{ "best_metric": 0.6592479348182678, "best_model_checkpoint": "V22/checkpoint-651", "epoch": 3.0, "eval_steps": 500, "global_step": 651, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.03225806451612903, "grad_norm": 109.45866394042969, "learning_rate": 1.3761467889908258e-06, "loss": 3.7822, "step": 7 }, { "epoch": 0.06451612903225806, "grad_norm": 48.60778045654297, "learning_rate": 4.587155963302753e-06, "loss": 2.8152, "step": 14 }, { "epoch": 0.0967741935483871, "grad_norm": 24.291868209838867, "learning_rate": 7.79816513761468e-06, "loss": 2.2555, "step": 21 }, { "epoch": 0.12903225806451613, "grad_norm": 66.26024627685547, "learning_rate": 1.1009174311926607e-05, "loss": 1.7015, "step": 28 }, { "epoch": 0.16129032258064516, "grad_norm": 21.380451202392578, "learning_rate": 1.4220183486238533e-05, "loss": 1.5816, "step": 35 }, { "epoch": 0.1935483870967742, "grad_norm": 33.05823516845703, "learning_rate": 1.743119266055046e-05, "loss": 0.9888, "step": 42 }, { "epoch": 0.22580645161290322, "grad_norm": 23.11123275756836, "learning_rate": 2.0642201834862388e-05, "loss": 0.9948, "step": 49 }, { "epoch": 0.25806451612903225, "grad_norm": 6.379981517791748, "learning_rate": 2.3853211009174313e-05, "loss": 1.1102, "step": 56 }, { "epoch": 0.2903225806451613, "grad_norm": 11.102519035339355, "learning_rate": 2.7064220183486238e-05, "loss": 0.8741, "step": 63 }, { "epoch": 0.3225806451612903, "grad_norm": 10.050251960754395, "learning_rate": 3.027522935779817e-05, "loss": 0.7438, "step": 70 }, { "epoch": 0.3548387096774194, "grad_norm": 5.337934494018555, "learning_rate": 3.348623853211009e-05, "loss": 0.7157, "step": 77 }, { "epoch": 0.3870967741935484, "grad_norm": 5.737290859222412, "learning_rate": 3.669724770642202e-05, "loss": 0.8164, "step": 84 }, { "epoch": 0.41935483870967744, "grad_norm": 11.127336502075195, "learning_rate": 3.990825688073395e-05, "loss": 0.7661, "step": 91 }, { "epoch": 0.45161290322580644, "grad_norm": 17.925769805908203, "learning_rate": 4.311926605504588e-05, "loss": 0.7558, "step": 98 }, { "epoch": 0.4838709677419355, "grad_norm": 23.771604537963867, "learning_rate": 4.6330275229357804e-05, "loss": 0.6493, "step": 105 }, { "epoch": 0.5161290322580645, "grad_norm": 4.815467834472656, "learning_rate": 4.954128440366973e-05, "loss": 0.766, "step": 112 }, { "epoch": 0.5483870967741935, "grad_norm": 16.425695419311523, "learning_rate": 4.969262295081967e-05, "loss": 0.8014, "step": 119 }, { "epoch": 0.5806451612903226, "grad_norm": 9.464469909667969, "learning_rate": 4.9334016393442626e-05, "loss": 0.6596, "step": 126 }, { "epoch": 0.6129032258064516, "grad_norm": 4.060314178466797, "learning_rate": 4.8975409836065575e-05, "loss": 0.6458, "step": 133 }, { "epoch": 0.6451612903225806, "grad_norm": 20.15302276611328, "learning_rate": 4.861680327868853e-05, "loss": 0.8691, "step": 140 }, { "epoch": 0.6774193548387096, "grad_norm": 11.513609886169434, "learning_rate": 4.825819672131148e-05, "loss": 0.9303, "step": 147 }, { "epoch": 0.7096774193548387, "grad_norm": 7.275956153869629, "learning_rate": 4.789959016393443e-05, "loss": 0.6554, "step": 154 }, { "epoch": 0.7419354838709677, "grad_norm": 28.995115280151367, "learning_rate": 4.754098360655738e-05, "loss": 0.7253, "step": 161 }, { "epoch": 0.7741935483870968, "grad_norm": 11.695868492126465, "learning_rate": 4.718237704918033e-05, "loss": 0.6601, "step": 168 }, { "epoch": 0.8064516129032258, "grad_norm": 11.97995376586914, "learning_rate": 4.682377049180328e-05, "loss": 0.6649, "step": 175 }, { "epoch": 0.8387096774193549, "grad_norm": 12.990303993225098, "learning_rate": 4.646516393442623e-05, "loss": 0.76, "step": 182 }, { "epoch": 0.8709677419354839, "grad_norm": 15.775096893310547, "learning_rate": 4.610655737704918e-05, "loss": 0.628, "step": 189 }, { "epoch": 0.9032258064516129, "grad_norm": 10.587272644042969, "learning_rate": 4.574795081967213e-05, "loss": 0.7037, "step": 196 }, { "epoch": 0.9354838709677419, "grad_norm": 8.605812072753906, "learning_rate": 4.5389344262295085e-05, "loss": 0.7753, "step": 203 }, { "epoch": 0.967741935483871, "grad_norm": 13.065276145935059, "learning_rate": 4.5030737704918034e-05, "loss": 0.8461, "step": 210 }, { "epoch": 1.0, "grad_norm": 10.025422096252441, "learning_rate": 4.467213114754098e-05, "loss": 0.6064, "step": 217 }, { "epoch": 1.0, "eval_accuracy": 0.6349206349206349, "eval_f1_macro": 0.5016859597651063, "eval_f1_micro": 0.6349206349206349, "eval_f1_weighted": 0.5356037461137318, "eval_loss": 0.6973989009857178, "eval_precision_macro": 0.6734914569966116, "eval_precision_micro": 0.6349206349206349, "eval_precision_weighted": 0.7076147220477118, "eval_recall_macro": 0.5572766548838334, "eval_recall_micro": 0.6349206349206349, "eval_recall_weighted": 0.6349206349206349, "eval_runtime": 0.8305, "eval_samples_per_second": 379.279, "eval_steps_per_second": 24.081, "step": 217 }, { "epoch": 1.032258064516129, "grad_norm": 12.027962684631348, "learning_rate": 4.431352459016393e-05, "loss": 0.7945, "step": 224 }, { "epoch": 1.064516129032258, "grad_norm": 9.709089279174805, "learning_rate": 4.395491803278689e-05, "loss": 0.4639, "step": 231 }, { "epoch": 1.096774193548387, "grad_norm": 12.356049537658691, "learning_rate": 4.359631147540984e-05, "loss": 0.6228, "step": 238 }, { "epoch": 1.129032258064516, "grad_norm": 15.807150840759277, "learning_rate": 4.323770491803279e-05, "loss": 0.6084, "step": 245 }, { "epoch": 1.1612903225806452, "grad_norm": 19.665773391723633, "learning_rate": 4.287909836065574e-05, "loss": 0.6107, "step": 252 }, { "epoch": 1.1935483870967742, "grad_norm": 8.455767631530762, "learning_rate": 4.252049180327869e-05, "loss": 0.4523, "step": 259 }, { "epoch": 1.2258064516129032, "grad_norm": 15.356058120727539, "learning_rate": 4.216188524590164e-05, "loss": 0.4767, "step": 266 }, { "epoch": 1.2580645161290323, "grad_norm": 8.633644104003906, "learning_rate": 4.1803278688524595e-05, "loss": 0.5657, "step": 273 }, { "epoch": 1.2903225806451613, "grad_norm": 9.788046836853027, "learning_rate": 4.1444672131147544e-05, "loss": 0.5636, "step": 280 }, { "epoch": 1.3225806451612903, "grad_norm": 14.03018856048584, "learning_rate": 4.108606557377049e-05, "loss": 0.7493, "step": 287 }, { "epoch": 1.3548387096774195, "grad_norm": 8.021461486816406, "learning_rate": 4.072745901639344e-05, "loss": 0.6374, "step": 294 }, { "epoch": 1.3870967741935485, "grad_norm": 36.801265716552734, "learning_rate": 4.036885245901639e-05, "loss": 0.833, "step": 301 }, { "epoch": 1.4193548387096775, "grad_norm": 9.589942932128906, "learning_rate": 4.001024590163935e-05, "loss": 0.4758, "step": 308 }, { "epoch": 1.4516129032258065, "grad_norm": 23.219730377197266, "learning_rate": 3.9651639344262296e-05, "loss": 0.6026, "step": 315 }, { "epoch": 1.4838709677419355, "grad_norm": 5.934767723083496, "learning_rate": 3.9293032786885245e-05, "loss": 0.5295, "step": 322 }, { "epoch": 1.5161290322580645, "grad_norm": 20.14740562438965, "learning_rate": 3.89344262295082e-05, "loss": 0.6674, "step": 329 }, { "epoch": 1.5483870967741935, "grad_norm": 29.94481658935547, "learning_rate": 3.857581967213115e-05, "loss": 0.5743, "step": 336 }, { "epoch": 1.5806451612903225, "grad_norm": 4.266819953918457, "learning_rate": 3.8217213114754106e-05, "loss": 0.5186, "step": 343 }, { "epoch": 1.6129032258064515, "grad_norm": 17.77097511291504, "learning_rate": 3.7858606557377055e-05, "loss": 0.5217, "step": 350 }, { "epoch": 1.6451612903225805, "grad_norm": 17.18390464782715, "learning_rate": 3.7500000000000003e-05, "loss": 0.4922, "step": 357 }, { "epoch": 1.6774193548387095, "grad_norm": 25.000282287597656, "learning_rate": 3.714139344262295e-05, "loss": 0.5006, "step": 364 }, { "epoch": 1.7096774193548387, "grad_norm": 6.265079021453857, "learning_rate": 3.67827868852459e-05, "loss": 0.5437, "step": 371 }, { "epoch": 1.7419354838709677, "grad_norm": 27.180173873901367, "learning_rate": 3.642418032786886e-05, "loss": 0.491, "step": 378 }, { "epoch": 1.7741935483870968, "grad_norm": 11.424750328063965, "learning_rate": 3.6065573770491806e-05, "loss": 0.4924, "step": 385 }, { "epoch": 1.8064516129032258, "grad_norm": 7.653794288635254, "learning_rate": 3.5706967213114755e-05, "loss": 0.5612, "step": 392 }, { "epoch": 1.838709677419355, "grad_norm": 7.977573394775391, "learning_rate": 3.5348360655737704e-05, "loss": 0.7469, "step": 399 }, { "epoch": 1.870967741935484, "grad_norm": 6.503817558288574, "learning_rate": 3.498975409836065e-05, "loss": 0.447, "step": 406 }, { "epoch": 1.903225806451613, "grad_norm": 53.40876388549805, "learning_rate": 3.463114754098361e-05, "loss": 0.5352, "step": 413 }, { "epoch": 1.935483870967742, "grad_norm": 10.495361328125, "learning_rate": 3.427254098360656e-05, "loss": 0.496, "step": 420 }, { "epoch": 1.967741935483871, "grad_norm": 17.43389320373535, "learning_rate": 3.391393442622951e-05, "loss": 0.506, "step": 427 }, { "epoch": 2.0, "grad_norm": 18.6962833404541, "learning_rate": 3.355532786885246e-05, "loss": 0.5325, "step": 434 }, { "epoch": 2.0, "eval_accuracy": 0.7206349206349206, "eval_f1_macro": 0.7194165458459646, "eval_f1_micro": 0.7206349206349206, "eval_f1_weighted": 0.723887518541532, "eval_loss": 0.6725990176200867, "eval_precision_macro": 0.6923798008747183, "eval_precision_micro": 0.7206349206349206, "eval_precision_weighted": 0.7642552562350153, "eval_recall_macro": 0.7907399015075983, "eval_recall_micro": 0.7206349206349206, "eval_recall_weighted": 0.7206349206349206, "eval_runtime": 0.8273, "eval_samples_per_second": 380.77, "eval_steps_per_second": 24.176, "step": 434 }, { "epoch": 2.032258064516129, "grad_norm": 15.976861953735352, "learning_rate": 3.319672131147541e-05, "loss": 0.3418, "step": 441 }, { "epoch": 2.064516129032258, "grad_norm": 12.423666000366211, "learning_rate": 3.283811475409837e-05, "loss": 0.539, "step": 448 }, { "epoch": 2.096774193548387, "grad_norm": 4.132760524749756, "learning_rate": 3.2479508196721316e-05, "loss": 0.4235, "step": 455 }, { "epoch": 2.129032258064516, "grad_norm": 7.873523712158203, "learning_rate": 3.2120901639344265e-05, "loss": 0.3752, "step": 462 }, { "epoch": 2.161290322580645, "grad_norm": 16.930147171020508, "learning_rate": 3.1762295081967214e-05, "loss": 0.3584, "step": 469 }, { "epoch": 2.193548387096774, "grad_norm": 13.781448364257812, "learning_rate": 3.1403688524590163e-05, "loss": 0.3122, "step": 476 }, { "epoch": 2.225806451612903, "grad_norm": 3.406670093536377, "learning_rate": 3.104508196721312e-05, "loss": 0.3588, "step": 483 }, { "epoch": 2.258064516129032, "grad_norm": 25.444080352783203, "learning_rate": 3.068647540983607e-05, "loss": 0.3171, "step": 490 }, { "epoch": 2.2903225806451615, "grad_norm": 4.011606216430664, "learning_rate": 3.0327868852459017e-05, "loss": 0.4139, "step": 497 }, { "epoch": 2.3225806451612905, "grad_norm": 25.006410598754883, "learning_rate": 2.996926229508197e-05, "loss": 0.4065, "step": 504 }, { "epoch": 2.3548387096774195, "grad_norm": 25.925764083862305, "learning_rate": 2.961065573770492e-05, "loss": 0.5893, "step": 511 }, { "epoch": 2.3870967741935485, "grad_norm": 62.814369201660156, "learning_rate": 2.9252049180327874e-05, "loss": 0.3352, "step": 518 }, { "epoch": 2.4193548387096775, "grad_norm": 7.263156890869141, "learning_rate": 2.8893442622950823e-05, "loss": 0.3747, "step": 525 }, { "epoch": 2.4516129032258065, "grad_norm": 17.882831573486328, "learning_rate": 2.8534836065573772e-05, "loss": 0.2106, "step": 532 }, { "epoch": 2.4838709677419355, "grad_norm": 58.87062072753906, "learning_rate": 2.817622950819672e-05, "loss": 0.4129, "step": 539 }, { "epoch": 2.5161290322580645, "grad_norm": 18.514890670776367, "learning_rate": 2.781762295081967e-05, "loss": 0.6121, "step": 546 }, { "epoch": 2.5483870967741935, "grad_norm": 48.30098342895508, "learning_rate": 2.7459016393442626e-05, "loss": 0.2255, "step": 553 }, { "epoch": 2.5806451612903225, "grad_norm": 1.9256900548934937, "learning_rate": 2.7100409836065575e-05, "loss": 0.2786, "step": 560 }, { "epoch": 2.6129032258064515, "grad_norm": 36.61083984375, "learning_rate": 2.6741803278688527e-05, "loss": 0.3542, "step": 567 }, { "epoch": 2.6451612903225805, "grad_norm": 12.661158561706543, "learning_rate": 2.6383196721311476e-05, "loss": 0.4469, "step": 574 }, { "epoch": 2.6774193548387095, "grad_norm": 40.683265686035156, "learning_rate": 2.6024590163934425e-05, "loss": 0.3134, "step": 581 }, { "epoch": 2.709677419354839, "grad_norm": 42.47820281982422, "learning_rate": 2.5665983606557374e-05, "loss": 0.3625, "step": 588 }, { "epoch": 2.741935483870968, "grad_norm": 21.992469787597656, "learning_rate": 2.530737704918033e-05, "loss": 0.3179, "step": 595 }, { "epoch": 2.774193548387097, "grad_norm": 31.075956344604492, "learning_rate": 2.494877049180328e-05, "loss": 0.7084, "step": 602 }, { "epoch": 2.806451612903226, "grad_norm": 11.713536262512207, "learning_rate": 2.459016393442623e-05, "loss": 0.5862, "step": 609 }, { "epoch": 2.838709677419355, "grad_norm": 10.594955444335938, "learning_rate": 2.4231557377049184e-05, "loss": 0.3564, "step": 616 }, { "epoch": 2.870967741935484, "grad_norm": 12.713808059692383, "learning_rate": 2.3872950819672133e-05, "loss": 0.465, "step": 623 }, { "epoch": 2.903225806451613, "grad_norm": 2.1826462745666504, "learning_rate": 2.3514344262295082e-05, "loss": 0.3214, "step": 630 }, { "epoch": 2.935483870967742, "grad_norm": 10.218241691589355, "learning_rate": 2.3155737704918034e-05, "loss": 0.3046, "step": 637 }, { "epoch": 2.967741935483871, "grad_norm": 3.6607110500335693, "learning_rate": 2.2797131147540983e-05, "loss": 0.5019, "step": 644 }, { "epoch": 3.0, "grad_norm": 6.482827663421631, "learning_rate": 2.2438524590163936e-05, "loss": 0.3717, "step": 651 }, { "epoch": 3.0, "eval_accuracy": 0.7365079365079366, "eval_f1_macro": 0.7200544867211534, "eval_f1_micro": 0.7365079365079366, "eval_f1_weighted": 0.7401985275001148, "eval_loss": 0.6592479348182678, "eval_precision_macro": 0.7198264429452804, "eval_precision_micro": 0.7365079365079366, "eval_precision_weighted": 0.7519327133494232, "eval_recall_macro": 0.7261276775733405, "eval_recall_micro": 0.7365079365079366, "eval_recall_weighted": 0.7365079365079366, "eval_runtime": 0.8246, "eval_samples_per_second": 381.986, "eval_steps_per_second": 24.253, "step": 651 } ], "logging_steps": 7, "max_steps": 1085, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 500, "total_flos": 341784329628672.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }