "loss": 0.5148, "step": 19000 }, { "epoch": 0.79, "learning_rate": 1.6823500651890484e-05, "loss": 0.516, "step": 19500 }, { "epoch": 0.81, "learning_rate": 1.6742014341590612e-05, "loss": 0.498, "step": 20000 }, { "epoch": 0.84, "learning_rate": 1.6660691003911344e-05, "loss": 0.5022, "step": 20500 }, { "epoch": 0.86, "learning_rate": 1.6579204693611475e-05, "loss": 0.5148, "step": 21000 }, { "epoch": 0.88, "learning_rate": 1.6497718383311606e-05, "loss": 0.5245, "step": 21500 }, { "epoch": 0.9, "learning_rate": 1.6416232073011734e-05, "loss": 0.4936, "step": 22000 }, { "epoch": 0.92, "learning_rate": 1.6334745762711865e-05, "loss": 0.5063, "step": 22500 }, { "epoch": 0.94, "learning_rate": 1.6253585397653195e-05, "loss": 0.514, "step": 23000 }, { "epoch": 0.96, "learning_rate": 1.6172099087353327e-05, "loss": 0.4997, "step": 23500 }, { "epoch": 0.98, "learning_rate": 1.6090612777053454e-05, "loss": 0.4968, "step": 24000 }, { "epoch": 1.0, "learning_rate": 1.6009126466753586e-05, "loss": 0.4964, "step": 24500 }, { "epoch": 1.02, "learning_rate": 1.5927640156453717e-05, "loss": 0.3777, "step": 25000 }, { "epoch": 1.04, "learning_rate": 1.5846153846153848e-05, "loss": 0.3681, "step": 25500 }, { "epoch": 1.06, "learning_rate": 1.5764667535853976e-05, "loss": 0.376, "step": 26000 }, { "epoch": 1.08, "learning_rate": 1.5683344198174708e-05, "loss": 0.3569, "step": 26500 }, { "epoch": 1.1, "learning_rate": 1.560185788787484e-05, "loss": 0.3708, "step": 27000 }, { "epoch": 1.12, "learning_rate": 1.5520371577574967e-05, "loss": 0.3585, "step": 27500 }, { "epoch": 1.14, "learning_rate": 1.5438885267275098e-05, "loss": 0.3835, "step": 28000 }, { "epoch": 1.16, "learning_rate": 1.535739895697523e-05, "loss": 0.3644, "step": 28500 }, { "epoch": 1.18, "learning_rate": 1.5276075619295962e-05, "loss": 0.3601, "step": 29000 }, { "epoch": 1.2, "learning_rate": 1.5194589308996091e-05, "loss": 0.3731, "step": 29500 }, { "epoch": 1.22, "learning_rate": 1.511310299869622e-05, "loss": 0.3642, "step": 30000 }, { "epoch": 1.24, "learning_rate": 1.503161668839635e-05, "loss": 0.3687, "step": 30500 }, { "epoch": 1.26, "learning_rate": 1.4950130378096481e-05, "loss": 0.3856, "step": 31000 }, { "epoch": 1.28, "learning_rate": 1.4868644067796611e-05, "loss": 0.3863, "step": 31500 }, { "epoch": 1.3, "learning_rate": 1.4787157757496742e-05, "loss": 0.3774, "step": 32000 }, { "epoch": 1.32, "learning_rate": 1.4705834419817473e-05, "loss": 0.3719, "step": 32500 }, { "epoch": 1.34, "learning_rate": 1.4624348109517604e-05, "loss": 0.3727, "step": 33000 }, { "epoch": 1.36, "learning_rate": 1.4542861799217733e-05, "loss": 0.3722, "step": 33500 }, { "epoch": 1.39, "learning_rate": 1.4461375488917863e-05, "loss": 0.3719, "step": 34000 }, { "epoch": 1.41, "learning_rate": 1.4379889178617994e-05, "loss": 0.3825, "step": 34500 }, { "epoch": 1.43, "learning_rate": 1.4298402868318124e-05, "loss": 0.3745, "step": 35000 }, { "epoch": 1.45, "learning_rate": 1.4217079530638852e-05, "loss": 0.3759, "step": 35500 }, { "epoch": 1.47, "learning_rate": 1.4135593220338984e-05, "loss": 0.3868, "step": 36000 }, { "epoch": 1.49, "learning_rate": 1.4054106910039113e-05, "loss": 0.3758, "step": 36500 }, { "epoch": 1.51, "learning_rate": 1.3972620599739244e-05, "loss": 0.3713, "step": 37000 }, { "epoch": 1.53, "learning_rate": 1.3891134289439376e-05, "loss": 0.3838, "step": 37500 }, { "epoch": 1.55, "learning_rate": 1.3809647979139507e-05, "loss": 0.3687, "step": 38000 }, { "epoch": 1.57, "learning_rate": 1.3728161668839636e-05, "loss": 0.3592, "step": 38500 }, { "epoch": 1.59, "learning_rate": 1.3646675358539767e-05, "loss": 0.3848, "step": 39000 }, { "epoch": 1.61, "learning_rate": 1.3565352020860496e-05, "loss": 0.3649, "step": 39500 }, { "epoch": 1.63, "learning_rate": 1.3483865710560626e-05, "loss": 0.3817, "step": 40000 }, { "epoch": 1.65, "learning_rate": 1.3402379400260757e-05, "loss": 0.3757, "step": 40500 }, { "epoch": 1.67, "learning_rate": 1.3320893089960886e-05, "loss": 0.3762, "step": 41000 }, { "epoch": 1.69, "learning_rate": 1.3239569752281619e-05, "loss": 0.3717, "step": 41500 }, { "epoch": 1.71, "learning_rate": 1.3158083441981748e-05, "loss": 0.3757, "step": 42000 }, { "epoch": 1.73, "learning_rate": 1.3076597131681878e-05, "loss": 0.3766, "step": 42500 }, { "epoch": 1.75, "learning_rate": 1.2995110821382009e-05, "loss": 0.3752, "step": 43000 }, { "epoch": 1.77, "learning_rate": 1.2913624511082138e-05, "loss": 0.3861, "step": 43500 }, { "epoch": 1.79, "learning_rate": 1.283213820078227e-05, "loss": 0.3809, "step": 44000 }, { "epoch": 1.81, "learning_rate": 1.27506518904824e-05, "loss": 0.3836, "step": 44500 }, { "epoch": 1.83, "learning_rate": 1.2669328552803131e-05, "loss": 0.3873, "step": 45000 }, { "epoch": 1.85, "learning_rate": 1.2587842242503261e-05, "loss": 0.3898, "step": 45500 }, { "epoch": 1.87, "learning_rate": 1.250635593220339e-05, "loss": 0.3892, "step": 46000 }, { "epoch": 1.89, "learning_rate": 1.2424869621903522e-05, "loss": 0.3828, "step": 46500 }, { "epoch": 1.91, "learning_rate": 1.2343383311603651e-05, "loss": 0.387, "step": 47000 }, { "epoch": 1.94, "learning_rate": 1.226205997392438e-05, "loss": 0.3691, "step": 47500 }, { "epoch": 1.96, "learning_rate": 1.2180573663624513e-05, "loss": 0.3551, "step": 48000 }, { "epoch": 1.98, "learning_rate": 1.2099087353324644e-05, "loss": 0.3846, "step": 48500 }, { "epoch": 2.0, "learning_rate": 1.2017601043024774e-05, "loss": 0.3766, "step": 49000 }, { "epoch": 2.02, "learning_rate": 1.1936277705345502e-05, "loss": 0.2658, "step": 49500 }, { "epoch": 2.04, "learning_rate": 1.1854791395045632e-05, "loss": 0.2262, "step": 50000 }, { "epoch": 2.06, "learning_rate": 1.1773468057366364e-05, "loss": 0.2343, "step": 50500 }, { "epoch": 2.08, "learning_rate": 1.1691981747066494e-05, "loss": 0.2277, "step": 51000 }, { "epoch": 2.1, "learning_rate": 1.1610495436766625e-05, "loss": 0.2209, "step": 51500 }, { "epoch": 2.12, "learning_rate": 1.1529009126466754e-05, "loss": 0.2317, "step": 52000 }, { "epoch": 2.14, "learning_rate": 1.1447522816166886e-05, "loss": 0.2373, "step": 52500 }, { "epoch": 2.16, "learning_rate": 1.1366036505867015e-05, "loss": 0.2308, "step": 53000 }, { "epoch": 2.18, "learning_rate": 1.1284550195567145e-05, "loss": 0.2403, "step": 53500 }, { "epoch": 2.2, "learning_rate": 1.1203063885267276e-05, "loss": 0.2493, "step": 54000 }, { "epoch": 2.22, "learning_rate": 1.1121577574967405e-05, "loss": 0.2362, "step": 54500 }, { "epoch": 2.24, "learning_rate": 1.1040254237288138e-05, "loss": 0.2233, "step": 55000 }, { "epoch": 2.26, "learning_rate": 1.0958767926988267e-05, "loss": 0.229, "step": 55500 }, { "epoch": 2.28, "learning_rate": 1.0877444589308996e-05, "loss": 0.2394, "step": 56000 }, { "epoch": 2.3, "learning_rate": 1.0795958279009127e-05, "loss": 0.2446, "step": 56500 }, { "epoch": 2.32, "learning_rate": 1.0714471968709257e-05, "loss": 0.2446, "step": 57000 }, { "epoch": 2.34, "learning_rate": 1.0632985658409388e-05, "loss": 0.2318, "step": 57500 }, { "epoch": 2.36, "learning_rate": 1.0551662320730118e-05, "loss": 0.2436, "step": 58000 }, { "epoch": 2.38, "learning_rate": 1.0470176010430248e-05, "loss": 0.2399, "step": 58500 }, { "epoch": 2.4, "learning_rate": 1.038868970013038e-05, "loss": 0.2446, "step": 59000 }, { "epoch": 2.42, "learning_rate": 1.0307203389830509e-05, "loss": 0.231, "step": 59500 }, { "epoch": 2.44, "learning_rate": 1.022571707953064e-05, "loss": 0.2399, "step": 60000 }, { "epoch": 2.46, "learning_rate": 1.014423076923077e-05, "loss": 0.2506, "step": 60500 }, { "epoch": 2.49, "learning_rate": 1.00629074315515e-05, "loss": 0.232, "step": 61000 }, { "epoch": 2.51, "learning_rate": 9.981421121251631e-06, "loss": 0.2571, "step": 61500 }, { "epoch": 2.53, "learning_rate": 9.89993481095176e-06, "loss": 0.2567, "step": 62000 }, { "epoch": 2.55, "learning_rate": 9.818448500651892e-06, "loss": 0.2457, "step": 62500 }, { "epoch": 2.57, "learning_rate": 9.736962190352021e-06, "loss": 0.2408, "step": 63000 }, { "epoch": 2.59, "learning_rate": 9.655475880052153e-06, "loss": 0.2538, "step": 63500 }, { "epoch": 2.61, "learning_rate": 9.574152542372881e-06, "loss": 0.2324, "step": 64000 }, { "epoch": 2.63, "learning_rate": 9.492666232073013e-06, "loss": 0.242, "step": 64500 }, { "epoch": 2.65, "learning_rate": 9.411179921773142e-06, "loss": 0.2416, "step": 65000 }, { "epoch": 2.67, "learning_rate": 9.329693611473273e-06, "loss": 0.2386, "step": 65500 }, { "epoch": 2.69, "learning_rate": 9.248207301173405e-06, "loss": 0.2355, "step": 66000 }, { "epoch": 2.71, "learning_rate": 9.166720990873534e-06, "loss": 0.2576, "step": 66500 }, { "epoch": 2.73, "learning_rate": 9.085234680573665e-06, "loss": 0.2472, "step": 67000 }, { "epoch": 2.75, "learning_rate": 9.003748370273795e-06, "loss": 0.243, "step": 67500 }, { "epoch": 2.77, "learning_rate": 8.922588005215125e-06, "loss": 0.247, "step": 68000 }, { "epoch": 2.79, "learning_rate": 8.841101694915256e-06, "loss": 0.2462, "step": 68500 }, { "epoch": 2.81, "learning_rate": 8.759615384615385e-06, "loss": 0.2614, "step": 69000 }, { "epoch": 2.83, "learning_rate": 8.678129074315515e-06, "loss": 0.2384, "step": 69500 }, { "epoch": 2.85, "learning_rate": 8.596642764015646e-06, "loss": 0.2417, "step": 70000 }, { "epoch": 2.87, "learning_rate": 8.515156453715776e-06, "loss": 0.255, "step": 70500 }, { "epoch": 2.89, "learning_rate": 8.433670143415907e-06, "loss": 0.2439, "step": 71000 }, { "epoch": 2.91, "learning_rate": 8.352183833116038e-06, "loss": 0.2473, "step": 71500 }, { "epoch": 2.93, "learning_rate": 8.270697522816167e-06, "loss": 0.2399, "step": 72000 }, { "epoch": 2.95, "learning_rate": 8.189211212516299e-06, "loss": 0.2503, "step": 72500 }, { "epoch": 2.97, "learning_rate": 8.107724902216428e-06, "loss": 0.2397, "step": 73000 }, { "epoch": 2.99, "learning_rate": 8.026401564537159e-06, "loss": 0.2373, "step": 73500 }, { "epoch": 3.01, "learning_rate": 7.944915254237288e-06, "loss": 0.1605, "step": 74000 }, { "epoch": 3.04, "learning_rate": 7.86342894393742e-06, "loss": 0.1379, "step": 74500 }, { "epoch": 3.06, "learning_rate": 7.781942633637549e-06, "loss": 0.1501, "step": 75000 }, { "epoch": 3.08, "learning_rate": 7.700782268578879e-06, "loss": 0.1415, "step": 75500 }, { "epoch": 3.1, "learning_rate": 7.619295958279009e-06, "loss": 0.1508, "step": 76000 }, { "epoch": 3.12, "learning_rate": 7.53780964797914e-06, "loss": 0.1495, "step": 76500 }, { "epoch": 3.14, "learning_rate": 7.456323337679271e-06, "loss": 0.1498, "step": 77000 }, { "epoch": 3.16, "learning_rate": 7.375000000000001e-06, "loss": 0.156, "step": 77500 }, { "epoch": 3.18, "learning_rate": 7.293513689700131e-06, "loss": 0.1609, "step": 78000 }, { "epoch": 3.2, "learning_rate": 7.212027379400261e-06, "loss": 0.1555, "step": 78500 }, { "epoch": 3.22, "learning_rate": 7.130541069100392e-06, "loss": 0.1483, "step": 79000 }, { "epoch": 3.24, "learning_rate": 7.049054758800522e-06, "loss": 0.1548, "step": 79500 }, { "epoch": 3.26, "learning_rate": 6.9677314211212525e-06, "loss": 0.1581, "step": 80000 }, { "epoch": 3.28, "learning_rate": 6.886245110821382e-06, "loss": 0.1608, "step": 80500 }, { "epoch": 3.3, "learning_rate": 6.804758800521512e-06, "loss": 0.1614, "step": 81000 }, { "epoch": 3.32, "learning_rate": 6.723272490221643e-06, "loss": 0.1535, "step": 81500 }, { "epoch": 3.34, "learning_rate": 6.641786179921774e-06, "loss": 0.1486, "step": 82000 }, { "epoch": 3.36, "learning_rate": 6.560299869621904e-06, "loss": 0.1483, "step": 82500 }, { "epoch": 3.38, "learning_rate": 6.478976531942635e-06, "loss": 0.156, "step": 83000 }, { "epoch": 3.4, "learning_rate": 6.397490221642765e-06, "loss": 0.1566, "step": 83500 }, { "epoch": 3.42, "learning_rate": 6.316003911342895e-06, "loss": 0.1535, "step": 84000 }, { "epoch": 3.44, "learning_rate": 6.234517601043025e-06, "loss": 0.1637, "step": 84500 }, { "epoch": 3.46, "learning_rate": 6.153357235984355e-06, "loss": 0.1636, "step": 85000 }, { "epoch": 3.48, "learning_rate": 6.071870925684485e-06, "loss": 0.1609, "step": 85500 }, { "epoch": 3.5, "learning_rate": 5.990384615384616e-06, "loss": 0.1631, "step": 86000 }, { "epoch": 3.52, "learning_rate": 5.908898305084746e-06, "loss": 0.167, "step": 86500 }, { "epoch": 3.54, "learning_rate": 5.827411994784876e-06, "loss": 0.1568, "step": 87000 }, { "epoch": 3.57, "learning_rate": 5.745925684485008e-06, "loss": 0.1624, "step": 87500 }, { "epoch": 3.59, "learning_rate": 5.664439374185138e-06, "loss": 0.1543, "step": 88000 }, { "epoch": 3.61, "learning_rate": 5.582953063885268e-06, "loss": 0.1579, "step": 88500 }, { "epoch": 3.63, "learning_rate": 5.501466753585399e-06, "loss": 0.1676, "step": 89000 }, { "epoch": 3.65, "learning_rate": 5.420143415906128e-06, "loss": 0.1508, "step": 89500 }, { "epoch": 3.67, "learning_rate": 5.338657105606259e-06, "loss": 0.1549, "step": 90000 }, { "epoch": 3.69, "learning_rate": 5.257170795306389e-06, "loss": 0.1642, "step": 90500 }, { "epoch": 3.71, "learning_rate": 5.175684485006519e-06, "loss": 0.1612, "step": 91000 }, { "epoch": 3.73, "learning_rate": 5.09419817470665e-06, "loss": 0.1476, "step": 91500 }, { "epoch": 3.75, "learning_rate": 5.012711864406779e-06, "loss": 0.1565, "step": 92000 }, { "epoch": 3.77, "learning_rate": 4.93155149934811e-06, "loss": 0.1526, "step": 92500 }, { "epoch": 3.79, "learning_rate": 4.85006518904824e-06, "loss": 0.1578, "step": 93000 }, { "epoch": 3.81, "learning_rate": 4.768578878748371e-06, "loss": 0.174, "step": 93500 }, { "epoch": 3.83, "learning_rate": 4.687092568448501e-06, "loss": 0.1643, "step": 94000 }, { "epoch": 3.85, "learning_rate": 4.6056062581486315e-06, "loss": 0.1519, "step": 94500 }, { "epoch": 3.87, "learning_rate": 4.524119947848762e-06, "loss": 0.1591, "step": 95000 }, { "epoch": 3.89, "learning_rate": 4.442633637548892e-06, "loss": 0.1641, "step": 95500 }, { "epoch": 3.91, "learning_rate": 4.3611473272490225e-06, "loss": 0.1533, "step": 96000 }, { "epoch": 3.93, "learning_rate": 4.279661016949153e-06, "loss": 0.1595, "step": 96500 }, { "epoch": 3.95, "learning_rate": 4.1983376792698834e-06, "loss": 0.1579, "step": 97000 }, { "epoch": 3.97, "learning_rate": 4.116851368970014e-06, "loss": 0.147, "step": 97500 }, { "epoch": 3.99, "learning_rate": 4.035365058670144e-06, "loss": 0.1501, "step": 98000 }, { "epoch": 4.01, "learning_rate": 3.953878748370274e-06, "loss": 0.0986, "step": 98500 }, { "epoch": 4.03, "learning_rate": 3.872392438070404e-06, "loss": 0.0773, "step": 99000 }, { "epoch": 4.05, "learning_rate": 3.791232073011734e-06, "loss": 0.0799, "step": 99500 }, { "epoch": 4.07, "learning_rate": 3.709745762711865e-06, "loss": 0.0843, "step": 100000 }, { "epoch": 4.09, "learning_rate": 3.628259452411995e-06, "loss": 0.0803, "step": 100500 }, { "epoch": 4.12, "learning_rate": 3.5467731421121254e-06, "loss": 0.0814, "step": 101000 }, { "epoch": 4.14, "learning_rate": 3.4652868318122558e-06, "loss": 0.0845, "step": 101500 }, { "epoch": 4.16, "learning_rate": 3.3838005215123865e-06, "loss": 0.0691, "step": 102000 }, { "epoch": 4.18, "learning_rate": 3.302314211212517e-06, "loss": 0.085, "step": 102500 }, { "epoch": 4.2, "learning_rate": 3.220827900912647e-06, "loss": 0.0833, "step": 103000 }, { "epoch": 4.22, "learning_rate": 3.139341590612777e-06, "loss": 0.0764, "step": 103500 }, { "epoch": 4.24, "learning_rate": 3.058181225554107e-06, "loss": 0.0726, "step": 104000 }, { "epoch": 4.26, "learning_rate": 2.9766949152542375e-06, "loss": 0.0887, "step": 104500 }, { "epoch": 4.28, "learning_rate": 2.8952086049543682e-06, "loss": 0.0855, "step": 105000 }, { "epoch": 4.3, "learning_rate": 2.8137222946544986e-06, "loss": 0.0792, "step": 105500 }, { "epoch": 4.32, "learning_rate": 2.7322359843546285e-06, "loss": 0.0821, "step": 106000 }, { "epoch": 4.34, "learning_rate": 2.6509126466753587e-06, "loss": 0.0857, "step": 106500 }, { "epoch": 4.36, "learning_rate": 2.569426336375489e-06, "loss": 0.071, "step": 107000 }, { "epoch": 4.38, "learning_rate": 2.4879400260756194e-06, "loss": 0.0887, "step": 107500 }, { "epoch": 4.4, "learning_rate": 2.4064537157757497e-06, "loss": 0.0793, "step": 108000 }, { "epoch": 4.42, "learning_rate": 2.3249674054758805e-06, "loss": 0.0709, "step": 108500 }, { "epoch": 4.44, "learning_rate": 2.2434810951760104e-06, "loss": 0.0694, "step": 109000 }, { "epoch": 4.46, "learning_rate": 2.161994784876141e-06, "loss": 0.0793, "step": 109500 }, { "epoch": 4.48, "learning_rate": 2.0805084745762715e-06, "loss": 0.0734, "step": 110000 }, { "epoch": 4.5, "learning_rate": 1.9991851368970013e-06, "loss": 0.0742, "step": 110500 }, { "epoch": 4.52, "learning_rate": 1.9178617992177314e-06, "loss": 0.0795, "step": 111000 }, { "epoch": 4.54, "learning_rate": 1.836375488917862e-06, "loss": 0.0871, "step": 111500 }, { "epoch": 4.56, "learning_rate": 1.7548891786179923e-06, "loss": 0.0776, "step": 112000 }, { "epoch": 4.58, "learning_rate": 1.6734028683181227e-06, "loss": 0.0764, "step": 112500 }, { "epoch": 4.6, "learning_rate": 1.5920795306388528e-06, "loss": 0.0794, "step": 113000 }, { "epoch": 4.62, "learning_rate": 1.5105932203389832e-06, "loss": 0.0766, "step": 113500 }, { "epoch": 4.64, "learning_rate": 1.4291069100391135e-06, "loss": 0.0818, "step": 114000 }, { "epoch": 4.67, "learning_rate": 1.3476205997392439e-06, "loss": 0.0794, "step": 114500 }, { "epoch": 4.69, "learning_rate": 1.2661342894393744e-06, "loss": 0.0702, "step": 115000 }, { "epoch": 4.71, "learning_rate": 1.1846479791395046e-06, "loss": 0.0897, "step": 115500 }, { "epoch": 4.73, "learning_rate": 1.1031616688396351e-06, "loss": 0.087, "step": 116000 }, { "epoch": 4.75, "learning_rate": 1.0216753585397655e-06, "loss": 0.0733, "step": 116500 }, { "epoch": 4.77, "learning_rate": 9.403520208604955e-07, "loss": 0.0778, "step": 117000 }, { "epoch": 4.79, "learning_rate": 8.588657105606259e-07, "loss": 0.0797, "step": 117500 }, { "epoch": 4.81, "learning_rate": 7.773794002607562e-07, "loss": 0.0718, "step": 118000 }, { "epoch": 4.83, "learning_rate": 6.958930899608867e-07, "loss": 0.0754, "step": 118500 }, { "epoch": 4.85, "learning_rate": 6.14406779661017e-07, "loss": 0.0686, "step": 119000 }, { "epoch": 4.87, "learning_rate": 5.330834419817471e-07, "loss": 0.0715, "step": 119500 }, { "epoch": 4.89, "learning_rate": 4.5159713168187747e-07, "loss": 0.0775, "step": 120000 }, { "epoch": 4.91, "learning_rate": 3.7011082138200787e-07, "loss": 0.072, "step": 120500 }, { "epoch": 4.93, "learning_rate": 2.886245110821382e-07, "loss": 0.0544, "step": 121000 }, { "epoch": 4.95, "learning_rate": 2.071382007822686e-07, "loss": 0.068, "step": 121500 }, { "epoch": 4.97, "learning_rate": 1.258148631029987e-07, "loss": 0.0628, "step": 122000 }, { "epoch": 4.99, "learning_rate": 4.432855280312908e-08, "loss": 0.065, "step": 122500 }, { "epoch": 5.0, "step": 122720, "total_flos": 4.5745873020125184e+17, "train_loss": 0.28029768033319846, "train_runtime": 11164.1016, "train_samples_per_second": 175.877, "train_steps_per_second": 10.992 } ], "max_steps": 122720, "num_train_epochs": 5, "total_flos": 4.5745873020125184e+17, "trial_name": null, "trial_params": null }