| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.8, |
| "eval_steps": 100, |
| "global_step": 1000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 0.6430254079401493, |
| "epoch": 0.008, |
| "grad_norm": 4.53125, |
| "learning_rate": 6.000000000000001e-07, |
| "loss": 1.8435, |
| "mean_token_accuracy": 0.6683766476809978, |
| "num_tokens": 133939.0, |
| "step": 10 |
| }, |
| { |
| "entropy": 0.6358832061290741, |
| "epoch": 0.016, |
| "grad_norm": 4.3125, |
| "learning_rate": 1.2666666666666669e-06, |
| "loss": 1.8392, |
| "mean_token_accuracy": 0.6651193369179964, |
| "num_tokens": 273600.0, |
| "step": 20 |
| }, |
| { |
| "entropy": 0.6754261270165444, |
| "epoch": 0.024, |
| "grad_norm": 4.9375, |
| "learning_rate": 1.9333333333333336e-06, |
| "loss": 1.9101, |
| "mean_token_accuracy": 0.6521823685616255, |
| "num_tokens": 408034.0, |
| "step": 30 |
| }, |
| { |
| "entropy": 0.6558344909921289, |
| "epoch": 0.032, |
| "grad_norm": 4.28125, |
| "learning_rate": 2.6e-06, |
| "loss": 1.8472, |
| "mean_token_accuracy": 0.6629668604582548, |
| "num_tokens": 542836.0, |
| "step": 40 |
| }, |
| { |
| "entropy": 0.6494104014709592, |
| "epoch": 0.04, |
| "grad_norm": 3.8125, |
| "learning_rate": 3.266666666666667e-06, |
| "loss": 1.8027, |
| "mean_token_accuracy": 0.6676773574203253, |
| "num_tokens": 678419.0, |
| "step": 50 |
| }, |
| { |
| "entropy": 0.662838158942759, |
| "epoch": 0.048, |
| "grad_norm": 4.03125, |
| "learning_rate": 3.9333333333333335e-06, |
| "loss": 1.8213, |
| "mean_token_accuracy": 0.6647731238976121, |
| "num_tokens": 813510.0, |
| "step": 60 |
| }, |
| { |
| "entropy": 0.6506876813247799, |
| "epoch": 0.056, |
| "grad_norm": 3.71875, |
| "learning_rate": 4.600000000000001e-06, |
| "loss": 1.7955, |
| "mean_token_accuracy": 0.6665123915299773, |
| "num_tokens": 954682.0, |
| "step": 70 |
| }, |
| { |
| "entropy": 0.6252049600705505, |
| "epoch": 0.064, |
| "grad_norm": 3.65625, |
| "learning_rate": 5.2666666666666665e-06, |
| "loss": 1.7861, |
| "mean_token_accuracy": 0.6728611093014478, |
| "num_tokens": 1096071.0, |
| "step": 80 |
| }, |
| { |
| "entropy": 0.6236795110628008, |
| "epoch": 0.072, |
| "grad_norm": 3.171875, |
| "learning_rate": 5.933333333333335e-06, |
| "loss": 1.6458, |
| "mean_token_accuracy": 0.6831523377448321, |
| "num_tokens": 1239910.0, |
| "step": 90 |
| }, |
| { |
| "entropy": 0.6565364092588425, |
| "epoch": 0.08, |
| "grad_norm": 3.125, |
| "learning_rate": 6.600000000000001e-06, |
| "loss": 1.6948, |
| "mean_token_accuracy": 0.6728412076830864, |
| "num_tokens": 1376045.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.08, |
| "eval_biology_entropy": 1.821127786874771, |
| "eval_biology_loss": 4.8697123527526855, |
| "eval_biology_mean_token_accuracy": 0.3407035154700279, |
| "eval_biology_num_tokens": 1376045.0, |
| "eval_biology_runtime": 26.9612, |
| "eval_biology_samples_per_second": 18.545, |
| "eval_biology_steps_per_second": 9.273, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.08, |
| "eval_chemistry_entropy": 1.1214371861219405, |
| "eval_chemistry_loss": 2.9583740234375, |
| "eval_chemistry_mean_token_accuracy": 0.5230503298044205, |
| "eval_chemistry_num_tokens": 1376045.0, |
| "eval_chemistry_runtime": 34.4743, |
| "eval_chemistry_samples_per_second": 14.504, |
| "eval_chemistry_steps_per_second": 7.252, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.08, |
| "eval_math_entropy": 0.6455623766183853, |
| "eval_math_loss": 1.6610009670257568, |
| "eval_math_mean_token_accuracy": 0.6816041300296783, |
| "eval_math_num_tokens": 1376045.0, |
| "eval_math_runtime": 34.9863, |
| "eval_math_samples_per_second": 14.291, |
| "eval_math_steps_per_second": 7.146, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.08, |
| "eval_physics_entropy": 0.8894976880550385, |
| "eval_physics_loss": 2.340852975845337, |
| "eval_physics_mean_token_accuracy": 0.5970065190792083, |
| "eval_physics_num_tokens": 1376045.0, |
| "eval_physics_runtime": 40.5871, |
| "eval_physics_samples_per_second": 12.319, |
| "eval_physics_steps_per_second": 6.16, |
| "step": 100 |
| }, |
| { |
| "entropy": 0.6108825646340847, |
| "epoch": 0.088, |
| "grad_norm": 3.0625, |
| "learning_rate": 7.266666666666668e-06, |
| "loss": 1.534, |
| "mean_token_accuracy": 0.6979989364743233, |
| "num_tokens": 1518257.0, |
| "step": 110 |
| }, |
| { |
| "entropy": 0.6626681972295045, |
| "epoch": 0.096, |
| "grad_norm": 3.0, |
| "learning_rate": 7.933333333333334e-06, |
| "loss": 1.4858, |
| "mean_token_accuracy": 0.6921830836683511, |
| "num_tokens": 1656215.0, |
| "step": 120 |
| }, |
| { |
| "entropy": 0.6807309044525027, |
| "epoch": 0.104, |
| "grad_norm": 2.4375, |
| "learning_rate": 8.6e-06, |
| "loss": 1.3897, |
| "mean_token_accuracy": 0.7023331839591265, |
| "num_tokens": 1791507.0, |
| "step": 130 |
| }, |
| { |
| "entropy": 0.7359636599197984, |
| "epoch": 0.112, |
| "grad_norm": 1.9765625, |
| "learning_rate": 9.266666666666667e-06, |
| "loss": 1.3356, |
| "mean_token_accuracy": 0.7036307867616415, |
| "num_tokens": 1928153.0, |
| "step": 140 |
| }, |
| { |
| "entropy": 0.8070579996332526, |
| "epoch": 0.12, |
| "grad_norm": 1.7265625, |
| "learning_rate": 9.933333333333334e-06, |
| "loss": 1.2682, |
| "mean_token_accuracy": 0.712115453183651, |
| "num_tokens": 2071008.0, |
| "step": 150 |
| }, |
| { |
| "entropy": 0.874446533434093, |
| "epoch": 0.128, |
| "grad_norm": 1.2734375, |
| "learning_rate": 1.0600000000000002e-05, |
| "loss": 1.1443, |
| "mean_token_accuracy": 0.7246435590088367, |
| "num_tokens": 2216042.0, |
| "step": 160 |
| }, |
| { |
| "entropy": 0.9639016635715961, |
| "epoch": 0.136, |
| "grad_norm": 1.125, |
| "learning_rate": 1.1266666666666668e-05, |
| "loss": 1.1005, |
| "mean_token_accuracy": 0.730505557358265, |
| "num_tokens": 2350589.0, |
| "step": 170 |
| }, |
| { |
| "entropy": 1.0533128878101707, |
| "epoch": 0.144, |
| "grad_norm": 0.90234375, |
| "learning_rate": 1.1933333333333335e-05, |
| "loss": 1.0494, |
| "mean_token_accuracy": 0.7364416709169745, |
| "num_tokens": 2483116.0, |
| "step": 180 |
| }, |
| { |
| "entropy": 1.0085792429745197, |
| "epoch": 0.152, |
| "grad_norm": 0.96484375, |
| "learning_rate": 1.2600000000000001e-05, |
| "loss": 0.9875, |
| "mean_token_accuracy": 0.749194710329175, |
| "num_tokens": 2617958.0, |
| "step": 190 |
| }, |
| { |
| "entropy": 1.000020370259881, |
| "epoch": 0.16, |
| "grad_norm": 0.90234375, |
| "learning_rate": 1.3266666666666668e-05, |
| "loss": 0.9952, |
| "mean_token_accuracy": 0.7486346989870072, |
| "num_tokens": 2752258.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_biology_entropy": 3.2312401638031005, |
| "eval_biology_loss": 3.4984476566314697, |
| "eval_biology_mean_token_accuracy": 0.4106140378713608, |
| "eval_biology_num_tokens": 2752258.0, |
| "eval_biology_runtime": 26.9494, |
| "eval_biology_samples_per_second": 18.553, |
| "eval_biology_steps_per_second": 9.277, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_chemistry_entropy": 1.9263038992881776, |
| "eval_chemistry_loss": 1.9608160257339478, |
| "eval_chemistry_mean_token_accuracy": 0.5982974277734756, |
| "eval_chemistry_num_tokens": 2752258.0, |
| "eval_chemistry_runtime": 34.4173, |
| "eval_chemistry_samples_per_second": 14.528, |
| "eval_chemistry_steps_per_second": 7.264, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_math_entropy": 1.0257348996400832, |
| "eval_math_loss": 1.029754638671875, |
| "eval_math_mean_token_accuracy": 0.7436505243778229, |
| "eval_math_num_tokens": 2752258.0, |
| "eval_math_runtime": 34.9579, |
| "eval_math_samples_per_second": 14.303, |
| "eval_math_steps_per_second": 7.151, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_physics_entropy": 1.5005105144977569, |
| "eval_physics_loss": 1.5203113555908203, |
| "eval_physics_mean_token_accuracy": 0.6669992462396622, |
| "eval_physics_num_tokens": 2752258.0, |
| "eval_physics_runtime": 40.5453, |
| "eval_physics_samples_per_second": 12.332, |
| "eval_physics_steps_per_second": 6.166, |
| "step": 200 |
| }, |
| { |
| "entropy": 1.0170325668528677, |
| "epoch": 0.168, |
| "grad_norm": 0.80859375, |
| "learning_rate": 1.3933333333333334e-05, |
| "loss": 1.0032, |
| "mean_token_accuracy": 0.7469347029924392, |
| "num_tokens": 2889249.0, |
| "step": 210 |
| }, |
| { |
| "entropy": 0.9709451261907815, |
| "epoch": 0.176, |
| "grad_norm": 0.7734375, |
| "learning_rate": 1.46e-05, |
| "loss": 0.9299, |
| "mean_token_accuracy": 0.7556545663625002, |
| "num_tokens": 3030305.0, |
| "step": 220 |
| }, |
| { |
| "entropy": 1.0022083457559348, |
| "epoch": 0.184, |
| "grad_norm": 0.90234375, |
| "learning_rate": 1.5266666666666667e-05, |
| "loss": 0.9815, |
| "mean_token_accuracy": 0.7520246226340532, |
| "num_tokens": 3167096.0, |
| "step": 230 |
| }, |
| { |
| "entropy": 0.9930278725922108, |
| "epoch": 0.192, |
| "grad_norm": 0.85546875, |
| "learning_rate": 1.5933333333333336e-05, |
| "loss": 0.9637, |
| "mean_token_accuracy": 0.7547307416796685, |
| "num_tokens": 3306293.0, |
| "step": 240 |
| }, |
| { |
| "entropy": 0.9768950846046209, |
| "epoch": 0.2, |
| "grad_norm": 0.7734375, |
| "learning_rate": 1.66e-05, |
| "loss": 0.9373, |
| "mean_token_accuracy": 0.7594478782266378, |
| "num_tokens": 3441895.0, |
| "step": 250 |
| }, |
| { |
| "entropy": 0.9833385257050395, |
| "epoch": 0.208, |
| "grad_norm": 0.8359375, |
| "learning_rate": 1.726666666666667e-05, |
| "loss": 0.9384, |
| "mean_token_accuracy": 0.7573206882923842, |
| "num_tokens": 3578349.0, |
| "step": 260 |
| }, |
| { |
| "entropy": 0.9361576547846198, |
| "epoch": 0.216, |
| "grad_norm": 0.74609375, |
| "learning_rate": 1.7933333333333333e-05, |
| "loss": 0.9004, |
| "mean_token_accuracy": 0.7651511203497648, |
| "num_tokens": 3718484.0, |
| "step": 270 |
| }, |
| { |
| "entropy": 0.9637048680335283, |
| "epoch": 0.224, |
| "grad_norm": 0.9140625, |
| "learning_rate": 1.86e-05, |
| "loss": 0.9349, |
| "mean_token_accuracy": 0.7605184197425843, |
| "num_tokens": 3851369.0, |
| "step": 280 |
| }, |
| { |
| "entropy": 0.9367688067257405, |
| "epoch": 0.232, |
| "grad_norm": 0.74609375, |
| "learning_rate": 1.926666666666667e-05, |
| "loss": 0.8962, |
| "mean_token_accuracy": 0.7659803122282028, |
| "num_tokens": 3992977.0, |
| "step": 290 |
| }, |
| { |
| "entropy": 0.9229113388806581, |
| "epoch": 0.24, |
| "grad_norm": 0.7890625, |
| "learning_rate": 1.9933333333333334e-05, |
| "loss": 0.9024, |
| "mean_token_accuracy": 0.77176120467484, |
| "num_tokens": 4133723.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.24, |
| "eval_biology_entropy": 3.018038348674774, |
| "eval_biology_loss": 3.2969858646392822, |
| "eval_biology_mean_token_accuracy": 0.43762873542308806, |
| "eval_biology_num_tokens": 4133723.0, |
| "eval_biology_runtime": 27.1593, |
| "eval_biology_samples_per_second": 18.41, |
| "eval_biology_steps_per_second": 9.205, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.24, |
| "eval_chemistry_entropy": 1.778470279932022, |
| "eval_chemistry_loss": 1.8064701557159424, |
| "eval_chemistry_mean_token_accuracy": 0.6220316555500031, |
| "eval_chemistry_num_tokens": 4133723.0, |
| "eval_chemistry_runtime": 34.6422, |
| "eval_chemistry_samples_per_second": 14.433, |
| "eval_chemistry_steps_per_second": 7.217, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.24, |
| "eval_math_entropy": 0.9541566734313964, |
| "eval_math_loss": 0.9326547980308533, |
| "eval_math_mean_token_accuracy": 0.7629979259967804, |
| "eval_math_num_tokens": 4133723.0, |
| "eval_math_runtime": 35.1904, |
| "eval_math_samples_per_second": 14.208, |
| "eval_math_steps_per_second": 7.104, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.24, |
| "eval_physics_entropy": 1.387200428724289, |
| "eval_physics_loss": 1.3932181596755981, |
| "eval_physics_mean_token_accuracy": 0.6886798795461655, |
| "eval_physics_num_tokens": 4133723.0, |
| "eval_physics_runtime": 40.7999, |
| "eval_physics_samples_per_second": 12.255, |
| "eval_physics_steps_per_second": 6.127, |
| "step": 300 |
| }, |
| { |
| "entropy": 0.9078424336388707, |
| "epoch": 0.248, |
| "grad_norm": 0.75, |
| "learning_rate": 1.9933333333333334e-05, |
| "loss": 0.8542, |
| "mean_token_accuracy": 0.7745326109230518, |
| "num_tokens": 4273949.0, |
| "step": 310 |
| }, |
| { |
| "entropy": 0.9262684918940067, |
| "epoch": 0.256, |
| "grad_norm": 0.84765625, |
| "learning_rate": 1.985925925925926e-05, |
| "loss": 0.8761, |
| "mean_token_accuracy": 0.7701829686760903, |
| "num_tokens": 4414102.0, |
| "step": 320 |
| }, |
| { |
| "entropy": 0.9303955681622028, |
| "epoch": 0.264, |
| "grad_norm": 0.78515625, |
| "learning_rate": 1.9785185185185187e-05, |
| "loss": 0.8765, |
| "mean_token_accuracy": 0.7714239094406367, |
| "num_tokens": 4548402.0, |
| "step": 330 |
| }, |
| { |
| "entropy": 0.9207985673099757, |
| "epoch": 0.272, |
| "grad_norm": 0.7109375, |
| "learning_rate": 1.971111111111111e-05, |
| "loss": 0.8728, |
| "mean_token_accuracy": 0.7721738632768392, |
| "num_tokens": 4685739.0, |
| "step": 340 |
| }, |
| { |
| "entropy": 0.9172879874706268, |
| "epoch": 0.28, |
| "grad_norm": 0.80078125, |
| "learning_rate": 1.963703703703704e-05, |
| "loss": 0.869, |
| "mean_token_accuracy": 0.7719811659306288, |
| "num_tokens": 4827264.0, |
| "step": 350 |
| }, |
| { |
| "entropy": 0.8906141923740506, |
| "epoch": 0.288, |
| "grad_norm": 0.76171875, |
| "learning_rate": 1.9562962962962964e-05, |
| "loss": 0.8676, |
| "mean_token_accuracy": 0.779513431340456, |
| "num_tokens": 4963900.0, |
| "step": 360 |
| }, |
| { |
| "entropy": 0.9126916229724884, |
| "epoch": 0.296, |
| "grad_norm": 0.671875, |
| "learning_rate": 1.948888888888889e-05, |
| "loss": 0.8727, |
| "mean_token_accuracy": 0.7722058985382318, |
| "num_tokens": 5106308.0, |
| "step": 370 |
| }, |
| { |
| "entropy": 0.9109785668551922, |
| "epoch": 0.304, |
| "grad_norm": 0.76171875, |
| "learning_rate": 1.9414814814814817e-05, |
| "loss": 0.8562, |
| "mean_token_accuracy": 0.7752636637538671, |
| "num_tokens": 5246290.0, |
| "step": 380 |
| }, |
| { |
| "entropy": 0.9183575309813022, |
| "epoch": 0.312, |
| "grad_norm": 0.71875, |
| "learning_rate": 1.9340740740740743e-05, |
| "loss": 0.8621, |
| "mean_token_accuracy": 0.7713671930134296, |
| "num_tokens": 5389824.0, |
| "step": 390 |
| }, |
| { |
| "entropy": 0.8591190708801151, |
| "epoch": 0.32, |
| "grad_norm": 0.82421875, |
| "learning_rate": 1.926666666666667e-05, |
| "loss": 0.8065, |
| "mean_token_accuracy": 0.7856244947761297, |
| "num_tokens": 5531902.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_biology_entropy": 2.974372909069061, |
| "eval_biology_loss": 3.2137467861175537, |
| "eval_biology_mean_token_accuracy": 0.44984077060222627, |
| "eval_biology_num_tokens": 5531902.0, |
| "eval_biology_runtime": 26.9787, |
| "eval_biology_samples_per_second": 18.533, |
| "eval_biology_steps_per_second": 9.267, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_chemistry_entropy": 1.7318408770561218, |
| "eval_chemistry_loss": 1.746923804283142, |
| "eval_chemistry_mean_token_accuracy": 0.632696654677391, |
| "eval_chemistry_num_tokens": 5531902.0, |
| "eval_chemistry_runtime": 34.5046, |
| "eval_chemistry_samples_per_second": 14.491, |
| "eval_chemistry_steps_per_second": 7.245, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_math_entropy": 0.9170646283626557, |
| "eval_math_loss": 0.8942311406135559, |
| "eval_math_mean_token_accuracy": 0.7709064717292786, |
| "eval_math_num_tokens": 5531902.0, |
| "eval_math_runtime": 35.0734, |
| "eval_math_samples_per_second": 14.256, |
| "eval_math_steps_per_second": 7.128, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_physics_entropy": 1.349606383562088, |
| "eval_physics_loss": 1.343300223350525, |
| "eval_physics_mean_token_accuracy": 0.6978710038661957, |
| "eval_physics_num_tokens": 5531902.0, |
| "eval_physics_runtime": 40.7692, |
| "eval_physics_samples_per_second": 12.264, |
| "eval_physics_steps_per_second": 6.132, |
| "step": 400 |
| }, |
| { |
| "entropy": 0.9086290217936039, |
| "epoch": 0.328, |
| "grad_norm": 0.796875, |
| "learning_rate": 1.9192592592592593e-05, |
| "loss": 0.847, |
| "mean_token_accuracy": 0.7745547294616699, |
| "num_tokens": 5669047.0, |
| "step": 410 |
| }, |
| { |
| "entropy": 0.9156009530648589, |
| "epoch": 0.336, |
| "grad_norm": 0.7578125, |
| "learning_rate": 1.911851851851852e-05, |
| "loss": 0.8769, |
| "mean_token_accuracy": 0.7692474499344826, |
| "num_tokens": 5806889.0, |
| "step": 420 |
| }, |
| { |
| "entropy": 0.8934111742302775, |
| "epoch": 0.344, |
| "grad_norm": 0.67578125, |
| "learning_rate": 1.9044444444444446e-05, |
| "loss": 0.8462, |
| "mean_token_accuracy": 0.7796071924269199, |
| "num_tokens": 5947019.0, |
| "step": 430 |
| }, |
| { |
| "entropy": 0.8856402298435568, |
| "epoch": 0.352, |
| "grad_norm": 0.7578125, |
| "learning_rate": 1.8970370370370372e-05, |
| "loss": 0.8362, |
| "mean_token_accuracy": 0.7777645040303469, |
| "num_tokens": 6083443.0, |
| "step": 440 |
| }, |
| { |
| "entropy": 0.8964365001767873, |
| "epoch": 0.36, |
| "grad_norm": 0.7109375, |
| "learning_rate": 1.8896296296296295e-05, |
| "loss": 0.8406, |
| "mean_token_accuracy": 0.776063571870327, |
| "num_tokens": 6215195.0, |
| "step": 450 |
| }, |
| { |
| "entropy": 0.8810398764908314, |
| "epoch": 0.368, |
| "grad_norm": 0.69921875, |
| "learning_rate": 1.8822222222222225e-05, |
| "loss": 0.8247, |
| "mean_token_accuracy": 0.7789717759937048, |
| "num_tokens": 6357305.0, |
| "step": 460 |
| }, |
| { |
| "entropy": 0.8900227572768926, |
| "epoch": 0.376, |
| "grad_norm": 0.6875, |
| "learning_rate": 1.874814814814815e-05, |
| "loss": 0.8424, |
| "mean_token_accuracy": 0.7792754102498293, |
| "num_tokens": 6498815.0, |
| "step": 470 |
| }, |
| { |
| "entropy": 0.8727197473868727, |
| "epoch": 0.384, |
| "grad_norm": 0.68359375, |
| "learning_rate": 1.8674074074074075e-05, |
| "loss": 0.8209, |
| "mean_token_accuracy": 0.7820516306906938, |
| "num_tokens": 6635429.0, |
| "step": 480 |
| }, |
| { |
| "entropy": 0.8815433232113719, |
| "epoch": 0.392, |
| "grad_norm": 0.73828125, |
| "learning_rate": 1.86e-05, |
| "loss": 0.8274, |
| "mean_token_accuracy": 0.7792554065585137, |
| "num_tokens": 6774408.0, |
| "step": 490 |
| }, |
| { |
| "entropy": 0.9402453400194645, |
| "epoch": 0.4, |
| "grad_norm": 0.78515625, |
| "learning_rate": 1.8525925925925928e-05, |
| "loss": 0.9268, |
| "mean_token_accuracy": 0.7666260946542025, |
| "num_tokens": 6914489.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.4, |
| "eval_biology_entropy": 2.984537806510925, |
| "eval_biology_loss": 3.180291175842285, |
| "eval_biology_mean_token_accuracy": 0.4551255764961243, |
| "eval_biology_num_tokens": 6914489.0, |
| "eval_biology_runtime": 26.9382, |
| "eval_biology_samples_per_second": 18.561, |
| "eval_biology_steps_per_second": 9.281, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.4, |
| "eval_chemistry_entropy": 1.735370470046997, |
| "eval_chemistry_loss": 1.7186591625213623, |
| "eval_chemistry_mean_token_accuracy": 0.6382190730571747, |
| "eval_chemistry_num_tokens": 6914489.0, |
| "eval_chemistry_runtime": 34.4276, |
| "eval_chemistry_samples_per_second": 14.523, |
| "eval_chemistry_steps_per_second": 7.262, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.4, |
| "eval_math_entropy": 0.9168894953727722, |
| "eval_math_loss": 0.8749539852142334, |
| "eval_math_mean_token_accuracy": 0.7746825203895569, |
| "eval_math_num_tokens": 6914489.0, |
| "eval_math_runtime": 34.9026, |
| "eval_math_samples_per_second": 14.326, |
| "eval_math_steps_per_second": 7.163, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.4, |
| "eval_physics_entropy": 1.34889164853096, |
| "eval_physics_loss": 1.3155016899108887, |
| "eval_physics_mean_token_accuracy": 0.7029694278240204, |
| "eval_physics_num_tokens": 6914489.0, |
| "eval_physics_runtime": 40.5298, |
| "eval_physics_samples_per_second": 12.337, |
| "eval_physics_steps_per_second": 6.168, |
| "step": 500 |
| }, |
| { |
| "entropy": 0.8826847270131111, |
| "epoch": 0.408, |
| "grad_norm": 0.6328125, |
| "learning_rate": 1.8451851851851855e-05, |
| "loss": 0.8191, |
| "mean_token_accuracy": 0.7815356869250536, |
| "num_tokens": 7053065.0, |
| "step": 510 |
| }, |
| { |
| "entropy": 0.888327713869512, |
| "epoch": 0.416, |
| "grad_norm": 0.67578125, |
| "learning_rate": 1.8377777777777778e-05, |
| "loss": 0.8458, |
| "mean_token_accuracy": 0.7794966764748097, |
| "num_tokens": 7187783.0, |
| "step": 520 |
| }, |
| { |
| "entropy": 0.8952456833794713, |
| "epoch": 0.424, |
| "grad_norm": 0.66015625, |
| "learning_rate": 1.8303703703703704e-05, |
| "loss": 0.8535, |
| "mean_token_accuracy": 0.7754339158535004, |
| "num_tokens": 7320375.0, |
| "step": 530 |
| }, |
| { |
| "entropy": 0.8508788656443358, |
| "epoch": 0.432, |
| "grad_norm": 0.65625, |
| "learning_rate": 1.822962962962963e-05, |
| "loss": 0.7973, |
| "mean_token_accuracy": 0.7881547212600708, |
| "num_tokens": 7454401.0, |
| "step": 540 |
| }, |
| { |
| "entropy": 0.8842971453443169, |
| "epoch": 0.44, |
| "grad_norm": 0.71484375, |
| "learning_rate": 1.8155555555555557e-05, |
| "loss": 0.8373, |
| "mean_token_accuracy": 0.7774394288659096, |
| "num_tokens": 7593764.0, |
| "step": 550 |
| }, |
| { |
| "entropy": 0.8545080715790391, |
| "epoch": 0.448, |
| "grad_norm": 0.75, |
| "learning_rate": 1.8081481481481484e-05, |
| "loss": 0.8067, |
| "mean_token_accuracy": 0.7855051398277283, |
| "num_tokens": 7732710.0, |
| "step": 560 |
| }, |
| { |
| "entropy": 0.8548783052712678, |
| "epoch": 0.456, |
| "grad_norm": 0.70703125, |
| "learning_rate": 1.800740740740741e-05, |
| "loss": 0.8214, |
| "mean_token_accuracy": 0.7835484184324741, |
| "num_tokens": 7873721.0, |
| "step": 570 |
| }, |
| { |
| "entropy": 0.8870714059099555, |
| "epoch": 0.464, |
| "grad_norm": 0.5859375, |
| "learning_rate": 1.7933333333333333e-05, |
| "loss": 0.8303, |
| "mean_token_accuracy": 0.7804129023104907, |
| "num_tokens": 8012478.0, |
| "step": 580 |
| }, |
| { |
| "entropy": 0.8966930942609906, |
| "epoch": 0.472, |
| "grad_norm": 0.64453125, |
| "learning_rate": 1.785925925925926e-05, |
| "loss": 0.8442, |
| "mean_token_accuracy": 0.7768670186400414, |
| "num_tokens": 8153429.0, |
| "step": 590 |
| }, |
| { |
| "entropy": 0.8990984376519918, |
| "epoch": 0.48, |
| "grad_norm": 0.7265625, |
| "learning_rate": 1.7785185185185186e-05, |
| "loss": 0.8363, |
| "mean_token_accuracy": 0.7791902955621481, |
| "num_tokens": 8288086.0, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_biology_entropy": 3.026648554325104, |
| "eval_biology_loss": 3.1774065494537354, |
| "eval_biology_mean_token_accuracy": 0.45753026235103605, |
| "eval_biology_num_tokens": 8288086.0, |
| "eval_biology_runtime": 26.9914, |
| "eval_biology_samples_per_second": 18.524, |
| "eval_biology_steps_per_second": 9.262, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_chemistry_entropy": 1.7211642112731933, |
| "eval_chemistry_loss": 1.7037526369094849, |
| "eval_chemistry_mean_token_accuracy": 0.6409431176185608, |
| "eval_chemistry_num_tokens": 8288086.0, |
| "eval_chemistry_runtime": 34.4519, |
| "eval_chemistry_samples_per_second": 14.513, |
| "eval_chemistry_steps_per_second": 7.256, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_math_entropy": 0.8927879617214203, |
| "eval_math_loss": 0.863281786441803, |
| "eval_math_mean_token_accuracy": 0.7770743372440339, |
| "eval_math_num_tokens": 8288086.0, |
| "eval_math_runtime": 34.9704, |
| "eval_math_samples_per_second": 14.298, |
| "eval_math_steps_per_second": 7.149, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_physics_entropy": 1.3300827517509461, |
| "eval_physics_loss": 1.3015602827072144, |
| "eval_physics_mean_token_accuracy": 0.7059740360975265, |
| "eval_physics_num_tokens": 8288086.0, |
| "eval_physics_runtime": 40.5426, |
| "eval_physics_samples_per_second": 12.333, |
| "eval_physics_steps_per_second": 6.166, |
| "step": 600 |
| }, |
| { |
| "entropy": 0.8605974804610014, |
| "epoch": 0.488, |
| "grad_norm": 0.65234375, |
| "learning_rate": 1.7711111111111113e-05, |
| "loss": 0.8123, |
| "mean_token_accuracy": 0.7824984133243561, |
| "num_tokens": 8424958.0, |
| "step": 610 |
| }, |
| { |
| "entropy": 0.8556231519207358, |
| "epoch": 0.496, |
| "grad_norm": 0.73828125, |
| "learning_rate": 1.763703703703704e-05, |
| "loss": 0.8073, |
| "mean_token_accuracy": 0.7834980070590973, |
| "num_tokens": 8562656.0, |
| "step": 620 |
| }, |
| { |
| "entropy": 0.8707238981500268, |
| "epoch": 0.504, |
| "grad_norm": 0.63671875, |
| "learning_rate": 1.7562962962962962e-05, |
| "loss": 0.8323, |
| "mean_token_accuracy": 0.7818343795835971, |
| "num_tokens": 8706097.0, |
| "step": 630 |
| }, |
| { |
| "entropy": 0.8535389142110944, |
| "epoch": 0.512, |
| "grad_norm": 0.6953125, |
| "learning_rate": 1.7488888888888892e-05, |
| "loss": 0.7829, |
| "mean_token_accuracy": 0.7894029818475247, |
| "num_tokens": 8844762.0, |
| "step": 640 |
| }, |
| { |
| "entropy": 0.8502879558131099, |
| "epoch": 0.52, |
| "grad_norm": 0.73828125, |
| "learning_rate": 1.7414814814814815e-05, |
| "loss": 0.8016, |
| "mean_token_accuracy": 0.789192034304142, |
| "num_tokens": 8983268.0, |
| "step": 650 |
| }, |
| { |
| "entropy": 0.841147867590189, |
| "epoch": 0.528, |
| "grad_norm": 0.76171875, |
| "learning_rate": 1.7340740740740742e-05, |
| "loss": 0.7936, |
| "mean_token_accuracy": 0.7880374405533075, |
| "num_tokens": 9132725.0, |
| "step": 660 |
| }, |
| { |
| "entropy": 0.8636940447613597, |
| "epoch": 0.536, |
| "grad_norm": 0.796875, |
| "learning_rate": 1.726666666666667e-05, |
| "loss": 0.841, |
| "mean_token_accuracy": 0.7832605082541704, |
| "num_tokens": 9271490.0, |
| "step": 670 |
| }, |
| { |
| "entropy": 0.8671527728438377, |
| "epoch": 0.544, |
| "grad_norm": 0.78125, |
| "learning_rate": 1.7192592592592595e-05, |
| "loss": 0.8081, |
| "mean_token_accuracy": 0.7864414434880018, |
| "num_tokens": 9412327.0, |
| "step": 680 |
| }, |
| { |
| "entropy": 0.9130071861669421, |
| "epoch": 0.552, |
| "grad_norm": 0.85546875, |
| "learning_rate": 1.711851851851852e-05, |
| "loss": 0.8653, |
| "mean_token_accuracy": 0.7723087813705206, |
| "num_tokens": 9541196.0, |
| "step": 690 |
| }, |
| { |
| "entropy": 0.8585997836664319, |
| "epoch": 0.56, |
| "grad_norm": 0.7265625, |
| "learning_rate": 1.7044444444444445e-05, |
| "loss": 0.794, |
| "mean_token_accuracy": 0.7848813243210315, |
| "num_tokens": 9676595.0, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.56, |
| "eval_biology_entropy": 2.9588126630783083, |
| "eval_biology_loss": 3.134213447570801, |
| "eval_biology_mean_token_accuracy": 0.46100987315177916, |
| "eval_biology_num_tokens": 9676595.0, |
| "eval_biology_runtime": 27.0044, |
| "eval_biology_samples_per_second": 18.516, |
| "eval_biology_steps_per_second": 9.258, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.56, |
| "eval_chemistry_entropy": 1.6786610894203187, |
| "eval_chemistry_loss": 1.6839625835418701, |
| "eval_chemistry_mean_token_accuracy": 0.6436767561435699, |
| "eval_chemistry_num_tokens": 9676595.0, |
| "eval_chemistry_runtime": 34.4136, |
| "eval_chemistry_samples_per_second": 14.529, |
| "eval_chemistry_steps_per_second": 7.265, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.56, |
| "eval_math_entropy": 0.8769606277942658, |
| "eval_math_loss": 0.8541995882987976, |
| "eval_math_mean_token_accuracy": 0.7790154383182526, |
| "eval_math_num_tokens": 9676595.0, |
| "eval_math_runtime": 34.9482, |
| "eval_math_samples_per_second": 14.307, |
| "eval_math_steps_per_second": 7.153, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.56, |
| "eval_physics_entropy": 1.302037154197693, |
| "eval_physics_loss": 1.289175271987915, |
| "eval_physics_mean_token_accuracy": 0.7078491554260254, |
| "eval_physics_num_tokens": 9676595.0, |
| "eval_physics_runtime": 40.5408, |
| "eval_physics_samples_per_second": 12.333, |
| "eval_physics_steps_per_second": 6.167, |
| "step": 700 |
| }, |
| { |
| "entropy": 0.8488044740632177, |
| "epoch": 0.568, |
| "grad_norm": 0.65625, |
| "learning_rate": 1.697037037037037e-05, |
| "loss": 0.8027, |
| "mean_token_accuracy": 0.7859589237719774, |
| "num_tokens": 9812856.0, |
| "step": 710 |
| }, |
| { |
| "entropy": 0.8703772563487291, |
| "epoch": 0.576, |
| "grad_norm": 0.76171875, |
| "learning_rate": 1.6896296296296298e-05, |
| "loss": 0.8196, |
| "mean_token_accuracy": 0.7833330363035202, |
| "num_tokens": 9953227.0, |
| "step": 720 |
| }, |
| { |
| "entropy": 0.8656591676175595, |
| "epoch": 0.584, |
| "grad_norm": 0.7109375, |
| "learning_rate": 1.6822222222222224e-05, |
| "loss": 0.8193, |
| "mean_token_accuracy": 0.7825856961309909, |
| "num_tokens": 10089276.0, |
| "step": 730 |
| }, |
| { |
| "entropy": 0.8539155850186944, |
| "epoch": 0.592, |
| "grad_norm": 0.62109375, |
| "learning_rate": 1.6748148148148147e-05, |
| "loss": 0.8051, |
| "mean_token_accuracy": 0.7846140246838331, |
| "num_tokens": 10234220.0, |
| "step": 740 |
| }, |
| { |
| "entropy": 0.8651883415877819, |
| "epoch": 0.6, |
| "grad_norm": 0.62109375, |
| "learning_rate": 1.6674074074074077e-05, |
| "loss": 0.8268, |
| "mean_token_accuracy": 0.7814744919538498, |
| "num_tokens": 10370667.0, |
| "step": 750 |
| }, |
| { |
| "entropy": 0.863803020492196, |
| "epoch": 0.608, |
| "grad_norm": 0.76171875, |
| "learning_rate": 1.66e-05, |
| "loss": 0.8229, |
| "mean_token_accuracy": 0.783318106085062, |
| "num_tokens": 10515760.0, |
| "step": 760 |
| }, |
| { |
| "entropy": 0.8657625541090965, |
| "epoch": 0.616, |
| "grad_norm": 0.609375, |
| "learning_rate": 1.6525925925925927e-05, |
| "loss": 0.8201, |
| "mean_token_accuracy": 0.781252521276474, |
| "num_tokens": 10660578.0, |
| "step": 770 |
| }, |
| { |
| "entropy": 0.8430647728964686, |
| "epoch": 0.624, |
| "grad_norm": 0.73828125, |
| "learning_rate": 1.6451851851851853e-05, |
| "loss": 0.786, |
| "mean_token_accuracy": 0.7895114913582801, |
| "num_tokens": 10797566.0, |
| "step": 780 |
| }, |
| { |
| "entropy": 0.8240785686299205, |
| "epoch": 0.632, |
| "grad_norm": 0.63671875, |
| "learning_rate": 1.637777777777778e-05, |
| "loss": 0.7805, |
| "mean_token_accuracy": 0.7922219831496478, |
| "num_tokens": 10938456.0, |
| "step": 790 |
| }, |
| { |
| "entropy": 0.8485349236056209, |
| "epoch": 0.64, |
| "grad_norm": 0.65625, |
| "learning_rate": 1.6303703703703706e-05, |
| "loss": 0.7989, |
| "mean_token_accuracy": 0.7866497144103051, |
| "num_tokens": 11079348.0, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_biology_entropy": 2.932964876174927, |
| "eval_biology_loss": 3.1179301738739014, |
| "eval_biology_mean_token_accuracy": 0.46290264546871185, |
| "eval_biology_num_tokens": 11079348.0, |
| "eval_biology_runtime": 27.0006, |
| "eval_biology_samples_per_second": 18.518, |
| "eval_biology_steps_per_second": 9.259, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_chemistry_entropy": 1.6625523691177369, |
| "eval_chemistry_loss": 1.6735429763793945, |
| "eval_chemistry_mean_token_accuracy": 0.645549393773079, |
| "eval_chemistry_num_tokens": 11079348.0, |
| "eval_chemistry_runtime": 34.504, |
| "eval_chemistry_samples_per_second": 14.491, |
| "eval_chemistry_steps_per_second": 7.246, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_math_entropy": 0.8715307763814926, |
| "eval_math_loss": 0.8475318551063538, |
| "eval_math_mean_token_accuracy": 0.780082686662674, |
| "eval_math_num_tokens": 11079348.0, |
| "eval_math_runtime": 34.9793, |
| "eval_math_samples_per_second": 14.294, |
| "eval_math_steps_per_second": 7.147, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_physics_entropy": 1.292870044708252, |
| "eval_physics_loss": 1.2783890962600708, |
| "eval_physics_mean_token_accuracy": 0.7100210604667664, |
| "eval_physics_num_tokens": 11079348.0, |
| "eval_physics_runtime": 40.5422, |
| "eval_physics_samples_per_second": 12.333, |
| "eval_physics_steps_per_second": 6.166, |
| "step": 800 |
| }, |
| { |
| "entropy": 0.8196414019912481, |
| "epoch": 0.648, |
| "grad_norm": 0.6484375, |
| "learning_rate": 1.622962962962963e-05, |
| "loss": 0.7785, |
| "mean_token_accuracy": 0.7903594326227903, |
| "num_tokens": 11221757.0, |
| "step": 810 |
| }, |
| { |
| "entropy": 0.8491896335035563, |
| "epoch": 0.656, |
| "grad_norm": 0.66796875, |
| "learning_rate": 1.6155555555555556e-05, |
| "loss": 0.8188, |
| "mean_token_accuracy": 0.7861174423247576, |
| "num_tokens": 11360239.0, |
| "step": 820 |
| }, |
| { |
| "entropy": 0.8492498191073536, |
| "epoch": 0.664, |
| "grad_norm": 0.83984375, |
| "learning_rate": 1.6081481481481482e-05, |
| "loss": 0.7993, |
| "mean_token_accuracy": 0.7853016339242458, |
| "num_tokens": 11498677.0, |
| "step": 830 |
| }, |
| { |
| "entropy": 0.8525434667244554, |
| "epoch": 0.672, |
| "grad_norm": 0.69140625, |
| "learning_rate": 1.600740740740741e-05, |
| "loss": 0.8167, |
| "mean_token_accuracy": 0.7853171724826098, |
| "num_tokens": 11644674.0, |
| "step": 840 |
| }, |
| { |
| "entropy": 0.8380661699920893, |
| "epoch": 0.68, |
| "grad_norm": 0.73046875, |
| "learning_rate": 1.5933333333333336e-05, |
| "loss": 0.7928, |
| "mean_token_accuracy": 0.788429806381464, |
| "num_tokens": 11783080.0, |
| "step": 850 |
| }, |
| { |
| "entropy": 0.8381451951339841, |
| "epoch": 0.688, |
| "grad_norm": 0.62109375, |
| "learning_rate": 1.5859259259259262e-05, |
| "loss": 0.7921, |
| "mean_token_accuracy": 0.7877590708434582, |
| "num_tokens": 11917713.0, |
| "step": 860 |
| }, |
| { |
| "entropy": 0.8567008957266807, |
| "epoch": 0.696, |
| "grad_norm": 0.6640625, |
| "learning_rate": 1.5785185185185185e-05, |
| "loss": 0.7966, |
| "mean_token_accuracy": 0.7848114792257548, |
| "num_tokens": 12058193.0, |
| "step": 870 |
| }, |
| { |
| "entropy": 0.8353428209200502, |
| "epoch": 0.704, |
| "grad_norm": 0.65234375, |
| "learning_rate": 1.571111111111111e-05, |
| "loss": 0.7868, |
| "mean_token_accuracy": 0.7881882593035698, |
| "num_tokens": 12199067.0, |
| "step": 880 |
| }, |
| { |
| "entropy": 0.8710033852607012, |
| "epoch": 0.712, |
| "grad_norm": 0.671875, |
| "learning_rate": 1.5637037037037038e-05, |
| "loss": 0.8088, |
| "mean_token_accuracy": 0.7816738404333592, |
| "num_tokens": 12335740.0, |
| "step": 890 |
| }, |
| { |
| "entropy": 0.8482802568003536, |
| "epoch": 0.72, |
| "grad_norm": 0.8671875, |
| "learning_rate": 1.5562962962962965e-05, |
| "loss": 0.8002, |
| "mean_token_accuracy": 0.7863489333540201, |
| "num_tokens": 12469017.0, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.72, |
| "eval_biology_entropy": 2.9198746318817137, |
| "eval_biology_loss": 3.095550775527954, |
| "eval_biology_mean_token_accuracy": 0.4650931693315506, |
| "eval_biology_num_tokens": 12469017.0, |
| "eval_biology_runtime": 26.9459, |
| "eval_biology_samples_per_second": 18.556, |
| "eval_biology_steps_per_second": 9.278, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.72, |
| "eval_chemistry_entropy": 1.654184314250946, |
| "eval_chemistry_loss": 1.6625065803527832, |
| "eval_chemistry_mean_token_accuracy": 0.6471669510602951, |
| "eval_chemistry_num_tokens": 12469017.0, |
| "eval_chemistry_runtime": 34.4235, |
| "eval_chemistry_samples_per_second": 14.525, |
| "eval_chemistry_steps_per_second": 7.262, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.72, |
| "eval_math_entropy": 0.8690718128681183, |
| "eval_math_loss": 0.8421436548233032, |
| "eval_math_mean_token_accuracy": 0.7811065847873688, |
| "eval_math_num_tokens": 12469017.0, |
| "eval_math_runtime": 34.9224, |
| "eval_math_samples_per_second": 14.317, |
| "eval_math_steps_per_second": 7.159, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.72, |
| "eval_physics_entropy": 1.2865873792171478, |
| "eval_physics_loss": 1.2697285413742065, |
| "eval_physics_mean_token_accuracy": 0.7113078062534333, |
| "eval_physics_num_tokens": 12469017.0, |
| "eval_physics_runtime": 40.5051, |
| "eval_physics_samples_per_second": 12.344, |
| "eval_physics_steps_per_second": 6.172, |
| "step": 900 |
| }, |
| { |
| "entropy": 0.8559863271191717, |
| "epoch": 0.728, |
| "grad_norm": 0.859375, |
| "learning_rate": 1.548888888888889e-05, |
| "loss": 0.8087, |
| "mean_token_accuracy": 0.7872065585106611, |
| "num_tokens": 12604524.0, |
| "step": 910 |
| }, |
| { |
| "entropy": 0.8351603880524635, |
| "epoch": 0.736, |
| "grad_norm": 0.69921875, |
| "learning_rate": 1.5414814814814814e-05, |
| "loss": 0.7815, |
| "mean_token_accuracy": 0.7907860331237316, |
| "num_tokens": 12746145.0, |
| "step": 920 |
| }, |
| { |
| "entropy": 0.8385633688420058, |
| "epoch": 0.744, |
| "grad_norm": 0.68359375, |
| "learning_rate": 1.5340740740740744e-05, |
| "loss": 0.7849, |
| "mean_token_accuracy": 0.7884344782680273, |
| "num_tokens": 12882140.0, |
| "step": 930 |
| }, |
| { |
| "entropy": 0.8265942361205816, |
| "epoch": 0.752, |
| "grad_norm": 0.671875, |
| "learning_rate": 1.5266666666666667e-05, |
| "loss": 0.7683, |
| "mean_token_accuracy": 0.7920810837298632, |
| "num_tokens": 13019785.0, |
| "step": 940 |
| }, |
| { |
| "entropy": 0.801921004243195, |
| "epoch": 0.76, |
| "grad_norm": 0.64453125, |
| "learning_rate": 1.5192592592592594e-05, |
| "loss": 0.7567, |
| "mean_token_accuracy": 0.7981402266770601, |
| "num_tokens": 13167975.0, |
| "step": 950 |
| }, |
| { |
| "entropy": 0.8200628351420164, |
| "epoch": 0.768, |
| "grad_norm": 0.6015625, |
| "learning_rate": 1.5118518518518519e-05, |
| "loss": 0.7712, |
| "mean_token_accuracy": 0.7937319375574589, |
| "num_tokens": 13307071.0, |
| "step": 960 |
| }, |
| { |
| "entropy": 0.844639228284359, |
| "epoch": 0.776, |
| "grad_norm": 0.68359375, |
| "learning_rate": 1.5044444444444445e-05, |
| "loss": 0.8066, |
| "mean_token_accuracy": 0.7887587692588568, |
| "num_tokens": 13444206.0, |
| "step": 970 |
| }, |
| { |
| "entropy": 0.8560901025310159, |
| "epoch": 0.784, |
| "grad_norm": 0.7109375, |
| "learning_rate": 1.497037037037037e-05, |
| "loss": 0.8159, |
| "mean_token_accuracy": 0.7839560698717832, |
| "num_tokens": 13586074.0, |
| "step": 980 |
| }, |
| { |
| "entropy": 0.8473082648590207, |
| "epoch": 0.792, |
| "grad_norm": 0.60546875, |
| "learning_rate": 1.4896296296296298e-05, |
| "loss": 0.8055, |
| "mean_token_accuracy": 0.785613476485014, |
| "num_tokens": 13727642.0, |
| "step": 990 |
| }, |
| { |
| "entropy": 0.8641716981306672, |
| "epoch": 0.8, |
| "grad_norm": 0.73828125, |
| "learning_rate": 1.4822222222222225e-05, |
| "loss": 0.8161, |
| "mean_token_accuracy": 0.7843089539557695, |
| "num_tokens": 13869252.0, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_biology_entropy": 2.927406606197357, |
| "eval_biology_loss": 3.0816211700439453, |
| "eval_biology_mean_token_accuracy": 0.46697849237918854, |
| "eval_biology_num_tokens": 13869252.0, |
| "eval_biology_runtime": 26.9151, |
| "eval_biology_samples_per_second": 18.577, |
| "eval_biology_steps_per_second": 9.288, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_chemistry_entropy": 1.6538746666908264, |
| "eval_chemistry_loss": 1.6552778482437134, |
| "eval_chemistry_mean_token_accuracy": 0.6485219527482986, |
| "eval_chemistry_num_tokens": 13869252.0, |
| "eval_chemistry_runtime": 34.4246, |
| "eval_chemistry_samples_per_second": 14.525, |
| "eval_chemistry_steps_per_second": 7.262, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_math_entropy": 0.8677517844438553, |
| "eval_math_loss": 0.8380001187324524, |
| "eval_math_mean_token_accuracy": 0.7821021761894226, |
| "eval_math_num_tokens": 13869252.0, |
| "eval_math_runtime": 34.92, |
| "eval_math_samples_per_second": 14.318, |
| "eval_math_steps_per_second": 7.159, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_physics_entropy": 1.2842487235069275, |
| "eval_physics_loss": 1.2633119821548462, |
| "eval_physics_mean_token_accuracy": 0.7128965456485749, |
| "eval_physics_num_tokens": 13869252.0, |
| "eval_physics_runtime": 40.4838, |
| "eval_physics_samples_per_second": 12.351, |
| "eval_physics_steps_per_second": 6.175, |
| "step": 1000 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 3000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "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": 7.281015907772006e+17, |
| "train_batch_size": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
