| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0588235294117647, | |
| "eval_steps": 500, | |
| "global_step": 36, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.029411764705882353, | |
| "grad_norm": 0.9765625, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0227, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.058823529411764705, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 4e-05, | |
| "loss": 1.6615, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.08823529411764706, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 6e-05, | |
| "loss": 1.9129, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.11764705882352941, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 8e-05, | |
| "loss": 2.2892, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.14705882352941177, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 0.0001, | |
| "loss": 2.0247, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.17647058823529413, | |
| "grad_norm": 0.98828125, | |
| "learning_rate": 0.00012, | |
| "loss": 1.8962, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.20588235294117646, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 0.00014, | |
| "loss": 1.703, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 0.00016, | |
| "loss": 1.5534, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.2647058823529412, | |
| "grad_norm": 0.93359375, | |
| "learning_rate": 0.00018, | |
| "loss": 1.4303, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.29411764705882354, | |
| "grad_norm": 0.83203125, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3701, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3235294117647059, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 0.00019985334138511237, | |
| "loss": 1.3449, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.35294117647058826, | |
| "grad_norm": 0.9765625, | |
| "learning_rate": 0.00019941379571543596, | |
| "loss": 1.3191, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.38235294117647056, | |
| "grad_norm": 0.94140625, | |
| "learning_rate": 0.00019868265225415265, | |
| "loss": 1.0243, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.4117647058823529, | |
| "grad_norm": 0.91015625, | |
| "learning_rate": 0.00019766205557100868, | |
| "loss": 1.0321, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.4411764705882353, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 0.0001963549992519223, | |
| "loss": 1.2083, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 0.00019476531711828027, | |
| "loss": 1.2976, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.125, | |
| "learning_rate": 0.00019289767198167916, | |
| "loss": 1.0929, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.5294117647058824, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 0.00019075754196709572, | |
| "loss": 0.9081, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5588235294117647, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 0.0001883512044446023, | |
| "loss": 0.9417, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 0.90234375, | |
| "learning_rate": 0.00018568571761675893, | |
| "loss": 0.9065, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.6176470588235294, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 0.00018276889981568906, | |
| "loss": 0.9419, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.6470588235294118, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 0.00017960930657056438, | |
| "loss": 0.9199, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.6764705882352942, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 0.00017621620551276366, | |
| "loss": 0.9139, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 0.0001725995491923131, | |
| "loss": 0.7512, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.7352941176470589, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 0.00016876994588534234, | |
| "loss": 0.7435, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.7647058823529411, | |
| "grad_norm": 0.95703125, | |
| "learning_rate": 0.00016473862847818277, | |
| "loss": 0.7502, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.7941176470588235, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 0.00016051742151937655, | |
| "loss": 0.956, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.8235294117647058, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 0.00015611870653623825, | |
| "loss": 0.8827, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.8529411764705882, | |
| "grad_norm": 0.94140625, | |
| "learning_rate": 0.00015155538571770218, | |
| "loss": 0.7322, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.8823529411764706, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 0.00014684084406997903, | |
| "loss": 0.7407, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9117647058823529, | |
| "grad_norm": 0.89453125, | |
| "learning_rate": 0.00014198891015602646, | |
| "loss": 0.8105, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 0.00013701381553399145, | |
| "loss": 0.9126, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.9705882352941176, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 0.000131930153013598, | |
| "loss": 0.9439, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 0.00012675283385292212, | |
| "loss": 0.7298, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.0294117647058822, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 0.00012149704402110243, | |
| "loss": 0.4975, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.0588235294117647, | |
| "grad_norm": 0.84765625, | |
| "learning_rate": 0.0001161781996552765, | |
| "loss": 0.8089, | |
| "step": 36 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 68, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 12, | |
| "total_flos": 1.0646082519092429e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |