| { | |
| "best_metric": 0.39345207810401917, | |
| "best_model_checkpoint": "mikhail-panzo/zlm-fil-ceb_b64_le5_s8000/checkpoint-3000", | |
| "epoch": 118.81188118811882, | |
| "eval_steps": 500, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.9801980198019802, | |
| "grad_norm": 1.2683665752410889, | |
| "learning_rate": 2.5000000000000004e-07, | |
| "loss": 0.4932, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.9603960396039604, | |
| "grad_norm": 1.1266463994979858, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 0.4898, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 5.9405940594059405, | |
| "grad_norm": 1.1800968647003174, | |
| "learning_rate": 7.5e-07, | |
| "loss": 0.4839, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 7.920792079207921, | |
| "grad_norm": 0.8235568404197693, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.4785, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 9.900990099009901, | |
| "grad_norm": 1.313211441040039, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.4767, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 11.881188118811881, | |
| "grad_norm": 0.7831560373306274, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.4681, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 13.861386138613861, | |
| "grad_norm": 0.7987237572669983, | |
| "learning_rate": 1.75e-06, | |
| "loss": 0.4658, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 15.841584158415841, | |
| "grad_norm": 0.7143813371658325, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.4627, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 17.821782178217823, | |
| "grad_norm": 0.8037531971931458, | |
| "learning_rate": 2.25e-06, | |
| "loss": 0.461, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 19.801980198019802, | |
| "grad_norm": 0.79031902551651, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.4592, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 19.801980198019802, | |
| "eval_loss": 0.42528408765792847, | |
| "eval_runtime": 7.8982, | |
| "eval_samples_per_second": 22.79, | |
| "eval_steps_per_second": 2.912, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 21.782178217821784, | |
| "grad_norm": 0.7966288328170776, | |
| "learning_rate": 2.7500000000000004e-06, | |
| "loss": 0.4551, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 23.762376237623762, | |
| "grad_norm": 0.8126183152198792, | |
| "learning_rate": 3e-06, | |
| "loss": 0.4542, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 25.742574257425744, | |
| "grad_norm": 0.6602805852890015, | |
| "learning_rate": 3.2500000000000002e-06, | |
| "loss": 0.4535, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 27.722772277227723, | |
| "grad_norm": 0.7177000641822815, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.4484, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 29.702970297029704, | |
| "grad_norm": 0.7912996411323547, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.4486, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 31.683168316831683, | |
| "grad_norm": 0.6251747608184814, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.4463, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 33.663366336633665, | |
| "grad_norm": 1.003780722618103, | |
| "learning_rate": 4.25e-06, | |
| "loss": 0.4474, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 35.64356435643565, | |
| "grad_norm": 0.742171585559845, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.4434, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 37.62376237623762, | |
| "grad_norm": 0.6034647822380066, | |
| "learning_rate": 4.75e-06, | |
| "loss": 0.4425, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 39.603960396039604, | |
| "grad_norm": 0.6990635395050049, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4381, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 39.603960396039604, | |
| "eval_loss": 0.41002073884010315, | |
| "eval_runtime": 6.7782, | |
| "eval_samples_per_second": 26.556, | |
| "eval_steps_per_second": 3.393, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 41.584158415841586, | |
| "grad_norm": 0.9420008659362793, | |
| "learning_rate": 5.2500000000000006e-06, | |
| "loss": 0.4421, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 43.56435643564357, | |
| "grad_norm": 0.7143107652664185, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 0.4389, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 45.54455445544554, | |
| "grad_norm": 0.7934616804122925, | |
| "learning_rate": 5.75e-06, | |
| "loss": 0.4369, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 47.524752475247524, | |
| "grad_norm": 0.9682563543319702, | |
| "learning_rate": 6e-06, | |
| "loss": 0.4355, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 49.504950495049506, | |
| "grad_norm": 0.7880070209503174, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.4359, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 51.48514851485149, | |
| "grad_norm": 0.8443828821182251, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 0.4324, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 53.46534653465346, | |
| "grad_norm": 0.724743127822876, | |
| "learning_rate": 6.750000000000001e-06, | |
| "loss": 0.4348, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 55.445544554455445, | |
| "grad_norm": 0.7557791471481323, | |
| "learning_rate": 7e-06, | |
| "loss": 0.433, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 57.42574257425743, | |
| "grad_norm": 0.7936097383499146, | |
| "learning_rate": 7.25e-06, | |
| "loss": 0.4273, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 59.40594059405941, | |
| "grad_norm": 0.7041624784469604, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.4281, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 59.40594059405941, | |
| "eval_loss": 0.40218469500541687, | |
| "eval_runtime": 7.4309, | |
| "eval_samples_per_second": 24.223, | |
| "eval_steps_per_second": 3.095, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 61.386138613861384, | |
| "grad_norm": 0.9194286465644836, | |
| "learning_rate": 7.75e-06, | |
| "loss": 0.4275, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 63.366336633663366, | |
| "grad_norm": 0.7239477634429932, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.4245, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 65.34653465346534, | |
| "grad_norm": 0.7235706448554993, | |
| "learning_rate": 8.25e-06, | |
| "loss": 0.4248, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 67.32673267326733, | |
| "grad_norm": 0.7281640768051147, | |
| "learning_rate": 8.5e-06, | |
| "loss": 0.425, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 69.3069306930693, | |
| "grad_norm": 1.011294960975647, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.4238, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 71.2871287128713, | |
| "grad_norm": 1.1249107122421265, | |
| "learning_rate": 9e-06, | |
| "loss": 0.4203, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 73.26732673267327, | |
| "grad_norm": 0.7369679808616638, | |
| "learning_rate": 9.250000000000001e-06, | |
| "loss": 0.422, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 75.24752475247524, | |
| "grad_norm": 0.704367995262146, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.4201, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 77.22772277227723, | |
| "grad_norm": 0.8307437896728516, | |
| "learning_rate": 9.75e-06, | |
| "loss": 0.4203, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 79.20792079207921, | |
| "grad_norm": 0.7827133536338806, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4195, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 79.20792079207921, | |
| "eval_loss": 0.39878609776496887, | |
| "eval_runtime": 8.1579, | |
| "eval_samples_per_second": 22.064, | |
| "eval_steps_per_second": 2.819, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 81.18811881188118, | |
| "grad_norm": 0.8909313082695007, | |
| "learning_rate": 9.916666666666668e-06, | |
| "loss": 0.4203, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 83.16831683168317, | |
| "grad_norm": 0.7147483825683594, | |
| "learning_rate": 9.833333333333333e-06, | |
| "loss": 0.418, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 85.14851485148515, | |
| "grad_norm": 0.6844545006752014, | |
| "learning_rate": 9.75e-06, | |
| "loss": 0.4167, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 87.12871287128714, | |
| "grad_norm": 0.88512122631073, | |
| "learning_rate": 9.666666666666667e-06, | |
| "loss": 0.4108, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 89.10891089108911, | |
| "grad_norm": 0.7097252607345581, | |
| "learning_rate": 9.583333333333335e-06, | |
| "loss": 0.4135, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 91.08910891089108, | |
| "grad_norm": 0.8400272727012634, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.4147, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 93.06930693069307, | |
| "grad_norm": 0.7614939212799072, | |
| "learning_rate": 9.416666666666667e-06, | |
| "loss": 0.4119, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 95.04950495049505, | |
| "grad_norm": 0.8099732995033264, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.4108, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 97.02970297029702, | |
| "grad_norm": 1.0697815418243408, | |
| "learning_rate": 9.250000000000001e-06, | |
| "loss": 0.4095, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 99.00990099009901, | |
| "grad_norm": 0.7235235571861267, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.4134, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 99.00990099009901, | |
| "eval_loss": 0.39545077085494995, | |
| "eval_runtime": 7.724, | |
| "eval_samples_per_second": 23.304, | |
| "eval_steps_per_second": 2.978, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 100.99009900990099, | |
| "grad_norm": 0.6166990399360657, | |
| "learning_rate": 9.083333333333333e-06, | |
| "loss": 0.4097, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 102.97029702970298, | |
| "grad_norm": 0.6944295763969421, | |
| "learning_rate": 9e-06, | |
| "loss": 0.4107, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 104.95049504950495, | |
| "grad_norm": 0.8469858765602112, | |
| "learning_rate": 8.916666666666667e-06, | |
| "loss": 0.4056, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 106.93069306930693, | |
| "grad_norm": 0.6261239647865295, | |
| "learning_rate": 8.833333333333334e-06, | |
| "loss": 0.409, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 108.91089108910892, | |
| "grad_norm": 0.6304221749305725, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.4081, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 110.89108910891089, | |
| "grad_norm": 0.6513387560844421, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 0.4071, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 112.87128712871286, | |
| "grad_norm": 0.9311261177062988, | |
| "learning_rate": 8.583333333333333e-06, | |
| "loss": 0.4091, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 114.85148514851485, | |
| "grad_norm": 0.7296612858772278, | |
| "learning_rate": 8.5e-06, | |
| "loss": 0.4055, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 116.83168316831683, | |
| "grad_norm": 0.8149676322937012, | |
| "learning_rate": 8.416666666666667e-06, | |
| "loss": 0.4052, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 118.81188118811882, | |
| "grad_norm": 0.8266366124153137, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.4049, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 118.81188118811882, | |
| "eval_loss": 0.39345207810401917, | |
| "eval_runtime": 7.2151, | |
| "eval_samples_per_second": 24.948, | |
| "eval_steps_per_second": 3.188, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 8000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 320, | |
| "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": 3.2419037301352416e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |