| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.8, |
| "eval_steps": 100, |
| "global_step": 500, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 1.0476444251835346, |
| "epoch": 0.016, |
| "grad_norm": 158.0, |
| "learning_rate": 6.000000000000001e-07, |
| "loss": 34.7404, |
| "mean_token_accuracy": 0.6322720043361187, |
| "num_tokens": 280941.0, |
| "step": 10 |
| }, |
| { |
| "entropy": 1.103129617869854, |
| "epoch": 0.032, |
| "grad_norm": 140.0, |
| "learning_rate": 1.2666666666666669e-06, |
| "loss": 35.2833, |
| "mean_token_accuracy": 0.627065640874207, |
| "num_tokens": 558056.0, |
| "step": 20 |
| }, |
| { |
| "entropy": 1.1186710916459561, |
| "epoch": 0.048, |
| "grad_norm": 125.5, |
| "learning_rate": 1.9333333333333336e-06, |
| "loss": 34.279, |
| "mean_token_accuracy": 0.6317529268562794, |
| "num_tokens": 836753.0, |
| "step": 30 |
| }, |
| { |
| "entropy": 1.2232312321662904, |
| "epoch": 0.064, |
| "grad_norm": 99.0, |
| "learning_rate": 2.6e-06, |
| "loss": 33.2898, |
| "mean_token_accuracy": 0.6292566264048218, |
| "num_tokens": 1126446.0, |
| "step": 40 |
| }, |
| { |
| "entropy": 1.3642345190048217, |
| "epoch": 0.08, |
| "grad_norm": 78.0, |
| "learning_rate": 3.266666666666667e-06, |
| "loss": 31.1081, |
| "mean_token_accuracy": 0.643590996041894, |
| "num_tokens": 1413596.0, |
| "step": 50 |
| }, |
| { |
| "entropy": 1.4666775342077016, |
| "epoch": 0.096, |
| "grad_norm": 50.0, |
| "learning_rate": 3.9333333333333335e-06, |
| "loss": 28.0855, |
| "mean_token_accuracy": 0.6677330181002616, |
| "num_tokens": 1701193.0, |
| "step": 60 |
| }, |
| { |
| "entropy": 1.479279049113393, |
| "epoch": 0.112, |
| "grad_norm": 44.5, |
| "learning_rate": 4.600000000000001e-06, |
| "loss": 25.9511, |
| "mean_token_accuracy": 0.6763931937515736, |
| "num_tokens": 1979232.0, |
| "step": 70 |
| }, |
| { |
| "entropy": 1.4356445044279098, |
| "epoch": 0.128, |
| "grad_norm": 31.5, |
| "learning_rate": 5.2666666666666665e-06, |
| "loss": 24.2383, |
| "mean_token_accuracy": 0.6900208964943886, |
| "num_tokens": 2274177.0, |
| "step": 80 |
| }, |
| { |
| "entropy": 1.3773791242390871, |
| "epoch": 0.144, |
| "grad_norm": 28.25, |
| "learning_rate": 5.933333333333335e-06, |
| "loss": 22.3128, |
| "mean_token_accuracy": 0.7042559664696455, |
| "num_tokens": 2548445.0, |
| "step": 90 |
| }, |
| { |
| "entropy": 1.3055105797946454, |
| "epoch": 0.16, |
| "grad_norm": 29.25, |
| "learning_rate": 6.600000000000001e-06, |
| "loss": 20.6858, |
| "mean_token_accuracy": 0.718151243403554, |
| "num_tokens": 2824418.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_biology_entropy": 3.5783245754241944, |
| "eval_biology_loss": 4.867801189422607, |
| "eval_biology_mean_token_accuracy": 0.3678655695915222, |
| "eval_biology_num_tokens": 2824418.0, |
| "eval_biology_runtime": 50.0015, |
| "eval_biology_samples_per_second": 10.0, |
| "eval_biology_steps_per_second": 2.5, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_chemistry_entropy": 2.541287685394287, |
| "eval_chemistry_loss": 2.7937169075012207, |
| "eval_chemistry_mean_token_accuracy": 0.539857518196106, |
| "eval_chemistry_num_tokens": 2824418.0, |
| "eval_chemistry_runtime": 61.8974, |
| "eval_chemistry_samples_per_second": 8.078, |
| "eval_chemistry_steps_per_second": 2.019, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_math_entropy": 1.325653440952301, |
| "eval_math_loss": 1.2963088750839233, |
| "eval_math_mean_token_accuracy": 0.7152829113006591, |
| "eval_math_num_tokens": 2824418.0, |
| "eval_math_runtime": 63.4831, |
| "eval_math_samples_per_second": 7.876, |
| "eval_math_steps_per_second": 1.969, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_physics_entropy": 2.0472509851455687, |
| "eval_physics_loss": 2.108372688293457, |
| "eval_physics_mean_token_accuracy": 0.6150235052108765, |
| "eval_physics_num_tokens": 2824418.0, |
| "eval_physics_runtime": 72.2792, |
| "eval_physics_samples_per_second": 6.918, |
| "eval_physics_steps_per_second": 1.729, |
| "step": 100 |
| }, |
| { |
| "entropy": 1.2799372777342797, |
| "epoch": 0.176, |
| "grad_norm": 23.25, |
| "learning_rate": 7.266666666666668e-06, |
| "loss": 20.1214, |
| "mean_token_accuracy": 0.7223256763070822, |
| "num_tokens": 3110313.0, |
| "step": 110 |
| }, |
| { |
| "entropy": 1.271880804002285, |
| "epoch": 0.192, |
| "grad_norm": 22.625, |
| "learning_rate": 7.933333333333334e-06, |
| "loss": 19.8735, |
| "mean_token_accuracy": 0.7208763293921947, |
| "num_tokens": 3394170.0, |
| "step": 120 |
| }, |
| { |
| "entropy": 1.2141485448926688, |
| "epoch": 0.208, |
| "grad_norm": 20.25, |
| "learning_rate": 8.6e-06, |
| "loss": 19.3285, |
| "mean_token_accuracy": 0.7284529402852058, |
| "num_tokens": 3673600.0, |
| "step": 130 |
| }, |
| { |
| "entropy": 1.1842644162476064, |
| "epoch": 0.224, |
| "grad_norm": 22.25, |
| "learning_rate": 9.266666666666667e-06, |
| "loss": 18.8222, |
| "mean_token_accuracy": 0.7318667653948069, |
| "num_tokens": 3953732.0, |
| "step": 140 |
| }, |
| { |
| "entropy": 1.1367911960929633, |
| "epoch": 0.24, |
| "grad_norm": 21.875, |
| "learning_rate": 9.933333333333334e-06, |
| "loss": 18.0331, |
| "mean_token_accuracy": 0.7411410234868526, |
| "num_tokens": 4243655.0, |
| "step": 150 |
| }, |
| { |
| "entropy": 1.099366496130824, |
| "epoch": 0.256, |
| "grad_norm": 20.875, |
| "learning_rate": 1.0600000000000002e-05, |
| "loss": 17.5038, |
| "mean_token_accuracy": 0.7470747366547584, |
| "num_tokens": 4531471.0, |
| "step": 160 |
| }, |
| { |
| "entropy": 1.1072874456644057, |
| "epoch": 0.272, |
| "grad_norm": 18.125, |
| "learning_rate": 1.1266666666666668e-05, |
| "loss": 17.6358, |
| "mean_token_accuracy": 0.7453399378806352, |
| "num_tokens": 4810284.0, |
| "step": 170 |
| }, |
| { |
| "entropy": 1.0879811342805623, |
| "epoch": 0.288, |
| "grad_norm": 16.375, |
| "learning_rate": 1.1933333333333335e-05, |
| "loss": 17.2142, |
| "mean_token_accuracy": 0.749498899653554, |
| "num_tokens": 5095104.0, |
| "step": 180 |
| }, |
| { |
| "entropy": 1.0824080746620894, |
| "epoch": 0.304, |
| "grad_norm": 16.625, |
| "learning_rate": 1.2600000000000001e-05, |
| "loss": 17.223, |
| "mean_token_accuracy": 0.7493264392018318, |
| "num_tokens": 5383732.0, |
| "step": 190 |
| }, |
| { |
| "entropy": 1.048988512158394, |
| "epoch": 0.32, |
| "grad_norm": 19.125, |
| "learning_rate": 1.3266666666666668e-05, |
| "loss": 16.7231, |
| "mean_token_accuracy": 0.7555124312639236, |
| "num_tokens": 5676334.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_biology_entropy": 3.5249868144989014, |
| "eval_biology_loss": 4.3318328857421875, |
| "eval_biology_mean_token_accuracy": 0.39426790857315064, |
| "eval_biology_num_tokens": 5676334.0, |
| "eval_biology_runtime": 49.8836, |
| "eval_biology_samples_per_second": 10.023, |
| "eval_biology_steps_per_second": 2.506, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_chemistry_entropy": 2.3051584205627442, |
| "eval_chemistry_loss": 2.4184799194335938, |
| "eval_chemistry_mean_token_accuracy": 0.5702217166423798, |
| "eval_chemistry_num_tokens": 5676334.0, |
| "eval_chemistry_runtime": 61.8203, |
| "eval_chemistry_samples_per_second": 8.088, |
| "eval_chemistry_steps_per_second": 2.022, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_math_entropy": 1.0760114483833314, |
| "eval_math_loss": 1.0703883171081543, |
| "eval_math_mean_token_accuracy": 0.7492856097221374, |
| "eval_math_num_tokens": 5676334.0, |
| "eval_math_runtime": 63.3918, |
| "eval_math_samples_per_second": 7.887, |
| "eval_math_steps_per_second": 1.972, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_physics_entropy": 1.7677014241218567, |
| "eval_physics_loss": 1.8187588453292847, |
| "eval_physics_mean_token_accuracy": 0.6450894718170166, |
| "eval_physics_num_tokens": 5676334.0, |
| "eval_physics_runtime": 72.1893, |
| "eval_physics_samples_per_second": 6.926, |
| "eval_physics_steps_per_second": 1.732, |
| "step": 200 |
| }, |
| { |
| "entropy": 1.0659823317080737, |
| "epoch": 0.336, |
| "grad_norm": 18.125, |
| "learning_rate": 1.3933333333333334e-05, |
| "loss": 17.0855, |
| "mean_token_accuracy": 0.7487949147820473, |
| "num_tokens": 5958480.0, |
| "step": 210 |
| }, |
| { |
| "entropy": 1.0319596778601408, |
| "epoch": 0.352, |
| "grad_norm": 16.625, |
| "learning_rate": 1.46e-05, |
| "loss": 16.3603, |
| "mean_token_accuracy": 0.7570414818823338, |
| "num_tokens": 6242161.0, |
| "step": 220 |
| }, |
| { |
| "entropy": 1.0307059772312641, |
| "epoch": 0.368, |
| "grad_norm": 19.125, |
| "learning_rate": 1.5266666666666667e-05, |
| "loss": 16.4803, |
| "mean_token_accuracy": 0.7558304250240326, |
| "num_tokens": 6523679.0, |
| "step": 230 |
| }, |
| { |
| "entropy": 1.008485846966505, |
| "epoch": 0.384, |
| "grad_norm": 17.75, |
| "learning_rate": 1.5933333333333336e-05, |
| "loss": 16.1119, |
| "mean_token_accuracy": 0.7602646600455045, |
| "num_tokens": 6810978.0, |
| "step": 240 |
| }, |
| { |
| "entropy": 1.0437373481690884, |
| "epoch": 0.4, |
| "grad_norm": 17.125, |
| "learning_rate": 1.66e-05, |
| "loss": 16.936, |
| "mean_token_accuracy": 0.7534951154142618, |
| "num_tokens": 7096903.0, |
| "step": 250 |
| }, |
| { |
| "entropy": 1.0108945474028588, |
| "epoch": 0.416, |
| "grad_norm": 16.75, |
| "learning_rate": 1.726666666666667e-05, |
| "loss": 16.165, |
| "mean_token_accuracy": 0.7585977960377932, |
| "num_tokens": 7377181.0, |
| "step": 260 |
| }, |
| { |
| "entropy": 0.9808746818453074, |
| "epoch": 0.432, |
| "grad_norm": 19.125, |
| "learning_rate": 1.7933333333333333e-05, |
| "loss": 15.708, |
| "mean_token_accuracy": 0.7660888392478228, |
| "num_tokens": 7650523.0, |
| "step": 270 |
| }, |
| { |
| "entropy": 0.9656771093606948, |
| "epoch": 0.448, |
| "grad_norm": 19.375, |
| "learning_rate": 1.86e-05, |
| "loss": 15.4934, |
| "mean_token_accuracy": 0.7665207251906395, |
| "num_tokens": 7936788.0, |
| "step": 280 |
| }, |
| { |
| "entropy": 0.970192850753665, |
| "epoch": 0.464, |
| "grad_norm": 21.125, |
| "learning_rate": 1.926666666666667e-05, |
| "loss": 15.697, |
| "mean_token_accuracy": 0.7657037891447545, |
| "num_tokens": 8223147.0, |
| "step": 290 |
| }, |
| { |
| "entropy": 0.9936408184468746, |
| "epoch": 0.48, |
| "grad_norm": 18.375, |
| "learning_rate": 1.9933333333333334e-05, |
| "loss": 15.8279, |
| "mean_token_accuracy": 0.7644913751631975, |
| "num_tokens": 8506030.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_biology_entropy": 3.547258296966553, |
| "eval_biology_loss": 4.036286354064941, |
| "eval_biology_mean_token_accuracy": 0.4092668824195862, |
| "eval_biology_num_tokens": 8506030.0, |
| "eval_biology_runtime": 49.8646, |
| "eval_biology_samples_per_second": 10.027, |
| "eval_biology_steps_per_second": 2.507, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_chemistry_entropy": 2.243741189956665, |
| "eval_chemistry_loss": 2.2735037803649902, |
| "eval_chemistry_mean_token_accuracy": 0.585998829126358, |
| "eval_chemistry_num_tokens": 8506030.0, |
| "eval_chemistry_runtime": 61.8087, |
| "eval_chemistry_samples_per_second": 8.089, |
| "eval_chemistry_steps_per_second": 2.022, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_math_entropy": 0.969578085899353, |
| "eval_math_loss": 0.9849457740783691, |
| "eval_math_mean_token_accuracy": 0.7633716745376586, |
| "eval_math_num_tokens": 8506030.0, |
| "eval_math_runtime": 63.4567, |
| "eval_math_samples_per_second": 7.879, |
| "eval_math_steps_per_second": 1.97, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_physics_entropy": 1.6970900883674622, |
| "eval_physics_loss": 1.7149206399917603, |
| "eval_physics_mean_token_accuracy": 0.6588896939754486, |
| "eval_physics_num_tokens": 8506030.0, |
| "eval_physics_runtime": 72.1904, |
| "eval_physics_samples_per_second": 6.926, |
| "eval_physics_steps_per_second": 1.732, |
| "step": 300 |
| }, |
| { |
| "entropy": 0.9509127460420131, |
| "epoch": 0.496, |
| "grad_norm": 27.125, |
| "learning_rate": 1.9933333333333334e-05, |
| "loss": 15.2509, |
| "mean_token_accuracy": 0.7699485655874014, |
| "num_tokens": 8788726.0, |
| "step": 310 |
| }, |
| { |
| "entropy": 0.9583811838179827, |
| "epoch": 0.512, |
| "grad_norm": 17.75, |
| "learning_rate": 1.985925925925926e-05, |
| "loss": 15.3156, |
| "mean_token_accuracy": 0.7701035752892494, |
| "num_tokens": 9078039.0, |
| "step": 320 |
| }, |
| { |
| "entropy": 0.9286689113825559, |
| "epoch": 0.528, |
| "grad_norm": 21.375, |
| "learning_rate": 1.9785185185185187e-05, |
| "loss": 14.9975, |
| "mean_token_accuracy": 0.7747602745890617, |
| "num_tokens": 9373860.0, |
| "step": 330 |
| }, |
| { |
| "entropy": 0.93574563190341, |
| "epoch": 0.544, |
| "grad_norm": 19.0, |
| "learning_rate": 1.971111111111111e-05, |
| "loss": 15.1982, |
| "mean_token_accuracy": 0.7723268665373325, |
| "num_tokens": 9660940.0, |
| "step": 340 |
| }, |
| { |
| "entropy": 0.9833962079137564, |
| "epoch": 0.56, |
| "grad_norm": 27.375, |
| "learning_rate": 1.963703703703704e-05, |
| "loss": 15.6566, |
| "mean_token_accuracy": 0.7635074742138386, |
| "num_tokens": 9932302.0, |
| "step": 350 |
| }, |
| { |
| "entropy": 0.9380292896181345, |
| "epoch": 0.576, |
| "grad_norm": 19.5, |
| "learning_rate": 1.9562962962962964e-05, |
| "loss": 15.0896, |
| "mean_token_accuracy": 0.771076611801982, |
| "num_tokens": 10215462.0, |
| "step": 360 |
| }, |
| { |
| "entropy": 0.9226088404655457, |
| "epoch": 0.592, |
| "grad_norm": 17.125, |
| "learning_rate": 1.948888888888889e-05, |
| "loss": 14.9585, |
| "mean_token_accuracy": 0.7726155135780572, |
| "num_tokens": 10504396.0, |
| "step": 370 |
| }, |
| { |
| "entropy": 0.9383664086461068, |
| "epoch": 0.608, |
| "grad_norm": 16.875, |
| "learning_rate": 1.9414814814814817e-05, |
| "loss": 15.0747, |
| "mean_token_accuracy": 0.772657984867692, |
| "num_tokens": 10793126.0, |
| "step": 380 |
| }, |
| { |
| "entropy": 0.9145499892532826, |
| "epoch": 0.624, |
| "grad_norm": 18.875, |
| "learning_rate": 1.9340740740740743e-05, |
| "loss": 14.7611, |
| "mean_token_accuracy": 0.7738783087581396, |
| "num_tokens": 11081768.0, |
| "step": 390 |
| }, |
| { |
| "entropy": 0.9076093930751086, |
| "epoch": 0.64, |
| "grad_norm": 17.5, |
| "learning_rate": 1.926666666666667e-05, |
| "loss": 14.491, |
| "mean_token_accuracy": 0.7761218074709177, |
| "num_tokens": 11370320.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_biology_entropy": 3.4644125595092774, |
| "eval_biology_loss": 3.9791781902313232, |
| "eval_biology_mean_token_accuracy": 0.4146076333522797, |
| "eval_biology_num_tokens": 11370320.0, |
| "eval_biology_runtime": 49.8972, |
| "eval_biology_samples_per_second": 10.021, |
| "eval_biology_steps_per_second": 2.505, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_chemistry_entropy": 2.1470180315971374, |
| "eval_chemistry_loss": 2.1781718730926514, |
| "eval_chemistry_mean_token_accuracy": 0.5962020962238311, |
| "eval_chemistry_num_tokens": 11370320.0, |
| "eval_chemistry_runtime": 61.8451, |
| "eval_chemistry_samples_per_second": 8.085, |
| "eval_chemistry_steps_per_second": 2.021, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_math_entropy": 0.9346609992980957, |
| "eval_math_loss": 0.9384942054748535, |
| "eval_math_mean_token_accuracy": 0.771290735244751, |
| "eval_math_num_tokens": 11370320.0, |
| "eval_math_runtime": 63.3885, |
| "eval_math_samples_per_second": 7.888, |
| "eval_math_steps_per_second": 1.972, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_physics_entropy": 1.6158739218711853, |
| "eval_physics_loss": 1.6468137502670288, |
| "eval_physics_mean_token_accuracy": 0.6678030126094818, |
| "eval_physics_num_tokens": 11370320.0, |
| "eval_physics_runtime": 72.1919, |
| "eval_physics_samples_per_second": 6.926, |
| "eval_physics_steps_per_second": 1.731, |
| "step": 400 |
| }, |
| { |
| "entropy": 0.8844340030103922, |
| "epoch": 0.656, |
| "grad_norm": 15.625, |
| "learning_rate": 1.9192592592592593e-05, |
| "loss": 14.2919, |
| "mean_token_accuracy": 0.7790201846510172, |
| "num_tokens": 11657835.0, |
| "step": 410 |
| }, |
| { |
| "entropy": 0.9230461683124304, |
| "epoch": 0.672, |
| "grad_norm": 17.25, |
| "learning_rate": 1.911851851851852e-05, |
| "loss": 14.8067, |
| "mean_token_accuracy": 0.7730407487601042, |
| "num_tokens": 11949262.0, |
| "step": 420 |
| }, |
| { |
| "entropy": 0.8864722024649382, |
| "epoch": 0.688, |
| "grad_norm": 14.5625, |
| "learning_rate": 1.9044444444444446e-05, |
| "loss": 14.2543, |
| "mean_token_accuracy": 0.7801163170486689, |
| "num_tokens": 12227640.0, |
| "step": 430 |
| }, |
| { |
| "entropy": 0.8921929065138101, |
| "epoch": 0.704, |
| "grad_norm": 17.5, |
| "learning_rate": 1.8970370370370372e-05, |
| "loss": 14.3854, |
| "mean_token_accuracy": 0.7790493600070476, |
| "num_tokens": 12516641.0, |
| "step": 440 |
| }, |
| { |
| "entropy": 0.9114352393895387, |
| "epoch": 0.72, |
| "grad_norm": 18.75, |
| "learning_rate": 1.8896296296296295e-05, |
| "loss": 14.7549, |
| "mean_token_accuracy": 0.7744539573788642, |
| "num_tokens": 12793343.0, |
| "step": 450 |
| }, |
| { |
| "entropy": 0.9004437670111656, |
| "epoch": 0.736, |
| "grad_norm": 16.625, |
| "learning_rate": 1.8822222222222225e-05, |
| "loss": 14.4296, |
| "mean_token_accuracy": 0.7780555929988623, |
| "num_tokens": 13077981.0, |
| "step": 460 |
| }, |
| { |
| "entropy": 0.8800053089857102, |
| "epoch": 0.752, |
| "grad_norm": 19.5, |
| "learning_rate": 1.874814814814815e-05, |
| "loss": 14.1044, |
| "mean_token_accuracy": 0.7824321106076241, |
| "num_tokens": 13358957.0, |
| "step": 470 |
| }, |
| { |
| "entropy": 0.8506957441568375, |
| "epoch": 0.768, |
| "grad_norm": 15.25, |
| "learning_rate": 1.8674074074074075e-05, |
| "loss": 13.7613, |
| "mean_token_accuracy": 0.7884838156402111, |
| "num_tokens": 13653412.0, |
| "step": 480 |
| }, |
| { |
| "entropy": 0.9173242174088955, |
| "epoch": 0.784, |
| "grad_norm": 18.125, |
| "learning_rate": 1.86e-05, |
| "loss": 14.7366, |
| "mean_token_accuracy": 0.776311819255352, |
| "num_tokens": 13940856.0, |
| "step": 490 |
| }, |
| { |
| "entropy": 0.9153516864404082, |
| "epoch": 0.8, |
| "grad_norm": 18.5, |
| "learning_rate": 1.8525925925925928e-05, |
| "loss": 14.84, |
| "mean_token_accuracy": 0.7749685924500227, |
| "num_tokens": 14230754.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_biology_entropy": 3.587346399307251, |
| "eval_biology_loss": 3.8396143913269043, |
| "eval_biology_mean_token_accuracy": 0.42098368620872495, |
| "eval_biology_num_tokens": 14230754.0, |
| "eval_biology_runtime": 49.8808, |
| "eval_biology_samples_per_second": 10.024, |
| "eval_biology_steps_per_second": 2.506, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_chemistry_entropy": 2.168390507221222, |
| "eval_chemistry_loss": 2.1044840812683105, |
| "eval_chemistry_mean_token_accuracy": 0.6037136344909668, |
| "eval_chemistry_num_tokens": 14230754.0, |
| "eval_chemistry_runtime": 61.4676, |
| "eval_chemistry_samples_per_second": 8.134, |
| "eval_chemistry_steps_per_second": 2.034, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_math_entropy": 0.9173241977691651, |
| "eval_math_loss": 0.911715030670166, |
| "eval_math_mean_token_accuracy": 0.7763807888031006, |
| "eval_math_num_tokens": 14230754.0, |
| "eval_math_runtime": 63.353, |
| "eval_math_samples_per_second": 7.892, |
| "eval_math_steps_per_second": 1.973, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_physics_entropy": 1.6311655960083007, |
| "eval_physics_loss": 1.610815167427063, |
| "eval_physics_mean_token_accuracy": 0.6733945074081421, |
| "eval_physics_num_tokens": 14230754.0, |
| "eval_physics_runtime": 72.3928, |
| "eval_physics_samples_per_second": 6.907, |
| "eval_physics_steps_per_second": 1.727, |
| "step": 500 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 3000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 5, |
| "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": 1.448451560178856e+18, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
