| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 2080, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.009615384615384616, | |
| "grad_norm": 9.013119895723426, | |
| "learning_rate": 4.807692307692308e-07, | |
| "loss": 1.3011, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.019230769230769232, | |
| "grad_norm": 8.170199017296364, | |
| "learning_rate": 9.615384615384617e-07, | |
| "loss": 1.1715, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.028846153846153848, | |
| "grad_norm": 5.232892874296514, | |
| "learning_rate": 1.4423076923076922e-06, | |
| "loss": 0.9015, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.038461538461538464, | |
| "grad_norm": 2.862322298571592, | |
| "learning_rate": 1.9230769230769234e-06, | |
| "loss": 0.7978, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04807692307692308, | |
| "grad_norm": 2.7249762448136132, | |
| "learning_rate": 2.403846153846154e-06, | |
| "loss": 0.7609, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.057692307692307696, | |
| "grad_norm": 3.169444209352466, | |
| "learning_rate": 2.8846153846153845e-06, | |
| "loss": 0.8146, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0673076923076923, | |
| "grad_norm": 2.6189394700073807, | |
| "learning_rate": 3.365384615384616e-06, | |
| "loss": 0.777, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 2.4397555428298077, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 0.78, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.08653846153846154, | |
| "grad_norm": 2.812934479310922, | |
| "learning_rate": 4.326923076923077e-06, | |
| "loss": 0.7721, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.09615384615384616, | |
| "grad_norm": 2.5334288416130746, | |
| "learning_rate": 4.807692307692308e-06, | |
| "loss": 0.7669, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10576923076923077, | |
| "grad_norm": 2.342690607930004, | |
| "learning_rate": 5.288461538461539e-06, | |
| "loss": 0.7459, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.11538461538461539, | |
| "grad_norm": 2.3269599095205264, | |
| "learning_rate": 5.769230769230769e-06, | |
| "loss": 0.7851, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 2.42709508438667, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.7703, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1346153846153846, | |
| "grad_norm": 2.6377643494039273, | |
| "learning_rate": 6.730769230769232e-06, | |
| "loss": 0.7496, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.14423076923076922, | |
| "grad_norm": 2.875686037935072, | |
| "learning_rate": 7.211538461538462e-06, | |
| "loss": 0.784, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 2.1768155795174917, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.8022, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.16346153846153846, | |
| "grad_norm": 2.46654983618583, | |
| "learning_rate": 8.173076923076923e-06, | |
| "loss": 0.7741, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.17307692307692307, | |
| "grad_norm": 2.4450582546281554, | |
| "learning_rate": 8.653846153846155e-06, | |
| "loss": 0.7827, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.18269230769230768, | |
| "grad_norm": 2.4657718631331567, | |
| "learning_rate": 9.134615384615384e-06, | |
| "loss": 0.7894, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.19230769230769232, | |
| "grad_norm": 2.54311809351163, | |
| "learning_rate": 9.615384615384616e-06, | |
| "loss": 0.8164, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.20192307692307693, | |
| "grad_norm": 2.097681181559205, | |
| "learning_rate": 9.999971836433636e-06, | |
| "loss": 0.8139, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.21153846153846154, | |
| "grad_norm": 2.2579921721680516, | |
| "learning_rate": 9.998986144924253e-06, | |
| "loss": 0.7919, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.22115384615384615, | |
| "grad_norm": 2.1097886183574466, | |
| "learning_rate": 9.996592592355083e-06, | |
| "loss": 0.8056, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 2.22507095955906, | |
| "learning_rate": 9.992791852820709e-06, | |
| "loss": 0.8108, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2403846153846154, | |
| "grad_norm": 2.1931221071365425, | |
| "learning_rate": 9.987584996720813e-06, | |
| "loss": 0.8041, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 2.1220375928827044, | |
| "learning_rate": 9.980973490458728e-06, | |
| "loss": 0.8149, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.25961538461538464, | |
| "grad_norm": 2.1068957326041904, | |
| "learning_rate": 9.972959196028456e-06, | |
| "loss": 0.7914, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2692307692307692, | |
| "grad_norm": 1.9372196057401814, | |
| "learning_rate": 9.96354437049027e-06, | |
| "loss": 0.8033, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.27884615384615385, | |
| "grad_norm": 2.2606013253960846, | |
| "learning_rate": 9.952731665335071e-06, | |
| "loss": 0.8039, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.28846153846153844, | |
| "grad_norm": 2.1614476516877406, | |
| "learning_rate": 9.940524125737641e-06, | |
| "loss": 0.7647, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2980769230769231, | |
| "grad_norm": 2.220222508508704, | |
| "learning_rate": 9.92692518969903e-06, | |
| "loss": 0.7901, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 2.201097779810873, | |
| "learning_rate": 9.911938687078324e-06, | |
| "loss": 0.7964, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.3173076923076923, | |
| "grad_norm": 2.029609328977134, | |
| "learning_rate": 9.895568838514042e-06, | |
| "loss": 0.7764, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3269230769230769, | |
| "grad_norm": 2.2142291176229834, | |
| "learning_rate": 9.87782025423547e-06, | |
| "loss": 0.8005, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.33653846153846156, | |
| "grad_norm": 2.084315605131572, | |
| "learning_rate": 9.8586979327643e-06, | |
| "loss": 0.7795, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.34615384615384615, | |
| "grad_norm": 2.3024285135825457, | |
| "learning_rate": 9.838207259506891e-06, | |
| "loss": 0.7935, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3557692307692308, | |
| "grad_norm": 2.5288625621043326, | |
| "learning_rate": 9.816354005237583e-06, | |
| "loss": 0.7749, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.36538461538461536, | |
| "grad_norm": 2.0010687704613743, | |
| "learning_rate": 9.793144324473473e-06, | |
| "loss": 0.7988, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 2.1536447186627496, | |
| "learning_rate": 9.768584753741134e-06, | |
| "loss": 0.7955, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 2.025116331685143, | |
| "learning_rate": 9.742682209735727e-06, | |
| "loss": 0.8014, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3942307692307692, | |
| "grad_norm": 2.027340612157234, | |
| "learning_rate": 9.715443987373062e-06, | |
| "loss": 0.8341, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.40384615384615385, | |
| "grad_norm": 1.968065863088579, | |
| "learning_rate": 9.686877757735126e-06, | |
| "loss": 0.7922, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.41346153846153844, | |
| "grad_norm": 2.016993760012052, | |
| "learning_rate": 9.656991565909703e-06, | |
| "loss": 0.8328, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4230769230769231, | |
| "grad_norm": 2.1902508822456213, | |
| "learning_rate": 9.62579382872462e-06, | |
| "loss": 0.8029, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4326923076923077, | |
| "grad_norm": 2.0187820173893534, | |
| "learning_rate": 9.593293332377325e-06, | |
| "loss": 0.7634, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.4423076923076923, | |
| "grad_norm": 2.1364408496498437, | |
| "learning_rate": 9.55949922996045e-06, | |
| "loss": 0.774, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.4519230769230769, | |
| "grad_norm": 2.0116381073985945, | |
| "learning_rate": 9.52442103888402e-06, | |
| "loss": 0.8331, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 2.100311438402205, | |
| "learning_rate": 9.488068638195072e-06, | |
| "loss": 0.7478, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.47115384615384615, | |
| "grad_norm": 2.059472131869095, | |
| "learning_rate": 9.450452265795423e-06, | |
| "loss": 0.8202, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4807692307692308, | |
| "grad_norm": 2.074026027178722, | |
| "learning_rate": 9.411582515558391e-06, | |
| "loss": 0.8177, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4807692307692308, | |
| "eval_loss": 0.7822802066802979, | |
| "eval_runtime": 31.3783, | |
| "eval_samples_per_second": 58.926, | |
| "eval_steps_per_second": 7.394, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.49038461538461536, | |
| "grad_norm": 1.9938002782487565, | |
| "learning_rate": 9.371470334345232e-06, | |
| "loss": 0.804, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.9596865504803214, | |
| "learning_rate": 9.330127018922195e-06, | |
| "loss": 0.7844, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5096153846153846, | |
| "grad_norm": 1.8726235477694055, | |
| "learning_rate": 9.287564212779012e-06, | |
| "loss": 0.756, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.5192307692307693, | |
| "grad_norm": 2.091468778530183, | |
| "learning_rate": 9.243793902849764e-06, | |
| "loss": 0.8057, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5288461538461539, | |
| "grad_norm": 2.032448528673805, | |
| "learning_rate": 9.198828416136991e-06, | |
| "loss": 0.8065, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 2.0648947345620137, | |
| "learning_rate": 9.152680416240059e-06, | |
| "loss": 0.8101, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5480769230769231, | |
| "grad_norm": 1.864661141720549, | |
| "learning_rate": 9.10536289978872e-06, | |
| "loss": 0.7597, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5576923076923077, | |
| "grad_norm": 1.8389259729678664, | |
| "learning_rate": 9.056889192782865e-06, | |
| "loss": 0.806, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5673076923076923, | |
| "grad_norm": 1.997501225318945, | |
| "learning_rate": 9.007272946839559e-06, | |
| "loss": 0.7696, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5769230769230769, | |
| "grad_norm": 1.9510249081212931, | |
| "learning_rate": 8.95652813534831e-06, | |
| "loss": 0.7779, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5865384615384616, | |
| "grad_norm": 2.043423090422744, | |
| "learning_rate": 8.90466904953579e-06, | |
| "loss": 0.8142, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5961538461538461, | |
| "grad_norm": 2.1520676063407, | |
| "learning_rate": 8.851710294440974e-06, | |
| "loss": 0.804, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6057692307692307, | |
| "grad_norm": 1.8353482991425896, | |
| "learning_rate": 8.797666784801954e-06, | |
| "loss": 0.7701, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 1.8647174071179153, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 0.7641, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 1.859731623435647, | |
| "learning_rate": 8.68638668405062e-06, | |
| "loss": 0.8293, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.6346153846153846, | |
| "grad_norm": 2.035763511676626, | |
| "learning_rate": 8.629181432677213e-06, | |
| "loss": 0.8386, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6442307692307693, | |
| "grad_norm": 1.9498309927555528, | |
| "learning_rate": 8.570954097411224e-06, | |
| "loss": 0.826, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.6538461538461539, | |
| "grad_norm": 2.157631582471241, | |
| "learning_rate": 8.511721076777388e-06, | |
| "loss": 0.7933, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.6634615384615384, | |
| "grad_norm": 1.817540210529388, | |
| "learning_rate": 8.451499052530923e-06, | |
| "loss": 0.794, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.6730769230769231, | |
| "grad_norm": 2.003699550055361, | |
| "learning_rate": 8.390304984959455e-06, | |
| "loss": 0.8152, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6826923076923077, | |
| "grad_norm": 1.7853207622424534, | |
| "learning_rate": 8.328156108106518e-06, | |
| "loss": 0.7869, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 1.951815011854146, | |
| "learning_rate": 8.265069924917925e-06, | |
| "loss": 0.8086, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7019230769230769, | |
| "grad_norm": 1.904892870037223, | |
| "learning_rate": 8.20106420231244e-06, | |
| "loss": 0.7959, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.7115384615384616, | |
| "grad_norm": 1.9417662866059235, | |
| "learning_rate": 8.136156966178082e-06, | |
| "loss": 0.8072, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.7211538461538461, | |
| "grad_norm": 1.8480817149859605, | |
| "learning_rate": 8.070366496295505e-06, | |
| "loss": 0.7814, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.7307692307692307, | |
| "grad_norm": 1.8327729681591678, | |
| "learning_rate": 8.003711321189895e-06, | |
| "loss": 0.7786, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.7403846153846154, | |
| "grad_norm": 2.1198206040027494, | |
| "learning_rate": 7.93621021291277e-06, | |
| "loss": 0.7485, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.9784614334378419, | |
| "learning_rate": 7.86788218175523e-06, | |
| "loss": 0.7736, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.7596153846153846, | |
| "grad_norm": 1.733613545143569, | |
| "learning_rate": 7.798746470894113e-06, | |
| "loss": 0.7613, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 1.7905885212318795, | |
| "learning_rate": 7.728822550972523e-06, | |
| "loss": 0.7902, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7788461538461539, | |
| "grad_norm": 2.02868421572116, | |
| "learning_rate": 7.658130114616364e-06, | |
| "loss": 0.7962, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.7884615384615384, | |
| "grad_norm": 1.8163329758157731, | |
| "learning_rate": 7.586689070888284e-06, | |
| "loss": 0.7625, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.7980769230769231, | |
| "grad_norm": 2.0352374441953542, | |
| "learning_rate": 7.5145195396807244e-06, | |
| "loss": 0.7836, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8076923076923077, | |
| "grad_norm": 2.0684855751019144, | |
| "learning_rate": 7.441641846049557e-06, | |
| "loss": 0.8016, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.8173076923076923, | |
| "grad_norm": 2.042151868500934, | |
| "learning_rate": 7.368076514489947e-06, | |
| "loss": 0.8012, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.8269230769230769, | |
| "grad_norm": 1.9253750246288859, | |
| "learning_rate": 7.2938442631560714e-06, | |
| "loss": 0.799, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.8365384615384616, | |
| "grad_norm": 1.8896143274467088, | |
| "learning_rate": 7.218965998026259e-06, | |
| "loss": 0.8099, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 1.7590979198633765, | |
| "learning_rate": 7.143462807015271e-06, | |
| "loss": 0.7726, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.8557692307692307, | |
| "grad_norm": 1.9332789610600787, | |
| "learning_rate": 7.067355954035316e-06, | |
| "loss": 0.808, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.8653846153846154, | |
| "grad_norm": 1.8265190820741775, | |
| "learning_rate": 6.990666873007506e-06, | |
| "loss": 0.7604, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 1.8486951872238018, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.7861, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.8846153846153846, | |
| "grad_norm": 2.0473377615419412, | |
| "learning_rate": 6.8356285762726385e-06, | |
| "loss": 0.7618, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.8942307692307693, | |
| "grad_norm": 1.7908009618022396, | |
| "learning_rate": 6.757323023895388e-06, | |
| "loss": 0.7547, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9038461538461539, | |
| "grad_norm": 1.7499349918180362, | |
| "learning_rate": 6.678522557833025e-06, | |
| "loss": 0.7595, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.9134615384615384, | |
| "grad_norm": 1.9193288351945685, | |
| "learning_rate": 6.599249370607083e-06, | |
| "loss": 0.7465, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 1.7382141318072688, | |
| "learning_rate": 6.519525787871235e-06, | |
| "loss": 0.7829, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.9326923076923077, | |
| "grad_norm": 1.6639815518816288, | |
| "learning_rate": 6.439374262123731e-06, | |
| "loss": 0.7483, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.9423076923076923, | |
| "grad_norm": 2.040809216745246, | |
| "learning_rate": 6.358817366384122e-06, | |
| "loss": 0.7695, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.9519230769230769, | |
| "grad_norm": 1.9765992795556684, | |
| "learning_rate": 6.277877787836034e-06, | |
| "loss": 0.8039, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.9615384615384616, | |
| "grad_norm": 1.7379165898055615, | |
| "learning_rate": 6.1965783214377895e-06, | |
| "loss": 0.7708, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9615384615384616, | |
| "eval_loss": 0.757188618183136, | |
| "eval_runtime": 30.718, | |
| "eval_samples_per_second": 60.193, | |
| "eval_steps_per_second": 7.553, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9711538461538461, | |
| "grad_norm": 1.91002204642886, | |
| "learning_rate": 6.114941863502682e-06, | |
| "loss": 0.791, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.9807692307692307, | |
| "grad_norm": 1.7547449600143168, | |
| "learning_rate": 6.032991405250702e-06, | |
| "loss": 0.7584, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.9903846153846154, | |
| "grad_norm": 1.7518146591240198, | |
| "learning_rate": 5.950750026333534e-06, | |
| "loss": 0.7911, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.8856712522978225, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.7762, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.0096153846153846, | |
| "grad_norm": 2.6930087290136417, | |
| "learning_rate": 5.785487228246339e-06, | |
| "loss": 0.6172, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.0192307692307692, | |
| "grad_norm": 2.0128509132494217, | |
| "learning_rate": 5.7025123519254644e-06, | |
| "loss": 0.5335, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.0288461538461537, | |
| "grad_norm": 2.081106502499814, | |
| "learning_rate": 5.619339627529876e-06, | |
| "loss": 0.5475, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.0384615384615385, | |
| "grad_norm": 2.089872779664623, | |
| "learning_rate": 5.53599247893724e-06, | |
| "loss": 0.5491, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.0480769230769231, | |
| "grad_norm": 2.487028007682742, | |
| "learning_rate": 5.45249437914819e-06, | |
| "loss": 0.5396, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.0576923076923077, | |
| "grad_norm": 2.1183702883792015, | |
| "learning_rate": 5.368868843675642e-06, | |
| "loss": 0.5602, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.0673076923076923, | |
| "grad_norm": 2.231623903215653, | |
| "learning_rate": 5.285139423922138e-06, | |
| "loss": 0.5341, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 1.8590648999967128, | |
| "learning_rate": 5.201329700547077e-06, | |
| "loss": 0.5499, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.0865384615384615, | |
| "grad_norm": 2.031456951588171, | |
| "learning_rate": 5.117463276825711e-06, | |
| "loss": 0.5541, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.0961538461538463, | |
| "grad_norm": 1.996025281675884, | |
| "learning_rate": 5.033563772001782e-06, | |
| "loss": 0.5229, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.1057692307692308, | |
| "grad_norm": 2.01803070793536, | |
| "learning_rate": 4.949654814635623e-06, | |
| "loss": 0.5534, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.1153846153846154, | |
| "grad_norm": 2.0131103606600163, | |
| "learning_rate": 4.865760035949695e-06, | |
| "loss": 0.5503, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.125, | |
| "grad_norm": 1.9109776540155083, | |
| "learning_rate": 4.781903063173321e-06, | |
| "loss": 0.5587, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.1346153846153846, | |
| "grad_norm": 2.240195731738998, | |
| "learning_rate": 4.69810751288857e-06, | |
| "loss": 0.5345, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.1442307692307692, | |
| "grad_norm": 2.21815291482556, | |
| "learning_rate": 4.61439698437914e-06, | |
| "loss": 0.5105, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 1.959424085301074, | |
| "learning_rate": 4.530795052984104e-06, | |
| "loss": 0.5472, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.1634615384615385, | |
| "grad_norm": 1.9724626057174217, | |
| "learning_rate": 4.447325263458401e-06, | |
| "loss": 0.5425, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.1730769230769231, | |
| "grad_norm": 1.9404277504052594, | |
| "learning_rate": 4.364011123341947e-06, | |
| "loss": 0.5383, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.1826923076923077, | |
| "grad_norm": 1.975957154057773, | |
| "learning_rate": 4.280876096339222e-06, | |
| "loss": 0.5635, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.1923076923076923, | |
| "grad_norm": 3.7893846099890203, | |
| "learning_rate": 4.1979435957111984e-06, | |
| "loss": 0.5508, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.2019230769230769, | |
| "grad_norm": 2.088379071314605, | |
| "learning_rate": 4.11523697768149e-06, | |
| "loss": 0.5917, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.2115384615384615, | |
| "grad_norm": 2.0699986543823914, | |
| "learning_rate": 4.032779534858544e-06, | |
| "loss": 0.5549, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.2211538461538463, | |
| "grad_norm": 2.4531750427167744, | |
| "learning_rate": 3.9505944896757635e-06, | |
| "loss": 0.5314, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 2.130560765169762, | |
| "learning_rate": 3.86870498785139e-06, | |
| "loss": 0.5327, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.2403846153846154, | |
| "grad_norm": 2.2163755882504947, | |
| "learning_rate": 3.7871340918699945e-06, | |
| "loss": 0.509, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 2.1107171131060136, | |
| "learning_rate": 3.705904774487396e-06, | |
| "loss": 0.5153, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.2596153846153846, | |
| "grad_norm": 2.1786622272309413, | |
| "learning_rate": 3.6250399122608713e-06, | |
| "loss": 0.5589, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.2692307692307692, | |
| "grad_norm": 1.9962608207111503, | |
| "learning_rate": 3.544562279106436e-06, | |
| "loss": 0.5425, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.2788461538461537, | |
| "grad_norm": 1.8464394968329445, | |
| "learning_rate": 3.464494539885047e-06, | |
| "loss": 0.5349, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.2884615384615383, | |
| "grad_norm": 2.138267726845245, | |
| "learning_rate": 3.3848592440195118e-06, | |
| "loss": 0.5461, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.2980769230769231, | |
| "grad_norm": 2.0087619862264767, | |
| "learning_rate": 3.3056788191439116e-06, | |
| "loss": 0.559, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 2.1318708987886854, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 0.5326, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.3173076923076923, | |
| "grad_norm": 2.0595039744355725, | |
| "learning_rate": 3.148771646093608e-06, | |
| "loss": 0.5215, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.3269230769230769, | |
| "grad_norm": 2.106731128516135, | |
| "learning_rate": 3.0710890875790745e-06, | |
| "loss": 0.5205, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.3365384615384617, | |
| "grad_norm": 2.015598237138461, | |
| "learning_rate": 2.993949766929711e-06, | |
| "loss": 0.5155, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.3461538461538463, | |
| "grad_norm": 2.089932096970866, | |
| "learning_rate": 2.917375408839803e-06, | |
| "loss": 0.5321, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.3557692307692308, | |
| "grad_norm": 2.1011121825872414, | |
| "learning_rate": 2.8413875788936067e-06, | |
| "loss": 0.5654, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.3653846153846154, | |
| "grad_norm": 2.5049128653170607, | |
| "learning_rate": 2.766007677491871e-06, | |
| "loss": 0.5332, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.375, | |
| "grad_norm": 2.199796386555421, | |
| "learning_rate": 2.6912569338248317e-06, | |
| "loss": 0.5129, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 2.122643085103428, | |
| "learning_rate": 2.6171563998934605e-06, | |
| "loss": 0.5315, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.3942307692307692, | |
| "grad_norm": 1.9680294145904305, | |
| "learning_rate": 2.5437269445806146e-06, | |
| "loss": 0.5388, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.4038461538461537, | |
| "grad_norm": 2.1649286481102505, | |
| "learning_rate": 2.4709892477737263e-06, | |
| "loss": 0.5449, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.4134615384615383, | |
| "grad_norm": 1.9108821348630196, | |
| "learning_rate": 2.3989637945407547e-06, | |
| "loss": 0.5124, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.4230769230769231, | |
| "grad_norm": 1.8705842302466216, | |
| "learning_rate": 2.3276708693609947e-06, | |
| "loss": 0.5271, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.4326923076923077, | |
| "grad_norm": 2.218892686030775, | |
| "learning_rate": 2.2571305504123547e-06, | |
| "loss": 0.536, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.4423076923076923, | |
| "grad_norm": 2.151762003437118, | |
| "learning_rate": 2.187362703916766e-06, | |
| "loss": 0.5513, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.4423076923076923, | |
| "eval_loss": 0.7693456411361694, | |
| "eval_runtime": 31.5006, | |
| "eval_samples_per_second": 58.697, | |
| "eval_steps_per_second": 7.365, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.4519230769230769, | |
| "grad_norm": 2.2335086500452768, | |
| "learning_rate": 2.1183869785452744e-06, | |
| "loss": 0.5343, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 2.201272848216364, | |
| "learning_rate": 2.050222799884387e-06, | |
| "loss": 0.5203, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.4711538461538463, | |
| "grad_norm": 2.245522812744825, | |
| "learning_rate": 1.9828893649652653e-06, | |
| "loss": 0.5492, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.4807692307692308, | |
| "grad_norm": 2.4219731931643134, | |
| "learning_rate": 1.9164056368572847e-06, | |
| "loss": 0.531, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.4903846153846154, | |
| "grad_norm": 1.9852458155211052, | |
| "learning_rate": 1.8507903393274622e-06, | |
| "loss": 0.5413, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 2.057532782936977, | |
| "learning_rate": 1.7860619515673034e-06, | |
| "loss": 0.5271, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.5096153846153846, | |
| "grad_norm": 2.3471986544512187, | |
| "learning_rate": 1.7222387029885268e-06, | |
| "loss": 0.5579, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.5192307692307692, | |
| "grad_norm": 2.12927805118844, | |
| "learning_rate": 1.6593385680891139e-06, | |
| "loss": 0.524, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.5288461538461537, | |
| "grad_norm": 2.3452167989714785, | |
| "learning_rate": 1.5973792613911698e-06, | |
| "loss": 0.5263, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 2.3303569386029115, | |
| "learning_rate": 1.5363782324520033e-06, | |
| "loss": 0.5075, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.5480769230769231, | |
| "grad_norm": 2.3477866769254514, | |
| "learning_rate": 1.476352660949802e-06, | |
| "loss": 0.5611, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.5576923076923077, | |
| "grad_norm": 1.9831558221073002, | |
| "learning_rate": 1.4173194518453415e-06, | |
| "loss": 0.5124, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.5673076923076923, | |
| "grad_norm": 2.5946859927202732, | |
| "learning_rate": 1.3592952306210589e-06, | |
| "loss": 0.5193, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.5769230769230769, | |
| "grad_norm": 2.3095838494615433, | |
| "learning_rate": 1.3022963385988153e-06, | |
| "loss": 0.5374, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.5865384615384617, | |
| "grad_norm": 2.0028485071747806, | |
| "learning_rate": 1.246338828337707e-06, | |
| "loss": 0.512, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.5961538461538463, | |
| "grad_norm": 2.029535183303286, | |
| "learning_rate": 1.1914384591132045e-06, | |
| "loss": 0.5128, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.6057692307692308, | |
| "grad_norm": 2.147413310284848, | |
| "learning_rate": 1.1376106924788594e-06, | |
| "loss": 0.5316, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 2.3899145312729693, | |
| "learning_rate": 1.0848706879118893e-06, | |
| "loss": 0.5658, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.625, | |
| "grad_norm": 2.0348123827167335, | |
| "learning_rate": 1.0332332985438248e-06, | |
| "loss": 0.5269, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.6346153846153846, | |
| "grad_norm": 1.8531683575644768, | |
| "learning_rate": 9.82713066977427e-07, | |
| "loss": 0.5405, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.6442307692307692, | |
| "grad_norm": 1.9592530320960306, | |
| "learning_rate": 9.333242211910687e-07, | |
| "loss": 0.5184, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.6538461538461537, | |
| "grad_norm": 1.9954409383029053, | |
| "learning_rate": 8.850806705317183e-07, | |
| "loss": 0.4852, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.6634615384615383, | |
| "grad_norm": 2.098998703354889, | |
| "learning_rate": 8.379960017976546e-07, | |
| "loss": 0.5139, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.6730769230769231, | |
| "grad_norm": 2.5491851954980658, | |
| "learning_rate": 7.920834754120305e-07, | |
| "loss": 0.5392, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.6826923076923077, | |
| "grad_norm": 2.3935228227975, | |
| "learning_rate": 7.473560216883524e-07, | |
| "loss": 0.533, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 2.170304242236588, | |
| "learning_rate": 7.03826237188916e-07, | |
| "loss": 0.5523, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.7019230769230769, | |
| "grad_norm": 2.238634395178553, | |
| "learning_rate": 6.615063811772532e-07, | |
| "loss": 0.5122, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.7115384615384617, | |
| "grad_norm": 2.0570639625612985, | |
| "learning_rate": 6.204083721655607e-07, | |
| "loss": 0.4928, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.7211538461538463, | |
| "grad_norm": 2.103504186819526, | |
| "learning_rate": 5.805437845580958e-07, | |
| "loss": 0.5203, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.7307692307692308, | |
| "grad_norm": 2.197105961623965, | |
| "learning_rate": 5.41923845391486e-07, | |
| "loss": 0.5124, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.7403846153846154, | |
| "grad_norm": 2.1152456056469227, | |
| "learning_rate": 5.045594311728708e-07, | |
| "loss": 0.4916, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 2.0939964989579973, | |
| "learning_rate": 4.6846106481675035e-07, | |
| "loss": 0.5111, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.7596153846153846, | |
| "grad_norm": 2.4677432405873256, | |
| "learning_rate": 4.336389126814311e-07, | |
| "loss": 0.5346, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 1.9790895263145134, | |
| "learning_rate": 4.001027817058789e-07, | |
| "loss": 0.5051, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.7788461538461537, | |
| "grad_norm": 2.2327516887468417, | |
| "learning_rate": 3.6786211664779583e-07, | |
| "loss": 0.5376, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.7884615384615383, | |
| "grad_norm": 2.0971604625086777, | |
| "learning_rate": 3.369259974236988e-07, | |
| "loss": 0.4795, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.7980769230769231, | |
| "grad_norm": 2.124716857194864, | |
| "learning_rate": 3.0730313655175647e-07, | |
| "loss": 0.5388, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.8076923076923077, | |
| "grad_norm": 2.1095913682765977, | |
| "learning_rate": 2.790018766980773e-07, | |
| "loss": 0.545, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.8173076923076923, | |
| "grad_norm": 2.0130617931416275, | |
| "learning_rate": 2.520301883271797e-07, | |
| "loss": 0.5169, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.8269230769230769, | |
| "grad_norm": 2.083745653036941, | |
| "learning_rate": 2.2639566745727203e-07, | |
| "loss": 0.4928, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.8365384615384617, | |
| "grad_norm": 2.220101162444105, | |
| "learning_rate": 2.0210553352098815e-07, | |
| "loss": 0.5075, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 2.308065945824805, | |
| "learning_rate": 1.7916662733218848e-07, | |
| "loss": 0.5116, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.8557692307692308, | |
| "grad_norm": 1.9607562596689283, | |
| "learning_rate": 1.575854091593837e-07, | |
| "loss": 0.5183, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.8653846153846154, | |
| "grad_norm": 2.1443103265105634, | |
| "learning_rate": 1.3736795690633353e-07, | |
| "loss": 0.5123, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 1.9320175413394869, | |
| "learning_rate": 1.185199644003332e-07, | |
| "loss": 0.4931, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.8846153846153846, | |
| "grad_norm": 2.2160725485465567, | |
| "learning_rate": 1.0104673978866164e-07, | |
| "loss": 0.513, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.8942307692307692, | |
| "grad_norm": 2.2787872837737697, | |
| "learning_rate": 8.495320404365348e-08, | |
| "loss": 0.5062, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.9038461538461537, | |
| "grad_norm": 2.168799751436692, | |
| "learning_rate": 7.024388957680705e-08, | |
| "loss": 0.5168, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.9134615384615383, | |
| "grad_norm": 2.064037817068646, | |
| "learning_rate": 5.6922938962329364e-08, | |
| "loss": 0.5154, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 2.085831001764275, | |
| "learning_rate": 4.499410377045765e-08, | |
| "loss": 0.5059, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "eval_loss": 0.7636520862579346, | |
| "eval_runtime": 31.002, | |
| "eval_samples_per_second": 59.641, | |
| "eval_steps_per_second": 7.483, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.9326923076923077, | |
| "grad_norm": 2.0866363192357245, | |
| "learning_rate": 3.446074351091566e-08, | |
| "loss": 0.535, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.9423076923076923, | |
| "grad_norm": 1.9303541701779643, | |
| "learning_rate": 2.5325824686772138e-08, | |
| "loss": 0.5193, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.9519230769230769, | |
| "grad_norm": 2.035058945568711, | |
| "learning_rate": 1.7591919958986348e-08, | |
| "loss": 0.5242, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 1.9615384615384617, | |
| "grad_norm": 1.906676822404482, | |
| "learning_rate": 1.1261207421874309e-08, | |
| "loss": 0.4826, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.9711538461538463, | |
| "grad_norm": 2.0734277093423965, | |
| "learning_rate": 6.335469989692255e-09, | |
| "loss": 0.5064, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.9807692307692308, | |
| "grad_norm": 2.320470177512831, | |
| "learning_rate": 2.816094894513843e-09, | |
| "loss": 0.5264, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.9903846153846154, | |
| "grad_norm": 2.2494248679357955, | |
| "learning_rate": 7.040732955487795e-10, | |
| "loss": 0.51, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.2699709074042307, | |
| "learning_rate": 0.0, | |
| "loss": 0.5211, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 2080, | |
| "total_flos": 34318199685120.0, | |
| "train_loss": 0.6641153321816371, | |
| "train_runtime": 2465.8736, | |
| "train_samples_per_second": 13.491, | |
| "train_steps_per_second": 0.844 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2080, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 50000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 34318199685120.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |