{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "global_step": 68592, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.02, "learning_rate": 4.963552600886401e-05, "loss": 1.8747, "step": 500 }, { "epoch": 0.04, "learning_rate": 4.927105201772802e-05, "loss": 1.7998, "step": 1000 }, { "epoch": 0.07, "learning_rate": 4.890657802659203e-05, "loss": 1.7788, "step": 1500 }, { "epoch": 0.09, "learning_rate": 4.8542104035456035e-05, "loss": 1.7681, "step": 2000 }, { "epoch": 0.11, "learning_rate": 4.8177630044320036e-05, "loss": 1.7504, "step": 2500 }, { "epoch": 0.13, "learning_rate": 4.781315605318405e-05, "loss": 1.7426, "step": 3000 }, { "epoch": 0.15, "learning_rate": 4.744868206204805e-05, "loss": 1.7515, "step": 3500 }, { "epoch": 0.17, "learning_rate": 4.708420807091206e-05, "loss": 1.7297, "step": 4000 }, { "epoch": 0.2, "learning_rate": 4.671973407977607e-05, "loss": 1.7266, "step": 4500 }, { "epoch": 0.22, "learning_rate": 4.635526008864008e-05, "loss": 1.716, "step": 5000 }, { "epoch": 0.24, "learning_rate": 4.5990786097504085e-05, "loss": 1.7168, "step": 5500 }, { "epoch": 0.26, "learning_rate": 4.5626312106368094e-05, "loss": 1.7136, "step": 6000 }, { "epoch": 0.28, "learning_rate": 4.5261838115232095e-05, "loss": 1.7018, "step": 6500 }, { "epoch": 0.31, "learning_rate": 4.489736412409611e-05, "loss": 1.6961, "step": 7000 }, { "epoch": 0.33, "learning_rate": 4.453289013296011e-05, "loss": 1.7138, "step": 7500 }, { "epoch": 0.35, "learning_rate": 4.416841614182412e-05, "loss": 1.7039, "step": 8000 }, { "epoch": 0.37, "learning_rate": 4.380394215068813e-05, "loss": 1.698, "step": 8500 }, { "epoch": 0.39, "learning_rate": 4.3439468159552136e-05, "loss": 1.6897, "step": 9000 }, { "epoch": 0.42, "learning_rate": 4.3074994168416144e-05, "loss": 1.6895, "step": 9500 }, { "epoch": 0.44, "learning_rate": 4.271052017728015e-05, "loss": 1.6914, "step": 10000 }, { "epoch": 0.46, "learning_rate": 4.234604618614416e-05, "loss": 1.6772, "step": 10500 }, { "epoch": 0.48, "learning_rate": 4.198157219500817e-05, "loss": 1.6738, "step": 11000 }, { "epoch": 0.5, "learning_rate": 4.161709820387218e-05, "loss": 1.6739, "step": 11500 }, { "epoch": 0.52, "learning_rate": 4.125262421273618e-05, "loss": 1.6645, "step": 12000 }, { "epoch": 0.55, "learning_rate": 4.088815022160019e-05, "loss": 1.6601, "step": 12500 }, { "epoch": 0.57, "learning_rate": 4.0523676230464194e-05, "loss": 1.6523, "step": 13000 }, { "epoch": 0.59, "learning_rate": 4.01592022393282e-05, "loss": 1.6626, "step": 13500 }, { "epoch": 0.61, "learning_rate": 3.979472824819221e-05, "loss": 1.6574, "step": 14000 }, { "epoch": 0.63, "learning_rate": 3.943025425705622e-05, "loss": 1.6748, "step": 14500 }, { "epoch": 0.66, "learning_rate": 3.906578026592023e-05, "loss": 1.668, "step": 15000 }, { "epoch": 0.68, "learning_rate": 3.8701306274784235e-05, "loss": 1.656, "step": 15500 }, { "epoch": 0.7, "learning_rate": 3.833683228364824e-05, "loss": 1.6487, "step": 16000 }, { "epoch": 0.72, "learning_rate": 3.797235829251225e-05, "loss": 1.6537, "step": 16500 }, { "epoch": 0.74, "learning_rate": 3.760788430137625e-05, "loss": 1.655, "step": 17000 }, { "epoch": 0.77, "learning_rate": 3.724341031024026e-05, "loss": 1.6492, "step": 17500 }, { "epoch": 0.79, "learning_rate": 3.687893631910427e-05, "loss": 1.6477, "step": 18000 }, { "epoch": 0.81, "learning_rate": 3.651446232796828e-05, "loss": 1.642, "step": 18500 }, { "epoch": 0.83, "learning_rate": 3.6149988336832286e-05, "loss": 1.6285, "step": 19000 }, { "epoch": 0.85, "learning_rate": 3.5785514345696294e-05, "loss": 1.6415, "step": 19500 }, { "epoch": 0.87, "learning_rate": 3.5421040354560295e-05, "loss": 1.6349, "step": 20000 }, { "epoch": 0.9, "learning_rate": 3.505656636342431e-05, "loss": 1.6272, "step": 20500 }, { "epoch": 0.92, "learning_rate": 3.469209237228832e-05, "loss": 1.6268, "step": 21000 }, { "epoch": 0.94, "learning_rate": 3.432761838115232e-05, "loss": 1.6244, "step": 21500 }, { "epoch": 0.96, "learning_rate": 3.3963144390016335e-05, "loss": 1.6253, "step": 22000 }, { "epoch": 0.98, "learning_rate": 3.3598670398880336e-05, "loss": 1.6337, "step": 22500 }, { "epoch": 1.01, "learning_rate": 3.3234196407744344e-05, "loss": 1.6131, "step": 23000 }, { "epoch": 1.03, "learning_rate": 3.286972241660835e-05, "loss": 1.6085, "step": 23500 }, { "epoch": 1.05, "learning_rate": 3.250524842547236e-05, "loss": 1.6141, "step": 24000 }, { "epoch": 1.07, "learning_rate": 3.214077443433637e-05, "loss": 1.6097, "step": 24500 }, { "epoch": 1.09, "learning_rate": 3.177630044320038e-05, "loss": 1.6068, "step": 25000 }, { "epoch": 1.12, "learning_rate": 3.141182645206438e-05, "loss": 1.5991, "step": 25500 }, { "epoch": 1.14, "learning_rate": 3.104735246092839e-05, "loss": 1.6001, "step": 26000 }, { "epoch": 1.16, "learning_rate": 3.0682878469792395e-05, "loss": 1.5873, "step": 26500 }, { "epoch": 1.18, "learning_rate": 3.0318404478656403e-05, "loss": 1.5979, "step": 27000 }, { "epoch": 1.2, "learning_rate": 2.995393048752041e-05, "loss": 1.5989, "step": 27500 }, { "epoch": 1.22, "learning_rate": 2.9589456496384422e-05, "loss": 1.5973, "step": 28000 }, { "epoch": 1.25, "learning_rate": 2.9224982505248427e-05, "loss": 1.5945, "step": 28500 }, { "epoch": 1.27, "learning_rate": 2.8860508514112432e-05, "loss": 1.581, "step": 29000 }, { "epoch": 1.29, "learning_rate": 2.849603452297644e-05, "loss": 1.5826, "step": 29500 }, { "epoch": 1.31, "learning_rate": 2.8131560531840452e-05, "loss": 1.594, "step": 30000 }, { "epoch": 1.33, "learning_rate": 2.7767086540704457e-05, "loss": 1.5812, "step": 30500 }, { "epoch": 1.36, "learning_rate": 2.740261254956846e-05, "loss": 1.5779, "step": 31000 }, { "epoch": 1.38, "learning_rate": 2.7038138558432473e-05, "loss": 1.5807, "step": 31500 }, { "epoch": 1.4, "learning_rate": 2.667366456729648e-05, "loss": 1.5815, "step": 32000 }, { "epoch": 1.42, "learning_rate": 2.6309190576160486e-05, "loss": 1.5877, "step": 32500 }, { "epoch": 1.44, "learning_rate": 2.594471658502449e-05, "loss": 1.5694, "step": 33000 }, { "epoch": 1.47, "learning_rate": 2.5580242593888502e-05, "loss": 1.5807, "step": 33500 }, { "epoch": 1.49, "learning_rate": 2.521576860275251e-05, "loss": 1.5794, "step": 34000 }, { "epoch": 1.51, "learning_rate": 2.4851294611616515e-05, "loss": 1.5798, "step": 34500 }, { "epoch": 1.53, "learning_rate": 2.4486820620480523e-05, "loss": 1.5702, "step": 35000 }, { "epoch": 1.55, "learning_rate": 2.412234662934453e-05, "loss": 1.5757, "step": 35500 }, { "epoch": 1.57, "learning_rate": 2.375787263820854e-05, "loss": 1.5699, "step": 36000 }, { "epoch": 1.6, "learning_rate": 2.3393398647072544e-05, "loss": 1.5691, "step": 36500 }, { "epoch": 1.62, "learning_rate": 2.3028924655936553e-05, "loss": 1.5566, "step": 37000 }, { "epoch": 1.64, "learning_rate": 2.266445066480056e-05, "loss": 1.562, "step": 37500 }, { "epoch": 1.66, "learning_rate": 2.229997667366457e-05, "loss": 1.5541, "step": 38000 }, { "epoch": 1.68, "learning_rate": 2.1935502682528574e-05, "loss": 1.5554, "step": 38500 }, { "epoch": 1.71, "learning_rate": 2.1571028691392585e-05, "loss": 1.5597, "step": 39000 }, { "epoch": 1.73, "learning_rate": 2.120655470025659e-05, "loss": 1.5508, "step": 39500 }, { "epoch": 1.75, "learning_rate": 2.0842080709120598e-05, "loss": 1.5551, "step": 40000 }, { "epoch": 1.77, "learning_rate": 2.0477606717984603e-05, "loss": 1.5637, "step": 40500 }, { "epoch": 1.79, "learning_rate": 2.0113132726848614e-05, "loss": 1.5442, "step": 41000 }, { "epoch": 1.82, "learning_rate": 1.9748658735712623e-05, "loss": 1.5521, "step": 41500 }, { "epoch": 1.84, "learning_rate": 1.9384184744576627e-05, "loss": 1.5468, "step": 42000 }, { "epoch": 1.86, "learning_rate": 1.9019710753440636e-05, "loss": 1.5492, "step": 42500 }, { "epoch": 1.88, "learning_rate": 1.8655236762304644e-05, "loss": 1.551, "step": 43000 }, { "epoch": 1.9, "learning_rate": 1.8290762771168652e-05, "loss": 1.5483, "step": 43500 }, { "epoch": 1.92, "learning_rate": 1.7926288780032657e-05, "loss": 1.5499, "step": 44000 }, { "epoch": 1.95, "learning_rate": 1.7561814788896665e-05, "loss": 1.5273, "step": 44500 }, { "epoch": 1.97, "learning_rate": 1.7197340797760673e-05, "loss": 1.5441, "step": 45000 }, { "epoch": 1.99, "learning_rate": 1.683286680662468e-05, "loss": 1.5401, "step": 45500 }, { "epoch": 2.01, "learning_rate": 1.6468392815488686e-05, "loss": 1.5262, "step": 46000 }, { "epoch": 2.03, "learning_rate": 1.6103918824352694e-05, "loss": 1.5231, "step": 46500 }, { "epoch": 2.06, "learning_rate": 1.5739444833216702e-05, "loss": 1.5282, "step": 47000 }, { "epoch": 2.08, "learning_rate": 1.537497084208071e-05, "loss": 1.5209, "step": 47500 }, { "epoch": 2.1, "learning_rate": 1.5010496850944717e-05, "loss": 1.5224, "step": 48000 }, { "epoch": 2.12, "learning_rate": 1.4646022859808725e-05, "loss": 1.5072, "step": 48500 }, { "epoch": 2.14, "learning_rate": 1.4281548868672732e-05, "loss": 1.5153, "step": 49000 }, { "epoch": 2.16, "learning_rate": 1.391707487753674e-05, "loss": 1.5189, "step": 49500 }, { "epoch": 2.19, "learning_rate": 1.3552600886400746e-05, "loss": 1.5185, "step": 50000 }, { "epoch": 2.21, "learning_rate": 1.3188126895264754e-05, "loss": 1.5073, "step": 50500 }, { "epoch": 2.23, "learning_rate": 1.282365290412876e-05, "loss": 1.522, "step": 51000 }, { "epoch": 2.25, "learning_rate": 1.2459178912992769e-05, "loss": 1.5158, "step": 51500 }, { "epoch": 2.27, "learning_rate": 1.2094704921856777e-05, "loss": 1.5102, "step": 52000 }, { "epoch": 2.3, "learning_rate": 1.1730230930720784e-05, "loss": 1.5093, "step": 52500 }, { "epoch": 2.32, "learning_rate": 1.1365756939584792e-05, "loss": 1.5101, "step": 53000 }, { "epoch": 2.34, "learning_rate": 1.1001282948448798e-05, "loss": 1.5062, "step": 53500 }, { "epoch": 2.36, "learning_rate": 1.0636808957312806e-05, "loss": 1.5039, "step": 54000 }, { "epoch": 2.38, "learning_rate": 1.0272334966176813e-05, "loss": 1.494, "step": 54500 }, { "epoch": 2.41, "learning_rate": 9.907860975040821e-06, "loss": 1.5095, "step": 55000 }, { "epoch": 2.43, "learning_rate": 9.54338698390483e-06, "loss": 1.5021, "step": 55500 }, { "epoch": 2.45, "learning_rate": 9.178912992768836e-06, "loss": 1.5046, "step": 56000 }, { "epoch": 2.47, "learning_rate": 8.814439001632844e-06, "loss": 1.4966, "step": 56500 }, { "epoch": 2.49, "learning_rate": 8.449965010496852e-06, "loss": 1.4875, "step": 57000 }, { "epoch": 2.51, "learning_rate": 8.085491019360859e-06, "loss": 1.5078, "step": 57500 }, { "epoch": 2.54, "learning_rate": 7.721017028224867e-06, "loss": 1.4892, "step": 58000 }, { "epoch": 2.56, "learning_rate": 7.356543037088874e-06, "loss": 1.4984, "step": 58500 }, { "epoch": 2.58, "learning_rate": 6.992069045952881e-06, "loss": 1.4963, "step": 59000 }, { "epoch": 2.6, "learning_rate": 6.627595054816889e-06, "loss": 1.4895, "step": 59500 }, { "epoch": 2.62, "learning_rate": 6.263121063680896e-06, "loss": 1.5005, "step": 60000 }, { "epoch": 2.65, "learning_rate": 5.898647072544903e-06, "loss": 1.4999, "step": 60500 }, { "epoch": 2.67, "learning_rate": 5.534173081408911e-06, "loss": 1.4832, "step": 61000 }, { "epoch": 2.69, "learning_rate": 5.169699090272919e-06, "loss": 1.495, "step": 61500 }, { "epoch": 2.71, "learning_rate": 4.805225099136926e-06, "loss": 1.499, "step": 62000 }, { "epoch": 2.73, "learning_rate": 4.4407511080009334e-06, "loss": 1.4831, "step": 62500 }, { "epoch": 2.76, "learning_rate": 4.076277116864941e-06, "loss": 1.4854, "step": 63000 }, { "epoch": 2.78, "learning_rate": 3.711803125728948e-06, "loss": 1.4918, "step": 63500 }, { "epoch": 2.8, "learning_rate": 3.3473291345929554e-06, "loss": 1.4914, "step": 64000 }, { "epoch": 2.82, "learning_rate": 2.982855143456963e-06, "loss": 1.4836, "step": 64500 }, { "epoch": 2.84, "learning_rate": 2.6183811523209705e-06, "loss": 1.4829, "step": 65000 }, { "epoch": 2.86, "learning_rate": 2.253907161184978e-06, "loss": 1.4824, "step": 65500 }, { "epoch": 2.89, "learning_rate": 1.8894331700489855e-06, "loss": 1.4856, "step": 66000 }, { "epoch": 2.91, "learning_rate": 1.5249591789129928e-06, "loss": 1.487, "step": 66500 }, { "epoch": 2.93, "learning_rate": 1.1604851877770002e-06, "loss": 1.4791, "step": 67000 }, { "epoch": 2.95, "learning_rate": 7.960111966410078e-07, "loss": 1.4779, "step": 67500 }, { "epoch": 2.97, "learning_rate": 4.3153720550501516e-07, "loss": 1.4794, "step": 68000 }, { "epoch": 3.0, "learning_rate": 6.706321436902263e-08, "loss": 1.476, "step": 68500 }, { "epoch": 3.0, "step": 68592, "total_flos": 1.805366555813376e+17, "train_loss": 1.5861018672255278, "train_runtime": 49915.9647, "train_samples_per_second": 13.741, "train_steps_per_second": 1.374 } ], "max_steps": 68592, "num_train_epochs": 3, "total_flos": 1.805366555813376e+17, "trial_name": null, "trial_params": null }