| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.995708154506438, | |
| "eval_steps": 500, | |
| "global_step": 349, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 5.055780862946804, | |
| "learning_rate": 9.997135147120633e-05, | |
| "loss": 2.3426, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 3.9099604179547867, | |
| "learning_rate": 9.98854387143534e-05, | |
| "loss": 1.9269, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 2.3917278619890237, | |
| "learning_rate": 9.974236018040474e-05, | |
| "loss": 1.6971, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 2.0300075071526407, | |
| "learning_rate": 9.954227982894034e-05, | |
| "loss": 1.5024, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 2.5503247326265703, | |
| "learning_rate": 9.928542694026862e-05, | |
| "loss": 1.4291, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 2.8860783306647813, | |
| "learning_rate": 9.897209585268458e-05, | |
| "loss": 1.3888, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.7598771943831562, | |
| "learning_rate": 9.86026456251757e-05, | |
| "loss": 1.3614, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 1.8472495254954506, | |
| "learning_rate": 9.817749962596115e-05, | |
| "loss": 1.3218, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 1.8685305808519006, | |
| "learning_rate": 9.769714504733694e-05, | |
| "loss": 1.3144, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.6934593149050314, | |
| "learning_rate": 9.716213234738215e-05, | |
| "loss": 1.3105, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.6536690261785199, | |
| "learning_rate": 9.657307461916635e-05, | |
| "loss": 1.287, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.8152365885852073, | |
| "learning_rate": 9.59306468881811e-05, | |
| "loss": 1.2877, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.6831619071424775, | |
| "learning_rate": 9.52355853388003e-05, | |
| "loss": 1.2504, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.8966578603527602, | |
| "learning_rate": 9.448868647065642e-05, | |
| "loss": 1.2782, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.9475911961026766, | |
| "learning_rate": 9.369080618589864e-05, | |
| "loss": 1.2517, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 1.7144643185356307, | |
| "learning_rate": 9.284285880837946e-05, | |
| "loss": 1.2442, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.7507182276034248, | |
| "learning_rate": 9.194581603589328e-05, | |
| "loss": 1.2174, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.782170598732355, | |
| "learning_rate": 9.100070582666795e-05, | |
| "loss": 1.2231, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 1.884782447980567, | |
| "learning_rate": 9.000861122138517e-05, | |
| "loss": 1.1901, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 1.9269085187568622, | |
| "learning_rate": 8.897066910207958e-05, | |
| "loss": 1.1959, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 2.0913092281668977, | |
| "learning_rate": 8.788806888933881e-05, | |
| "loss": 1.1953, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.6532109062340223, | |
| "learning_rate": 8.676205117929752e-05, | |
| "loss": 1.2135, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 1.9499528694077117, | |
| "learning_rate": 8.559390632198723e-05, | |
| "loss": 1.2141, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.2135138511657715, | |
| "eval_runtime": 39.1621, | |
| "eval_samples_per_second": 7.66, | |
| "eval_steps_per_second": 0.97, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 1.6359783575613835, | |
| "learning_rate": 8.438497294267117e-05, | |
| "loss": 1.1555, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 1.9376629758093418, | |
| "learning_rate": 8.313663640785839e-05, | |
| "loss": 1.123, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 1.8078455578062511, | |
| "learning_rate": 8.185032723775539e-05, | |
| "loss": 1.1545, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 1.78298002573264, | |
| "learning_rate": 8.052751946697403e-05, | |
| "loss": 1.0917, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 1.8433442214623892, | |
| "learning_rate": 7.916972895537471e-05, | |
| "loss": 1.1507, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 1.7754964498370087, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 1.1461, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 1.9524999874543483, | |
| "learning_rate": 7.635546180695038e-05, | |
| "loss": 1.0643, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 1.7260817926145178, | |
| "learning_rate": 7.490221015466279e-05, | |
| "loss": 1.1075, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 2.253848011637847, | |
| "learning_rate": 7.342042203498951e-05, | |
| "loss": 1.1714, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 1.9774736597698745, | |
| "learning_rate": 7.191179548991507e-05, | |
| "loss": 1.1076, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 2.0714596334335673, | |
| "learning_rate": 7.037805931668005e-05, | |
| "loss": 1.1452, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.7678144581225992, | |
| "learning_rate": 6.882097108668132e-05, | |
| "loss": 1.1356, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 1.709740560594397, | |
| "learning_rate": 6.724231513139852e-05, | |
| "loss": 1.1126, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 1.7902700297237863, | |
| "learning_rate": 6.564390049765528e-05, | |
| "loss": 1.0937, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 1.8180035012469915, | |
| "learning_rate": 6.402755887455792e-05, | |
| "loss": 1.1122, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 1.8387833131421265, | |
| "learning_rate": 6.239514249448767e-05, | |
| "loss": 1.1156, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 1.7206104055323197, | |
| "learning_rate": 6.0748522010551215e-05, | |
| "loss": 1.0624, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.96126387756175, | |
| "learning_rate": 5.908958435292241e-05, | |
| "loss": 1.0896, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.7505678845455834, | |
| "learning_rate": 5.742023056653131e-05, | |
| "loss": 1.1093, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 1.88819291262154, | |
| "learning_rate": 5.574237363257858e-05, | |
| "loss": 1.1057, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 1.981696800102056, | |
| "learning_rate": 5.4057936276371565e-05, | |
| "loss": 1.0809, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 2.0184048752186734, | |
| "learning_rate": 5.236884876399429e-05, | |
| "loss": 1.1245, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 1.7913247269002772, | |
| "learning_rate": 5.0677046690336096e-05, | |
| "loss": 1.1107, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.147458553314209, | |
| "eval_runtime": 38.5013, | |
| "eval_samples_per_second": 7.792, | |
| "eval_steps_per_second": 0.987, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 1.573903357064543, | |
| "learning_rate": 4.898446876101379e-05, | |
| "loss": 1.0289, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 1.886593037989885, | |
| "learning_rate": 4.729305457072913e-05, | |
| "loss": 1.0287, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 2.3850699585618855, | |
| "learning_rate": 4.560474238060739e-05, | |
| "loss": 1.0079, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 1.8897004088761564, | |
| "learning_rate": 4.392146689706425e-05, | |
| "loss": 1.0176, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 1.6637180057615024, | |
| "learning_rate": 4.224515705474603e-05, | |
| "loss": 1.0393, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 1.9082972238615958, | |
| "learning_rate": 4.057773380608411e-05, | |
| "loss": 0.9883, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 1.6248508379655304, | |
| "learning_rate": 3.892110791999649e-05, | |
| "loss": 1.0335, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 2.057742891690489, | |
| "learning_rate": 3.7277177792259114e-05, | |
| "loss": 1.0055, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 2.1082750754536725, | |
| "learning_rate": 3.5647827270055945e-05, | |
| "loss": 1.0363, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.954319411917034, | |
| "learning_rate": 3.403492349320101e-05, | |
| "loss": 1.0049, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 1.835816144902083, | |
| "learning_rate": 3.244031475450599e-05, | |
| "loss": 1.0382, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 1.905542574366058, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 1.0012, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 1.9640737297497406, | |
| "learning_rate": 2.9313268643646986e-05, | |
| "loss": 1.0183, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 1.8759801566914525, | |
| "learning_rate": 2.7784414682304832e-05, | |
| "loss": 1.0513, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 1.7782596177203436, | |
| "learning_rate": 2.628101847438835e-05, | |
| "loss": 1.0053, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 1.8683159360283232, | |
| "learning_rate": 2.4804802823479613e-05, | |
| "loss": 1.0159, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 2.234906406234719, | |
| "learning_rate": 2.3357459385841823e-05, | |
| "loss": 1.0041, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 1.8995547320155095, | |
| "learning_rate": 2.194064673188089e-05, | |
| "loss": 1.0403, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 1.8065108976199886, | |
| "learning_rate": 2.055598844552129e-05, | |
| "loss": 1.0004, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 1.7757725659621284, | |
| "learning_rate": 1.920507126367448e-05, | |
| "loss": 1.0075, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 1.7478001464098316, | |
| "learning_rate": 1.7889443257931737e-05, | |
| "loss": 1.0163, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 1.7453331826305865, | |
| "learning_rate": 1.6610612060565234e-05, | |
| "loss": 1.0246, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 2.015231393689711, | |
| "learning_rate": 1.5370043136870148e-05, | |
| "loss": 1.0339, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.123692274093628, | |
| "eval_runtime": 38.4705, | |
| "eval_samples_per_second": 7.798, | |
| "eval_steps_per_second": 0.988, | |
| "step": 349 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 464, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 1663515060338688.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |