| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 27741, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 24.947961807250977, | |
| "learning_rate": 4.909880682022999e-05, | |
| "loss": 1.2957, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 18.850772857666016, | |
| "learning_rate": 4.819761364045997e-05, | |
| "loss": 0.9299, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 23.09844398498535, | |
| "learning_rate": 4.7296420460689956e-05, | |
| "loss": 0.8582, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 15.375988006591797, | |
| "learning_rate": 4.639522728091994e-05, | |
| "loss": 0.7935, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 15.962867736816406, | |
| "learning_rate": 4.5494034101149925e-05, | |
| "loss": 0.7439, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 16.872922897338867, | |
| "learning_rate": 4.459284092137991e-05, | |
| "loss": 0.6964, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 21.044837951660156, | |
| "learning_rate": 4.3691647741609894e-05, | |
| "loss": 0.6701, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 15.188350677490234, | |
| "learning_rate": 4.279045456183988e-05, | |
| "loss": 0.6339, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 12.721724510192871, | |
| "learning_rate": 4.188926138206986e-05, | |
| "loss": 0.6221, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 13.383834838867188, | |
| "learning_rate": 4.098806820229985e-05, | |
| "loss": 0.5805, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 15.311894416809082, | |
| "learning_rate": 4.008687502252983e-05, | |
| "loss": 0.561, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 16.439502716064453, | |
| "learning_rate": 3.9185681842759816e-05, | |
| "loss": 0.5484, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 16.220827102661133, | |
| "learning_rate": 3.82844886629898e-05, | |
| "loss": 0.5265, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 14.172871589660645, | |
| "learning_rate": 3.7383295483219785e-05, | |
| "loss": 0.5022, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 13.379059791564941, | |
| "learning_rate": 3.648210230344977e-05, | |
| "loss": 0.4878, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 21.462778091430664, | |
| "learning_rate": 3.558090912367975e-05, | |
| "loss": 0.4657, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 9.849862098693848, | |
| "learning_rate": 3.467971594390974e-05, | |
| "loss": 0.4566, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 11.99971866607666, | |
| "learning_rate": 3.3778522764139723e-05, | |
| "loss": 0.4384, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 12.555150985717773, | |
| "learning_rate": 3.287732958436971e-05, | |
| "loss": 0.3905, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 14.604622840881348, | |
| "learning_rate": 3.197613640459969e-05, | |
| "loss": 0.3573, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 9.172225952148438, | |
| "learning_rate": 3.107494322482968e-05, | |
| "loss": 0.3486, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 10.893487930297852, | |
| "learning_rate": 3.017375004505966e-05, | |
| "loss": 0.3413, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 9.724763870239258, | |
| "learning_rate": 2.9272556865289646e-05, | |
| "loss": 0.3405, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 9.83249568939209, | |
| "learning_rate": 2.837136368551963e-05, | |
| "loss": 0.3397, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 10.763433456420898, | |
| "learning_rate": 2.7470170505749615e-05, | |
| "loss": 0.3312, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 10.305283546447754, | |
| "learning_rate": 2.65689773259796e-05, | |
| "loss": 0.3228, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 8.631694793701172, | |
| "learning_rate": 2.5667784146209584e-05, | |
| "loss": 0.3162, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 7.709570407867432, | |
| "learning_rate": 2.476659096643957e-05, | |
| "loss": 0.3033, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 8.473445892333984, | |
| "learning_rate": 2.3865397786669553e-05, | |
| "loss": 0.2987, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 5.8854217529296875, | |
| "learning_rate": 2.2964204606899537e-05, | |
| "loss": 0.292, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 5.954896926879883, | |
| "learning_rate": 2.2063011427129522e-05, | |
| "loss": 0.2948, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 6.9837164878845215, | |
| "learning_rate": 2.1161818247359506e-05, | |
| "loss": 0.2848, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 9.53173542022705, | |
| "learning_rate": 2.026062506758949e-05, | |
| "loss": 0.279, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 8.36210823059082, | |
| "learning_rate": 1.9359431887819472e-05, | |
| "loss": 0.2692, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 8.573978424072266, | |
| "learning_rate": 1.8458238708049457e-05, | |
| "loss": 0.265, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 7.375257968902588, | |
| "learning_rate": 1.755704552827944e-05, | |
| "loss": 0.2683, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 6.663829803466797, | |
| "learning_rate": 1.6655852348509426e-05, | |
| "loss": 0.2539, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 5.3059492111206055, | |
| "learning_rate": 1.575465916873941e-05, | |
| "loss": 0.1978, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 6.28593635559082, | |
| "learning_rate": 1.4853465988969395e-05, | |
| "loss": 0.2004, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 7.351054668426514, | |
| "learning_rate": 1.3952272809199379e-05, | |
| "loss": 0.1926, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 6.031446933746338, | |
| "learning_rate": 1.3051079629429367e-05, | |
| "loss": 0.1903, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 4.734748840332031, | |
| "learning_rate": 1.214988644965935e-05, | |
| "loss": 0.1852, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 6.121973514556885, | |
| "learning_rate": 1.1248693269889334e-05, | |
| "loss": 0.1764, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 4.245573043823242, | |
| "learning_rate": 1.0347500090119317e-05, | |
| "loss": 0.1808, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 3.6098544597625732, | |
| "learning_rate": 9.446306910349303e-06, | |
| "loss": 0.1781, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 7.525985240936279, | |
| "learning_rate": 8.545113730579288e-06, | |
| "loss": 0.1837, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 4.0185546875, | |
| "learning_rate": 7.643920550809272e-06, | |
| "loss": 0.1734, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 5.020835876464844, | |
| "learning_rate": 6.742727371039257e-06, | |
| "loss": 0.171, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 5.588749408721924, | |
| "learning_rate": 5.84153419126924e-06, | |
| "loss": 0.1699, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 3.7061429023742676, | |
| "learning_rate": 4.940341011499225e-06, | |
| "loss": 0.1621, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 4.335537433624268, | |
| "learning_rate": 4.03914783172921e-06, | |
| "loss": 0.1626, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 5.078334331512451, | |
| "learning_rate": 3.1379546519591943e-06, | |
| "loss": 0.156, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 5.9595947265625, | |
| "learning_rate": 2.2367614721891784e-06, | |
| "loss": 0.1617, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 5.210651397705078, | |
| "learning_rate": 1.335568292419163e-06, | |
| "loss": 0.1533, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 3.798105478286743, | |
| "learning_rate": 4.3437511264914753e-07, | |
| "loss": 0.1521, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 27741, | |
| "total_flos": 2.3702775297552e+17, | |
| "train_loss": 0.3772407876341264, | |
| "train_runtime": 10846.9143, | |
| "train_samples_per_second": 163.665, | |
| "train_steps_per_second": 2.558 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 27741, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 2.3702775297552e+17, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |