| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 58, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.017241379310344827, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 0.0, | |
| "loss": 2.5553, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.034482758620689655, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.25e-06, | |
| "loss": 2.5516, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05172413793103448, | |
| "grad_norm": 1.25, | |
| "learning_rate": 2.5e-06, | |
| "loss": 2.5156, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06896551724137931, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 2.5199, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08620689655172414, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 5e-06, | |
| "loss": 2.5059, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10344827586206896, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 6.25e-06, | |
| "loss": 2.4909, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.1206896551724138, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 2.5399, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.13793103448275862, | |
| "grad_norm": 0.91796875, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 2.5355, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.15517241379310345, | |
| "grad_norm": 0.828125, | |
| "learning_rate": 1e-05, | |
| "loss": 2.5359, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1724137931034483, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 9.99910461334869e-06, | |
| "loss": 2.5257, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1896551724137931, | |
| "grad_norm": 0.6875, | |
| "learning_rate": 9.996418774081658e-06, | |
| "loss": 2.5138, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.20689655172413793, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 9.991943444144758e-06, | |
| "loss": 2.5139, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.22413793103448276, | |
| "grad_norm": 0.87109375, | |
| "learning_rate": 9.985680226398261e-06, | |
| "loss": 2.4983, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2413793103448276, | |
| "grad_norm": 0.69921875, | |
| "learning_rate": 9.977631364042796e-06, | |
| "loss": 2.5036, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.25862068965517243, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 9.967799739815925e-06, | |
| "loss": 2.5121, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 9.956188874959686e-06, | |
| "loss": 2.539, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.29310344827586204, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 9.942802927959444e-06, | |
| "loss": 2.5107, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3103448275862069, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 9.927646693054498e-06, | |
| "loss": 2.4907, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3275862068965517, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 9.910725598521014e-06, | |
| "loss": 2.4836, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 9.892045704727864e-06, | |
| "loss": 2.494, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3620689655172414, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 9.871613701966067e-06, | |
| "loss": 2.4733, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3793103448275862, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 9.849436908052636e-06, | |
| "loss": 2.4141, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.39655172413793105, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 9.825523265709667e-06, | |
| "loss": 2.521, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.41379310344827586, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 9.799881339719615e-06, | |
| "loss": 2.4647, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.43103448275862066, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 9.772520313857777e-06, | |
| "loss": 2.4892, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4482758620689655, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 9.743449987603082e-06, | |
| "loss": 2.4734, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.46551724137931033, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 9.712680772628365e-06, | |
| "loss": 2.5229, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.4827586206896552, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 9.680223689071364e-06, | |
| "loss": 2.5502, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 9.646090361587828e-06, | |
| "loss": 2.5008, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5172413793103449, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 9.610293015188067e-06, | |
| "loss": 2.5319, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5344827586206896, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 9.572844470858537e-06, | |
| "loss": 2.5004, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5517241379310345, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 9.533758140969913e-06, | |
| "loss": 2.4611, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5689655172413793, | |
| "grad_norm": 0.625, | |
| "learning_rate": 9.493048024473413e-06, | |
| "loss": 2.4883, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5862068965517241, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 9.450728701886985e-06, | |
| "loss": 2.5014, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.603448275862069, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 9.406815330073244e-06, | |
| "loss": 2.4959, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6206896551724138, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 9.36132363681097e-06, | |
| "loss": 2.4901, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6379310344827587, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 9.314269915162115e-06, | |
| "loss": 2.54, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6551724137931034, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 9.265671017636384e-06, | |
| "loss": 2.5077, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6724137931034483, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 9.215544350155423e-06, | |
| "loss": 2.5093, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 9.163907865818806e-06, | |
| "loss": 2.5134, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7068965517241379, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 9.110780058474052e-06, | |
| "loss": 2.5651, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7241379310344828, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 9.056179956092961e-06, | |
| "loss": 2.4615, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7413793103448276, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 9.000127113956673e-06, | |
| "loss": 2.4763, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7586206896551724, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 8.94264160765183e-06, | |
| "loss": 2.4732, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7758620689655172, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 8.883744025880429e-06, | |
| "loss": 2.49, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7931034482758621, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 8.823455463085873e-06, | |
| "loss": 2.4944, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8103448275862069, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 8.761797511897907e-06, | |
| "loss": 2.5177, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8275862068965517, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 8.698792255399104e-06, | |
| "loss": 2.5083, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8448275862068966, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 8.634462259215719e-06, | |
| "loss": 2.5096, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8620689655172413, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 8.568830563435695e-06, | |
| "loss": 2.5267, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8793103448275862, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 8.501920674356755e-06, | |
| "loss": 2.5041, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.896551724137931, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 8.433756556067506e-06, | |
| "loss": 2.5081, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.9137931034482759, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 8.364362621864595e-06, | |
| "loss": 2.4329, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.9310344827586207, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 8.29376372550897e-06, | |
| "loss": 2.4784, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.9482758620689655, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 8.221985152324385e-06, | |
| "loss": 2.5528, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9655172413793104, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 8.149052610141357e-06, | |
| "loss": 2.4803, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.9827586206896551, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 8.07499222008977e-06, | |
| "loss": 2.5519, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 7.625, | |
| "learning_rate": 7.999830507243478e-06, | |
| "loss": 3.0347, | |
| "step": 58 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 174, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 58, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.587894225456333e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |