| { | |
| "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": 3.838202126167235, | |
| "learning_rate": 9.997135147120633e-05, | |
| "loss": 2.1255, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.1177204859561205, | |
| "learning_rate": 9.98854387143534e-05, | |
| "loss": 1.7295, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.4704605855835866, | |
| "learning_rate": 9.974236018040474e-05, | |
| "loss": 1.4756, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.4984947152628136, | |
| "learning_rate": 9.954227982894034e-05, | |
| "loss": 1.3201, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.3923423053547814, | |
| "learning_rate": 9.928542694026862e-05, | |
| "loss": 1.28, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.3739605464965448, | |
| "learning_rate": 9.897209585268458e-05, | |
| "loss": 1.2543, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.275550734636967, | |
| "learning_rate": 9.86026456251757e-05, | |
| "loss": 1.2261, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 1.3278212210523785, | |
| "learning_rate": 9.817749962596115e-05, | |
| "loss": 1.1967, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 1.407416916788669, | |
| "learning_rate": 9.769714504733694e-05, | |
| "loss": 1.1876, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.4541097126165785, | |
| "learning_rate": 9.716213234738215e-05, | |
| "loss": 1.1845, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.4706662049445451, | |
| "learning_rate": 9.657307461916635e-05, | |
| "loss": 1.1681, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.3719130683126355, | |
| "learning_rate": 9.59306468881811e-05, | |
| "loss": 1.1622, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.3153375215490886, | |
| "learning_rate": 9.52355853388003e-05, | |
| "loss": 1.1298, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 1.488930807431193, | |
| "learning_rate": 9.448868647065642e-05, | |
| "loss": 1.1598, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.2649466524851178, | |
| "learning_rate": 9.369080618589864e-05, | |
| "loss": 1.135, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 1.4161938356640078, | |
| "learning_rate": 9.284285880837946e-05, | |
| "loss": 1.1325, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.418651947979571, | |
| "learning_rate": 9.194581603589328e-05, | |
| "loss": 1.1078, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.2926106311909347, | |
| "learning_rate": 9.100070582666795e-05, | |
| "loss": 1.1134, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 1.387359204219206, | |
| "learning_rate": 9.000861122138517e-05, | |
| "loss": 1.0808, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 1.5961791185416505, | |
| "learning_rate": 8.897066910207958e-05, | |
| "loss": 1.0832, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 1.476620650207239, | |
| "learning_rate": 8.788806888933881e-05, | |
| "loss": 1.0843, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.270735111137924, | |
| "learning_rate": 8.676205117929752e-05, | |
| "loss": 1.1029, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 1.425141501784203, | |
| "learning_rate": 8.559390632198723e-05, | |
| "loss": 1.101, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.1002874374389648, | |
| "eval_runtime": 69.0354, | |
| "eval_samples_per_second": 4.346, | |
| "eval_steps_per_second": 0.55, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 1.1791383340555186, | |
| "learning_rate": 8.438497294267117e-05, | |
| "loss": 1.0411, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 1.333740137623119, | |
| "learning_rate": 8.313663640785839e-05, | |
| "loss": 1.0033, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 1.4447991614503835, | |
| "learning_rate": 8.185032723775539e-05, | |
| "loss": 1.035, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 1.2844125525377348, | |
| "learning_rate": 8.052751946697403e-05, | |
| "loss": 0.9736, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 1.2700768647818643, | |
| "learning_rate": 7.916972895537471e-05, | |
| "loss": 1.0296, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 1.3154218666274633, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 1.0291, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 1.1964311130128484, | |
| "learning_rate": 7.635546180695038e-05, | |
| "loss": 0.9498, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 1.5080490468085683, | |
| "learning_rate": 7.490221015466279e-05, | |
| "loss": 0.9916, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 1.4371536477549405, | |
| "learning_rate": 7.342042203498951e-05, | |
| "loss": 1.0462, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 1.633026684473331, | |
| "learning_rate": 7.191179548991507e-05, | |
| "loss": 0.9906, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 1.5739785726895548, | |
| "learning_rate": 7.037805931668005e-05, | |
| "loss": 1.0287, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.429154502412058, | |
| "learning_rate": 6.882097108668132e-05, | |
| "loss": 1.0154, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 1.1785345522362023, | |
| "learning_rate": 6.724231513139852e-05, | |
| "loss": 1.0018, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 1.6392707933966606, | |
| "learning_rate": 6.564390049765528e-05, | |
| "loss": 0.9824, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 1.5381793424250634, | |
| "learning_rate": 6.402755887455792e-05, | |
| "loss": 0.9953, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 1.2412583778357746, | |
| "learning_rate": 6.239514249448767e-05, | |
| "loss": 0.9974, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 1.5172544424431558, | |
| "learning_rate": 6.0748522010551215e-05, | |
| "loss": 0.9531, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.3293697921192489, | |
| "learning_rate": 5.908958435292241e-05, | |
| "loss": 0.9842, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.4362543686797302, | |
| "learning_rate": 5.742023056653131e-05, | |
| "loss": 0.9993, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 1.3658639776213064, | |
| "learning_rate": 5.574237363257858e-05, | |
| "loss": 0.9936, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 1.583719578671925, | |
| "learning_rate": 5.4057936276371565e-05, | |
| "loss": 0.9686, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 1.2596980255346535, | |
| "learning_rate": 5.236884876399429e-05, | |
| "loss": 1.0095, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 1.3013094438190969, | |
| "learning_rate": 5.0677046690336096e-05, | |
| "loss": 0.999, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.0417044162750244, | |
| "eval_runtime": 68.6388, | |
| "eval_samples_per_second": 4.371, | |
| "eval_steps_per_second": 0.554, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 1.2842579006806363, | |
| "learning_rate": 4.898446876101379e-05, | |
| "loss": 0.9169, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 1.3799527209106432, | |
| "learning_rate": 4.729305457072913e-05, | |
| "loss": 0.9044, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.332914263778945, | |
| "learning_rate": 4.560474238060739e-05, | |
| "loss": 0.8905, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 1.4483104158765132, | |
| "learning_rate": 4.392146689706425e-05, | |
| "loss": 0.895, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 1.2300017765319582, | |
| "learning_rate": 4.224515705474603e-05, | |
| "loss": 0.9144, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 1.5671642418549678, | |
| "learning_rate": 4.057773380608411e-05, | |
| "loss": 0.8692, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 1.1553793609487584, | |
| "learning_rate": 3.892110791999649e-05, | |
| "loss": 0.9092, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 1.3993702614119785, | |
| "learning_rate": 3.7277177792259114e-05, | |
| "loss": 0.8869, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 1.512174508677905, | |
| "learning_rate": 3.5647827270055945e-05, | |
| "loss": 0.9104, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.4126799043823064, | |
| "learning_rate": 3.403492349320101e-05, | |
| "loss": 0.8837, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 1.3750228704249976, | |
| "learning_rate": 3.244031475450599e-05, | |
| "loss": 0.9164, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 1.5268920387463116, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.8805, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 1.4206059480865851, | |
| "learning_rate": 2.9313268643646986e-05, | |
| "loss": 0.8942, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 1.4073831093700344, | |
| "learning_rate": 2.7784414682304832e-05, | |
| "loss": 0.9289, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 1.4564634651821706, | |
| "learning_rate": 2.628101847438835e-05, | |
| "loss": 0.8864, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 1.4217201527416836, | |
| "learning_rate": 2.4804802823479613e-05, | |
| "loss": 0.8947, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 1.5100802994235183, | |
| "learning_rate": 2.3357459385841823e-05, | |
| "loss": 0.8829, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 1.4578776901613366, | |
| "learning_rate": 2.194064673188089e-05, | |
| "loss": 0.922, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 1.3183332342703558, | |
| "learning_rate": 2.055598844552129e-05, | |
| "loss": 0.8785, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 1.3169540028278115, | |
| "learning_rate": 1.920507126367448e-05, | |
| "loss": 0.8853, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 1.4997550219337101, | |
| "learning_rate": 1.7889443257931737e-05, | |
| "loss": 0.8932, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 1.2868613064278502, | |
| "learning_rate": 1.6610612060565234e-05, | |
| "loss": 0.9047, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 1.3566421396199244, | |
| "learning_rate": 1.5370043136870148e-05, | |
| "loss": 0.9106, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.0241451263427734, | |
| "eval_runtime": 68.638, | |
| "eval_samples_per_second": 4.371, | |
| "eval_steps_per_second": 0.554, | |
| "step": 349 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 464, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 2581092118822912.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |