| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.2, |
| "eval_steps": 100, |
| "global_step": 2000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 0.9806057829409838, |
| "epoch": 0.016, |
| "grad_norm": 157.0, |
| "learning_rate": 1.8e-07, |
| "loss": 21.7295, |
| "mean_token_accuracy": 0.695003604888916, |
| "num_tokens": 280941.0, |
| "step": 10 |
| }, |
| { |
| "entropy": 1.0097382467240095, |
| "epoch": 0.032, |
| "grad_norm": 171.0, |
| "learning_rate": 3.8e-07, |
| "loss": 21.935, |
| "mean_token_accuracy": 0.6911116372793913, |
| "num_tokens": 558056.0, |
| "step": 20 |
| }, |
| { |
| "entropy": 1.0059957668185233, |
| "epoch": 0.048, |
| "grad_norm": 146.0, |
| "learning_rate": 5.800000000000001e-07, |
| "loss": 21.4243, |
| "mean_token_accuracy": 0.6972350142896175, |
| "num_tokens": 836753.0, |
| "step": 30 |
| }, |
| { |
| "entropy": 1.0606450594961643, |
| "epoch": 0.064, |
| "grad_norm": 129.0, |
| "learning_rate": 7.8e-07, |
| "loss": 22.2256, |
| "mean_token_accuracy": 0.6895487096160651, |
| "num_tokens": 1126446.0, |
| "step": 40 |
| }, |
| { |
| "entropy": 1.1183118436485529, |
| "epoch": 0.08, |
| "grad_norm": 99.5, |
| "learning_rate": 9.800000000000001e-07, |
| "loss": 21.2808, |
| "mean_token_accuracy": 0.6964452721178531, |
| "num_tokens": 1413596.0, |
| "step": 50 |
| }, |
| { |
| "entropy": 1.1463136691600084, |
| "epoch": 0.096, |
| "grad_norm": 83.0, |
| "learning_rate": 1.1800000000000001e-06, |
| "loss": 20.5501, |
| "mean_token_accuracy": 0.7033276192843914, |
| "num_tokens": 1701193.0, |
| "step": 60 |
| }, |
| { |
| "entropy": 1.1966488853096962, |
| "epoch": 0.112, |
| "grad_norm": 75.5, |
| "learning_rate": 1.3800000000000001e-06, |
| "loss": 20.5984, |
| "mean_token_accuracy": 0.7012545391917229, |
| "num_tokens": 1979232.0, |
| "step": 70 |
| }, |
| { |
| "entropy": 1.2089795324951411, |
| "epoch": 0.128, |
| "grad_norm": 69.0, |
| "learning_rate": 1.5800000000000001e-06, |
| "loss": 20.2753, |
| "mean_token_accuracy": 0.7075193412601948, |
| "num_tokens": 2274177.0, |
| "step": 80 |
| }, |
| { |
| "entropy": 1.1860636565834284, |
| "epoch": 0.144, |
| "grad_norm": 70.0, |
| "learning_rate": 1.7800000000000001e-06, |
| "loss": 19.7829, |
| "mean_token_accuracy": 0.710378497838974, |
| "num_tokens": 2548445.0, |
| "step": 90 |
| }, |
| { |
| "entropy": 1.1838629063218833, |
| "epoch": 0.16, |
| "grad_norm": 85.5, |
| "learning_rate": 1.98e-06, |
| "loss": 19.4179, |
| "mean_token_accuracy": 0.7149847097694874, |
| "num_tokens": 2824418.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_biology_entropy": 1.474690469264984, |
| "eval_biology_loss": 1.3576592206954956, |
| "eval_biology_mean_token_accuracy": 0.6709716210365295, |
| "eval_biology_num_tokens": 2824418.0, |
| "eval_biology_runtime": 22.204, |
| "eval_biology_samples_per_second": 22.518, |
| "eval_biology_steps_per_second": 5.63, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_chemistry_entropy": 1.247537253856659, |
| "eval_chemistry_loss": 1.1488269567489624, |
| "eval_chemistry_mean_token_accuracy": 0.7154391117095947, |
| "eval_chemistry_num_tokens": 2824418.0, |
| "eval_chemistry_runtime": 26.8594, |
| "eval_chemistry_samples_per_second": 18.615, |
| "eval_chemistry_steps_per_second": 4.654, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_math_entropy": 1.2081865873336792, |
| "eval_math_loss": 1.2300448417663574, |
| "eval_math_mean_token_accuracy": 0.710663827419281, |
| "eval_math_num_tokens": 2824418.0, |
| "eval_math_runtime": 27.5187, |
| "eval_math_samples_per_second": 18.169, |
| "eval_math_steps_per_second": 4.542, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_cyber_entropy": 3.049607857465744, |
| "eval_cyber_loss": 3.3160624504089355, |
| "eval_cyber_mean_token_accuracy": 0.4259996695816517, |
| "eval_cyber_num_tokens": 2824418.0, |
| "eval_cyber_runtime": 26.1305, |
| "eval_cyber_samples_per_second": 15.193, |
| "eval_cyber_steps_per_second": 3.827, |
| "step": 100 |
| }, |
| { |
| "entropy": 1.1945379309356212, |
| "epoch": 0.176, |
| "grad_norm": 78.5, |
| "learning_rate": 2.1800000000000003e-06, |
| "loss": 19.3659, |
| "mean_token_accuracy": 0.7157268539071083, |
| "num_tokens": 3110313.0, |
| "step": 110 |
| }, |
| { |
| "entropy": 1.1879339709877967, |
| "epoch": 0.192, |
| "grad_norm": 62.5, |
| "learning_rate": 2.38e-06, |
| "loss": 19.2957, |
| "mean_token_accuracy": 0.7150671608746052, |
| "num_tokens": 3394170.0, |
| "step": 120 |
| }, |
| { |
| "entropy": 1.1658376209437846, |
| "epoch": 0.208, |
| "grad_norm": 64.5, |
| "learning_rate": 2.5800000000000003e-06, |
| "loss": 18.9243, |
| "mean_token_accuracy": 0.7201551966369152, |
| "num_tokens": 3673600.0, |
| "step": 130 |
| }, |
| { |
| "entropy": 1.1712732832878827, |
| "epoch": 0.224, |
| "grad_norm": 62.25, |
| "learning_rate": 2.7800000000000005e-06, |
| "loss": 18.8985, |
| "mean_token_accuracy": 0.7195578265935183, |
| "num_tokens": 3953732.0, |
| "step": 140 |
| }, |
| { |
| "entropy": 1.1372143357992173, |
| "epoch": 0.24, |
| "grad_norm": 59.25, |
| "learning_rate": 2.9800000000000003e-06, |
| "loss": 18.3388, |
| "mean_token_accuracy": 0.7276442721486092, |
| "num_tokens": 4243655.0, |
| "step": 150 |
| }, |
| { |
| "entropy": 1.1075575590133666, |
| "epoch": 0.256, |
| "grad_norm": 55.5, |
| "learning_rate": 3.1800000000000005e-06, |
| "loss": 17.7317, |
| "mean_token_accuracy": 0.7332122329622507, |
| "num_tokens": 4531471.0, |
| "step": 160 |
| }, |
| { |
| "entropy": 1.1275247156620025, |
| "epoch": 0.272, |
| "grad_norm": 53.0, |
| "learning_rate": 3.3800000000000007e-06, |
| "loss": 18.2164, |
| "mean_token_accuracy": 0.7298861864954234, |
| "num_tokens": 4810284.0, |
| "step": 170 |
| }, |
| { |
| "entropy": 1.1071187134832143, |
| "epoch": 0.288, |
| "grad_norm": 49.25, |
| "learning_rate": 3.58e-06, |
| "loss": 17.7625, |
| "mean_token_accuracy": 0.7337470225989818, |
| "num_tokens": 5095104.0, |
| "step": 180 |
| }, |
| { |
| "entropy": 1.1132069051265716, |
| "epoch": 0.304, |
| "grad_norm": 54.25, |
| "learning_rate": 3.7800000000000002e-06, |
| "loss": 17.7962, |
| "mean_token_accuracy": 0.7336070898920297, |
| "num_tokens": 5383732.0, |
| "step": 190 |
| }, |
| { |
| "entropy": 1.0761282254010438, |
| "epoch": 0.32, |
| "grad_norm": 51.0, |
| "learning_rate": 3.980000000000001e-06, |
| "loss": 17.3026, |
| "mean_token_accuracy": 0.7393171060830355, |
| "num_tokens": 5676334.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_biology_entropy": 1.4338086099624634, |
| "eval_biology_loss": 1.3705029487609863, |
| "eval_biology_mean_token_accuracy": 0.6675603828430176, |
| "eval_biology_num_tokens": 5676334.0, |
| "eval_biology_runtime": 22.0314, |
| "eval_biology_samples_per_second": 22.695, |
| "eval_biology_steps_per_second": 5.674, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_chemistry_entropy": 1.1902209458351136, |
| "eval_chemistry_loss": 1.1403467655181885, |
| "eval_chemistry_mean_token_accuracy": 0.7157996978759765, |
| "eval_chemistry_num_tokens": 5676334.0, |
| "eval_chemistry_runtime": 26.8655, |
| "eval_chemistry_samples_per_second": 18.611, |
| "eval_chemistry_steps_per_second": 4.653, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_math_entropy": 1.1017972359657286, |
| "eval_math_loss": 1.1044487953186035, |
| "eval_math_mean_token_accuracy": 0.7332207527160645, |
| "eval_math_num_tokens": 5676334.0, |
| "eval_math_runtime": 27.5365, |
| "eval_math_samples_per_second": 18.158, |
| "eval_math_steps_per_second": 4.539, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_cyber_entropy": 2.916593015193939, |
| "eval_cyber_loss": 3.2743771076202393, |
| "eval_cyber_mean_token_accuracy": 0.4284310387074947, |
| "eval_cyber_num_tokens": 5676334.0, |
| "eval_cyber_runtime": 26.1262, |
| "eval_cyber_samples_per_second": 15.195, |
| "eval_cyber_steps_per_second": 3.828, |
| "step": 200 |
| }, |
| { |
| "entropy": 1.0972535338252782, |
| "epoch": 0.336, |
| "grad_norm": 48.0, |
| "learning_rate": 4.18e-06, |
| "loss": 17.6967, |
| "mean_token_accuracy": 0.7339770793914795, |
| "num_tokens": 5958480.0, |
| "step": 210 |
| }, |
| { |
| "entropy": 1.0661637954413892, |
| "epoch": 0.352, |
| "grad_norm": 52.0, |
| "learning_rate": 4.38e-06, |
| "loss": 17.0451, |
| "mean_token_accuracy": 0.7420999370515347, |
| "num_tokens": 6242161.0, |
| "step": 220 |
| }, |
| { |
| "entropy": 1.0574691709131003, |
| "epoch": 0.368, |
| "grad_norm": 43.75, |
| "learning_rate": 4.58e-06, |
| "loss": 17.0476, |
| "mean_token_accuracy": 0.7413399379700423, |
| "num_tokens": 6523679.0, |
| "step": 230 |
| }, |
| { |
| "entropy": 1.0383559666574, |
| "epoch": 0.384, |
| "grad_norm": 52.75, |
| "learning_rate": 4.78e-06, |
| "loss": 16.6413, |
| "mean_token_accuracy": 0.7455471660941839, |
| "num_tokens": 6810978.0, |
| "step": 240 |
| }, |
| { |
| "entropy": 1.0844071809202434, |
| "epoch": 0.4, |
| "grad_norm": 54.25, |
| "learning_rate": 4.980000000000001e-06, |
| "loss": 17.5456, |
| "mean_token_accuracy": 0.7374324273318053, |
| "num_tokens": 7096903.0, |
| "step": 250 |
| }, |
| { |
| "entropy": 1.0462343256920577, |
| "epoch": 0.416, |
| "grad_norm": 46.25, |
| "learning_rate": 5.18e-06, |
| "loss": 16.7333, |
| "mean_token_accuracy": 0.7437998823821544, |
| "num_tokens": 7377181.0, |
| "step": 260 |
| }, |
| { |
| "entropy": 1.0309648185968399, |
| "epoch": 0.432, |
| "grad_norm": 53.5, |
| "learning_rate": 5.380000000000001e-06, |
| "loss": 16.3635, |
| "mean_token_accuracy": 0.7494884602725506, |
| "num_tokens": 7650523.0, |
| "step": 270 |
| }, |
| { |
| "entropy": 1.011606451496482, |
| "epoch": 0.448, |
| "grad_norm": 55.5, |
| "learning_rate": 5.580000000000001e-06, |
| "loss": 16.2941, |
| "mean_token_accuracy": 0.7498481426388025, |
| "num_tokens": 7936788.0, |
| "step": 280 |
| }, |
| { |
| "entropy": 1.0099118243902923, |
| "epoch": 0.464, |
| "grad_norm": 46.25, |
| "learning_rate": 5.78e-06, |
| "loss": 16.2711, |
| "mean_token_accuracy": 0.7521415069699288, |
| "num_tokens": 8223147.0, |
| "step": 290 |
| }, |
| { |
| "entropy": 1.0334131706506013, |
| "epoch": 0.48, |
| "grad_norm": 43.25, |
| "learning_rate": 5.98e-06, |
| "loss": 16.4655, |
| "mean_token_accuracy": 0.748864620923996, |
| "num_tokens": 8506030.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_biology_entropy": 1.422005220413208, |
| "eval_biology_loss": 1.3784860372543335, |
| "eval_biology_mean_token_accuracy": 0.6655148763656616, |
| "eval_biology_num_tokens": 8506030.0, |
| "eval_biology_runtime": 22.0223, |
| "eval_biology_samples_per_second": 22.704, |
| "eval_biology_steps_per_second": 5.676, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_chemistry_entropy": 1.1605761876106262, |
| "eval_chemistry_loss": 1.1291848421096802, |
| "eval_chemistry_mean_token_accuracy": 0.7179730429649352, |
| "eval_chemistry_num_tokens": 8506030.0, |
| "eval_chemistry_runtime": 26.8904, |
| "eval_chemistry_samples_per_second": 18.594, |
| "eval_chemistry_steps_per_second": 4.648, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_math_entropy": 1.0216443300247193, |
| "eval_math_loss": 1.0201935768127441, |
| "eval_math_mean_token_accuracy": 0.749815523147583, |
| "eval_math_num_tokens": 8506030.0, |
| "eval_math_runtime": 27.5335, |
| "eval_math_samples_per_second": 18.16, |
| "eval_math_steps_per_second": 4.54, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_cyber_entropy": 2.940396952629089, |
| "eval_cyber_loss": 3.153137683868408, |
| "eval_cyber_mean_token_accuracy": 0.4395553506910801, |
| "eval_cyber_num_tokens": 8506030.0, |
| "eval_cyber_runtime": 26.1452, |
| "eval_cyber_samples_per_second": 15.184, |
| "eval_cyber_steps_per_second": 3.825, |
| "step": 300 |
| }, |
| { |
| "entropy": 0.9862126674503088, |
| "epoch": 0.496, |
| "grad_norm": 56.25, |
| "learning_rate": 6.18e-06, |
| "loss": 15.7919, |
| "mean_token_accuracy": 0.7557632889598608, |
| "num_tokens": 8788726.0, |
| "step": 310 |
| }, |
| { |
| "entropy": 0.9811401419341564, |
| "epoch": 0.512, |
| "grad_norm": 46.25, |
| "learning_rate": 6.380000000000001e-06, |
| "loss": 15.7454, |
| "mean_token_accuracy": 0.7576492365449667, |
| "num_tokens": 9078039.0, |
| "step": 320 |
| }, |
| { |
| "entropy": 0.9647420089691877, |
| "epoch": 0.528, |
| "grad_norm": 44.5, |
| "learning_rate": 6.5800000000000005e-06, |
| "loss": 15.4619, |
| "mean_token_accuracy": 0.7620009411126375, |
| "num_tokens": 9373860.0, |
| "step": 330 |
| }, |
| { |
| "entropy": 0.9784351203590631, |
| "epoch": 0.544, |
| "grad_norm": 51.75, |
| "learning_rate": 6.780000000000001e-06, |
| "loss": 15.6775, |
| "mean_token_accuracy": 0.7592026349157095, |
| "num_tokens": 9660940.0, |
| "step": 340 |
| }, |
| { |
| "entropy": 0.9894711822271347, |
| "epoch": 0.56, |
| "grad_norm": 62.25, |
| "learning_rate": 6.98e-06, |
| "loss": 15.8268, |
| "mean_token_accuracy": 0.754706758633256, |
| "num_tokens": 9932302.0, |
| "step": 350 |
| }, |
| { |
| "entropy": 0.9637547507882118, |
| "epoch": 0.576, |
| "grad_norm": 45.0, |
| "learning_rate": 7.180000000000001e-06, |
| "loss": 15.4676, |
| "mean_token_accuracy": 0.7605381533503532, |
| "num_tokens": 10215462.0, |
| "step": 360 |
| }, |
| { |
| "entropy": 0.9582533340901136, |
| "epoch": 0.592, |
| "grad_norm": 42.5, |
| "learning_rate": 7.3800000000000005e-06, |
| "loss": 15.343, |
| "mean_token_accuracy": 0.7618525486439467, |
| "num_tokens": 10504396.0, |
| "step": 370 |
| }, |
| { |
| "entropy": 0.9651506002992392, |
| "epoch": 0.608, |
| "grad_norm": 42.75, |
| "learning_rate": 7.58e-06, |
| "loss": 15.4775, |
| "mean_token_accuracy": 0.7616991735994816, |
| "num_tokens": 10793126.0, |
| "step": 380 |
| }, |
| { |
| "entropy": 0.951299836859107, |
| "epoch": 0.624, |
| "grad_norm": 46.0, |
| "learning_rate": 7.78e-06, |
| "loss": 15.2327, |
| "mean_token_accuracy": 0.7628108691424131, |
| "num_tokens": 11081768.0, |
| "step": 390 |
| }, |
| { |
| "entropy": 0.9370664428919554, |
| "epoch": 0.64, |
| "grad_norm": 45.75, |
| "learning_rate": 7.980000000000002e-06, |
| "loss": 14.9056, |
| "mean_token_accuracy": 0.7668638564646244, |
| "num_tokens": 11370320.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_biology_entropy": 1.4101261868476869, |
| "eval_biology_loss": 1.385290503501892, |
| "eval_biology_mean_token_accuracy": 0.6643283500671386, |
| "eval_biology_num_tokens": 11370320.0, |
| "eval_biology_runtime": 22.0311, |
| "eval_biology_samples_per_second": 22.695, |
| "eval_biology_steps_per_second": 5.674, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_chemistry_entropy": 1.134839651107788, |
| "eval_chemistry_loss": 1.117846965789795, |
| "eval_chemistry_mean_token_accuracy": 0.721014030456543, |
| "eval_chemistry_num_tokens": 11370320.0, |
| "eval_chemistry_runtime": 26.9193, |
| "eval_chemistry_samples_per_second": 18.574, |
| "eval_chemistry_steps_per_second": 4.644, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_math_entropy": 0.9476065578460693, |
| "eval_math_loss": 0.9578045606613159, |
| "eval_math_mean_token_accuracy": 0.760874231338501, |
| "eval_math_num_tokens": 11370320.0, |
| "eval_math_runtime": 27.5198, |
| "eval_math_samples_per_second": 18.169, |
| "eval_math_steps_per_second": 4.542, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_cyber_entropy": 2.8283654510974885, |
| "eval_cyber_loss": 3.0569260120391846, |
| "eval_cyber_mean_token_accuracy": 0.4431245893239975, |
| "eval_cyber_num_tokens": 11370320.0, |
| "eval_cyber_runtime": 26.384, |
| "eval_cyber_samples_per_second": 15.047, |
| "eval_cyber_steps_per_second": 3.79, |
| "step": 400 |
| }, |
| { |
| "entropy": 0.9207184508442878, |
| "epoch": 0.656, |
| "grad_norm": 44.5, |
| "learning_rate": 8.18e-06, |
| "loss": 14.748, |
| "mean_token_accuracy": 0.7685097701847553, |
| "num_tokens": 11657835.0, |
| "step": 410 |
| }, |
| { |
| "entropy": 0.9384444292634726, |
| "epoch": 0.672, |
| "grad_norm": 44.5, |
| "learning_rate": 8.380000000000001e-06, |
| "loss": 15.0013, |
| "mean_token_accuracy": 0.7654238797724247, |
| "num_tokens": 11949262.0, |
| "step": 420 |
| }, |
| { |
| "entropy": 0.9148579228669405, |
| "epoch": 0.688, |
| "grad_norm": 35.75, |
| "learning_rate": 8.580000000000001e-06, |
| "loss": 14.6923, |
| "mean_token_accuracy": 0.7695376992225647, |
| "num_tokens": 12227640.0, |
| "step": 430 |
| }, |
| { |
| "entropy": 0.9156919397413731, |
| "epoch": 0.704, |
| "grad_norm": 36.5, |
| "learning_rate": 8.78e-06, |
| "loss": 14.6672, |
| "mean_token_accuracy": 0.7705338027328252, |
| "num_tokens": 12516641.0, |
| "step": 440 |
| }, |
| { |
| "entropy": 0.9278485044836998, |
| "epoch": 0.72, |
| "grad_norm": 42.5, |
| "learning_rate": 8.98e-06, |
| "loss": 14.8435, |
| "mean_token_accuracy": 0.7673114899545908, |
| "num_tokens": 12793343.0, |
| "step": 450 |
| }, |
| { |
| "entropy": 0.9052219696342945, |
| "epoch": 0.736, |
| "grad_norm": 41.0, |
| "learning_rate": 9.180000000000002e-06, |
| "loss": 14.5171, |
| "mean_token_accuracy": 0.7729556966573, |
| "num_tokens": 13077981.0, |
| "step": 460 |
| }, |
| { |
| "entropy": 0.8888643320649863, |
| "epoch": 0.752, |
| "grad_norm": 41.25, |
| "learning_rate": 9.38e-06, |
| "loss": 14.1497, |
| "mean_token_accuracy": 0.776068452000618, |
| "num_tokens": 13358957.0, |
| "step": 470 |
| }, |
| { |
| "entropy": 0.8620530396699906, |
| "epoch": 0.768, |
| "grad_norm": 42.5, |
| "learning_rate": 9.58e-06, |
| "loss": 13.834, |
| "mean_token_accuracy": 0.7819891981780529, |
| "num_tokens": 13653412.0, |
| "step": 480 |
| }, |
| { |
| "entropy": 0.9176285572350025, |
| "epoch": 0.784, |
| "grad_norm": 39.0, |
| "learning_rate": 9.780000000000001e-06, |
| "loss": 14.6357, |
| "mean_token_accuracy": 0.7709478087723255, |
| "num_tokens": 13940856.0, |
| "step": 490 |
| }, |
| { |
| "entropy": 0.9128546692430973, |
| "epoch": 0.8, |
| "grad_norm": 40.25, |
| "learning_rate": 9.980000000000001e-06, |
| "loss": 14.5874, |
| "mean_token_accuracy": 0.7713334109634161, |
| "num_tokens": 14230754.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_biology_entropy": 1.4446520280838013, |
| "eval_biology_loss": 1.392669439315796, |
| "eval_biology_mean_token_accuracy": 0.6637401723861694, |
| "eval_biology_num_tokens": 14230754.0, |
| "eval_biology_runtime": 22.2391, |
| "eval_biology_samples_per_second": 22.483, |
| "eval_biology_steps_per_second": 5.621, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_chemistry_entropy": 1.1566239352226257, |
| "eval_chemistry_loss": 1.1047961711883545, |
| "eval_chemistry_mean_token_accuracy": 0.7233014287948608, |
| "eval_chemistry_num_tokens": 14230754.0, |
| "eval_chemistry_runtime": 26.8042, |
| "eval_chemistry_samples_per_second": 18.654, |
| "eval_chemistry_steps_per_second": 4.663, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_math_entropy": 0.9341866765022278, |
| "eval_math_loss": 0.9057817459106445, |
| "eval_math_mean_token_accuracy": 0.7704657621383667, |
| "eval_math_num_tokens": 14230754.0, |
| "eval_math_runtime": 27.5093, |
| "eval_math_samples_per_second": 18.176, |
| "eval_math_steps_per_second": 4.544, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_cyber_entropy": 2.902767553329468, |
| "eval_cyber_loss": 2.947915554046631, |
| "eval_cyber_mean_token_accuracy": 0.45555895671248436, |
| "eval_cyber_num_tokens": 14230754.0, |
| "eval_cyber_runtime": 26.1612, |
| "eval_cyber_samples_per_second": 15.175, |
| "eval_cyber_steps_per_second": 3.822, |
| "step": 500 |
| }, |
| { |
| "entropy": 0.8841698631644249, |
| "epoch": 0.816, |
| "grad_norm": 38.0, |
| "learning_rate": 1.018e-05, |
| "loss": 14.1477, |
| "mean_token_accuracy": 0.7752781912684441, |
| "num_tokens": 14519893.0, |
| "step": 510 |
| }, |
| { |
| "entropy": 0.875461632013321, |
| "epoch": 0.832, |
| "grad_norm": 39.0, |
| "learning_rate": 1.038e-05, |
| "loss": 14.0361, |
| "mean_token_accuracy": 0.7779335591942071, |
| "num_tokens": 14805088.0, |
| "step": 520 |
| }, |
| { |
| "entropy": 0.899658627063036, |
| "epoch": 0.848, |
| "grad_norm": 37.25, |
| "learning_rate": 1.0580000000000002e-05, |
| "loss": 14.3925, |
| "mean_token_accuracy": 0.7728543490171432, |
| "num_tokens": 15086306.0, |
| "step": 530 |
| }, |
| { |
| "entropy": 0.8889949310570955, |
| "epoch": 0.864, |
| "grad_norm": 41.0, |
| "learning_rate": 1.0780000000000002e-05, |
| "loss": 14.1314, |
| "mean_token_accuracy": 0.7759746141731739, |
| "num_tokens": 15370985.0, |
| "step": 540 |
| }, |
| { |
| "entropy": 0.8936371214687824, |
| "epoch": 0.88, |
| "grad_norm": 36.5, |
| "learning_rate": 1.0980000000000002e-05, |
| "loss": 14.2807, |
| "mean_token_accuracy": 0.7754444174468518, |
| "num_tokens": 15653836.0, |
| "step": 550 |
| }, |
| { |
| "entropy": 0.8719608142971993, |
| "epoch": 0.896, |
| "grad_norm": 34.75, |
| "learning_rate": 1.1180000000000001e-05, |
| "loss": 13.9767, |
| "mean_token_accuracy": 0.7785432428121567, |
| "num_tokens": 15932179.0, |
| "step": 560 |
| }, |
| { |
| "entropy": 0.8601628458127379, |
| "epoch": 0.912, |
| "grad_norm": 35.25, |
| "learning_rate": 1.138e-05, |
| "loss": 13.7333, |
| "mean_token_accuracy": 0.77998266518116, |
| "num_tokens": 16219842.0, |
| "step": 570 |
| }, |
| { |
| "entropy": 0.848052042350173, |
| "epoch": 0.928, |
| "grad_norm": 34.25, |
| "learning_rate": 1.1580000000000001e-05, |
| "loss": 13.5598, |
| "mean_token_accuracy": 0.783201026916504, |
| "num_tokens": 16499842.0, |
| "step": 580 |
| }, |
| { |
| "entropy": 0.8647568510845304, |
| "epoch": 0.944, |
| "grad_norm": 98.5, |
| "learning_rate": 1.178e-05, |
| "loss": 13.9513, |
| "mean_token_accuracy": 0.7791629247367382, |
| "num_tokens": 16781882.0, |
| "step": 590 |
| }, |
| { |
| "entropy": 0.8681454580277205, |
| "epoch": 0.96, |
| "grad_norm": 28.375, |
| "learning_rate": 1.198e-05, |
| "loss": 13.8619, |
| "mean_token_accuracy": 0.7806598395109177, |
| "num_tokens": 17067407.0, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.96, |
| "eval_biology_entropy": 1.403841501235962, |
| "eval_biology_loss": 1.3980714082717896, |
| "eval_biology_mean_token_accuracy": 0.6626365647315979, |
| "eval_biology_num_tokens": 17067407.0, |
| "eval_biology_runtime": 22.2675, |
| "eval_biology_samples_per_second": 22.454, |
| "eval_biology_steps_per_second": 5.614, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.96, |
| "eval_chemistry_entropy": 1.099705493927002, |
| "eval_chemistry_loss": 1.0918222665786743, |
| "eval_chemistry_mean_token_accuracy": 0.7260931057929992, |
| "eval_chemistry_num_tokens": 17067407.0, |
| "eval_chemistry_runtime": 26.9078, |
| "eval_chemistry_samples_per_second": 18.582, |
| "eval_chemistry_steps_per_second": 4.645, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.96, |
| "eval_math_entropy": 0.8664147562980652, |
| "eval_math_loss": 0.8665754795074463, |
| "eval_math_mean_token_accuracy": 0.7780344748497009, |
| "eval_math_num_tokens": 17067407.0, |
| "eval_math_runtime": 27.5569, |
| "eval_math_samples_per_second": 18.144, |
| "eval_math_steps_per_second": 4.536, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.96, |
| "eval_cyber_entropy": 2.6499598491191865, |
| "eval_cyber_loss": 2.8601412773132324, |
| "eval_cyber_mean_token_accuracy": 0.4642623996734619, |
| "eval_cyber_num_tokens": 17067407.0, |
| "eval_cyber_runtime": 26.1765, |
| "eval_cyber_samples_per_second": 15.166, |
| "eval_cyber_steps_per_second": 3.82, |
| "step": 600 |
| }, |
| { |
| "entropy": 0.8494564741849899, |
| "epoch": 0.976, |
| "grad_norm": 40.25, |
| "learning_rate": 1.218e-05, |
| "loss": 13.6376, |
| "mean_token_accuracy": 0.7825958080589771, |
| "num_tokens": 17350994.0, |
| "step": 610 |
| }, |
| { |
| "entropy": 0.8730685204267502, |
| "epoch": 0.992, |
| "grad_norm": 38.75, |
| "learning_rate": 1.2380000000000002e-05, |
| "loss": 13.8595, |
| "mean_token_accuracy": 0.7786977473646403, |
| "num_tokens": 17637514.0, |
| "step": 620 |
| }, |
| { |
| "entropy": 0.8468878531828523, |
| "epoch": 1.008, |
| "grad_norm": 32.5, |
| "learning_rate": 1.2580000000000002e-05, |
| "loss": 13.534, |
| "mean_token_accuracy": 0.7822502862662077, |
| "num_tokens": 17926570.0, |
| "step": 630 |
| }, |
| { |
| "entropy": 0.8290399981662631, |
| "epoch": 1.024, |
| "grad_norm": 29.375, |
| "learning_rate": 1.2780000000000001e-05, |
| "loss": 13.2779, |
| "mean_token_accuracy": 0.7864516761153937, |
| "num_tokens": 18207652.0, |
| "step": 640 |
| }, |
| { |
| "entropy": 0.8298395985737443, |
| "epoch": 1.04, |
| "grad_norm": 31.0, |
| "learning_rate": 1.2980000000000001e-05, |
| "loss": 13.1281, |
| "mean_token_accuracy": 0.7878676626831294, |
| "num_tokens": 18484931.0, |
| "step": 650 |
| }, |
| { |
| "entropy": 0.8254991695284843, |
| "epoch": 1.056, |
| "grad_norm": 49.0, |
| "learning_rate": 1.3180000000000001e-05, |
| "loss": 13.2747, |
| "mean_token_accuracy": 0.7866876818239689, |
| "num_tokens": 18773457.0, |
| "step": 660 |
| }, |
| { |
| "entropy": 0.8410865612328052, |
| "epoch": 1.072, |
| "grad_norm": 38.75, |
| "learning_rate": 1.3380000000000002e-05, |
| "loss": 13.3101, |
| "mean_token_accuracy": 0.7853217396885157, |
| "num_tokens": 19055365.0, |
| "step": 670 |
| }, |
| { |
| "entropy": 0.8255538143217563, |
| "epoch": 1.088, |
| "grad_norm": 28.875, |
| "learning_rate": 1.3580000000000002e-05, |
| "loss": 13.2174, |
| "mean_token_accuracy": 0.7872007485479117, |
| "num_tokens": 19345730.0, |
| "step": 680 |
| }, |
| { |
| "entropy": 0.823124579153955, |
| "epoch": 1.104, |
| "grad_norm": 29.375, |
| "learning_rate": 1.378e-05, |
| "loss": 13.1696, |
| "mean_token_accuracy": 0.7877223126590251, |
| "num_tokens": 19637390.0, |
| "step": 690 |
| }, |
| { |
| "entropy": 0.8028364922851324, |
| "epoch": 1.12, |
| "grad_norm": 34.5, |
| "learning_rate": 1.398e-05, |
| "loss": 12.7597, |
| "mean_token_accuracy": 0.792793495580554, |
| "num_tokens": 19923914.0, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.12, |
| "eval_biology_entropy": 1.3986766724586486, |
| "eval_biology_loss": 1.407199740409851, |
| "eval_biology_mean_token_accuracy": 0.6613863172531128, |
| "eval_biology_num_tokens": 19923914.0, |
| "eval_biology_runtime": 21.9946, |
| "eval_biology_samples_per_second": 22.733, |
| "eval_biology_steps_per_second": 5.683, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.12, |
| "eval_chemistry_entropy": 1.0769947800636293, |
| "eval_chemistry_loss": 1.0871174335479736, |
| "eval_chemistry_mean_token_accuracy": 0.7282235732078552, |
| "eval_chemistry_num_tokens": 19923914.0, |
| "eval_chemistry_runtime": 26.8846, |
| "eval_chemistry_samples_per_second": 18.598, |
| "eval_chemistry_steps_per_second": 4.65, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.12, |
| "eval_math_entropy": 0.8362808737754822, |
| "eval_math_loss": 0.8373622894287109, |
| "eval_math_mean_token_accuracy": 0.7839989976882935, |
| "eval_math_num_tokens": 19923914.0, |
| "eval_math_runtime": 27.4992, |
| "eval_math_samples_per_second": 18.182, |
| "eval_math_steps_per_second": 4.546, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.12, |
| "eval_cyber_entropy": 2.5681421542167664, |
| "eval_cyber_loss": 2.8721120357513428, |
| "eval_cyber_mean_token_accuracy": 0.4651792038977146, |
| "eval_cyber_num_tokens": 19923914.0, |
| "eval_cyber_runtime": 26.1144, |
| "eval_cyber_samples_per_second": 15.202, |
| "eval_cyber_steps_per_second": 3.829, |
| "step": 700 |
| }, |
| { |
| "entropy": 0.8095206459984183, |
| "epoch": 1.1360000000000001, |
| "grad_norm": 38.5, |
| "learning_rate": 1.418e-05, |
| "loss": 12.8823, |
| "mean_token_accuracy": 0.7904491990804672, |
| "num_tokens": 20201892.0, |
| "step": 710 |
| }, |
| { |
| "entropy": 0.8196006739512086, |
| "epoch": 1.152, |
| "grad_norm": 30.75, |
| "learning_rate": 1.4380000000000001e-05, |
| "loss": 13.0652, |
| "mean_token_accuracy": 0.7910903133451939, |
| "num_tokens": 20490282.0, |
| "step": 720 |
| }, |
| { |
| "entropy": 0.8046272564679384, |
| "epoch": 1.168, |
| "grad_norm": 29.5, |
| "learning_rate": 1.4580000000000001e-05, |
| "loss": 12.8751, |
| "mean_token_accuracy": 0.7915120176970959, |
| "num_tokens": 20785786.0, |
| "step": 730 |
| }, |
| { |
| "entropy": 0.8037027461454272, |
| "epoch": 1.184, |
| "grad_norm": 30.5, |
| "learning_rate": 1.478e-05, |
| "loss": 12.8554, |
| "mean_token_accuracy": 0.7912269696593285, |
| "num_tokens": 21074205.0, |
| "step": 740 |
| }, |
| { |
| "entropy": 0.79942841604352, |
| "epoch": 1.2, |
| "grad_norm": 30.75, |
| "learning_rate": 1.498e-05, |
| "loss": 12.7343, |
| "mean_token_accuracy": 0.7923291265964508, |
| "num_tokens": 21369159.0, |
| "step": 750 |
| }, |
| { |
| "entropy": 0.807464637234807, |
| "epoch": 1.216, |
| "grad_norm": 35.0, |
| "learning_rate": 1.5180000000000002e-05, |
| "loss": 12.8367, |
| "mean_token_accuracy": 0.7913754984736443, |
| "num_tokens": 21649178.0, |
| "step": 760 |
| }, |
| { |
| "entropy": 0.7876615423709155, |
| "epoch": 1.232, |
| "grad_norm": 30.25, |
| "learning_rate": 1.5380000000000002e-05, |
| "loss": 12.556, |
| "mean_token_accuracy": 0.7947213523089885, |
| "num_tokens": 21930239.0, |
| "step": 770 |
| }, |
| { |
| "entropy": 0.7889078231528401, |
| "epoch": 1.248, |
| "grad_norm": 29.75, |
| "learning_rate": 1.5580000000000003e-05, |
| "loss": 12.5585, |
| "mean_token_accuracy": 0.7938403252512216, |
| "num_tokens": 22216387.0, |
| "step": 780 |
| }, |
| { |
| "entropy": 0.8203166201710701, |
| "epoch": 1.264, |
| "grad_norm": 30.0, |
| "learning_rate": 1.578e-05, |
| "loss": 13.0401, |
| "mean_token_accuracy": 0.7889407943934202, |
| "num_tokens": 22501002.0, |
| "step": 790 |
| }, |
| { |
| "entropy": 0.7915117274969816, |
| "epoch": 1.28, |
| "grad_norm": 33.0, |
| "learning_rate": 1.5980000000000003e-05, |
| "loss": 12.6967, |
| "mean_token_accuracy": 0.7933882053941488, |
| "num_tokens": 22779682.0, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.28, |
| "eval_biology_entropy": 1.4146036610603332, |
| "eval_biology_loss": 1.413214921951294, |
| "eval_biology_mean_token_accuracy": 0.659837914943695, |
| "eval_biology_num_tokens": 22779682.0, |
| "eval_biology_runtime": 22.0253, |
| "eval_biology_samples_per_second": 22.701, |
| "eval_biology_steps_per_second": 5.675, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.28, |
| "eval_chemistry_entropy": 1.0714722080230712, |
| "eval_chemistry_loss": 1.0812031030654907, |
| "eval_chemistry_mean_token_accuracy": 0.7291303877830505, |
| "eval_chemistry_num_tokens": 22779682.0, |
| "eval_chemistry_runtime": 26.8892, |
| "eval_chemistry_samples_per_second": 18.595, |
| "eval_chemistry_steps_per_second": 4.649, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.28, |
| "eval_math_entropy": 0.8192751173973083, |
| "eval_math_loss": 0.8122938275337219, |
| "eval_math_mean_token_accuracy": 0.789606306552887, |
| "eval_math_num_tokens": 22779682.0, |
| "eval_math_runtime": 27.5274, |
| "eval_math_samples_per_second": 18.164, |
| "eval_math_steps_per_second": 4.541, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.28, |
| "eval_cyber_entropy": 2.5670096004009246, |
| "eval_cyber_loss": 2.8539652824401855, |
| "eval_cyber_mean_token_accuracy": 0.46533648878335954, |
| "eval_cyber_num_tokens": 22779682.0, |
| "eval_cyber_runtime": 26.1723, |
| "eval_cyber_samples_per_second": 15.169, |
| "eval_cyber_steps_per_second": 3.821, |
| "step": 800 |
| }, |
| { |
| "entropy": 0.786592660844326, |
| "epoch": 1.296, |
| "grad_norm": 35.5, |
| "learning_rate": 1.618e-05, |
| "loss": 12.5036, |
| "mean_token_accuracy": 0.7954838387668133, |
| "num_tokens": 23057744.0, |
| "step": 810 |
| }, |
| { |
| "entropy": 0.7977361943572759, |
| "epoch": 1.312, |
| "grad_norm": 34.5, |
| "learning_rate": 1.638e-05, |
| "loss": 12.7511, |
| "mean_token_accuracy": 0.7935790359973908, |
| "num_tokens": 23344644.0, |
| "step": 820 |
| }, |
| { |
| "entropy": 0.8038571482524276, |
| "epoch": 1.328, |
| "grad_norm": 38.75, |
| "learning_rate": 1.658e-05, |
| "loss": 12.847, |
| "mean_token_accuracy": 0.7909597154706717, |
| "num_tokens": 23622405.0, |
| "step": 830 |
| }, |
| { |
| "entropy": 0.781531005539, |
| "epoch": 1.3439999999999999, |
| "grad_norm": 28.75, |
| "learning_rate": 1.6780000000000002e-05, |
| "loss": 12.4551, |
| "mean_token_accuracy": 0.795590429380536, |
| "num_tokens": 23899771.0, |
| "step": 840 |
| }, |
| { |
| "entropy": 0.7783096175640821, |
| "epoch": 1.3599999999999999, |
| "grad_norm": 27.0, |
| "learning_rate": 1.698e-05, |
| "loss": 12.4462, |
| "mean_token_accuracy": 0.7967745348811149, |
| "num_tokens": 24187023.0, |
| "step": 850 |
| }, |
| { |
| "entropy": 0.8302321504801512, |
| "epoch": 1.376, |
| "grad_norm": 30.625, |
| "learning_rate": 1.718e-05, |
| "loss": 13.2594, |
| "mean_token_accuracy": 0.7850385505706072, |
| "num_tokens": 24466132.0, |
| "step": 860 |
| }, |
| { |
| "entropy": 0.7808034917339682, |
| "epoch": 1.392, |
| "grad_norm": 34.5, |
| "learning_rate": 1.7380000000000003e-05, |
| "loss": 12.4747, |
| "mean_token_accuracy": 0.7949298892170191, |
| "num_tokens": 24748043.0, |
| "step": 870 |
| }, |
| { |
| "entropy": 0.7715026669204235, |
| "epoch": 1.408, |
| "grad_norm": 36.0, |
| "learning_rate": 1.758e-05, |
| "loss": 12.3399, |
| "mean_token_accuracy": 0.7984749253839254, |
| "num_tokens": 25036674.0, |
| "step": 880 |
| }, |
| { |
| "entropy": 0.7645593881607056, |
| "epoch": 1.424, |
| "grad_norm": 27.25, |
| "learning_rate": 1.7780000000000003e-05, |
| "loss": 12.1973, |
| "mean_token_accuracy": 0.7993213057518005, |
| "num_tokens": 25324579.0, |
| "step": 890 |
| }, |
| { |
| "entropy": 0.7820997565984726, |
| "epoch": 1.44, |
| "grad_norm": 33.0, |
| "learning_rate": 1.798e-05, |
| "loss": 12.5051, |
| "mean_token_accuracy": 0.7951443370431661, |
| "num_tokens": 25606824.0, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.44, |
| "eval_biology_entropy": 1.3808940649032593, |
| "eval_biology_loss": 1.4220765829086304, |
| "eval_biology_mean_token_accuracy": 0.6588086094856263, |
| "eval_biology_num_tokens": 25606824.0, |
| "eval_biology_runtime": 22.0118, |
| "eval_biology_samples_per_second": 22.715, |
| "eval_biology_steps_per_second": 5.679, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.44, |
| "eval_chemistry_entropy": 1.0482762174606324, |
| "eval_chemistry_loss": 1.073889136314392, |
| "eval_chemistry_mean_token_accuracy": 0.7307762913703918, |
| "eval_chemistry_num_tokens": 25606824.0, |
| "eval_chemistry_runtime": 26.8657, |
| "eval_chemistry_samples_per_second": 18.611, |
| "eval_chemistry_steps_per_second": 4.653, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.44, |
| "eval_math_entropy": 0.7846709032058716, |
| "eval_math_loss": 0.7932249903678894, |
| "eval_math_mean_token_accuracy": 0.792679114818573, |
| "eval_math_num_tokens": 25606824.0, |
| "eval_math_runtime": 27.514, |
| "eval_math_samples_per_second": 18.173, |
| "eval_math_steps_per_second": 4.543, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.44, |
| "eval_cyber_entropy": 2.5483840811252594, |
| "eval_cyber_loss": 2.8718831539154053, |
| "eval_cyber_mean_token_accuracy": 0.4638554835319519, |
| "eval_cyber_num_tokens": 25606824.0, |
| "eval_cyber_runtime": 26.216, |
| "eval_cyber_samples_per_second": 15.143, |
| "eval_cyber_steps_per_second": 3.814, |
| "step": 900 |
| }, |
| { |
| "entropy": 0.7723285494372248, |
| "epoch": 1.456, |
| "grad_norm": 28.625, |
| "learning_rate": 1.8180000000000002e-05, |
| "loss": 12.303, |
| "mean_token_accuracy": 0.7964108034968376, |
| "num_tokens": 25886396.0, |
| "step": 910 |
| }, |
| { |
| "entropy": 0.7762986140325665, |
| "epoch": 1.472, |
| "grad_norm": 28.875, |
| "learning_rate": 1.8380000000000004e-05, |
| "loss": 12.4134, |
| "mean_token_accuracy": 0.7955747056752444, |
| "num_tokens": 26163618.0, |
| "step": 920 |
| }, |
| { |
| "entropy": 0.7938198037445545, |
| "epoch": 1.488, |
| "grad_norm": 29.625, |
| "learning_rate": 1.858e-05, |
| "loss": 12.75, |
| "mean_token_accuracy": 0.7917455974966288, |
| "num_tokens": 26438338.0, |
| "step": 930 |
| }, |
| { |
| "entropy": 0.7594615155830979, |
| "epoch": 1.504, |
| "grad_norm": 34.0, |
| "learning_rate": 1.878e-05, |
| "loss": 12.14, |
| "mean_token_accuracy": 0.8009778898209333, |
| "num_tokens": 26729255.0, |
| "step": 940 |
| }, |
| { |
| "entropy": 0.7861603863537312, |
| "epoch": 1.52, |
| "grad_norm": 27.125, |
| "learning_rate": 1.898e-05, |
| "loss": 12.4626, |
| "mean_token_accuracy": 0.7956234533339739, |
| "num_tokens": 27017935.0, |
| "step": 950 |
| }, |
| { |
| "entropy": 0.7631909586489201, |
| "epoch": 1.536, |
| "grad_norm": 24.625, |
| "learning_rate": 1.918e-05, |
| "loss": 12.1955, |
| "mean_token_accuracy": 0.7989141892641782, |
| "num_tokens": 27306339.0, |
| "step": 960 |
| }, |
| { |
| "entropy": 0.7708934009075165, |
| "epoch": 1.552, |
| "grad_norm": 27.75, |
| "learning_rate": 1.938e-05, |
| "loss": 12.1963, |
| "mean_token_accuracy": 0.7984277427196502, |
| "num_tokens": 27591959.0, |
| "step": 970 |
| }, |
| { |
| "entropy": 0.7459486592561007, |
| "epoch": 1.568, |
| "grad_norm": 28.625, |
| "learning_rate": 1.9580000000000002e-05, |
| "loss": 11.9228, |
| "mean_token_accuracy": 0.8039638720452785, |
| "num_tokens": 27884398.0, |
| "step": 980 |
| }, |
| { |
| "entropy": 0.7573445823043585, |
| "epoch": 1.584, |
| "grad_norm": 27.0, |
| "learning_rate": 1.978e-05, |
| "loss": 12.0883, |
| "mean_token_accuracy": 0.8011246718466282, |
| "num_tokens": 28171274.0, |
| "step": 990 |
| }, |
| { |
| "entropy": 0.7612122105434537, |
| "epoch": 1.6, |
| "grad_norm": 27.5, |
| "learning_rate": 1.9980000000000002e-05, |
| "loss": 12.0981, |
| "mean_token_accuracy": 0.7986274570226669, |
| "num_tokens": 28457624.0, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.6, |
| "eval_biology_entropy": 1.3829385170936583, |
| "eval_biology_loss": 1.4260554313659668, |
| "eval_biology_mean_token_accuracy": 0.6589009766578674, |
| "eval_biology_num_tokens": 28457624.0, |
| "eval_biology_runtime": 22.024, |
| "eval_biology_samples_per_second": 22.703, |
| "eval_biology_steps_per_second": 5.676, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.6, |
| "eval_chemistry_entropy": 1.0377137541770936, |
| "eval_chemistry_loss": 1.0700007677078247, |
| "eval_chemistry_mean_token_accuracy": 0.7318583874702453, |
| "eval_chemistry_num_tokens": 28457624.0, |
| "eval_chemistry_runtime": 26.9114, |
| "eval_chemistry_samples_per_second": 18.579, |
| "eval_chemistry_steps_per_second": 4.645, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.6, |
| "eval_math_entropy": 0.775481684923172, |
| "eval_math_loss": 0.7736496329307556, |
| "eval_math_mean_token_accuracy": 0.7959285154342651, |
| "eval_math_num_tokens": 28457624.0, |
| "eval_math_runtime": 27.53, |
| "eval_math_samples_per_second": 18.162, |
| "eval_math_steps_per_second": 4.541, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.6, |
| "eval_cyber_entropy": 2.5157601726055145, |
| "eval_cyber_loss": 2.8350298404693604, |
| "eval_cyber_mean_token_accuracy": 0.4679137858748436, |
| "eval_cyber_num_tokens": 28457624.0, |
| "eval_cyber_runtime": 26.1345, |
| "eval_cyber_samples_per_second": 15.191, |
| "eval_cyber_steps_per_second": 3.826, |
| "step": 1000 |
| }, |
| { |
| "entropy": 0.7650468161329627, |
| "epoch": 1.616, |
| "grad_norm": 24.875, |
| "learning_rate": 1.9980000000000002e-05, |
| "loss": 12.1883, |
| "mean_token_accuracy": 0.798528803884983, |
| "num_tokens": 28743099.0, |
| "step": 1010 |
| }, |
| { |
| "entropy": 0.7713425377383828, |
| "epoch": 1.6320000000000001, |
| "grad_norm": 25.75, |
| "learning_rate": 1.995777777777778e-05, |
| "loss": 12.2948, |
| "mean_token_accuracy": 0.7963836405426263, |
| "num_tokens": 29017297.0, |
| "step": 1020 |
| }, |
| { |
| "entropy": 0.7499153949320316, |
| "epoch": 1.6480000000000001, |
| "grad_norm": 25.75, |
| "learning_rate": 1.9935555555555557e-05, |
| "loss": 11.9856, |
| "mean_token_accuracy": 0.803160610422492, |
| "num_tokens": 29303707.0, |
| "step": 1030 |
| }, |
| { |
| "entropy": 0.7566261947154999, |
| "epoch": 1.6640000000000001, |
| "grad_norm": 25.0, |
| "learning_rate": 1.9913333333333335e-05, |
| "loss": 12.034, |
| "mean_token_accuracy": 0.7999875675886869, |
| "num_tokens": 29597156.0, |
| "step": 1040 |
| }, |
| { |
| "entropy": 0.7669804213568568, |
| "epoch": 1.6800000000000002, |
| "grad_norm": 24.25, |
| "learning_rate": 1.9891111111111112e-05, |
| "loss": 12.2025, |
| "mean_token_accuracy": 0.7990686308592558, |
| "num_tokens": 29883879.0, |
| "step": 1050 |
| }, |
| { |
| "entropy": 0.7553620956838131, |
| "epoch": 1.696, |
| "grad_norm": 28.125, |
| "learning_rate": 1.986888888888889e-05, |
| "loss": 12.1827, |
| "mean_token_accuracy": 0.8000101692974567, |
| "num_tokens": 30165760.0, |
| "step": 1060 |
| }, |
| { |
| "entropy": 0.7463795414194465, |
| "epoch": 1.712, |
| "grad_norm": 22.375, |
| "learning_rate": 1.9846666666666668e-05, |
| "loss": 11.9561, |
| "mean_token_accuracy": 0.8028988271951676, |
| "num_tokens": 30460367.0, |
| "step": 1070 |
| }, |
| { |
| "entropy": 0.7401833109557628, |
| "epoch": 1.728, |
| "grad_norm": 27.375, |
| "learning_rate": 1.9824444444444445e-05, |
| "loss": 11.7133, |
| "mean_token_accuracy": 0.8052776392549277, |
| "num_tokens": 30739137.0, |
| "step": 1080 |
| }, |
| { |
| "entropy": 0.7436290748417378, |
| "epoch": 1.744, |
| "grad_norm": 27.375, |
| "learning_rate": 1.9802222222222226e-05, |
| "loss": 11.8806, |
| "mean_token_accuracy": 0.8036583166569471, |
| "num_tokens": 31022663.0, |
| "step": 1090 |
| }, |
| { |
| "entropy": 0.7478637570515275, |
| "epoch": 1.76, |
| "grad_norm": 25.25, |
| "learning_rate": 1.978e-05, |
| "loss": 11.9202, |
| "mean_token_accuracy": 0.8017146904021502, |
| "num_tokens": 31306494.0, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.76, |
| "eval_biology_entropy": 1.3845259475708007, |
| "eval_biology_loss": 1.4283864498138428, |
| "eval_biology_mean_token_accuracy": 0.657891107082367, |
| "eval_biology_num_tokens": 31306494.0, |
| "eval_biology_runtime": 21.9927, |
| "eval_biology_samples_per_second": 22.735, |
| "eval_biology_steps_per_second": 5.684, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.76, |
| "eval_chemistry_entropy": 1.0231492972373963, |
| "eval_chemistry_loss": 1.063183307647705, |
| "eval_chemistry_mean_token_accuracy": 0.7330445971488952, |
| "eval_chemistry_num_tokens": 31306494.0, |
| "eval_chemistry_runtime": 26.8519, |
| "eval_chemistry_samples_per_second": 18.621, |
| "eval_chemistry_steps_per_second": 4.655, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.76, |
| "eval_math_entropy": 0.7524698441028594, |
| "eval_math_loss": 0.7613377571105957, |
| "eval_math_mean_token_accuracy": 0.798446418762207, |
| "eval_math_num_tokens": 31306494.0, |
| "eval_math_runtime": 27.5284, |
| "eval_math_samples_per_second": 18.163, |
| "eval_math_steps_per_second": 4.541, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.76, |
| "eval_cyber_entropy": 2.3844272685050965, |
| "eval_cyber_loss": 2.8584094047546387, |
| "eval_cyber_mean_token_accuracy": 0.47087193533778193, |
| "eval_cyber_num_tokens": 31306494.0, |
| "eval_cyber_runtime": 26.2154, |
| "eval_cyber_samples_per_second": 15.144, |
| "eval_cyber_steps_per_second": 3.815, |
| "step": 1100 |
| }, |
| { |
| "entropy": 0.7506109833717346, |
| "epoch": 1.776, |
| "grad_norm": 22.875, |
| "learning_rate": 1.975777777777778e-05, |
| "loss": 11.957, |
| "mean_token_accuracy": 0.803263409435749, |
| "num_tokens": 31595542.0, |
| "step": 1110 |
| }, |
| { |
| "entropy": 0.7545284632593393, |
| "epoch": 1.792, |
| "grad_norm": 25.25, |
| "learning_rate": 1.9735555555555556e-05, |
| "loss": 12.055, |
| "mean_token_accuracy": 0.8008169520646333, |
| "num_tokens": 31881189.0, |
| "step": 1120 |
| }, |
| { |
| "entropy": 0.7454792723059654, |
| "epoch": 1.808, |
| "grad_norm": 22.625, |
| "learning_rate": 1.9713333333333337e-05, |
| "loss": 11.8818, |
| "mean_token_accuracy": 0.8028201397508383, |
| "num_tokens": 32164196.0, |
| "step": 1130 |
| }, |
| { |
| "entropy": 0.7103133289143443, |
| "epoch": 1.8239999999999998, |
| "grad_norm": 23.625, |
| "learning_rate": 1.969111111111111e-05, |
| "loss": 11.3018, |
| "mean_token_accuracy": 0.8097421944141387, |
| "num_tokens": 32441530.0, |
| "step": 1140 |
| }, |
| { |
| "entropy": 0.7296694969758392, |
| "epoch": 1.8399999999999999, |
| "grad_norm": 23.125, |
| "learning_rate": 1.9668888888888892e-05, |
| "loss": 11.722, |
| "mean_token_accuracy": 0.8063848353922367, |
| "num_tokens": 32723145.0, |
| "step": 1150 |
| }, |
| { |
| "entropy": 0.746064018085599, |
| "epoch": 1.8559999999999999, |
| "grad_norm": 26.375, |
| "learning_rate": 1.9646666666666666e-05, |
| "loss": 11.9198, |
| "mean_token_accuracy": 0.8034628454595805, |
| "num_tokens": 33011263.0, |
| "step": 1160 |
| }, |
| { |
| "entropy": 0.7246854526922106, |
| "epoch": 1.8719999999999999, |
| "grad_norm": 25.125, |
| "learning_rate": 1.9624444444444447e-05, |
| "loss": 11.6702, |
| "mean_token_accuracy": 0.8065225839614868, |
| "num_tokens": 33298921.0, |
| "step": 1170 |
| }, |
| { |
| "entropy": 0.7422073289752007, |
| "epoch": 1.888, |
| "grad_norm": 33.0, |
| "learning_rate": 1.9602222222222225e-05, |
| "loss": 11.8223, |
| "mean_token_accuracy": 0.8029078282415867, |
| "num_tokens": 33576243.0, |
| "step": 1180 |
| }, |
| { |
| "entropy": 0.7377389714121818, |
| "epoch": 1.904, |
| "grad_norm": 21.875, |
| "learning_rate": 1.9580000000000002e-05, |
| "loss": 11.7626, |
| "mean_token_accuracy": 0.803905576467514, |
| "num_tokens": 33850968.0, |
| "step": 1190 |
| }, |
| { |
| "entropy": 0.720432554371655, |
| "epoch": 1.92, |
| "grad_norm": 24.125, |
| "learning_rate": 1.955777777777778e-05, |
| "loss": 11.4648, |
| "mean_token_accuracy": 0.8074128460139036, |
| "num_tokens": 34128558.0, |
| "step": 1200 |
| }, |
| { |
| "epoch": 1.92, |
| "eval_biology_entropy": 1.426067009449005, |
| "eval_biology_loss": 1.4298174381256104, |
| "eval_biology_mean_token_accuracy": 0.6572316522598267, |
| "eval_biology_num_tokens": 34128558.0, |
| "eval_biology_runtime": 21.9751, |
| "eval_biology_samples_per_second": 22.753, |
| "eval_biology_steps_per_second": 5.688, |
| "step": 1200 |
| }, |
| { |
| "epoch": 1.92, |
| "eval_chemistry_entropy": 1.052424753189087, |
| "eval_chemistry_loss": 1.058487057685852, |
| "eval_chemistry_mean_token_accuracy": 0.7344707479476928, |
| "eval_chemistry_num_tokens": 34128558.0, |
| "eval_chemistry_runtime": 26.8451, |
| "eval_chemistry_samples_per_second": 18.625, |
| "eval_chemistry_steps_per_second": 4.656, |
| "step": 1200 |
| }, |
| { |
| "epoch": 1.92, |
| "eval_math_entropy": 0.7641568143367767, |
| "eval_math_loss": 0.7467027306556702, |
| "eval_math_mean_token_accuracy": 0.8014610476493835, |
| "eval_math_num_tokens": 34128558.0, |
| "eval_math_runtime": 27.5216, |
| "eval_math_samples_per_second": 18.168, |
| "eval_math_steps_per_second": 4.542, |
| "step": 1200 |
| }, |
| { |
| "epoch": 1.92, |
| "eval_cyber_entropy": 2.4904652881622313, |
| "eval_cyber_loss": 2.6954379081726074, |
| "eval_cyber_mean_token_accuracy": 0.4830169627070427, |
| "eval_cyber_num_tokens": 34128558.0, |
| "eval_cyber_runtime": 26.1923, |
| "eval_cyber_samples_per_second": 15.157, |
| "eval_cyber_steps_per_second": 3.818, |
| "step": 1200 |
| }, |
| { |
| "entropy": 0.731533533334732, |
| "epoch": 1.936, |
| "grad_norm": 25.125, |
| "learning_rate": 1.9535555555555557e-05, |
| "loss": 11.6804, |
| "mean_token_accuracy": 0.8052534744143486, |
| "num_tokens": 34408056.0, |
| "step": 1210 |
| }, |
| { |
| "entropy": 0.7303263584151864, |
| "epoch": 1.952, |
| "grad_norm": 23.375, |
| "learning_rate": 1.9513333333333335e-05, |
| "loss": 11.5676, |
| "mean_token_accuracy": 0.8064503286033868, |
| "num_tokens": 34684679.0, |
| "step": 1220 |
| }, |
| { |
| "entropy": 0.7569911142811179, |
| "epoch": 1.968, |
| "grad_norm": 25.375, |
| "learning_rate": 1.9491111111111113e-05, |
| "loss": 12.1009, |
| "mean_token_accuracy": 0.8005167040973902, |
| "num_tokens": 34971038.0, |
| "step": 1230 |
| }, |
| { |
| "entropy": 0.7218442076817155, |
| "epoch": 1.984, |
| "grad_norm": 23.375, |
| "learning_rate": 1.946888888888889e-05, |
| "loss": 11.5014, |
| "mean_token_accuracy": 0.808349072188139, |
| "num_tokens": 35262281.0, |
| "step": 1240 |
| }, |
| { |
| "entropy": 0.7173755820840597, |
| "epoch": 2.0, |
| "grad_norm": 24.125, |
| "learning_rate": 1.9446666666666668e-05, |
| "loss": 11.4742, |
| "mean_token_accuracy": 0.8086670659482479, |
| "num_tokens": 35560864.0, |
| "step": 1250 |
| }, |
| { |
| "entropy": 0.6936481088399887, |
| "epoch": 2.016, |
| "grad_norm": 23.125, |
| "learning_rate": 1.9424444444444446e-05, |
| "loss": 10.8847, |
| "mean_token_accuracy": 0.8153551481664181, |
| "num_tokens": 35846704.0, |
| "step": 1260 |
| }, |
| { |
| "entropy": 0.6694988587871193, |
| "epoch": 2.032, |
| "grad_norm": 22.5, |
| "learning_rate": 1.9402222222222223e-05, |
| "loss": 10.711, |
| "mean_token_accuracy": 0.816493459790945, |
| "num_tokens": 36128775.0, |
| "step": 1270 |
| }, |
| { |
| "entropy": 0.6576118635013699, |
| "epoch": 2.048, |
| "grad_norm": 23.25, |
| "learning_rate": 1.938e-05, |
| "loss": 10.4997, |
| "mean_token_accuracy": 0.8206516925245524, |
| "num_tokens": 36419504.0, |
| "step": 1280 |
| }, |
| { |
| "entropy": 0.6648308178409934, |
| "epoch": 2.064, |
| "grad_norm": 22.375, |
| "learning_rate": 1.935777777777778e-05, |
| "loss": 10.5449, |
| "mean_token_accuracy": 0.8189927719533443, |
| "num_tokens": 36706816.0, |
| "step": 1290 |
| }, |
| { |
| "entropy": 0.6633218213915825, |
| "epoch": 2.08, |
| "grad_norm": 24.125, |
| "learning_rate": 1.9335555555555556e-05, |
| "loss": 10.5216, |
| "mean_token_accuracy": 0.8182109944522381, |
| "num_tokens": 36988475.0, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.08, |
| "eval_biology_entropy": 1.240901198387146, |
| "eval_biology_loss": 1.4575155973434448, |
| "eval_biology_mean_token_accuracy": 0.6538263387680053, |
| "eval_biology_num_tokens": 36988475.0, |
| "eval_biology_runtime": 22.0164, |
| "eval_biology_samples_per_second": 22.71, |
| "eval_biology_steps_per_second": 5.678, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.08, |
| "eval_chemistry_entropy": 0.9047480673789978, |
| "eval_chemistry_loss": 1.077072262763977, |
| "eval_chemistry_mean_token_accuracy": 0.7334306511878967, |
| "eval_chemistry_num_tokens": 36988475.0, |
| "eval_chemistry_runtime": 26.8608, |
| "eval_chemistry_samples_per_second": 18.614, |
| "eval_chemistry_steps_per_second": 4.654, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.08, |
| "eval_math_entropy": 0.6772888927459717, |
| "eval_math_loss": 0.7423775792121887, |
| "eval_math_mean_token_accuracy": 0.8029003148078918, |
| "eval_math_num_tokens": 36988475.0, |
| "eval_math_runtime": 27.5039, |
| "eval_math_samples_per_second": 18.179, |
| "eval_math_steps_per_second": 4.545, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.08, |
| "eval_cyber_entropy": 2.240402947664261, |
| "eval_cyber_loss": 2.8395872116088867, |
| "eval_cyber_mean_token_accuracy": 0.4778119161725044, |
| "eval_cyber_num_tokens": 36988475.0, |
| "eval_cyber_runtime": 26.1457, |
| "eval_cyber_samples_per_second": 15.184, |
| "eval_cyber_steps_per_second": 3.825, |
| "step": 1300 |
| }, |
| { |
| "entropy": 0.6708741160109639, |
| "epoch": 2.096, |
| "grad_norm": 22.375, |
| "learning_rate": 1.9313333333333334e-05, |
| "loss": 10.6881, |
| "mean_token_accuracy": 0.8182591505348682, |
| "num_tokens": 37270131.0, |
| "step": 1310 |
| }, |
| { |
| "entropy": 0.6532387970015406, |
| "epoch": 2.112, |
| "grad_norm": 24.0, |
| "learning_rate": 1.9291111111111115e-05, |
| "loss": 10.4792, |
| "mean_token_accuracy": 0.8189583510160446, |
| "num_tokens": 37563537.0, |
| "step": 1320 |
| }, |
| { |
| "entropy": 0.6555240735411644, |
| "epoch": 2.128, |
| "grad_norm": 22.375, |
| "learning_rate": 1.926888888888889e-05, |
| "loss": 10.369, |
| "mean_token_accuracy": 0.8218111298978329, |
| "num_tokens": 37843959.0, |
| "step": 1330 |
| }, |
| { |
| "entropy": 0.6665401035919786, |
| "epoch": 2.144, |
| "grad_norm": 24.25, |
| "learning_rate": 1.924666666666667e-05, |
| "loss": 10.5665, |
| "mean_token_accuracy": 0.8194981347769499, |
| "num_tokens": 38133092.0, |
| "step": 1340 |
| }, |
| { |
| "entropy": 0.6592796456068755, |
| "epoch": 2.16, |
| "grad_norm": 21.5, |
| "learning_rate": 1.9224444444444444e-05, |
| "loss": 10.5062, |
| "mean_token_accuracy": 0.8200013760477305, |
| "num_tokens": 38421229.0, |
| "step": 1350 |
| }, |
| { |
| "entropy": 0.639416103810072, |
| "epoch": 2.176, |
| "grad_norm": 23.0, |
| "learning_rate": 1.9202222222222225e-05, |
| "loss": 10.1779, |
| "mean_token_accuracy": 0.8243647638708353, |
| "num_tokens": 38708043.0, |
| "step": 1360 |
| }, |
| { |
| "entropy": 0.667258214391768, |
| "epoch": 2.192, |
| "grad_norm": 23.375, |
| "learning_rate": 1.918e-05, |
| "loss": 10.6186, |
| "mean_token_accuracy": 0.8171255987137556, |
| "num_tokens": 38996930.0, |
| "step": 1370 |
| }, |
| { |
| "entropy": 0.6627653013914824, |
| "epoch": 2.208, |
| "grad_norm": 23.75, |
| "learning_rate": 1.915777777777778e-05, |
| "loss": 10.6227, |
| "mean_token_accuracy": 0.818722078576684, |
| "num_tokens": 39279481.0, |
| "step": 1380 |
| }, |
| { |
| "entropy": 0.6583162900060415, |
| "epoch": 2.224, |
| "grad_norm": 24.0, |
| "learning_rate": 1.9135555555555555e-05, |
| "loss": 10.4429, |
| "mean_token_accuracy": 0.8204963516443968, |
| "num_tokens": 39569030.0, |
| "step": 1390 |
| }, |
| { |
| "entropy": 0.6620556140318513, |
| "epoch": 2.24, |
| "grad_norm": 23.625, |
| "learning_rate": 1.9113333333333336e-05, |
| "loss": 10.591, |
| "mean_token_accuracy": 0.8184640970081091, |
| "num_tokens": 39854873.0, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.24, |
| "eval_biology_entropy": 1.1904019894599915, |
| "eval_biology_loss": 1.473069667816162, |
| "eval_biology_mean_token_accuracy": 0.6529810581207275, |
| "eval_biology_num_tokens": 39854873.0, |
| "eval_biology_runtime": 22.0119, |
| "eval_biology_samples_per_second": 22.715, |
| "eval_biology_steps_per_second": 5.679, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.24, |
| "eval_chemistry_entropy": 0.8868081021308899, |
| "eval_chemistry_loss": 1.0847948789596558, |
| "eval_chemistry_mean_token_accuracy": 0.7322362198829651, |
| "eval_chemistry_num_tokens": 39854873.0, |
| "eval_chemistry_runtime": 27.1533, |
| "eval_chemistry_samples_per_second": 18.414, |
| "eval_chemistry_steps_per_second": 4.603, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.24, |
| "eval_math_entropy": 0.6937474160194397, |
| "eval_math_loss": 0.7352772951126099, |
| "eval_math_mean_token_accuracy": 0.8037003560066223, |
| "eval_math_num_tokens": 39854873.0, |
| "eval_math_runtime": 27.5567, |
| "eval_math_samples_per_second": 18.144, |
| "eval_math_steps_per_second": 4.536, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.24, |
| "eval_cyber_entropy": 2.280633035302162, |
| "eval_cyber_loss": 2.848487615585327, |
| "eval_cyber_mean_token_accuracy": 0.46778192803263663, |
| "eval_cyber_num_tokens": 39854873.0, |
| "eval_cyber_runtime": 26.2399, |
| "eval_cyber_samples_per_second": 15.13, |
| "eval_cyber_steps_per_second": 3.811, |
| "step": 1400 |
| }, |
| { |
| "entropy": 0.6573172532021999, |
| "epoch": 2.2560000000000002, |
| "grad_norm": 22.75, |
| "learning_rate": 1.9091111111111113e-05, |
| "loss": 10.4628, |
| "mean_token_accuracy": 0.8198426373302936, |
| "num_tokens": 40141190.0, |
| "step": 1410 |
| }, |
| { |
| "entropy": 0.6748596677556634, |
| "epoch": 2.2720000000000002, |
| "grad_norm": 26.25, |
| "learning_rate": 1.906888888888889e-05, |
| "loss": 10.7759, |
| "mean_token_accuracy": 0.8157621681690216, |
| "num_tokens": 40415203.0, |
| "step": 1420 |
| }, |
| { |
| "entropy": 0.6660139387473464, |
| "epoch": 2.288, |
| "grad_norm": 25.0, |
| "learning_rate": 1.904666666666667e-05, |
| "loss": 10.5722, |
| "mean_token_accuracy": 0.8170963436365127, |
| "num_tokens": 40702393.0, |
| "step": 1430 |
| }, |
| { |
| "entropy": 0.6447202865034342, |
| "epoch": 2.304, |
| "grad_norm": 24.875, |
| "learning_rate": 1.9024444444444446e-05, |
| "loss": 10.2772, |
| "mean_token_accuracy": 0.8228711977601051, |
| "num_tokens": 40982775.0, |
| "step": 1440 |
| }, |
| { |
| "entropy": 0.667175211571157, |
| "epoch": 2.32, |
| "grad_norm": 25.375, |
| "learning_rate": 1.9002222222222224e-05, |
| "loss": 10.6322, |
| "mean_token_accuracy": 0.817449289560318, |
| "num_tokens": 41263356.0, |
| "step": 1450 |
| }, |
| { |
| "entropy": 0.6582919212058187, |
| "epoch": 2.336, |
| "grad_norm": 24.625, |
| "learning_rate": 1.898e-05, |
| "loss": 10.5061, |
| "mean_token_accuracy": 0.8195808235555887, |
| "num_tokens": 41545235.0, |
| "step": 1460 |
| }, |
| { |
| "entropy": 0.683755399286747, |
| "epoch": 2.352, |
| "grad_norm": 24.25, |
| "learning_rate": 1.895777777777778e-05, |
| "loss": 10.8267, |
| "mean_token_accuracy": 0.8143463153392076, |
| "num_tokens": 41833417.0, |
| "step": 1470 |
| }, |
| { |
| "entropy": 0.6577698297798633, |
| "epoch": 2.368, |
| "grad_norm": 24.625, |
| "learning_rate": 1.8935555555555556e-05, |
| "loss": 10.5268, |
| "mean_token_accuracy": 0.8191198598593473, |
| "num_tokens": 42117030.0, |
| "step": 1480 |
| }, |
| { |
| "entropy": 0.6793028621003032, |
| "epoch": 2.384, |
| "grad_norm": 28.25, |
| "learning_rate": 1.8913333333333334e-05, |
| "loss": 10.7829, |
| "mean_token_accuracy": 0.8163190931081772, |
| "num_tokens": 42410990.0, |
| "step": 1490 |
| }, |
| { |
| "entropy": 0.6641744881868362, |
| "epoch": 2.4, |
| "grad_norm": 24.5, |
| "learning_rate": 1.8891111111111115e-05, |
| "loss": 10.5965, |
| "mean_token_accuracy": 0.8189876776188612, |
| "num_tokens": 42691890.0, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.4, |
| "eval_biology_entropy": 1.2293707489967347, |
| "eval_biology_loss": 1.4714155197143555, |
| "eval_biology_mean_token_accuracy": 0.6514331855773926, |
| "eval_biology_num_tokens": 42691890.0, |
| "eval_biology_runtime": 21.9938, |
| "eval_biology_samples_per_second": 22.734, |
| "eval_biology_steps_per_second": 5.683, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.4, |
| "eval_chemistry_entropy": 0.9099567327499389, |
| "eval_chemistry_loss": 1.0838559865951538, |
| "eval_chemistry_mean_token_accuracy": 0.7319689979553222, |
| "eval_chemistry_num_tokens": 42691890.0, |
| "eval_chemistry_runtime": 26.8774, |
| "eval_chemistry_samples_per_second": 18.603, |
| "eval_chemistry_steps_per_second": 4.651, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.4, |
| "eval_math_entropy": 0.6841141791343689, |
| "eval_math_loss": 0.726224958896637, |
| "eval_math_mean_token_accuracy": 0.8050775575637817, |
| "eval_math_num_tokens": 42691890.0, |
| "eval_math_runtime": 27.5383, |
| "eval_math_samples_per_second": 18.157, |
| "eval_math_steps_per_second": 4.539, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.4, |
| "eval_cyber_entropy": 2.1712605100870133, |
| "eval_cyber_loss": 2.742515802383423, |
| "eval_cyber_mean_token_accuracy": 0.48678219854831695, |
| "eval_cyber_num_tokens": 42691890.0, |
| "eval_cyber_runtime": 26.1725, |
| "eval_cyber_samples_per_second": 15.169, |
| "eval_cyber_steps_per_second": 3.821, |
| "step": 1500 |
| }, |
| { |
| "entropy": 0.681360544078052, |
| "epoch": 2.416, |
| "grad_norm": 23.875, |
| "learning_rate": 1.886888888888889e-05, |
| "loss": 10.8018, |
| "mean_token_accuracy": 0.8153177864849568, |
| "num_tokens": 42971588.0, |
| "step": 1510 |
| }, |
| { |
| "entropy": 0.6504227627068758, |
| "epoch": 2.432, |
| "grad_norm": 23.625, |
| "learning_rate": 1.884666666666667e-05, |
| "loss": 10.4029, |
| "mean_token_accuracy": 0.8209474917501212, |
| "num_tokens": 43253821.0, |
| "step": 1520 |
| }, |
| { |
| "entropy": 0.6501767633482813, |
| "epoch": 2.448, |
| "grad_norm": 24.375, |
| "learning_rate": 1.8824444444444445e-05, |
| "loss": 10.3415, |
| "mean_token_accuracy": 0.8225077040493488, |
| "num_tokens": 43550902.0, |
| "step": 1530 |
| }, |
| { |
| "entropy": 0.6623956672847271, |
| "epoch": 2.464, |
| "grad_norm": 24.125, |
| "learning_rate": 1.8802222222222226e-05, |
| "loss": 10.5246, |
| "mean_token_accuracy": 0.8181775715202093, |
| "num_tokens": 43844259.0, |
| "step": 1540 |
| }, |
| { |
| "entropy": 0.6859173832461238, |
| "epoch": 2.48, |
| "grad_norm": 23.125, |
| "learning_rate": 1.878e-05, |
| "loss": 10.9239, |
| "mean_token_accuracy": 0.8137104224413634, |
| "num_tokens": 44115701.0, |
| "step": 1550 |
| }, |
| { |
| "entropy": 0.6711945479735733, |
| "epoch": 2.496, |
| "grad_norm": 21.875, |
| "learning_rate": 1.875777777777778e-05, |
| "loss": 10.6685, |
| "mean_token_accuracy": 0.8173221621662379, |
| "num_tokens": 44405520.0, |
| "step": 1560 |
| }, |
| { |
| "entropy": 0.6624481493607164, |
| "epoch": 2.512, |
| "grad_norm": 23.375, |
| "learning_rate": 1.873555555555556e-05, |
| "loss": 10.4945, |
| "mean_token_accuracy": 0.8196087624877691, |
| "num_tokens": 44686477.0, |
| "step": 1570 |
| }, |
| { |
| "entropy": 0.6599110793322325, |
| "epoch": 2.528, |
| "grad_norm": 22.75, |
| "learning_rate": 1.8713333333333336e-05, |
| "loss": 10.4873, |
| "mean_token_accuracy": 0.8193403802812099, |
| "num_tokens": 44969760.0, |
| "step": 1580 |
| }, |
| { |
| "entropy": 0.6503490032628179, |
| "epoch": 2.544, |
| "grad_norm": 22.375, |
| "learning_rate": 1.8691111111111114e-05, |
| "loss": 10.4441, |
| "mean_token_accuracy": 0.8206405211240053, |
| "num_tokens": 45255326.0, |
| "step": 1590 |
| }, |
| { |
| "entropy": 0.6536685146391392, |
| "epoch": 2.56, |
| "grad_norm": 22.875, |
| "learning_rate": 1.866888888888889e-05, |
| "loss": 10.4091, |
| "mean_token_accuracy": 0.8202110458165407, |
| "num_tokens": 45532525.0, |
| "step": 1600 |
| }, |
| { |
| "epoch": 2.56, |
| "eval_biology_entropy": 1.2134844155311584, |
| "eval_biology_loss": 1.478155255317688, |
| "eval_biology_mean_token_accuracy": 0.6511134657859802, |
| "eval_biology_num_tokens": 45532525.0, |
| "eval_biology_runtime": 21.9488, |
| "eval_biology_samples_per_second": 22.78, |
| "eval_biology_steps_per_second": 5.695, |
| "step": 1600 |
| }, |
| { |
| "epoch": 2.56, |
| "eval_chemistry_entropy": 0.8961781821250916, |
| "eval_chemistry_loss": 1.0829113721847534, |
| "eval_chemistry_mean_token_accuracy": 0.7327736926078796, |
| "eval_chemistry_num_tokens": 45532525.0, |
| "eval_chemistry_runtime": 26.8437, |
| "eval_chemistry_samples_per_second": 18.626, |
| "eval_chemistry_steps_per_second": 4.657, |
| "step": 1600 |
| }, |
| { |
| "epoch": 2.56, |
| "eval_math_entropy": 0.6742748956680298, |
| "eval_math_loss": 0.7170487642288208, |
| "eval_math_mean_token_accuracy": 0.8080672206878662, |
| "eval_math_num_tokens": 45532525.0, |
| "eval_math_runtime": 27.5073, |
| "eval_math_samples_per_second": 18.177, |
| "eval_math_steps_per_second": 4.544, |
| "step": 1600 |
| }, |
| { |
| "epoch": 2.56, |
| "eval_cyber_entropy": 2.198235506415367, |
| "eval_cyber_loss": 2.8126513957977295, |
| "eval_cyber_mean_token_accuracy": 0.4754572454094887, |
| "eval_cyber_num_tokens": 45532525.0, |
| "eval_cyber_runtime": 26.3904, |
| "eval_cyber_samples_per_second": 15.043, |
| "eval_cyber_steps_per_second": 3.789, |
| "step": 1600 |
| }, |
| { |
| "entropy": 0.659245578199625, |
| "epoch": 2.576, |
| "grad_norm": 23.125, |
| "learning_rate": 1.864666666666667e-05, |
| "loss": 10.4939, |
| "mean_token_accuracy": 0.8194193851202727, |
| "num_tokens": 45817478.0, |
| "step": 1610 |
| }, |
| { |
| "entropy": 0.6425925368443132, |
| "epoch": 2.592, |
| "grad_norm": 22.375, |
| "learning_rate": 1.8624444444444446e-05, |
| "loss": 10.2816, |
| "mean_token_accuracy": 0.8222486432641745, |
| "num_tokens": 46109575.0, |
| "step": 1620 |
| }, |
| { |
| "entropy": 0.6715872915461659, |
| "epoch": 2.608, |
| "grad_norm": 25.25, |
| "learning_rate": 1.8602222222222224e-05, |
| "loss": 10.6569, |
| "mean_token_accuracy": 0.8173692885786294, |
| "num_tokens": 46391461.0, |
| "step": 1630 |
| }, |
| { |
| "entropy": 0.6294447083026171, |
| "epoch": 2.624, |
| "grad_norm": 22.5, |
| "learning_rate": 1.858e-05, |
| "loss": 10.0396, |
| "mean_token_accuracy": 0.8279551289975643, |
| "num_tokens": 46683117.0, |
| "step": 1640 |
| }, |
| { |
| "entropy": 0.6628140497952699, |
| "epoch": 2.64, |
| "grad_norm": 23.375, |
| "learning_rate": 1.855777777777778e-05, |
| "loss": 10.5068, |
| "mean_token_accuracy": 0.8184260647743941, |
| "num_tokens": 46965534.0, |
| "step": 1650 |
| }, |
| { |
| "entropy": 0.6376811485737562, |
| "epoch": 2.656, |
| "grad_norm": 23.0, |
| "learning_rate": 1.8535555555555557e-05, |
| "loss": 10.0941, |
| "mean_token_accuracy": 0.8235533174127341, |
| "num_tokens": 47245852.0, |
| "step": 1660 |
| }, |
| { |
| "entropy": 0.6615891676396132, |
| "epoch": 2.672, |
| "grad_norm": 24.625, |
| "learning_rate": 1.8513333333333335e-05, |
| "loss": 10.5681, |
| "mean_token_accuracy": 0.8179556384682656, |
| "num_tokens": 47524916.0, |
| "step": 1670 |
| }, |
| { |
| "entropy": 0.6560400146991014, |
| "epoch": 2.6879999999999997, |
| "grad_norm": 22.0, |
| "learning_rate": 1.8491111111111112e-05, |
| "loss": 10.4122, |
| "mean_token_accuracy": 0.820205406472087, |
| "num_tokens": 47807131.0, |
| "step": 1680 |
| }, |
| { |
| "entropy": 0.6595821080729365, |
| "epoch": 2.7039999999999997, |
| "grad_norm": 23.875, |
| "learning_rate": 1.846888888888889e-05, |
| "loss": 10.5383, |
| "mean_token_accuracy": 0.8188040845096112, |
| "num_tokens": 48099654.0, |
| "step": 1690 |
| }, |
| { |
| "entropy": 0.6555765904486179, |
| "epoch": 2.7199999999999998, |
| "grad_norm": 24.125, |
| "learning_rate": 1.8446666666666667e-05, |
| "loss": 10.3768, |
| "mean_token_accuracy": 0.8199018821120262, |
| "num_tokens": 48375019.0, |
| "step": 1700 |
| }, |
| { |
| "epoch": 2.7199999999999998, |
| "eval_biology_entropy": 1.1961839065551758, |
| "eval_biology_loss": 1.474403977394104, |
| "eval_biology_mean_token_accuracy": 0.6521351528167725, |
| "eval_biology_num_tokens": 48375019.0, |
| "eval_biology_runtime": 22.2564, |
| "eval_biology_samples_per_second": 22.465, |
| "eval_biology_steps_per_second": 5.616, |
| "step": 1700 |
| }, |
| { |
| "epoch": 2.7199999999999998, |
| "eval_chemistry_entropy": 0.8893180413246154, |
| "eval_chemistry_loss": 1.0786951780319214, |
| "eval_chemistry_mean_token_accuracy": 0.7333630976676941, |
| "eval_chemistry_num_tokens": 48375019.0, |
| "eval_chemistry_runtime": 26.8892, |
| "eval_chemistry_samples_per_second": 18.595, |
| "eval_chemistry_steps_per_second": 4.649, |
| "step": 1700 |
| }, |
| { |
| "epoch": 2.7199999999999998, |
| "eval_math_entropy": 0.6758732006549836, |
| "eval_math_loss": 0.7107370495796204, |
| "eval_math_mean_token_accuracy": 0.8093916850090027, |
| "eval_math_num_tokens": 48375019.0, |
| "eval_math_runtime": 27.5333, |
| "eval_math_samples_per_second": 18.16, |
| "eval_math_steps_per_second": 4.54, |
| "step": 1700 |
| }, |
| { |
| "epoch": 2.7199999999999998, |
| "eval_cyber_entropy": 2.2349812412261962, |
| "eval_cyber_loss": 2.779714822769165, |
| "eval_cyber_mean_token_accuracy": 0.4819659352302551, |
| "eval_cyber_num_tokens": 48375019.0, |
| "eval_cyber_runtime": 26.1417, |
| "eval_cyber_samples_per_second": 15.186, |
| "eval_cyber_steps_per_second": 3.825, |
| "step": 1700 |
| }, |
| { |
| "entropy": 0.6559532942250371, |
| "epoch": 2.7359999999999998, |
| "grad_norm": 24.0, |
| "learning_rate": 1.842444444444445e-05, |
| "loss": 10.4594, |
| "mean_token_accuracy": 0.8196242570877075, |
| "num_tokens": 48659284.0, |
| "step": 1710 |
| }, |
| { |
| "entropy": 0.6569495009258389, |
| "epoch": 2.752, |
| "grad_norm": 22.25, |
| "learning_rate": 1.8402222222222223e-05, |
| "loss": 10.4467, |
| "mean_token_accuracy": 0.8201438017189503, |
| "num_tokens": 48943804.0, |
| "step": 1720 |
| }, |
| { |
| "entropy": 0.6529736818745733, |
| "epoch": 2.768, |
| "grad_norm": 21.875, |
| "learning_rate": 1.8380000000000004e-05, |
| "loss": 10.4356, |
| "mean_token_accuracy": 0.8206765007227659, |
| "num_tokens": 49230939.0, |
| "step": 1730 |
| }, |
| { |
| "entropy": 0.6720895063132047, |
| "epoch": 2.784, |
| "grad_norm": 23.0, |
| "learning_rate": 1.8357777777777778e-05, |
| "loss": 10.6977, |
| "mean_token_accuracy": 0.8159397479146719, |
| "num_tokens": 49504425.0, |
| "step": 1740 |
| }, |
| { |
| "entropy": 0.6508316185325385, |
| "epoch": 2.8, |
| "grad_norm": 23.125, |
| "learning_rate": 1.833555555555556e-05, |
| "loss": 10.351, |
| "mean_token_accuracy": 0.8222934223711491, |
| "num_tokens": 49782661.0, |
| "step": 1750 |
| }, |
| { |
| "entropy": 0.6627168050035834, |
| "epoch": 2.816, |
| "grad_norm": 24.5, |
| "learning_rate": 1.8313333333333333e-05, |
| "loss": 10.445, |
| "mean_token_accuracy": 0.8191927138715982, |
| "num_tokens": 50073632.0, |
| "step": 1760 |
| }, |
| { |
| "entropy": 0.6389238258823753, |
| "epoch": 2.832, |
| "grad_norm": 25.875, |
| "learning_rate": 1.8291111111111114e-05, |
| "loss": 10.2761, |
| "mean_token_accuracy": 0.8222359851002693, |
| "num_tokens": 50356964.0, |
| "step": 1770 |
| }, |
| { |
| "entropy": 0.6667038291692734, |
| "epoch": 2.848, |
| "grad_norm": 21.375, |
| "learning_rate": 1.8268888888888888e-05, |
| "loss": 10.586, |
| "mean_token_accuracy": 0.818210769072175, |
| "num_tokens": 50644535.0, |
| "step": 1780 |
| }, |
| { |
| "entropy": 0.6509249521419406, |
| "epoch": 2.864, |
| "grad_norm": 24.125, |
| "learning_rate": 1.824666666666667e-05, |
| "loss": 10.4338, |
| "mean_token_accuracy": 0.8203214287757874, |
| "num_tokens": 50925653.0, |
| "step": 1790 |
| }, |
| { |
| "entropy": 0.6507569069042802, |
| "epoch": 2.88, |
| "grad_norm": 23.625, |
| "learning_rate": 1.8224444444444447e-05, |
| "loss": 10.285, |
| "mean_token_accuracy": 0.8213449958711863, |
| "num_tokens": 51204374.0, |
| "step": 1800 |
| }, |
| { |
| "epoch": 2.88, |
| "eval_biology_entropy": 1.1794821062088012, |
| "eval_biology_loss": 1.4819698333740234, |
| "eval_biology_mean_token_accuracy": 0.6517698068618775, |
| "eval_biology_num_tokens": 51204374.0, |
| "eval_biology_runtime": 21.9924, |
| "eval_biology_samples_per_second": 22.735, |
| "eval_biology_steps_per_second": 5.684, |
| "step": 1800 |
| }, |
| { |
| "epoch": 2.88, |
| "eval_chemistry_entropy": 0.8722342405319213, |
| "eval_chemistry_loss": 1.0784211158752441, |
| "eval_chemistry_mean_token_accuracy": 0.7344747610092163, |
| "eval_chemistry_num_tokens": 51204374.0, |
| "eval_chemistry_runtime": 26.8909, |
| "eval_chemistry_samples_per_second": 18.594, |
| "eval_chemistry_steps_per_second": 4.648, |
| "step": 1800 |
| }, |
| { |
| "epoch": 2.88, |
| "eval_math_entropy": 0.6604897229671478, |
| "eval_math_loss": 0.7050039768218994, |
| "eval_math_mean_token_accuracy": 0.8106719055175782, |
| "eval_math_num_tokens": 51204374.0, |
| "eval_math_runtime": 27.5351, |
| "eval_math_samples_per_second": 18.159, |
| "eval_math_steps_per_second": 4.54, |
| "step": 1800 |
| }, |
| { |
| "epoch": 2.88, |
| "eval_cyber_entropy": 2.112373055815697, |
| "eval_cyber_loss": 2.8215885162353516, |
| "eval_cyber_mean_token_accuracy": 0.4802186432480812, |
| "eval_cyber_num_tokens": 51204374.0, |
| "eval_cyber_runtime": 26.1331, |
| "eval_cyber_samples_per_second": 15.191, |
| "eval_cyber_steps_per_second": 3.827, |
| "step": 1800 |
| }, |
| { |
| "entropy": 0.6363504879176617, |
| "epoch": 2.896, |
| "grad_norm": 23.75, |
| "learning_rate": 1.8202222222222225e-05, |
| "loss": 10.224, |
| "mean_token_accuracy": 0.8235703807324171, |
| "num_tokens": 51483944.0, |
| "step": 1810 |
| }, |
| { |
| "entropy": 0.6513262124732137, |
| "epoch": 2.912, |
| "grad_norm": 23.0, |
| "learning_rate": 1.8180000000000002e-05, |
| "loss": 10.3132, |
| "mean_token_accuracy": 0.8217961758375167, |
| "num_tokens": 51765755.0, |
| "step": 1820 |
| }, |
| { |
| "entropy": 0.6529291735962033, |
| "epoch": 2.928, |
| "grad_norm": 23.25, |
| "learning_rate": 1.815777777777778e-05, |
| "loss": 10.3766, |
| "mean_token_accuracy": 0.8222961116582155, |
| "num_tokens": 52056379.0, |
| "step": 1830 |
| }, |
| { |
| "entropy": 0.6346785051748156, |
| "epoch": 2.944, |
| "grad_norm": 23.75, |
| "learning_rate": 1.8135555555555557e-05, |
| "loss": 10.14, |
| "mean_token_accuracy": 0.8229989748448133, |
| "num_tokens": 52346232.0, |
| "step": 1840 |
| }, |
| { |
| "entropy": 0.6526656987145543, |
| "epoch": 2.96, |
| "grad_norm": 21.75, |
| "learning_rate": 1.8113333333333335e-05, |
| "loss": 10.38, |
| "mean_token_accuracy": 0.8202067915350199, |
| "num_tokens": 52633789.0, |
| "step": 1850 |
| }, |
| { |
| "entropy": 0.6502787992358208, |
| "epoch": 2.976, |
| "grad_norm": 21.625, |
| "learning_rate": 1.8091111111111113e-05, |
| "loss": 10.2748, |
| "mean_token_accuracy": 0.8214363507926464, |
| "num_tokens": 52911755.0, |
| "step": 1860 |
| }, |
| { |
| "entropy": 0.6417823160067201, |
| "epoch": 2.992, |
| "grad_norm": 22.875, |
| "learning_rate": 1.806888888888889e-05, |
| "loss": 10.2309, |
| "mean_token_accuracy": 0.8228288643062115, |
| "num_tokens": 53198176.0, |
| "step": 1870 |
| }, |
| { |
| "entropy": 0.6243049314245581, |
| "epoch": 3.008, |
| "grad_norm": 26.25, |
| "learning_rate": 1.8046666666666668e-05, |
| "loss": 9.7241, |
| "mean_token_accuracy": 0.8291641604155302, |
| "num_tokens": 53481893.0, |
| "step": 1880 |
| }, |
| { |
| "entropy": 0.5638323642313481, |
| "epoch": 3.024, |
| "grad_norm": 27.125, |
| "learning_rate": 1.8024444444444445e-05, |
| "loss": 8.9824, |
| "mean_token_accuracy": 0.8390695653855801, |
| "num_tokens": 53771717.0, |
| "step": 1890 |
| }, |
| { |
| "entropy": 0.558561889640987, |
| "epoch": 3.04, |
| "grad_norm": 24.125, |
| "learning_rate": 1.8002222222222223e-05, |
| "loss": 8.9038, |
| "mean_token_accuracy": 0.8408384509384632, |
| "num_tokens": 54058045.0, |
| "step": 1900 |
| }, |
| { |
| "epoch": 3.04, |
| "eval_biology_entropy": 1.0243187880516051, |
| "eval_biology_loss": 1.581600546836853, |
| "eval_biology_mean_token_accuracy": 0.6445312175750733, |
| "eval_biology_num_tokens": 54058045.0, |
| "eval_biology_runtime": 21.9531, |
| "eval_biology_samples_per_second": 22.776, |
| "eval_biology_steps_per_second": 5.694, |
| "step": 1900 |
| }, |
| { |
| "epoch": 3.04, |
| "eval_chemistry_entropy": 0.7532488117218018, |
| "eval_chemistry_loss": 1.1571515798568726, |
| "eval_chemistry_mean_token_accuracy": 0.7285859537124634, |
| "eval_chemistry_num_tokens": 54058045.0, |
| "eval_chemistry_runtime": 26.8231, |
| "eval_chemistry_samples_per_second": 18.641, |
| "eval_chemistry_steps_per_second": 4.66, |
| "step": 1900 |
| }, |
| { |
| "epoch": 3.04, |
| "eval_math_entropy": 0.5930595288276672, |
| "eval_math_loss": 0.7225678563117981, |
| "eval_math_mean_token_accuracy": 0.8094966850280761, |
| "eval_math_num_tokens": 54058045.0, |
| "eval_math_runtime": 27.497, |
| "eval_math_samples_per_second": 18.184, |
| "eval_math_steps_per_second": 4.546, |
| "step": 1900 |
| }, |
| { |
| "epoch": 3.04, |
| "eval_cyber_entropy": 1.8411164230108261, |
| "eval_cyber_loss": 3.063901424407959, |
| "eval_cyber_mean_token_accuracy": 0.47065704002976416, |
| "eval_cyber_num_tokens": 54058045.0, |
| "eval_cyber_runtime": 26.1343, |
| "eval_cyber_samples_per_second": 15.191, |
| "eval_cyber_steps_per_second": 3.826, |
| "step": 1900 |
| }, |
| { |
| "entropy": 0.552313212864101, |
| "epoch": 3.056, |
| "grad_norm": 26.625, |
| "learning_rate": 1.798e-05, |
| "loss": 8.7715, |
| "mean_token_accuracy": 0.8414296887814998, |
| "num_tokens": 54334332.0, |
| "step": 1910 |
| }, |
| { |
| "entropy": 0.5570558808743954, |
| "epoch": 3.072, |
| "grad_norm": 27.125, |
| "learning_rate": 1.7957777777777778e-05, |
| "loss": 8.7638, |
| "mean_token_accuracy": 0.8418575689196587, |
| "num_tokens": 54624543.0, |
| "step": 1920 |
| }, |
| { |
| "entropy": 0.5351565392687917, |
| "epoch": 3.088, |
| "grad_norm": 27.625, |
| "learning_rate": 1.7935555555555556e-05, |
| "loss": 8.5286, |
| "mean_token_accuracy": 0.8457726195454598, |
| "num_tokens": 54907550.0, |
| "step": 1930 |
| }, |
| { |
| "entropy": 0.5519297284074127, |
| "epoch": 3.104, |
| "grad_norm": 27.75, |
| "learning_rate": 1.7913333333333337e-05, |
| "loss": 8.7504, |
| "mean_token_accuracy": 0.8423502463847399, |
| "num_tokens": 55190959.0, |
| "step": 1940 |
| }, |
| { |
| "entropy": 0.544937571324408, |
| "epoch": 3.12, |
| "grad_norm": 27.0, |
| "learning_rate": 1.789111111111111e-05, |
| "loss": 8.6451, |
| "mean_token_accuracy": 0.8444420550018549, |
| "num_tokens": 55481635.0, |
| "step": 1950 |
| }, |
| { |
| "entropy": 0.546911165677011, |
| "epoch": 3.136, |
| "grad_norm": 26.75, |
| "learning_rate": 1.7868888888888892e-05, |
| "loss": 8.6997, |
| "mean_token_accuracy": 0.8431835647672414, |
| "num_tokens": 55769010.0, |
| "step": 1960 |
| }, |
| { |
| "entropy": 0.5542461348697543, |
| "epoch": 3.152, |
| "grad_norm": 26.5, |
| "learning_rate": 1.7846666666666666e-05, |
| "loss": 8.7996, |
| "mean_token_accuracy": 0.842128399387002, |
| "num_tokens": 56053160.0, |
| "step": 1970 |
| }, |
| { |
| "entropy": 0.5676137331873179, |
| "epoch": 3.168, |
| "grad_norm": 26.75, |
| "learning_rate": 1.7824444444444447e-05, |
| "loss": 8.9893, |
| "mean_token_accuracy": 0.8386933848261833, |
| "num_tokens": 56337066.0, |
| "step": 1980 |
| }, |
| { |
| "entropy": 0.5574962265789509, |
| "epoch": 3.184, |
| "grad_norm": 30.0, |
| "learning_rate": 1.780222222222222e-05, |
| "loss": 8.8126, |
| "mean_token_accuracy": 0.8404616348445415, |
| "num_tokens": 56618899.0, |
| "step": 1990 |
| }, |
| { |
| "entropy": 0.5407627185806632, |
| "epoch": 3.2, |
| "grad_norm": 26.0, |
| "learning_rate": 1.7780000000000003e-05, |
| "loss": 8.6062, |
| "mean_token_accuracy": 0.8444454524666071, |
| "num_tokens": 56910071.0, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.2, |
| "eval_biology_entropy": 0.9433750138282776, |
| "eval_biology_loss": 1.6609779596328735, |
| "eval_biology_mean_token_accuracy": 0.6404439859390259, |
| "eval_biology_num_tokens": 56910071.0, |
| "eval_biology_runtime": 21.9804, |
| "eval_biology_samples_per_second": 22.748, |
| "eval_biology_steps_per_second": 5.687, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.2, |
| "eval_chemistry_entropy": 0.697946096420288, |
| "eval_chemistry_loss": 1.214890718460083, |
| "eval_chemistry_mean_token_accuracy": 0.7252082509994506, |
| "eval_chemistry_num_tokens": 56910071.0, |
| "eval_chemistry_runtime": 26.8518, |
| "eval_chemistry_samples_per_second": 18.621, |
| "eval_chemistry_steps_per_second": 4.655, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.2, |
| "eval_math_entropy": 0.5889736828804016, |
| "eval_math_loss": 0.72825688123703, |
| "eval_math_mean_token_accuracy": 0.809436321735382, |
| "eval_math_num_tokens": 56910071.0, |
| "eval_math_runtime": 27.524, |
| "eval_math_samples_per_second": 18.166, |
| "eval_math_steps_per_second": 4.541, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.2, |
| "eval_cyber_entropy": 1.8025889378786086, |
| "eval_cyber_loss": 3.0566282272338867, |
| "eval_cyber_mean_token_accuracy": 0.46734614998102186, |
| "eval_cyber_num_tokens": 56910071.0, |
| "eval_cyber_runtime": 26.2014, |
| "eval_cyber_samples_per_second": 15.152, |
| "eval_cyber_steps_per_second": 3.817, |
| "step": 2000 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 10000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 16, |
| "save_steps": 1000, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 5.791878293573609e+18, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
