| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 40.99733333333333, | |
| "eval_steps": 500, | |
| "global_step": 7687, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 0.275871217250824, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8545, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9973333333333333, | |
| "eval_accuracy": 0.6037399103139014, | |
| "eval_loss": 1.7123007774353027, | |
| "eval_runtime": 6.2976, | |
| "eval_samples_per_second": 79.395, | |
| "eval_steps_per_second": 10.004, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.9973333333333333, | |
| "eval_exact_match": 14.4, | |
| "eval_f1": 22.843015873015876, | |
| "eval_qa_bleu": 9.471299126615676, | |
| "eval_qa_exact_match": 0.112, | |
| "eval_recite_bleu": 12.267251092702137, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.32765188813209534, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7408, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.3532409369945526, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7063, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6058565022421525, | |
| "eval_loss": 1.69453763961792, | |
| "eval_runtime": 6.5928, | |
| "eval_samples_per_second": 75.841, | |
| "eval_steps_per_second": 9.556, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_exact_match": 14.4, | |
| "eval_f1": 23.29731601731603, | |
| "eval_qa_bleu": 9.4408132575913, | |
| "eval_qa_exact_match": 0.11, | |
| "eval_recite_bleu": 13.027794116434837, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.3777780532836914, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7094, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.3971036374568939, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6702, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.997333333333333, | |
| "eval_accuracy": 0.6065919282511211, | |
| "eval_loss": 1.6851236820220947, | |
| "eval_runtime": 6.3104, | |
| "eval_samples_per_second": 79.234, | |
| "eval_steps_per_second": 9.984, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.997333333333333, | |
| "eval_exact_match": 14.0, | |
| "eval_f1": 23.306825396825417, | |
| "eval_qa_bleu": 8.683248465366908, | |
| "eval_qa_exact_match": 0.11, | |
| "eval_recite_bleu": 12.39336052958371, | |
| "eval_recite_exact_match": 0.006, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.4782978594303131, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6574, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.7333333333333334, | |
| "grad_norm": 0.52582186460495, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6356, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6074977578475337, | |
| "eval_loss": 1.682567834854126, | |
| "eval_runtime": 6.65, | |
| "eval_samples_per_second": 75.188, | |
| "eval_steps_per_second": 9.474, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_exact_match": 15.0, | |
| "eval_f1": 23.366825396825416, | |
| "eval_qa_bleu": 9.05994637010322, | |
| "eval_qa_exact_match": 0.114, | |
| "eval_recite_bleu": 13.060415058384661, | |
| "eval_recite_exact_match": 0.006, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 0.6418241262435913, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6005, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.6805307269096375, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5775, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.997333333333334, | |
| "eval_accuracy": 0.6071031390134529, | |
| "eval_loss": 1.69106125831604, | |
| "eval_runtime": 5.6598, | |
| "eval_samples_per_second": 88.343, | |
| "eval_steps_per_second": 11.131, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 4.997333333333334, | |
| "eval_exact_match": 15.2, | |
| "eval_f1": 23.38539682539684, | |
| "eval_qa_bleu": 11.854591678140883, | |
| "eval_qa_exact_match": 0.112, | |
| "eval_recite_bleu": 12.694062454678122, | |
| "eval_recite_exact_match": 0.004, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 5.333333333333333, | |
| "grad_norm": 0.8362048864364624, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5173, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.866666666666667, | |
| "grad_norm": 0.8521270155906677, | |
| "learning_rate": 3e-05, | |
| "loss": 1.529, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6069417040358744, | |
| "eval_loss": 1.7066757678985596, | |
| "eval_runtime": 6.2922, | |
| "eval_samples_per_second": 79.464, | |
| "eval_steps_per_second": 10.012, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_exact_match": 15.2, | |
| "eval_f1": 24.04682539682541, | |
| "eval_qa_bleu": 12.150293484705452, | |
| "eval_qa_exact_match": 0.106, | |
| "eval_recite_bleu": 12.732083161926754, | |
| "eval_recite_exact_match": 0.002, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 0.9372308254241943, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4613, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 6.933333333333334, | |
| "grad_norm": 1.0236693620681763, | |
| "learning_rate": 3e-05, | |
| "loss": 1.457, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 6.997333333333334, | |
| "eval_accuracy": 0.6055515695067265, | |
| "eval_loss": 1.72792649269104, | |
| "eval_runtime": 6.3278, | |
| "eval_samples_per_second": 79.016, | |
| "eval_steps_per_second": 9.956, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 6.997333333333334, | |
| "eval_exact_match": 15.0, | |
| "eval_f1": 23.28492063492065, | |
| "eval_qa_bleu": 9.849658422779212, | |
| "eval_qa_exact_match": 0.114, | |
| "eval_recite_bleu": 12.3997433702838, | |
| "eval_recite_exact_match": 0.006, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 7.466666666666667, | |
| "grad_norm": 1.0773049592971802, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3996, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.0705862045288086, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3907, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6046188340807175, | |
| "eval_loss": 1.7511730194091797, | |
| "eval_runtime": 5.5895, | |
| "eval_samples_per_second": 89.453, | |
| "eval_steps_per_second": 11.271, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_exact_match": 13.0, | |
| "eval_f1": 19.93238095238096, | |
| "eval_qa_bleu": 11.93942669425287, | |
| "eval_qa_exact_match": 0.098, | |
| "eval_recite_bleu": 11.949414001720045, | |
| "eval_recite_exact_match": 0.006, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 8.533333333333333, | |
| "grad_norm": 1.2424529790878296, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3309, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 8.997333333333334, | |
| "eval_accuracy": 0.6025022421524664, | |
| "eval_loss": 1.7774019241333008, | |
| "eval_runtime": 6.4013, | |
| "eval_samples_per_second": 78.109, | |
| "eval_steps_per_second": 9.842, | |
| "step": 1687 | |
| }, | |
| { | |
| "epoch": 8.997333333333334, | |
| "eval_exact_match": 13.6, | |
| "eval_f1": 21.089523809523815, | |
| "eval_qa_bleu": 10.094627644241111, | |
| "eval_qa_exact_match": 0.094, | |
| "eval_recite_bleu": 12.197681191541987, | |
| "eval_recite_exact_match": 0.006, | |
| "step": 1687 | |
| }, | |
| { | |
| "epoch": 9.066666666666666, | |
| "grad_norm": 1.535229206085205, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3336, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 1.4936225414276123, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2841, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.6012645739910314, | |
| "eval_loss": 1.804250717163086, | |
| "eval_runtime": 6.385, | |
| "eval_samples_per_second": 78.309, | |
| "eval_steps_per_second": 9.867, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_exact_match": 13.2, | |
| "eval_f1": 21.172832722832737, | |
| "eval_qa_bleu": 9.841075001407239, | |
| "eval_qa_exact_match": 0.096, | |
| "eval_recite_bleu": 11.691398764835105, | |
| "eval_recite_exact_match": 0.004, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 10.133333333333333, | |
| "grad_norm": 1.6687729358673096, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2562, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 10.666666666666666, | |
| "grad_norm": 1.437538504600525, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2308, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 10.997333333333334, | |
| "eval_accuracy": 0.6000717488789238, | |
| "eval_loss": 1.8528188467025757, | |
| "eval_runtime": 6.6644, | |
| "eval_samples_per_second": 75.026, | |
| "eval_steps_per_second": 9.453, | |
| "step": 2062 | |
| }, | |
| { | |
| "epoch": 10.997333333333334, | |
| "eval_exact_match": 12.6, | |
| "eval_f1": 20.865213675213692, | |
| "eval_qa_bleu": 7.784433449171618, | |
| "eval_qa_exact_match": 0.108, | |
| "eval_recite_bleu": 12.305413068460787, | |
| "eval_recite_exact_match": 0.008, | |
| "step": 2062 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 1.708260416984558, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2011, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 11.733333333333333, | |
| "grad_norm": 1.7026830911636353, | |
| "learning_rate": 3e-05, | |
| "loss": 1.1722, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.5987892376681614, | |
| "eval_loss": 1.8851206302642822, | |
| "eval_runtime": 5.5764, | |
| "eval_samples_per_second": 89.663, | |
| "eval_steps_per_second": 11.298, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_exact_match": 13.0, | |
| "eval_f1": 21.17809523809525, | |
| "eval_qa_bleu": 11.19937598521782, | |
| "eval_qa_exact_match": 0.096, | |
| "eval_recite_bleu": 12.500534642121808, | |
| "eval_recite_exact_match": 0.01, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 12.266666666666667, | |
| "grad_norm": 1.6281794309616089, | |
| "learning_rate": 3e-05, | |
| "loss": 1.151, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 1.686490535736084, | |
| "learning_rate": 3e-05, | |
| "loss": 1.1354, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 12.997333333333334, | |
| "eval_accuracy": 0.5979910313901345, | |
| "eval_loss": 1.9114351272583008, | |
| "eval_runtime": 5.5718, | |
| "eval_samples_per_second": 89.737, | |
| "eval_steps_per_second": 11.307, | |
| "step": 2437 | |
| }, | |
| { | |
| "epoch": 12.997333333333334, | |
| "eval_exact_match": 13.8, | |
| "eval_f1": 21.856666666666673, | |
| "eval_qa_bleu": 10.206087008658956, | |
| "eval_qa_exact_match": 0.116, | |
| "eval_recite_bleu": 12.669262328097592, | |
| "eval_recite_exact_match": 0.008, | |
| "step": 2437 | |
| }, | |
| { | |
| "epoch": 13.333333333333334, | |
| "grad_norm": 1.8666125535964966, | |
| "learning_rate": 3e-05, | |
| "loss": 1.097, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 13.866666666666667, | |
| "grad_norm": 1.9091932773590088, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0793, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.5960986547085202, | |
| "eval_loss": 1.9584920406341553, | |
| "eval_runtime": 6.3056, | |
| "eval_samples_per_second": 79.294, | |
| "eval_steps_per_second": 9.991, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_exact_match": 10.6, | |
| "eval_f1": 18.697936507936518, | |
| "eval_qa_bleu": 6.4581555766931436, | |
| "eval_qa_exact_match": 0.082, | |
| "eval_recite_bleu": 12.843526576123885, | |
| "eval_recite_exact_match": 0.01, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "grad_norm": 2.042712688446045, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0376, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 14.933333333333334, | |
| "grad_norm": 1.8875784873962402, | |
| "learning_rate": 3e-05, | |
| "loss": 1.037, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 14.997333333333334, | |
| "eval_accuracy": 0.5948161434977578, | |
| "eval_loss": 1.9966908693313599, | |
| "eval_runtime": 6.0322, | |
| "eval_samples_per_second": 82.888, | |
| "eval_steps_per_second": 10.444, | |
| "step": 2812 | |
| }, | |
| { | |
| "epoch": 14.997333333333334, | |
| "eval_exact_match": 12.8, | |
| "eval_f1": 20.438917748917753, | |
| "eval_qa_bleu": 10.19364322169534, | |
| "eval_qa_exact_match": 0.104, | |
| "eval_recite_bleu": 12.69742056629334, | |
| "eval_recite_exact_match": 0.012, | |
| "step": 2812 | |
| }, | |
| { | |
| "epoch": 15.466666666666667, | |
| "grad_norm": 2.027426242828369, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9844, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 2.0708699226379395, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9901, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.5934080717488789, | |
| "eval_loss": 2.0336477756500244, | |
| "eval_runtime": 5.9811, | |
| "eval_samples_per_second": 83.597, | |
| "eval_steps_per_second": 10.533, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_exact_match": 12.4, | |
| "eval_f1": 20.427142857142865, | |
| "eval_qa_bleu": 9.13217752730269, | |
| "eval_qa_exact_match": 0.096, | |
| "eval_recite_bleu": 12.688250071816965, | |
| "eval_recite_exact_match": 0.014, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 16.533333333333335, | |
| "grad_norm": 1.995875358581543, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9316, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 16.997333333333334, | |
| "eval_accuracy": 0.591372197309417, | |
| "eval_loss": 2.088000535964966, | |
| "eval_runtime": 5.9631, | |
| "eval_samples_per_second": 83.849, | |
| "eval_steps_per_second": 10.565, | |
| "step": 3187 | |
| }, | |
| { | |
| "epoch": 16.997333333333334, | |
| "eval_exact_match": 11.4, | |
| "eval_f1": 18.528571428571443, | |
| "eval_qa_bleu": 7.115393148941145, | |
| "eval_qa_exact_match": 0.084, | |
| "eval_recite_bleu": 12.219544702526637, | |
| "eval_recite_exact_match": 0.014, | |
| "step": 3187 | |
| }, | |
| { | |
| "epoch": 17.066666666666666, | |
| "grad_norm": 2.623262643814087, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9354, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "grad_norm": 2.4125397205352783, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8802, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.5900627802690583, | |
| "eval_loss": 2.1439507007598877, | |
| "eval_runtime": 5.8546, | |
| "eval_samples_per_second": 85.404, | |
| "eval_steps_per_second": 10.761, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_exact_match": 10.8, | |
| "eval_f1": 17.508095238095244, | |
| "eval_qa_bleu": 7.769457934505562, | |
| "eval_qa_exact_match": 0.09, | |
| "eval_recite_bleu": 12.155079189562132, | |
| "eval_recite_exact_match": 0.016, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 18.133333333333333, | |
| "grad_norm": 2.3584065437316895, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8913, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 18.666666666666668, | |
| "grad_norm": 2.130357265472412, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8382, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 18.997333333333334, | |
| "eval_accuracy": 0.5892645739910314, | |
| "eval_loss": 2.171532154083252, | |
| "eval_runtime": 6.6257, | |
| "eval_samples_per_second": 75.464, | |
| "eval_steps_per_second": 9.508, | |
| "step": 3562 | |
| }, | |
| { | |
| "epoch": 18.997333333333334, | |
| "eval_exact_match": 13.0, | |
| "eval_f1": 18.838095238095242, | |
| "eval_qa_bleu": 7.682885911703284, | |
| "eval_qa_exact_match": 0.106, | |
| "eval_recite_bleu": 12.350855746597011, | |
| "eval_recite_exact_match": 0.014, | |
| "step": 3562 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "grad_norm": 2.6397740840911865, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8248, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 19.733333333333334, | |
| "grad_norm": 3.049431085586548, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7962, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.5878834080717489, | |
| "eval_loss": 2.2236711978912354, | |
| "eval_runtime": 6.764, | |
| "eval_samples_per_second": 73.921, | |
| "eval_steps_per_second": 9.314, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_exact_match": 12.2, | |
| "eval_f1": 18.373809523809534, | |
| "eval_qa_bleu": 8.448974659181921, | |
| "eval_qa_exact_match": 0.096, | |
| "eval_recite_bleu": 11.298504528713615, | |
| "eval_recite_exact_match": 0.014, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 20.266666666666666, | |
| "grad_norm": 2.647123098373413, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7797, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 20.8, | |
| "grad_norm": 2.798877239227295, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7553, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 20.997333333333334, | |
| "eval_accuracy": 0.5860896860986547, | |
| "eval_loss": 2.295654535293579, | |
| "eval_runtime": 5.5847, | |
| "eval_samples_per_second": 89.531, | |
| "eval_steps_per_second": 11.281, | |
| "step": 3937 | |
| }, | |
| { | |
| "epoch": 20.997333333333334, | |
| "eval_exact_match": 10.2, | |
| "eval_f1": 16.601428571428578, | |
| "eval_qa_bleu": 7.90021052694059, | |
| "eval_qa_exact_match": 0.08, | |
| "eval_recite_bleu": 11.716166611151118, | |
| "eval_recite_exact_match": 0.01, | |
| "step": 3937 | |
| }, | |
| { | |
| "epoch": 21.333333333333332, | |
| "grad_norm": 2.7208638191223145, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7306, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 21.866666666666667, | |
| "grad_norm": 2.9887781143188477, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7238, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.5850941704035875, | |
| "eval_loss": 2.331202507019043, | |
| "eval_runtime": 6.6871, | |
| "eval_samples_per_second": 74.771, | |
| "eval_steps_per_second": 9.421, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_exact_match": 10.8, | |
| "eval_f1": 17.400692640692647, | |
| "eval_qa_bleu": 4.77963863209427, | |
| "eval_qa_exact_match": 0.084, | |
| "eval_recite_bleu": 11.404518212072734, | |
| "eval_recite_exact_match": 0.014, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 22.4, | |
| "grad_norm": 2.706368923187256, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6813, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 22.933333333333334, | |
| "grad_norm": 2.850799798965454, | |
| "learning_rate": 3e-05, | |
| "loss": 0.676, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 22.997333333333334, | |
| "eval_accuracy": 0.583237668161435, | |
| "eval_loss": 2.4042670726776123, | |
| "eval_runtime": 6.5747, | |
| "eval_samples_per_second": 76.049, | |
| "eval_steps_per_second": 9.582, | |
| "step": 4312 | |
| }, | |
| { | |
| "epoch": 22.997333333333334, | |
| "eval_exact_match": 8.6, | |
| "eval_f1": 16.314285714285724, | |
| "eval_qa_bleu": 7.466573197185606, | |
| "eval_qa_exact_match": 0.064, | |
| "eval_recite_bleu": 11.568247151151132, | |
| "eval_recite_exact_match": 0.014, | |
| "step": 4312 | |
| }, | |
| { | |
| "epoch": 23.466666666666665, | |
| "grad_norm": 2.95532488822937, | |
| "learning_rate": 3e-05, | |
| "loss": 0.632, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 3.36663556098938, | |
| "learning_rate": 3e-05, | |
| "loss": 0.644, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.5824125560538117, | |
| "eval_loss": 2.443950891494751, | |
| "eval_runtime": 5.7069, | |
| "eval_samples_per_second": 87.614, | |
| "eval_steps_per_second": 11.039, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_exact_match": 8.8, | |
| "eval_f1": 15.491948051948057, | |
| "eval_qa_bleu": 4.531789460915551, | |
| "eval_qa_exact_match": 0.076, | |
| "eval_recite_bleu": 10.65729884465588, | |
| "eval_recite_exact_match": 0.014, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 24.533333333333335, | |
| "grad_norm": 3.074622631072998, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5939, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 24.997333333333334, | |
| "eval_accuracy": 0.5818385650224215, | |
| "eval_loss": 2.5127182006835938, | |
| "eval_runtime": 5.6588, | |
| "eval_samples_per_second": 88.357, | |
| "eval_steps_per_second": 11.133, | |
| "step": 4687 | |
| }, | |
| { | |
| "epoch": 24.997333333333334, | |
| "eval_exact_match": 9.0, | |
| "eval_f1": 16.361587301587303, | |
| "eval_qa_bleu": 7.691728089554307, | |
| "eval_qa_exact_match": 0.068, | |
| "eval_recite_bleu": 11.016738678559781, | |
| "eval_recite_exact_match": 0.008, | |
| "step": 4687 | |
| }, | |
| { | |
| "epoch": 25.066666666666666, | |
| "grad_norm": 2.702502727508545, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5904, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 25.6, | |
| "grad_norm": 3.233320951461792, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5551, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 0.5809596412556054, | |
| "eval_loss": 2.538985252380371, | |
| "eval_runtime": 6.6838, | |
| "eval_samples_per_second": 74.808, | |
| "eval_steps_per_second": 9.426, | |
| "step": 4875 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_exact_match": 9.2, | |
| "eval_f1": 16.819884559884564, | |
| "eval_qa_bleu": 4.658892987257927, | |
| "eval_qa_exact_match": 0.078, | |
| "eval_recite_bleu": 10.853797139069188, | |
| "eval_recite_exact_match": 0.012, | |
| "step": 4875 | |
| }, | |
| { | |
| "epoch": 26.133333333333333, | |
| "grad_norm": 3.280097246170044, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5503, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 26.666666666666668, | |
| "grad_norm": 3.1984827518463135, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5163, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 26.997333333333334, | |
| "eval_accuracy": 0.579829596412556, | |
| "eval_loss": 2.5808944702148438, | |
| "eval_runtime": 6.6339, | |
| "eval_samples_per_second": 75.37, | |
| "eval_steps_per_second": 9.497, | |
| "step": 5062 | |
| }, | |
| { | |
| "epoch": 26.997333333333334, | |
| "eval_exact_match": 8.0, | |
| "eval_f1": 16.641197691197693, | |
| "eval_qa_bleu": 5.575351494624609, | |
| "eval_qa_exact_match": 0.066, | |
| "eval_recite_bleu": 11.099361911324253, | |
| "eval_recite_exact_match": 0.01, | |
| "step": 5062 | |
| }, | |
| { | |
| "epoch": 27.2, | |
| "grad_norm": 3.5599937438964844, | |
| "learning_rate": 3e-05, | |
| "loss": 0.52, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 27.733333333333334, | |
| "grad_norm": 2.9382145404815674, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4892, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.5788609865470852, | |
| "eval_loss": 2.6670191287994385, | |
| "eval_runtime": 6.0585, | |
| "eval_samples_per_second": 82.529, | |
| "eval_steps_per_second": 10.399, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_exact_match": 8.6, | |
| "eval_f1": 15.144920634920638, | |
| "eval_qa_bleu": 5.387088284739065, | |
| "eval_qa_exact_match": 0.074, | |
| "eval_recite_bleu": 11.457740165981269, | |
| "eval_recite_exact_match": 0.012, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 28.266666666666666, | |
| "grad_norm": 4.097602367401123, | |
| "learning_rate": 3e-05, | |
| "loss": 0.47, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 28.8, | |
| "grad_norm": 3.9295685291290283, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4669, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 28.997333333333334, | |
| "eval_accuracy": 0.578609865470852, | |
| "eval_loss": 2.669480085372925, | |
| "eval_runtime": 6.6992, | |
| "eval_samples_per_second": 74.636, | |
| "eval_steps_per_second": 9.404, | |
| "step": 5437 | |
| }, | |
| { | |
| "epoch": 28.997333333333334, | |
| "eval_exact_match": 9.2, | |
| "eval_f1": 16.65718614718615, | |
| "eval_qa_bleu": 4.637892076000507, | |
| "eval_qa_exact_match": 0.074, | |
| "eval_recite_bleu": 11.77754451628722, | |
| "eval_recite_exact_match": 0.012, | |
| "step": 5437 | |
| }, | |
| { | |
| "epoch": 29.333333333333332, | |
| "grad_norm": 3.4413976669311523, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4398, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 29.866666666666667, | |
| "grad_norm": 2.88114857673645, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4353, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_accuracy": 0.5786995515695067, | |
| "eval_loss": 2.7645530700683594, | |
| "eval_runtime": 6.3214, | |
| "eval_samples_per_second": 79.096, | |
| "eval_steps_per_second": 9.966, | |
| "step": 5625 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_exact_match": 8.0, | |
| "eval_f1": 14.828455988455993, | |
| "eval_qa_bleu": 7.042137393349148, | |
| "eval_qa_exact_match": 0.068, | |
| "eval_recite_bleu": 11.413791110087148, | |
| "eval_recite_exact_match": 0.01, | |
| "step": 5625 | |
| }, | |
| { | |
| "epoch": 30.4, | |
| "grad_norm": 4.096495628356934, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4032, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 30.933333333333334, | |
| "grad_norm": 3.567011833190918, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4104, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 30.997333333333334, | |
| "eval_accuracy": 0.5774887892376681, | |
| "eval_loss": 2.829072952270508, | |
| "eval_runtime": 6.655, | |
| "eval_samples_per_second": 75.131, | |
| "eval_steps_per_second": 9.467, | |
| "step": 5812 | |
| }, | |
| { | |
| "epoch": 30.997333333333334, | |
| "eval_exact_match": 8.8, | |
| "eval_f1": 14.71683982683983, | |
| "eval_qa_bleu": 7.548806097282496, | |
| "eval_qa_exact_match": 0.072, | |
| "eval_recite_bleu": 10.549616368516189, | |
| "eval_recite_exact_match": 0.01, | |
| "step": 5812 | |
| }, | |
| { | |
| "epoch": 31.466666666666665, | |
| "grad_norm": 3.1184980869293213, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3695, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 3.125758647918701, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3885, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_accuracy": 0.5763587443946189, | |
| "eval_loss": 2.893293857574463, | |
| "eval_runtime": 5.9693, | |
| "eval_samples_per_second": 83.762, | |
| "eval_steps_per_second": 10.554, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_exact_match": 8.6, | |
| "eval_f1": 15.2548051948052, | |
| "eval_qa_bleu": 5.915179328505316, | |
| "eval_qa_exact_match": 0.072, | |
| "eval_recite_bleu": 10.647049650020403, | |
| "eval_recite_exact_match": 0.008, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 32.53333333333333, | |
| "grad_norm": 2.6737849712371826, | |
| "learning_rate": 3e-05, | |
| "loss": 0.342, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 32.99733333333333, | |
| "eval_accuracy": 0.5755964125560538, | |
| "eval_loss": 2.9434149265289307, | |
| "eval_runtime": 6.0396, | |
| "eval_samples_per_second": 82.787, | |
| "eval_steps_per_second": 10.431, | |
| "step": 6187 | |
| }, | |
| { | |
| "epoch": 32.99733333333333, | |
| "eval_exact_match": 8.2, | |
| "eval_f1": 14.65269841269841, | |
| "eval_qa_bleu": 8.616100659809678, | |
| "eval_qa_exact_match": 0.068, | |
| "eval_recite_bleu": 11.341262248022645, | |
| "eval_recite_exact_match": 0.014, | |
| "step": 6187 | |
| }, | |
| { | |
| "epoch": 33.06666666666667, | |
| "grad_norm": 2.490464448928833, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3528, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 33.6, | |
| "grad_norm": 3.0587589740753174, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3213, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_accuracy": 0.5755874439461883, | |
| "eval_loss": 2.934598207473755, | |
| "eval_runtime": 5.9452, | |
| "eval_samples_per_second": 84.101, | |
| "eval_steps_per_second": 10.597, | |
| "step": 6375 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_exact_match": 8.2, | |
| "eval_f1": 13.88111111111111, | |
| "eval_qa_bleu": 8.221633718322723, | |
| "eval_qa_exact_match": 0.066, | |
| "eval_recite_bleu": 11.218940899233823, | |
| "eval_recite_exact_match": 0.018, | |
| "step": 6375 | |
| }, | |
| { | |
| "epoch": 34.13333333333333, | |
| "grad_norm": 3.1941940784454346, | |
| "learning_rate": 3e-05, | |
| "loss": 0.328, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 34.666666666666664, | |
| "grad_norm": 3.2870864868164062, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3065, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 34.99733333333333, | |
| "eval_accuracy": 0.5757757847533632, | |
| "eval_loss": 3.008178949356079, | |
| "eval_runtime": 6.5953, | |
| "eval_samples_per_second": 75.811, | |
| "eval_steps_per_second": 9.552, | |
| "step": 6562 | |
| }, | |
| { | |
| "epoch": 34.99733333333333, | |
| "eval_exact_match": 9.6, | |
| "eval_f1": 15.528095238095238, | |
| "eval_qa_bleu": 6.040085184981805, | |
| "eval_qa_exact_match": 0.066, | |
| "eval_recite_bleu": 11.118518942785343, | |
| "eval_recite_exact_match": 0.016, | |
| "step": 6562 | |
| }, | |
| { | |
| "epoch": 35.2, | |
| "grad_norm": 3.2354586124420166, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3006, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 35.733333333333334, | |
| "grad_norm": 3.262101173400879, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2842, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_accuracy": 0.5739013452914798, | |
| "eval_loss": 3.0947136878967285, | |
| "eval_runtime": 6.5534, | |
| "eval_samples_per_second": 76.297, | |
| "eval_steps_per_second": 9.613, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_exact_match": 10.0, | |
| "eval_f1": 17.241428571428575, | |
| "eval_qa_bleu": 8.55409762541613, | |
| "eval_qa_exact_match": 0.08, | |
| "eval_recite_bleu": 11.315935831255379, | |
| "eval_recite_exact_match": 0.014, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 36.266666666666666, | |
| "grad_norm": 2.951231002807617, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2794, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 36.8, | |
| "grad_norm": 3.770897626876831, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2695, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 36.99733333333333, | |
| "eval_accuracy": 0.5751569506726457, | |
| "eval_loss": 3.0905086994171143, | |
| "eval_runtime": 5.6522, | |
| "eval_samples_per_second": 88.46, | |
| "eval_steps_per_second": 11.146, | |
| "step": 6937 | |
| }, | |
| { | |
| "epoch": 36.99733333333333, | |
| "eval_exact_match": 7.0, | |
| "eval_f1": 14.115916305916304, | |
| "eval_qa_bleu": 7.552037183523391, | |
| "eval_qa_exact_match": 0.054, | |
| "eval_recite_bleu": 10.863808387375153, | |
| "eval_recite_exact_match": 0.01, | |
| "step": 6937 | |
| }, | |
| { | |
| "epoch": 37.333333333333336, | |
| "grad_norm": 3.664700984954834, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2606, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 37.86666666666667, | |
| "grad_norm": 3.442922830581665, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2541, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_accuracy": 0.5738385650224215, | |
| "eval_loss": 3.1831231117248535, | |
| "eval_runtime": 6.7239, | |
| "eval_samples_per_second": 74.361, | |
| "eval_steps_per_second": 9.37, | |
| "step": 7125 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_exact_match": 8.8, | |
| "eval_f1": 14.986277056277055, | |
| "eval_qa_bleu": 6.307928648492083, | |
| "eval_qa_exact_match": 0.072, | |
| "eval_recite_bleu": 10.683207288869733, | |
| "eval_recite_exact_match": 0.014, | |
| "step": 7125 | |
| }, | |
| { | |
| "epoch": 38.4, | |
| "grad_norm": 2.9627795219421387, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2391, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 38.93333333333333, | |
| "grad_norm": 3.517446517944336, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2411, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 38.99733333333333, | |
| "eval_accuracy": 0.5740358744394619, | |
| "eval_loss": 3.213481903076172, | |
| "eval_runtime": 5.6174, | |
| "eval_samples_per_second": 89.009, | |
| "eval_steps_per_second": 11.215, | |
| "step": 7312 | |
| }, | |
| { | |
| "epoch": 38.99733333333333, | |
| "eval_exact_match": 9.4, | |
| "eval_f1": 15.155079365079372, | |
| "eval_qa_bleu": 5.289271493416481, | |
| "eval_qa_exact_match": 0.078, | |
| "eval_recite_bleu": 11.161407462076504, | |
| "eval_recite_exact_match": 0.018, | |
| "step": 7312 | |
| }, | |
| { | |
| "epoch": 39.46666666666667, | |
| "grad_norm": 2.9763762950897217, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2232, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 2.818361520767212, | |
| "learning_rate": 3e-05, | |
| "loss": 0.228, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_accuracy": 0.5739192825112107, | |
| "eval_loss": 3.2505199909210205, | |
| "eval_runtime": 6.0109, | |
| "eval_samples_per_second": 83.182, | |
| "eval_steps_per_second": 10.481, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_exact_match": 8.6, | |
| "eval_f1": 14.30142857142857, | |
| "eval_qa_bleu": 6.210824473363826, | |
| "eval_qa_exact_match": 0.074, | |
| "eval_recite_bleu": 11.205695530168638, | |
| "eval_recite_exact_match": 0.014, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 40.53333333333333, | |
| "grad_norm": 3.813812494277954, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2067, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 40.99733333333333, | |
| "eval_accuracy": 0.5743408071748879, | |
| "eval_loss": 3.286703586578369, | |
| "eval_runtime": 5.9771, | |
| "eval_samples_per_second": 83.652, | |
| "eval_steps_per_second": 10.54, | |
| "step": 7687 | |
| }, | |
| { | |
| "epoch": 40.99733333333333, | |
| "eval_exact_match": 7.8, | |
| "eval_f1": 13.974487734487733, | |
| "eval_qa_bleu": 9.0024711252003, | |
| "eval_qa_exact_match": 0.058, | |
| "eval_recite_bleu": 11.159844095103711, | |
| "eval_recite_exact_match": 0.016, | |
| "step": 7687 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 9350, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "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": 8.48357874231935e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |