| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 87, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03508771929824561, | |
| "grad_norm": 2.067223362182073, | |
| "learning_rate": 0.0, | |
| "loss": 0.866, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07017543859649122, | |
| "grad_norm": 2.2905484024803084, | |
| "learning_rate": 7.142857142857143e-07, | |
| "loss": 1.043, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 2.0823489332612506, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 0.9479, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.14035087719298245, | |
| "grad_norm": 2.155054227940796, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 1.0054, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.17543859649122806, | |
| "grad_norm": 2.122195910115268, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 1.0227, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 2.0002296419363823, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 0.9994, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.24561403508771928, | |
| "grad_norm": 1.7794168286652636, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.9833, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.2807017543859649, | |
| "grad_norm": 1.608102047372758, | |
| "learning_rate": 5e-06, | |
| "loss": 1.0059, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 1.5096684699130434, | |
| "learning_rate": 4.999302593030069e-06, | |
| "loss": 1.0199, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3508771929824561, | |
| "grad_norm": 1.2553700926461733, | |
| "learning_rate": 4.99721076122146e-06, | |
| "loss": 0.9945, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.38596491228070173, | |
| "grad_norm": 1.0588682250755448, | |
| "learning_rate": 4.99372567166064e-06, | |
| "loss": 0.8896, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 1.2403025527385267, | |
| "learning_rate": 4.98884926876821e-06, | |
| "loss": 0.9064, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.45614035087719296, | |
| "grad_norm": 1.3277416791779741, | |
| "learning_rate": 4.982584273214061e-06, | |
| "loss": 0.9728, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.49122807017543857, | |
| "grad_norm": 1.200452342066303, | |
| "learning_rate": 4.974934180399447e-06, | |
| "loss": 0.8801, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 1.2276015693500428, | |
| "learning_rate": 4.965903258506806e-06, | |
| "loss": 0.9056, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.5614035087719298, | |
| "grad_norm": 0.9588298096291193, | |
| "learning_rate": 4.955496546118439e-06, | |
| "loss": 0.851, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5964912280701754, | |
| "grad_norm": 0.8961934214540342, | |
| "learning_rate": 4.943719849405347e-06, | |
| "loss": 0.9588, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 0.7752982692883841, | |
| "learning_rate": 4.930579738887827e-06, | |
| "loss": 0.8592, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.7598215199572412, | |
| "learning_rate": 4.9160835457696075e-06, | |
| "loss": 0.9224, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7017543859649122, | |
| "grad_norm": 0.7947043775095588, | |
| "learning_rate": 4.900239357847582e-06, | |
| "loss": 0.9183, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 0.8131278296751281, | |
| "learning_rate": 4.883056014999423e-06, | |
| "loss": 0.8854, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.7719298245614035, | |
| "grad_norm": 0.720597258722373, | |
| "learning_rate": 4.864543104251587e-06, | |
| "loss": 0.837, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.8070175438596491, | |
| "grad_norm": 0.698175660395316, | |
| "learning_rate": 4.844710954430464e-06, | |
| "loss": 0.9006, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.6061246443427316, | |
| "learning_rate": 4.823570630399665e-06, | |
| "loss": 0.8713, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.8771929824561403, | |
| "grad_norm": 0.5800728862173004, | |
| "learning_rate": 4.8011339268866505e-06, | |
| "loss": 0.8572, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.9122807017543859, | |
| "grad_norm": 0.6881850987832964, | |
| "learning_rate": 4.777413361902152e-06, | |
| "loss": 0.9243, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 0.6013453853835422, | |
| "learning_rate": 4.752422169756048e-06, | |
| "loss": 0.8215, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.9824561403508771, | |
| "grad_norm": 0.5049917945559669, | |
| "learning_rate": 4.726174293673612e-06, | |
| "loss": 0.8498, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.5049917945559669, | |
| "learning_rate": 4.698684378016223e-06, | |
| "loss": 0.7765, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.0350877192982457, | |
| "grad_norm": 0.8391221152195739, | |
| "learning_rate": 4.669967760110908e-06, | |
| "loss": 0.8698, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.0701754385964912, | |
| "grad_norm": 0.5071651643461081, | |
| "learning_rate": 4.6400404616932505e-06, | |
| "loss": 0.8376, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 0.4758228095407797, | |
| "learning_rate": 4.608919179968457e-06, | |
| "loss": 0.883, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.1403508771929824, | |
| "grad_norm": 0.4648364085184619, | |
| "learning_rate": 4.576621278295558e-06, | |
| "loss": 0.8628, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.1754385964912282, | |
| "grad_norm": 0.4712254209084955, | |
| "learning_rate": 4.543164776499945e-06, | |
| "loss": 0.8281, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 0.4013954971972578, | |
| "learning_rate": 4.508568340819654e-06, | |
| "loss": 0.8867, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.2456140350877192, | |
| "grad_norm": 0.41192488021825646, | |
| "learning_rate": 4.472851273490985e-06, | |
| "loss": 0.8675, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.280701754385965, | |
| "grad_norm": 0.4239007109950775, | |
| "learning_rate": 4.436033501979299e-06, | |
| "loss": 0.8591, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 0.41628314789696425, | |
| "learning_rate": 4.398135567860972e-06, | |
| "loss": 0.8254, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.3508771929824561, | |
| "grad_norm": 0.3923309284581745, | |
| "learning_rate": 4.359178615362725e-06, | |
| "loss": 0.7828, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.3859649122807016, | |
| "grad_norm": 0.47649896532625685, | |
| "learning_rate": 4.319184379564716e-06, | |
| "loss": 0.8922, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 0.47963769728260897, | |
| "learning_rate": 4.278175174273989e-06, | |
| "loss": 0.8179, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.456140350877193, | |
| "grad_norm": 0.44693221984156073, | |
| "learning_rate": 4.236173879575022e-06, | |
| "loss": 0.7873, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.4912280701754386, | |
| "grad_norm": 0.3915854729078168, | |
| "learning_rate": 4.1932039290643534e-06, | |
| "loss": 0.7786, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 0.3379023862301008, | |
| "learning_rate": 4.149289296776369e-06, | |
| "loss": 0.9101, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.5614035087719298, | |
| "grad_norm": 0.3691862472080892, | |
| "learning_rate": 4.104454483807579e-06, | |
| "loss": 0.8606, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.5964912280701755, | |
| "grad_norm": 0.34291927069087574, | |
| "learning_rate": 4.058724504646834e-06, | |
| "loss": 0.7938, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 0.4003834992970626, | |
| "learning_rate": 4.012124873219094e-06, | |
| "loss": 0.7814, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.3392403324591956, | |
| "learning_rate": 3.964681588650562e-06, | |
| "loss": 0.8389, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.7017543859649122, | |
| "grad_norm": 0.3377270590051618, | |
| "learning_rate": 3.916421120763106e-06, | |
| "loss": 0.85, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 0.38740961950138847, | |
| "learning_rate": 3.8673703953060685e-06, | |
| "loss": 0.8348, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.7719298245614035, | |
| "grad_norm": 0.35219126914631843, | |
| "learning_rate": 3.817556778933697e-06, | |
| "loss": 0.8577, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.807017543859649, | |
| "grad_norm": 0.3774463349958884, | |
| "learning_rate": 3.7670080639366e-06, | |
| "loss": 0.8769, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 0.35130371579419645, | |
| "learning_rate": 3.7157524527357036e-06, | |
| "loss": 0.7163, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.8771929824561404, | |
| "grad_norm": 0.3551942508437019, | |
| "learning_rate": 3.663818542147409e-06, | |
| "loss": 0.8756, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.912280701754386, | |
| "grad_norm": 0.3537337899404262, | |
| "learning_rate": 3.61123530742869e-06, | |
| "loss": 0.8219, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 0.327364809217546, | |
| "learning_rate": 3.5580320861110627e-06, | |
| "loss": 0.8696, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.9824561403508771, | |
| "grad_norm": 0.3367652725930036, | |
| "learning_rate": 3.5042385616324243e-06, | |
| "loss": 0.872, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.5052810208370435, | |
| "learning_rate": 3.4498847467759e-06, | |
| "loss": 0.8607, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.0350877192982457, | |
| "grad_norm": 0.3177161861315085, | |
| "learning_rate": 3.3950009669249502e-06, | |
| "loss": 0.7808, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.0701754385964914, | |
| "grad_norm": 0.3450713484869485, | |
| "learning_rate": 3.3396178431440572e-06, | |
| "loss": 0.9217, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.32271078088898175, | |
| "learning_rate": 3.283766275094454e-06, | |
| "loss": 0.7982, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.1403508771929824, | |
| "grad_norm": 0.33864939828309026, | |
| "learning_rate": 3.227477423794412e-06, | |
| "loss": 0.7907, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.175438596491228, | |
| "grad_norm": 0.32084183022976825, | |
| "learning_rate": 3.1707826942337124e-06, | |
| "loss": 0.8362, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "grad_norm": 0.31523598018882065, | |
| "learning_rate": 3.1137137178519983e-06, | |
| "loss": 0.8603, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.245614035087719, | |
| "grad_norm": 0.3534452142386561, | |
| "learning_rate": 3.056302334890786e-06, | |
| "loss": 0.7473, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.280701754385965, | |
| "grad_norm": 0.3065081387248856, | |
| "learning_rate": 2.9985805766289815e-06, | |
| "loss": 0.8076, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 0.31593346837979236, | |
| "learning_rate": 2.940580647511805e-06, | |
| "loss": 0.7911, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.3508771929824563, | |
| "grad_norm": 0.33421077857364934, | |
| "learning_rate": 2.8823349071831154e-06, | |
| "loss": 0.8567, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.3859649122807016, | |
| "grad_norm": 0.36084453724888643, | |
| "learning_rate": 2.8238758524311316e-06, | |
| "loss": 0.7815, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 0.31104114669216265, | |
| "learning_rate": 2.7652360990576457e-06, | |
| "loss": 0.9035, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.456140350877193, | |
| "grad_norm": 0.3283662140442227, | |
| "learning_rate": 2.7064483636808314e-06, | |
| "loss": 0.815, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.4912280701754383, | |
| "grad_norm": 0.28913740772142477, | |
| "learning_rate": 2.6475454454818072e-06, | |
| "loss": 0.8369, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 0.30507064501325465, | |
| "learning_rate": 2.5885602079051354e-06, | |
| "loss": 0.7603, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.56140350877193, | |
| "grad_norm": 0.3531329822293121, | |
| "learning_rate": 2.529525560323462e-06, | |
| "loss": 0.8004, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.5964912280701755, | |
| "grad_norm": 0.3210792288970346, | |
| "learning_rate": 2.470474439676539e-06, | |
| "loss": 0.8168, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 0.36897603215114533, | |
| "learning_rate": 2.411439792094866e-06, | |
| "loss": 0.8398, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.3217143259950112, | |
| "learning_rate": 2.3524545545181936e-06, | |
| "loss": 0.866, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.7017543859649122, | |
| "grad_norm": 0.3126909067501635, | |
| "learning_rate": 2.2935516363191695e-06, | |
| "loss": 0.8341, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 0.3006660553493533, | |
| "learning_rate": 2.234763900942355e-06, | |
| "loss": 0.8081, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.7719298245614032, | |
| "grad_norm": 0.316850226629777, | |
| "learning_rate": 2.1761241475688697e-06, | |
| "loss": 0.7562, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.807017543859649, | |
| "grad_norm": 0.2903780216034827, | |
| "learning_rate": 2.117665092816885e-06, | |
| "loss": 0.7419, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 0.3424299492049514, | |
| "learning_rate": 2.059419352488196e-06, | |
| "loss": 0.8289, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.8771929824561404, | |
| "grad_norm": 0.3113278689750571, | |
| "learning_rate": 2.0014194233710193e-06, | |
| "loss": 0.823, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.912280701754386, | |
| "grad_norm": 0.3375235251647405, | |
| "learning_rate": 1.9436976651092143e-06, | |
| "loss": 0.8292, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.3084098102163688, | |
| "learning_rate": 1.8862862821480023e-06, | |
| "loss": 0.7702, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.982456140350877, | |
| "grad_norm": 0.30920611879243615, | |
| "learning_rate": 1.829217305766289e-06, | |
| "loss": 0.8181, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.30920611879243615, | |
| "learning_rate": 1.772522576205589e-06, | |
| "loss": 0.6831, | |
| "step": 87 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 140, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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": 39139677044736.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |