| { | |
| "best_metric": 1.1520808935165405, | |
| "best_model_checkpoint": "data/Mistral-7B_task-1_180-samples_config-2/checkpoint-34", | |
| "epoch": 10.941176470588236, | |
| "eval_steps": 500, | |
| "global_step": 93, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11764705882352941, | |
| "grad_norm": 4.991026401519775, | |
| "learning_rate": 2.5e-06, | |
| "loss": 2.1553, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 5.117615699768066, | |
| "learning_rate": 5e-06, | |
| "loss": 2.1988, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 5.423519134521484, | |
| "learning_rate": 1e-05, | |
| "loss": 2.2826, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 4.905459403991699, | |
| "learning_rate": 1.5e-05, | |
| "loss": 2.1742, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 4.527350425720215, | |
| "learning_rate": 2e-05, | |
| "loss": 2.053, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "eval_loss": 1.8889563083648682, | |
| "eval_runtime": 14.5533, | |
| "eval_samples_per_second": 2.474, | |
| "eval_steps_per_second": 2.474, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 3.475832223892212, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.8598, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 2.5801305770874023, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6493, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.6470588235294117, | |
| "grad_norm": 1.8207895755767822, | |
| "learning_rate": 3.5e-05, | |
| "loss": 1.4821, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 1.4556984901428223, | |
| "learning_rate": 4e-05, | |
| "loss": 1.3317, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.3413469791412354, | |
| "eval_runtime": 14.5537, | |
| "eval_samples_per_second": 2.474, | |
| "eval_steps_per_second": 2.474, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 2.1176470588235294, | |
| "grad_norm": 1.04146409034729, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.2367, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 1.2536530494689941, | |
| "learning_rate": 5e-05, | |
| "loss": 1.2383, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.588235294117647, | |
| "grad_norm": 1.1814831495285034, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 1.2719, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.8235294117647056, | |
| "grad_norm": 1.0946425199508667, | |
| "learning_rate": 6e-05, | |
| "loss": 1.1994, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "eval_loss": 1.2018373012542725, | |
| "eval_runtime": 14.5629, | |
| "eval_samples_per_second": 2.472, | |
| "eval_steps_per_second": 2.472, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.0588235294117645, | |
| "grad_norm": 1.0643476247787476, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 1.1306, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 3.2941176470588234, | |
| "grad_norm": 1.1662391424179077, | |
| "learning_rate": 7e-05, | |
| "loss": 1.0472, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 1.2767999172210693, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.0429, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.764705882352941, | |
| "grad_norm": 1.1417080163955688, | |
| "learning_rate": 8e-05, | |
| "loss": 0.9668, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.3573873043060303, | |
| "learning_rate": 8.5e-05, | |
| "loss": 0.9607, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.1520808935165405, | |
| "eval_runtime": 14.5602, | |
| "eval_samples_per_second": 2.472, | |
| "eval_steps_per_second": 2.472, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.235294117647059, | |
| "grad_norm": 1.1198458671569824, | |
| "learning_rate": 9e-05, | |
| "loss": 0.8712, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 4.470588235294118, | |
| "grad_norm": 1.2332489490509033, | |
| "learning_rate": 9.5e-05, | |
| "loss": 0.8081, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 1.2247358560562134, | |
| "learning_rate": 0.0001, | |
| "loss": 0.7503, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "grad_norm": 1.4213950634002686, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 0.7457, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "eval_loss": 1.1806774139404297, | |
| "eval_runtime": 14.5514, | |
| "eval_samples_per_second": 2.474, | |
| "eval_steps_per_second": 2.474, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 5.176470588235294, | |
| "grad_norm": 1.2573533058166504, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 0.5997, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 5.411764705882353, | |
| "grad_norm": 1.7378907203674316, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 0.5895, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 5.647058823529412, | |
| "grad_norm": 1.972646951675415, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.4286, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 2.6307058334350586, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 0.5441, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 1.3877015113830566, | |
| "eval_runtime": 14.5569, | |
| "eval_samples_per_second": 2.473, | |
| "eval_steps_per_second": 2.473, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 6.117647058823529, | |
| "grad_norm": 1.6350064277648926, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 0.3851, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 6.352941176470588, | |
| "grad_norm": 2.2904036045074463, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 0.2898, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 6.588235294117647, | |
| "grad_norm": 2.2712900638580322, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 0.2675, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 6.823529411764706, | |
| "grad_norm": 2.793653726577759, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 0.262, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.9411764705882355, | |
| "eval_loss": 1.616546630859375, | |
| "eval_runtime": 14.5577, | |
| "eval_samples_per_second": 2.473, | |
| "eval_steps_per_second": 2.473, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 7.0588235294117645, | |
| "grad_norm": 1.9732766151428223, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 0.2309, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.294117647058823, | |
| "grad_norm": 1.9136290550231934, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 0.1441, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 7.529411764705882, | |
| "grad_norm": 2.7887516021728516, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.1155, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 7.764705882352941, | |
| "grad_norm": 3.061058282852173, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 0.1023, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 3.621833324432373, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 0.1407, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.9109816551208496, | |
| "eval_runtime": 14.5636, | |
| "eval_samples_per_second": 2.472, | |
| "eval_steps_per_second": 2.472, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.235294117647058, | |
| "grad_norm": 2.3172683715820312, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 0.0603, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.470588235294118, | |
| "grad_norm": 2.268026351928711, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.0708, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 8.705882352941176, | |
| "grad_norm": 2.028542995452881, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 0.0715, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "grad_norm": 1.5232679843902588, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.0589, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "eval_loss": 1.9830735921859741, | |
| "eval_runtime": 14.5581, | |
| "eval_samples_per_second": 2.473, | |
| "eval_steps_per_second": 2.473, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 9.176470588235293, | |
| "grad_norm": 1.2906736135482788, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 0.0401, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 9.411764705882353, | |
| "grad_norm": 2.3489370346069336, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.0418, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 9.647058823529411, | |
| "grad_norm": 2.143594741821289, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 0.0481, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 9.882352941176471, | |
| "grad_norm": 1.7526501417160034, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 0.053, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 2.023517370223999, | |
| "eval_runtime": 14.5596, | |
| "eval_samples_per_second": 2.473, | |
| "eval_steps_per_second": 2.473, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 10.117647058823529, | |
| "grad_norm": 1.686197280883789, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 0.0401, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 10.352941176470589, | |
| "grad_norm": 1.4375970363616943, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.026, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 10.588235294117647, | |
| "grad_norm": 1.4091485738754272, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 0.0297, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.823529411764707, | |
| "grad_norm": 1.1159406900405884, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 0.0268, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 10.941176470588236, | |
| "eval_loss": 2.1510894298553467, | |
| "eval_runtime": 14.5798, | |
| "eval_samples_per_second": 2.469, | |
| "eval_steps_per_second": 2.469, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 10.941176470588236, | |
| "step": 93, | |
| "total_flos": 2.3461897328132096e+16, | |
| "train_loss": 0.709120698954149, | |
| "train_runtime": 1774.2454, | |
| "train_samples_per_second": 3.833, | |
| "train_steps_per_second": 0.225 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.3461897328132096e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |