| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.4451121126133645, |
| "eval_steps": 100, |
| "global_step": 500, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 2.6993201158940794, |
| "epoch": 0.00890224225226729, |
| "grad_norm": 264.0, |
| "learning_rate": 1.8e-07, |
| "loss": 6.8158, |
| "mean_token_accuracy": 0.21653764923103153, |
| "num_tokens": 129942.0, |
| "step": 10 |
| }, |
| { |
| "entropy": 2.748416267335415, |
| "epoch": 0.01780448450453458, |
| "grad_norm": 342.0, |
| "learning_rate": 3.8e-07, |
| "loss": 6.842, |
| "mean_token_accuracy": 0.21704859719611705, |
| "num_tokens": 259048.0, |
| "step": 20 |
| }, |
| { |
| "entropy": 2.856283613294363, |
| "epoch": 0.026706726756801868, |
| "grad_norm": 260.0, |
| "learning_rate": 5.800000000000001e-07, |
| "loss": 6.5796, |
| "mean_token_accuracy": 0.21546755363233389, |
| "num_tokens": 397518.0, |
| "step": 30 |
| }, |
| { |
| "entropy": 3.1255638763308524, |
| "epoch": 0.03560896900906916, |
| "grad_norm": 235.0, |
| "learning_rate": 7.8e-07, |
| "loss": 6.0702, |
| "mean_token_accuracy": 0.22630739556625484, |
| "num_tokens": 532019.0, |
| "step": 40 |
| }, |
| { |
| "entropy": 3.437747097015381, |
| "epoch": 0.04451121126133645, |
| "grad_norm": 120.0, |
| "learning_rate": 9.800000000000001e-07, |
| "loss": 5.3567, |
| "mean_token_accuracy": 0.24060304677113892, |
| "num_tokens": 667753.0, |
| "step": 50 |
| }, |
| { |
| "entropy": 3.8268849313259126, |
| "epoch": 0.053413453513603736, |
| "grad_norm": 69.0, |
| "learning_rate": 1.1800000000000001e-06, |
| "loss": 4.7458, |
| "mean_token_accuracy": 0.25712613044306637, |
| "num_tokens": 807220.0, |
| "step": 60 |
| }, |
| { |
| "entropy": 4.098993194103241, |
| "epoch": 0.06231569576587103, |
| "grad_norm": 44.0, |
| "learning_rate": 1.3800000000000001e-06, |
| "loss": 4.1873, |
| "mean_token_accuracy": 0.2863635550253093, |
| "num_tokens": 938024.0, |
| "step": 70 |
| }, |
| { |
| "entropy": 3.7632771015167235, |
| "epoch": 0.07121793801813832, |
| "grad_norm": 39.75, |
| "learning_rate": 1.5800000000000001e-06, |
| "loss": 3.8191, |
| "mean_token_accuracy": 0.3185409564524889, |
| "num_tokens": 1068429.0, |
| "step": 80 |
| }, |
| { |
| "entropy": 3.4629461631178855, |
| "epoch": 0.08012018027040561, |
| "grad_norm": 32.0, |
| "learning_rate": 1.7800000000000001e-06, |
| "loss": 3.5033, |
| "mean_token_accuracy": 0.3425675223581493, |
| "num_tokens": 1209109.0, |
| "step": 90 |
| }, |
| { |
| "entropy": 3.3570878595113753, |
| "epoch": 0.0890224225226729, |
| "grad_norm": 22.375, |
| "learning_rate": 1.98e-06, |
| "loss": 3.3283, |
| "mean_token_accuracy": 0.36780647775158287, |
| "num_tokens": 1336698.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.0890224225226729, |
| "eval_coding_entropy": 2.025357067108154, |
| "eval_coding_loss": 2.231159210205078, |
| "eval_coding_mean_token_accuracy": 0.49899543941020963, |
| "eval_coding_num_tokens": 1336698.0, |
| "eval_coding_runtime": 81.3311, |
| "eval_coding_samples_per_second": 6.148, |
| "eval_coding_steps_per_second": 3.074, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.0890224225226729, |
| "eval_chemistry_entropy": 3.242468201637268, |
| "eval_chemistry_loss": 3.3159916400909424, |
| "eval_chemistry_mean_token_accuracy": 0.3768811701536179, |
| "eval_chemistry_num_tokens": 1336698.0, |
| "eval_chemistry_runtime": 44.3638, |
| "eval_chemistry_samples_per_second": 11.27, |
| "eval_chemistry_steps_per_second": 5.635, |
| "step": 100 |
| }, |
| { |
| "entropy": 3.208672270178795, |
| "epoch": 0.09792466477494019, |
| "grad_norm": 26.625, |
| "learning_rate": 2.1800000000000003e-06, |
| "loss": 3.2262, |
| "mean_token_accuracy": 0.38037093859165905, |
| "num_tokens": 1470326.0, |
| "step": 110 |
| }, |
| { |
| "entropy": 3.073743937909603, |
| "epoch": 0.10682690702720747, |
| "grad_norm": 27.75, |
| "learning_rate": 2.38e-06, |
| "loss": 3.0418, |
| "mean_token_accuracy": 0.4051793536171317, |
| "num_tokens": 1608574.0, |
| "step": 120 |
| }, |
| { |
| "entropy": 3.0620595172047613, |
| "epoch": 0.11572914927947477, |
| "grad_norm": 23.0, |
| "learning_rate": 2.5800000000000003e-06, |
| "loss": 3.0619, |
| "mean_token_accuracy": 0.407612837664783, |
| "num_tokens": 1740382.0, |
| "step": 130 |
| }, |
| { |
| "entropy": 2.9336021929979323, |
| "epoch": 0.12463139153174206, |
| "grad_norm": 24.0, |
| "learning_rate": 2.7800000000000005e-06, |
| "loss": 2.8918, |
| "mean_token_accuracy": 0.4315342627465725, |
| "num_tokens": 1872747.0, |
| "step": 140 |
| }, |
| { |
| "entropy": 2.933651026338339, |
| "epoch": 0.13353363378400934, |
| "grad_norm": 21.125, |
| "learning_rate": 2.9800000000000003e-06, |
| "loss": 2.8977, |
| "mean_token_accuracy": 0.4332870215177536, |
| "num_tokens": 1999786.0, |
| "step": 150 |
| }, |
| { |
| "entropy": 2.78707619830966, |
| "epoch": 0.14243587603627664, |
| "grad_norm": 26.75, |
| "learning_rate": 3.1800000000000005e-06, |
| "loss": 2.713, |
| "mean_token_accuracy": 0.45073309261351824, |
| "num_tokens": 2133173.0, |
| "step": 160 |
| }, |
| { |
| "entropy": 2.686560308933258, |
| "epoch": 0.15133811828854393, |
| "grad_norm": 21.0, |
| "learning_rate": 3.3800000000000007e-06, |
| "loss": 2.6608, |
| "mean_token_accuracy": 0.45698707569390534, |
| "num_tokens": 2268889.0, |
| "step": 170 |
| }, |
| { |
| "entropy": 2.688250133395195, |
| "epoch": 0.16024036054081123, |
| "grad_norm": 23.0, |
| "learning_rate": 3.58e-06, |
| "loss": 2.6085, |
| "mean_token_accuracy": 0.4693053809925914, |
| "num_tokens": 2406145.0, |
| "step": 180 |
| }, |
| { |
| "entropy": 2.606059934198856, |
| "epoch": 0.1691426027930785, |
| "grad_norm": 19.0, |
| "learning_rate": 3.7800000000000002e-06, |
| "loss": 2.5776, |
| "mean_token_accuracy": 0.4732652172446251, |
| "num_tokens": 2533703.0, |
| "step": 190 |
| }, |
| { |
| "entropy": 2.6039498060941697, |
| "epoch": 0.1780448450453458, |
| "grad_norm": 20.625, |
| "learning_rate": 3.980000000000001e-06, |
| "loss": 2.5669, |
| "mean_token_accuracy": 0.4830343261361122, |
| "num_tokens": 2663709.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.1780448450453458, |
| "eval_coding_entropy": 1.940526816368103, |
| "eval_coding_loss": 2.1769461631774902, |
| "eval_coding_mean_token_accuracy": 0.5133515602350235, |
| "eval_coding_num_tokens": 2663709.0, |
| "eval_coding_runtime": 80.7684, |
| "eval_coding_samples_per_second": 6.191, |
| "eval_coding_steps_per_second": 3.095, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.1780448450453458, |
| "eval_chemistry_entropy": 2.574181836128235, |
| "eval_chemistry_loss": 2.5678696632385254, |
| "eval_chemistry_mean_token_accuracy": 0.4837975095510483, |
| "eval_chemistry_num_tokens": 2663709.0, |
| "eval_chemistry_runtime": 44.3338, |
| "eval_chemistry_samples_per_second": 11.278, |
| "eval_chemistry_steps_per_second": 5.639, |
| "step": 200 |
| }, |
| { |
| "entropy": 2.517921609431505, |
| "epoch": 0.18694708729761308, |
| "grad_norm": 21.875, |
| "learning_rate": 4.18e-06, |
| "loss": 2.4289, |
| "mean_token_accuracy": 0.4946174016222358, |
| "num_tokens": 2803638.0, |
| "step": 210 |
| }, |
| { |
| "entropy": 2.4712012231349947, |
| "epoch": 0.19584932954988038, |
| "grad_norm": 20.625, |
| "learning_rate": 4.38e-06, |
| "loss": 2.4073, |
| "mean_token_accuracy": 0.4968462478369474, |
| "num_tokens": 2934980.0, |
| "step": 220 |
| }, |
| { |
| "entropy": 2.402497890591621, |
| "epoch": 0.20475157180214767, |
| "grad_norm": 19.375, |
| "learning_rate": 4.58e-06, |
| "loss": 2.3435, |
| "mean_token_accuracy": 0.507471177726984, |
| "num_tokens": 3071339.0, |
| "step": 230 |
| }, |
| { |
| "entropy": 2.4326496072113515, |
| "epoch": 0.21365381405441494, |
| "grad_norm": 19.875, |
| "learning_rate": 4.78e-06, |
| "loss": 2.3438, |
| "mean_token_accuracy": 0.5086988668888808, |
| "num_tokens": 3208702.0, |
| "step": 240 |
| }, |
| { |
| "entropy": 2.321472028642893, |
| "epoch": 0.22255605630668224, |
| "grad_norm": 20.0, |
| "learning_rate": 4.980000000000001e-06, |
| "loss": 2.2871, |
| "mean_token_accuracy": 0.5164231477305293, |
| "num_tokens": 3347298.0, |
| "step": 250 |
| }, |
| { |
| "entropy": 2.342081367969513, |
| "epoch": 0.23145829855894953, |
| "grad_norm": 20.375, |
| "learning_rate": 5.18e-06, |
| "loss": 2.3225, |
| "mean_token_accuracy": 0.5174826141446829, |
| "num_tokens": 3480119.0, |
| "step": 260 |
| }, |
| { |
| "entropy": 2.30421741977334, |
| "epoch": 0.24036054081121683, |
| "grad_norm": 25.125, |
| "learning_rate": 5.380000000000001e-06, |
| "loss": 2.2163, |
| "mean_token_accuracy": 0.5236617928370834, |
| "num_tokens": 3615777.0, |
| "step": 270 |
| }, |
| { |
| "entropy": 2.3181479632854463, |
| "epoch": 0.24926278306348412, |
| "grad_norm": 19.625, |
| "learning_rate": 5.580000000000001e-06, |
| "loss": 2.2445, |
| "mean_token_accuracy": 0.5264638463035226, |
| "num_tokens": 3746392.0, |
| "step": 280 |
| }, |
| { |
| "entropy": 2.3754931248724462, |
| "epoch": 0.2581650253157514, |
| "grad_norm": 19.5, |
| "learning_rate": 5.78e-06, |
| "loss": 2.3122, |
| "mean_token_accuracy": 0.5208129966631532, |
| "num_tokens": 3880425.0, |
| "step": 290 |
| }, |
| { |
| "entropy": 2.2281612612307073, |
| "epoch": 0.2670672675680187, |
| "grad_norm": 21.5, |
| "learning_rate": 5.98e-06, |
| "loss": 2.1504, |
| "mean_token_accuracy": 0.5389395957812667, |
| "num_tokens": 4019487.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.2670672675680187, |
| "eval_coding_entropy": 1.8798885374069214, |
| "eval_coding_loss": 2.1364073753356934, |
| "eval_coding_mean_token_accuracy": 0.5203545212745666, |
| "eval_coding_num_tokens": 4019487.0, |
| "eval_coding_runtime": 80.6635, |
| "eval_coding_samples_per_second": 6.199, |
| "eval_coding_steps_per_second": 3.099, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.2670672675680187, |
| "eval_chemistry_entropy": 2.227835373401642, |
| "eval_chemistry_loss": 2.2248311042785645, |
| "eval_chemistry_mean_token_accuracy": 0.5376261820793152, |
| "eval_chemistry_num_tokens": 4019487.0, |
| "eval_chemistry_runtime": 44.3841, |
| "eval_chemistry_samples_per_second": 11.265, |
| "eval_chemistry_steps_per_second": 5.633, |
| "step": 300 |
| }, |
| { |
| "entropy": 2.2015350952744486, |
| "epoch": 0.275969509820286, |
| "grad_norm": 21.0, |
| "learning_rate": 6.18e-06, |
| "loss": 2.1543, |
| "mean_token_accuracy": 0.5409426406025887, |
| "num_tokens": 4162935.0, |
| "step": 310 |
| }, |
| { |
| "entropy": 2.148462438583374, |
| "epoch": 0.2848717520725533, |
| "grad_norm": 23.5, |
| "learning_rate": 6.380000000000001e-06, |
| "loss": 2.0957, |
| "mean_token_accuracy": 0.5452342376112937, |
| "num_tokens": 4298100.0, |
| "step": 320 |
| }, |
| { |
| "entropy": 2.0984656289219856, |
| "epoch": 0.29377399432482054, |
| "grad_norm": 19.5, |
| "learning_rate": 6.5800000000000005e-06, |
| "loss": 2.0404, |
| "mean_token_accuracy": 0.5558066863566637, |
| "num_tokens": 4431990.0, |
| "step": 330 |
| }, |
| { |
| "entropy": 2.1295887492597103, |
| "epoch": 0.30267623657708786, |
| "grad_norm": 17.625, |
| "learning_rate": 6.780000000000001e-06, |
| "loss": 2.0432, |
| "mean_token_accuracy": 0.5554485151544213, |
| "num_tokens": 4568116.0, |
| "step": 340 |
| }, |
| { |
| "entropy": 2.0176764823496343, |
| "epoch": 0.31157847882935513, |
| "grad_norm": 15.6875, |
| "learning_rate": 6.98e-06, |
| "loss": 1.9609, |
| "mean_token_accuracy": 0.5690282396972179, |
| "num_tokens": 4696575.0, |
| "step": 350 |
| }, |
| { |
| "entropy": 2.138897743076086, |
| "epoch": 0.32048072108162245, |
| "grad_norm": 20.0, |
| "learning_rate": 7.180000000000001e-06, |
| "loss": 2.0699, |
| "mean_token_accuracy": 0.5525342913344502, |
| "num_tokens": 4831334.0, |
| "step": 360 |
| }, |
| { |
| "entropy": 1.9979716561734677, |
| "epoch": 0.3293829633338897, |
| "grad_norm": 19.875, |
| "learning_rate": 7.3800000000000005e-06, |
| "loss": 1.9417, |
| "mean_token_accuracy": 0.5726395657286047, |
| "num_tokens": 4968132.0, |
| "step": 370 |
| }, |
| { |
| "entropy": 2.0710657715797423, |
| "epoch": 0.338285205586157, |
| "grad_norm": 23.125, |
| "learning_rate": 7.58e-06, |
| "loss": 1.9983, |
| "mean_token_accuracy": 0.5640093663707375, |
| "num_tokens": 5103301.0, |
| "step": 380 |
| }, |
| { |
| "entropy": 2.0690263971686362, |
| "epoch": 0.3471874478384243, |
| "grad_norm": 19.125, |
| "learning_rate": 7.78e-06, |
| "loss": 2.0099, |
| "mean_token_accuracy": 0.5661305790767074, |
| "num_tokens": 5230700.0, |
| "step": 390 |
| }, |
| { |
| "entropy": 2.0010731495916843, |
| "epoch": 0.3560896900906916, |
| "grad_norm": 15.625, |
| "learning_rate": 7.980000000000002e-06, |
| "loss": 1.9302, |
| "mean_token_accuracy": 0.5734745427966118, |
| "num_tokens": 5362915.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.3560896900906916, |
| "eval_coding_entropy": 1.9175666098594666, |
| "eval_coding_loss": 2.1198952198028564, |
| "eval_coding_mean_token_accuracy": 0.5218169560432434, |
| "eval_coding_num_tokens": 5362915.0, |
| "eval_coding_runtime": 80.6411, |
| "eval_coding_samples_per_second": 6.2, |
| "eval_coding_steps_per_second": 3.1, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.3560896900906916, |
| "eval_chemistry_entropy": 2.0164283666610716, |
| "eval_chemistry_loss": 1.9814609289169312, |
| "eval_chemistry_mean_token_accuracy": 0.5761862490177154, |
| "eval_chemistry_num_tokens": 5362915.0, |
| "eval_chemistry_runtime": 44.318, |
| "eval_chemistry_samples_per_second": 11.282, |
| "eval_chemistry_steps_per_second": 5.641, |
| "step": 400 |
| }, |
| { |
| "entropy": 1.9029628101736307, |
| "epoch": 0.3649919323429589, |
| "grad_norm": 20.625, |
| "learning_rate": 8.18e-06, |
| "loss": 1.861, |
| "mean_token_accuracy": 0.5885544015094638, |
| "num_tokens": 5497551.0, |
| "step": 410 |
| }, |
| { |
| "entropy": 1.8858691919595003, |
| "epoch": 0.37389417459522617, |
| "grad_norm": 18.75, |
| "learning_rate": 8.380000000000001e-06, |
| "loss": 1.8473, |
| "mean_token_accuracy": 0.5886353930458427, |
| "num_tokens": 5625455.0, |
| "step": 420 |
| }, |
| { |
| "entropy": 1.917143502831459, |
| "epoch": 0.38279641684749344, |
| "grad_norm": 18.0, |
| "learning_rate": 8.580000000000001e-06, |
| "loss": 1.8792, |
| "mean_token_accuracy": 0.5864967351779342, |
| "num_tokens": 5757241.0, |
| "step": 430 |
| }, |
| { |
| "entropy": 1.9342061571776867, |
| "epoch": 0.39169865909976076, |
| "grad_norm": 18.25, |
| "learning_rate": 8.78e-06, |
| "loss": 1.8538, |
| "mean_token_accuracy": 0.5882492190226912, |
| "num_tokens": 5890806.0, |
| "step": 440 |
| }, |
| { |
| "entropy": 1.8766419854015113, |
| "epoch": 0.400600901352028, |
| "grad_norm": 19.25, |
| "learning_rate": 8.98e-06, |
| "loss": 1.8162, |
| "mean_token_accuracy": 0.5928706232458353, |
| "num_tokens": 6023088.0, |
| "step": 450 |
| }, |
| { |
| "entropy": 1.8140752121806145, |
| "epoch": 0.40950314360429535, |
| "grad_norm": 17.375, |
| "learning_rate": 9.180000000000002e-06, |
| "loss": 1.7678, |
| "mean_token_accuracy": 0.6031433455646038, |
| "num_tokens": 6158751.0, |
| "step": 460 |
| }, |
| { |
| "entropy": 1.8589354373514653, |
| "epoch": 0.4184053858565626, |
| "grad_norm": 17.25, |
| "learning_rate": 9.38e-06, |
| "loss": 1.8104, |
| "mean_token_accuracy": 0.5982734605669975, |
| "num_tokens": 6289699.0, |
| "step": 470 |
| }, |
| { |
| "entropy": 1.8204911682754754, |
| "epoch": 0.4273076281088299, |
| "grad_norm": 18.125, |
| "learning_rate": 9.58e-06, |
| "loss": 1.7427, |
| "mean_token_accuracy": 0.6014625774696469, |
| "num_tokens": 6423747.0, |
| "step": 480 |
| }, |
| { |
| "entropy": 1.8097763508558273, |
| "epoch": 0.4362098703610972, |
| "grad_norm": 18.0, |
| "learning_rate": 9.780000000000001e-06, |
| "loss": 1.7556, |
| "mean_token_accuracy": 0.606043404340744, |
| "num_tokens": 6556651.0, |
| "step": 490 |
| }, |
| { |
| "entropy": 1.7938172847032547, |
| "epoch": 0.4451121126133645, |
| "grad_norm": 16.125, |
| "learning_rate": 9.980000000000001e-06, |
| "loss": 1.7349, |
| "mean_token_accuracy": 0.6063602799549699, |
| "num_tokens": 6684517.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.4451121126133645, |
| "eval_coding_entropy": 1.9086761941909791, |
| "eval_coding_loss": 2.1499152183532715, |
| "eval_coding_mean_token_accuracy": 0.5191502428054809, |
| "eval_coding_num_tokens": 6684517.0, |
| "eval_coding_runtime": 80.761, |
| "eval_coding_samples_per_second": 6.191, |
| "eval_coding_steps_per_second": 3.096, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.4451121126133645, |
| "eval_chemistry_entropy": 1.8382492520809173, |
| "eval_chemistry_loss": 1.8092410564422607, |
| "eval_chemistry_mean_token_accuracy": 0.6044963012933731, |
| "eval_chemistry_num_tokens": 6684517.0, |
| "eval_chemistry_runtime": 43.9817, |
| "eval_chemistry_samples_per_second": 11.368, |
| "eval_chemistry_steps_per_second": 5.684, |
| "step": 500 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 10000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 9, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 4.3005744249719194e+17, |
| "train_batch_size": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
