| { | |
| "best_metric": 0.41038990020751953, | |
| "best_model_checkpoint": "mikhail-panzo/fil_b64_le5_s4000/checkpoint-3500", | |
| "epoch": 155.55555555555554, | |
| "eval_steps": 500, | |
| "global_step": 3500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 5.408687114715576, | |
| "learning_rate": 2.5000000000000004e-07, | |
| "loss": 0.8111, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.444444444444445, | |
| "grad_norm": 3.0843052864074707, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 0.7694, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 5.368913650512695, | |
| "learning_rate": 7.5e-07, | |
| "loss": 0.7562, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 8.88888888888889, | |
| "grad_norm": 1.200212836265564, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.7184, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 11.11111111111111, | |
| "grad_norm": 1.2329410314559937, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.7022, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 13.333333333333334, | |
| "grad_norm": 1.7799458503723145, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.6823, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 15.555555555555555, | |
| "grad_norm": 2.0204672813415527, | |
| "learning_rate": 1.75e-06, | |
| "loss": 0.6635, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 17.77777777777778, | |
| "grad_norm": 1.8028016090393066, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.6499, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 1.5485812425613403, | |
| "learning_rate": 2.25e-06, | |
| "loss": 0.6216, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 22.22222222222222, | |
| "grad_norm": 1.0476795434951782, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.575, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 22.22222222222222, | |
| "eval_loss": 0.49667900800704956, | |
| "eval_runtime": 8.3577, | |
| "eval_samples_per_second": 19.024, | |
| "eval_steps_per_second": 2.393, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 24.444444444444443, | |
| "grad_norm": 1.134787678718567, | |
| "learning_rate": 2.7500000000000004e-06, | |
| "loss": 0.5464, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 26.666666666666668, | |
| "grad_norm": 1.543503999710083, | |
| "learning_rate": 3e-06, | |
| "loss": 0.5432, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 28.88888888888889, | |
| "grad_norm": 1.0410478115081787, | |
| "learning_rate": 3.2500000000000002e-06, | |
| "loss": 0.5355, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 31.11111111111111, | |
| "grad_norm": 1.5184495449066162, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.5289, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 33.333333333333336, | |
| "grad_norm": 1.490964651107788, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.5225, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 35.55555555555556, | |
| "grad_norm": 1.2980331182479858, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.5122, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 37.77777777777778, | |
| "grad_norm": 1.2520424127578735, | |
| "learning_rate": 4.25e-06, | |
| "loss": 0.51, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 1.2899489402770996, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.5051, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 42.22222222222222, | |
| "grad_norm": 1.2064433097839355, | |
| "learning_rate": 4.75e-06, | |
| "loss": 0.5006, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 44.44444444444444, | |
| "grad_norm": 1.1178175210952759, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4945, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 44.44444444444444, | |
| "eval_loss": 0.4459993243217468, | |
| "eval_runtime": 7.7575, | |
| "eval_samples_per_second": 20.496, | |
| "eval_steps_per_second": 2.578, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 46.666666666666664, | |
| "grad_norm": 1.2157412767410278, | |
| "learning_rate": 5.2500000000000006e-06, | |
| "loss": 0.4976, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 48.888888888888886, | |
| "grad_norm": 1.0625967979431152, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 0.485, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 51.111111111111114, | |
| "grad_norm": 1.2712880373001099, | |
| "learning_rate": 5.75e-06, | |
| "loss": 0.4835, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 53.333333333333336, | |
| "grad_norm": 1.1251996755599976, | |
| "learning_rate": 6e-06, | |
| "loss": 0.482, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 55.55555555555556, | |
| "grad_norm": 1.1670546531677246, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.4847, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 57.77777777777778, | |
| "grad_norm": 1.2118104696273804, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 0.4804, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "grad_norm": 1.352782964706421, | |
| "learning_rate": 6.750000000000001e-06, | |
| "loss": 0.4763, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 62.22222222222222, | |
| "grad_norm": 0.9930539727210999, | |
| "learning_rate": 7e-06, | |
| "loss": 0.4776, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 64.44444444444444, | |
| "grad_norm": 0.9236252903938293, | |
| "learning_rate": 7.25e-06, | |
| "loss": 0.4646, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 66.66666666666667, | |
| "grad_norm": 1.295523762702942, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.4681, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 66.66666666666667, | |
| "eval_loss": 0.43010735511779785, | |
| "eval_runtime": 7.7488, | |
| "eval_samples_per_second": 20.519, | |
| "eval_steps_per_second": 2.581, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 68.88888888888889, | |
| "grad_norm": 1.0008094310760498, | |
| "learning_rate": 7.75e-06, | |
| "loss": 0.4673, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 71.11111111111111, | |
| "grad_norm": 1.275070071220398, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.4654, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 73.33333333333333, | |
| "grad_norm": 1.0170423984527588, | |
| "learning_rate": 8.25e-06, | |
| "loss": 0.4571, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 75.55555555555556, | |
| "grad_norm": 1.0809487104415894, | |
| "learning_rate": 8.5e-06, | |
| "loss": 0.4578, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 77.77777777777777, | |
| "grad_norm": 1.9192956686019897, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.4574, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "grad_norm": 2.315152406692505, | |
| "learning_rate": 9e-06, | |
| "loss": 0.4566, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 82.22222222222223, | |
| "grad_norm": 1.0425012111663818, | |
| "learning_rate": 9.250000000000001e-06, | |
| "loss": 0.4583, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 84.44444444444444, | |
| "grad_norm": 0.896725058555603, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.4542, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 86.66666666666667, | |
| "grad_norm": 1.2378209829330444, | |
| "learning_rate": 9.75e-06, | |
| "loss": 0.455, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 88.88888888888889, | |
| "grad_norm": 1.078316330909729, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4514, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 88.88888888888889, | |
| "eval_loss": 0.4194343686103821, | |
| "eval_runtime": 8.6226, | |
| "eval_samples_per_second": 18.44, | |
| "eval_steps_per_second": 2.319, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 91.11111111111111, | |
| "grad_norm": 1.1163235902786255, | |
| "learning_rate": 9.75e-06, | |
| "loss": 0.456, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 93.33333333333333, | |
| "grad_norm": 1.3120802640914917, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.4444, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 95.55555555555556, | |
| "grad_norm": 1.4265310764312744, | |
| "learning_rate": 9.250000000000001e-06, | |
| "loss": 0.449, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 97.77777777777777, | |
| "grad_norm": 1.206393837928772, | |
| "learning_rate": 9e-06, | |
| "loss": 0.4441, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "grad_norm": 1.218729853630066, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.4432, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 102.22222222222223, | |
| "grad_norm": 1.154038906097412, | |
| "learning_rate": 8.5e-06, | |
| "loss": 0.448, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 104.44444444444444, | |
| "grad_norm": 0.9223791360855103, | |
| "learning_rate": 8.25e-06, | |
| "loss": 0.4442, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 106.66666666666667, | |
| "grad_norm": 1.0341440439224243, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.4424, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 108.88888888888889, | |
| "grad_norm": 1.072590708732605, | |
| "learning_rate": 7.75e-06, | |
| "loss": 0.4365, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 111.11111111111111, | |
| "grad_norm": 0.8107730150222778, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.4396, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 111.11111111111111, | |
| "eval_loss": 0.41287755966186523, | |
| "eval_runtime": 8.2486, | |
| "eval_samples_per_second": 19.276, | |
| "eval_steps_per_second": 2.425, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 113.33333333333333, | |
| "grad_norm": 1.7002819776535034, | |
| "learning_rate": 7.25e-06, | |
| "loss": 0.4389, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 115.55555555555556, | |
| "grad_norm": 0.957378089427948, | |
| "learning_rate": 7e-06, | |
| "loss": 0.4345, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 117.77777777777777, | |
| "grad_norm": 1.361790657043457, | |
| "learning_rate": 6.750000000000001e-06, | |
| "loss": 0.4351, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 120.0, | |
| "grad_norm": 1.0980664491653442, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 0.4371, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 122.22222222222223, | |
| "grad_norm": 0.8956480026245117, | |
| "learning_rate": 6.255e-06, | |
| "loss": 0.4413, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 124.44444444444444, | |
| "grad_norm": 1.1513115167617798, | |
| "learning_rate": 6.005000000000001e-06, | |
| "loss": 0.4332, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 126.66666666666667, | |
| "grad_norm": 0.7204779386520386, | |
| "learning_rate": 5.755000000000001e-06, | |
| "loss": 0.4364, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 128.88888888888889, | |
| "grad_norm": 1.0155471563339233, | |
| "learning_rate": 5.505000000000001e-06, | |
| "loss": 0.4425, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 131.11111111111111, | |
| "grad_norm": 0.9366397261619568, | |
| "learning_rate": 5.2550000000000005e-06, | |
| "loss": 0.4358, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 133.33333333333334, | |
| "grad_norm": 1.350168228149414, | |
| "learning_rate": 5.0049999999999995e-06, | |
| "loss": 0.432, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 133.33333333333334, | |
| "eval_loss": 0.412423700094223, | |
| "eval_runtime": 8.2357, | |
| "eval_samples_per_second": 19.306, | |
| "eval_steps_per_second": 2.428, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 135.55555555555554, | |
| "grad_norm": 0.9131026268005371, | |
| "learning_rate": 4.755e-06, | |
| "loss": 0.4311, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 137.77777777777777, | |
| "grad_norm": 1.1815382242202759, | |
| "learning_rate": 4.505e-06, | |
| "loss": 0.4295, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 140.0, | |
| "grad_norm": 1.2635456323623657, | |
| "learning_rate": 4.255e-06, | |
| "loss": 0.4301, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 142.22222222222223, | |
| "grad_norm": 0.8922115564346313, | |
| "learning_rate": 4.005000000000001e-06, | |
| "loss": 0.4308, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 144.44444444444446, | |
| "grad_norm": 0.9796882271766663, | |
| "learning_rate": 3.7550000000000005e-06, | |
| "loss": 0.4303, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 146.66666666666666, | |
| "grad_norm": 0.9250928163528442, | |
| "learning_rate": 3.505e-06, | |
| "loss": 0.429, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 148.88888888888889, | |
| "grad_norm": 1.269997477531433, | |
| "learning_rate": 3.255e-06, | |
| "loss": 0.4271, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 151.11111111111111, | |
| "grad_norm": 1.0031682252883911, | |
| "learning_rate": 3.005e-06, | |
| "loss": 0.4289, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 153.33333333333334, | |
| "grad_norm": 1.0691368579864502, | |
| "learning_rate": 2.7550000000000003e-06, | |
| "loss": 0.4281, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 155.55555555555554, | |
| "grad_norm": 0.9480464458465576, | |
| "learning_rate": 2.505e-06, | |
| "loss": 0.43, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 155.55555555555554, | |
| "eval_loss": 0.41038990020751953, | |
| "eval_runtime": 8.2981, | |
| "eval_samples_per_second": 19.161, | |
| "eval_steps_per_second": 2.41, | |
| "step": 3500 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 4000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 182, | |
| "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": 4.822102582474954e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |