| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.4, |
| "eval_steps": 100, |
| "global_step": 1000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 0.7581225659698247, |
| "epoch": 0.004, |
| "grad_norm": 101.5, |
| "learning_rate": 1.8e-07, |
| "loss": 18.1141, |
| "mean_token_accuracy": 0.7393306585028767, |
| "num_tokens": 72542.0, |
| "step": 10 |
| }, |
| { |
| "entropy": 0.7520309612154961, |
| "epoch": 0.008, |
| "grad_norm": 162.0, |
| "learning_rate": 3.8e-07, |
| "loss": 18.2441, |
| "mean_token_accuracy": 0.7392390985041857, |
| "num_tokens": 138146.0, |
| "step": 20 |
| }, |
| { |
| "entropy": 0.741816989146173, |
| "epoch": 0.012, |
| "grad_norm": 92.5, |
| "learning_rate": 5.800000000000001e-07, |
| "loss": 17.8539, |
| "mean_token_accuracy": 0.7405363619327545, |
| "num_tokens": 208409.0, |
| "step": 30 |
| }, |
| { |
| "entropy": 0.7636421315371991, |
| "epoch": 0.016, |
| "grad_norm": 88.0, |
| "learning_rate": 7.8e-07, |
| "loss": 18.142, |
| "mean_token_accuracy": 0.7400334309786558, |
| "num_tokens": 280941.0, |
| "step": 40 |
| }, |
| { |
| "entropy": 0.7905357209965587, |
| "epoch": 0.02, |
| "grad_norm": 108.0, |
| "learning_rate": 9.800000000000001e-07, |
| "loss": 18.1041, |
| "mean_token_accuracy": 0.7354144401848316, |
| "num_tokens": 349731.0, |
| "step": 50 |
| }, |
| { |
| "entropy": 0.8223027169704438, |
| "epoch": 0.024, |
| "grad_norm": 112.5, |
| "learning_rate": 1.1800000000000001e-06, |
| "loss": 18.4599, |
| "mean_token_accuracy": 0.731158659607172, |
| "num_tokens": 419127.0, |
| "step": 60 |
| }, |
| { |
| "entropy": 0.7908559931442142, |
| "epoch": 0.028, |
| "grad_norm": 106.0, |
| "learning_rate": 1.3800000000000001e-06, |
| "loss": 17.5053, |
| "mean_token_accuracy": 0.7440328601747751, |
| "num_tokens": 489016.0, |
| "step": 70 |
| }, |
| { |
| "entropy": 0.8229548260569572, |
| "epoch": 0.032, |
| "grad_norm": 86.0, |
| "learning_rate": 1.5800000000000001e-06, |
| "loss": 17.3714, |
| "mean_token_accuracy": 0.7438380591571331, |
| "num_tokens": 558056.0, |
| "step": 80 |
| }, |
| { |
| "entropy": 0.84512098710984, |
| "epoch": 0.036, |
| "grad_norm": 100.5, |
| "learning_rate": 1.7800000000000001e-06, |
| "loss": 17.2456, |
| "mean_token_accuracy": 0.7478047309443354, |
| "num_tokens": 627804.0, |
| "step": 90 |
| }, |
| { |
| "entropy": 0.8738276792690158, |
| "epoch": 0.04, |
| "grad_norm": 67.0, |
| "learning_rate": 1.98e-06, |
| "loss": 16.6896, |
| "mean_token_accuracy": 0.748936639726162, |
| "num_tokens": 697556.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.04, |
| "eval_biology_entropy": 1.1379984815716744, |
| "eval_biology_loss": 1.2692387104034424, |
| "eval_biology_mean_token_accuracy": 0.6880569371581078, |
| "eval_biology_num_tokens": 697556.0, |
| "eval_biology_runtime": 40.5573, |
| "eval_biology_samples_per_second": 12.328, |
| "eval_biology_steps_per_second": 12.328, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.04, |
| "eval_math_entropy": 0.9095822716951371, |
| "eval_math_loss": 1.0899367332458496, |
| "eval_math_mean_token_accuracy": 0.7395683622956276, |
| "eval_math_num_tokens": 697556.0, |
| "eval_math_runtime": 48.0541, |
| "eval_math_samples_per_second": 10.405, |
| "eval_math_steps_per_second": 10.405, |
| "step": 100 |
| }, |
| { |
| "entropy": 0.9193563124164938, |
| "epoch": 0.044, |
| "grad_norm": 88.5, |
| "learning_rate": 2.1800000000000003e-06, |
| "loss": 17.3763, |
| "mean_token_accuracy": 0.7420263335108757, |
| "num_tokens": 765262.0, |
| "step": 110 |
| }, |
| { |
| "entropy": 0.8906641118228436, |
| "epoch": 0.048, |
| "grad_norm": 78.0, |
| "learning_rate": 2.38e-06, |
| "loss": 16.2142, |
| "mean_token_accuracy": 0.7552934620529413, |
| "num_tokens": 836753.0, |
| "step": 120 |
| }, |
| { |
| "entropy": 0.9178624864667654, |
| "epoch": 0.052, |
| "grad_norm": 70.0, |
| "learning_rate": 2.5800000000000003e-06, |
| "loss": 16.3594, |
| "mean_token_accuracy": 0.7507667759433388, |
| "num_tokens": 909947.0, |
| "step": 130 |
| }, |
| { |
| "entropy": 0.9556385111063719, |
| "epoch": 0.056, |
| "grad_norm": 70.0, |
| "learning_rate": 2.7800000000000005e-06, |
| "loss": 16.6438, |
| "mean_token_accuracy": 0.748611630871892, |
| "num_tokens": 981959.0, |
| "step": 140 |
| }, |
| { |
| "entropy": 0.9541467605158687, |
| "epoch": 0.06, |
| "grad_norm": 86.0, |
| "learning_rate": 2.9800000000000003e-06, |
| "loss": 16.5304, |
| "mean_token_accuracy": 0.7486465960741043, |
| "num_tokens": 1053624.0, |
| "step": 150 |
| }, |
| { |
| "entropy": 0.9960285406559706, |
| "epoch": 0.064, |
| "grad_norm": 78.0, |
| "learning_rate": 3.1800000000000005e-06, |
| "loss": 16.9435, |
| "mean_token_accuracy": 0.7448480188846588, |
| "num_tokens": 1126446.0, |
| "step": 160 |
| }, |
| { |
| "entropy": 0.9677659103646874, |
| "epoch": 0.068, |
| "grad_norm": 106.5, |
| "learning_rate": 3.3800000000000007e-06, |
| "loss": 16.1013, |
| "mean_token_accuracy": 0.7561902595683933, |
| "num_tokens": 1196286.0, |
| "step": 170 |
| }, |
| { |
| "entropy": 0.9270429328083992, |
| "epoch": 0.072, |
| "grad_norm": 62.0, |
| "learning_rate": 3.58e-06, |
| "loss": 15.3741, |
| "mean_token_accuracy": 0.763100459985435, |
| "num_tokens": 1273905.0, |
| "step": 180 |
| }, |
| { |
| "entropy": 0.9678281666710973, |
| "epoch": 0.076, |
| "grad_norm": 51.0, |
| "learning_rate": 3.7800000000000002e-06, |
| "loss": 15.8618, |
| "mean_token_accuracy": 0.7534618094563484, |
| "num_tokens": 1346454.0, |
| "step": 190 |
| }, |
| { |
| "entropy": 0.9695806819945574, |
| "epoch": 0.08, |
| "grad_norm": 61.25, |
| "learning_rate": 3.980000000000001e-06, |
| "loss": 15.7874, |
| "mean_token_accuracy": 0.7560878146439791, |
| "num_tokens": 1413596.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.08, |
| "eval_biology_entropy": 1.214613477051258, |
| "eval_biology_loss": 1.2630934715270996, |
| "eval_biology_mean_token_accuracy": 0.6885634814500808, |
| "eval_biology_num_tokens": 1413596.0, |
| "eval_biology_runtime": 40.4294, |
| "eval_biology_samples_per_second": 12.367, |
| "eval_biology_steps_per_second": 12.367, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.08, |
| "eval_math_entropy": 0.9574190213680267, |
| "eval_math_loss": 0.9871246218681335, |
| "eval_math_mean_token_accuracy": 0.7557044946551323, |
| "eval_math_num_tokens": 1413596.0, |
| "eval_math_runtime": 48.0895, |
| "eval_math_samples_per_second": 10.397, |
| "eval_math_steps_per_second": 10.397, |
| "step": 200 |
| }, |
| { |
| "entropy": 0.8888338362798095, |
| "epoch": 0.084, |
| "grad_norm": 61.25, |
| "learning_rate": 4.18e-06, |
| "loss": 14.623, |
| "mean_token_accuracy": 0.7706882227212191, |
| "num_tokens": 1485147.0, |
| "step": 210 |
| }, |
| { |
| "entropy": 0.906145141646266, |
| "epoch": 0.088, |
| "grad_norm": 53.5, |
| "learning_rate": 4.38e-06, |
| "loss": 14.8128, |
| "mean_token_accuracy": 0.7661576565355063, |
| "num_tokens": 1559450.0, |
| "step": 220 |
| }, |
| { |
| "entropy": 0.95110107883811, |
| "epoch": 0.092, |
| "grad_norm": 63.25, |
| "learning_rate": 4.58e-06, |
| "loss": 15.2457, |
| "mean_token_accuracy": 0.761219710484147, |
| "num_tokens": 1631535.0, |
| "step": 230 |
| }, |
| { |
| "entropy": 0.9411580273881555, |
| "epoch": 0.096, |
| "grad_norm": 51.0, |
| "learning_rate": 4.78e-06, |
| "loss": 14.938, |
| "mean_token_accuracy": 0.7646324578672647, |
| "num_tokens": 1701193.0, |
| "step": 240 |
| }, |
| { |
| "entropy": 0.9254088591784239, |
| "epoch": 0.1, |
| "grad_norm": 56.25, |
| "learning_rate": 4.980000000000001e-06, |
| "loss": 14.6029, |
| "mean_token_accuracy": 0.7678503680974245, |
| "num_tokens": 1773596.0, |
| "step": 250 |
| }, |
| { |
| "entropy": 0.9214777417480946, |
| "epoch": 0.104, |
| "grad_norm": 54.0, |
| "learning_rate": 5.18e-06, |
| "loss": 15.043, |
| "mean_token_accuracy": 0.7618352189660073, |
| "num_tokens": 1840165.0, |
| "step": 260 |
| }, |
| { |
| "entropy": 0.9363527970388532, |
| "epoch": 0.108, |
| "grad_norm": 111.5, |
| "learning_rate": 5.380000000000001e-06, |
| "loss": 15.6052, |
| "mean_token_accuracy": 0.7564919672906398, |
| "num_tokens": 1908149.0, |
| "step": 270 |
| }, |
| { |
| "entropy": 0.9113753678277134, |
| "epoch": 0.112, |
| "grad_norm": 53.0, |
| "learning_rate": 5.580000000000001e-06, |
| "loss": 14.4645, |
| "mean_token_accuracy": 0.7736385434865951, |
| "num_tokens": 1979232.0, |
| "step": 280 |
| }, |
| { |
| "entropy": 0.916975767724216, |
| "epoch": 0.116, |
| "grad_norm": 54.5, |
| "learning_rate": 5.78e-06, |
| "loss": 15.0318, |
| "mean_token_accuracy": 0.7656373519450426, |
| "num_tokens": 2052039.0, |
| "step": 290 |
| }, |
| { |
| "entropy": 0.915022067911923, |
| "epoch": 0.12, |
| "grad_norm": 46.25, |
| "learning_rate": 5.98e-06, |
| "loss": 14.7236, |
| "mean_token_accuracy": 0.7691222932189703, |
| "num_tokens": 2125815.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.12, |
| "eval_biology_entropy": 1.2286294117569923, |
| "eval_biology_loss": 1.268429160118103, |
| "eval_biology_mean_token_accuracy": 0.6861296361088752, |
| "eval_biology_num_tokens": 2125815.0, |
| "eval_biology_runtime": 40.4544, |
| "eval_biology_samples_per_second": 12.36, |
| "eval_biology_steps_per_second": 12.36, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.12, |
| "eval_math_entropy": 0.9302992832064628, |
| "eval_math_loss": 0.9273415803909302, |
| "eval_math_mean_token_accuracy": 0.7664831226468086, |
| "eval_math_num_tokens": 2125815.0, |
| "eval_math_runtime": 48.047, |
| "eval_math_samples_per_second": 10.406, |
| "eval_math_steps_per_second": 10.406, |
| "step": 300 |
| }, |
| { |
| "entropy": 0.9336849415674806, |
| "epoch": 0.124, |
| "grad_norm": 55.25, |
| "learning_rate": 6.18e-06, |
| "loss": 14.9674, |
| "mean_token_accuracy": 0.7661335166543722, |
| "num_tokens": 2201991.0, |
| "step": 310 |
| }, |
| { |
| "entropy": 0.8756070390343667, |
| "epoch": 0.128, |
| "grad_norm": 52.25, |
| "learning_rate": 6.380000000000001e-06, |
| "loss": 14.0, |
| "mean_token_accuracy": 0.7786776568740607, |
| "num_tokens": 2274177.0, |
| "step": 320 |
| }, |
| { |
| "entropy": 0.892756593041122, |
| "epoch": 0.132, |
| "grad_norm": 45.75, |
| "learning_rate": 6.5800000000000005e-06, |
| "loss": 14.1896, |
| "mean_token_accuracy": 0.7731969062238931, |
| "num_tokens": 2342173.0, |
| "step": 330 |
| }, |
| { |
| "entropy": 0.8899674592539668, |
| "epoch": 0.136, |
| "grad_norm": 56.5, |
| "learning_rate": 6.780000000000001e-06, |
| "loss": 14.2167, |
| "mean_token_accuracy": 0.7683363553136587, |
| "num_tokens": 2412435.0, |
| "step": 340 |
| }, |
| { |
| "entropy": 0.9237651886418462, |
| "epoch": 0.14, |
| "grad_norm": 49.5, |
| "learning_rate": 6.98e-06, |
| "loss": 14.7203, |
| "mean_token_accuracy": 0.7684664275497198, |
| "num_tokens": 2481105.0, |
| "step": 350 |
| }, |
| { |
| "entropy": 0.8494639024138451, |
| "epoch": 0.144, |
| "grad_norm": 49.0, |
| "learning_rate": 7.180000000000001e-06, |
| "loss": 13.6956, |
| "mean_token_accuracy": 0.7817885402590037, |
| "num_tokens": 2548445.0, |
| "step": 360 |
| }, |
| { |
| "entropy": 0.8341110894456506, |
| "epoch": 0.148, |
| "grad_norm": 44.0, |
| "learning_rate": 7.3800000000000005e-06, |
| "loss": 13.2223, |
| "mean_token_accuracy": 0.789648124948144, |
| "num_tokens": 2621364.0, |
| "step": 370 |
| }, |
| { |
| "entropy": 0.8608193326741457, |
| "epoch": 0.152, |
| "grad_norm": 52.25, |
| "learning_rate": 7.58e-06, |
| "loss": 13.8861, |
| "mean_token_accuracy": 0.7754677496850491, |
| "num_tokens": 2686462.0, |
| "step": 380 |
| }, |
| { |
| "entropy": 0.8447565982118249, |
| "epoch": 0.156, |
| "grad_norm": 50.0, |
| "learning_rate": 7.78e-06, |
| "loss": 13.5199, |
| "mean_token_accuracy": 0.7812942665070295, |
| "num_tokens": 2754241.0, |
| "step": 390 |
| }, |
| { |
| "entropy": 0.8711243279278278, |
| "epoch": 0.16, |
| "grad_norm": 79.0, |
| "learning_rate": 7.980000000000002e-06, |
| "loss": 14.1276, |
| "mean_token_accuracy": 0.7742310840636492, |
| "num_tokens": 2824418.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_biology_entropy": 1.22758097666502, |
| "eval_biology_loss": 1.2762199640274048, |
| "eval_biology_mean_token_accuracy": 0.6836535354852676, |
| "eval_biology_num_tokens": 2824418.0, |
| "eval_biology_runtime": 40.7281, |
| "eval_biology_samples_per_second": 12.277, |
| "eval_biology_steps_per_second": 12.277, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_math_entropy": 0.9047691844701767, |
| "eval_math_loss": 0.8841754198074341, |
| "eval_math_mean_token_accuracy": 0.775034920334816, |
| "eval_math_num_tokens": 2824418.0, |
| "eval_math_runtime": 47.9197, |
| "eval_math_samples_per_second": 10.434, |
| "eval_math_steps_per_second": 10.434, |
| "step": 400 |
| }, |
| { |
| "entropy": 0.8859184484928846, |
| "epoch": 0.164, |
| "grad_norm": 127.0, |
| "learning_rate": 8.18e-06, |
| "loss": 14.142, |
| "mean_token_accuracy": 0.7765266574919224, |
| "num_tokens": 2895667.0, |
| "step": 410 |
| }, |
| { |
| "entropy": 0.8501209424808621, |
| "epoch": 0.168, |
| "grad_norm": 40.0, |
| "learning_rate": 8.380000000000001e-06, |
| "loss": 13.8435, |
| "mean_token_accuracy": 0.7774271417409182, |
| "num_tokens": 2966031.0, |
| "step": 420 |
| }, |
| { |
| "entropy": 0.8510567891411484, |
| "epoch": 0.172, |
| "grad_norm": 45.75, |
| "learning_rate": 8.580000000000001e-06, |
| "loss": 13.3673, |
| "mean_token_accuracy": 0.7858892437070608, |
| "num_tokens": 3039780.0, |
| "step": 430 |
| }, |
| { |
| "entropy": 0.8355857279151678, |
| "epoch": 0.176, |
| "grad_norm": 44.5, |
| "learning_rate": 8.78e-06, |
| "loss": 13.7383, |
| "mean_token_accuracy": 0.7784004233777523, |
| "num_tokens": 3110313.0, |
| "step": 440 |
| }, |
| { |
| "entropy": 0.8662491450086236, |
| "epoch": 0.18, |
| "grad_norm": 43.75, |
| "learning_rate": 8.98e-06, |
| "loss": 13.6485, |
| "mean_token_accuracy": 0.7806661173701286, |
| "num_tokens": 3179963.0, |
| "step": 450 |
| }, |
| { |
| "entropy": 0.8596813324838877, |
| "epoch": 0.184, |
| "grad_norm": 46.0, |
| "learning_rate": 9.180000000000002e-06, |
| "loss": 13.7717, |
| "mean_token_accuracy": 0.7796813258901238, |
| "num_tokens": 3251139.0, |
| "step": 460 |
| }, |
| { |
| "entropy": 0.8454847941175103, |
| "epoch": 0.188, |
| "grad_norm": 56.25, |
| "learning_rate": 9.38e-06, |
| "loss": 13.4682, |
| "mean_token_accuracy": 0.7803339377045632, |
| "num_tokens": 3323519.0, |
| "step": 470 |
| }, |
| { |
| "entropy": 0.8228314617648721, |
| "epoch": 0.192, |
| "grad_norm": 49.75, |
| "learning_rate": 9.58e-06, |
| "loss": 13.3842, |
| "mean_token_accuracy": 0.7827629748731851, |
| "num_tokens": 3394170.0, |
| "step": 480 |
| }, |
| { |
| "entropy": 0.8164506537839771, |
| "epoch": 0.196, |
| "grad_norm": 54.25, |
| "learning_rate": 9.780000000000001e-06, |
| "loss": 13.1965, |
| "mean_token_accuracy": 0.7878578577190638, |
| "num_tokens": 3465341.0, |
| "step": 490 |
| }, |
| { |
| "entropy": 0.8838530456647277, |
| "epoch": 0.2, |
| "grad_norm": 51.0, |
| "learning_rate": 9.980000000000001e-06, |
| "loss": 13.9771, |
| "mean_token_accuracy": 0.7734297584742308, |
| "num_tokens": 3533571.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.2, |
| "eval_biology_entropy": 1.2344867270588875, |
| "eval_biology_loss": 1.2817195653915405, |
| "eval_biology_mean_token_accuracy": 0.6832386125922203, |
| "eval_biology_num_tokens": 3533571.0, |
| "eval_biology_runtime": 40.4092, |
| "eval_biology_samples_per_second": 12.373, |
| "eval_biology_steps_per_second": 12.373, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.2, |
| "eval_math_entropy": 0.8546719763875008, |
| "eval_math_loss": 0.8547734618186951, |
| "eval_math_mean_token_accuracy": 0.7804626071453095, |
| "eval_math_num_tokens": 3533571.0, |
| "eval_math_runtime": 48.0961, |
| "eval_math_samples_per_second": 10.396, |
| "eval_math_steps_per_second": 10.396, |
| "step": 500 |
| }, |
| { |
| "entropy": 0.8578155474737287, |
| "epoch": 0.204, |
| "grad_norm": 45.75, |
| "learning_rate": 1.018e-05, |
| "loss": 13.8097, |
| "mean_token_accuracy": 0.776990968361497, |
| "num_tokens": 3602983.0, |
| "step": 510 |
| }, |
| { |
| "entropy": 0.8306718476116657, |
| "epoch": 0.208, |
| "grad_norm": 50.25, |
| "learning_rate": 1.038e-05, |
| "loss": 13.4186, |
| "mean_token_accuracy": 0.7863588180392981, |
| "num_tokens": 3673600.0, |
| "step": 520 |
| }, |
| { |
| "entropy": 0.8772016573697329, |
| "epoch": 0.212, |
| "grad_norm": 45.75, |
| "learning_rate": 1.0580000000000002e-05, |
| "loss": 13.9353, |
| "mean_token_accuracy": 0.7783012766391039, |
| "num_tokens": 3743641.0, |
| "step": 530 |
| }, |
| { |
| "entropy": 0.8009312467649579, |
| "epoch": 0.216, |
| "grad_norm": 55.25, |
| "learning_rate": 1.0780000000000002e-05, |
| "loss": 12.9356, |
| "mean_token_accuracy": 0.7893661856651306, |
| "num_tokens": 3817397.0, |
| "step": 540 |
| }, |
| { |
| "entropy": 0.8427618410438299, |
| "epoch": 0.22, |
| "grad_norm": 51.0, |
| "learning_rate": 1.0980000000000002e-05, |
| "loss": 13.3831, |
| "mean_token_accuracy": 0.7832117900252342, |
| "num_tokens": 3886866.0, |
| "step": 550 |
| }, |
| { |
| "entropy": 0.8758323887363076, |
| "epoch": 0.224, |
| "grad_norm": 41.5, |
| "learning_rate": 1.1180000000000001e-05, |
| "loss": 13.8448, |
| "mean_token_accuracy": 0.7745687130838632, |
| "num_tokens": 3953732.0, |
| "step": 560 |
| }, |
| { |
| "entropy": 0.8352954359725118, |
| "epoch": 0.228, |
| "grad_norm": 41.75, |
| "learning_rate": 1.138e-05, |
| "loss": 13.4861, |
| "mean_token_accuracy": 0.7828061260282994, |
| "num_tokens": 4025795.0, |
| "step": 570 |
| }, |
| { |
| "entropy": 0.7947553901001811, |
| "epoch": 0.232, |
| "grad_norm": 43.25, |
| "learning_rate": 1.1580000000000001e-05, |
| "loss": 12.8366, |
| "mean_token_accuracy": 0.7880385994911194, |
| "num_tokens": 4098989.0, |
| "step": 580 |
| }, |
| { |
| "entropy": 0.8285046070814133, |
| "epoch": 0.236, |
| "grad_norm": 39.25, |
| "learning_rate": 1.178e-05, |
| "loss": 13.0789, |
| "mean_token_accuracy": 0.7872951343655586, |
| "num_tokens": 4170878.0, |
| "step": 590 |
| }, |
| { |
| "entropy": 0.7863880429416895, |
| "epoch": 0.24, |
| "grad_norm": 45.0, |
| "learning_rate": 1.198e-05, |
| "loss": 12.7479, |
| "mean_token_accuracy": 0.7928395148366689, |
| "num_tokens": 4243655.0, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.24, |
| "eval_biology_entropy": 1.2514844969511032, |
| "eval_biology_loss": 1.2855682373046875, |
| "eval_biology_mean_token_accuracy": 0.6817903207540512, |
| "eval_biology_num_tokens": 4243655.0, |
| "eval_biology_runtime": 40.7639, |
| "eval_biology_samples_per_second": 12.266, |
| "eval_biology_steps_per_second": 12.266, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.24, |
| "eval_math_entropy": 0.843876126229763, |
| "eval_math_loss": 0.8332411646842957, |
| "eval_math_mean_token_accuracy": 0.783966186761856, |
| "eval_math_num_tokens": 4243655.0, |
| "eval_math_runtime": 48.1965, |
| "eval_math_samples_per_second": 10.374, |
| "eval_math_steps_per_second": 10.374, |
| "step": 600 |
| }, |
| { |
| "entropy": 0.7838549023494125, |
| "epoch": 0.244, |
| "grad_norm": 41.75, |
| "learning_rate": 1.218e-05, |
| "loss": 12.4042, |
| "mean_token_accuracy": 0.7971927735954523, |
| "num_tokens": 4317321.0, |
| "step": 610 |
| }, |
| { |
| "entropy": 0.7797296568751335, |
| "epoch": 0.248, |
| "grad_norm": 51.75, |
| "learning_rate": 1.2380000000000002e-05, |
| "loss": 12.5895, |
| "mean_token_accuracy": 0.7916548330336809, |
| "num_tokens": 4387651.0, |
| "step": 620 |
| }, |
| { |
| "entropy": 0.8094759196043014, |
| "epoch": 0.252, |
| "grad_norm": 84.5, |
| "learning_rate": 1.2580000000000002e-05, |
| "loss": 12.9446, |
| "mean_token_accuracy": 0.7892590902745724, |
| "num_tokens": 4461909.0, |
| "step": 630 |
| }, |
| { |
| "entropy": 0.7797345662489533, |
| "epoch": 0.256, |
| "grad_norm": 62.0, |
| "learning_rate": 1.2780000000000001e-05, |
| "loss": 12.462, |
| "mean_token_accuracy": 0.7935512393712998, |
| "num_tokens": 4531471.0, |
| "step": 640 |
| }, |
| { |
| "entropy": 0.7911159221082926, |
| "epoch": 0.26, |
| "grad_norm": 40.0, |
| "learning_rate": 1.2980000000000001e-05, |
| "loss": 12.8515, |
| "mean_token_accuracy": 0.7928617469966411, |
| "num_tokens": 4600757.0, |
| "step": 650 |
| }, |
| { |
| "entropy": 0.8169831104576588, |
| "epoch": 0.264, |
| "grad_norm": 49.5, |
| "learning_rate": 1.3180000000000001e-05, |
| "loss": 13.117, |
| "mean_token_accuracy": 0.7893466159701348, |
| "num_tokens": 4669389.0, |
| "step": 660 |
| }, |
| { |
| "entropy": 0.8287884982302784, |
| "epoch": 0.268, |
| "grad_norm": 55.25, |
| "learning_rate": 1.3380000000000002e-05, |
| "loss": 13.3618, |
| "mean_token_accuracy": 0.7853043630719185, |
| "num_tokens": 4740492.0, |
| "step": 670 |
| }, |
| { |
| "entropy": 0.7938818972557783, |
| "epoch": 0.272, |
| "grad_norm": 42.5, |
| "learning_rate": 1.3580000000000002e-05, |
| "loss": 12.6092, |
| "mean_token_accuracy": 0.7941738195717335, |
| "num_tokens": 4810284.0, |
| "step": 680 |
| }, |
| { |
| "entropy": 0.7964617364108563, |
| "epoch": 0.276, |
| "grad_norm": 37.25, |
| "learning_rate": 1.378e-05, |
| "loss": 12.6495, |
| "mean_token_accuracy": 0.7921258114278317, |
| "num_tokens": 4885701.0, |
| "step": 690 |
| }, |
| { |
| "entropy": 0.7932526333257556, |
| "epoch": 0.28, |
| "grad_norm": 48.5, |
| "learning_rate": 1.398e-05, |
| "loss": 12.8488, |
| "mean_token_accuracy": 0.7890512771904469, |
| "num_tokens": 4955375.0, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.28, |
| "eval_biology_entropy": 1.2677810794115068, |
| "eval_biology_loss": 1.2936662435531616, |
| "eval_biology_mean_token_accuracy": 0.6814599596261978, |
| "eval_biology_num_tokens": 4955375.0, |
| "eval_biology_runtime": 40.9028, |
| "eval_biology_samples_per_second": 12.224, |
| "eval_biology_steps_per_second": 12.224, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.28, |
| "eval_math_entropy": 0.8274374225139618, |
| "eval_math_loss": 0.8181362748146057, |
| "eval_math_mean_token_accuracy": 0.7868153600692749, |
| "eval_math_num_tokens": 4955375.0, |
| "eval_math_runtime": 48.2074, |
| "eval_math_samples_per_second": 10.372, |
| "eval_math_steps_per_second": 10.372, |
| "step": 700 |
| }, |
| { |
| "entropy": 0.7877025598660111, |
| "epoch": 0.284, |
| "grad_norm": 39.25, |
| "learning_rate": 1.418e-05, |
| "loss": 12.6293, |
| "mean_token_accuracy": 0.7944098394364119, |
| "num_tokens": 5025638.0, |
| "step": 710 |
| }, |
| { |
| "entropy": 0.7564179511740804, |
| "epoch": 0.288, |
| "grad_norm": 38.75, |
| "learning_rate": 1.4380000000000001e-05, |
| "loss": 12.0294, |
| "mean_token_accuracy": 0.8003660906106234, |
| "num_tokens": 5095104.0, |
| "step": 720 |
| }, |
| { |
| "entropy": 0.7647898152470589, |
| "epoch": 0.292, |
| "grad_norm": 56.0, |
| "learning_rate": 1.4580000000000001e-05, |
| "loss": 12.4713, |
| "mean_token_accuracy": 0.7947716873139143, |
| "num_tokens": 5167049.0, |
| "step": 730 |
| }, |
| { |
| "entropy": 0.8119000235572458, |
| "epoch": 0.296, |
| "grad_norm": 36.75, |
| "learning_rate": 1.478e-05, |
| "loss": 13.0034, |
| "mean_token_accuracy": 0.7876484204083681, |
| "num_tokens": 5240988.0, |
| "step": 740 |
| }, |
| { |
| "entropy": 0.7802535065449774, |
| "epoch": 0.3, |
| "grad_norm": 40.5, |
| "learning_rate": 1.498e-05, |
| "loss": 12.4371, |
| "mean_token_accuracy": 0.795463678240776, |
| "num_tokens": 5312258.0, |
| "step": 750 |
| }, |
| { |
| "entropy": 0.8087782407179475, |
| "epoch": 0.304, |
| "grad_norm": 44.75, |
| "learning_rate": 1.5180000000000002e-05, |
| "loss": 12.9661, |
| "mean_token_accuracy": 0.7857579983770847, |
| "num_tokens": 5383732.0, |
| "step": 760 |
| }, |
| { |
| "entropy": 0.8243673056364059, |
| "epoch": 0.308, |
| "grad_norm": 39.5, |
| "learning_rate": 1.5380000000000002e-05, |
| "loss": 13.1334, |
| "mean_token_accuracy": 0.7867084331810474, |
| "num_tokens": 5456650.0, |
| "step": 770 |
| }, |
| { |
| "entropy": 0.7901528097689152, |
| "epoch": 0.312, |
| "grad_norm": 40.0, |
| "learning_rate": 1.5580000000000003e-05, |
| "loss": 12.6608, |
| "mean_token_accuracy": 0.7905277730897069, |
| "num_tokens": 5530666.0, |
| "step": 780 |
| }, |
| { |
| "entropy": 0.7515985878184438, |
| "epoch": 0.316, |
| "grad_norm": 44.75, |
| "learning_rate": 1.578e-05, |
| "loss": 11.9655, |
| "mean_token_accuracy": 0.8011779848486185, |
| "num_tokens": 5605269.0, |
| "step": 790 |
| }, |
| { |
| "entropy": 0.7314818985760212, |
| "epoch": 0.32, |
| "grad_norm": 42.5, |
| "learning_rate": 1.5980000000000003e-05, |
| "loss": 11.7244, |
| "mean_token_accuracy": 0.8059452049434185, |
| "num_tokens": 5676334.0, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_biology_entropy": 1.2663928886651994, |
| "eval_biology_loss": 1.2994052171707153, |
| "eval_biology_mean_token_accuracy": 0.6803199350833893, |
| "eval_biology_num_tokens": 5676334.0, |
| "eval_biology_runtime": 40.7632, |
| "eval_biology_samples_per_second": 12.266, |
| "eval_biology_steps_per_second": 12.266, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_math_entropy": 0.7805578476190567, |
| "eval_math_loss": 0.8004927635192871, |
| "eval_math_mean_token_accuracy": 0.7903167794942856, |
| "eval_math_num_tokens": 5676334.0, |
| "eval_math_runtime": 48.1783, |
| "eval_math_samples_per_second": 10.378, |
| "eval_math_steps_per_second": 10.378, |
| "step": 800 |
| }, |
| { |
| "entropy": 0.7661854576319456, |
| "epoch": 0.324, |
| "grad_norm": 40.5, |
| "learning_rate": 1.618e-05, |
| "loss": 12.5767, |
| "mean_token_accuracy": 0.7944763712584972, |
| "num_tokens": 5747779.0, |
| "step": 810 |
| }, |
| { |
| "entropy": 0.803387301787734, |
| "epoch": 0.328, |
| "grad_norm": 42.0, |
| "learning_rate": 1.638e-05, |
| "loss": 12.825, |
| "mean_token_accuracy": 0.7907922245562077, |
| "num_tokens": 5816488.0, |
| "step": 820 |
| }, |
| { |
| "entropy": 0.8049221463501454, |
| "epoch": 0.332, |
| "grad_norm": 44.0, |
| "learning_rate": 1.658e-05, |
| "loss": 12.9669, |
| "mean_token_accuracy": 0.7854703675955534, |
| "num_tokens": 5887180.0, |
| "step": 830 |
| }, |
| { |
| "entropy": 0.8102014064788818, |
| "epoch": 0.336, |
| "grad_norm": 38.5, |
| "learning_rate": 1.6780000000000002e-05, |
| "loss": 12.9814, |
| "mean_token_accuracy": 0.7886964596807957, |
| "num_tokens": 5958480.0, |
| "step": 840 |
| }, |
| { |
| "entropy": 0.7344874950125814, |
| "epoch": 0.34, |
| "grad_norm": 50.25, |
| "learning_rate": 1.698e-05, |
| "loss": 11.7609, |
| "mean_token_accuracy": 0.8062219381332397, |
| "num_tokens": 6031020.0, |
| "step": 850 |
| }, |
| { |
| "entropy": 0.802255941182375, |
| "epoch": 0.344, |
| "grad_norm": 39.0, |
| "learning_rate": 1.718e-05, |
| "loss": 12.8439, |
| "mean_token_accuracy": 0.7877463910728693, |
| "num_tokens": 6102081.0, |
| "step": 860 |
| }, |
| { |
| "entropy": 0.7980229431763292, |
| "epoch": 0.348, |
| "grad_norm": 36.25, |
| "learning_rate": 1.7380000000000003e-05, |
| "loss": 12.7774, |
| "mean_token_accuracy": 0.7900870595127344, |
| "num_tokens": 6172871.0, |
| "step": 870 |
| }, |
| { |
| "entropy": 0.7401213280856609, |
| "epoch": 0.352, |
| "grad_norm": 40.75, |
| "learning_rate": 1.758e-05, |
| "loss": 11.9409, |
| "mean_token_accuracy": 0.801040057092905, |
| "num_tokens": 6242161.0, |
| "step": 880 |
| }, |
| { |
| "entropy": 0.8095649138092995, |
| "epoch": 0.356, |
| "grad_norm": 46.5, |
| "learning_rate": 1.7780000000000003e-05, |
| "loss": 12.9134, |
| "mean_token_accuracy": 0.7872339155524969, |
| "num_tokens": 6312187.0, |
| "step": 890 |
| }, |
| { |
| "entropy": 0.7853369496762752, |
| "epoch": 0.36, |
| "grad_norm": 37.75, |
| "learning_rate": 1.798e-05, |
| "loss": 12.4062, |
| "mean_token_accuracy": 0.7960268270224333, |
| "num_tokens": 6377327.0, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.36, |
| "eval_biology_entropy": 1.250610340833664, |
| "eval_biology_loss": 1.3047173023223877, |
| "eval_biology_mean_token_accuracy": 0.6794998321533203, |
| "eval_biology_num_tokens": 6377327.0, |
| "eval_biology_runtime": 40.5105, |
| "eval_biology_samples_per_second": 12.342, |
| "eval_biology_steps_per_second": 12.342, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.36, |
| "eval_math_entropy": 0.773322255730629, |
| "eval_math_loss": 0.789590060710907, |
| "eval_math_mean_token_accuracy": 0.7914780566692352, |
| "eval_math_num_tokens": 6377327.0, |
| "eval_math_runtime": 48.04, |
| "eval_math_samples_per_second": 10.408, |
| "eval_math_steps_per_second": 10.408, |
| "step": 900 |
| }, |
| { |
| "entropy": 0.763472201116383, |
| "epoch": 0.364, |
| "grad_norm": 41.5, |
| "learning_rate": 1.8180000000000002e-05, |
| "loss": 12.3288, |
| "mean_token_accuracy": 0.7959023587405681, |
| "num_tokens": 6453254.0, |
| "step": 910 |
| }, |
| { |
| "entropy": 0.7712945869192481, |
| "epoch": 0.368, |
| "grad_norm": 40.25, |
| "learning_rate": 1.8380000000000004e-05, |
| "loss": 12.2854, |
| "mean_token_accuracy": 0.795868669450283, |
| "num_tokens": 6523679.0, |
| "step": 920 |
| }, |
| { |
| "entropy": 0.7641221083700657, |
| "epoch": 0.372, |
| "grad_norm": 37.25, |
| "learning_rate": 1.858e-05, |
| "loss": 12.2867, |
| "mean_token_accuracy": 0.7949323236942292, |
| "num_tokens": 6594651.0, |
| "step": 930 |
| }, |
| { |
| "entropy": 0.7518373979255557, |
| "epoch": 0.376, |
| "grad_norm": 42.0, |
| "learning_rate": 1.878e-05, |
| "loss": 12.1179, |
| "mean_token_accuracy": 0.798585458099842, |
| "num_tokens": 6669278.0, |
| "step": 940 |
| }, |
| { |
| "entropy": 0.7543134400621057, |
| "epoch": 0.38, |
| "grad_norm": 33.75, |
| "learning_rate": 1.898e-05, |
| "loss": 12.023, |
| "mean_token_accuracy": 0.7991646058857441, |
| "num_tokens": 6740491.0, |
| "step": 950 |
| }, |
| { |
| "entropy": 0.7203706834465265, |
| "epoch": 0.384, |
| "grad_norm": 35.5, |
| "learning_rate": 1.918e-05, |
| "loss": 11.6634, |
| "mean_token_accuracy": 0.8036353968083858, |
| "num_tokens": 6810978.0, |
| "step": 960 |
| }, |
| { |
| "entropy": 0.7761434537358582, |
| "epoch": 0.388, |
| "grad_norm": 39.5, |
| "learning_rate": 1.938e-05, |
| "loss": 12.4387, |
| "mean_token_accuracy": 0.7944561909884215, |
| "num_tokens": 6882077.0, |
| "step": 970 |
| }, |
| { |
| "entropy": 0.7785217678174376, |
| "epoch": 0.392, |
| "grad_norm": 33.75, |
| "learning_rate": 1.9580000000000002e-05, |
| "loss": 12.607, |
| "mean_token_accuracy": 0.7944069147109986, |
| "num_tokens": 6953929.0, |
| "step": 980 |
| }, |
| { |
| "entropy": 0.8195465696975589, |
| "epoch": 0.396, |
| "grad_norm": 65.5, |
| "learning_rate": 1.978e-05, |
| "loss": 13.2202, |
| "mean_token_accuracy": 0.7838506288826466, |
| "num_tokens": 7023835.0, |
| "step": 990 |
| }, |
| { |
| "entropy": 0.8217529645189643, |
| "epoch": 0.4, |
| "grad_norm": 39.75, |
| "learning_rate": 1.9980000000000002e-05, |
| "loss": 13.1698, |
| "mean_token_accuracy": 0.7848756689578295, |
| "num_tokens": 7096903.0, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.4, |
| "eval_biology_entropy": 1.266352553486824, |
| "eval_biology_loss": 1.312301516532898, |
| "eval_biology_mean_token_accuracy": 0.6775026069879532, |
| "eval_biology_num_tokens": 7096903.0, |
| "eval_biology_runtime": 40.7343, |
| "eval_biology_samples_per_second": 12.275, |
| "eval_biology_steps_per_second": 12.275, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.4, |
| "eval_math_entropy": 0.7719556384682655, |
| "eval_math_loss": 0.7821868062019348, |
| "eval_math_mean_token_accuracy": 0.7932658619880676, |
| "eval_math_num_tokens": 7096903.0, |
| "eval_math_runtime": 47.8946, |
| "eval_math_samples_per_second": 10.44, |
| "eval_math_steps_per_second": 10.44, |
| "step": 1000 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 10000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 4, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 4.7587833384015814e+17, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
