| { | |
| "best_global_step": 27, | |
| "best_metric": 0.5779597759246826, | |
| "best_model_checkpoint": "//outputs/task7_microsoft/Phi-3.5-mini-instruct/checkpoint-27", | |
| "epoch": 18.0, | |
| "eval_steps": 500, | |
| "global_step": 162, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 0.2979797124862671, | |
| "learning_rate": 0.00015, | |
| "loss": 0.7889, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.6737979650497437, | |
| "eval_runtime": 3.3675, | |
| "eval_samples_per_second": 4.454, | |
| "eval_steps_per_second": 0.594, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.1176470588235294, | |
| "grad_norm": 0.20464661717414856, | |
| "learning_rate": 0.0003, | |
| "loss": 0.6771, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.7058823529411766, | |
| "grad_norm": 0.1992536187171936, | |
| "learning_rate": 0.00029948767395100045, | |
| "loss": 0.6255, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.6006895303726196, | |
| "eval_runtime": 3.3642, | |
| "eval_samples_per_second": 4.459, | |
| "eval_steps_per_second": 0.594, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.235294117647059, | |
| "grad_norm": 0.19050121307373047, | |
| "learning_rate": 0.00029795419551040833, | |
| "loss": 0.5302, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.8235294117647056, | |
| "grad_norm": 0.17813433706760406, | |
| "learning_rate": 0.0002954100398908995, | |
| "loss": 0.4439, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.5779597759246826, | |
| "eval_runtime": 3.3656, | |
| "eval_samples_per_second": 4.457, | |
| "eval_steps_per_second": 0.594, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 3.3529411764705883, | |
| "grad_norm": 0.23658457398414612, | |
| "learning_rate": 0.00029187258625509513, | |
| "loss": 0.4257, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.9411764705882355, | |
| "grad_norm": 0.27489331364631653, | |
| "learning_rate": 0.00028736599899825856, | |
| "loss": 0.3254, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.5997567772865295, | |
| "eval_runtime": 3.3651, | |
| "eval_samples_per_second": 4.458, | |
| "eval_steps_per_second": 0.594, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 4.470588235294118, | |
| "grad_norm": 0.3510420620441437, | |
| "learning_rate": 0.00028192106268097334, | |
| "loss": 0.2985, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.4653756320476532, | |
| "learning_rate": 0.00027557497173937923, | |
| "loss": 0.2828, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.6584295034408569, | |
| "eval_runtime": 3.3668, | |
| "eval_samples_per_second": 4.455, | |
| "eval_steps_per_second": 0.594, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 5.588235294117647, | |
| "grad_norm": 0.3469710052013397, | |
| "learning_rate": 0.00026837107640945905, | |
| "loss": 0.2038, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.751620888710022, | |
| "eval_runtime": 3.3651, | |
| "eval_samples_per_second": 4.458, | |
| "eval_steps_per_second": 0.594, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 6.117647058823529, | |
| "grad_norm": 0.3903103470802307, | |
| "learning_rate": 0.0002603585866009697, | |
| "loss": 0.1656, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 6.705882352941177, | |
| "grad_norm": 0.38518330454826355, | |
| "learning_rate": 0.00025159223574386114, | |
| "loss": 0.1296, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.8753386735916138, | |
| "eval_runtime": 3.3654, | |
| "eval_samples_per_second": 4.457, | |
| "eval_steps_per_second": 0.594, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 7.235294117647059, | |
| "grad_norm": 0.24312527477741241, | |
| "learning_rate": 0.00024213190690345018, | |
| "loss": 0.085, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 7.823529411764706, | |
| "grad_norm": 0.3267911970615387, | |
| "learning_rate": 0.00023204222371836405, | |
| "loss": 0.0751, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.8750011920928955, | |
| "eval_runtime": 3.3678, | |
| "eval_samples_per_second": 4.454, | |
| "eval_steps_per_second": 0.594, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 8.352941176470589, | |
| "grad_norm": 0.24549448490142822, | |
| "learning_rate": 0.00022139210895556104, | |
| "loss": 0.0576, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "grad_norm": 0.3071894347667694, | |
| "learning_rate": 0.0002102543136979454, | |
| "loss": 0.0475, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.9990332126617432, | |
| "eval_runtime": 3.3671, | |
| "eval_samples_per_second": 4.455, | |
| "eval_steps_per_second": 0.594, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 9.470588235294118, | |
| "grad_norm": 0.17012739181518555, | |
| "learning_rate": 0.00019870492038070252, | |
| "loss": 0.0282, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.2809393107891083, | |
| "learning_rate": 0.00018682282307111987, | |
| "loss": 0.0249, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 1.022998571395874, | |
| "eval_runtime": 3.3665, | |
| "eval_samples_per_second": 4.456, | |
| "eval_steps_per_second": 0.594, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.588235294117647, | |
| "grad_norm": 0.16731834411621094, | |
| "learning_rate": 0.00017468918854211007, | |
| "loss": 0.0151, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 1.0694772005081177, | |
| "eval_runtime": 3.3662, | |
| "eval_samples_per_second": 4.456, | |
| "eval_steps_per_second": 0.594, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 11.117647058823529, | |
| "grad_norm": 0.149811789393425, | |
| "learning_rate": 0.00016238690182084986, | |
| "loss": 0.0166, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 11.705882352941176, | |
| "grad_norm": 0.13291925191879272, | |
| "learning_rate": 0.00015, | |
| "loss": 0.0095, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 1.0867458581924438, | |
| "eval_runtime": 3.3664, | |
| "eval_samples_per_second": 4.456, | |
| "eval_steps_per_second": 0.594, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 12.235294117647058, | |
| "grad_norm": 0.08576802164316177, | |
| "learning_rate": 0.00013761309817915014, | |
| "loss": 0.009, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 12.823529411764707, | |
| "grad_norm": 0.07071960717439651, | |
| "learning_rate": 0.00012531081145788987, | |
| "loss": 0.0075, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 1.1065537929534912, | |
| "eval_runtime": 3.3672, | |
| "eval_samples_per_second": 4.455, | |
| "eval_steps_per_second": 0.594, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 13.352941176470589, | |
| "grad_norm": 0.1268472969532013, | |
| "learning_rate": 0.00011317717692888012, | |
| "loss": 0.0125, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 13.941176470588236, | |
| "grad_norm": 0.08666448295116425, | |
| "learning_rate": 0.00010129507961929748, | |
| "loss": 0.0058, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 1.0826694965362549, | |
| "eval_runtime": 3.3663, | |
| "eval_samples_per_second": 4.456, | |
| "eval_steps_per_second": 0.594, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 14.470588235294118, | |
| "grad_norm": 0.20298424363136292, | |
| "learning_rate": 8.97456863020546e-05, | |
| "loss": 0.0063, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.07050978392362595, | |
| "learning_rate": 7.860789104443896e-05, | |
| "loss": 0.0059, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 1.107211947441101, | |
| "eval_runtime": 3.3663, | |
| "eval_samples_per_second": 4.456, | |
| "eval_steps_per_second": 0.594, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 15.588235294117647, | |
| "grad_norm": 0.042294692248106, | |
| "learning_rate": 6.795777628163599e-05, | |
| "loss": 0.0045, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 1.1277925968170166, | |
| "eval_runtime": 3.3661, | |
| "eval_samples_per_second": 4.456, | |
| "eval_steps_per_second": 0.594, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 16.11764705882353, | |
| "grad_norm": 0.03719251602888107, | |
| "learning_rate": 5.786809309654982e-05, | |
| "loss": 0.007, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 16.705882352941178, | |
| "grad_norm": 0.03469066694378853, | |
| "learning_rate": 4.840776425613886e-05, | |
| "loss": 0.0051, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_loss": 1.1320844888687134, | |
| "eval_runtime": 3.3665, | |
| "eval_samples_per_second": 4.456, | |
| "eval_steps_per_second": 0.594, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 17.235294117647058, | |
| "grad_norm": 0.04715495929121971, | |
| "learning_rate": 3.964141339903026e-05, | |
| "loss": 0.0039, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 17.823529411764707, | |
| "grad_norm": 0.10318444669246674, | |
| "learning_rate": 3.162892359054098e-05, | |
| "loss": 0.0031, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 1.1403193473815918, | |
| "eval_runtime": 3.3665, | |
| "eval_samples_per_second": 4.456, | |
| "eval_steps_per_second": 0.594, | |
| "step": 162 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.395354535852851e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |