| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 8.0, |
| "eval_steps": 100, |
| "global_step": 5000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 0.6421742392703891, |
| "epoch": 0.016, |
| "grad_norm": 4.4375, |
| "learning_rate": 1.8e-07, |
| "loss": 0.953, |
| "mean_token_accuracy": 0.792544937506318, |
| "num_tokens": 282849.0, |
| "step": 10 |
| }, |
| { |
| "entropy": 0.6719903893768787, |
| "epoch": 0.032, |
| "grad_norm": 4.25, |
| "learning_rate": 3.8e-07, |
| "loss": 0.9758, |
| "mean_token_accuracy": 0.7857426963746548, |
| "num_tokens": 561726.0, |
| "step": 20 |
| }, |
| { |
| "entropy": 0.6603780068457127, |
| "epoch": 0.048, |
| "grad_norm": 4.40625, |
| "learning_rate": 5.800000000000001e-07, |
| "loss": 0.9582, |
| "mean_token_accuracy": 0.7894285101443529, |
| "num_tokens": 842092.0, |
| "step": 30 |
| }, |
| { |
| "entropy": 0.6775440879166126, |
| "epoch": 0.064, |
| "grad_norm": 3.890625, |
| "learning_rate": 7.8e-07, |
| "loss": 0.9776, |
| "mean_token_accuracy": 0.785004175081849, |
| "num_tokens": 1133349.0, |
| "step": 40 |
| }, |
| { |
| "entropy": 0.6926743468269706, |
| "epoch": 0.08, |
| "grad_norm": 4.6875, |
| "learning_rate": 9.800000000000001e-07, |
| "loss": 0.9497, |
| "mean_token_accuracy": 0.7896979171782732, |
| "num_tokens": 1422340.0, |
| "step": 50 |
| }, |
| { |
| "entropy": 0.7043748194351792, |
| "epoch": 0.096, |
| "grad_norm": 3.484375, |
| "learning_rate": 1.1800000000000001e-06, |
| "loss": 0.9217, |
| "mean_token_accuracy": 0.7949878334999084, |
| "num_tokens": 1711821.0, |
| "step": 60 |
| }, |
| { |
| "entropy": 0.7551965553313493, |
| "epoch": 0.112, |
| "grad_norm": 2.71875, |
| "learning_rate": 1.3800000000000001e-06, |
| "loss": 0.931, |
| "mean_token_accuracy": 0.7912098452448845, |
| "num_tokens": 1991855.0, |
| "step": 70 |
| }, |
| { |
| "entropy": 0.7913430579006672, |
| "epoch": 0.128, |
| "grad_norm": 2.359375, |
| "learning_rate": 1.5800000000000001e-06, |
| "loss": 0.9172, |
| "mean_token_accuracy": 0.7935858864337206, |
| "num_tokens": 2288720.0, |
| "step": 80 |
| }, |
| { |
| "entropy": 0.7881425807252527, |
| "epoch": 0.144, |
| "grad_norm": 2.125, |
| "learning_rate": 1.7800000000000001e-06, |
| "loss": 0.917, |
| "mean_token_accuracy": 0.7932404045015573, |
| "num_tokens": 2564352.0, |
| "step": 90 |
| }, |
| { |
| "entropy": 0.7627748234197498, |
| "epoch": 0.16, |
| "grad_norm": 2.125, |
| "learning_rate": 1.98e-06, |
| "loss": 0.9014, |
| "mean_token_accuracy": 0.7970093633979559, |
| "num_tokens": 2841876.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_math_entropy": 0.7972113332748413, |
| "eval_math_loss": 0.9417253136634827, |
| "eval_math_mean_token_accuracy": 0.7907100429534912, |
| "eval_math_num_tokens": 2841876.0, |
| "eval_math_runtime": 35.8728, |
| "eval_math_samples_per_second": 13.938, |
| "eval_math_steps_per_second": 3.485, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_physics_entropy": 0.7183168392181396, |
| "eval_physics_loss": 0.7405859231948853, |
| "eval_physics_mean_token_accuracy": 0.822186062335968, |
| "eval_physics_num_tokens": 2841876.0, |
| "eval_physics_runtime": 41.4814, |
| "eval_physics_samples_per_second": 12.054, |
| "eval_physics_steps_per_second": 3.013, |
| "step": 100 |
| }, |
| { |
| "entropy": 0.7770459342747926, |
| "epoch": 0.176, |
| "grad_norm": 1.9765625, |
| "learning_rate": 2.1800000000000003e-06, |
| "loss": 0.8952, |
| "mean_token_accuracy": 0.7969978783279658, |
| "num_tokens": 3129237.0, |
| "step": 110 |
| }, |
| { |
| "entropy": 0.7958783155307174, |
| "epoch": 0.192, |
| "grad_norm": 2.3125, |
| "learning_rate": 2.38e-06, |
| "loss": 0.9119, |
| "mean_token_accuracy": 0.793396720290184, |
| "num_tokens": 3414875.0, |
| "step": 120 |
| }, |
| { |
| "entropy": 0.7911858700215817, |
| "epoch": 0.208, |
| "grad_norm": 2.4375, |
| "learning_rate": 2.5800000000000003e-06, |
| "loss": 0.8979, |
| "mean_token_accuracy": 0.7956417176872492, |
| "num_tokens": 3695623.0, |
| "step": 130 |
| }, |
| { |
| "entropy": 0.7871318949386478, |
| "epoch": 0.224, |
| "grad_norm": 2.09375, |
| "learning_rate": 2.7800000000000005e-06, |
| "loss": 0.8882, |
| "mean_token_accuracy": 0.7958744909614325, |
| "num_tokens": 3977666.0, |
| "step": 140 |
| }, |
| { |
| "entropy": 0.7629930060356855, |
| "epoch": 0.24, |
| "grad_norm": 1.9453125, |
| "learning_rate": 2.9800000000000003e-06, |
| "loss": 0.8654, |
| "mean_token_accuracy": 0.8001766365021468, |
| "num_tokens": 4269287.0, |
| "step": 150 |
| }, |
| { |
| "entropy": 0.7563591374084353, |
| "epoch": 0.256, |
| "grad_norm": 2.015625, |
| "learning_rate": 3.1800000000000005e-06, |
| "loss": 0.846, |
| "mean_token_accuracy": 0.8012917254120111, |
| "num_tokens": 4558632.0, |
| "step": 160 |
| }, |
| { |
| "entropy": 0.7706880155950785, |
| "epoch": 0.272, |
| "grad_norm": 1.859375, |
| "learning_rate": 3.3800000000000007e-06, |
| "loss": 0.8589, |
| "mean_token_accuracy": 0.8001386649906636, |
| "num_tokens": 4839270.0, |
| "step": 170 |
| }, |
| { |
| "entropy": 0.7551738245412707, |
| "epoch": 0.288, |
| "grad_norm": 1.609375, |
| "learning_rate": 3.58e-06, |
| "loss": 0.8468, |
| "mean_token_accuracy": 0.8015156831592322, |
| "num_tokens": 5125447.0, |
| "step": 180 |
| }, |
| { |
| "entropy": 0.7806578669697046, |
| "epoch": 0.304, |
| "grad_norm": 1.7734375, |
| "learning_rate": 3.7800000000000002e-06, |
| "loss": 0.8405, |
| "mean_token_accuracy": 0.7998461078852415, |
| "num_tokens": 5415925.0, |
| "step": 190 |
| }, |
| { |
| "entropy": 0.7316340124234557, |
| "epoch": 0.32, |
| "grad_norm": 1.6953125, |
| "learning_rate": 3.980000000000001e-06, |
| "loss": 0.7942, |
| "mean_token_accuracy": 0.8076768793165684, |
| "num_tokens": 5710178.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_math_entropy": 0.7726770024299622, |
| "eval_math_loss": 0.84871506690979, |
| "eval_math_mean_token_accuracy": 0.7979933972358704, |
| "eval_math_num_tokens": 5710178.0, |
| "eval_math_runtime": 35.765, |
| "eval_math_samples_per_second": 13.98, |
| "eval_math_steps_per_second": 3.495, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_physics_entropy": 0.6965276994705201, |
| "eval_physics_loss": 0.7036627531051636, |
| "eval_physics_mean_token_accuracy": 0.820941798210144, |
| "eval_physics_num_tokens": 5710178.0, |
| "eval_physics_runtime": 41.4597, |
| "eval_physics_samples_per_second": 12.06, |
| "eval_physics_steps_per_second": 3.015, |
| "step": 200 |
| }, |
| { |
| "entropy": 0.7737893080338836, |
| "epoch": 0.336, |
| "grad_norm": 1.828125, |
| "learning_rate": 4.18e-06, |
| "loss": 0.8126, |
| "mean_token_accuracy": 0.800426610186696, |
| "num_tokens": 5993843.0, |
| "step": 210 |
| }, |
| { |
| "entropy": 0.7482354704290628, |
| "epoch": 0.352, |
| "grad_norm": 1.9609375, |
| "learning_rate": 4.38e-06, |
| "loss": 0.7874, |
| "mean_token_accuracy": 0.8061643201857805, |
| "num_tokens": 6279137.0, |
| "step": 220 |
| }, |
| { |
| "entropy": 0.7608217550441623, |
| "epoch": 0.368, |
| "grad_norm": 1.65625, |
| "learning_rate": 4.58e-06, |
| "loss": 0.7742, |
| "mean_token_accuracy": 0.803000221401453, |
| "num_tokens": 6562387.0, |
| "step": 230 |
| }, |
| { |
| "entropy": 0.748550821095705, |
| "epoch": 0.384, |
| "grad_norm": 1.84375, |
| "learning_rate": 4.78e-06, |
| "loss": 0.7565, |
| "mean_token_accuracy": 0.8081529751420021, |
| "num_tokens": 6851520.0, |
| "step": 240 |
| }, |
| { |
| "entropy": 0.7808488950133323, |
| "epoch": 0.4, |
| "grad_norm": 1.75, |
| "learning_rate": 4.980000000000001e-06, |
| "loss": 0.7953, |
| "mean_token_accuracy": 0.7999043371528387, |
| "num_tokens": 7138932.0, |
| "step": 250 |
| }, |
| { |
| "entropy": 0.7674930268898607, |
| "epoch": 0.416, |
| "grad_norm": 1.6796875, |
| "learning_rate": 5.18e-06, |
| "loss": 0.7597, |
| "mean_token_accuracy": 0.8044768057763576, |
| "num_tokens": 7420720.0, |
| "step": 260 |
| }, |
| { |
| "entropy": 0.7496220650151372, |
| "epoch": 0.432, |
| "grad_norm": 1.5390625, |
| "learning_rate": 5.380000000000001e-06, |
| "loss": 0.7468, |
| "mean_token_accuracy": 0.8073872365057468, |
| "num_tokens": 7695518.0, |
| "step": 270 |
| }, |
| { |
| "entropy": 0.7439631385728717, |
| "epoch": 0.448, |
| "grad_norm": 2.21875, |
| "learning_rate": 5.580000000000001e-06, |
| "loss": 0.7339, |
| "mean_token_accuracy": 0.8076559219509363, |
| "num_tokens": 7983543.0, |
| "step": 280 |
| }, |
| { |
| "entropy": 0.7457249015569687, |
| "epoch": 0.464, |
| "grad_norm": 2.65625, |
| "learning_rate": 5.78e-06, |
| "loss": 0.741, |
| "mean_token_accuracy": 0.8087894931435585, |
| "num_tokens": 8271787.0, |
| "step": 290 |
| }, |
| { |
| "entropy": 0.7761816097423434, |
| "epoch": 0.48, |
| "grad_norm": 1.78125, |
| "learning_rate": 5.98e-06, |
| "loss": 0.7548, |
| "mean_token_accuracy": 0.8063571337610483, |
| "num_tokens": 8556170.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_math_entropy": 0.7653279647827148, |
| "eval_math_loss": 0.7636919021606445, |
| "eval_math_mean_token_accuracy": 0.8056883907318115, |
| "eval_math_num_tokens": 8556170.0, |
| "eval_math_runtime": 35.8013, |
| "eval_math_samples_per_second": 13.966, |
| "eval_math_steps_per_second": 3.491, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_physics_entropy": 0.6890090732574463, |
| "eval_physics_loss": 0.6574941277503967, |
| "eval_physics_mean_token_accuracy": 0.8227382025718689, |
| "eval_physics_num_tokens": 8556170.0, |
| "eval_physics_runtime": 41.4883, |
| "eval_physics_samples_per_second": 12.052, |
| "eval_physics_steps_per_second": 3.013, |
| "step": 300 |
| }, |
| { |
| "entropy": 0.7349074123427272, |
| "epoch": 0.496, |
| "grad_norm": 1.828125, |
| "learning_rate": 6.18e-06, |
| "loss": 0.7172, |
| "mean_token_accuracy": 0.8132212206721305, |
| "num_tokens": 8840590.0, |
| "step": 310 |
| }, |
| { |
| "entropy": 0.7443849379196763, |
| "epoch": 0.512, |
| "grad_norm": 1.6640625, |
| "learning_rate": 6.380000000000001e-06, |
| "loss": 0.7181, |
| "mean_token_accuracy": 0.812743404135108, |
| "num_tokens": 9131903.0, |
| "step": 320 |
| }, |
| { |
| "entropy": 0.7170431386679411, |
| "epoch": 0.528, |
| "grad_norm": 1.7109375, |
| "learning_rate": 6.5800000000000005e-06, |
| "loss": 0.7045, |
| "mean_token_accuracy": 0.8160003747791051, |
| "num_tokens": 9429519.0, |
| "step": 330 |
| }, |
| { |
| "entropy": 0.7328394843265414, |
| "epoch": 0.544, |
| "grad_norm": 2.234375, |
| "learning_rate": 6.780000000000001e-06, |
| "loss": 0.722, |
| "mean_token_accuracy": 0.8143766384571791, |
| "num_tokens": 9718247.0, |
| "step": 340 |
| }, |
| { |
| "entropy": 0.7620531113818287, |
| "epoch": 0.56, |
| "grad_norm": 2.453125, |
| "learning_rate": 6.98e-06, |
| "loss": 0.7395, |
| "mean_token_accuracy": 0.8070962373167276, |
| "num_tokens": 9991100.0, |
| "step": 350 |
| }, |
| { |
| "entropy": 0.7331368897110224, |
| "epoch": 0.576, |
| "grad_norm": 1.8515625, |
| "learning_rate": 7.180000000000001e-06, |
| "loss": 0.7173, |
| "mean_token_accuracy": 0.8115411166101694, |
| "num_tokens": 10276090.0, |
| "step": 360 |
| }, |
| { |
| "entropy": 0.7342592483386398, |
| "epoch": 0.592, |
| "grad_norm": 1.546875, |
| "learning_rate": 7.3800000000000005e-06, |
| "loss": 0.7141, |
| "mean_token_accuracy": 0.8124774299561978, |
| "num_tokens": 10566762.0, |
| "step": 370 |
| }, |
| { |
| "entropy": 0.7440765531733632, |
| "epoch": 0.608, |
| "grad_norm": 1.59375, |
| "learning_rate": 7.58e-06, |
| "loss": 0.7237, |
| "mean_token_accuracy": 0.8122259072959424, |
| "num_tokens": 10857002.0, |
| "step": 380 |
| }, |
| { |
| "entropy": 0.7216648861765862, |
| "epoch": 0.624, |
| "grad_norm": 2.25, |
| "learning_rate": 7.78e-06, |
| "loss": 0.7057, |
| "mean_token_accuracy": 0.8139940846711398, |
| "num_tokens": 11147326.0, |
| "step": 390 |
| }, |
| { |
| "entropy": 0.7119627617299557, |
| "epoch": 0.64, |
| "grad_norm": 1.5234375, |
| "learning_rate": 7.980000000000002e-06, |
| "loss": 0.6947, |
| "mean_token_accuracy": 0.8168974690139293, |
| "num_tokens": 11437710.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_math_entropy": 0.7429183626174927, |
| "eval_math_loss": 0.7330864071846008, |
| "eval_math_mean_token_accuracy": 0.8096566972732544, |
| "eval_math_num_tokens": 11437710.0, |
| "eval_math_runtime": 35.7944, |
| "eval_math_samples_per_second": 13.969, |
| "eval_math_steps_per_second": 3.492, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_physics_entropy": 0.6820782516002655, |
| "eval_physics_loss": 0.6511842012405396, |
| "eval_physics_mean_token_accuracy": 0.8221303143501282, |
| "eval_physics_num_tokens": 11437710.0, |
| "eval_physics_runtime": 41.4584, |
| "eval_physics_samples_per_second": 12.06, |
| "eval_physics_steps_per_second": 3.015, |
| "step": 400 |
| }, |
| { |
| "entropy": 0.712121813185513, |
| "epoch": 0.656, |
| "grad_norm": 1.40625, |
| "learning_rate": 8.18e-06, |
| "loss": 0.6914, |
| "mean_token_accuracy": 0.8179911982268095, |
| "num_tokens": 11727268.0, |
| "step": 410 |
| }, |
| { |
| "entropy": 0.7159143410623073, |
| "epoch": 0.672, |
| "grad_norm": 1.65625, |
| "learning_rate": 8.380000000000001e-06, |
| "loss": 0.7016, |
| "mean_token_accuracy": 0.813546035438776, |
| "num_tokens": 12020478.0, |
| "step": 420 |
| }, |
| { |
| "entropy": 0.7149255678057671, |
| "epoch": 0.688, |
| "grad_norm": 1.5078125, |
| "learning_rate": 8.580000000000001e-06, |
| "loss": 0.6908, |
| "mean_token_accuracy": 0.8162215895950794, |
| "num_tokens": 12300380.0, |
| "step": 430 |
| }, |
| { |
| "entropy": 0.7053035136312247, |
| "epoch": 0.704, |
| "grad_norm": 1.78125, |
| "learning_rate": 8.78e-06, |
| "loss": 0.6839, |
| "mean_token_accuracy": 0.8172036681324244, |
| "num_tokens": 12591366.0, |
| "step": 440 |
| }, |
| { |
| "entropy": 0.7228720126673579, |
| "epoch": 0.72, |
| "grad_norm": 1.7421875, |
| "learning_rate": 8.98e-06, |
| "loss": 0.6945, |
| "mean_token_accuracy": 0.8145164046436548, |
| "num_tokens": 12869618.0, |
| "step": 450 |
| }, |
| { |
| "entropy": 0.7052439017221331, |
| "epoch": 0.736, |
| "grad_norm": 1.75, |
| "learning_rate": 9.180000000000002e-06, |
| "loss": 0.6866, |
| "mean_token_accuracy": 0.8176059562712907, |
| "num_tokens": 13156311.0, |
| "step": 460 |
| }, |
| { |
| "entropy": 0.692553037032485, |
| "epoch": 0.752, |
| "grad_norm": 1.5625, |
| "learning_rate": 9.38e-06, |
| "loss": 0.6644, |
| "mean_token_accuracy": 0.821098442748189, |
| "num_tokens": 13438911.0, |
| "step": 470 |
| }, |
| { |
| "entropy": 0.6714967370033265, |
| "epoch": 0.768, |
| "grad_norm": 1.3203125, |
| "learning_rate": 9.58e-06, |
| "loss": 0.6595, |
| "mean_token_accuracy": 0.825907539948821, |
| "num_tokens": 13735397.0, |
| "step": 480 |
| }, |
| { |
| "entropy": 0.7148594724014401, |
| "epoch": 0.784, |
| "grad_norm": 1.7890625, |
| "learning_rate": 9.780000000000001e-06, |
| "loss": 0.6954, |
| "mean_token_accuracy": 0.8156244982033968, |
| "num_tokens": 14024497.0, |
| "step": 490 |
| }, |
| { |
| "entropy": 0.7110015576705336, |
| "epoch": 0.8, |
| "grad_norm": 1.6328125, |
| "learning_rate": 9.980000000000001e-06, |
| "loss": 0.69, |
| "mean_token_accuracy": 0.8156023971736431, |
| "num_tokens": 14316427.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_math_entropy": 0.7275117125511169, |
| "eval_math_loss": 0.7106672525405884, |
| "eval_math_mean_token_accuracy": 0.8132652158737183, |
| "eval_math_num_tokens": 14316427.0, |
| "eval_math_runtime": 35.7797, |
| "eval_math_samples_per_second": 13.974, |
| "eval_math_steps_per_second": 3.494, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_physics_entropy": 0.6752480027675629, |
| "eval_physics_loss": 0.6474179029464722, |
| "eval_physics_mean_token_accuracy": 0.8222041974067688, |
| "eval_physics_num_tokens": 14316427.0, |
| "eval_physics_runtime": 41.1592, |
| "eval_physics_samples_per_second": 12.148, |
| "eval_physics_steps_per_second": 3.037, |
| "step": 500 |
| }, |
| { |
| "entropy": 0.7031571570783853, |
| "epoch": 0.816, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.018e-05, |
| "loss": 0.6756, |
| "mean_token_accuracy": 0.8171336345374585, |
| "num_tokens": 14607136.0, |
| "step": 510 |
| }, |
| { |
| "entropy": 0.693414007127285, |
| "epoch": 0.832, |
| "grad_norm": 1.5859375, |
| "learning_rate": 1.038e-05, |
| "loss": 0.6706, |
| "mean_token_accuracy": 0.8211553748697042, |
| "num_tokens": 14894084.0, |
| "step": 520 |
| }, |
| { |
| "entropy": 0.7212638273835182, |
| "epoch": 0.848, |
| "grad_norm": 1.59375, |
| "learning_rate": 1.0580000000000002e-05, |
| "loss": 0.6965, |
| "mean_token_accuracy": 0.8145442601293326, |
| "num_tokens": 15177039.0, |
| "step": 530 |
| }, |
| { |
| "entropy": 0.7037423193454743, |
| "epoch": 0.864, |
| "grad_norm": 1.4296875, |
| "learning_rate": 1.0780000000000002e-05, |
| "loss": 0.6824, |
| "mean_token_accuracy": 0.816459146887064, |
| "num_tokens": 15463393.0, |
| "step": 540 |
| }, |
| { |
| "entropy": 0.7084541182965041, |
| "epoch": 0.88, |
| "grad_norm": 1.5546875, |
| "learning_rate": 1.0980000000000002e-05, |
| "loss": 0.6861, |
| "mean_token_accuracy": 0.8153413504362106, |
| "num_tokens": 15747943.0, |
| "step": 550 |
| }, |
| { |
| "entropy": 0.7014794830232859, |
| "epoch": 0.896, |
| "grad_norm": 1.5078125, |
| "learning_rate": 1.1180000000000001e-05, |
| "loss": 0.6844, |
| "mean_token_accuracy": 0.8173291139304638, |
| "num_tokens": 16027931.0, |
| "step": 560 |
| }, |
| { |
| "entropy": 0.6914853570982814, |
| "epoch": 0.912, |
| "grad_norm": 1.546875, |
| "learning_rate": 1.138e-05, |
| "loss": 0.6643, |
| "mean_token_accuracy": 0.8196637153625488, |
| "num_tokens": 16317127.0, |
| "step": 570 |
| }, |
| { |
| "entropy": 0.6866581376641989, |
| "epoch": 0.928, |
| "grad_norm": 1.46875, |
| "learning_rate": 1.1580000000000001e-05, |
| "loss": 0.661, |
| "mean_token_accuracy": 0.8199340119957924, |
| "num_tokens": 16598757.0, |
| "step": 580 |
| }, |
| { |
| "entropy": 0.7026519874110818, |
| "epoch": 0.944, |
| "grad_norm": 7.0625, |
| "learning_rate": 1.178e-05, |
| "loss": 0.6838, |
| "mean_token_accuracy": 0.8174264524132013, |
| "num_tokens": 16882236.0, |
| "step": 590 |
| }, |
| { |
| "entropy": 0.7053765447810292, |
| "epoch": 0.96, |
| "grad_norm": 1.515625, |
| "learning_rate": 1.198e-05, |
| "loss": 0.6832, |
| "mean_token_accuracy": 0.8171489015221596, |
| "num_tokens": 17169495.0, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.96, |
| "eval_math_entropy": 0.7025088634490967, |
| "eval_math_loss": 0.6926642656326294, |
| "eval_math_mean_token_accuracy": 0.8157670006752015, |
| "eval_math_num_tokens": 17169495.0, |
| "eval_math_runtime": 35.7558, |
| "eval_math_samples_per_second": 13.984, |
| "eval_math_steps_per_second": 3.496, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.96, |
| "eval_physics_entropy": 0.6707950072288513, |
| "eval_physics_loss": 0.6470632553100586, |
| "eval_physics_mean_token_accuracy": 0.8212249627113343, |
| "eval_physics_num_tokens": 17169495.0, |
| "eval_physics_runtime": 41.7459, |
| "eval_physics_samples_per_second": 11.977, |
| "eval_physics_steps_per_second": 2.994, |
| "step": 600 |
| }, |
| { |
| "entropy": 0.684873866662383, |
| "epoch": 0.976, |
| "grad_norm": 1.8515625, |
| "learning_rate": 1.218e-05, |
| "loss": 0.6671, |
| "mean_token_accuracy": 0.8196386702358722, |
| "num_tokens": 17454147.0, |
| "step": 610 |
| }, |
| { |
| "entropy": 0.7006405493244529, |
| "epoch": 0.992, |
| "grad_norm": 1.5390625, |
| "learning_rate": 1.2380000000000002e-05, |
| "loss": 0.672, |
| "mean_token_accuracy": 0.8165230687707663, |
| "num_tokens": 17742596.0, |
| "step": 620 |
| }, |
| { |
| "entropy": 0.6851861752569676, |
| "epoch": 1.008, |
| "grad_norm": 1.3828125, |
| "learning_rate": 1.2580000000000002e-05, |
| "loss": 0.6635, |
| "mean_token_accuracy": 0.8197994366288185, |
| "num_tokens": 18033540.0, |
| "step": 630 |
| }, |
| { |
| "entropy": 0.6624577900394797, |
| "epoch": 1.024, |
| "grad_norm": 1.5, |
| "learning_rate": 1.2780000000000001e-05, |
| "loss": 0.6405, |
| "mean_token_accuracy": 0.823651035502553, |
| "num_tokens": 18316346.0, |
| "step": 640 |
| }, |
| { |
| "entropy": 0.6590202970430254, |
| "epoch": 1.04, |
| "grad_norm": 1.6484375, |
| "learning_rate": 1.2980000000000001e-05, |
| "loss": 0.6392, |
| "mean_token_accuracy": 0.8249083694070578, |
| "num_tokens": 18595230.0, |
| "step": 650 |
| }, |
| { |
| "entropy": 0.6720681777223945, |
| "epoch": 1.056, |
| "grad_norm": 1.5625, |
| "learning_rate": 1.3180000000000001e-05, |
| "loss": 0.6481, |
| "mean_token_accuracy": 0.8219482194632292, |
| "num_tokens": 18885547.0, |
| "step": 660 |
| }, |
| { |
| "entropy": 0.666364430822432, |
| "epoch": 1.072, |
| "grad_norm": 1.5625, |
| "learning_rate": 1.3380000000000002e-05, |
| "loss": 0.6453, |
| "mean_token_accuracy": 0.821395493671298, |
| "num_tokens": 19169083.0, |
| "step": 670 |
| }, |
| { |
| "entropy": 0.6612174535170198, |
| "epoch": 1.088, |
| "grad_norm": 1.515625, |
| "learning_rate": 1.3580000000000002e-05, |
| "loss": 0.6328, |
| "mean_token_accuracy": 0.823925781622529, |
| "num_tokens": 19461465.0, |
| "step": 680 |
| }, |
| { |
| "entropy": 0.6591472556814552, |
| "epoch": 1.104, |
| "grad_norm": 1.3125, |
| "learning_rate": 1.378e-05, |
| "loss": 0.6414, |
| "mean_token_accuracy": 0.8235542017966508, |
| "num_tokens": 19755027.0, |
| "step": 690 |
| }, |
| { |
| "entropy": 0.6446556953713298, |
| "epoch": 1.12, |
| "grad_norm": 1.5390625, |
| "learning_rate": 1.398e-05, |
| "loss": 0.6235, |
| "mean_token_accuracy": 0.8280277159065008, |
| "num_tokens": 20043293.0, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.12, |
| "eval_math_entropy": 0.6619428229331971, |
| "eval_math_loss": 0.6814246773719788, |
| "eval_math_mean_token_accuracy": 0.8176272907257081, |
| "eval_math_num_tokens": 20043293.0, |
| "eval_math_runtime": 35.7663, |
| "eval_math_samples_per_second": 13.98, |
| "eval_math_steps_per_second": 3.495, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.12, |
| "eval_physics_entropy": 0.6304002203941346, |
| "eval_physics_loss": 0.6498542428016663, |
| "eval_physics_mean_token_accuracy": 0.8209347081184387, |
| "eval_physics_num_tokens": 20043293.0, |
| "eval_physics_runtime": 41.4297, |
| "eval_physics_samples_per_second": 12.069, |
| "eval_physics_steps_per_second": 3.017, |
| "step": 700 |
| }, |
| { |
| "entropy": 0.6416833782568574, |
| "epoch": 1.1360000000000001, |
| "grad_norm": 1.46875, |
| "learning_rate": 1.418e-05, |
| "loss": 0.6298, |
| "mean_token_accuracy": 0.8263315539807081, |
| "num_tokens": 20323010.0, |
| "step": 710 |
| }, |
| { |
| "entropy": 0.6587465461343527, |
| "epoch": 1.152, |
| "grad_norm": 1.578125, |
| "learning_rate": 1.4380000000000001e-05, |
| "loss": 0.6376, |
| "mean_token_accuracy": 0.8255084995180368, |
| "num_tokens": 20613266.0, |
| "step": 720 |
| }, |
| { |
| "entropy": 0.6476276513189078, |
| "epoch": 1.168, |
| "grad_norm": 1.40625, |
| "learning_rate": 1.4580000000000001e-05, |
| "loss": 0.6316, |
| "mean_token_accuracy": 0.8264493472874165, |
| "num_tokens": 20910404.0, |
| "step": 730 |
| }, |
| { |
| "entropy": 0.6569468161091209, |
| "epoch": 1.184, |
| "grad_norm": 1.546875, |
| "learning_rate": 1.478e-05, |
| "loss": 0.6336, |
| "mean_token_accuracy": 0.8247423082590103, |
| "num_tokens": 21200197.0, |
| "step": 740 |
| }, |
| { |
| "entropy": 0.6399336410686374, |
| "epoch": 1.2, |
| "grad_norm": 1.40625, |
| "learning_rate": 1.498e-05, |
| "loss": 0.622, |
| "mean_token_accuracy": 0.8271243523806333, |
| "num_tokens": 21496441.0, |
| "step": 750 |
| }, |
| { |
| "entropy": 0.6582916142418981, |
| "epoch": 1.216, |
| "grad_norm": 1.609375, |
| "learning_rate": 1.5180000000000002e-05, |
| "loss": 0.6353, |
| "mean_token_accuracy": 0.8234689589589834, |
| "num_tokens": 21778030.0, |
| "step": 760 |
| }, |
| { |
| "entropy": 0.637734330072999, |
| "epoch": 1.232, |
| "grad_norm": 1.5703125, |
| "learning_rate": 1.5380000000000002e-05, |
| "loss": 0.6159, |
| "mean_token_accuracy": 0.8288460947573185, |
| "num_tokens": 22060851.0, |
| "step": 770 |
| }, |
| { |
| "entropy": 0.636210915632546, |
| "epoch": 1.248, |
| "grad_norm": 1.5078125, |
| "learning_rate": 1.5580000000000003e-05, |
| "loss": 0.6165, |
| "mean_token_accuracy": 0.82775485701859, |
| "num_tokens": 22348738.0, |
| "step": 780 |
| }, |
| { |
| "entropy": 0.6564677763730288, |
| "epoch": 1.264, |
| "grad_norm": 1.3671875, |
| "learning_rate": 1.578e-05, |
| "loss": 0.6349, |
| "mean_token_accuracy": 0.8237839676439762, |
| "num_tokens": 22635039.0, |
| "step": 790 |
| }, |
| { |
| "entropy": 0.6571361353620887, |
| "epoch": 1.28, |
| "grad_norm": 1.421875, |
| "learning_rate": 1.5980000000000003e-05, |
| "loss": 0.6366, |
| "mean_token_accuracy": 0.8244204126298428, |
| "num_tokens": 22915190.0, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.28, |
| "eval_math_entropy": 0.6791466112136841, |
| "eval_math_loss": 0.669770359992981, |
| "eval_math_mean_token_accuracy": 0.8196616005897522, |
| "eval_math_num_tokens": 22915190.0, |
| "eval_math_runtime": 35.7574, |
| "eval_math_samples_per_second": 13.983, |
| "eval_math_steps_per_second": 3.496, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.28, |
| "eval_physics_entropy": 0.6547831742763519, |
| "eval_physics_loss": 0.650165319442749, |
| "eval_physics_mean_token_accuracy": 0.82009166097641, |
| "eval_physics_num_tokens": 22915190.0, |
| "eval_physics_runtime": 41.4354, |
| "eval_physics_samples_per_second": 12.067, |
| "eval_physics_steps_per_second": 3.017, |
| "step": 800 |
| }, |
| { |
| "entropy": 0.6429510954767466, |
| "epoch": 1.296, |
| "grad_norm": 1.40625, |
| "learning_rate": 1.618e-05, |
| "loss": 0.6172, |
| "mean_token_accuracy": 0.8271825641393662, |
| "num_tokens": 23194947.0, |
| "step": 810 |
| }, |
| { |
| "entropy": 0.6526322312653065, |
| "epoch": 1.312, |
| "grad_norm": 1.5390625, |
| "learning_rate": 1.638e-05, |
| "loss": 0.6414, |
| "mean_token_accuracy": 0.8243110898882151, |
| "num_tokens": 23483496.0, |
| "step": 820 |
| }, |
| { |
| "entropy": 0.6679502466693521, |
| "epoch": 1.328, |
| "grad_norm": 1.6171875, |
| "learning_rate": 1.658e-05, |
| "loss": 0.6441, |
| "mean_token_accuracy": 0.8223706830292941, |
| "num_tokens": 23762589.0, |
| "step": 830 |
| }, |
| { |
| "entropy": 0.6403571154922247, |
| "epoch": 1.3439999999999999, |
| "grad_norm": 1.4921875, |
| "learning_rate": 1.6780000000000002e-05, |
| "loss": 0.6222, |
| "mean_token_accuracy": 0.8268204662948847, |
| "num_tokens": 24041497.0, |
| "step": 840 |
| }, |
| { |
| "entropy": 0.6487369466573, |
| "epoch": 1.3599999999999999, |
| "grad_norm": 1.203125, |
| "learning_rate": 1.698e-05, |
| "loss": 0.6282, |
| "mean_token_accuracy": 0.8273422639817, |
| "num_tokens": 24330591.0, |
| "step": 850 |
| }, |
| { |
| "entropy": 0.688577077165246, |
| "epoch": 1.376, |
| "grad_norm": 1.4453125, |
| "learning_rate": 1.718e-05, |
| "loss": 0.6651, |
| "mean_token_accuracy": 0.8164406325668097, |
| "num_tokens": 24611420.0, |
| "step": 860 |
| }, |
| { |
| "entropy": 0.643337251059711, |
| "epoch": 1.392, |
| "grad_norm": 1.4375, |
| "learning_rate": 1.7380000000000003e-05, |
| "loss": 0.6209, |
| "mean_token_accuracy": 0.825714922696352, |
| "num_tokens": 24894694.0, |
| "step": 870 |
| }, |
| { |
| "entropy": 0.6371515756472945, |
| "epoch": 1.408, |
| "grad_norm": 1.453125, |
| "learning_rate": 1.758e-05, |
| "loss": 0.6156, |
| "mean_token_accuracy": 0.8278428696095943, |
| "num_tokens": 25184946.0, |
| "step": 880 |
| }, |
| { |
| "entropy": 0.6312845427542925, |
| "epoch": 1.424, |
| "grad_norm": 1.25, |
| "learning_rate": 1.7780000000000003e-05, |
| "loss": 0.615, |
| "mean_token_accuracy": 0.8290112767368555, |
| "num_tokens": 25474746.0, |
| "step": 890 |
| }, |
| { |
| "entropy": 0.6522130774334073, |
| "epoch": 1.44, |
| "grad_norm": 1.3671875, |
| "learning_rate": 1.798e-05, |
| "loss": 0.6295, |
| "mean_token_accuracy": 0.8248222030699253, |
| "num_tokens": 25758550.0, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.44, |
| "eval_math_entropy": 0.6579540917873382, |
| "eval_math_loss": 0.6607276797294617, |
| "eval_math_mean_token_accuracy": 0.8207882242202759, |
| "eval_math_num_tokens": 25758550.0, |
| "eval_math_runtime": 35.7442, |
| "eval_math_samples_per_second": 13.988, |
| "eval_math_steps_per_second": 3.497, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.44, |
| "eval_physics_entropy": 0.6424285123348236, |
| "eval_physics_loss": 0.651524543762207, |
| "eval_physics_mean_token_accuracy": 0.8197560949325562, |
| "eval_physics_num_tokens": 25758550.0, |
| "eval_physics_runtime": 41.435, |
| "eval_physics_samples_per_second": 12.067, |
| "eval_physics_steps_per_second": 3.017, |
| "step": 900 |
| }, |
| { |
| "entropy": 0.6359886834397912, |
| "epoch": 1.456, |
| "grad_norm": 1.3203125, |
| "learning_rate": 1.8180000000000002e-05, |
| "loss": 0.6177, |
| "mean_token_accuracy": 0.8268891926854849, |
| "num_tokens": 26039862.0, |
| "step": 910 |
| }, |
| { |
| "entropy": 0.6429944572970271, |
| "epoch": 1.472, |
| "grad_norm": 1.28125, |
| "learning_rate": 1.8380000000000004e-05, |
| "loss": 0.6266, |
| "mean_token_accuracy": 0.8256016962230206, |
| "num_tokens": 26318666.0, |
| "step": 920 |
| }, |
| { |
| "entropy": 0.666933044232428, |
| "epoch": 1.488, |
| "grad_norm": 1.265625, |
| "learning_rate": 1.858e-05, |
| "loss": 0.6455, |
| "mean_token_accuracy": 0.8214054699987173, |
| "num_tokens": 26594867.0, |
| "step": 930 |
| }, |
| { |
| "entropy": 0.6266062622889876, |
| "epoch": 1.504, |
| "grad_norm": 1.5, |
| "learning_rate": 1.878e-05, |
| "loss": 0.6092, |
| "mean_token_accuracy": 0.8303067740052938, |
| "num_tokens": 26887233.0, |
| "step": 940 |
| }, |
| { |
| "entropy": 0.659474827721715, |
| "epoch": 1.52, |
| "grad_norm": 1.140625, |
| "learning_rate": 1.898e-05, |
| "loss": 0.6344, |
| "mean_token_accuracy": 0.8236033659428358, |
| "num_tokens": 27177448.0, |
| "step": 950 |
| }, |
| { |
| "entropy": 0.6371477670967579, |
| "epoch": 1.536, |
| "grad_norm": 1.1484375, |
| "learning_rate": 1.918e-05, |
| "loss": 0.6245, |
| "mean_token_accuracy": 0.8263927459716797, |
| "num_tokens": 27467857.0, |
| "step": 960 |
| }, |
| { |
| "entropy": 0.6363119596615434, |
| "epoch": 1.552, |
| "grad_norm": 1.265625, |
| "learning_rate": 1.938e-05, |
| "loss": 0.6172, |
| "mean_token_accuracy": 0.8277775753289461, |
| "num_tokens": 27755138.0, |
| "step": 970 |
| }, |
| { |
| "entropy": 0.625205036252737, |
| "epoch": 1.568, |
| "grad_norm": 1.15625, |
| "learning_rate": 1.9580000000000002e-05, |
| "loss": 0.6047, |
| "mean_token_accuracy": 0.8324479281902313, |
| "num_tokens": 28049363.0, |
| "step": 980 |
| }, |
| { |
| "entropy": 0.6341174798086285, |
| "epoch": 1.584, |
| "grad_norm": 1.375, |
| "learning_rate": 1.978e-05, |
| "loss": 0.6153, |
| "mean_token_accuracy": 0.8282931160181761, |
| "num_tokens": 28337890.0, |
| "step": 990 |
| }, |
| { |
| "entropy": 0.6248919799923897, |
| "epoch": 1.6, |
| "grad_norm": 1.3359375, |
| "learning_rate": 1.9980000000000002e-05, |
| "loss": 0.6112, |
| "mean_token_accuracy": 0.827129240334034, |
| "num_tokens": 28626068.0, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.6, |
| "eval_math_entropy": 0.6310200927257538, |
| "eval_math_loss": 0.651292085647583, |
| "eval_math_mean_token_accuracy": 0.8225455293655396, |
| "eval_math_num_tokens": 28626068.0, |
| "eval_math_runtime": 35.7544, |
| "eval_math_samples_per_second": 13.984, |
| "eval_math_steps_per_second": 3.496, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.6, |
| "eval_physics_entropy": 0.6265229551792145, |
| "eval_physics_loss": 0.6552274227142334, |
| "eval_physics_mean_token_accuracy": 0.8185877275466918, |
| "eval_physics_num_tokens": 28626068.0, |
| "eval_physics_runtime": 41.1356, |
| "eval_physics_samples_per_second": 12.155, |
| "eval_physics_steps_per_second": 3.039, |
| "step": 1000 |
| }, |
| { |
| "entropy": 0.6397603360936046, |
| "epoch": 1.616, |
| "grad_norm": 1.25, |
| "learning_rate": 1.9980000000000002e-05, |
| "loss": 0.6266, |
| "mean_token_accuracy": 0.8261015623807907, |
| "num_tokens": 28913208.0, |
| "step": 1010 |
| }, |
| { |
| "entropy": 0.6517996810376644, |
| "epoch": 1.6320000000000001, |
| "grad_norm": 1.265625, |
| "learning_rate": 1.995777777777778e-05, |
| "loss": 0.6284, |
| "mean_token_accuracy": 0.8228654317557812, |
| "num_tokens": 29188917.0, |
| "step": 1020 |
| }, |
| { |
| "entropy": 0.6239277983084321, |
| "epoch": 1.6480000000000001, |
| "grad_norm": 1.2421875, |
| "learning_rate": 1.9935555555555557e-05, |
| "loss": 0.6068, |
| "mean_token_accuracy": 0.8318834796547889, |
| "num_tokens": 29477292.0, |
| "step": 1030 |
| }, |
| { |
| "entropy": 0.6378339532762766, |
| "epoch": 1.6640000000000001, |
| "grad_norm": 1.2109375, |
| "learning_rate": 1.9913333333333335e-05, |
| "loss": 0.6129, |
| "mean_token_accuracy": 0.8274775009602309, |
| "num_tokens": 29772480.0, |
| "step": 1040 |
| }, |
| { |
| "entropy": 0.64110322073102, |
| "epoch": 1.6800000000000002, |
| "grad_norm": 1.15625, |
| "learning_rate": 1.9891111111111112e-05, |
| "loss": 0.6244, |
| "mean_token_accuracy": 0.8272217728197575, |
| "num_tokens": 30061007.0, |
| "step": 1050 |
| }, |
| { |
| "entropy": 0.6449864653870463, |
| "epoch": 1.696, |
| "grad_norm": 1.1875, |
| "learning_rate": 1.986888888888889e-05, |
| "loss": 0.6243, |
| "mean_token_accuracy": 0.8258783508092165, |
| "num_tokens": 30344415.0, |
| "step": 1060 |
| }, |
| { |
| "entropy": 0.6212725475430488, |
| "epoch": 1.712, |
| "grad_norm": 1.1640625, |
| "learning_rate": 1.9846666666666668e-05, |
| "loss": 0.6076, |
| "mean_token_accuracy": 0.8308158092200756, |
| "num_tokens": 30640891.0, |
| "step": 1070 |
| }, |
| { |
| "entropy": 0.618477713689208, |
| "epoch": 1.728, |
| "grad_norm": 1.2109375, |
| "learning_rate": 1.9824444444444445e-05, |
| "loss": 0.5981, |
| "mean_token_accuracy": 0.8317991364747286, |
| "num_tokens": 30921295.0, |
| "step": 1080 |
| }, |
| { |
| "entropy": 0.6253316847607493, |
| "epoch": 1.744, |
| "grad_norm": 1.3125, |
| "learning_rate": 1.9802222222222226e-05, |
| "loss": 0.608, |
| "mean_token_accuracy": 0.8302719250321389, |
| "num_tokens": 31206375.0, |
| "step": 1090 |
| }, |
| { |
| "entropy": 0.6203237606212497, |
| "epoch": 1.76, |
| "grad_norm": 1.140625, |
| "learning_rate": 1.978e-05, |
| "loss": 0.6016, |
| "mean_token_accuracy": 0.829212823137641, |
| "num_tokens": 31491856.0, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.76, |
| "eval_math_entropy": 0.6470205163955689, |
| "eval_math_loss": 0.6425853371620178, |
| "eval_math_mean_token_accuracy": 0.8245493578910827, |
| "eval_math_num_tokens": 31491856.0, |
| "eval_math_runtime": 35.7795, |
| "eval_math_samples_per_second": 13.974, |
| "eval_math_steps_per_second": 3.494, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.76, |
| "eval_physics_entropy": 0.6618543210029602, |
| "eval_physics_loss": 0.6551723480224609, |
| "eval_physics_mean_token_accuracy": 0.8183454179763794, |
| "eval_physics_num_tokens": 31491856.0, |
| "eval_physics_runtime": 41.448, |
| "eval_physics_samples_per_second": 12.063, |
| "eval_physics_steps_per_second": 3.016, |
| "step": 1100 |
| }, |
| { |
| "entropy": 0.6285459484905005, |
| "epoch": 1.776, |
| "grad_norm": 1.203125, |
| "learning_rate": 1.975777777777778e-05, |
| "loss": 0.608, |
| "mean_token_accuracy": 0.8288080658763647, |
| "num_tokens": 31782829.0, |
| "step": 1110 |
| }, |
| { |
| "entropy": 0.6362737115472555, |
| "epoch": 1.792, |
| "grad_norm": 1.1484375, |
| "learning_rate": 1.9735555555555556e-05, |
| "loss": 0.6168, |
| "mean_token_accuracy": 0.8268269985914231, |
| "num_tokens": 32070187.0, |
| "step": 1120 |
| }, |
| { |
| "entropy": 0.6320221189409494, |
| "epoch": 1.808, |
| "grad_norm": 1.203125, |
| "learning_rate": 1.9713333333333337e-05, |
| "loss": 0.6148, |
| "mean_token_accuracy": 0.8273419640958309, |
| "num_tokens": 32355072.0, |
| "step": 1130 |
| }, |
| { |
| "entropy": 0.6012085109949112, |
| "epoch": 1.8239999999999998, |
| "grad_norm": 1.0859375, |
| "learning_rate": 1.969111111111111e-05, |
| "loss": 0.5802, |
| "mean_token_accuracy": 0.8352209012955427, |
| "num_tokens": 32634013.0, |
| "step": 1140 |
| }, |
| { |
| "entropy": 0.6171128546819091, |
| "epoch": 1.8399999999999999, |
| "grad_norm": 1.328125, |
| "learning_rate": 1.9668888888888892e-05, |
| "loss": 0.6044, |
| "mean_token_accuracy": 0.8311630170792341, |
| "num_tokens": 32917222.0, |
| "step": 1150 |
| }, |
| { |
| "entropy": 0.6315351065248251, |
| "epoch": 1.8559999999999999, |
| "grad_norm": 1.171875, |
| "learning_rate": 1.9646666666666666e-05, |
| "loss": 0.6066, |
| "mean_token_accuracy": 0.8292891420423985, |
| "num_tokens": 33207147.0, |
| "step": 1160 |
| }, |
| { |
| "entropy": 0.6153346884995698, |
| "epoch": 1.8719999999999999, |
| "grad_norm": 1.234375, |
| "learning_rate": 1.9624444444444447e-05, |
| "loss": 0.5954, |
| "mean_token_accuracy": 0.8320576064288616, |
| "num_tokens": 33496572.0, |
| "step": 1170 |
| }, |
| { |
| "entropy": 0.624875632673502, |
| "epoch": 1.888, |
| "grad_norm": 1.203125, |
| "learning_rate": 1.9602222222222225e-05, |
| "loss": 0.6059, |
| "mean_token_accuracy": 0.8299812860786915, |
| "num_tokens": 33775788.0, |
| "step": 1180 |
| }, |
| { |
| "entropy": 0.6256145890802145, |
| "epoch": 1.904, |
| "grad_norm": 1.1640625, |
| "learning_rate": 1.9580000000000002e-05, |
| "loss": 0.6084, |
| "mean_token_accuracy": 0.8293462209403515, |
| "num_tokens": 34052322.0, |
| "step": 1190 |
| }, |
| { |
| "entropy": 0.6057112148031593, |
| "epoch": 1.92, |
| "grad_norm": 1.109375, |
| "learning_rate": 1.955777777777778e-05, |
| "loss": 0.5888, |
| "mean_token_accuracy": 0.8329556312412023, |
| "num_tokens": 34331468.0, |
| "step": 1200 |
| }, |
| { |
| "epoch": 1.92, |
| "eval_math_entropy": 0.6444816522598267, |
| "eval_math_loss": 0.6337655782699585, |
| "eval_math_mean_token_accuracy": 0.8254525041580201, |
| "eval_math_num_tokens": 34331468.0, |
| "eval_math_runtime": 35.7615, |
| "eval_math_samples_per_second": 13.982, |
| "eval_math_steps_per_second": 3.495, |
| "step": 1200 |
| }, |
| { |
| "epoch": 1.92, |
| "eval_physics_entropy": 0.6525136830806733, |
| "eval_physics_loss": 0.6549710035324097, |
| "eval_physics_mean_token_accuracy": 0.8188423981666565, |
| "eval_physics_num_tokens": 34331468.0, |
| "eval_physics_runtime": 41.437, |
| "eval_physics_samples_per_second": 12.067, |
| "eval_physics_steps_per_second": 3.017, |
| "step": 1200 |
| }, |
| { |
| "entropy": 0.622246976941824, |
| "epoch": 1.936, |
| "grad_norm": 1.2109375, |
| "learning_rate": 1.9535555555555557e-05, |
| "loss": 0.6052, |
| "mean_token_accuracy": 0.8293883047997952, |
| "num_tokens": 34612743.0, |
| "step": 1210 |
| }, |
| { |
| "entropy": 0.6145633114501834, |
| "epoch": 1.952, |
| "grad_norm": 1.078125, |
| "learning_rate": 1.9513333333333335e-05, |
| "loss": 0.5899, |
| "mean_token_accuracy": 0.8315575309097767, |
| "num_tokens": 34891115.0, |
| "step": 1220 |
| }, |
| { |
| "entropy": 0.6348631424829364, |
| "epoch": 1.968, |
| "grad_norm": 3.625, |
| "learning_rate": 1.9491111111111113e-05, |
| "loss": 0.6247, |
| "mean_token_accuracy": 0.8266950808465481, |
| "num_tokens": 35179505.0, |
| "step": 1230 |
| }, |
| { |
| "entropy": 0.6054144280031324, |
| "epoch": 1.984, |
| "grad_norm": 1.0703125, |
| "learning_rate": 1.946888888888889e-05, |
| "loss": 0.5878, |
| "mean_token_accuracy": 0.8339846853166819, |
| "num_tokens": 35472596.0, |
| "step": 1240 |
| }, |
| { |
| "entropy": 0.614653673581779, |
| "epoch": 2.0, |
| "grad_norm": 1.109375, |
| "learning_rate": 1.9446666666666668e-05, |
| "loss": 0.5966, |
| "mean_token_accuracy": 0.8319407891482115, |
| "num_tokens": 35772848.0, |
| "step": 1250 |
| }, |
| { |
| "entropy": 0.5460906764492393, |
| "epoch": 2.016, |
| "grad_norm": 1.3515625, |
| "learning_rate": 1.9424444444444446e-05, |
| "loss": 0.5207, |
| "mean_token_accuracy": 0.844809464737773, |
| "num_tokens": 36060330.0, |
| "step": 1260 |
| }, |
| { |
| "entropy": 0.5267174040898681, |
| "epoch": 2.032, |
| "grad_norm": 1.3671875, |
| "learning_rate": 1.9402222222222223e-05, |
| "loss": 0.5098, |
| "mean_token_accuracy": 0.8475256130099297, |
| "num_tokens": 36343894.0, |
| "step": 1270 |
| }, |
| { |
| "entropy": 0.5175346418283879, |
| "epoch": 2.048, |
| "grad_norm": 1.3671875, |
| "learning_rate": 1.938e-05, |
| "loss": 0.501, |
| "mean_token_accuracy": 0.8504213903099298, |
| "num_tokens": 36636410.0, |
| "step": 1280 |
| }, |
| { |
| "entropy": 0.5137943865731358, |
| "epoch": 2.064, |
| "grad_norm": 1.28125, |
| "learning_rate": 1.935777777777778e-05, |
| "loss": 0.4962, |
| "mean_token_accuracy": 0.8507547531276941, |
| "num_tokens": 36925382.0, |
| "step": 1290 |
| }, |
| { |
| "entropy": 0.5208146266639233, |
| "epoch": 2.08, |
| "grad_norm": 1.375, |
| "learning_rate": 1.9335555555555556e-05, |
| "loss": 0.4954, |
| "mean_token_accuracy": 0.849713422730565, |
| "num_tokens": 37208672.0, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.08, |
| "eval_math_entropy": 0.5398842208385467, |
| "eval_math_loss": 0.6511635780334473, |
| "eval_math_mean_token_accuracy": 0.8244616360664367, |
| "eval_math_num_tokens": 37208672.0, |
| "eval_math_runtime": 35.7792, |
| "eval_math_samples_per_second": 13.975, |
| "eval_math_steps_per_second": 3.494, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.08, |
| "eval_physics_entropy": 0.5216713752746582, |
| "eval_physics_loss": 0.6896871328353882, |
| "eval_physics_mean_token_accuracy": 0.8157849154472351, |
| "eval_physics_num_tokens": 37208672.0, |
| "eval_physics_runtime": 41.4274, |
| "eval_physics_samples_per_second": 12.069, |
| "eval_physics_steps_per_second": 3.017, |
| "step": 1300 |
| }, |
| { |
| "entropy": 0.5262768002226949, |
| "epoch": 2.096, |
| "grad_norm": 1.359375, |
| "learning_rate": 1.9313333333333334e-05, |
| "loss": 0.5071, |
| "mean_token_accuracy": 0.8485719960182905, |
| "num_tokens": 37491954.0, |
| "step": 1310 |
| }, |
| { |
| "entropy": 0.5144134562462568, |
| "epoch": 2.112, |
| "grad_norm": 1.390625, |
| "learning_rate": 1.9291111111111115e-05, |
| "loss": 0.4932, |
| "mean_token_accuracy": 0.8491108123213053, |
| "num_tokens": 37787033.0, |
| "step": 1320 |
| }, |
| { |
| "entropy": 0.5125698197633028, |
| "epoch": 2.128, |
| "grad_norm": 1.390625, |
| "learning_rate": 1.926888888888889e-05, |
| "loss": 0.4936, |
| "mean_token_accuracy": 0.8515845112502575, |
| "num_tokens": 38068955.0, |
| "step": 1330 |
| }, |
| { |
| "entropy": 0.5147779244929552, |
| "epoch": 2.144, |
| "grad_norm": 1.359375, |
| "learning_rate": 1.924666666666667e-05, |
| "loss": 0.4998, |
| "mean_token_accuracy": 0.8497466515749693, |
| "num_tokens": 38359914.0, |
| "step": 1340 |
| }, |
| { |
| "entropy": 0.5150290129706263, |
| "epoch": 2.16, |
| "grad_norm": 1.28125, |
| "learning_rate": 1.9224444444444444e-05, |
| "loss": 0.4976, |
| "mean_token_accuracy": 0.8506554007530213, |
| "num_tokens": 38649684.0, |
| "step": 1350 |
| }, |
| { |
| "entropy": 0.49578714426606896, |
| "epoch": 2.176, |
| "grad_norm": 1.375, |
| "learning_rate": 1.9202222222222225e-05, |
| "loss": 0.4768, |
| "mean_token_accuracy": 0.8545470163226128, |
| "num_tokens": 38938155.0, |
| "step": 1360 |
| }, |
| { |
| "entropy": 0.5279102198779583, |
| "epoch": 2.192, |
| "grad_norm": 1.2578125, |
| "learning_rate": 1.918e-05, |
| "loss": 0.5109, |
| "mean_token_accuracy": 0.8461268071085215, |
| "num_tokens": 39228800.0, |
| "step": 1370 |
| }, |
| { |
| "entropy": 0.520401804894209, |
| "epoch": 2.208, |
| "grad_norm": 1.4296875, |
| "learning_rate": 1.915777777777778e-05, |
| "loss": 0.5095, |
| "mean_token_accuracy": 0.8481683298945427, |
| "num_tokens": 39513118.0, |
| "step": 1380 |
| }, |
| { |
| "entropy": 0.517220887541771, |
| "epoch": 2.224, |
| "grad_norm": 1.3828125, |
| "learning_rate": 1.9135555555555555e-05, |
| "loss": 0.4947, |
| "mean_token_accuracy": 0.8505345996469259, |
| "num_tokens": 39804212.0, |
| "step": 1390 |
| }, |
| { |
| "entropy": 0.5203588601201773, |
| "epoch": 2.24, |
| "grad_norm": 1.359375, |
| "learning_rate": 1.9113333333333336e-05, |
| "loss": 0.5071, |
| "mean_token_accuracy": 0.8482481308281422, |
| "num_tokens": 40092012.0, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.24, |
| "eval_math_entropy": 0.5498721191883087, |
| "eval_math_loss": 0.6494505405426025, |
| "eval_math_mean_token_accuracy": 0.8248468618392945, |
| "eval_math_num_tokens": 40092012.0, |
| "eval_math_runtime": 35.7606, |
| "eval_math_samples_per_second": 13.982, |
| "eval_math_steps_per_second": 3.495, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.24, |
| "eval_physics_entropy": 0.5240244953632355, |
| "eval_physics_loss": 0.6975257992744446, |
| "eval_physics_mean_token_accuracy": 0.8148801970481873, |
| "eval_physics_num_tokens": 40092012.0, |
| "eval_physics_runtime": 41.7753, |
| "eval_physics_samples_per_second": 11.969, |
| "eval_physics_steps_per_second": 2.992, |
| "step": 1400 |
| }, |
| { |
| "entropy": 0.5205017613247037, |
| "epoch": 2.2560000000000002, |
| "grad_norm": 1.3671875, |
| "learning_rate": 1.9091111111111113e-05, |
| "loss": 0.5042, |
| "mean_token_accuracy": 0.8483030397444964, |
| "num_tokens": 40380069.0, |
| "step": 1410 |
| }, |
| { |
| "entropy": 0.5286959325894713, |
| "epoch": 2.2720000000000002, |
| "grad_norm": 1.5625, |
| "learning_rate": 1.906888888888889e-05, |
| "loss": 0.5089, |
| "mean_token_accuracy": 0.8469363551586866, |
| "num_tokens": 40655903.0, |
| "step": 1420 |
| }, |
| { |
| "entropy": 0.5245801825076342, |
| "epoch": 2.288, |
| "grad_norm": 1.421875, |
| "learning_rate": 1.904666666666667e-05, |
| "loss": 0.505, |
| "mean_token_accuracy": 0.8467433623969555, |
| "num_tokens": 40944791.0, |
| "step": 1430 |
| }, |
| { |
| "entropy": 0.5089570162817836, |
| "epoch": 2.304, |
| "grad_norm": 1.453125, |
| "learning_rate": 1.9024444444444446e-05, |
| "loss": 0.4927, |
| "mean_token_accuracy": 0.851346742734313, |
| "num_tokens": 41227146.0, |
| "step": 1440 |
| }, |
| { |
| "entropy": 0.5277796785347164, |
| "epoch": 2.32, |
| "grad_norm": 1.4296875, |
| "learning_rate": 1.9002222222222224e-05, |
| "loss": 0.5143, |
| "mean_token_accuracy": 0.8476031333208084, |
| "num_tokens": 41509322.0, |
| "step": 1450 |
| }, |
| { |
| "entropy": 0.5237356763333082, |
| "epoch": 2.336, |
| "grad_norm": 1.3984375, |
| "learning_rate": 1.898e-05, |
| "loss": 0.5015, |
| "mean_token_accuracy": 0.8484678287059069, |
| "num_tokens": 41792799.0, |
| "step": 1460 |
| }, |
| { |
| "entropy": 0.5347706735134125, |
| "epoch": 2.352, |
| "grad_norm": 1.3828125, |
| "learning_rate": 1.895777777777778e-05, |
| "loss": 0.5169, |
| "mean_token_accuracy": 0.8446083668619394, |
| "num_tokens": 42083080.0, |
| "step": 1470 |
| }, |
| { |
| "entropy": 0.5238756524398923, |
| "epoch": 2.368, |
| "grad_norm": 1.40625, |
| "learning_rate": 1.8935555555555556e-05, |
| "loss": 0.5074, |
| "mean_token_accuracy": 0.8478495571762323, |
| "num_tokens": 42368442.0, |
| "step": 1480 |
| }, |
| { |
| "entropy": 0.5398916637524962, |
| "epoch": 2.384, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.8913333333333334e-05, |
| "loss": 0.5247, |
| "mean_token_accuracy": 0.8446283016353846, |
| "num_tokens": 42663926.0, |
| "step": 1490 |
| }, |
| { |
| "entropy": 0.5313681181520223, |
| "epoch": 2.4, |
| "grad_norm": 1.6015625, |
| "learning_rate": 1.8891111111111115e-05, |
| "loss": 0.5162, |
| "mean_token_accuracy": 0.8469122361391783, |
| "num_tokens": 42946351.0, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.4, |
| "eval_math_entropy": 0.5470261330604553, |
| "eval_math_loss": 0.6441261172294617, |
| "eval_math_mean_token_accuracy": 0.8258598875999451, |
| "eval_math_num_tokens": 42946351.0, |
| "eval_math_runtime": 35.7282, |
| "eval_math_samples_per_second": 13.995, |
| "eval_math_steps_per_second": 3.499, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.4, |
| "eval_physics_entropy": 0.5238613975048065, |
| "eval_physics_loss": 0.6977487802505493, |
| "eval_physics_mean_token_accuracy": 0.81443390417099, |
| "eval_physics_num_tokens": 42946351.0, |
| "eval_physics_runtime": 41.457, |
| "eval_physics_samples_per_second": 12.061, |
| "eval_physics_steps_per_second": 3.015, |
| "step": 1500 |
| }, |
| { |
| "entropy": 0.5377832099795341, |
| "epoch": 2.416, |
| "grad_norm": 1.4609375, |
| "learning_rate": 1.886888888888889e-05, |
| "loss": 0.5215, |
| "mean_token_accuracy": 0.8441088248044253, |
| "num_tokens": 43227782.0, |
| "step": 1510 |
| }, |
| { |
| "entropy": 0.5219327821396291, |
| "epoch": 2.432, |
| "grad_norm": 1.3671875, |
| "learning_rate": 1.884666666666667e-05, |
| "loss": 0.5065, |
| "mean_token_accuracy": 0.8481958381831646, |
| "num_tokens": 43511939.0, |
| "step": 1520 |
| }, |
| { |
| "entropy": 0.5146825276315212, |
| "epoch": 2.448, |
| "grad_norm": 1.53125, |
| "learning_rate": 1.8824444444444445e-05, |
| "loss": 0.4976, |
| "mean_token_accuracy": 0.8503648530691862, |
| "num_tokens": 43811005.0, |
| "step": 1530 |
| }, |
| { |
| "entropy": 0.5279206298291683, |
| "epoch": 2.464, |
| "grad_norm": 1.484375, |
| "learning_rate": 1.8802222222222226e-05, |
| "loss": 0.5124, |
| "mean_token_accuracy": 0.8466742359101772, |
| "num_tokens": 44106223.0, |
| "step": 1540 |
| }, |
| { |
| "entropy": 0.548205547593534, |
| "epoch": 2.48, |
| "grad_norm": 1.515625, |
| "learning_rate": 1.878e-05, |
| "loss": 0.5272, |
| "mean_token_accuracy": 0.8421384803950787, |
| "num_tokens": 44379095.0, |
| "step": 1550 |
| }, |
| { |
| "entropy": 0.5311242727562785, |
| "epoch": 2.496, |
| "grad_norm": 1.3515625, |
| "learning_rate": 1.875777777777778e-05, |
| "loss": 0.5146, |
| "mean_token_accuracy": 0.8455508589744568, |
| "num_tokens": 44670316.0, |
| "step": 1560 |
| }, |
| { |
| "entropy": 0.5254273321479559, |
| "epoch": 2.512, |
| "grad_norm": 1.4765625, |
| "learning_rate": 1.873555555555556e-05, |
| "loss": 0.5098, |
| "mean_token_accuracy": 0.8476663187146187, |
| "num_tokens": 44952877.0, |
| "step": 1570 |
| }, |
| { |
| "entropy": 0.5215749599039554, |
| "epoch": 2.528, |
| "grad_norm": 1.390625, |
| "learning_rate": 1.8713333333333336e-05, |
| "loss": 0.5036, |
| "mean_token_accuracy": 0.8488380990922451, |
| "num_tokens": 45237653.0, |
| "step": 1580 |
| }, |
| { |
| "entropy": 0.5184456476941705, |
| "epoch": 2.544, |
| "grad_norm": 1.46875, |
| "learning_rate": 1.8691111111111114e-05, |
| "loss": 0.5051, |
| "mean_token_accuracy": 0.8494234204292297, |
| "num_tokens": 45524866.0, |
| "step": 1590 |
| }, |
| { |
| "entropy": 0.5188256312161684, |
| "epoch": 2.56, |
| "grad_norm": 1.3359375, |
| "learning_rate": 1.866888888888889e-05, |
| "loss": 0.4994, |
| "mean_token_accuracy": 0.8480058163404465, |
| "num_tokens": 45803783.0, |
| "step": 1600 |
| }, |
| { |
| "epoch": 2.56, |
| "eval_math_entropy": 0.5415253615379334, |
| "eval_math_loss": 0.6402944326400757, |
| "eval_math_mean_token_accuracy": 0.8268161854743957, |
| "eval_math_num_tokens": 45803783.0, |
| "eval_math_runtime": 35.8351, |
| "eval_math_samples_per_second": 13.953, |
| "eval_math_steps_per_second": 3.488, |
| "step": 1600 |
| }, |
| { |
| "epoch": 2.56, |
| "eval_physics_entropy": 0.5146132915019989, |
| "eval_physics_loss": 0.7024555802345276, |
| "eval_physics_mean_token_accuracy": 0.8143429985046386, |
| "eval_physics_num_tokens": 45803783.0, |
| "eval_physics_runtime": 41.4942, |
| "eval_physics_samples_per_second": 12.05, |
| "eval_physics_steps_per_second": 3.012, |
| "step": 1600 |
| }, |
| { |
| "entropy": 0.5206789819523692, |
| "epoch": 2.576, |
| "grad_norm": 1.40625, |
| "learning_rate": 1.864666666666667e-05, |
| "loss": 0.5037, |
| "mean_token_accuracy": 0.8492428727447987, |
| "num_tokens": 46090713.0, |
| "step": 1610 |
| }, |
| { |
| "entropy": 0.5139110781252384, |
| "epoch": 2.592, |
| "grad_norm": 1.25, |
| "learning_rate": 1.8624444444444446e-05, |
| "loss": 0.4983, |
| "mean_token_accuracy": 0.8500846106559038, |
| "num_tokens": 46384084.0, |
| "step": 1620 |
| }, |
| { |
| "entropy": 0.5300323162227869, |
| "epoch": 2.608, |
| "grad_norm": 1.59375, |
| "learning_rate": 1.8602222222222224e-05, |
| "loss": 0.5166, |
| "mean_token_accuracy": 0.8454120934009552, |
| "num_tokens": 46667552.0, |
| "step": 1630 |
| }, |
| { |
| "entropy": 0.498782429471612, |
| "epoch": 2.624, |
| "grad_norm": 1.421875, |
| "learning_rate": 1.858e-05, |
| "loss": 0.4798, |
| "mean_token_accuracy": 0.8554852791130543, |
| "num_tokens": 46961124.0, |
| "step": 1640 |
| }, |
| { |
| "entropy": 0.5273189852014184, |
| "epoch": 2.64, |
| "grad_norm": 1.46875, |
| "learning_rate": 1.855777777777778e-05, |
| "loss": 0.5138, |
| "mean_token_accuracy": 0.8468208122998476, |
| "num_tokens": 47245326.0, |
| "step": 1650 |
| }, |
| { |
| "entropy": 0.5071797709912061, |
| "epoch": 2.656, |
| "grad_norm": 1.5, |
| "learning_rate": 1.8535555555555557e-05, |
| "loss": 0.4904, |
| "mean_token_accuracy": 0.8513832915574312, |
| "num_tokens": 47527347.0, |
| "step": 1660 |
| }, |
| { |
| "entropy": 0.5276256233453751, |
| "epoch": 2.672, |
| "grad_norm": 1.6484375, |
| "learning_rate": 1.8513333333333335e-05, |
| "loss": 0.5105, |
| "mean_token_accuracy": 0.8466553438454867, |
| "num_tokens": 47808167.0, |
| "step": 1670 |
| }, |
| { |
| "entropy": 0.5247279232367873, |
| "epoch": 2.6879999999999997, |
| "grad_norm": 1.390625, |
| "learning_rate": 1.8491111111111112e-05, |
| "loss": 0.5049, |
| "mean_token_accuracy": 0.8479447506368161, |
| "num_tokens": 48092035.0, |
| "step": 1680 |
| }, |
| { |
| "entropy": 0.5260777238756418, |
| "epoch": 2.7039999999999997, |
| "grad_norm": 1.5234375, |
| "learning_rate": 1.846888888888889e-05, |
| "loss": 0.5153, |
| "mean_token_accuracy": 0.8464129194617271, |
| "num_tokens": 48386231.0, |
| "step": 1690 |
| }, |
| { |
| "entropy": 0.5193733531981707, |
| "epoch": 2.7199999999999998, |
| "grad_norm": 1.4609375, |
| "learning_rate": 1.8446666666666667e-05, |
| "loss": 0.5044, |
| "mean_token_accuracy": 0.8482546083629131, |
| "num_tokens": 48663262.0, |
| "step": 1700 |
| }, |
| { |
| "epoch": 2.7199999999999998, |
| "eval_math_entropy": 0.555855593919754, |
| "eval_math_loss": 0.6335127949714661, |
| "eval_math_mean_token_accuracy": 0.8283462662696839, |
| "eval_math_num_tokens": 48663262.0, |
| "eval_math_runtime": 35.7836, |
| "eval_math_samples_per_second": 13.973, |
| "eval_math_steps_per_second": 3.493, |
| "step": 1700 |
| }, |
| { |
| "epoch": 2.7199999999999998, |
| "eval_physics_entropy": 0.5213737454414368, |
| "eval_physics_loss": 0.6999514698982239, |
| "eval_physics_mean_token_accuracy": 0.8152164473533631, |
| "eval_physics_num_tokens": 48663262.0, |
| "eval_physics_runtime": 41.4953, |
| "eval_physics_samples_per_second": 12.05, |
| "eval_physics_steps_per_second": 3.012, |
| "step": 1700 |
| }, |
| { |
| "entropy": 0.5167077284306287, |
| "epoch": 2.7359999999999998, |
| "grad_norm": 1.5546875, |
| "learning_rate": 1.842444444444445e-05, |
| "loss": 0.5014, |
| "mean_token_accuracy": 0.8498601946979761, |
| "num_tokens": 48949456.0, |
| "step": 1710 |
| }, |
| { |
| "entropy": 0.522574995085597, |
| "epoch": 2.752, |
| "grad_norm": 1.4921875, |
| "learning_rate": 1.8402222222222223e-05, |
| "loss": 0.5015, |
| "mean_token_accuracy": 0.8494253471493721, |
| "num_tokens": 49235751.0, |
| "step": 1720 |
| }, |
| { |
| "entropy": 0.5236585384234786, |
| "epoch": 2.768, |
| "grad_norm": 1.3515625, |
| "learning_rate": 1.8380000000000004e-05, |
| "loss": 0.5081, |
| "mean_token_accuracy": 0.8480564054101706, |
| "num_tokens": 49524628.0, |
| "step": 1730 |
| }, |
| { |
| "entropy": 0.536912607960403, |
| "epoch": 2.784, |
| "grad_norm": 1.484375, |
| "learning_rate": 1.8357777777777778e-05, |
| "loss": 0.5166, |
| "mean_token_accuracy": 0.8444275613874197, |
| "num_tokens": 49799604.0, |
| "step": 1740 |
| }, |
| { |
| "entropy": 0.5194279940798878, |
| "epoch": 2.8, |
| "grad_norm": 1.234375, |
| "learning_rate": 1.833555555555556e-05, |
| "loss": 0.5044, |
| "mean_token_accuracy": 0.8491090513765812, |
| "num_tokens": 50079307.0, |
| "step": 1750 |
| }, |
| { |
| "entropy": 0.5226334543898702, |
| "epoch": 2.816, |
| "grad_norm": 1.46875, |
| "learning_rate": 1.8313333333333333e-05, |
| "loss": 0.5013, |
| "mean_token_accuracy": 0.8487468950450421, |
| "num_tokens": 50372216.0, |
| "step": 1760 |
| }, |
| { |
| "entropy": 0.5100100792944431, |
| "epoch": 2.832, |
| "grad_norm": 1.5390625, |
| "learning_rate": 1.8291111111111114e-05, |
| "loss": 0.4977, |
| "mean_token_accuracy": 0.8502487774938345, |
| "num_tokens": 50657211.0, |
| "step": 1770 |
| }, |
| { |
| "entropy": 0.5339175367727875, |
| "epoch": 2.848, |
| "grad_norm": 1.3828125, |
| "learning_rate": 1.8268888888888888e-05, |
| "loss": 0.5176, |
| "mean_token_accuracy": 0.8462332025170326, |
| "num_tokens": 50946503.0, |
| "step": 1780 |
| }, |
| { |
| "entropy": 0.5188839716836811, |
| "epoch": 2.864, |
| "grad_norm": 1.5234375, |
| "learning_rate": 1.824666666666667e-05, |
| "loss": 0.5046, |
| "mean_token_accuracy": 0.8482739225029945, |
| "num_tokens": 51229298.0, |
| "step": 1790 |
| }, |
| { |
| "entropy": 0.52184556145221, |
| "epoch": 2.88, |
| "grad_norm": 1.4609375, |
| "learning_rate": 1.8224444444444447e-05, |
| "loss": 0.5, |
| "mean_token_accuracy": 0.8485745534300804, |
| "num_tokens": 51509622.0, |
| "step": 1800 |
| }, |
| { |
| "epoch": 2.88, |
| "eval_math_entropy": 0.5322999174594879, |
| "eval_math_loss": 0.6332678198814392, |
| "eval_math_mean_token_accuracy": 0.8286352276802063, |
| "eval_math_num_tokens": 51509622.0, |
| "eval_math_runtime": 35.803, |
| "eval_math_samples_per_second": 13.965, |
| "eval_math_steps_per_second": 3.491, |
| "step": 1800 |
| }, |
| { |
| "epoch": 2.88, |
| "eval_physics_entropy": 0.5120384588241577, |
| "eval_physics_loss": 0.7020947337150574, |
| "eval_physics_mean_token_accuracy": 0.8149435911178589, |
| "eval_physics_num_tokens": 51509622.0, |
| "eval_physics_runtime": 41.5069, |
| "eval_physics_samples_per_second": 12.046, |
| "eval_physics_steps_per_second": 3.012, |
| "step": 1800 |
| }, |
| { |
| "entropy": 0.5138611877337098, |
| "epoch": 2.896, |
| "grad_norm": 1.4765625, |
| "learning_rate": 1.8202222222222225e-05, |
| "loss": 0.496, |
| "mean_token_accuracy": 0.8509018927812576, |
| "num_tokens": 51790856.0, |
| "step": 1810 |
| }, |
| { |
| "entropy": 0.5193969309329987, |
| "epoch": 2.912, |
| "grad_norm": 1.484375, |
| "learning_rate": 1.8180000000000002e-05, |
| "loss": 0.5039, |
| "mean_token_accuracy": 0.8491050932556391, |
| "num_tokens": 52074393.0, |
| "step": 1820 |
| }, |
| { |
| "entropy": 0.5282592326402664, |
| "epoch": 2.928, |
| "grad_norm": 1.4140625, |
| "learning_rate": 1.815777777777778e-05, |
| "loss": 0.514, |
| "mean_token_accuracy": 0.8482961695641279, |
| "num_tokens": 52366467.0, |
| "step": 1830 |
| }, |
| { |
| "entropy": 0.512021827697754, |
| "epoch": 2.944, |
| "grad_norm": 1.4375, |
| "learning_rate": 1.8135555555555557e-05, |
| "loss": 0.4995, |
| "mean_token_accuracy": 0.8494243275374174, |
| "num_tokens": 52657863.0, |
| "step": 1840 |
| }, |
| { |
| "entropy": 0.5220472222194076, |
| "epoch": 2.96, |
| "grad_norm": 1.40625, |
| "learning_rate": 1.8113333333333335e-05, |
| "loss": 0.5045, |
| "mean_token_accuracy": 0.848219645395875, |
| "num_tokens": 52947183.0, |
| "step": 1850 |
| }, |
| { |
| "entropy": 0.5195147603750229, |
| "epoch": 2.976, |
| "grad_norm": 1.3203125, |
| "learning_rate": 1.8091111111111113e-05, |
| "loss": 0.5009, |
| "mean_token_accuracy": 0.8488207392394542, |
| "num_tokens": 53226984.0, |
| "step": 1860 |
| }, |
| { |
| "entropy": 0.5123168766498566, |
| "epoch": 2.992, |
| "grad_norm": 1.3828125, |
| "learning_rate": 1.806888888888889e-05, |
| "loss": 0.4941, |
| "mean_token_accuracy": 0.8505234856158495, |
| "num_tokens": 53515204.0, |
| "step": 1870 |
| }, |
| { |
| "entropy": 0.4752411070279777, |
| "epoch": 3.008, |
| "grad_norm": 1.875, |
| "learning_rate": 1.8046666666666668e-05, |
| "loss": 0.4333, |
| "mean_token_accuracy": 0.8668791756033898, |
| "num_tokens": 53800592.0, |
| "step": 1880 |
| }, |
| { |
| "entropy": 0.3587687503546476, |
| "epoch": 3.024, |
| "grad_norm": 1.8828125, |
| "learning_rate": 1.8024444444444445e-05, |
| "loss": 0.3423, |
| "mean_token_accuracy": 0.888729453831911, |
| "num_tokens": 54092017.0, |
| "step": 1890 |
| }, |
| { |
| "entropy": 0.3438520409166813, |
| "epoch": 3.04, |
| "grad_norm": 1.8359375, |
| "learning_rate": 1.8002222222222223e-05, |
| "loss": 0.3233, |
| "mean_token_accuracy": 0.8938188333064317, |
| "num_tokens": 54380158.0, |
| "step": 1900 |
| }, |
| { |
| "epoch": 3.04, |
| "eval_math_entropy": 0.406534277677536, |
| "eval_math_loss": 0.7412194609642029, |
| "eval_math_mean_token_accuracy": 0.8225762667655945, |
| "eval_math_num_tokens": 54380158.0, |
| "eval_math_runtime": 35.7958, |
| "eval_math_samples_per_second": 13.968, |
| "eval_math_steps_per_second": 3.492, |
| "step": 1900 |
| }, |
| { |
| "epoch": 3.04, |
| "eval_physics_entropy": 0.38361185932159425, |
| "eval_physics_loss": 0.8625921607017517, |
| "eval_physics_mean_token_accuracy": 0.8045027327537536, |
| "eval_physics_num_tokens": 54380158.0, |
| "eval_physics_runtime": 41.4955, |
| "eval_physics_samples_per_second": 12.05, |
| "eval_physics_steps_per_second": 3.012, |
| "step": 1900 |
| }, |
| { |
| "entropy": 0.34686512807384134, |
| "epoch": 3.056, |
| "grad_norm": 1.8984375, |
| "learning_rate": 1.798e-05, |
| "loss": 0.3202, |
| "mean_token_accuracy": 0.8946099061518907, |
| "num_tokens": 54658093.0, |
| "step": 1910 |
| }, |
| { |
| "entropy": 0.34148691557347777, |
| "epoch": 3.072, |
| "grad_norm": 1.953125, |
| "learning_rate": 1.7957777777777778e-05, |
| "loss": 0.3202, |
| "mean_token_accuracy": 0.8947911463677883, |
| "num_tokens": 54949938.0, |
| "step": 1920 |
| }, |
| { |
| "entropy": 0.3277126123197377, |
| "epoch": 3.088, |
| "grad_norm": 2.21875, |
| "learning_rate": 1.7935555555555556e-05, |
| "loss": 0.3064, |
| "mean_token_accuracy": 0.8989322237670422, |
| "num_tokens": 55234643.0, |
| "step": 1930 |
| }, |
| { |
| "entropy": 0.34841074971482155, |
| "epoch": 3.104, |
| "grad_norm": 1.9609375, |
| "learning_rate": 1.7913333333333337e-05, |
| "loss": 0.3262, |
| "mean_token_accuracy": 0.8924810353666544, |
| "num_tokens": 55519679.0, |
| "step": 1940 |
| }, |
| { |
| "entropy": 0.33432568768039345, |
| "epoch": 3.12, |
| "grad_norm": 1.9453125, |
| "learning_rate": 1.789111111111111e-05, |
| "loss": 0.3163, |
| "mean_token_accuracy": 0.896980819106102, |
| "num_tokens": 55812397.0, |
| "step": 1950 |
| }, |
| { |
| "entropy": 0.33816060032695533, |
| "epoch": 3.136, |
| "grad_norm": 1.8671875, |
| "learning_rate": 1.7868888888888892e-05, |
| "loss": 0.3182, |
| "mean_token_accuracy": 0.8948573391884566, |
| "num_tokens": 56101550.0, |
| "step": 1960 |
| }, |
| { |
| "entropy": 0.3384258272126317, |
| "epoch": 3.152, |
| "grad_norm": 2.078125, |
| "learning_rate": 1.7846666666666666e-05, |
| "loss": 0.3209, |
| "mean_token_accuracy": 0.8947299372404813, |
| "num_tokens": 56387662.0, |
| "step": 1970 |
| }, |
| { |
| "entropy": 0.3522730448283255, |
| "epoch": 3.168, |
| "grad_norm": 2.03125, |
| "learning_rate": 1.7824444444444447e-05, |
| "loss": 0.333, |
| "mean_token_accuracy": 0.892043673619628, |
| "num_tokens": 56673426.0, |
| "step": 1980 |
| }, |
| { |
| "entropy": 0.3521577626466751, |
| "epoch": 3.184, |
| "grad_norm": 2.25, |
| "learning_rate": 1.780222222222222e-05, |
| "loss": 0.3268, |
| "mean_token_accuracy": 0.89161705262959, |
| "num_tokens": 56957044.0, |
| "step": 1990 |
| }, |
| { |
| "entropy": 0.3363432651385665, |
| "epoch": 3.2, |
| "grad_norm": 1.8671875, |
| "learning_rate": 1.7780000000000003e-05, |
| "loss": 0.3174, |
| "mean_token_accuracy": 0.8954976741224527, |
| "num_tokens": 57250097.0, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.2, |
| "eval_math_entropy": 0.3876710822582245, |
| "eval_math_loss": 0.7634654641151428, |
| "eval_math_mean_token_accuracy": 0.8221899409294129, |
| "eval_math_num_tokens": 57250097.0, |
| "eval_math_runtime": 35.8054, |
| "eval_math_samples_per_second": 13.964, |
| "eval_math_steps_per_second": 3.491, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.2, |
| "eval_physics_entropy": 0.35663985908031465, |
| "eval_physics_loss": 0.8939647674560547, |
| "eval_physics_mean_token_accuracy": 0.8043588905334472, |
| "eval_physics_num_tokens": 57250097.0, |
| "eval_physics_runtime": 41.1357, |
| "eval_physics_samples_per_second": 12.155, |
| "eval_physics_steps_per_second": 3.039, |
| "step": 2000 |
| }, |
| { |
| "entropy": 0.3453281221911311, |
| "epoch": 3.216, |
| "grad_norm": 2.046875, |
| "learning_rate": 1.7757777777777777e-05, |
| "loss": 0.3293, |
| "mean_token_accuracy": 0.8925925586372614, |
| "num_tokens": 57532026.0, |
| "step": 2010 |
| }, |
| { |
| "entropy": 0.34009323371574285, |
| "epoch": 3.232, |
| "grad_norm": 2.21875, |
| "learning_rate": 1.7735555555555558e-05, |
| "loss": 0.3216, |
| "mean_token_accuracy": 0.895359569042921, |
| "num_tokens": 57814853.0, |
| "step": 2020 |
| }, |
| { |
| "entropy": 0.32144517404958606, |
| "epoch": 3.248, |
| "grad_norm": 2.046875, |
| "learning_rate": 1.7713333333333335e-05, |
| "loss": 0.3013, |
| "mean_token_accuracy": 0.8993485007435084, |
| "num_tokens": 58094190.0, |
| "step": 2030 |
| }, |
| { |
| "entropy": 0.3362208865582943, |
| "epoch": 3.2640000000000002, |
| "grad_norm": 2.15625, |
| "learning_rate": 1.7691111111111113e-05, |
| "loss": 0.3179, |
| "mean_token_accuracy": 0.8959619626402855, |
| "num_tokens": 58374378.0, |
| "step": 2040 |
| }, |
| { |
| "entropy": 0.34206875674426557, |
| "epoch": 3.2800000000000002, |
| "grad_norm": 1.96875, |
| "learning_rate": 1.766888888888889e-05, |
| "loss": 0.3233, |
| "mean_token_accuracy": 0.8937642563134431, |
| "num_tokens": 58652789.0, |
| "step": 2050 |
| }, |
| { |
| "entropy": 0.3506787528283894, |
| "epoch": 3.296, |
| "grad_norm": 2.078125, |
| "learning_rate": 1.7646666666666668e-05, |
| "loss": 0.331, |
| "mean_token_accuracy": 0.8917676325887441, |
| "num_tokens": 58930014.0, |
| "step": 2060 |
| }, |
| { |
| "entropy": 0.345881634298712, |
| "epoch": 3.312, |
| "grad_norm": 2.015625, |
| "learning_rate": 1.7624444444444446e-05, |
| "loss": 0.3262, |
| "mean_token_accuracy": 0.8924973297864198, |
| "num_tokens": 59210270.0, |
| "step": 2070 |
| }, |
| { |
| "entropy": 0.34120792699977753, |
| "epoch": 3.328, |
| "grad_norm": 2.140625, |
| "learning_rate": 1.7602222222222223e-05, |
| "loss": 0.3192, |
| "mean_token_accuracy": 0.8952000044286251, |
| "num_tokens": 59496408.0, |
| "step": 2080 |
| }, |
| { |
| "entropy": 0.3278652913868427, |
| "epoch": 3.344, |
| "grad_norm": 2.15625, |
| "learning_rate": 1.758e-05, |
| "loss": 0.3112, |
| "mean_token_accuracy": 0.8966204669326544, |
| "num_tokens": 59779722.0, |
| "step": 2090 |
| }, |
| { |
| "entropy": 0.3657307820394635, |
| "epoch": 3.36, |
| "grad_norm": 2.21875, |
| "learning_rate": 1.755777777777778e-05, |
| "loss": 0.3419, |
| "mean_token_accuracy": 0.8882275156676769, |
| "num_tokens": 60061562.0, |
| "step": 2100 |
| }, |
| { |
| "epoch": 3.36, |
| "eval_math_entropy": 0.3830914467573166, |
| "eval_math_loss": 0.7686993479728699, |
| "eval_math_mean_token_accuracy": 0.8224415044784545, |
| "eval_math_num_tokens": 60061562.0, |
| "eval_math_runtime": 35.8062, |
| "eval_math_samples_per_second": 13.964, |
| "eval_math_steps_per_second": 3.491, |
| "step": 2100 |
| }, |
| { |
| "epoch": 3.36, |
| "eval_physics_entropy": 0.3522386626005173, |
| "eval_physics_loss": 0.9039635062217712, |
| "eval_physics_mean_token_accuracy": 0.8043717379570008, |
| "eval_physics_num_tokens": 60061562.0, |
| "eval_physics_runtime": 41.4883, |
| "eval_physics_samples_per_second": 12.052, |
| "eval_physics_steps_per_second": 3.013, |
| "step": 2100 |
| }, |
| { |
| "entropy": 0.35135511131957176, |
| "epoch": 3.376, |
| "grad_norm": 2.015625, |
| "learning_rate": 1.7535555555555556e-05, |
| "loss": 0.3328, |
| "mean_token_accuracy": 0.8913474876433611, |
| "num_tokens": 60346574.0, |
| "step": 2110 |
| }, |
| { |
| "entropy": 0.3290985696949065, |
| "epoch": 3.392, |
| "grad_norm": 2.015625, |
| "learning_rate": 1.7513333333333334e-05, |
| "loss": 0.3122, |
| "mean_token_accuracy": 0.8976335555315018, |
| "num_tokens": 60636820.0, |
| "step": 2120 |
| }, |
| { |
| "entropy": 0.3349749938584864, |
| "epoch": 3.408, |
| "grad_norm": 2.09375, |
| "learning_rate": 1.749111111111111e-05, |
| "loss": 0.3148, |
| "mean_token_accuracy": 0.8964381668716669, |
| "num_tokens": 60922033.0, |
| "step": 2130 |
| }, |
| { |
| "entropy": 0.3461784217506647, |
| "epoch": 3.424, |
| "grad_norm": 2.25, |
| "learning_rate": 1.746888888888889e-05, |
| "loss": 0.327, |
| "mean_token_accuracy": 0.8919767573475837, |
| "num_tokens": 61209156.0, |
| "step": 2140 |
| }, |
| { |
| "entropy": 0.3424711484462023, |
| "epoch": 3.44, |
| "grad_norm": 1.9609375, |
| "learning_rate": 1.7446666666666667e-05, |
| "loss": 0.3233, |
| "mean_token_accuracy": 0.8935233734548091, |
| "num_tokens": 61491228.0, |
| "step": 2150 |
| }, |
| { |
| "entropy": 0.3479755653068423, |
| "epoch": 3.456, |
| "grad_norm": 2.140625, |
| "learning_rate": 1.7424444444444444e-05, |
| "loss": 0.3292, |
| "mean_token_accuracy": 0.8924077559262514, |
| "num_tokens": 61773033.0, |
| "step": 2160 |
| }, |
| { |
| "entropy": 0.3536012376658618, |
| "epoch": 3.472, |
| "grad_norm": 2.0625, |
| "learning_rate": 1.7402222222222222e-05, |
| "loss": 0.3402, |
| "mean_token_accuracy": 0.8894314222037792, |
| "num_tokens": 62061802.0, |
| "step": 2170 |
| }, |
| { |
| "entropy": 0.34079490676522256, |
| "epoch": 3.488, |
| "grad_norm": 2.078125, |
| "learning_rate": 1.7380000000000003e-05, |
| "loss": 0.3233, |
| "mean_token_accuracy": 0.8940149489790201, |
| "num_tokens": 62347490.0, |
| "step": 2180 |
| }, |
| { |
| "entropy": 0.34653188129886986, |
| "epoch": 3.504, |
| "grad_norm": 2.0, |
| "learning_rate": 1.735777777777778e-05, |
| "loss": 0.3243, |
| "mean_token_accuracy": 0.8935138266533613, |
| "num_tokens": 62636811.0, |
| "step": 2190 |
| }, |
| { |
| "entropy": 0.3420496225357056, |
| "epoch": 3.52, |
| "grad_norm": 2.03125, |
| "learning_rate": 1.7335555555555558e-05, |
| "loss": 0.3255, |
| "mean_token_accuracy": 0.8933457151055336, |
| "num_tokens": 62928536.0, |
| "step": 2200 |
| }, |
| { |
| "epoch": 3.52, |
| "eval_math_entropy": 0.39784613919258116, |
| "eval_math_loss": 0.7561159133911133, |
| "eval_math_mean_token_accuracy": 0.8226356348991394, |
| "eval_math_num_tokens": 62928536.0, |
| "eval_math_runtime": 35.8487, |
| "eval_math_samples_per_second": 13.948, |
| "eval_math_steps_per_second": 3.487, |
| "step": 2200 |
| }, |
| { |
| "epoch": 3.52, |
| "eval_physics_entropy": 0.3542559086084366, |
| "eval_physics_loss": 0.8993895053863525, |
| "eval_physics_mean_token_accuracy": 0.8046151103973389, |
| "eval_physics_num_tokens": 62928536.0, |
| "eval_physics_runtime": 41.5136, |
| "eval_physics_samples_per_second": 12.044, |
| "eval_physics_steps_per_second": 3.011, |
| "step": 2200 |
| }, |
| { |
| "entropy": 0.3441290996037424, |
| "epoch": 3.536, |
| "grad_norm": 2.234375, |
| "learning_rate": 1.7313333333333336e-05, |
| "loss": 0.3257, |
| "mean_token_accuracy": 0.8930490992963314, |
| "num_tokens": 63212496.0, |
| "step": 2210 |
| }, |
| { |
| "entropy": 0.3405540899373591, |
| "epoch": 3.552, |
| "grad_norm": 2.0, |
| "learning_rate": 1.7291111111111113e-05, |
| "loss": 0.3211, |
| "mean_token_accuracy": 0.8948382496833801, |
| "num_tokens": 63506921.0, |
| "step": 2220 |
| }, |
| { |
| "entropy": 0.33624334512278437, |
| "epoch": 3.568, |
| "grad_norm": 1.9375, |
| "learning_rate": 1.726888888888889e-05, |
| "loss": 0.3159, |
| "mean_token_accuracy": 0.8952401872724295, |
| "num_tokens": 63803208.0, |
| "step": 2230 |
| }, |
| { |
| "entropy": 0.34232397861778735, |
| "epoch": 3.584, |
| "grad_norm": 2.25, |
| "learning_rate": 1.724666666666667e-05, |
| "loss": 0.3238, |
| "mean_token_accuracy": 0.8939516879618168, |
| "num_tokens": 64084867.0, |
| "step": 2240 |
| }, |
| { |
| "entropy": 0.33109099259600044, |
| "epoch": 3.6, |
| "grad_norm": 2.109375, |
| "learning_rate": 1.7224444444444446e-05, |
| "loss": 0.3153, |
| "mean_token_accuracy": 0.8967003002762794, |
| "num_tokens": 64377046.0, |
| "step": 2250 |
| }, |
| { |
| "entropy": 0.35013388488441705, |
| "epoch": 3.616, |
| "grad_norm": 2.109375, |
| "learning_rate": 1.7202222222222224e-05, |
| "loss": 0.331, |
| "mean_token_accuracy": 0.8905543919652701, |
| "num_tokens": 64664301.0, |
| "step": 2260 |
| }, |
| { |
| "entropy": 0.3353199644945562, |
| "epoch": 3.632, |
| "grad_norm": 2.140625, |
| "learning_rate": 1.718e-05, |
| "loss": 0.316, |
| "mean_token_accuracy": 0.895779125764966, |
| "num_tokens": 64946172.0, |
| "step": 2270 |
| }, |
| { |
| "entropy": 0.3319018458016217, |
| "epoch": 3.648, |
| "grad_norm": 2.0, |
| "learning_rate": 1.715777777777778e-05, |
| "loss": 0.3138, |
| "mean_token_accuracy": 0.8972285516560078, |
| "num_tokens": 65230709.0, |
| "step": 2280 |
| }, |
| { |
| "entropy": 0.36236064694821835, |
| "epoch": 3.664, |
| "grad_norm": 2.109375, |
| "learning_rate": 1.7135555555555557e-05, |
| "loss": 0.344, |
| "mean_token_accuracy": 0.8877242401242256, |
| "num_tokens": 65515738.0, |
| "step": 2290 |
| }, |
| { |
| "entropy": 0.3466769499704242, |
| "epoch": 3.68, |
| "grad_norm": 2.0625, |
| "learning_rate": 1.7113333333333334e-05, |
| "loss": 0.3276, |
| "mean_token_accuracy": 0.8919828839600086, |
| "num_tokens": 65802680.0, |
| "step": 2300 |
| }, |
| { |
| "epoch": 3.68, |
| "eval_math_entropy": 0.3900481255054474, |
| "eval_math_loss": 0.7627390623092651, |
| "eval_math_mean_token_accuracy": 0.8218873319625855, |
| "eval_math_num_tokens": 65802680.0, |
| "eval_math_runtime": 35.8389, |
| "eval_math_samples_per_second": 13.951, |
| "eval_math_steps_per_second": 3.488, |
| "step": 2300 |
| }, |
| { |
| "epoch": 3.68, |
| "eval_physics_entropy": 0.3501615208387375, |
| "eval_physics_loss": 0.9098384380340576, |
| "eval_physics_mean_token_accuracy": 0.8042201838493347, |
| "eval_physics_num_tokens": 65802680.0, |
| "eval_physics_runtime": 41.525, |
| "eval_physics_samples_per_second": 12.041, |
| "eval_physics_steps_per_second": 3.01, |
| "step": 2300 |
| }, |
| { |
| "entropy": 0.3516978560946882, |
| "epoch": 3.6959999999999997, |
| "grad_norm": 2.0625, |
| "learning_rate": 1.7091111111111112e-05, |
| "loss": 0.3334, |
| "mean_token_accuracy": 0.8908947184681892, |
| "num_tokens": 66089104.0, |
| "step": 2310 |
| }, |
| { |
| "entropy": 0.3427901498042047, |
| "epoch": 3.7119999999999997, |
| "grad_norm": 2.0, |
| "learning_rate": 1.706888888888889e-05, |
| "loss": 0.3228, |
| "mean_token_accuracy": 0.8933374501764775, |
| "num_tokens": 66369895.0, |
| "step": 2320 |
| }, |
| { |
| "entropy": 0.3467458762228489, |
| "epoch": 3.7279999999999998, |
| "grad_norm": 2.21875, |
| "learning_rate": 1.704666666666667e-05, |
| "loss": 0.327, |
| "mean_token_accuracy": 0.8924272887408733, |
| "num_tokens": 66653165.0, |
| "step": 2330 |
| }, |
| { |
| "entropy": 0.34478482212871314, |
| "epoch": 3.7439999999999998, |
| "grad_norm": 2.109375, |
| "learning_rate": 1.7024444444444445e-05, |
| "loss": 0.3245, |
| "mean_token_accuracy": 0.8931741625070572, |
| "num_tokens": 66941476.0, |
| "step": 2340 |
| }, |
| { |
| "entropy": 0.34707736801356076, |
| "epoch": 3.76, |
| "grad_norm": 2.140625, |
| "learning_rate": 1.7002222222222226e-05, |
| "loss": 0.3306, |
| "mean_token_accuracy": 0.8919444873929023, |
| "num_tokens": 67225380.0, |
| "step": 2350 |
| }, |
| { |
| "entropy": 0.35979456212371586, |
| "epoch": 3.776, |
| "grad_norm": 2.234375, |
| "learning_rate": 1.698e-05, |
| "loss": 0.3425, |
| "mean_token_accuracy": 0.888241409137845, |
| "num_tokens": 67516641.0, |
| "step": 2360 |
| }, |
| { |
| "entropy": 0.337408459559083, |
| "epoch": 3.792, |
| "grad_norm": 2.234375, |
| "learning_rate": 1.695777777777778e-05, |
| "loss": 0.3207, |
| "mean_token_accuracy": 0.8952056996524334, |
| "num_tokens": 67808009.0, |
| "step": 2370 |
| }, |
| { |
| "entropy": 0.35541942873969673, |
| "epoch": 3.808, |
| "grad_norm": 2.0625, |
| "learning_rate": 1.6935555555555555e-05, |
| "loss": 0.3354, |
| "mean_token_accuracy": 0.8904807876795531, |
| "num_tokens": 68092996.0, |
| "step": 2380 |
| }, |
| { |
| "entropy": 0.338986696396023, |
| "epoch": 3.824, |
| "grad_norm": 2.03125, |
| "learning_rate": 1.6913333333333336e-05, |
| "loss": 0.3217, |
| "mean_token_accuracy": 0.8948666993528605, |
| "num_tokens": 68380631.0, |
| "step": 2390 |
| }, |
| { |
| "entropy": 0.34510293062776326, |
| "epoch": 3.84, |
| "grad_norm": 2.140625, |
| "learning_rate": 1.689111111111111e-05, |
| "loss": 0.327, |
| "mean_token_accuracy": 0.8922771938145161, |
| "num_tokens": 68666505.0, |
| "step": 2400 |
| }, |
| { |
| "epoch": 3.84, |
| "eval_math_entropy": 0.3931693367958069, |
| "eval_math_loss": 0.7599309682846069, |
| "eval_math_mean_token_accuracy": 0.8226110072135925, |
| "eval_math_num_tokens": 68666505.0, |
| "eval_math_runtime": 35.819, |
| "eval_math_samples_per_second": 13.959, |
| "eval_math_steps_per_second": 3.49, |
| "step": 2400 |
| }, |
| { |
| "epoch": 3.84, |
| "eval_physics_entropy": 0.3486952732801437, |
| "eval_physics_loss": 0.9106001257896423, |
| "eval_physics_mean_token_accuracy": 0.8040544500350952, |
| "eval_physics_num_tokens": 68666505.0, |
| "eval_physics_runtime": 41.4809, |
| "eval_physics_samples_per_second": 12.054, |
| "eval_physics_steps_per_second": 3.013, |
| "step": 2400 |
| }, |
| { |
| "entropy": 0.341458183247596, |
| "epoch": 3.856, |
| "grad_norm": 2.1875, |
| "learning_rate": 1.686888888888889e-05, |
| "loss": 0.3259, |
| "mean_token_accuracy": 0.8934879627078771, |
| "num_tokens": 68951258.0, |
| "step": 2410 |
| }, |
| { |
| "entropy": 0.35674845492467283, |
| "epoch": 3.872, |
| "grad_norm": 2.234375, |
| "learning_rate": 1.684666666666667e-05, |
| "loss": 0.3383, |
| "mean_token_accuracy": 0.8894801624119282, |
| "num_tokens": 69243705.0, |
| "step": 2420 |
| }, |
| { |
| "entropy": 0.3472815524786711, |
| "epoch": 3.888, |
| "grad_norm": 2.171875, |
| "learning_rate": 1.6824444444444447e-05, |
| "loss": 0.3299, |
| "mean_token_accuracy": 0.8912571128457785, |
| "num_tokens": 69529393.0, |
| "step": 2430 |
| }, |
| { |
| "entropy": 0.345150712877512, |
| "epoch": 3.904, |
| "grad_norm": 2.28125, |
| "learning_rate": 1.6802222222222224e-05, |
| "loss": 0.3269, |
| "mean_token_accuracy": 0.8925980541855096, |
| "num_tokens": 69820950.0, |
| "step": 2440 |
| }, |
| { |
| "entropy": 0.3369919521734118, |
| "epoch": 3.92, |
| "grad_norm": 2.21875, |
| "learning_rate": 1.6780000000000002e-05, |
| "loss": 0.317, |
| "mean_token_accuracy": 0.8948526129126548, |
| "num_tokens": 70097736.0, |
| "step": 2450 |
| }, |
| { |
| "entropy": 0.3590950695797801, |
| "epoch": 3.936, |
| "grad_norm": 2.046875, |
| "learning_rate": 1.675777777777778e-05, |
| "loss": 0.34, |
| "mean_token_accuracy": 0.8890990480780602, |
| "num_tokens": 70395123.0, |
| "step": 2460 |
| }, |
| { |
| "entropy": 0.345857547596097, |
| "epoch": 3.952, |
| "grad_norm": 2.015625, |
| "learning_rate": 1.6735555555555557e-05, |
| "loss": 0.3301, |
| "mean_token_accuracy": 0.8929780598729848, |
| "num_tokens": 70678153.0, |
| "step": 2470 |
| }, |
| { |
| "entropy": 0.3558691343292594, |
| "epoch": 3.968, |
| "grad_norm": 2.078125, |
| "learning_rate": 1.6713333333333335e-05, |
| "loss": 0.3375, |
| "mean_token_accuracy": 0.8891723170876503, |
| "num_tokens": 70963030.0, |
| "step": 2480 |
| }, |
| { |
| "entropy": 0.33871191181242466, |
| "epoch": 3.984, |
| "grad_norm": 1.953125, |
| "learning_rate": 1.6691111111111112e-05, |
| "loss": 0.3215, |
| "mean_token_accuracy": 0.8948656614869833, |
| "num_tokens": 71264525.0, |
| "step": 2490 |
| }, |
| { |
| "entropy": 0.3556901432573795, |
| "epoch": 4.0, |
| "grad_norm": 2.265625, |
| "learning_rate": 1.666888888888889e-05, |
| "loss": 0.337, |
| "mean_token_accuracy": 0.8900394972413779, |
| "num_tokens": 71545696.0, |
| "step": 2500 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_math_entropy": 0.3949365179538727, |
| "eval_math_loss": 0.7542520761489868, |
| "eval_math_mean_token_accuracy": 0.8234356184005738, |
| "eval_math_num_tokens": 71545696.0, |
| "eval_math_runtime": 35.7966, |
| "eval_math_samples_per_second": 13.968, |
| "eval_math_steps_per_second": 3.492, |
| "step": 2500 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_physics_entropy": 0.3519064650535583, |
| "eval_physics_loss": 0.9068982005119324, |
| "eval_physics_mean_token_accuracy": 0.8042608909606933, |
| "eval_physics_num_tokens": 71545696.0, |
| "eval_physics_runtime": 41.1579, |
| "eval_physics_samples_per_second": 12.148, |
| "eval_physics_steps_per_second": 3.037, |
| "step": 2500 |
| }, |
| { |
| "entropy": 0.22543487511575222, |
| "epoch": 4.016, |
| "grad_norm": 2.359375, |
| "learning_rate": 1.6646666666666668e-05, |
| "loss": 0.1799, |
| "mean_token_accuracy": 0.9439745012670755, |
| "num_tokens": 71839909.0, |
| "step": 2510 |
| }, |
| { |
| "entropy": 0.2174504542723298, |
| "epoch": 4.032, |
| "grad_norm": 2.203125, |
| "learning_rate": 1.6624444444444445e-05, |
| "loss": 0.1708, |
| "mean_token_accuracy": 0.9454751797020435, |
| "num_tokens": 72128860.0, |
| "step": 2520 |
| }, |
| { |
| "entropy": 0.20308812106959523, |
| "epoch": 4.048, |
| "grad_norm": 1.75, |
| "learning_rate": 1.6602222222222223e-05, |
| "loss": 0.1635, |
| "mean_token_accuracy": 0.9484880570322275, |
| "num_tokens": 72415886.0, |
| "step": 2530 |
| }, |
| { |
| "entropy": 0.19942862098105252, |
| "epoch": 4.064, |
| "grad_norm": 2.3125, |
| "learning_rate": 1.658e-05, |
| "loss": 0.1599, |
| "mean_token_accuracy": 0.9482992067933083, |
| "num_tokens": 72695806.0, |
| "step": 2540 |
| }, |
| { |
| "entropy": 0.19497004388831557, |
| "epoch": 4.08, |
| "grad_norm": 1.984375, |
| "learning_rate": 1.6557777777777778e-05, |
| "loss": 0.1524, |
| "mean_token_accuracy": 0.9501827869564294, |
| "num_tokens": 72986978.0, |
| "step": 2550 |
| }, |
| { |
| "entropy": 0.1978817671071738, |
| "epoch": 4.096, |
| "grad_norm": 2.328125, |
| "learning_rate": 1.6535555555555556e-05, |
| "loss": 0.1597, |
| "mean_token_accuracy": 0.9481502700597048, |
| "num_tokens": 73271299.0, |
| "step": 2560 |
| }, |
| { |
| "entropy": 0.19556209091097115, |
| "epoch": 4.112, |
| "grad_norm": 2.15625, |
| "learning_rate": 1.6513333333333333e-05, |
| "loss": 0.1569, |
| "mean_token_accuracy": 0.9491645645350217, |
| "num_tokens": 73562479.0, |
| "step": 2570 |
| }, |
| { |
| "entropy": 0.19597707632929087, |
| "epoch": 4.128, |
| "grad_norm": 2.09375, |
| "learning_rate": 1.6491111111111114e-05, |
| "loss": 0.1588, |
| "mean_token_accuracy": 0.9492202006280422, |
| "num_tokens": 73845445.0, |
| "step": 2580 |
| }, |
| { |
| "entropy": 0.19680100358091296, |
| "epoch": 4.144, |
| "grad_norm": 1.9296875, |
| "learning_rate": 1.646888888888889e-05, |
| "loss": 0.1611, |
| "mean_token_accuracy": 0.9478188194334507, |
| "num_tokens": 74135425.0, |
| "step": 2590 |
| }, |
| { |
| "entropy": 0.20560383140109478, |
| "epoch": 4.16, |
| "grad_norm": 1.8828125, |
| "learning_rate": 1.644666666666667e-05, |
| "loss": 0.1678, |
| "mean_token_accuracy": 0.945937181636691, |
| "num_tokens": 74431943.0, |
| "step": 2600 |
| }, |
| { |
| "epoch": 4.16, |
| "eval_math_entropy": 0.2884341893196106, |
| "eval_math_loss": 0.9303049445152283, |
| "eval_math_mean_token_accuracy": 0.8163944878578187, |
| "eval_math_num_tokens": 74431943.0, |
| "eval_math_runtime": 35.7808, |
| "eval_math_samples_per_second": 13.974, |
| "eval_math_steps_per_second": 3.493, |
| "step": 2600 |
| }, |
| { |
| "epoch": 4.16, |
| "eval_physics_entropy": 0.2818371082544327, |
| "eval_physics_loss": 1.0674083232879639, |
| "eval_physics_mean_token_accuracy": 0.7992216634750366, |
| "eval_physics_num_tokens": 74431943.0, |
| "eval_physics_runtime": 41.4876, |
| "eval_physics_samples_per_second": 12.052, |
| "eval_physics_steps_per_second": 3.013, |
| "step": 2600 |
| }, |
| { |
| "entropy": 0.20040173502638936, |
| "epoch": 4.176, |
| "grad_norm": 2.21875, |
| "learning_rate": 1.6424444444444444e-05, |
| "loss": 0.1641, |
| "mean_token_accuracy": 0.947614687308669, |
| "num_tokens": 74713404.0, |
| "step": 2610 |
| }, |
| { |
| "entropy": 0.19675334333442152, |
| "epoch": 4.192, |
| "grad_norm": 2.0625, |
| "learning_rate": 1.6402222222222225e-05, |
| "loss": 0.1587, |
| "mean_token_accuracy": 0.948599212616682, |
| "num_tokens": 74997493.0, |
| "step": 2620 |
| }, |
| { |
| "entropy": 0.19941296209581197, |
| "epoch": 4.208, |
| "grad_norm": 2.0625, |
| "learning_rate": 1.638e-05, |
| "loss": 0.1591, |
| "mean_token_accuracy": 0.94874595887959, |
| "num_tokens": 75275375.0, |
| "step": 2630 |
| }, |
| { |
| "entropy": 0.19698526146821677, |
| "epoch": 4.224, |
| "grad_norm": 2.09375, |
| "learning_rate": 1.635777777777778e-05, |
| "loss": 0.1579, |
| "mean_token_accuracy": 0.9489138159900904, |
| "num_tokens": 75563461.0, |
| "step": 2640 |
| }, |
| { |
| "entropy": 0.2004603309556842, |
| "epoch": 4.24, |
| "grad_norm": 2.078125, |
| "learning_rate": 1.6335555555555558e-05, |
| "loss": 0.1662, |
| "mean_token_accuracy": 0.9469812735915184, |
| "num_tokens": 75854808.0, |
| "step": 2650 |
| }, |
| { |
| "entropy": 0.19907335569150747, |
| "epoch": 4.256, |
| "grad_norm": 2.21875, |
| "learning_rate": 1.6313333333333335e-05, |
| "loss": 0.1629, |
| "mean_token_accuracy": 0.9474829658865929, |
| "num_tokens": 76148449.0, |
| "step": 2660 |
| }, |
| { |
| "entropy": 0.19561562612652778, |
| "epoch": 4.272, |
| "grad_norm": 2.140625, |
| "learning_rate": 1.6291111111111113e-05, |
| "loss": 0.1579, |
| "mean_token_accuracy": 0.9487213175743818, |
| "num_tokens": 76435793.0, |
| "step": 2670 |
| }, |
| { |
| "entropy": 0.19496299759484828, |
| "epoch": 4.288, |
| "grad_norm": 2.203125, |
| "learning_rate": 1.626888888888889e-05, |
| "loss": 0.1597, |
| "mean_token_accuracy": 0.9485157284885645, |
| "num_tokens": 76726162.0, |
| "step": 2680 |
| }, |
| { |
| "entropy": 0.20333012314513327, |
| "epoch": 4.304, |
| "grad_norm": 2.125, |
| "learning_rate": 1.6246666666666668e-05, |
| "loss": 0.1668, |
| "mean_token_accuracy": 0.94562135823071, |
| "num_tokens": 77016858.0, |
| "step": 2690 |
| }, |
| { |
| "entropy": 0.19900708929635585, |
| "epoch": 4.32, |
| "grad_norm": 2.15625, |
| "learning_rate": 1.6224444444444446e-05, |
| "loss": 0.1639, |
| "mean_token_accuracy": 0.9465276099741459, |
| "num_tokens": 77311424.0, |
| "step": 2700 |
| }, |
| { |
| "epoch": 4.32, |
| "eval_math_entropy": 0.2942132350206375, |
| "eval_math_loss": 0.9334467053413391, |
| "eval_math_mean_token_accuracy": 0.8159489192962647, |
| "eval_math_num_tokens": 77311424.0, |
| "eval_math_runtime": 35.8238, |
| "eval_math_samples_per_second": 13.957, |
| "eval_math_steps_per_second": 3.489, |
| "step": 2700 |
| }, |
| { |
| "epoch": 4.32, |
| "eval_physics_entropy": 0.28400694143772126, |
| "eval_physics_loss": 1.0711963176727295, |
| "eval_physics_mean_token_accuracy": 0.7983153300285339, |
| "eval_physics_num_tokens": 77311424.0, |
| "eval_physics_runtime": 41.5147, |
| "eval_physics_samples_per_second": 12.044, |
| "eval_physics_steps_per_second": 3.011, |
| "step": 2700 |
| }, |
| { |
| "entropy": 0.20124221742153167, |
| "epoch": 4.336, |
| "grad_norm": 2.1875, |
| "learning_rate": 1.6202222222222223e-05, |
| "loss": 0.1659, |
| "mean_token_accuracy": 0.9459259379655123, |
| "num_tokens": 77596829.0, |
| "step": 2710 |
| }, |
| { |
| "entropy": 0.20279881260357796, |
| "epoch": 4.352, |
| "grad_norm": 2.34375, |
| "learning_rate": 1.618e-05, |
| "loss": 0.1676, |
| "mean_token_accuracy": 0.9458406940102577, |
| "num_tokens": 77889107.0, |
| "step": 2720 |
| }, |
| { |
| "entropy": 0.20401687268167734, |
| "epoch": 4.368, |
| "grad_norm": 2.265625, |
| "learning_rate": 1.615777777777778e-05, |
| "loss": 0.1683, |
| "mean_token_accuracy": 0.9460482109338045, |
| "num_tokens": 78176273.0, |
| "step": 2730 |
| }, |
| { |
| "entropy": 0.20603159624151884, |
| "epoch": 4.384, |
| "grad_norm": 2.25, |
| "learning_rate": 1.6135555555555556e-05, |
| "loss": 0.171, |
| "mean_token_accuracy": 0.9445526108145714, |
| "num_tokens": 78461335.0, |
| "step": 2740 |
| }, |
| { |
| "entropy": 0.20936214146204293, |
| "epoch": 4.4, |
| "grad_norm": 2.296875, |
| "learning_rate": 1.6113333333333334e-05, |
| "loss": 0.1746, |
| "mean_token_accuracy": 0.943215336278081, |
| "num_tokens": 78748521.0, |
| "step": 2750 |
| }, |
| { |
| "entropy": 0.20285641457885503, |
| "epoch": 4.416, |
| "grad_norm": 2.3125, |
| "learning_rate": 1.609111111111111e-05, |
| "loss": 0.1679, |
| "mean_token_accuracy": 0.9455374360084534, |
| "num_tokens": 79028488.0, |
| "step": 2760 |
| }, |
| { |
| "entropy": 0.1989856549538672, |
| "epoch": 4.432, |
| "grad_norm": 2.171875, |
| "learning_rate": 1.606888888888889e-05, |
| "loss": 0.162, |
| "mean_token_accuracy": 0.9475978959351778, |
| "num_tokens": 79315788.0, |
| "step": 2770 |
| }, |
| { |
| "entropy": 0.20125137683935462, |
| "epoch": 4.448, |
| "grad_norm": 2.265625, |
| "learning_rate": 1.6046666666666667e-05, |
| "loss": 0.1667, |
| "mean_token_accuracy": 0.9461831945925951, |
| "num_tokens": 79595473.0, |
| "step": 2780 |
| }, |
| { |
| "entropy": 0.20731261624023317, |
| "epoch": 4.464, |
| "grad_norm": 2.296875, |
| "learning_rate": 1.6024444444444444e-05, |
| "loss": 0.1719, |
| "mean_token_accuracy": 0.9443279825150966, |
| "num_tokens": 79884376.0, |
| "step": 2790 |
| }, |
| { |
| "entropy": 0.20476281307637692, |
| "epoch": 4.48, |
| "grad_norm": 2.28125, |
| "learning_rate": 1.6002222222222222e-05, |
| "loss": 0.1676, |
| "mean_token_accuracy": 0.9456661235541105, |
| "num_tokens": 80163433.0, |
| "step": 2800 |
| }, |
| { |
| "epoch": 4.48, |
| "eval_math_entropy": 0.2912301456928253, |
| "eval_math_loss": 0.9387769103050232, |
| "eval_math_mean_token_accuracy": 0.8159988284111023, |
| "eval_math_num_tokens": 80163433.0, |
| "eval_math_runtime": 35.8056, |
| "eval_math_samples_per_second": 13.964, |
| "eval_math_steps_per_second": 3.491, |
| "step": 2800 |
| }, |
| { |
| "epoch": 4.48, |
| "eval_physics_entropy": 0.283928528547287, |
| "eval_physics_loss": 1.0744123458862305, |
| "eval_physics_mean_token_accuracy": 0.79825497341156, |
| "eval_physics_num_tokens": 80163433.0, |
| "eval_physics_runtime": 41.5025, |
| "eval_physics_samples_per_second": 12.047, |
| "eval_physics_steps_per_second": 3.012, |
| "step": 2800 |
| }, |
| { |
| "entropy": 0.21272651255130767, |
| "epoch": 4.496, |
| "grad_norm": 2.484375, |
| "learning_rate": 1.5980000000000003e-05, |
| "loss": 0.1786, |
| "mean_token_accuracy": 0.94269258081913, |
| "num_tokens": 80443676.0, |
| "step": 2810 |
| }, |
| { |
| "entropy": 0.20025675450451672, |
| "epoch": 4.5120000000000005, |
| "grad_norm": 2.28125, |
| "learning_rate": 1.5957777777777777e-05, |
| "loss": 0.165, |
| "mean_token_accuracy": 0.9459723506122828, |
| "num_tokens": 80731924.0, |
| "step": 2820 |
| }, |
| { |
| "entropy": 0.20602681376039983, |
| "epoch": 4.5280000000000005, |
| "grad_norm": 2.203125, |
| "learning_rate": 1.5935555555555558e-05, |
| "loss": 0.1714, |
| "mean_token_accuracy": 0.9448773771524429, |
| "num_tokens": 81022957.0, |
| "step": 2830 |
| }, |
| { |
| "entropy": 0.19887811122462154, |
| "epoch": 4.5440000000000005, |
| "grad_norm": 2.03125, |
| "learning_rate": 1.5913333333333332e-05, |
| "loss": 0.1649, |
| "mean_token_accuracy": 0.9466291036456823, |
| "num_tokens": 81309913.0, |
| "step": 2840 |
| }, |
| { |
| "entropy": 0.20165206529200078, |
| "epoch": 4.5600000000000005, |
| "grad_norm": 2.390625, |
| "learning_rate": 1.5891111111111113e-05, |
| "loss": 0.1685, |
| "mean_token_accuracy": 0.9454752709716558, |
| "num_tokens": 81589202.0, |
| "step": 2850 |
| }, |
| { |
| "entropy": 0.2059997975360602, |
| "epoch": 4.576, |
| "grad_norm": 2.375, |
| "learning_rate": 1.5868888888888888e-05, |
| "loss": 0.1704, |
| "mean_token_accuracy": 0.9445979256182909, |
| "num_tokens": 81869804.0, |
| "step": 2860 |
| }, |
| { |
| "entropy": 0.2071733211632818, |
| "epoch": 4.592, |
| "grad_norm": 2.296875, |
| "learning_rate": 1.584666666666667e-05, |
| "loss": 0.1733, |
| "mean_token_accuracy": 0.9441272236406804, |
| "num_tokens": 82148464.0, |
| "step": 2870 |
| }, |
| { |
| "entropy": 0.20506689832545816, |
| "epoch": 4.608, |
| "grad_norm": 2.40625, |
| "learning_rate": 1.5824444444444446e-05, |
| "loss": 0.1696, |
| "mean_token_accuracy": 0.9445500578731298, |
| "num_tokens": 82427716.0, |
| "step": 2880 |
| }, |
| { |
| "entropy": 0.20488002435304226, |
| "epoch": 4.624, |
| "grad_norm": 2.359375, |
| "learning_rate": 1.5802222222222224e-05, |
| "loss": 0.1696, |
| "mean_token_accuracy": 0.9450069412589073, |
| "num_tokens": 82710604.0, |
| "step": 2890 |
| }, |
| { |
| "entropy": 0.20467088604345918, |
| "epoch": 4.64, |
| "grad_norm": 2.296875, |
| "learning_rate": 1.578e-05, |
| "loss": 0.1688, |
| "mean_token_accuracy": 0.944780471175909, |
| "num_tokens": 82990401.0, |
| "step": 2900 |
| }, |
| { |
| "epoch": 4.64, |
| "eval_math_entropy": 0.29218543314933776, |
| "eval_math_loss": 0.9344849586486816, |
| "eval_math_mean_token_accuracy": 0.816554964542389, |
| "eval_math_num_tokens": 82990401.0, |
| "eval_math_runtime": 35.7784, |
| "eval_math_samples_per_second": 13.975, |
| "eval_math_steps_per_second": 3.494, |
| "step": 2900 |
| }, |
| { |
| "epoch": 4.64, |
| "eval_physics_entropy": 0.2805333559513092, |
| "eval_physics_loss": 1.0783778429031372, |
| "eval_physics_mean_token_accuracy": 0.7986170620918274, |
| "eval_physics_num_tokens": 82990401.0, |
| "eval_physics_runtime": 41.499, |
| "eval_physics_samples_per_second": 12.048, |
| "eval_physics_steps_per_second": 3.012, |
| "step": 2900 |
| }, |
| { |
| "entropy": 0.20116342874243856, |
| "epoch": 4.656, |
| "grad_norm": 2.421875, |
| "learning_rate": 1.575777777777778e-05, |
| "loss": 0.1639, |
| "mean_token_accuracy": 0.9467363331466914, |
| "num_tokens": 83269847.0, |
| "step": 2910 |
| }, |
| { |
| "entropy": 0.20632342710159718, |
| "epoch": 4.672, |
| "grad_norm": 2.40625, |
| "learning_rate": 1.5735555555555557e-05, |
| "loss": 0.1709, |
| "mean_token_accuracy": 0.9442934289574623, |
| "num_tokens": 83550833.0, |
| "step": 2920 |
| }, |
| { |
| "entropy": 0.2032525334507227, |
| "epoch": 4.688, |
| "grad_norm": 2.15625, |
| "learning_rate": 1.5713333333333334e-05, |
| "loss": 0.1693, |
| "mean_token_accuracy": 0.9453159097582102, |
| "num_tokens": 83840860.0, |
| "step": 2930 |
| }, |
| { |
| "entropy": 0.20695003918372096, |
| "epoch": 4.704, |
| "grad_norm": 2.234375, |
| "learning_rate": 1.5691111111111112e-05, |
| "loss": 0.1735, |
| "mean_token_accuracy": 0.9437373761087656, |
| "num_tokens": 84128625.0, |
| "step": 2940 |
| }, |
| { |
| "entropy": 0.208350771619007, |
| "epoch": 4.72, |
| "grad_norm": 2.140625, |
| "learning_rate": 1.5668888888888893e-05, |
| "loss": 0.1736, |
| "mean_token_accuracy": 0.9436807181686163, |
| "num_tokens": 84423857.0, |
| "step": 2950 |
| }, |
| { |
| "entropy": 0.20460730344057082, |
| "epoch": 4.736, |
| "grad_norm": 2.359375, |
| "learning_rate": 1.5646666666666667e-05, |
| "loss": 0.1697, |
| "mean_token_accuracy": 0.9452991545200348, |
| "num_tokens": 84695402.0, |
| "step": 2960 |
| }, |
| { |
| "entropy": 0.19873600411228837, |
| "epoch": 4.752, |
| "grad_norm": 2.390625, |
| "learning_rate": 1.5624444444444448e-05, |
| "loss": 0.1635, |
| "mean_token_accuracy": 0.9467461623251439, |
| "num_tokens": 84980763.0, |
| "step": 2970 |
| }, |
| { |
| "entropy": 0.203874301770702, |
| "epoch": 4.768, |
| "grad_norm": 2.09375, |
| "learning_rate": 1.5602222222222222e-05, |
| "loss": 0.1679, |
| "mean_token_accuracy": 0.945096442475915, |
| "num_tokens": 85272835.0, |
| "step": 2980 |
| }, |
| { |
| "entropy": 0.2079331438988447, |
| "epoch": 4.784, |
| "grad_norm": 2.359375, |
| "learning_rate": 1.5580000000000003e-05, |
| "loss": 0.1767, |
| "mean_token_accuracy": 0.9423614408820867, |
| "num_tokens": 85568674.0, |
| "step": 2990 |
| }, |
| { |
| "entropy": 0.20461545852012933, |
| "epoch": 4.8, |
| "grad_norm": 2.09375, |
| "learning_rate": 1.5557777777777778e-05, |
| "loss": 0.1701, |
| "mean_token_accuracy": 0.9448350362479687, |
| "num_tokens": 85856157.0, |
| "step": 3000 |
| }, |
| { |
| "epoch": 4.8, |
| "eval_math_entropy": 0.29029106283187867, |
| "eval_math_loss": 0.9365277290344238, |
| "eval_math_mean_token_accuracy": 0.8171119575500488, |
| "eval_math_num_tokens": 85856157.0, |
| "eval_math_runtime": 35.799, |
| "eval_math_samples_per_second": 13.967, |
| "eval_math_steps_per_second": 3.492, |
| "step": 3000 |
| }, |
| { |
| "epoch": 4.8, |
| "eval_physics_entropy": 0.2788742837905884, |
| "eval_physics_loss": 1.0801695585250854, |
| "eval_physics_mean_token_accuracy": 0.7987381658554077, |
| "eval_physics_num_tokens": 85856157.0, |
| "eval_physics_runtime": 41.1763, |
| "eval_physics_samples_per_second": 12.143, |
| "eval_physics_steps_per_second": 3.036, |
| "step": 3000 |
| }, |
| { |
| "entropy": 0.2044725831132382, |
| "epoch": 4.816, |
| "grad_norm": 2.4375, |
| "learning_rate": 1.553555555555556e-05, |
| "loss": 0.1707, |
| "mean_token_accuracy": 0.9451114404946566, |
| "num_tokens": 86148164.0, |
| "step": 3010 |
| }, |
| { |
| "entropy": 0.20232565151527523, |
| "epoch": 4.832, |
| "grad_norm": 2.25, |
| "learning_rate": 1.5513333333333333e-05, |
| "loss": 0.1684, |
| "mean_token_accuracy": 0.9456779569387436, |
| "num_tokens": 86439959.0, |
| "step": 3020 |
| }, |
| { |
| "entropy": 0.20742189674638212, |
| "epoch": 4.848, |
| "grad_norm": 2.3125, |
| "learning_rate": 1.5491111111111114e-05, |
| "loss": 0.172, |
| "mean_token_accuracy": 0.9437137056142092, |
| "num_tokens": 86721776.0, |
| "step": 3030 |
| }, |
| { |
| "entropy": 0.204735844489187, |
| "epoch": 4.864, |
| "grad_norm": 2.28125, |
| "learning_rate": 1.546888888888889e-05, |
| "loss": 0.168, |
| "mean_token_accuracy": 0.9456948045641184, |
| "num_tokens": 87004064.0, |
| "step": 3040 |
| }, |
| { |
| "entropy": 0.20204079821705817, |
| "epoch": 4.88, |
| "grad_norm": 2.34375, |
| "learning_rate": 1.544666666666667e-05, |
| "loss": 0.1691, |
| "mean_token_accuracy": 0.9455896012485028, |
| "num_tokens": 87300720.0, |
| "step": 3050 |
| }, |
| { |
| "entropy": 0.20231212503276766, |
| "epoch": 4.896, |
| "grad_norm": 2.296875, |
| "learning_rate": 1.5424444444444447e-05, |
| "loss": 0.1673, |
| "mean_token_accuracy": 0.9458602901548148, |
| "num_tokens": 87593142.0, |
| "step": 3060 |
| }, |
| { |
| "entropy": 0.19865587293170392, |
| "epoch": 4.912, |
| "grad_norm": 2.046875, |
| "learning_rate": 1.5402222222222224e-05, |
| "loss": 0.1644, |
| "mean_token_accuracy": 0.9465924687683582, |
| "num_tokens": 87881200.0, |
| "step": 3070 |
| }, |
| { |
| "entropy": 0.201658932864666, |
| "epoch": 4.928, |
| "grad_norm": 2.296875, |
| "learning_rate": 1.5380000000000002e-05, |
| "loss": 0.1669, |
| "mean_token_accuracy": 0.9455430846661329, |
| "num_tokens": 88161276.0, |
| "step": 3080 |
| }, |
| { |
| "entropy": 0.20509861945174634, |
| "epoch": 4.944, |
| "grad_norm": 2.171875, |
| "learning_rate": 1.535777777777778e-05, |
| "loss": 0.1739, |
| "mean_token_accuracy": 0.9440295215696096, |
| "num_tokens": 88442647.0, |
| "step": 3090 |
| }, |
| { |
| "entropy": 0.20068702194839716, |
| "epoch": 4.96, |
| "grad_norm": 2.328125, |
| "learning_rate": 1.5335555555555557e-05, |
| "loss": 0.1638, |
| "mean_token_accuracy": 0.9464703939855099, |
| "num_tokens": 88727638.0, |
| "step": 3100 |
| }, |
| { |
| "epoch": 4.96, |
| "eval_math_entropy": 0.2880592755079269, |
| "eval_math_loss": 0.9388962388038635, |
| "eval_math_mean_token_accuracy": 0.8168627028465271, |
| "eval_math_num_tokens": 88727638.0, |
| "eval_math_runtime": 35.7707, |
| "eval_math_samples_per_second": 13.978, |
| "eval_math_steps_per_second": 3.494, |
| "step": 3100 |
| }, |
| { |
| "epoch": 4.96, |
| "eval_physics_entropy": 0.2795389975309372, |
| "eval_physics_loss": 1.0800063610076904, |
| "eval_physics_mean_token_accuracy": 0.7986403579711914, |
| "eval_physics_num_tokens": 88727638.0, |
| "eval_physics_runtime": 41.4648, |
| "eval_physics_samples_per_second": 12.058, |
| "eval_physics_steps_per_second": 3.015, |
| "step": 3100 |
| }, |
| { |
| "entropy": 0.20669097322970628, |
| "epoch": 4.976, |
| "grad_norm": 2.40625, |
| "learning_rate": 1.5313333333333335e-05, |
| "loss": 0.1734, |
| "mean_token_accuracy": 0.9433898750692606, |
| "num_tokens": 89014226.0, |
| "step": 3110 |
| }, |
| { |
| "entropy": 0.20596866989508272, |
| "epoch": 4.992, |
| "grad_norm": 2.328125, |
| "learning_rate": 1.5291111111111112e-05, |
| "loss": 0.1728, |
| "mean_token_accuracy": 0.9439191181212664, |
| "num_tokens": 89291373.0, |
| "step": 3120 |
| }, |
| { |
| "entropy": 0.17975524491630496, |
| "epoch": 5.008, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.526888888888889e-05, |
| "loss": 0.1294, |
| "mean_token_accuracy": 0.9600633040070534, |
| "num_tokens": 89575899.0, |
| "step": 3130 |
| }, |
| { |
| "entropy": 0.12737601874396204, |
| "epoch": 5.024, |
| "grad_norm": 1.5234375, |
| "learning_rate": 1.5246666666666668e-05, |
| "loss": 0.0799, |
| "mean_token_accuracy": 0.9776071958243847, |
| "num_tokens": 89860193.0, |
| "step": 3140 |
| }, |
| { |
| "entropy": 0.12302283146418631, |
| "epoch": 5.04, |
| "grad_norm": 1.546875, |
| "learning_rate": 1.5224444444444447e-05, |
| "loss": 0.0763, |
| "mean_token_accuracy": 0.9784069903194904, |
| "num_tokens": 90155473.0, |
| "step": 3150 |
| }, |
| { |
| "entropy": 0.1233028472866863, |
| "epoch": 5.056, |
| "grad_norm": 1.5703125, |
| "learning_rate": 1.5202222222222223e-05, |
| "loss": 0.076, |
| "mean_token_accuracy": 0.9784812565892935, |
| "num_tokens": 90447777.0, |
| "step": 3160 |
| }, |
| { |
| "entropy": 0.12666487065143883, |
| "epoch": 5.072, |
| "grad_norm": 1.5625, |
| "learning_rate": 1.5180000000000002e-05, |
| "loss": 0.0801, |
| "mean_token_accuracy": 0.9770117931067943, |
| "num_tokens": 90733638.0, |
| "step": 3170 |
| }, |
| { |
| "entropy": 0.12284240419976414, |
| "epoch": 5.088, |
| "grad_norm": 1.46875, |
| "learning_rate": 1.5157777777777778e-05, |
| "loss": 0.078, |
| "mean_token_accuracy": 0.9779970541596412, |
| "num_tokens": 91024199.0, |
| "step": 3180 |
| }, |
| { |
| "entropy": 0.11969875160139054, |
| "epoch": 5.104, |
| "grad_norm": 1.453125, |
| "learning_rate": 1.5135555555555557e-05, |
| "loss": 0.0758, |
| "mean_token_accuracy": 0.9783815696835518, |
| "num_tokens": 91314842.0, |
| "step": 3190 |
| }, |
| { |
| "entropy": 0.11972461631521583, |
| "epoch": 5.12, |
| "grad_norm": 1.53125, |
| "learning_rate": 1.5113333333333335e-05, |
| "loss": 0.0758, |
| "mean_token_accuracy": 0.9785178400576114, |
| "num_tokens": 91602437.0, |
| "step": 3200 |
| }, |
| { |
| "epoch": 5.12, |
| "eval_math_entropy": 0.24255421686172485, |
| "eval_math_loss": 1.0493830442428589, |
| "eval_math_mean_token_accuracy": 0.8152036366462707, |
| "eval_math_num_tokens": 91602437.0, |
| "eval_math_runtime": 35.752, |
| "eval_math_samples_per_second": 13.985, |
| "eval_math_steps_per_second": 3.496, |
| "step": 3200 |
| }, |
| { |
| "epoch": 5.12, |
| "eval_physics_entropy": 0.25365390503406526, |
| "eval_physics_loss": 1.161871314048767, |
| "eval_physics_mean_token_accuracy": 0.7974244031906128, |
| "eval_physics_num_tokens": 91602437.0, |
| "eval_physics_runtime": 41.4324, |
| "eval_physics_samples_per_second": 12.068, |
| "eval_physics_steps_per_second": 3.017, |
| "step": 3200 |
| }, |
| { |
| "entropy": 0.11904104156419634, |
| "epoch": 5.136, |
| "grad_norm": 1.5703125, |
| "learning_rate": 1.5091111111111113e-05, |
| "loss": 0.0766, |
| "mean_token_accuracy": 0.9780814602971077, |
| "num_tokens": 91887940.0, |
| "step": 3210 |
| }, |
| { |
| "entropy": 0.12109884731471539, |
| "epoch": 5.152, |
| "grad_norm": 1.6171875, |
| "learning_rate": 1.506888888888889e-05, |
| "loss": 0.0776, |
| "mean_token_accuracy": 0.9777121197432279, |
| "num_tokens": 92169269.0, |
| "step": 3220 |
| }, |
| { |
| "entropy": 0.11855497197248041, |
| "epoch": 5.168, |
| "grad_norm": 1.5703125, |
| "learning_rate": 1.5046666666666668e-05, |
| "loss": 0.0761, |
| "mean_token_accuracy": 0.978103457018733, |
| "num_tokens": 92463326.0, |
| "step": 3230 |
| }, |
| { |
| "entropy": 0.12662705583497882, |
| "epoch": 5.184, |
| "grad_norm": 1.5390625, |
| "learning_rate": 1.5024444444444445e-05, |
| "loss": 0.081, |
| "mean_token_accuracy": 0.9766223061829805, |
| "num_tokens": 92753062.0, |
| "step": 3240 |
| }, |
| { |
| "entropy": 0.12166502382606267, |
| "epoch": 5.2, |
| "grad_norm": 1.6015625, |
| "learning_rate": 1.5002222222222223e-05, |
| "loss": 0.0773, |
| "mean_token_accuracy": 0.9779473695904016, |
| "num_tokens": 93039536.0, |
| "step": 3250 |
| }, |
| { |
| "entropy": 0.12320108511485159, |
| "epoch": 5.216, |
| "grad_norm": 1.6328125, |
| "learning_rate": 1.498e-05, |
| "loss": 0.0801, |
| "mean_token_accuracy": 0.9769222058355809, |
| "num_tokens": 93324998.0, |
| "step": 3260 |
| }, |
| { |
| "entropy": 0.12509258496575057, |
| "epoch": 5.232, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.495777777777778e-05, |
| "loss": 0.0811, |
| "mean_token_accuracy": 0.9771292954683304, |
| "num_tokens": 93607538.0, |
| "step": 3270 |
| }, |
| { |
| "entropy": 0.12425780035555363, |
| "epoch": 5.248, |
| "grad_norm": 1.8515625, |
| "learning_rate": 1.4935555555555556e-05, |
| "loss": 0.08, |
| "mean_token_accuracy": 0.976675358787179, |
| "num_tokens": 93889153.0, |
| "step": 3280 |
| }, |
| { |
| "entropy": 0.12271388149820268, |
| "epoch": 5.264, |
| "grad_norm": 1.5625, |
| "learning_rate": 1.4913333333333335e-05, |
| "loss": 0.0789, |
| "mean_token_accuracy": 0.977228007465601, |
| "num_tokens": 94174451.0, |
| "step": 3290 |
| }, |
| { |
| "entropy": 0.1203123148996383, |
| "epoch": 5.28, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.4891111111111111e-05, |
| "loss": 0.0771, |
| "mean_token_accuracy": 0.9775407303124666, |
| "num_tokens": 94463178.0, |
| "step": 3300 |
| }, |
| { |
| "epoch": 5.28, |
| "eval_math_entropy": 0.23942122876644134, |
| "eval_math_loss": 1.0654295682907104, |
| "eval_math_mean_token_accuracy": 0.8145109858512879, |
| "eval_math_num_tokens": 94463178.0, |
| "eval_math_runtime": 35.7653, |
| "eval_math_samples_per_second": 13.98, |
| "eval_math_steps_per_second": 3.495, |
| "step": 3300 |
| }, |
| { |
| "epoch": 5.28, |
| "eval_physics_entropy": 0.25065979206562045, |
| "eval_physics_loss": 1.1764365434646606, |
| "eval_physics_mean_token_accuracy": 0.7967135186195373, |
| "eval_physics_num_tokens": 94463178.0, |
| "eval_physics_runtime": 41.4451, |
| "eval_physics_samples_per_second": 12.064, |
| "eval_physics_steps_per_second": 3.016, |
| "step": 3300 |
| }, |
| { |
| "entropy": 0.12212947504594922, |
| "epoch": 5.296, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.486888888888889e-05, |
| "loss": 0.0797, |
| "mean_token_accuracy": 0.9769365161657333, |
| "num_tokens": 94754637.0, |
| "step": 3310 |
| }, |
| { |
| "entropy": 0.12625247961841524, |
| "epoch": 5.312, |
| "grad_norm": 1.8125, |
| "learning_rate": 1.4846666666666666e-05, |
| "loss": 0.0833, |
| "mean_token_accuracy": 0.9761669352650643, |
| "num_tokens": 95034199.0, |
| "step": 3320 |
| }, |
| { |
| "entropy": 0.12353909476660192, |
| "epoch": 5.328, |
| "grad_norm": 1.890625, |
| "learning_rate": 1.4824444444444446e-05, |
| "loss": 0.0809, |
| "mean_token_accuracy": 0.9766261216253043, |
| "num_tokens": 95324901.0, |
| "step": 3330 |
| }, |
| { |
| "entropy": 0.12188207178842277, |
| "epoch": 5.344, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.4802222222222225e-05, |
| "loss": 0.0782, |
| "mean_token_accuracy": 0.9771796323359012, |
| "num_tokens": 95615328.0, |
| "step": 3340 |
| }, |
| { |
| "entropy": 0.1235438220668584, |
| "epoch": 5.36, |
| "grad_norm": 1.796875, |
| "learning_rate": 1.478e-05, |
| "loss": 0.0818, |
| "mean_token_accuracy": 0.9761600013822317, |
| "num_tokens": 95906341.0, |
| "step": 3350 |
| }, |
| { |
| "entropy": 0.1280383275821805, |
| "epoch": 5.376, |
| "grad_norm": 1.6328125, |
| "learning_rate": 1.475777777777778e-05, |
| "loss": 0.0827, |
| "mean_token_accuracy": 0.9760391939431429, |
| "num_tokens": 96188769.0, |
| "step": 3360 |
| }, |
| { |
| "entropy": 0.12412518821656704, |
| "epoch": 5.392, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.4735555555555556e-05, |
| "loss": 0.082, |
| "mean_token_accuracy": 0.9760548025369644, |
| "num_tokens": 96472047.0, |
| "step": 3370 |
| }, |
| { |
| "entropy": 0.12685924135148524, |
| "epoch": 5.408, |
| "grad_norm": 1.5546875, |
| "learning_rate": 1.4713333333333335e-05, |
| "loss": 0.0828, |
| "mean_token_accuracy": 0.9758801158517599, |
| "num_tokens": 96752754.0, |
| "step": 3380 |
| }, |
| { |
| "entropy": 0.12485403921455145, |
| "epoch": 5.424, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.4691111111111111e-05, |
| "loss": 0.0829, |
| "mean_token_accuracy": 0.9760515049099923, |
| "num_tokens": 97042615.0, |
| "step": 3390 |
| }, |
| { |
| "entropy": 0.12233043625019491, |
| "epoch": 5.44, |
| "grad_norm": 1.65625, |
| "learning_rate": 1.466888888888889e-05, |
| "loss": 0.0815, |
| "mean_token_accuracy": 0.9764630265533925, |
| "num_tokens": 97324337.0, |
| "step": 3400 |
| }, |
| { |
| "epoch": 5.44, |
| "eval_math_entropy": 0.24116922390460968, |
| "eval_math_loss": 1.0628420114517212, |
| "eval_math_mean_token_accuracy": 0.8142255573272705, |
| "eval_math_num_tokens": 97324337.0, |
| "eval_math_runtime": 35.7647, |
| "eval_math_samples_per_second": 13.98, |
| "eval_math_steps_per_second": 3.495, |
| "step": 3400 |
| }, |
| { |
| "epoch": 5.44, |
| "eval_physics_entropy": 0.2520143152475357, |
| "eval_physics_loss": 1.171416997909546, |
| "eval_physics_mean_token_accuracy": 0.7969081449508667, |
| "eval_physics_num_tokens": 97324337.0, |
| "eval_physics_runtime": 41.7765, |
| "eval_physics_samples_per_second": 11.968, |
| "eval_physics_steps_per_second": 2.992, |
| "step": 3400 |
| }, |
| { |
| "entropy": 0.12518517016433178, |
| "epoch": 5.456, |
| "grad_norm": 1.8515625, |
| "learning_rate": 1.4646666666666666e-05, |
| "loss": 0.0819, |
| "mean_token_accuracy": 0.975970969721675, |
| "num_tokens": 97610848.0, |
| "step": 3410 |
| }, |
| { |
| "entropy": 0.12347186738625168, |
| "epoch": 5.4719999999999995, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.4624444444444446e-05, |
| "loss": 0.0809, |
| "mean_token_accuracy": 0.9765568733215332, |
| "num_tokens": 97894197.0, |
| "step": 3420 |
| }, |
| { |
| "entropy": 0.12211509156040848, |
| "epoch": 5.4879999999999995, |
| "grad_norm": 1.6015625, |
| "learning_rate": 1.4602222222222225e-05, |
| "loss": 0.0801, |
| "mean_token_accuracy": 0.9765133555978537, |
| "num_tokens": 98178647.0, |
| "step": 3430 |
| }, |
| { |
| "entropy": 0.1255460273474455, |
| "epoch": 5.504, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.4580000000000001e-05, |
| "loss": 0.0832, |
| "mean_token_accuracy": 0.975815936923027, |
| "num_tokens": 98466145.0, |
| "step": 3440 |
| }, |
| { |
| "entropy": 0.12366767269559205, |
| "epoch": 5.52, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.455777777777778e-05, |
| "loss": 0.0833, |
| "mean_token_accuracy": 0.9755478825420141, |
| "num_tokens": 98747917.0, |
| "step": 3450 |
| }, |
| { |
| "entropy": 0.1271245769225061, |
| "epoch": 5.536, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.4535555555555556e-05, |
| "loss": 0.0854, |
| "mean_token_accuracy": 0.975119874253869, |
| "num_tokens": 99030810.0, |
| "step": 3460 |
| }, |
| { |
| "entropy": 0.12403408349491656, |
| "epoch": 5.552, |
| "grad_norm": 1.8828125, |
| "learning_rate": 1.4513333333333336e-05, |
| "loss": 0.0829, |
| "mean_token_accuracy": 0.9762228842824697, |
| "num_tokens": 99318677.0, |
| "step": 3470 |
| }, |
| { |
| "entropy": 0.1286336882505566, |
| "epoch": 5.568, |
| "grad_norm": 1.84375, |
| "learning_rate": 1.4491111111111111e-05, |
| "loss": 0.084, |
| "mean_token_accuracy": 0.9754358228296042, |
| "num_tokens": 99592537.0, |
| "step": 3480 |
| }, |
| { |
| "entropy": 0.12519475985318423, |
| "epoch": 5.584, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.446888888888889e-05, |
| "loss": 0.0836, |
| "mean_token_accuracy": 0.9757570996880531, |
| "num_tokens": 99875992.0, |
| "step": 3490 |
| }, |
| { |
| "entropy": 0.12342443629167973, |
| "epoch": 5.6, |
| "grad_norm": 1.65625, |
| "learning_rate": 1.4446666666666668e-05, |
| "loss": 0.0802, |
| "mean_token_accuracy": 0.9769513830542564, |
| "num_tokens": 100168027.0, |
| "step": 3500 |
| }, |
| { |
| "epoch": 5.6, |
| "eval_math_entropy": 0.23975398194789888, |
| "eval_math_loss": 1.071073055267334, |
| "eval_math_mean_token_accuracy": 0.814323438167572, |
| "eval_math_num_tokens": 100168027.0, |
| "eval_math_runtime": 35.7663, |
| "eval_math_samples_per_second": 13.98, |
| "eval_math_steps_per_second": 3.495, |
| "step": 3500 |
| }, |
| { |
| "epoch": 5.6, |
| "eval_physics_entropy": 0.2488333419561386, |
| "eval_physics_loss": 1.1844922304153442, |
| "eval_physics_mean_token_accuracy": 0.7968628706932068, |
| "eval_physics_num_tokens": 100168027.0, |
| "eval_physics_runtime": 41.1644, |
| "eval_physics_samples_per_second": 12.146, |
| "eval_physics_steps_per_second": 3.037, |
| "step": 3500 |
| }, |
| { |
| "entropy": 0.11995140444487333, |
| "epoch": 5.616, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.4424444444444446e-05, |
| "loss": 0.0804, |
| "mean_token_accuracy": 0.976571949198842, |
| "num_tokens": 100458800.0, |
| "step": 3510 |
| }, |
| { |
| "entropy": 0.12303966791369021, |
| "epoch": 5.632, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.4402222222222224e-05, |
| "loss": 0.083, |
| "mean_token_accuracy": 0.9757887721061707, |
| "num_tokens": 100750138.0, |
| "step": 3520 |
| }, |
| { |
| "entropy": 0.12365907938219607, |
| "epoch": 5.648, |
| "grad_norm": 1.8671875, |
| "learning_rate": 1.4380000000000001e-05, |
| "loss": 0.0815, |
| "mean_token_accuracy": 0.9761408776044845, |
| "num_tokens": 101035408.0, |
| "step": 3530 |
| }, |
| { |
| "entropy": 0.12068781261332333, |
| "epoch": 5.664, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.4357777777777779e-05, |
| "loss": 0.0801, |
| "mean_token_accuracy": 0.9766639798879624, |
| "num_tokens": 101331186.0, |
| "step": 3540 |
| }, |
| { |
| "entropy": 0.12283993698656559, |
| "epoch": 5.68, |
| "grad_norm": 1.9609375, |
| "learning_rate": 1.4335555555555556e-05, |
| "loss": 0.0805, |
| "mean_token_accuracy": 0.9763456650078297, |
| "num_tokens": 101621993.0, |
| "step": 3550 |
| }, |
| { |
| "entropy": 0.12474635057151318, |
| "epoch": 5.696, |
| "grad_norm": 1.8515625, |
| "learning_rate": 1.4313333333333334e-05, |
| "loss": 0.0825, |
| "mean_token_accuracy": 0.9756041605025529, |
| "num_tokens": 101904112.0, |
| "step": 3560 |
| }, |
| { |
| "entropy": 0.12090231254696845, |
| "epoch": 5.712, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.4291111111111113e-05, |
| "loss": 0.082, |
| "mean_token_accuracy": 0.9761635400354862, |
| "num_tokens": 102196831.0, |
| "step": 3570 |
| }, |
| { |
| "entropy": 0.12512639928609132, |
| "epoch": 5.728, |
| "grad_norm": 2.0, |
| "learning_rate": 1.426888888888889e-05, |
| "loss": 0.0839, |
| "mean_token_accuracy": 0.9757933501154185, |
| "num_tokens": 102482122.0, |
| "step": 3580 |
| }, |
| { |
| "entropy": 0.12669197719078512, |
| "epoch": 5.744, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.4246666666666669e-05, |
| "loss": 0.0845, |
| "mean_token_accuracy": 0.9752174407243729, |
| "num_tokens": 102754317.0, |
| "step": 3590 |
| }, |
| { |
| "entropy": 0.12015235424041748, |
| "epoch": 5.76, |
| "grad_norm": 1.6015625, |
| "learning_rate": 1.4224444444444445e-05, |
| "loss": 0.0795, |
| "mean_token_accuracy": 0.9764200799167156, |
| "num_tokens": 103040247.0, |
| "step": 3600 |
| }, |
| { |
| "epoch": 5.76, |
| "eval_math_entropy": 0.23906498980522156, |
| "eval_math_loss": 1.0675584077835083, |
| "eval_math_mean_token_accuracy": 0.8141876225471496, |
| "eval_math_num_tokens": 103040247.0, |
| "eval_math_runtime": 35.8224, |
| "eval_math_samples_per_second": 13.958, |
| "eval_math_steps_per_second": 3.489, |
| "step": 3600 |
| }, |
| { |
| "epoch": 5.76, |
| "eval_physics_entropy": 0.24975030255317687, |
| "eval_physics_loss": 1.178022027015686, |
| "eval_physics_mean_token_accuracy": 0.7968517804145813, |
| "eval_physics_num_tokens": 103040247.0, |
| "eval_physics_runtime": 41.4817, |
| "eval_physics_samples_per_second": 12.054, |
| "eval_physics_steps_per_second": 3.013, |
| "step": 3600 |
| }, |
| { |
| "entropy": 0.12267617424950003, |
| "epoch": 5.776, |
| "grad_norm": 1.8046875, |
| "learning_rate": 1.4202222222222224e-05, |
| "loss": 0.0814, |
| "mean_token_accuracy": 0.9759464427828789, |
| "num_tokens": 103322676.0, |
| "step": 3610 |
| }, |
| { |
| "entropy": 0.12787257870659233, |
| "epoch": 5.792, |
| "grad_norm": 1.9296875, |
| "learning_rate": 1.418e-05, |
| "loss": 0.0866, |
| "mean_token_accuracy": 0.9747348535805941, |
| "num_tokens": 103601932.0, |
| "step": 3620 |
| }, |
| { |
| "entropy": 0.12417071931995452, |
| "epoch": 5.808, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.4157777777777779e-05, |
| "loss": 0.0832, |
| "mean_token_accuracy": 0.9756274610757828, |
| "num_tokens": 103899790.0, |
| "step": 3630 |
| }, |
| { |
| "entropy": 0.12282869541086257, |
| "epoch": 5.824, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.4135555555555555e-05, |
| "loss": 0.0815, |
| "mean_token_accuracy": 0.9763524547219277, |
| "num_tokens": 104178127.0, |
| "step": 3640 |
| }, |
| { |
| "entropy": 0.12581051201559604, |
| "epoch": 5.84, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.4113333333333334e-05, |
| "loss": 0.0853, |
| "mean_token_accuracy": 0.9752345487475396, |
| "num_tokens": 104468018.0, |
| "step": 3650 |
| }, |
| { |
| "entropy": 0.12290314708370716, |
| "epoch": 5.856, |
| "grad_norm": 2.0, |
| "learning_rate": 1.4091111111111114e-05, |
| "loss": 0.0824, |
| "mean_token_accuracy": 0.9761314608156682, |
| "num_tokens": 104760904.0, |
| "step": 3660 |
| }, |
| { |
| "entropy": 0.12341588027775288, |
| "epoch": 5.872, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.406888888888889e-05, |
| "loss": 0.0819, |
| "mean_token_accuracy": 0.9755884770303964, |
| "num_tokens": 105040296.0, |
| "step": 3670 |
| }, |
| { |
| "entropy": 0.12345146541483701, |
| "epoch": 5.888, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.4046666666666669e-05, |
| "loss": 0.0829, |
| "mean_token_accuracy": 0.9762541994452476, |
| "num_tokens": 105334020.0, |
| "step": 3680 |
| }, |
| { |
| "entropy": 0.12768929479643704, |
| "epoch": 5.904, |
| "grad_norm": 1.8359375, |
| "learning_rate": 1.4024444444444445e-05, |
| "loss": 0.0844, |
| "mean_token_accuracy": 0.9747776690870523, |
| "num_tokens": 105618594.0, |
| "step": 3690 |
| }, |
| { |
| "entropy": 0.1248483968898654, |
| "epoch": 5.92, |
| "grad_norm": 2.046875, |
| "learning_rate": 1.4002222222222224e-05, |
| "loss": 0.0842, |
| "mean_token_accuracy": 0.9751092988997698, |
| "num_tokens": 105908769.0, |
| "step": 3700 |
| }, |
| { |
| "epoch": 5.92, |
| "eval_math_entropy": 0.23840607726573945, |
| "eval_math_loss": 1.0716767311096191, |
| "eval_math_mean_token_accuracy": 0.8149038062095643, |
| "eval_math_num_tokens": 105908769.0, |
| "eval_math_runtime": 35.7392, |
| "eval_math_samples_per_second": 13.99, |
| "eval_math_steps_per_second": 3.498, |
| "step": 3700 |
| }, |
| { |
| "epoch": 5.92, |
| "eval_physics_entropy": 0.24841285252571105, |
| "eval_physics_loss": 1.184152603149414, |
| "eval_physics_mean_token_accuracy": 0.7970950255393981, |
| "eval_physics_num_tokens": 105908769.0, |
| "eval_physics_runtime": 41.4426, |
| "eval_physics_samples_per_second": 12.065, |
| "eval_physics_steps_per_second": 3.016, |
| "step": 3700 |
| }, |
| { |
| "entropy": 0.12872918481007217, |
| "epoch": 5.936, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.398e-05, |
| "loss": 0.0859, |
| "mean_token_accuracy": 0.9748044356703758, |
| "num_tokens": 106184251.0, |
| "step": 3710 |
| }, |
| { |
| "entropy": 0.12503819265402855, |
| "epoch": 5.952, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.395777777777778e-05, |
| "loss": 0.0827, |
| "mean_token_accuracy": 0.9756678476929664, |
| "num_tokens": 106475238.0, |
| "step": 3720 |
| }, |
| { |
| "entropy": 0.12310672523453832, |
| "epoch": 5.968, |
| "grad_norm": 1.84375, |
| "learning_rate": 1.3935555555555557e-05, |
| "loss": 0.0832, |
| "mean_token_accuracy": 0.9751558441668748, |
| "num_tokens": 106761563.0, |
| "step": 3730 |
| }, |
| { |
| "entropy": 0.12658624798059465, |
| "epoch": 5.984, |
| "grad_norm": 1.8203125, |
| "learning_rate": 1.3913333333333335e-05, |
| "loss": 0.0848, |
| "mean_token_accuracy": 0.9749173305928707, |
| "num_tokens": 107044983.0, |
| "step": 3740 |
| }, |
| { |
| "entropy": 0.1277452927082777, |
| "epoch": 6.0, |
| "grad_norm": 1.7890625, |
| "learning_rate": 1.3891111111111114e-05, |
| "loss": 0.0857, |
| "mean_token_accuracy": 0.9748894535005093, |
| "num_tokens": 107318544.0, |
| "step": 3750 |
| }, |
| { |
| "entropy": 0.08838868229649961, |
| "epoch": 6.016, |
| "grad_norm": 2.265625, |
| "learning_rate": 1.386888888888889e-05, |
| "loss": 0.0488, |
| "mean_token_accuracy": 0.9871940586715937, |
| "num_tokens": 107597955.0, |
| "step": 3760 |
| }, |
| { |
| "entropy": 0.08257517793681472, |
| "epoch": 6.032, |
| "grad_norm": 1.203125, |
| "learning_rate": 1.3846666666666669e-05, |
| "loss": 0.0478, |
| "mean_token_accuracy": 0.9871934026479721, |
| "num_tokens": 107882792.0, |
| "step": 3770 |
| }, |
| { |
| "entropy": 0.0828174332389608, |
| "epoch": 6.048, |
| "grad_norm": 1.203125, |
| "learning_rate": 1.3824444444444445e-05, |
| "loss": 0.0465, |
| "mean_token_accuracy": 0.9873949829488993, |
| "num_tokens": 108172211.0, |
| "step": 3780 |
| }, |
| { |
| "entropy": 0.0837429367704317, |
| "epoch": 6.064, |
| "grad_norm": 1.3671875, |
| "learning_rate": 1.3802222222222224e-05, |
| "loss": 0.0476, |
| "mean_token_accuracy": 0.9872307199984789, |
| "num_tokens": 108456311.0, |
| "step": 3790 |
| }, |
| { |
| "entropy": 0.08258170303888619, |
| "epoch": 6.08, |
| "grad_norm": 0.90234375, |
| "learning_rate": 1.378e-05, |
| "loss": 0.0469, |
| "mean_token_accuracy": 0.9875688355416059, |
| "num_tokens": 108741840.0, |
| "step": 3800 |
| }, |
| { |
| "epoch": 6.08, |
| "eval_math_entropy": 0.2157260871529579, |
| "eval_math_loss": 1.1480746269226074, |
| "eval_math_mean_token_accuracy": 0.8150639281272888, |
| "eval_math_num_tokens": 108741840.0, |
| "eval_math_runtime": 35.7792, |
| "eval_math_samples_per_second": 13.975, |
| "eval_math_steps_per_second": 3.494, |
| "step": 3800 |
| }, |
| { |
| "epoch": 6.08, |
| "eval_physics_entropy": 0.2341501749753952, |
| "eval_physics_loss": 1.242053747177124, |
| "eval_physics_mean_token_accuracy": 0.7970363121032715, |
| "eval_physics_num_tokens": 108741840.0, |
| "eval_physics_runtime": 41.45, |
| "eval_physics_samples_per_second": 12.063, |
| "eval_physics_steps_per_second": 3.016, |
| "step": 3800 |
| }, |
| { |
| "entropy": 0.08380906586535275, |
| "epoch": 6.096, |
| "grad_norm": 1.0078125, |
| "learning_rate": 1.375777777777778e-05, |
| "loss": 0.0485, |
| "mean_token_accuracy": 0.9867967016994953, |
| "num_tokens": 109019117.0, |
| "step": 3810 |
| }, |
| { |
| "entropy": 0.08307188230101019, |
| "epoch": 6.112, |
| "grad_norm": 1.0546875, |
| "learning_rate": 1.3735555555555557e-05, |
| "loss": 0.048, |
| "mean_token_accuracy": 0.9867318440228701, |
| "num_tokens": 109311190.0, |
| "step": 3820 |
| }, |
| { |
| "entropy": 0.08322459494229406, |
| "epoch": 6.128, |
| "grad_norm": 0.9296875, |
| "learning_rate": 1.3713333333333335e-05, |
| "loss": 0.0488, |
| "mean_token_accuracy": 0.986859867721796, |
| "num_tokens": 109597456.0, |
| "step": 3830 |
| }, |
| { |
| "entropy": 0.0813226085389033, |
| "epoch": 6.144, |
| "grad_norm": 1.15625, |
| "learning_rate": 1.3691111111111112e-05, |
| "loss": 0.0473, |
| "mean_token_accuracy": 0.9870293717831373, |
| "num_tokens": 109889318.0, |
| "step": 3840 |
| }, |
| { |
| "entropy": 0.08239689606707543, |
| "epoch": 6.16, |
| "grad_norm": 1.171875, |
| "learning_rate": 1.366888888888889e-05, |
| "loss": 0.049, |
| "mean_token_accuracy": 0.9869503367692232, |
| "num_tokens": 110172819.0, |
| "step": 3850 |
| }, |
| { |
| "entropy": 0.08577790288254619, |
| "epoch": 6.176, |
| "grad_norm": 1.3828125, |
| "learning_rate": 1.3646666666666668e-05, |
| "loss": 0.0492, |
| "mean_token_accuracy": 0.9867170631885529, |
| "num_tokens": 110459481.0, |
| "step": 3860 |
| }, |
| { |
| "entropy": 0.08229333041235805, |
| "epoch": 6.192, |
| "grad_norm": 1.03125, |
| "learning_rate": 1.3624444444444445e-05, |
| "loss": 0.049, |
| "mean_token_accuracy": 0.9867132067680359, |
| "num_tokens": 110747009.0, |
| "step": 3870 |
| }, |
| { |
| "entropy": 0.08104511541314423, |
| "epoch": 6.208, |
| "grad_norm": 1.1328125, |
| "learning_rate": 1.3602222222222223e-05, |
| "loss": 0.0477, |
| "mean_token_accuracy": 0.9871339503675699, |
| "num_tokens": 111034380.0, |
| "step": 3880 |
| }, |
| { |
| "entropy": 0.08298772852867842, |
| "epoch": 6.224, |
| "grad_norm": 1.203125, |
| "learning_rate": 1.3580000000000002e-05, |
| "loss": 0.0489, |
| "mean_token_accuracy": 0.9868796184659004, |
| "num_tokens": 111326385.0, |
| "step": 3890 |
| }, |
| { |
| "entropy": 0.08269706610590219, |
| "epoch": 6.24, |
| "grad_norm": 1.0546875, |
| "learning_rate": 1.3557777777777778e-05, |
| "loss": 0.0481, |
| "mean_token_accuracy": 0.9872248943895101, |
| "num_tokens": 111617177.0, |
| "step": 3900 |
| }, |
| { |
| "epoch": 6.24, |
| "eval_math_entropy": 0.2116558358669281, |
| "eval_math_loss": 1.1540143489837646, |
| "eval_math_mean_token_accuracy": 0.8150960731506348, |
| "eval_math_num_tokens": 111617177.0, |
| "eval_math_runtime": 35.7843, |
| "eval_math_samples_per_second": 13.973, |
| "eval_math_steps_per_second": 3.493, |
| "step": 3900 |
| }, |
| { |
| "epoch": 6.24, |
| "eval_physics_entropy": 0.23068565714359285, |
| "eval_physics_loss": 1.244989275932312, |
| "eval_physics_mean_token_accuracy": 0.797289421081543, |
| "eval_physics_num_tokens": 111617177.0, |
| "eval_physics_runtime": 41.443, |
| "eval_physics_samples_per_second": 12.065, |
| "eval_physics_steps_per_second": 3.016, |
| "step": 3900 |
| }, |
| { |
| "entropy": 0.08489166686777025, |
| "epoch": 6.256, |
| "grad_norm": 1.1875, |
| "learning_rate": 1.3535555555555557e-05, |
| "loss": 0.0506, |
| "mean_token_accuracy": 0.9859484441578388, |
| "num_tokens": 111893281.0, |
| "step": 3910 |
| }, |
| { |
| "entropy": 0.08082030359655619, |
| "epoch": 6.272, |
| "grad_norm": 1.15625, |
| "learning_rate": 1.3513333333333333e-05, |
| "loss": 0.0476, |
| "mean_token_accuracy": 0.9871520031243562, |
| "num_tokens": 112189253.0, |
| "step": 3920 |
| }, |
| { |
| "entropy": 0.08449386192951351, |
| "epoch": 6.288, |
| "grad_norm": 1.15625, |
| "learning_rate": 1.3491111111111113e-05, |
| "loss": 0.0498, |
| "mean_token_accuracy": 0.986621393635869, |
| "num_tokens": 112473705.0, |
| "step": 3930 |
| }, |
| { |
| "entropy": 0.08417739674914629, |
| "epoch": 6.304, |
| "grad_norm": 0.94921875, |
| "learning_rate": 1.3468888888888888e-05, |
| "loss": 0.0503, |
| "mean_token_accuracy": 0.9866305027157068, |
| "num_tokens": 112753183.0, |
| "step": 3940 |
| }, |
| { |
| "entropy": 0.08426821974571794, |
| "epoch": 6.32, |
| "grad_norm": 1.0546875, |
| "learning_rate": 1.3446666666666668e-05, |
| "loss": 0.0496, |
| "mean_token_accuracy": 0.9866181764751673, |
| "num_tokens": 113040028.0, |
| "step": 3950 |
| }, |
| { |
| "entropy": 0.08344817301258445, |
| "epoch": 6.336, |
| "grad_norm": 1.0859375, |
| "learning_rate": 1.3424444444444447e-05, |
| "loss": 0.05, |
| "mean_token_accuracy": 0.9863654665648938, |
| "num_tokens": 113323947.0, |
| "step": 3960 |
| }, |
| { |
| "entropy": 0.08314311665017157, |
| "epoch": 6.352, |
| "grad_norm": 1.2578125, |
| "learning_rate": 1.3402222222222223e-05, |
| "loss": 0.0507, |
| "mean_token_accuracy": 0.9866466347128153, |
| "num_tokens": 113603514.0, |
| "step": 3970 |
| }, |
| { |
| "entropy": 0.08517711046151817, |
| "epoch": 6.368, |
| "grad_norm": 1.203125, |
| "learning_rate": 1.3380000000000002e-05, |
| "loss": 0.0498, |
| "mean_token_accuracy": 0.9866021793335676, |
| "num_tokens": 113894528.0, |
| "step": 3980 |
| }, |
| { |
| "entropy": 0.08286653154063969, |
| "epoch": 6.384, |
| "grad_norm": 1.0859375, |
| "learning_rate": 1.3357777777777778e-05, |
| "loss": 0.0498, |
| "mean_token_accuracy": 0.986630416661501, |
| "num_tokens": 114172836.0, |
| "step": 3990 |
| }, |
| { |
| "entropy": 0.082943707681261, |
| "epoch": 6.4, |
| "grad_norm": 1.1171875, |
| "learning_rate": 1.3335555555555558e-05, |
| "loss": 0.0491, |
| "mean_token_accuracy": 0.9870663437992334, |
| "num_tokens": 114460847.0, |
| "step": 4000 |
| }, |
| { |
| "epoch": 6.4, |
| "eval_math_entropy": 0.21333465337753296, |
| "eval_math_loss": 1.1600124835968018, |
| "eval_math_mean_token_accuracy": 0.815093065738678, |
| "eval_math_num_tokens": 114460847.0, |
| "eval_math_runtime": 35.7657, |
| "eval_math_samples_per_second": 13.98, |
| "eval_math_steps_per_second": 3.495, |
| "step": 4000 |
| }, |
| { |
| "epoch": 6.4, |
| "eval_physics_entropy": 0.23318842422962188, |
| "eval_physics_loss": 1.2491592168807983, |
| "eval_physics_mean_token_accuracy": 0.7966313948631286, |
| "eval_physics_num_tokens": 114460847.0, |
| "eval_physics_runtime": 41.22, |
| "eval_physics_samples_per_second": 12.13, |
| "eval_physics_steps_per_second": 3.033, |
| "step": 4000 |
| }, |
| { |
| "entropy": 0.08455195005517453, |
| "epoch": 6.416, |
| "grad_norm": 1.09375, |
| "learning_rate": 1.3313333333333333e-05, |
| "loss": 0.0504, |
| "mean_token_accuracy": 0.9863806873559952, |
| "num_tokens": 114744933.0, |
| "step": 4010 |
| }, |
| { |
| "entropy": 0.08246206922922283, |
| "epoch": 6.432, |
| "grad_norm": 1.03125, |
| "learning_rate": 1.3291111111111113e-05, |
| "loss": 0.0486, |
| "mean_token_accuracy": 0.9869922034442424, |
| "num_tokens": 115035799.0, |
| "step": 4020 |
| }, |
| { |
| "entropy": 0.08484494448639453, |
| "epoch": 6.448, |
| "grad_norm": 1.0390625, |
| "learning_rate": 1.3268888888888889e-05, |
| "loss": 0.0497, |
| "mean_token_accuracy": 0.9866210386157036, |
| "num_tokens": 115324970.0, |
| "step": 4030 |
| }, |
| { |
| "entropy": 0.08431054749526083, |
| "epoch": 6.464, |
| "grad_norm": 1.265625, |
| "learning_rate": 1.3246666666666668e-05, |
| "loss": 0.0502, |
| "mean_token_accuracy": 0.9865565162152052, |
| "num_tokens": 115617802.0, |
| "step": 4040 |
| }, |
| { |
| "entropy": 0.0853661192348227, |
| "epoch": 6.48, |
| "grad_norm": 1.421875, |
| "learning_rate": 1.3224444444444446e-05, |
| "loss": 0.051, |
| "mean_token_accuracy": 0.9861403401941061, |
| "num_tokens": 115895920.0, |
| "step": 4050 |
| }, |
| { |
| "entropy": 0.0857542624231428, |
| "epoch": 6.496, |
| "grad_norm": 1.21875, |
| "learning_rate": 1.3202222222222223e-05, |
| "loss": 0.0505, |
| "mean_token_accuracy": 0.9863781563937664, |
| "num_tokens": 116183103.0, |
| "step": 4060 |
| }, |
| { |
| "entropy": 0.08211216740310193, |
| "epoch": 6.5120000000000005, |
| "grad_norm": 1.046875, |
| "learning_rate": 1.3180000000000001e-05, |
| "loss": 0.0489, |
| "mean_token_accuracy": 0.9870047964155674, |
| "num_tokens": 116470867.0, |
| "step": 4070 |
| }, |
| { |
| "entropy": 0.08192372117191553, |
| "epoch": 6.5280000000000005, |
| "grad_norm": 1.3984375, |
| "learning_rate": 1.3157777777777778e-05, |
| "loss": 0.0488, |
| "mean_token_accuracy": 0.9866585023701191, |
| "num_tokens": 116761140.0, |
| "step": 4080 |
| }, |
| { |
| "entropy": 0.08760172158945352, |
| "epoch": 6.5440000000000005, |
| "grad_norm": 1.1640625, |
| "learning_rate": 1.3135555555555558e-05, |
| "loss": 0.0513, |
| "mean_token_accuracy": 0.9858183633536101, |
| "num_tokens": 117044558.0, |
| "step": 4090 |
| }, |
| { |
| "entropy": 0.0826570155331865, |
| "epoch": 6.5600000000000005, |
| "grad_norm": 1.15625, |
| "learning_rate": 1.3113333333333334e-05, |
| "loss": 0.05, |
| "mean_token_accuracy": 0.9865829836577177, |
| "num_tokens": 117338304.0, |
| "step": 4100 |
| }, |
| { |
| "epoch": 6.5600000000000005, |
| "eval_math_entropy": 0.21119803923368455, |
| "eval_math_loss": 1.1628667116165161, |
| "eval_math_mean_token_accuracy": 0.8154761862754821, |
| "eval_math_num_tokens": 117338304.0, |
| "eval_math_runtime": 35.8268, |
| "eval_math_samples_per_second": 13.956, |
| "eval_math_steps_per_second": 3.489, |
| "step": 4100 |
| }, |
| { |
| "epoch": 6.5600000000000005, |
| "eval_physics_entropy": 0.22983593618869783, |
| "eval_physics_loss": 1.2532014846801758, |
| "eval_physics_mean_token_accuracy": 0.7969202942848206, |
| "eval_physics_num_tokens": 117338304.0, |
| "eval_physics_runtime": 41.5321, |
| "eval_physics_samples_per_second": 12.039, |
| "eval_physics_steps_per_second": 3.01, |
| "step": 4100 |
| }, |
| { |
| "entropy": 0.08333607679232954, |
| "epoch": 6.576, |
| "grad_norm": 1.1953125, |
| "learning_rate": 1.3091111111111113e-05, |
| "loss": 0.0497, |
| "mean_token_accuracy": 0.9862550985068083, |
| "num_tokens": 117633007.0, |
| "step": 4110 |
| }, |
| { |
| "entropy": 0.08117786860093475, |
| "epoch": 6.592, |
| "grad_norm": 1.15625, |
| "learning_rate": 1.306888888888889e-05, |
| "loss": 0.0494, |
| "mean_token_accuracy": 0.9866597782820463, |
| "num_tokens": 117927414.0, |
| "step": 4120 |
| }, |
| { |
| "entropy": 0.08642549277283251, |
| "epoch": 6.608, |
| "grad_norm": 1.1015625, |
| "learning_rate": 1.3046666666666668e-05, |
| "loss": 0.0512, |
| "mean_token_accuracy": 0.9859737418591976, |
| "num_tokens": 118209455.0, |
| "step": 4130 |
| }, |
| { |
| "entropy": 0.08443117390852421, |
| "epoch": 6.624, |
| "grad_norm": 1.15625, |
| "learning_rate": 1.3024444444444446e-05, |
| "loss": 0.0505, |
| "mean_token_accuracy": 0.9862124625593424, |
| "num_tokens": 118495992.0, |
| "step": 4140 |
| }, |
| { |
| "entropy": 0.08461201293393969, |
| "epoch": 6.64, |
| "grad_norm": 1.3125, |
| "learning_rate": 1.3002222222222223e-05, |
| "loss": 0.0516, |
| "mean_token_accuracy": 0.9858912594616414, |
| "num_tokens": 118779177.0, |
| "step": 4150 |
| }, |
| { |
| "entropy": 0.08362286686897277, |
| "epoch": 6.656, |
| "grad_norm": 1.171875, |
| "learning_rate": 1.2980000000000001e-05, |
| "loss": 0.0501, |
| "mean_token_accuracy": 0.9861939534544945, |
| "num_tokens": 119065831.0, |
| "step": 4160 |
| }, |
| { |
| "entropy": 0.08377992056775838, |
| "epoch": 6.672, |
| "grad_norm": 1.140625, |
| "learning_rate": 1.2957777777777779e-05, |
| "loss": 0.0505, |
| "mean_token_accuracy": 0.9862146992236376, |
| "num_tokens": 119349974.0, |
| "step": 4170 |
| }, |
| { |
| "entropy": 0.08517354882787913, |
| "epoch": 6.688, |
| "grad_norm": 1.2578125, |
| "learning_rate": 1.2935555555555556e-05, |
| "loss": 0.0517, |
| "mean_token_accuracy": 0.9860025811940432, |
| "num_tokens": 119634067.0, |
| "step": 4180 |
| }, |
| { |
| "entropy": 0.08468333540949971, |
| "epoch": 6.704, |
| "grad_norm": 1.390625, |
| "learning_rate": 1.2913333333333336e-05, |
| "loss": 0.0506, |
| "mean_token_accuracy": 0.9862395841628313, |
| "num_tokens": 119922640.0, |
| "step": 4190 |
| }, |
| { |
| "entropy": 0.08581791603937745, |
| "epoch": 6.72, |
| "grad_norm": 1.1640625, |
| "learning_rate": 1.2891111111111112e-05, |
| "loss": 0.0516, |
| "mean_token_accuracy": 0.9857941895723343, |
| "num_tokens": 120206747.0, |
| "step": 4200 |
| }, |
| { |
| "epoch": 6.72, |
| "eval_math_entropy": 0.2110741074681282, |
| "eval_math_loss": 1.1673368215560913, |
| "eval_math_mean_token_accuracy": 0.8153284401893616, |
| "eval_math_num_tokens": 120206747.0, |
| "eval_math_runtime": 35.8248, |
| "eval_math_samples_per_second": 13.957, |
| "eval_math_steps_per_second": 3.489, |
| "step": 4200 |
| }, |
| { |
| "epoch": 6.72, |
| "eval_physics_entropy": 0.22907479107379913, |
| "eval_physics_loss": 1.260624885559082, |
| "eval_physics_mean_token_accuracy": 0.7967779293060303, |
| "eval_physics_num_tokens": 120206747.0, |
| "eval_physics_runtime": 41.5156, |
| "eval_physics_samples_per_second": 12.044, |
| "eval_physics_steps_per_second": 3.011, |
| "step": 4200 |
| }, |
| { |
| "entropy": 0.08579737725667655, |
| "epoch": 6.736, |
| "grad_norm": 1.1640625, |
| "learning_rate": 1.2868888888888891e-05, |
| "loss": 0.0516, |
| "mean_token_accuracy": 0.9860372472554445, |
| "num_tokens": 120486581.0, |
| "step": 4210 |
| }, |
| { |
| "entropy": 0.08400838901288807, |
| "epoch": 6.752, |
| "grad_norm": 1.1875, |
| "learning_rate": 1.2846666666666667e-05, |
| "loss": 0.05, |
| "mean_token_accuracy": 0.9863463308662176, |
| "num_tokens": 120779617.0, |
| "step": 4220 |
| }, |
| { |
| "entropy": 0.08478187646251172, |
| "epoch": 6.768, |
| "grad_norm": 1.0625, |
| "learning_rate": 1.2824444444444446e-05, |
| "loss": 0.0509, |
| "mean_token_accuracy": 0.9862184945493937, |
| "num_tokens": 121064384.0, |
| "step": 4230 |
| }, |
| { |
| "entropy": 0.08424062917474658, |
| "epoch": 6.784, |
| "grad_norm": 1.0703125, |
| "learning_rate": 1.2802222222222222e-05, |
| "loss": 0.0517, |
| "mean_token_accuracy": 0.9859384100884199, |
| "num_tokens": 121351299.0, |
| "step": 4240 |
| }, |
| { |
| "entropy": 0.08782856916077435, |
| "epoch": 6.8, |
| "grad_norm": 1.1875, |
| "learning_rate": 1.2780000000000001e-05, |
| "loss": 0.0531, |
| "mean_token_accuracy": 0.9860101863741875, |
| "num_tokens": 121626812.0, |
| "step": 4250 |
| }, |
| { |
| "entropy": 0.08583347962703555, |
| "epoch": 6.816, |
| "grad_norm": 1.3203125, |
| "learning_rate": 1.2757777777777777e-05, |
| "loss": 0.0512, |
| "mean_token_accuracy": 0.9861073594540357, |
| "num_tokens": 121917563.0, |
| "step": 4260 |
| }, |
| { |
| "entropy": 0.08525952904019504, |
| "epoch": 6.832, |
| "grad_norm": 1.140625, |
| "learning_rate": 1.2735555555555557e-05, |
| "loss": 0.0509, |
| "mean_token_accuracy": 0.985946049168706, |
| "num_tokens": 122205507.0, |
| "step": 4270 |
| }, |
| { |
| "entropy": 0.08095580453518778, |
| "epoch": 6.848, |
| "grad_norm": 1.1171875, |
| "learning_rate": 1.2713333333333336e-05, |
| "loss": 0.0479, |
| "mean_token_accuracy": 0.9870608881115913, |
| "num_tokens": 122502960.0, |
| "step": 4280 |
| }, |
| { |
| "entropy": 0.0856839137384668, |
| "epoch": 6.864, |
| "grad_norm": 1.1875, |
| "learning_rate": 1.2691111111111112e-05, |
| "loss": 0.0517, |
| "mean_token_accuracy": 0.9857846073806286, |
| "num_tokens": 122783468.0, |
| "step": 4290 |
| }, |
| { |
| "entropy": 0.08315675526391715, |
| "epoch": 6.88, |
| "grad_norm": 1.109375, |
| "learning_rate": 1.2668888888888891e-05, |
| "loss": 0.0498, |
| "mean_token_accuracy": 0.9864332474768162, |
| "num_tokens": 123073078.0, |
| "step": 4300 |
| }, |
| { |
| "epoch": 6.88, |
| "eval_math_entropy": 0.2092110061645508, |
| "eval_math_loss": 1.1747575998306274, |
| "eval_math_mean_token_accuracy": 0.8151103315353394, |
| "eval_math_num_tokens": 123073078.0, |
| "eval_math_runtime": 35.8349, |
| "eval_math_samples_per_second": 13.953, |
| "eval_math_steps_per_second": 3.488, |
| "step": 4300 |
| }, |
| { |
| "epoch": 6.88, |
| "eval_physics_entropy": 0.22831890201568603, |
| "eval_physics_loss": 1.264940619468689, |
| "eval_physics_mean_token_accuracy": 0.7968414192199706, |
| "eval_physics_num_tokens": 123073078.0, |
| "eval_physics_runtime": 41.5084, |
| "eval_physics_samples_per_second": 12.046, |
| "eval_physics_steps_per_second": 3.011, |
| "step": 4300 |
| }, |
| { |
| "entropy": 0.08344000696670265, |
| "epoch": 6.896, |
| "grad_norm": 1.2109375, |
| "learning_rate": 1.2646666666666667e-05, |
| "loss": 0.0501, |
| "mean_token_accuracy": 0.9864881403744221, |
| "num_tokens": 123362103.0, |
| "step": 4310 |
| }, |
| { |
| "entropy": 0.08804852021858096, |
| "epoch": 6.912, |
| "grad_norm": 1.2578125, |
| "learning_rate": 1.2624444444444446e-05, |
| "loss": 0.0534, |
| "mean_token_accuracy": 0.9851112041622401, |
| "num_tokens": 123640151.0, |
| "step": 4320 |
| }, |
| { |
| "entropy": 0.08389839944429696, |
| "epoch": 6.928, |
| "grad_norm": 1.3125, |
| "learning_rate": 1.2602222222222222e-05, |
| "loss": 0.05, |
| "mean_token_accuracy": 0.9865112222731114, |
| "num_tokens": 123930914.0, |
| "step": 4330 |
| }, |
| { |
| "entropy": 0.08494017843622714, |
| "epoch": 6.944, |
| "grad_norm": 1.0703125, |
| "learning_rate": 1.2580000000000002e-05, |
| "loss": 0.0512, |
| "mean_token_accuracy": 0.9860570751130581, |
| "num_tokens": 124216965.0, |
| "step": 4340 |
| }, |
| { |
| "entropy": 0.08553386291023343, |
| "epoch": 6.96, |
| "grad_norm": 1.0859375, |
| "learning_rate": 1.2557777777777779e-05, |
| "loss": 0.0517, |
| "mean_token_accuracy": 0.9858683951199054, |
| "num_tokens": 124495374.0, |
| "step": 4350 |
| }, |
| { |
| "entropy": 0.0846124368486926, |
| "epoch": 6.976, |
| "grad_norm": 1.203125, |
| "learning_rate": 1.2535555555555557e-05, |
| "loss": 0.0507, |
| "mean_token_accuracy": 0.9861722022294999, |
| "num_tokens": 124779153.0, |
| "step": 4360 |
| }, |
| { |
| "entropy": 0.08296751752495765, |
| "epoch": 6.992, |
| "grad_norm": 1.2734375, |
| "learning_rate": 1.2513333333333334e-05, |
| "loss": 0.0502, |
| "mean_token_accuracy": 0.9865076046437025, |
| "num_tokens": 125065811.0, |
| "step": 4370 |
| }, |
| { |
| "entropy": 0.0782727177022025, |
| "epoch": 7.008, |
| "grad_norm": 0.6640625, |
| "learning_rate": 1.2491111111111112e-05, |
| "loss": 0.0455, |
| "mean_token_accuracy": 0.987744751945138, |
| "num_tokens": 125343099.0, |
| "step": 4380 |
| }, |
| { |
| "entropy": 0.06690982989966869, |
| "epoch": 7.024, |
| "grad_norm": 0.828125, |
| "learning_rate": 1.246888888888889e-05, |
| "loss": 0.0414, |
| "mean_token_accuracy": 0.9888394929468631, |
| "num_tokens": 125628503.0, |
| "step": 4390 |
| }, |
| { |
| "entropy": 0.06452394074294716, |
| "epoch": 7.04, |
| "grad_norm": 0.75, |
| "learning_rate": 1.2446666666666667e-05, |
| "loss": 0.0382, |
| "mean_token_accuracy": 0.9892929658293724, |
| "num_tokens": 125916477.0, |
| "step": 4400 |
| }, |
| { |
| "epoch": 7.04, |
| "eval_math_entropy": 0.19902950882911682, |
| "eval_math_loss": 1.2132738828659058, |
| "eval_math_mean_token_accuracy": 0.8153792238235473, |
| "eval_math_num_tokens": 125916477.0, |
| "eval_math_runtime": 35.8544, |
| "eval_math_samples_per_second": 13.945, |
| "eval_math_steps_per_second": 3.486, |
| "step": 4400 |
| }, |
| { |
| "epoch": 7.04, |
| "eval_physics_entropy": 0.22179586195945739, |
| "eval_physics_loss": 1.2990820407867432, |
| "eval_physics_mean_token_accuracy": 0.7966229004859924, |
| "eval_physics_num_tokens": 125916477.0, |
| "eval_physics_runtime": 41.5422, |
| "eval_physics_samples_per_second": 12.036, |
| "eval_physics_steps_per_second": 3.009, |
| "step": 4400 |
| }, |
| { |
| "entropy": 0.06385192952584476, |
| "epoch": 7.056, |
| "grad_norm": 0.671875, |
| "learning_rate": 1.2424444444444445e-05, |
| "loss": 0.0379, |
| "mean_token_accuracy": 0.9894341301172972, |
| "num_tokens": 126207007.0, |
| "step": 4410 |
| }, |
| { |
| "entropy": 0.06478747753426432, |
| "epoch": 7.072, |
| "grad_norm": 0.89453125, |
| "learning_rate": 1.2402222222222222e-05, |
| "loss": 0.0378, |
| "mean_token_accuracy": 0.9893794670701027, |
| "num_tokens": 126498674.0, |
| "step": 4420 |
| }, |
| { |
| "entropy": 0.06380244307219982, |
| "epoch": 7.088, |
| "grad_norm": 0.75390625, |
| "learning_rate": 1.2380000000000002e-05, |
| "loss": 0.0383, |
| "mean_token_accuracy": 0.9894326873123646, |
| "num_tokens": 126789042.0, |
| "step": 4430 |
| }, |
| { |
| "entropy": 0.06424271930009126, |
| "epoch": 7.104, |
| "grad_norm": 1.0625, |
| "learning_rate": 1.235777777777778e-05, |
| "loss": 0.039, |
| "mean_token_accuracy": 0.9894317038357258, |
| "num_tokens": 127078313.0, |
| "step": 4440 |
| }, |
| { |
| "entropy": 0.06499261818826199, |
| "epoch": 7.12, |
| "grad_norm": 0.73828125, |
| "learning_rate": 1.2335555555555557e-05, |
| "loss": 0.0382, |
| "mean_token_accuracy": 0.9892727185040713, |
| "num_tokens": 127363541.0, |
| "step": 4450 |
| }, |
| { |
| "entropy": 0.06562398292589933, |
| "epoch": 7.136, |
| "grad_norm": 0.66015625, |
| "learning_rate": 1.2313333333333335e-05, |
| "loss": 0.0389, |
| "mean_token_accuracy": 0.9892641551792621, |
| "num_tokens": 127642777.0, |
| "step": 4460 |
| }, |
| { |
| "entropy": 0.06509807226248085, |
| "epoch": 7.152, |
| "grad_norm": 0.73828125, |
| "learning_rate": 1.2291111111111112e-05, |
| "loss": 0.0387, |
| "mean_token_accuracy": 0.9890487994998693, |
| "num_tokens": 127928479.0, |
| "step": 4470 |
| }, |
| { |
| "entropy": 0.06522750512231142, |
| "epoch": 7.168, |
| "grad_norm": 0.734375, |
| "learning_rate": 1.226888888888889e-05, |
| "loss": 0.0392, |
| "mean_token_accuracy": 0.989065657928586, |
| "num_tokens": 128209825.0, |
| "step": 4480 |
| }, |
| { |
| "entropy": 0.06474495439324528, |
| "epoch": 7.184, |
| "grad_norm": 0.9296875, |
| "learning_rate": 1.2246666666666667e-05, |
| "loss": 0.0388, |
| "mean_token_accuracy": 0.9894650772213935, |
| "num_tokens": 128493833.0, |
| "step": 4490 |
| }, |
| { |
| "entropy": 0.06559755264315754, |
| "epoch": 7.2, |
| "grad_norm": 0.88671875, |
| "learning_rate": 1.2224444444444445e-05, |
| "loss": 0.0394, |
| "mean_token_accuracy": 0.9890734907239676, |
| "num_tokens": 128773901.0, |
| "step": 4500 |
| }, |
| { |
| "epoch": 7.2, |
| "eval_math_entropy": 0.1962946730852127, |
| "eval_math_loss": 1.2238725423812866, |
| "eval_math_mean_token_accuracy": 0.8156712355613709, |
| "eval_math_num_tokens": 128773901.0, |
| "eval_math_runtime": 35.8136, |
| "eval_math_samples_per_second": 13.961, |
| "eval_math_steps_per_second": 3.49, |
| "step": 4500 |
| }, |
| { |
| "epoch": 7.2, |
| "eval_physics_entropy": 0.2198963850736618, |
| "eval_physics_loss": 1.3048741817474365, |
| "eval_physics_mean_token_accuracy": 0.7971375894546509, |
| "eval_physics_num_tokens": 128773901.0, |
| "eval_physics_runtime": 41.1717, |
| "eval_physics_samples_per_second": 12.144, |
| "eval_physics_steps_per_second": 3.036, |
| "step": 4500 |
| }, |
| { |
| "entropy": 0.06589861372485757, |
| "epoch": 7.216, |
| "grad_norm": 0.7890625, |
| "learning_rate": 1.2202222222222224e-05, |
| "loss": 0.039, |
| "mean_token_accuracy": 0.989148149639368, |
| "num_tokens": 129057801.0, |
| "step": 4510 |
| }, |
| { |
| "entropy": 0.06478882993105799, |
| "epoch": 7.232, |
| "grad_norm": 0.73828125, |
| "learning_rate": 1.218e-05, |
| "loss": 0.0386, |
| "mean_token_accuracy": 0.989036962017417, |
| "num_tokens": 129340781.0, |
| "step": 4520 |
| }, |
| { |
| "entropy": 0.06308786221779883, |
| "epoch": 7.248, |
| "grad_norm": 0.70703125, |
| "learning_rate": 1.215777777777778e-05, |
| "loss": 0.0377, |
| "mean_token_accuracy": 0.9893654063344002, |
| "num_tokens": 129631132.0, |
| "step": 4530 |
| }, |
| { |
| "entropy": 0.06536523974500597, |
| "epoch": 7.264, |
| "grad_norm": 0.82421875, |
| "learning_rate": 1.2135555555555556e-05, |
| "loss": 0.0394, |
| "mean_token_accuracy": 0.9890879034996033, |
| "num_tokens": 129913307.0, |
| "step": 4540 |
| }, |
| { |
| "entropy": 0.06405791421420873, |
| "epoch": 7.28, |
| "grad_norm": 0.70703125, |
| "learning_rate": 1.2113333333333335e-05, |
| "loss": 0.0382, |
| "mean_token_accuracy": 0.9891516473144293, |
| "num_tokens": 130201744.0, |
| "step": 4550 |
| }, |
| { |
| "entropy": 0.06376761447172612, |
| "epoch": 7.296, |
| "grad_norm": 0.75, |
| "learning_rate": 1.209111111111111e-05, |
| "loss": 0.0384, |
| "mean_token_accuracy": 0.9893031839281321, |
| "num_tokens": 130488187.0, |
| "step": 4560 |
| }, |
| { |
| "entropy": 0.0670982622075826, |
| "epoch": 7.312, |
| "grad_norm": 0.8828125, |
| "learning_rate": 1.206888888888889e-05, |
| "loss": 0.0399, |
| "mean_token_accuracy": 0.9886458057910203, |
| "num_tokens": 130761481.0, |
| "step": 4570 |
| }, |
| { |
| "entropy": 0.06522584960330277, |
| "epoch": 7.328, |
| "grad_norm": 0.72265625, |
| "learning_rate": 1.204666666666667e-05, |
| "loss": 0.0388, |
| "mean_token_accuracy": 0.9892189968377352, |
| "num_tokens": 131046136.0, |
| "step": 4580 |
| }, |
| { |
| "entropy": 0.064781365939416, |
| "epoch": 7.344, |
| "grad_norm": 0.57421875, |
| "learning_rate": 1.2024444444444445e-05, |
| "loss": 0.039, |
| "mean_token_accuracy": 0.9891346648335457, |
| "num_tokens": 131333800.0, |
| "step": 4590 |
| }, |
| { |
| "entropy": 0.06483790301717818, |
| "epoch": 7.36, |
| "grad_norm": 0.91015625, |
| "learning_rate": 1.2002222222222225e-05, |
| "loss": 0.0383, |
| "mean_token_accuracy": 0.989232637360692, |
| "num_tokens": 131626450.0, |
| "step": 4600 |
| }, |
| { |
| "epoch": 7.36, |
| "eval_math_entropy": 0.19576846724748612, |
| "eval_math_loss": 1.226768136024475, |
| "eval_math_mean_token_accuracy": 0.8159392666816712, |
| "eval_math_num_tokens": 131626450.0, |
| "eval_math_runtime": 35.7948, |
| "eval_math_samples_per_second": 13.969, |
| "eval_math_steps_per_second": 3.492, |
| "step": 4600 |
| }, |
| { |
| "epoch": 7.36, |
| "eval_physics_entropy": 0.21936397910118102, |
| "eval_physics_loss": 1.3070223331451416, |
| "eval_physics_mean_token_accuracy": 0.7969147138595581, |
| "eval_physics_num_tokens": 131626450.0, |
| "eval_physics_runtime": 41.4955, |
| "eval_physics_samples_per_second": 12.05, |
| "eval_physics_steps_per_second": 3.012, |
| "step": 4600 |
| }, |
| { |
| "entropy": 0.06515917279757559, |
| "epoch": 7.376, |
| "grad_norm": 0.64453125, |
| "learning_rate": 1.198e-05, |
| "loss": 0.0392, |
| "mean_token_accuracy": 0.9892047554254532, |
| "num_tokens": 131906649.0, |
| "step": 4610 |
| }, |
| { |
| "entropy": 0.06519709336571396, |
| "epoch": 7.392, |
| "grad_norm": 0.73828125, |
| "learning_rate": 1.195777777777778e-05, |
| "loss": 0.0387, |
| "mean_token_accuracy": 0.9889419995248317, |
| "num_tokens": 132190328.0, |
| "step": 4620 |
| }, |
| { |
| "entropy": 0.06457304577343166, |
| "epoch": 7.408, |
| "grad_norm": 0.58984375, |
| "learning_rate": 1.1935555555555556e-05, |
| "loss": 0.0383, |
| "mean_token_accuracy": 0.9893009889870882, |
| "num_tokens": 132473728.0, |
| "step": 4630 |
| }, |
| { |
| "entropy": 0.06410650881007314, |
| "epoch": 7.424, |
| "grad_norm": 0.859375, |
| "learning_rate": 1.1913333333333335e-05, |
| "loss": 0.039, |
| "mean_token_accuracy": 0.9890226166695356, |
| "num_tokens": 132758657.0, |
| "step": 4640 |
| }, |
| { |
| "entropy": 0.0660351888742298, |
| "epoch": 7.44, |
| "grad_norm": 0.796875, |
| "learning_rate": 1.1891111111111111e-05, |
| "loss": 0.0394, |
| "mean_token_accuracy": 0.9889182843267917, |
| "num_tokens": 133037748.0, |
| "step": 4650 |
| }, |
| { |
| "entropy": 0.06468923375941812, |
| "epoch": 7.456, |
| "grad_norm": 0.66796875, |
| "learning_rate": 1.186888888888889e-05, |
| "loss": 0.0384, |
| "mean_token_accuracy": 0.9891764044761657, |
| "num_tokens": 133326761.0, |
| "step": 4660 |
| }, |
| { |
| "entropy": 0.06298574220854788, |
| "epoch": 7.4719999999999995, |
| "grad_norm": 0.7890625, |
| "learning_rate": 1.1846666666666668e-05, |
| "loss": 0.0381, |
| "mean_token_accuracy": 0.989344285428524, |
| "num_tokens": 133617018.0, |
| "step": 4670 |
| }, |
| { |
| "entropy": 0.06321970208082348, |
| "epoch": 7.4879999999999995, |
| "grad_norm": 0.80859375, |
| "learning_rate": 1.1824444444444445e-05, |
| "loss": 0.0384, |
| "mean_token_accuracy": 0.9892230208963155, |
| "num_tokens": 133904602.0, |
| "step": 4680 |
| }, |
| { |
| "entropy": 0.06326428384054453, |
| "epoch": 7.504, |
| "grad_norm": 1.390625, |
| "learning_rate": 1.1802222222222223e-05, |
| "loss": 0.0384, |
| "mean_token_accuracy": 0.9892394132912159, |
| "num_tokens": 134198376.0, |
| "step": 4690 |
| }, |
| { |
| "entropy": 0.0634507178561762, |
| "epoch": 7.52, |
| "grad_norm": 0.7265625, |
| "learning_rate": 1.178e-05, |
| "loss": 0.0383, |
| "mean_token_accuracy": 0.9894102096557618, |
| "num_tokens": 134491162.0, |
| "step": 4700 |
| }, |
| { |
| "epoch": 7.52, |
| "eval_math_entropy": 0.19578898119926452, |
| "eval_math_loss": 1.2300976514816284, |
| "eval_math_mean_token_accuracy": 0.8157597880363464, |
| "eval_math_num_tokens": 134491162.0, |
| "eval_math_runtime": 35.7866, |
| "eval_math_samples_per_second": 13.972, |
| "eval_math_steps_per_second": 3.493, |
| "step": 4700 |
| }, |
| { |
| "epoch": 7.52, |
| "eval_physics_entropy": 0.21874468064308167, |
| "eval_physics_loss": 1.310891032218933, |
| "eval_physics_mean_token_accuracy": 0.7968743557929993, |
| "eval_physics_num_tokens": 134491162.0, |
| "eval_physics_runtime": 41.4806, |
| "eval_physics_samples_per_second": 12.054, |
| "eval_physics_steps_per_second": 3.013, |
| "step": 4700 |
| }, |
| { |
| "entropy": 0.06613360948394983, |
| "epoch": 7.536, |
| "grad_norm": 0.84765625, |
| "learning_rate": 1.1757777777777778e-05, |
| "loss": 0.0397, |
| "mean_token_accuracy": 0.9887725580483675, |
| "num_tokens": 134772647.0, |
| "step": 4710 |
| }, |
| { |
| "entropy": 0.06250753803178669, |
| "epoch": 7.552, |
| "grad_norm": 0.6484375, |
| "learning_rate": 1.1735555555555556e-05, |
| "loss": 0.0381, |
| "mean_token_accuracy": 0.9894171375781298, |
| "num_tokens": 135062793.0, |
| "step": 4720 |
| }, |
| { |
| "entropy": 0.06458577974699438, |
| "epoch": 7.568, |
| "grad_norm": 0.9453125, |
| "learning_rate": 1.1713333333333334e-05, |
| "loss": 0.0382, |
| "mean_token_accuracy": 0.9891802627593279, |
| "num_tokens": 135352954.0, |
| "step": 4730 |
| }, |
| { |
| "entropy": 0.06382402395829559, |
| "epoch": 7.584, |
| "grad_norm": 1.15625, |
| "learning_rate": 1.1691111111111113e-05, |
| "loss": 0.0383, |
| "mean_token_accuracy": 0.9893467653542757, |
| "num_tokens": 135643709.0, |
| "step": 4740 |
| }, |
| { |
| "entropy": 0.062360757333226505, |
| "epoch": 7.6, |
| "grad_norm": 0.6640625, |
| "learning_rate": 1.1668888888888889e-05, |
| "loss": 0.0382, |
| "mean_token_accuracy": 0.9892903961241245, |
| "num_tokens": 135933730.0, |
| "step": 4750 |
| }, |
| { |
| "entropy": 0.06553376447409391, |
| "epoch": 7.616, |
| "grad_norm": 0.8359375, |
| "learning_rate": 1.1646666666666668e-05, |
| "loss": 0.0396, |
| "mean_token_accuracy": 0.9887435000389815, |
| "num_tokens": 136215545.0, |
| "step": 4760 |
| }, |
| { |
| "entropy": 0.06385163033846766, |
| "epoch": 7.632, |
| "grad_norm": 0.72265625, |
| "learning_rate": 1.1624444444444446e-05, |
| "loss": 0.0379, |
| "mean_token_accuracy": 0.989441429823637, |
| "num_tokens": 136509697.0, |
| "step": 4770 |
| }, |
| { |
| "entropy": 0.06461194895673543, |
| "epoch": 7.648, |
| "grad_norm": 0.82421875, |
| "learning_rate": 1.1602222222222223e-05, |
| "loss": 0.0398, |
| "mean_token_accuracy": 0.9889575779438019, |
| "num_tokens": 136794834.0, |
| "step": 4780 |
| }, |
| { |
| "entropy": 0.06364777330309153, |
| "epoch": 7.664, |
| "grad_norm": 0.74609375, |
| "learning_rate": 1.1580000000000001e-05, |
| "loss": 0.0387, |
| "mean_token_accuracy": 0.9890321586281061, |
| "num_tokens": 137077698.0, |
| "step": 4790 |
| }, |
| { |
| "entropy": 0.06495753165800125, |
| "epoch": 7.68, |
| "grad_norm": 0.7890625, |
| "learning_rate": 1.1557777777777779e-05, |
| "loss": 0.0407, |
| "mean_token_accuracy": 0.9888042062520981, |
| "num_tokens": 137354100.0, |
| "step": 4800 |
| }, |
| { |
| "epoch": 7.68, |
| "eval_math_entropy": 0.1961505770087242, |
| "eval_math_loss": 1.2280024290084839, |
| "eval_math_mean_token_accuracy": 0.8158598508834839, |
| "eval_math_num_tokens": 137354100.0, |
| "eval_math_runtime": 35.7851, |
| "eval_math_samples_per_second": 13.972, |
| "eval_math_steps_per_second": 3.493, |
| "step": 4800 |
| }, |
| { |
| "epoch": 7.68, |
| "eval_physics_entropy": 0.21946947872638703, |
| "eval_physics_loss": 1.3097177743911743, |
| "eval_physics_mean_token_accuracy": 0.7968477377891541, |
| "eval_physics_num_tokens": 137354100.0, |
| "eval_physics_runtime": 41.4785, |
| "eval_physics_samples_per_second": 12.054, |
| "eval_physics_steps_per_second": 3.014, |
| "step": 4800 |
| }, |
| { |
| "entropy": 0.06376796101685614, |
| "epoch": 7.696, |
| "grad_norm": 0.66015625, |
| "learning_rate": 1.1535555555555556e-05, |
| "loss": 0.0387, |
| "mean_token_accuracy": 0.9891933016479015, |
| "num_tokens": 137646688.0, |
| "step": 4810 |
| }, |
| { |
| "entropy": 0.06496104435063899, |
| "epoch": 7.712, |
| "grad_norm": 0.734375, |
| "learning_rate": 1.1513333333333334e-05, |
| "loss": 0.0391, |
| "mean_token_accuracy": 0.9888025749474764, |
| "num_tokens": 137930478.0, |
| "step": 4820 |
| }, |
| { |
| "entropy": 0.06421369009185582, |
| "epoch": 7.728, |
| "grad_norm": 0.76953125, |
| "learning_rate": 1.1491111111111113e-05, |
| "loss": 0.0389, |
| "mean_token_accuracy": 0.9891324803233147, |
| "num_tokens": 138219528.0, |
| "step": 4830 |
| }, |
| { |
| "entropy": 0.06491015092469751, |
| "epoch": 7.744, |
| "grad_norm": 0.6796875, |
| "learning_rate": 1.1468888888888889e-05, |
| "loss": 0.0389, |
| "mean_token_accuracy": 0.9887373331934214, |
| "num_tokens": 138505271.0, |
| "step": 4840 |
| }, |
| { |
| "entropy": 0.06559491397347302, |
| "epoch": 7.76, |
| "grad_norm": 0.76953125, |
| "learning_rate": 1.1446666666666668e-05, |
| "loss": 0.0396, |
| "mean_token_accuracy": 0.9887735281139612, |
| "num_tokens": 138790612.0, |
| "step": 4850 |
| }, |
| { |
| "entropy": 0.06484551811590791, |
| "epoch": 7.776, |
| "grad_norm": 0.79296875, |
| "learning_rate": 1.1424444444444444e-05, |
| "loss": 0.0383, |
| "mean_token_accuracy": 0.9889496352523566, |
| "num_tokens": 139081356.0, |
| "step": 4860 |
| }, |
| { |
| "entropy": 0.06331797884777188, |
| "epoch": 7.792, |
| "grad_norm": 0.70703125, |
| "learning_rate": 1.1402222222222224e-05, |
| "loss": 0.0372, |
| "mean_token_accuracy": 0.9895282920449973, |
| "num_tokens": 139380135.0, |
| "step": 4870 |
| }, |
| { |
| "entropy": 0.06256569258403033, |
| "epoch": 7.808, |
| "grad_norm": 0.69921875, |
| "learning_rate": 1.138e-05, |
| "loss": 0.0388, |
| "mean_token_accuracy": 0.9891048073768616, |
| "num_tokens": 139670791.0, |
| "step": 4880 |
| }, |
| { |
| "entropy": 0.06474397722631693, |
| "epoch": 7.824, |
| "grad_norm": 0.91015625, |
| "learning_rate": 1.1357777777777779e-05, |
| "loss": 0.0389, |
| "mean_token_accuracy": 0.988869558647275, |
| "num_tokens": 139954935.0, |
| "step": 4890 |
| }, |
| { |
| "entropy": 0.06261393134482204, |
| "epoch": 7.84, |
| "grad_norm": 0.72265625, |
| "learning_rate": 1.1335555555555558e-05, |
| "loss": 0.0395, |
| "mean_token_accuracy": 0.9891028236597776, |
| "num_tokens": 140243499.0, |
| "step": 4900 |
| }, |
| { |
| "epoch": 7.84, |
| "eval_math_entropy": 0.19476979553699494, |
| "eval_math_loss": 1.2364224195480347, |
| "eval_math_mean_token_accuracy": 0.8157722191810608, |
| "eval_math_num_tokens": 140243499.0, |
| "eval_math_runtime": 35.7988, |
| "eval_math_samples_per_second": 13.967, |
| "eval_math_steps_per_second": 3.492, |
| "step": 4900 |
| }, |
| { |
| "epoch": 7.84, |
| "eval_physics_entropy": 0.2179366738796234, |
| "eval_physics_loss": 1.3180210590362549, |
| "eval_physics_mean_token_accuracy": 0.7968727760314941, |
| "eval_physics_num_tokens": 140243499.0, |
| "eval_physics_runtime": 41.4652, |
| "eval_physics_samples_per_second": 12.058, |
| "eval_physics_steps_per_second": 3.015, |
| "step": 4900 |
| }, |
| { |
| "entropy": 0.0660986915230751, |
| "epoch": 7.856, |
| "grad_norm": 0.796875, |
| "learning_rate": 1.1313333333333334e-05, |
| "loss": 0.0395, |
| "mean_token_accuracy": 0.9887882888317108, |
| "num_tokens": 140523048.0, |
| "step": 4910 |
| }, |
| { |
| "entropy": 0.06733772295992821, |
| "epoch": 7.872, |
| "grad_norm": 0.80078125, |
| "learning_rate": 1.1291111111111113e-05, |
| "loss": 0.0411, |
| "mean_token_accuracy": 0.9885444976389408, |
| "num_tokens": 140798987.0, |
| "step": 4920 |
| }, |
| { |
| "entropy": 0.06379716179799289, |
| "epoch": 7.888, |
| "grad_norm": 0.7734375, |
| "learning_rate": 1.126888888888889e-05, |
| "loss": 0.0394, |
| "mean_token_accuracy": 0.9888154394924641, |
| "num_tokens": 141086821.0, |
| "step": 4930 |
| }, |
| { |
| "entropy": 0.06268583498895168, |
| "epoch": 7.904, |
| "grad_norm": 0.60546875, |
| "learning_rate": 1.1246666666666669e-05, |
| "loss": 0.0377, |
| "mean_token_accuracy": 0.9894758500158787, |
| "num_tokens": 141381172.0, |
| "step": 4940 |
| }, |
| { |
| "entropy": 0.06482646651566029, |
| "epoch": 7.92, |
| "grad_norm": 0.71875, |
| "learning_rate": 1.1224444444444444e-05, |
| "loss": 0.039, |
| "mean_token_accuracy": 0.9889786690473557, |
| "num_tokens": 141668933.0, |
| "step": 4950 |
| }, |
| { |
| "entropy": 0.06496006473898888, |
| "epoch": 7.936, |
| "grad_norm": 0.71875, |
| "learning_rate": 1.1202222222222224e-05, |
| "loss": 0.0393, |
| "mean_token_accuracy": 0.9888488940894604, |
| "num_tokens": 141951435.0, |
| "step": 4960 |
| }, |
| { |
| "entropy": 0.06452734193298966, |
| "epoch": 7.952, |
| "grad_norm": 0.66796875, |
| "learning_rate": 1.1180000000000001e-05, |
| "loss": 0.0394, |
| "mean_token_accuracy": 0.9889313533902169, |
| "num_tokens": 142238865.0, |
| "step": 4970 |
| }, |
| { |
| "entropy": 0.06580421230755747, |
| "epoch": 7.968, |
| "grad_norm": 0.70703125, |
| "learning_rate": 1.1157777777777779e-05, |
| "loss": 0.04, |
| "mean_token_accuracy": 0.9887514386326075, |
| "num_tokens": 142519525.0, |
| "step": 4980 |
| }, |
| { |
| "entropy": 0.06502947660628706, |
| "epoch": 7.984, |
| "grad_norm": 0.6953125, |
| "learning_rate": 1.1135555555555557e-05, |
| "loss": 0.0391, |
| "mean_token_accuracy": 0.9888614844530821, |
| "num_tokens": 142805078.0, |
| "step": 4990 |
| }, |
| { |
| "entropy": 0.06425679824315012, |
| "epoch": 8.0, |
| "grad_norm": 0.72265625, |
| "learning_rate": 1.1113333333333334e-05, |
| "loss": 0.0394, |
| "mean_token_accuracy": 0.9888634238392114, |
| "num_tokens": 143091392.0, |
| "step": 5000 |
| }, |
| { |
| "epoch": 8.0, |
| "eval_math_entropy": 0.1938511391878128, |
| "eval_math_loss": 1.2406725883483887, |
| "eval_math_mean_token_accuracy": 0.8157460641860962, |
| "eval_math_num_tokens": 143091392.0, |
| "eval_math_runtime": 35.8187, |
| "eval_math_samples_per_second": 13.959, |
| "eval_math_steps_per_second": 3.49, |
| "step": 5000 |
| }, |
| { |
| "epoch": 8.0, |
| "eval_physics_entropy": 0.21758921098709105, |
| "eval_physics_loss": 1.3192702531814575, |
| "eval_physics_mean_token_accuracy": 0.7965651154518127, |
| "eval_physics_num_tokens": 143091392.0, |
| "eval_physics_runtime": 41.2015, |
| "eval_physics_samples_per_second": 12.135, |
| "eval_physics_steps_per_second": 3.034, |
| "step": 5000 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 10000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 16, |
| "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": 1.0853280143146635e+19, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
