| { | |
| "best_metric": 0.4075862765312195, | |
| "best_model_checkpoint": "mikhail-panzo/fil_b32_le5_s8000/checkpoint-4000", | |
| "epoch": 88.88888888888889, | |
| "eval_steps": 500, | |
| "global_step": 4000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 3.579704523086548, | |
| "learning_rate": 2.4500000000000004e-07, | |
| "loss": 0.8282, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 2.0918073654174805, | |
| "learning_rate": 4.95e-07, | |
| "loss": 0.7857, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 1.283610224723816, | |
| "learning_rate": 7.450000000000001e-07, | |
| "loss": 0.7356, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.444444444444445, | |
| "grad_norm": 1.5381903648376465, | |
| "learning_rate": 9.950000000000002e-07, | |
| "loss": 0.7606, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 5.555555555555555, | |
| "grad_norm": 1.613647222518921, | |
| "learning_rate": 1.2450000000000002e-06, | |
| "loss": 0.7314, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 2.6762027740478516, | |
| "learning_rate": 1.495e-06, | |
| "loss": 0.6933, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 7.777777777777778, | |
| "grad_norm": 2.667527675628662, | |
| "learning_rate": 1.745e-06, | |
| "loss": 0.6863, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 8.88888888888889, | |
| "grad_norm": 1.9682470560073853, | |
| "learning_rate": 1.9950000000000004e-06, | |
| "loss": 0.6649, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 2.0042927265167236, | |
| "learning_rate": 2.245e-06, | |
| "loss": 0.6507, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 11.11111111111111, | |
| "grad_norm": 2.6399147510528564, | |
| "learning_rate": 2.4950000000000003e-06, | |
| "loss": 0.632, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 11.11111111111111, | |
| "eval_loss": 0.5323343276977539, | |
| "eval_runtime": 7.6175, | |
| "eval_samples_per_second": 20.873, | |
| "eval_steps_per_second": 2.626, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 12.222222222222221, | |
| "grad_norm": 2.593517780303955, | |
| "learning_rate": 2.7450000000000004e-06, | |
| "loss": 0.5956, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 13.333333333333334, | |
| "grad_norm": 5.524081707000732, | |
| "learning_rate": 2.99e-06, | |
| "loss": 0.577, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 14.444444444444445, | |
| "grad_norm": 2.2165024280548096, | |
| "learning_rate": 3.2400000000000003e-06, | |
| "loss": 0.5614, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 15.555555555555555, | |
| "grad_norm": 2.4440901279449463, | |
| "learning_rate": 3.49e-06, | |
| "loss": 0.5654, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 16.666666666666668, | |
| "grad_norm": 1.8401445150375366, | |
| "learning_rate": 3.74e-06, | |
| "loss": 0.5329, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 17.77777777777778, | |
| "grad_norm": 2.101787567138672, | |
| "learning_rate": 3.990000000000001e-06, | |
| "loss": 0.5293, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 18.88888888888889, | |
| "grad_norm": 2.1338887214660645, | |
| "learning_rate": 4.24e-06, | |
| "loss": 0.5345, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 1.8499983549118042, | |
| "learning_rate": 4.49e-06, | |
| "loss": 0.5148, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 21.11111111111111, | |
| "grad_norm": 1.3982588052749634, | |
| "learning_rate": 4.74e-06, | |
| "loss": 0.5085, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 22.22222222222222, | |
| "grad_norm": 1.7611488103866577, | |
| "learning_rate": 4.9900000000000005e-06, | |
| "loss": 0.519, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 22.22222222222222, | |
| "eval_loss": 0.4493897557258606, | |
| "eval_runtime": 7.4167, | |
| "eval_samples_per_second": 21.438, | |
| "eval_steps_per_second": 2.697, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 23.333333333333332, | |
| "grad_norm": 2.625955581665039, | |
| "learning_rate": 5.240000000000001e-06, | |
| "loss": 0.5194, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 24.444444444444443, | |
| "grad_norm": 1.75005304813385, | |
| "learning_rate": 5.490000000000001e-06, | |
| "loss": 0.5152, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 25.555555555555557, | |
| "grad_norm": 2.279965877532959, | |
| "learning_rate": 5.74e-06, | |
| "loss": 0.5129, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 26.666666666666668, | |
| "grad_norm": 1.5506455898284912, | |
| "learning_rate": 5.99e-06, | |
| "loss": 0.4941, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 27.77777777777778, | |
| "grad_norm": 1.830841064453125, | |
| "learning_rate": 6.24e-06, | |
| "loss": 0.4899, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 28.88888888888889, | |
| "grad_norm": 2.0044639110565186, | |
| "learning_rate": 6.4900000000000005e-06, | |
| "loss": 0.496, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 2.272815465927124, | |
| "learning_rate": 6.740000000000001e-06, | |
| "loss": 0.4906, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 31.11111111111111, | |
| "grad_norm": 1.460723638534546, | |
| "learning_rate": 6.99e-06, | |
| "loss": 0.4741, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 32.22222222222222, | |
| "grad_norm": 1.7440383434295654, | |
| "learning_rate": 7.24e-06, | |
| "loss": 0.4821, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 33.333333333333336, | |
| "grad_norm": 1.4221312999725342, | |
| "learning_rate": 7.49e-06, | |
| "loss": 0.4816, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 33.333333333333336, | |
| "eval_loss": 0.42913728952407837, | |
| "eval_runtime": 7.4808, | |
| "eval_samples_per_second": 21.254, | |
| "eval_steps_per_second": 2.674, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 34.44444444444444, | |
| "grad_norm": 1.8739147186279297, | |
| "learning_rate": 7.74e-06, | |
| "loss": 0.4794, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 35.55555555555556, | |
| "grad_norm": 1.1802700757980347, | |
| "learning_rate": 7.990000000000001e-06, | |
| "loss": 0.4824, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 36.666666666666664, | |
| "grad_norm": 1.364354133605957, | |
| "learning_rate": 8.24e-06, | |
| "loss": 0.4739, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 37.77777777777778, | |
| "grad_norm": 1.5628408193588257, | |
| "learning_rate": 8.49e-06, | |
| "loss": 0.4697, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 38.888888888888886, | |
| "grad_norm": 4.399195671081543, | |
| "learning_rate": 8.740000000000001e-06, | |
| "loss": 0.4685, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 2.055992841720581, | |
| "learning_rate": 8.99e-06, | |
| "loss": 0.4681, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 41.111111111111114, | |
| "grad_norm": 4.06639289855957, | |
| "learning_rate": 9.240000000000001e-06, | |
| "loss": 0.4708, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 42.22222222222222, | |
| "grad_norm": 1.4587314128875732, | |
| "learning_rate": 9.49e-06, | |
| "loss": 0.4665, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 43.333333333333336, | |
| "grad_norm": 1.4163354635238647, | |
| "learning_rate": 9.74e-06, | |
| "loss": 0.4648, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 44.44444444444444, | |
| "grad_norm": 2.2223002910614014, | |
| "learning_rate": 9.990000000000001e-06, | |
| "loss": 0.481, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 44.44444444444444, | |
| "eval_loss": 0.42109793424606323, | |
| "eval_runtime": 7.5115, | |
| "eval_samples_per_second": 21.168, | |
| "eval_steps_per_second": 2.663, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 45.55555555555556, | |
| "grad_norm": 1.7627613544464111, | |
| "learning_rate": 9.920000000000002e-06, | |
| "loss": 0.4598, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 46.666666666666664, | |
| "grad_norm": 1.7885198593139648, | |
| "learning_rate": 9.836666666666668e-06, | |
| "loss": 0.4697, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 47.77777777777778, | |
| "grad_norm": 1.8856687545776367, | |
| "learning_rate": 9.753333333333335e-06, | |
| "loss": 0.4503, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 48.888888888888886, | |
| "grad_norm": 2.0350635051727295, | |
| "learning_rate": 9.67e-06, | |
| "loss": 0.4573, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 1.7863447666168213, | |
| "learning_rate": 9.586666666666667e-06, | |
| "loss": 0.454, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 51.111111111111114, | |
| "grad_norm": 1.5656064748764038, | |
| "learning_rate": 9.503333333333334e-06, | |
| "loss": 0.4488, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 52.22222222222222, | |
| "grad_norm": 2.449720859527588, | |
| "learning_rate": 9.42e-06, | |
| "loss": 0.4552, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 53.333333333333336, | |
| "grad_norm": 1.5119515657424927, | |
| "learning_rate": 9.336666666666666e-06, | |
| "loss": 0.4472, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 54.44444444444444, | |
| "grad_norm": 1.2574548721313477, | |
| "learning_rate": 9.253333333333333e-06, | |
| "loss": 0.4623, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 55.55555555555556, | |
| "grad_norm": 1.1045562028884888, | |
| "learning_rate": 9.17e-06, | |
| "loss": 0.4459, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 55.55555555555556, | |
| "eval_loss": 0.4139043688774109, | |
| "eval_runtime": 7.7673, | |
| "eval_samples_per_second": 20.471, | |
| "eval_steps_per_second": 2.575, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 56.666666666666664, | |
| "grad_norm": 2.62166428565979, | |
| "learning_rate": 9.086666666666667e-06, | |
| "loss": 0.4548, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 57.77777777777778, | |
| "grad_norm": 1.6748510599136353, | |
| "learning_rate": 9.003333333333334e-06, | |
| "loss": 0.4505, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 58.888888888888886, | |
| "grad_norm": 1.5706872940063477, | |
| "learning_rate": 8.920000000000001e-06, | |
| "loss": 0.4512, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "grad_norm": 1.9278029203414917, | |
| "learning_rate": 8.836666666666668e-06, | |
| "loss": 0.4546, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 61.111111111111114, | |
| "grad_norm": 1.5691492557525635, | |
| "learning_rate": 8.753333333333333e-06, | |
| "loss": 0.4407, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 62.22222222222222, | |
| "grad_norm": 1.2561596632003784, | |
| "learning_rate": 8.67e-06, | |
| "loss": 0.4407, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 63.333333333333336, | |
| "grad_norm": 1.7894113063812256, | |
| "learning_rate": 8.586666666666667e-06, | |
| "loss": 0.4407, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 64.44444444444444, | |
| "grad_norm": 1.5325671434402466, | |
| "learning_rate": 8.503333333333334e-06, | |
| "loss": 0.4409, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 65.55555555555556, | |
| "grad_norm": 2.345316171646118, | |
| "learning_rate": 8.42e-06, | |
| "loss": 0.4502, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 66.66666666666667, | |
| "grad_norm": 1.5134572982788086, | |
| "learning_rate": 8.336666666666668e-06, | |
| "loss": 0.4484, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 66.66666666666667, | |
| "eval_loss": 0.41138172149658203, | |
| "eval_runtime": 7.5275, | |
| "eval_samples_per_second": 21.122, | |
| "eval_steps_per_second": 2.657, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 67.77777777777777, | |
| "grad_norm": 2.0079166889190674, | |
| "learning_rate": 8.253333333333334e-06, | |
| "loss": 0.4391, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 68.88888888888889, | |
| "grad_norm": 1.5494792461395264, | |
| "learning_rate": 8.17e-06, | |
| "loss": 0.4426, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "grad_norm": 2.7749524116516113, | |
| "learning_rate": 8.086666666666667e-06, | |
| "loss": 0.4489, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 71.11111111111111, | |
| "grad_norm": 3.4985108375549316, | |
| "learning_rate": 8.003333333333334e-06, | |
| "loss": 0.4382, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 72.22222222222223, | |
| "grad_norm": 2.1767873764038086, | |
| "learning_rate": 7.92e-06, | |
| "loss": 0.456, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 73.33333333333333, | |
| "grad_norm": 1.7623385190963745, | |
| "learning_rate": 7.836666666666667e-06, | |
| "loss": 0.4312, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 74.44444444444444, | |
| "grad_norm": 6.053190231323242, | |
| "learning_rate": 7.753333333333334e-06, | |
| "loss": 0.441, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 75.55555555555556, | |
| "grad_norm": 1.4667917490005493, | |
| "learning_rate": 7.670000000000001e-06, | |
| "loss": 0.4407, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 76.66666666666667, | |
| "grad_norm": 1.4021631479263306, | |
| "learning_rate": 7.586666666666668e-06, | |
| "loss": 0.4332, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 77.77777777777777, | |
| "grad_norm": 2.186607599258423, | |
| "learning_rate": 7.503333333333333e-06, | |
| "loss": 0.4317, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 77.77777777777777, | |
| "eval_loss": 0.4080665707588196, | |
| "eval_runtime": 7.5364, | |
| "eval_samples_per_second": 21.098, | |
| "eval_steps_per_second": 2.654, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 78.88888888888889, | |
| "grad_norm": 2.0717358589172363, | |
| "learning_rate": 7.420000000000001e-06, | |
| "loss": 0.429, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "grad_norm": 1.5765066146850586, | |
| "learning_rate": 7.336666666666668e-06, | |
| "loss": 0.4382, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 81.11111111111111, | |
| "grad_norm": 1.5274180173873901, | |
| "learning_rate": 7.253333333333335e-06, | |
| "loss": 0.4365, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 82.22222222222223, | |
| "grad_norm": 1.8260159492492676, | |
| "learning_rate": 7.17e-06, | |
| "loss": 0.4331, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 83.33333333333333, | |
| "grad_norm": 1.5595194101333618, | |
| "learning_rate": 7.086666666666667e-06, | |
| "loss": 0.4314, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 84.44444444444444, | |
| "grad_norm": 1.8033090829849243, | |
| "learning_rate": 7.003333333333335e-06, | |
| "loss": 0.4339, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 85.55555555555556, | |
| "grad_norm": 1.943740725517273, | |
| "learning_rate": 6.92e-06, | |
| "loss": 0.4356, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 86.66666666666667, | |
| "grad_norm": 1.802024006843567, | |
| "learning_rate": 6.836666666666667e-06, | |
| "loss": 0.4307, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 87.77777777777777, | |
| "grad_norm": 1.2584848403930664, | |
| "learning_rate": 6.753333333333334e-06, | |
| "loss": 0.4228, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 88.88888888888889, | |
| "grad_norm": 1.6099965572357178, | |
| "learning_rate": 6.6700000000000005e-06, | |
| "loss": 0.4301, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 88.88888888888889, | |
| "eval_loss": 0.4075862765312195, | |
| "eval_runtime": 7.5332, | |
| "eval_samples_per_second": 21.107, | |
| "eval_steps_per_second": 2.655, | |
| "step": 4000 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 8000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 178, | |
| "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": 2.919129905785104e+16, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |