| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 6, | |
| "global_step": 57, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 211.82034301757812, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5234, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 330236.46875, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.4781, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 3773.773681640625, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.5327, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 8766.265625, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.4726, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 633.0609130859375, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.4571, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 609995.625, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.5539, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "eval_loss": 0.3873564898967743, | |
| "eval_runtime": 0.8662, | |
| "eval_samples_per_second": 2.309, | |
| "eval_steps_per_second": 1.154, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 578.82763671875, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.3949, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 227617.40625, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.5256, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 785.8093872070312, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.4265, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 354838.09375, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.6413, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.5789473684210527, | |
| "grad_norm": 3835.27197265625, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.5344, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 4633.39013671875, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.5069, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "eval_loss": 0.42546606063842773, | |
| "eval_runtime": 0.8681, | |
| "eval_samples_per_second": 2.304, | |
| "eval_steps_per_second": 1.152, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.6842105263157895, | |
| "grad_norm": 2101.546142578125, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.5058, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 175295.453125, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.4587, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 3184.1103515625, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.614, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 496073.1875, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.644, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.8947368421052632, | |
| "grad_norm": 11245.0107421875, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5248, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 600.7078857421875, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.5634, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "eval_loss": 0.47000423073768616, | |
| "eval_runtime": 0.8698, | |
| "eval_samples_per_second": 2.299, | |
| "eval_steps_per_second": 1.15, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1067464.375, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.655, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 7849.232421875, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.5225, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 5074.62060546875, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.5764, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 1.1578947368421053, | |
| "grad_norm": 755949.375, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.5878, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 1128.499267578125, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.6382, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 955.5560913085938, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.7207, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "eval_loss": 0.5061320066452026, | |
| "eval_runtime": 0.8698, | |
| "eval_samples_per_second": 2.299, | |
| "eval_steps_per_second": 1.15, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 10727177.0, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.5727, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.368421052631579, | |
| "grad_norm": 835351.6875, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.6236, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 24849.080078125, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.5477, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 1144.915283203125, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.7404, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 43178.171875, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5133, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 105640.3828125, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.6525, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "eval_loss": 0.5270132422447205, | |
| "eval_runtime": 0.8703, | |
| "eval_samples_per_second": 2.298, | |
| "eval_steps_per_second": 1.149, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 2033.227783203125, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.538, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 679948.1875, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.6714, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 220607.5, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.7117, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.7894736842105263, | |
| "grad_norm": 13599.486328125, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.5645, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 651100.1875, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.716, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 575592.1875, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.7435, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "eval_loss": 0.5324785113334656, | |
| "eval_runtime": 0.8699, | |
| "eval_samples_per_second": 2.299, | |
| "eval_steps_per_second": 1.149, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 163141.171875, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.7309, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3034.2685546875, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.5396, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 2.0526315789473686, | |
| "grad_norm": 124152.5, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.6829, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 3723410.75, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.5327, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.1578947368421053, | |
| "grad_norm": 539427.625, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.532, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "grad_norm": 20275.404296875, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.7488, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "eval_loss": 0.5383380055427551, | |
| "eval_runtime": 0.8703, | |
| "eval_samples_per_second": 2.298, | |
| "eval_steps_per_second": 1.149, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 2.263157894736842, | |
| "grad_norm": 505885.78125, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.5959, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 74390.25, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.5717, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 2.3684210526315788, | |
| "grad_norm": 1850279.375, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.682, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 7982978.5, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.6275, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 2.473684210526316, | |
| "grad_norm": 132265160.0, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.6225, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 96502.90625, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.5786, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "eval_loss": 0.5181117057800293, | |
| "eval_runtime": 0.8701, | |
| "eval_samples_per_second": 2.299, | |
| "eval_steps_per_second": 1.149, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 2.5789473684210527, | |
| "grad_norm": 697082.1875, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.6937, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 43987.10546875, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.675, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.6842105263157894, | |
| "grad_norm": 47614.27734375, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.676, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 79889032.0, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.6332, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.7894736842105265, | |
| "grad_norm": 1658783.125, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.6212, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 1998786.75, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.7325, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "eval_loss": 0.5145716667175293, | |
| "eval_runtime": 0.871, | |
| "eval_samples_per_second": 2.296, | |
| "eval_steps_per_second": 1.148, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.8947368421052633, | |
| "grad_norm": 58182.7734375, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.5281, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 854980864.0, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.7652, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 4794.498046875, | |
| "learning_rate": 0.0, | |
| "loss": 0.66, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 57, | |
| "total_flos": 8.713084645618483e+16, | |
| "train_loss": 0.597966728503244, | |
| "train_runtime": 375.3657, | |
| "train_samples_per_second": 1.183, | |
| "train_steps_per_second": 0.152 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 57, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 57, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.713084645618483e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |