| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 3922, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 5.089058524173028e-07, | |
| "loss": 1.4534, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 2.544529262086514e-06, | |
| "loss": 1.2906, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 5.089058524173028e-06, | |
| "loss": 1.253, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.78515625, | |
| "learning_rate": 7.633587786259543e-06, | |
| "loss": 1.2664, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 1.0178117048346055e-05, | |
| "loss": 1.1842, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 1.2722646310432571e-05, | |
| "loss": 1.1703, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 1.5267175572519086e-05, | |
| "loss": 1.1021, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 1.78117048346056e-05, | |
| "loss": 1.05, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 2.035623409669211e-05, | |
| "loss": 0.9612, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 2.2900763358778628e-05, | |
| "loss": 0.9693, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 2.5445292620865142e-05, | |
| "loss": 0.9687, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 2.7989821882951656e-05, | |
| "loss": 0.8825, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 3.053435114503817e-05, | |
| "loss": 0.8668, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 3.307888040712468e-05, | |
| "loss": 0.8335, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 3.56234096692112e-05, | |
| "loss": 0.8146, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 3.816793893129771e-05, | |
| "loss": 0.8094, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 4.071246819338422e-05, | |
| "loss": 0.7349, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 4.325699745547074e-05, | |
| "loss": 0.7013, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.375, | |
| "learning_rate": 4.5801526717557256e-05, | |
| "loss": 0.7324, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 4.8346055979643766e-05, | |
| "loss": 0.7071, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 5.0890585241730283e-05, | |
| "loss": 0.7077, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 5.3435114503816794e-05, | |
| "loss": 0.7136, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 5.597964376590331e-05, | |
| "loss": 0.6851, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 5.852417302798983e-05, | |
| "loss": 0.6655, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 6.106870229007635e-05, | |
| "loss": 0.7056, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 6.361323155216285e-05, | |
| "loss": 0.6643, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 6.615776081424937e-05, | |
| "loss": 0.6525, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 6.870229007633588e-05, | |
| "loss": 0.6733, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 7.12468193384224e-05, | |
| "loss": 0.6742, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 7.379134860050892e-05, | |
| "loss": 0.6985, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 7.633587786259542e-05, | |
| "loss": 0.61, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 7.888040712468194e-05, | |
| "loss": 0.6655, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 8.142493638676844e-05, | |
| "loss": 0.6217, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 8.396946564885496e-05, | |
| "loss": 0.5982, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 8.651399491094148e-05, | |
| "loss": 0.6715, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 8.9058524173028e-05, | |
| "loss": 0.6818, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 9.160305343511451e-05, | |
| "loss": 0.7241, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 9.414758269720102e-05, | |
| "loss": 0.6986, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 9.669211195928753e-05, | |
| "loss": 0.6512, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 9.923664122137405e-05, | |
| "loss": 0.6016, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00010178117048346057, | |
| "loss": 0.6165, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00010432569974554708, | |
| "loss": 0.6832, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00010687022900763359, | |
| "loss": 0.6087, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00010941475826972009, | |
| "loss": 0.6266, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00011195928753180662, | |
| "loss": 0.6328, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00011450381679389313, | |
| "loss": 0.6222, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00011704834605597966, | |
| "loss": 0.6983, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00011959287531806616, | |
| "loss": 0.5978, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.0001221374045801527, | |
| "loss": 0.6174, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00012468193384223918, | |
| "loss": 0.6544, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.0001272264631043257, | |
| "loss": 0.5861, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00012977099236641222, | |
| "loss": 0.6288, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00013231552162849873, | |
| "loss": 0.6122, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00013486005089058525, | |
| "loss": 0.6029, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00013740458015267177, | |
| "loss": 0.5742, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00013994910941475828, | |
| "loss": 0.6598, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.0001424936386768448, | |
| "loss": 0.6627, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.0001450381679389313, | |
| "loss": 0.6372, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00014758269720101784, | |
| "loss": 0.6057, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00015012722646310433, | |
| "loss": 0.5944, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00015267175572519084, | |
| "loss": 0.5982, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00015521628498727736, | |
| "loss": 0.5889, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00015776081424936388, | |
| "loss": 0.6026, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0001603053435114504, | |
| "loss": 0.5539, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00016284987277353689, | |
| "loss": 0.6334, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00016539440203562343, | |
| "loss": 0.612, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.00016793893129770992, | |
| "loss": 0.6216, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00017048346055979644, | |
| "loss": 0.6328, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.00017302798982188295, | |
| "loss": 0.5892, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00017557251908396947, | |
| "loss": 0.5984, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.000178117048346056, | |
| "loss": 0.6217, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.0001806615776081425, | |
| "loss": 0.5835, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00018320610687022902, | |
| "loss": 0.6361, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00018575063613231554, | |
| "loss": 0.5791, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00018829516539440203, | |
| "loss": 0.6529, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.00019083969465648857, | |
| "loss": 0.6094, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 0.00019338422391857506, | |
| "loss": 0.5728, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00019592875318066158, | |
| "loss": 0.6022, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0001984732824427481, | |
| "loss": 0.5601, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00019999984150107918, | |
| "loss": 0.6034, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00019999805839398995, | |
| "loss": 0.6251, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0001999942940916057, | |
| "loss": 0.6018, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.0001999885486685061, | |
| "loss": 0.5812, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00019998082223852138, | |
| "loss": 0.6252, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00019997111495473018, | |
| "loss": 0.592, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0001999594270094565, | |
| "loss": 0.5532, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00019994575863426585, | |
| "loss": 0.5884, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.0001999301100999607, | |
| "loss": 0.634, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00019991248171657508, | |
| "loss": 0.6146, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00019989287383336853, | |
| "loss": 0.5627, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 0.00019987128683881896, | |
| "loss": 0.529, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00019984772116061523, | |
| "loss": 0.6148, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00019982217726564856, | |
| "loss": 0.5908, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00019979465566000317, | |
| "loss": 0.5796, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 0.00019976515688894638, | |
| "loss": 0.6234, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.0001997336815369179, | |
| "loss": 0.6051, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00019970023022751793, | |
| "loss": 0.5882, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00019966480362349515, | |
| "loss": 0.5858, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00019962740242673342, | |
| "loss": 0.5626, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00019958802737823786, | |
| "loss": 0.5921, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.0001995466792581202, | |
| "loss": 0.5472, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00019950335888558338, | |
| "loss": 0.5575, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00019945806711890526, | |
| "loss": 0.5166, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00019941080485542156, | |
| "loss": 0.6087, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00019936157303150823, | |
| "loss": 0.562, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.0001993103726225628, | |
| "loss": 0.6184, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00019925720464298511, | |
| "loss": 0.6244, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 0.00019920207014615703, | |
| "loss": 0.5759, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00019914497022442186, | |
| "loss": 0.5807, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.0001990859060090625, | |
| "loss": 0.5753, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00019902487867027918, | |
| "loss": 0.562, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00019896188941716613, | |
| "loss": 0.55, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00019889693949768766, | |
| "loss": 0.564, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 0.00019883003019865357, | |
| "loss": 0.5283, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 0.00019876116284569338, | |
| "loss": 0.5591, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.00019869033880323044, | |
| "loss": 0.5353, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00019861755947445456, | |
| "loss": 0.5084, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00019854282630129438, | |
| "loss": 0.5518, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 0.0001984661407643887, | |
| "loss": 0.5609, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00019838750438305733, | |
| "loss": 0.5606, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00019830691871527078, | |
| "loss": 0.5606, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00019822438535761953, | |
| "loss": 0.5632, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00019813990594528234, | |
| "loss": 0.5923, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00019805348215199377, | |
| "loss": 0.5813, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0001979651156900113, | |
| "loss": 0.5534, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00019787480831008102, | |
| "loss": 0.4909, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00019778256180140327, | |
| "loss": 0.5642, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00019768837799159704, | |
| "loss": 0.6093, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.00019759225874666373, | |
| "loss": 0.4626, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00019749420597095033, | |
| "loss": 0.5873, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00019739422160711147, | |
| "loss": 0.5864, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0001972923076360712, | |
| "loss": 0.5286, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00019718846607698343, | |
| "loss": 0.5755, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00019708269898719225, | |
| "loss": 0.53, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00019697500846219094, | |
| "loss": 0.5735, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 0.00019686539663558053, | |
| "loss": 0.5132, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00019675386567902757, | |
| "loss": 0.5416, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00019664041780222103, | |
| "loss": 0.6035, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00019652505525282848, | |
| "loss": 0.5411, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00019640778031645176, | |
| "loss": 0.5178, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0001962885953165815, | |
| "loss": 0.5137, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.0001961675026145511, | |
| "loss": 0.5723, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00019604450460949013, | |
| "loss": 0.5401, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.00019591960373827657, | |
| "loss": 0.5215, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00019579280247548865, | |
| "loss": 0.5151, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00019566410333335578, | |
| "loss": 0.5204, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00019553350886170883, | |
| "loss": 0.5436, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 0.00019540102164792956, | |
| "loss": 0.5291, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0001952666443168994, | |
| "loss": 0.5917, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0001951303795309474, | |
| "loss": 0.5848, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.0001949922299897975, | |
| "loss": 0.5621, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.00019485219843051502, | |
| "loss": 0.5872, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00019471028762745252, | |
| "loss": 0.5723, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00019456650039219474, | |
| "loss": 0.6229, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00019442083957350295, | |
| "loss": 0.5566, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00019427330805725843, | |
| "loss": 0.4698, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00019412390876640543, | |
| "loss": 0.52, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00019397264466089313, | |
| "loss": 0.4791, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00019381951873761707, | |
| "loss": 0.5702, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.0001936645340303597, | |
| "loss": 0.5061, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00019350769360973038, | |
| "loss": 0.5498, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.0001933490005831045, | |
| "loss": 0.4755, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00019318845809456186, | |
| "loss": 0.5354, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00019302606932482448, | |
| "loss": 0.5161, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00019286183749119346, | |
| "loss": 0.4965, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00019269576584748536, | |
| "loss": 0.5485, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0001925278576839676, | |
| "loss": 0.4787, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00019235811632729353, | |
| "loss": 0.5091, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00019218654514043608, | |
| "loss": 0.5315, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0001920131475226216, | |
| "loss": 0.5157, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00019183792690926222, | |
| "loss": 0.5745, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00019166088677188793, | |
| "loss": 0.5265, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00019148203061807766, | |
| "loss": 0.5473, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00019130136199138994, | |
| "loss": 0.4799, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00019111888447129256, | |
| "loss": 0.552, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.00019093460167309177, | |
| "loss": 0.4779, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.00019074851724786064, | |
| "loss": 0.5315, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00019056063488236661, | |
| "loss": 0.4807, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00019037095829899849, | |
| "loss": 0.5849, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00019017949125569285, | |
| "loss": 0.5343, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00018998623754585935, | |
| "loss": 0.5233, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00018979120099830578, | |
| "loss": 0.4892, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00018959438547716203, | |
| "loss": 0.5311, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.0001893957948818037, | |
| "loss": 0.5872, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00018919543314677463, | |
| "loss": 0.5303, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.00018899330424170926, | |
| "loss": 0.5294, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.00018878941217125367, | |
| "loss": 0.5178, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.00018858376097498637, | |
| "loss": 0.4638, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00018837635472733832, | |
| "loss": 0.5604, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00018816719753751213, | |
| "loss": 0.5109, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00018795629354940064, | |
| "loss": 0.5107, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00018774364694150488, | |
| "loss": 0.5133, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00018752926192685125, | |
| "loss": 0.4816, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.000187313142752908, | |
| "loss": 0.5037, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00018709529370150124, | |
| "loss": 0.5207, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0001868757190887299, | |
| "loss": 0.4827, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00018665442326488033, | |
| "loss": 0.5168, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.0001864314106143401, | |
| "loss": 0.5157, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.00018620668555551116, | |
| "loss": 0.5182, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00018598025254072225, | |
| "loss": 0.5141, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00018575211605614066, | |
| "loss": 0.4837, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00018552228062168355, | |
| "loss": 0.4921, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00018529075079092803, | |
| "loss": 0.4074, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.0001850575311510214, | |
| "loss": 0.5471, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.00018482262632258975, | |
| "loss": 0.552, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.000184586040959647, | |
| "loss": 0.4955, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00018434777974950218, | |
| "loss": 0.5257, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.0001841078474126668, | |
| "loss": 0.4816, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00018386624870276138, | |
| "loss": 0.474, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00018362298840642107, | |
| "loss": 0.4893, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.00018337807134320103, | |
| "loss": 0.4894, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00018313150236548082, | |
| "loss": 0.5137, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00018288328635836826, | |
| "loss": 0.4951, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00018263342823960269, | |
| "loss": 0.514, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.0001823819329594575, | |
| "loss": 0.5667, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00018212880550064214, | |
| "loss": 0.5274, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.0001818740508782032, | |
| "loss": 0.5278, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00018161767413942537, | |
| "loss": 0.4962, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.0001813596803637311, | |
| "loss": 0.4785, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.00018110007466258017, | |
| "loss": 0.5365, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0001808388621793684, | |
| "loss": 0.4773, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00018057604808932576, | |
| "loss": 0.5222, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00018031163759941362, | |
| "loss": 0.5292, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00018004563594822197, | |
| "loss": 0.5277, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.0001797780484058653, | |
| "loss": 0.4917, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00017950888027387833, | |
| "loss": 0.501, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.000179238136885111, | |
| "loss": 0.4991, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00017896582360362273, | |
| "loss": 0.4903, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00017869194582457616, | |
| "loss": 0.5216, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00017841650897413035, | |
| "loss": 0.5009, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.0001781395185093332, | |
| "loss": 0.56, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00017786097991801328, | |
| "loss": 0.5148, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00017758089871867117, | |
| "loss": 0.5133, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00017729928046037024, | |
| "loss": 0.5191, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00017701613072262644, | |
| "loss": 0.4322, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.000176731455115298, | |
| "loss": 0.5283, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.00017644525927847416, | |
| "loss": 0.5175, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00017615754888236347, | |
| "loss": 0.4731, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00017586832962718143, | |
| "loss": 0.4347, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.0001755776072430376, | |
| "loss": 0.5024, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00017528538748982198, | |
| "loss": 0.4423, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00017499167615709095, | |
| "loss": 0.4574, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.0001746964790639527, | |
| "loss": 0.4691, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00017439980205895158, | |
| "loss": 0.4838, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00017410165101995267, | |
| "loss": 0.4605, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00017380203185402503, | |
| "loss": 0.4571, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00017350095049732466, | |
| "loss": 0.5028, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.0001731984129149771, | |
| "loss": 0.4862, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00017289442510095906, | |
| "loss": 0.4735, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00017258899307797976, | |
| "loss": 0.4943, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00017228212289736153, | |
| "loss": 0.4617, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00017197382063892006, | |
| "loss": 0.4763, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.00017166409241084374, | |
| "loss": 0.4584, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00017135294434957282, | |
| "loss": 0.4857, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00017104038261967768, | |
| "loss": 0.4834, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00017072641341373692, | |
| "loss": 0.5171, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.00017041104295221438, | |
| "loss": 0.4695, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0001700942774833362, | |
| "loss": 0.5051, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00016977612328296674, | |
| "loss": 0.4571, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00016945658665448447, | |
| "loss": 0.467, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00016913567392865697, | |
| "loss": 0.5049, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00016881339146351544, | |
| "loss": 0.4694, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00016848974564422894, | |
| "loss": 0.4315, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00016816474288297773, | |
| "loss": 0.4595, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.0001678383896188262, | |
| "loss": 0.4894, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00016751069231759548, | |
| "loss": 0.471, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00016718165747173507, | |
| "loss": 0.4694, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00016685129160019444, | |
| "loss": 0.4806, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00016651960124829378, | |
| "loss": 0.4655, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00016618659298759436, | |
| "loss": 0.4996, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00016585227341576824, | |
| "loss": 0.4961, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00016551664915646767, | |
| "loss": 0.4674, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.0001651797268591938, | |
| "loss": 0.5274, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00016484151319916488, | |
| "loss": 0.447, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00016450201487718422, | |
| "loss": 0.4504, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00016416123861950714, | |
| "loss": 0.4656, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.0001638191911777079, | |
| "loss": 0.5016, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.000163475879328546, | |
| "loss": 0.4829, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00016313130987383157, | |
| "loss": 0.4662, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00016278548964029108, | |
| "loss": 0.4765, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.0001624384254794317, | |
| "loss": 0.4522, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00016209012426740582, | |
| "loss": 0.4363, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.0001617405929048746, | |
| "loss": 0.4888, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.00016138983831687142, | |
| "loss": 0.4398, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.0001610378674526646, | |
| "loss": 0.5025, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00016068468728561977, | |
| "loss": 0.4561, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.0001603303048130616, | |
| "loss": 0.4637, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.0001599747270561353, | |
| "loss": 0.5011, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00015961796105966745, | |
| "loss": 0.4757, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00015926001389202638, | |
| "loss": 0.4707, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.0001589008926449823, | |
| "loss": 0.4305, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00015854060443356658, | |
| "loss": 0.4506, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.0001581791563959309, | |
| "loss": 0.4594, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00015781655569320577, | |
| "loss": 0.4565, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00015745280950935883, | |
| "loss": 0.4848, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.00015708792505105223, | |
| "loss": 0.5282, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00015672190954750005, | |
| "loss": 0.4261, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00015635477025032504, | |
| "loss": 0.486, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00015598651443341491, | |
| "loss": 0.462, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00015561714939277822, | |
| "loss": 0.4475, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.00015524668244639993, | |
| "loss": 0.4765, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00015487512093409616, | |
| "loss": 0.4819, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.0001545024722173691, | |
| "loss": 0.4338, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00015412874367926093, | |
| "loss": 0.4556, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.00015375394272420763, | |
| "loss": 0.4704, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00015337807677789228, | |
| "loss": 0.4999, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.0001530011532870979, | |
| "loss": 0.4351, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.2333984375, | |
| "learning_rate": 0.00015262317971955993, | |
| "loss": 0.444, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.00015224416356381833, | |
| "loss": 0.4488, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00015186411232906912, | |
| "loss": 0.4704, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.0001514830335450157, | |
| "loss": 0.4406, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.0001511009347617196, | |
| "loss": 0.458, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00015071782354945094, | |
| "loss": 0.4432, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00015033370749853835, | |
| "loss": 0.4713, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00014994859421921876, | |
| "loss": 0.4574, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00014956249134148646, | |
| "loss": 0.4234, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00014917540651494205, | |
| "loss": 0.471, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.0001487873474086408, | |
| "loss": 0.4152, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00014839832171094074, | |
| "loss": 0.4751, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00014800833712935033, | |
| "loss": 0.5005, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00014761740139037578, | |
| "loss": 0.4411, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00014722552223936792, | |
| "loss": 0.4463, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.00014683270744036882, | |
| "loss": 0.4677, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.00014643896477595792, | |
| "loss": 0.4811, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.0001460443020470978, | |
| "loss": 0.4741, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.00014564872707297966, | |
| "loss": 0.4751, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.00014525224769086853, | |
| "loss": 0.4897, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0001448548717559477, | |
| "loss": 0.4843, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 0.00014445660714116344, | |
| "loss": 0.4625, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.0001440574617370687, | |
| "loss": 0.4661, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00014365744345166704, | |
| "loss": 0.4235, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.00014325656021025573, | |
| "loss": 0.4567, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00014285481995526892, | |
| "loss": 0.4371, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.0001424522306461202, | |
| "loss": 0.456, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.0001420488002590449, | |
| "loss": 0.4579, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.000141644536786942, | |
| "loss": 0.5085, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00014123944823921596, | |
| "loss": 0.4936, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 0.00014083354264161782, | |
| "loss": 0.4667, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00014042682803608637, | |
| "loss": 0.4848, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.0001400193124805886, | |
| "loss": 0.432, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.00013961100404896035, | |
| "loss": 0.4491, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.00013920191083074612, | |
| "loss": 0.4099, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.00013879204093103885, | |
| "loss": 0.4092, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.00013838140247031944, | |
| "loss": 0.4587, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00013797000358429563, | |
| "loss": 0.4563, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00013755785242374112, | |
| "loss": 0.5002, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.0001371449571543338, | |
| "loss": 0.4465, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00013673132595649418, | |
| "loss": 0.4596, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.0001363169670252232, | |
| "loss": 0.484, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00013590188856993992, | |
| "loss": 0.4489, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00013548609881431883, | |
| "loss": 0.4362, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.0001350696059961269, | |
| "loss": 0.4953, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00013465241836706056, | |
| "loss": 0.377, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 0.0001342345441925819, | |
| "loss": 0.3919, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00013381599175175517, | |
| "loss": 0.4668, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00013339676933708266, | |
| "loss": 0.4401, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.00013297688525434035, | |
| "loss": 0.4791, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00013255634782241349, | |
| "loss": 0.3973, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00013213516537313163, | |
| "loss": 0.4692, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.0001317133462511037, | |
| "loss": 0.456, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00013129089881355256, | |
| "loss": 0.4319, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00013086783143014942, | |
| "loss": 0.3933, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.0001304441524828482, | |
| "loss": 0.427, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.00013001987036571916, | |
| "loss": 0.4683, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00012959499348478293, | |
| "loss": 0.4522, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.00012916953025784368, | |
| "loss": 0.4255, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00012874348911432251, | |
| "loss": 0.4362, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00012831687849509043, | |
| "loss": 0.4532, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00012788970685230106, | |
| "loss": 0.4177, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 0.00012746198264922319, | |
| "loss": 0.4137, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00012703371436007313, | |
| "loss": 0.4552, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00012660491046984686, | |
| "loss": 0.4682, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00012617557947415179, | |
| "loss": 0.4947, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.0001257457298790386, | |
| "loss": 0.4205, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00012531537020083258, | |
| "loss": 0.4037, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.000124884508965965, | |
| "loss": 0.4267, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00012445315471080402, | |
| "loss": 0.3942, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.0001240213159814859, | |
| "loss": 0.4453, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.0001235890013337453, | |
| "loss": 0.4269, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00012315621933274597, | |
| "loss": 0.4226, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00012272297855291103, | |
| "loss": 0.452, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.0001222892875777531, | |
| "loss": 0.4421, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.00012185515499970421, | |
| "loss": 0.4633, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00012142058941994556, | |
| "loss": 0.4258, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00012098559944823714, | |
| "loss": 0.4136, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00012055019370274714, | |
| "loss": 0.3866, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.00012011438080988118, | |
| "loss": 0.4743, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00011967816940411145, | |
| "loss": 0.4307, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00011924156812780558, | |
| "loss": 0.4401, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.0001188045856310555, | |
| "loss": 0.4417, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.00011836723057150594, | |
| "loss": 0.4071, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00011792951161418301, | |
| "loss": 0.4008, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.00011749143743132249, | |
| "loss": 0.442, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.000117053016702198, | |
| "loss": 0.4082, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00011661425811294902, | |
| "loss": 0.4163, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00011617517035640888, | |
| "loss": 0.4438, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.00011573576213193245, | |
| "loss": 0.46, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00011529604214522385, | |
| "loss": 0.5053, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00011485601910816388, | |
| "loss": 0.4295, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00011441570173863756, | |
| "loss": 0.4254, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00011397509876036125, | |
| "loss": 0.4208, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00011353421890270992, | |
| "loss": 0.4441, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 0.00011309307090054415, | |
| "loss": 0.4299, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.00011265166349403713, | |
| "loss": 0.4684, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00011221000542850142, | |
| "loss": 0.4475, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00011176810545421572, | |
| "loss": 0.4259, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00011132597232625153, | |
| "loss": 0.4661, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00011088361480429965, | |
| "loss": 0.4305, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.00011044104165249665, | |
| "loss": 0.4598, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00010999826163925127, | |
| "loss": 0.4168, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00010955528353707059, | |
| "loss": 0.4045, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.00010911211612238632, | |
| "loss": 0.4082, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00010866876817538097, | |
| "loss": 0.4321, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0001082252484798137, | |
| "loss": 0.4071, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00010778156582284649, | |
| "loss": 0.4588, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00010733772899486991, | |
| "loss": 0.4414, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00010689374678932906, | |
| "loss": 0.393, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 0.00010644962800254932, | |
| "loss": 0.4724, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.00010600538143356201, | |
| "loss": 0.4428, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.00010556101588393015, | |
| "loss": 0.391, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00010511654015757407, | |
| "loss": 0.3976, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.00010467196306059687, | |
| "loss": 0.4012, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00010422729340111015, | |
| "loss": 0.468, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00010378253998905932, | |
| "loss": 0.431, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00010333771163604911, | |
| "loss": 0.4248, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.0001028928171551691, | |
| "loss": 0.389, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.0001024478653608189, | |
| "loss": 0.418, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.0001020028650685337, | |
| "loss": 0.4409, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00010155782509480954, | |
| "loss": 0.4182, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.00010111275425692863, | |
| "loss": 0.3616, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.0001006676613727847, | |
| "loss": 0.384, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00010022255526070823, | |
| "loss": 0.4133, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.97774447392918e-05, | |
| "loss": 0.4635, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 9.933233862721532e-05, | |
| "loss": 0.4389, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.888724574307139e-05, | |
| "loss": 0.4488, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 9.844217490519049e-05, | |
| "loss": 0.4116, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.799713493146633e-05, | |
| "loss": 0.3794, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.755213463918114e-05, | |
| "loss": 0.3887, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.710718284483094e-05, | |
| "loss": 0.4222, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 9.66622883639509e-05, | |
| "loss": 0.4451, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 9.621746001094072e-05, | |
| "loss": 0.4213, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.577270659888988e-05, | |
| "loss": 0.4378, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.532803693940315e-05, | |
| "loss": 0.4315, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 9.488345984242597e-05, | |
| "loss": 0.4118, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 9.443898411606988e-05, | |
| "loss": 0.4118, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.399461856643802e-05, | |
| "loss": 0.5094, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.355037199745072e-05, | |
| "loss": 0.391, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 9.310625321067096e-05, | |
| "loss": 0.3856, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.26622710051301e-05, | |
| "loss": 0.4223, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 9.221843417715352e-05, | |
| "loss": 0.3672, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 9.17747515201863e-05, | |
| "loss": 0.433, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 9.133123182461904e-05, | |
| "loss": 0.3983, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 9.088788387761366e-05, | |
| "loss": 0.4376, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 9.044471646292942e-05, | |
| "loss": 0.4773, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.000173836074874e-05, | |
| "loss": 0.461, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 8.955895834750334e-05, | |
| "loss": 0.4507, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 8.911638519570036e-05, | |
| "loss": 0.3797, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 8.867402767374848e-05, | |
| "loss": 0.3717, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 8.823189454578428e-05, | |
| "loss": 0.3888, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 8.778999457149863e-05, | |
| "loss": 0.4496, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 8.73483365059629e-05, | |
| "loss": 0.3736, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 8.69069290994559e-05, | |
| "loss": 0.4593, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 8.646578109729014e-05, | |
| "loss": 0.4097, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 8.60249012396388e-05, | |
| "loss": 0.4032, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 8.558429826136248e-05, | |
| "loss": 0.3813, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 8.514398089183616e-05, | |
| "loss": 0.4135, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 8.470395785477619e-05, | |
| "loss": 0.3782, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 8.426423786806756e-05, | |
| "loss": 0.4147, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 8.382482964359115e-05, | |
| "loss": 0.4261, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 8.338574188705101e-05, | |
| "loss": 0.4219, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 8.294698329780203e-05, | |
| "loss": 0.3911, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 8.250856256867753e-05, | |
| "loss": 0.376, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 8.207048838581701e-05, | |
| "loss": 0.3934, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 8.163276942849408e-05, | |
| "loss": 0.3958, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 8.119541436894452e-05, | |
| "loss": 0.396, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 8.075843187219443e-05, | |
| "loss": 0.4779, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 8.032183059588858e-05, | |
| "loss": 0.4172, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 7.988561919011884e-05, | |
| "loss": 0.3797, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 7.94498062972529e-05, | |
| "loss": 0.3962, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 7.901440055176287e-05, | |
| "loss": 0.3965, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 7.857941058005447e-05, | |
| "loss": 0.444, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.375, | |
| "learning_rate": 7.814484500029581e-05, | |
| "loss": 0.3988, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 7.771071242224693e-05, | |
| "loss": 0.4253, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 7.727702144708899e-05, | |
| "loss": 0.445, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 7.684378066725406e-05, | |
| "loss": 0.4292, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 7.641099866625472e-05, | |
| "loss": 0.4002, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 7.597868401851411e-05, | |
| "loss": 0.388, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 7.554684528919598e-05, | |
| "loss": 0.4417, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 7.511549103403505e-05, | |
| "loss": 0.3966, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 7.468462979916744e-05, | |
| "loss": 0.3563, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 7.425427012096142e-05, | |
| "loss": 0.3639, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 7.382442052584821e-05, | |
| "loss": 0.4633, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 7.339508953015316e-05, | |
| "loss": 0.3974, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 7.296628563992689e-05, | |
| "loss": 0.4446, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 7.253801735077684e-05, | |
| "loss": 0.3914, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 7.211029314769897e-05, | |
| "loss": 0.4249, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 7.168312150490959e-05, | |
| "loss": 0.41, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 7.125651088567751e-05, | |
| "loss": 0.4528, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 7.083046974215634e-05, | |
| "loss": 0.38, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 7.040500651521708e-05, | |
| "loss": 0.3793, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 6.998012963428084e-05, | |
| "loss": 0.3859, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 6.95558475171518e-05, | |
| "loss": 0.3929, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 6.913216856985056e-05, | |
| "loss": 0.3616, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 6.870910118644744e-05, | |
| "loss": 0.3599, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 6.828665374889629e-05, | |
| "loss": 0.4315, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 6.786483462686835e-05, | |
| "loss": 0.5134, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 6.744365217758651e-05, | |
| "loss": 0.3792, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 6.702311474565965e-05, | |
| "loss": 0.419, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 6.660323066291735e-05, | |
| "loss": 0.4255, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 6.618400824824482e-05, | |
| "loss": 0.4608, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 6.576545580741811e-05, | |
| "loss": 0.4734, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 6.534758163293944e-05, | |
| "loss": 0.4022, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.375, | |
| "learning_rate": 6.493039400387312e-05, | |
| "loss": 0.393, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 6.451390118568124e-05, | |
| "loss": 0.4285, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 6.409811143006012e-05, | |
| "loss": 0.3768, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 6.368303297477685e-05, | |
| "loss": 0.4258, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 6.326867404350587e-05, | |
| "loss": 0.4105, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 6.285504284566623e-05, | |
| "loss": 0.3375, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 6.244214757625891e-05, | |
| "loss": 0.4004, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 6.202999641570438e-05, | |
| "loss": 0.4723, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.375, | |
| "learning_rate": 6.161859752968059e-05, | |
| "loss": 0.3726, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 6.120795906896116e-05, | |
| "loss": 0.3782, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 6.079808916925391e-05, | |
| "loss": 0.3922, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 6.038899595103968e-05, | |
| "loss": 0.3924, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 5.9980687519411436e-05, | |
| "loss": 0.4182, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 5.95731719639137e-05, | |
| "loss": 0.483, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 5.91664573583822e-05, | |
| "loss": 0.4156, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 5.876055176078407e-05, | |
| "loss": 0.4438, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 5.8355463213058025e-05, | |
| "loss": 0.4392, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 5.795119974095514e-05, | |
| "loss": 0.4561, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 5.754776935387982e-05, | |
| "loss": 0.3761, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 5.714518004473111e-05, | |
| "loss": 0.3753, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 5.674343978974431e-05, | |
| "loss": 0.3991, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 5.634255654833297e-05, | |
| "loss": 0.3749, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 5.594253826293129e-05, | |
| "loss": 0.3925, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 5.554339285883656e-05, | |
| "loss": 0.3903, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 5.514512824405228e-05, | |
| "loss": 0.3997, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 5.474775230913147e-05, | |
| "loss": 0.4459, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 5.435127292702033e-05, | |
| "loss": 0.3924, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 5.395569795290221e-05, | |
| "loss": 0.4301, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 5.356103522404208e-05, | |
| "loss": 0.4262, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 5.316729255963118e-05, | |
| "loss": 0.4261, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 5.277447776063208e-05, | |
| "loss": 0.3991, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 5.238259860962423e-05, | |
| "loss": 0.3679, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 5.199166287064967e-05, | |
| "loss": 0.3751, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 5.160167828905926e-05, | |
| "loss": 0.4476, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 5.121265259135921e-05, | |
| "loss": 0.3787, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 5.0824593485057946e-05, | |
| "loss": 0.362, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 5.043750865851354e-05, | |
| "loss": 0.3852, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 5.005140578078127e-05, | |
| "loss": 0.3985, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 4.966629250146167e-05, | |
| "loss": 0.3807, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 4.928217645054909e-05, | |
| "loss": 0.3646, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 4.889906523828041e-05, | |
| "loss": 0.3996, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 4.851696645498429e-05, | |
| "loss": 0.4218, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 4.813588767093088e-05, | |
| "loss": 0.4394, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 4.7755836436181676e-05, | |
| "loss": 0.3549, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 4.737682028044007e-05, | |
| "loss": 0.4035, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 4.699884671290211e-05, | |
| "loss": 0.3727, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 4.662192322210771e-05, | |
| "loss": 0.3926, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 4.6246057275792356e-05, | |
| "loss": 0.3738, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 4.587125632073906e-05, | |
| "loss": 0.4018, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 4.54975277826309e-05, | |
| "loss": 0.4074, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 4.512487906590385e-05, | |
| "loss": 0.4087, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 4.4753317553600084e-05, | |
| "loss": 0.3621, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 4.438285060722176e-05, | |
| "loss": 0.4324, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 4.401348556658509e-05, | |
| "loss": 0.4352, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 4.3645229749674987e-05, | |
| "loss": 0.4221, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 4.3278090452499984e-05, | |
| "loss": 0.3938, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 4.291207494894781e-05, | |
| "loss": 0.4412, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 4.254719049064121e-05, | |
| "loss": 0.3785, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 4.218344430679426e-05, | |
| "loss": 0.4147, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 4.182084360406917e-05, | |
| "loss": 0.3998, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 4.145939556643348e-05, | |
| "loss": 0.3945, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 4.109910735501774e-05, | |
| "loss": 0.422, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 4.073998610797366e-05, | |
| "loss": 0.4348, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 4.0382038940332615e-05, | |
| "loss": 0.408, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 4.002527294386476e-05, | |
| "loss": 0.3963, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.966969518693845e-05, | |
| "loss": 0.3804, | |
| "step": 2885 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 3.931531271438027e-05, | |
| "loss": 0.4101, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 3.896213254733543e-05, | |
| "loss": 0.3588, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 3.861016168312861e-05, | |
| "loss": 0.3458, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 3.8259407095125446e-05, | |
| "loss": 0.4511, | |
| "step": 2905 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 3.790987573259422e-05, | |
| "loss": 0.44, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 3.756157452056832e-05, | |
| "loss": 0.4312, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 3.7214510359708934e-05, | |
| "loss": 0.4258, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 3.6868690126168425e-05, | |
| "loss": 0.4092, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 3.652412067145402e-05, | |
| "loss": 0.4247, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 3.6180808822292077e-05, | |
| "loss": 0.3827, | |
| "step": 2935 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 3.583876138049287e-05, | |
| "loss": 0.4062, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 3.549798512281579e-05, | |
| "loss": 0.392, | |
| "step": 2945 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 3.5158486800835114e-05, | |
| "loss": 0.3869, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 3.482027314080621e-05, | |
| "loss": 0.4215, | |
| "step": 2955 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 3.4483350843532335e-05, | |
| "loss": 0.3947, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 3.414772658423176e-05, | |
| "loss": 0.4174, | |
| "step": 2965 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 3.381340701240564e-05, | |
| "loss": 0.4834, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 3.348039875170622e-05, | |
| "loss": 0.3834, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 3.3148708399805574e-05, | |
| "loss": 0.4661, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 3.2818342528264944e-05, | |
| "loss": 0.3878, | |
| "step": 2985 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.375, | |
| "learning_rate": 3.248930768240454e-05, | |
| "loss": 0.4531, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 3.216161038117379e-05, | |
| "loss": 0.388, | |
| "step": 2995 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 3.1835257117022276e-05, | |
| "loss": 0.4092, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 3.151025435577106e-05, | |
| "loss": 0.4103, | |
| "step": 3005 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 3.118660853648457e-05, | |
| "loss": 0.4309, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 3.0864326071343064e-05, | |
| "loss": 0.361, | |
| "step": 3015 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 3.0543413345515526e-05, | |
| "loss": 0.3627, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 3.022387671703325e-05, | |
| "loss": 0.444, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 2.9905722516663803e-05, | |
| "loss": 0.44, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 2.9588957047785614e-05, | |
| "loss": 0.3949, | |
| "step": 3035 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 2.9273586586263103e-05, | |
| "loss": 0.4309, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 2.8959617380322336e-05, | |
| "loss": 0.4052, | |
| "step": 3045 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 2.8647055650427214e-05, | |
| "loss": 0.4693, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 2.8335907589156264e-05, | |
| "loss": 0.3854, | |
| "step": 3055 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 2.802617936107993e-05, | |
| "loss": 0.3751, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 2.7717877102638446e-05, | |
| "loss": 0.3518, | |
| "step": 3065 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.7411006922020244e-05, | |
| "loss": 0.3935, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 2.7105574899040942e-05, | |
| "loss": 0.3661, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 2.6801587085022906e-05, | |
| "loss": 0.3446, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 2.6499049502675344e-05, | |
| "loss": 0.4103, | |
| "step": 3085 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 2.6197968145974993e-05, | |
| "loss": 0.4104, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 2.589834898004735e-05, | |
| "loss": 0.3658, | |
| "step": 3095 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 2.5600197941048442e-05, | |
| "loss": 0.4091, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 2.5303520936047366e-05, | |
| "loss": 0.4013, | |
| "step": 3105 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 2.5008323842909075e-05, | |
| "loss": 0.3581, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 2.471461251017808e-05, | |
| "loss": 0.3613, | |
| "step": 3115 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 2.4422392756962452e-05, | |
| "loss": 0.3939, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 2.4131670372818603e-05, | |
| "loss": 0.402, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.3842451117636566e-05, | |
| "loss": 0.4436, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 2.355474072152588e-05, | |
| "loss": 0.3995, | |
| "step": 3135 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 2.3268544884702037e-05, | |
| "loss": 0.3673, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 2.2983869277373604e-05, | |
| "loss": 0.3658, | |
| "step": 3145 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 2.270071953962981e-05, | |
| "loss": 0.3789, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 2.2419101281328857e-05, | |
| "loss": 0.442, | |
| "step": 3155 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 2.2139020081986783e-05, | |
| "loss": 0.3913, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 2.186048149066684e-05, | |
| "loss": 0.3532, | |
| "step": 3165 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 2.158349102586964e-05, | |
| "loss": 0.4629, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 2.130805417542384e-05, | |
| "loss": 0.3224, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 2.103417639637729e-05, | |
| "loss": 0.4609, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 2.0761863114889002e-05, | |
| "loss": 0.4257, | |
| "step": 3185 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 2.0491119726121667e-05, | |
| "loss": 0.4391, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 2.0221951594134713e-05, | |
| "loss": 0.3992, | |
| "step": 3195 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.995436405177804e-05, | |
| "loss": 0.3755, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 1.968836240058638e-05, | |
| "loss": 0.3586, | |
| "step": 3205 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 1.9423951910674265e-05, | |
| "loss": 0.4388, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 1.916113782063159e-05, | |
| "loss": 0.422, | |
| "step": 3215 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 1.8899925337419844e-05, | |
| "loss": 0.4077, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 1.864031963626893e-05, | |
| "loss": 0.3461, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 1.8382325860574656e-05, | |
| "loss": 0.3394, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.375, | |
| "learning_rate": 1.8125949121796805e-05, | |
| "loss": 0.3918, | |
| "step": 3235 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 1.787119449935789e-05, | |
| "loss": 0.4278, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 1.7618067040542507e-05, | |
| "loss": 0.4128, | |
| "step": 3245 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 1.736657176039732e-05, | |
| "loss": 0.3997, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 1.7116713641631744e-05, | |
| "loss": 0.3856, | |
| "step": 3255 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 1.6868497634519185e-05, | |
| "loss": 0.3838, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.6621928656798967e-05, | |
| "loss": 0.4358, | |
| "step": 3265 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 1.637701159357895e-05, | |
| "loss": 0.4258, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 1.6133751297238654e-05, | |
| "loss": 0.4668, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 1.589215258733322e-05, | |
| "loss": 0.4056, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 1.565222025049785e-05, | |
| "loss": 0.3928, | |
| "step": 3285 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 1.541395904035301e-05, | |
| "loss": 0.3993, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 1.5177373677410245e-05, | |
| "loss": 0.4256, | |
| "step": 3295 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 1.4942468848978642e-05, | |
| "loss": 0.3931, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 1.4709249209071974e-05, | |
| "loss": 0.4088, | |
| "step": 3305 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 1.4477719378316469e-05, | |
| "loss": 0.376, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.5, | |
| "learning_rate": 1.424788394385933e-05, | |
| "loss": 0.4155, | |
| "step": 3315 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 1.4019747459277777e-05, | |
| "loss": 0.4336, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 1.379331444448886e-05, | |
| "loss": 0.4389, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.3568589385659914e-05, | |
| "loss": 0.3889, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.334557673511969e-05, | |
| "loss": 0.3689, | |
| "step": 3335 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 1.3124280911270114e-05, | |
| "loss": 0.4072, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 1.290470629849876e-05, | |
| "loss": 0.3905, | |
| "step": 3345 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 1.2686857247091987e-05, | |
| "loss": 0.3706, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 1.2470738073148769e-05, | |
| "loss": 0.3561, | |
| "step": 3355 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 1.225635305849513e-05, | |
| "loss": 0.3786, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 1.2043706450599391e-05, | |
| "loss": 0.3541, | |
| "step": 3365 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.1832802462487902e-05, | |
| "loss": 0.3837, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.1623645272661698e-05, | |
| "loss": 0.3629, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 1.1416239025013653e-05, | |
| "loss": 0.371, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.1210587828746355e-05, | |
| "loss": 0.3975, | |
| "step": 3385 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 1.1006695758290753e-05, | |
| "loss": 0.3682, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 1.0804566853225383e-05, | |
| "loss": 0.4048, | |
| "step": 3395 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.0604205118196342e-05, | |
| "loss": 0.3566, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 1.0405614522837992e-05, | |
| "loss": 0.3956, | |
| "step": 3405 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 1.0208799001694247e-05, | |
| "loss": 0.4019, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 1.0013762454140663e-05, | |
| "loss": 0.4616, | |
| "step": 3415 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.820508744307177e-06, | |
| "loss": 0.3845, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 9.629041701001507e-06, | |
| "loss": 0.4296, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.439365117633404e-06, | |
| "loss": 0.3861, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 9.251482752139352e-06, | |
| "loss": 0.3431, | |
| "step": 3435 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 9.065398326908215e-06, | |
| "loss": 0.3941, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 8.881115528707462e-06, | |
| "loss": 0.3596, | |
| "step": 3445 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 8.698638008610094e-06, | |
| "loss": 0.4148, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 8.517969381922364e-06, | |
| "loss": 0.3717, | |
| "step": 3455 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 8.339113228112083e-06, | |
| "loss": 0.4585, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 8.162073090737776e-06, | |
| "loss": 0.4115, | |
| "step": 3465 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 7.986852477378404e-06, | |
| "loss": 0.3699, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 7.813454859563929e-06, | |
| "loss": 0.4215, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 7.641883672706496e-06, | |
| "loss": 0.3779, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.375, | |
| "learning_rate": 7.472142316032382e-06, | |
| "loss": 0.3675, | |
| "step": 3485 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 7.304234152514655e-06, | |
| "loss": 0.4172, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 7.138162508806556e-06, | |
| "loss": 0.423, | |
| "step": 3495 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 6.973930675175533e-06, | |
| "loss": 0.368, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 6.811541905438145e-06, | |
| "loss": 0.4112, | |
| "step": 3505 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 6.6509994168955204e-06, | |
| "loss": 0.4059, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 6.492306390269642e-06, | |
| "loss": 0.3527, | |
| "step": 3515 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 6.335465969640331e-06, | |
| "loss": 0.3475, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 6.180481262382964e-06, | |
| "loss": 0.3912, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 6.027355339106877e-06, | |
| "loss": 0.428, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 5.876091233594572e-06, | |
| "loss": 0.4253, | |
| "step": 3535 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 5.726691942741569e-06, | |
| "loss": 0.4701, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 5.579160426497065e-06, | |
| "loss": 0.3716, | |
| "step": 3545 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 5.4334996078052705e-06, | |
| "loss": 0.3864, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 5.289712372547495e-06, | |
| "loss": 0.4586, | |
| "step": 3555 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 5.147801569485e-06, | |
| "loss": 0.4011, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 5.007770010202528e-06, | |
| "loss": 0.3688, | |
| "step": 3565 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 4.8696204690526155e-06, | |
| "loss": 0.3555, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 4.733355683100604e-06, | |
| "loss": 0.3783, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 4.598978352070437e-06, | |
| "loss": 0.386, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 4.466491138291185e-06, | |
| "loss": 0.3876, | |
| "step": 3585 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 4.3358966666442455e-06, | |
| "loss": 0.4083, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 4.207197524511375e-06, | |
| "loss": 0.3974, | |
| "step": 3595 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 4.0803962617234405e-06, | |
| "loss": 0.3811, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 3.955495390509878e-06, | |
| "loss": 0.3775, | |
| "step": 3605 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 3.8324973854489074e-06, | |
| "loss": 0.446, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 3.71140468341854e-06, | |
| "loss": 0.4419, | |
| "step": 3615 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 3.592219683548259e-06, | |
| "loss": 0.3783, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 3.474944747171538e-06, | |
| "loss": 0.4398, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.35958219777901e-06, | |
| "loss": 0.4204, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 3.246134320972438e-06, | |
| "loss": 0.4019, | |
| "step": 3635 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 3.134603364419475e-06, | |
| "loss": 0.3955, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 3.0249915378090854e-06, | |
| "loss": 0.414, | |
| "step": 3645 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 2.9173010128077826e-06, | |
| "loss": 0.3865, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 2.811533923016607e-06, | |
| "loss": 0.4735, | |
| "step": 3655 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 2.7076923639288486e-06, | |
| "loss": 0.3737, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 2.605778392888547e-06, | |
| "loss": 0.3935, | |
| "step": 3665 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.505794029049702e-06, | |
| "loss": 0.3963, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.4077412533362707e-06, | |
| "loss": 0.4122, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 2.3116220084029694e-06, | |
| "loss": 0.3266, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.217438198596733e-06, | |
| "loss": 0.4134, | |
| "step": 3685 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 2.125191689918993e-06, | |
| "loss": 0.4304, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 2.034884309988727e-06, | |
| "loss": 0.4402, | |
| "step": 3695 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 1.9465178480062353e-06, | |
| "loss": 0.4104, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 1.860094054717687e-06, | |
| "loss": 0.4293, | |
| "step": 3705 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 1.7756146423804742e-06, | |
| "loss": 0.317, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 1.6930812847292255e-06, | |
| "loss": 0.3738, | |
| "step": 3715 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.6124956169426886e-06, | |
| "loss": 0.4004, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 1.5338592356113124e-06, | |
| "loss": 0.4346, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 1.4571736987056495e-06, | |
| "loss": 0.4255, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 1.3824405255454475e-06, | |
| "loss": 0.3787, | |
| "step": 3735 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 1.3096611967695517e-06, | |
| "loss": 0.3445, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 1.2388371543066045e-06, | |
| "loss": 0.4093, | |
| "step": 3745 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 1.1699698013464488e-06, | |
| "loss": 0.3977, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 1.1030605023123364e-06, | |
| "loss": 0.373, | |
| "step": 3755 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 1.0381105828338844e-06, | |
| "loss": 0.3833, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 9.751213297208184e-07, | |
| "loss": 0.434, | |
| "step": 3765 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.140939909375034e-07, | |
| "loss": 0.3993, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 8.550297755781644e-07, | |
| "loss": 0.3842, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 7.979298538429935e-07, | |
| "loss": 0.462, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 7.427953570149027e-07, | |
| "loss": 0.3928, | |
| "step": 3785 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 6.896273774371742e-07, | |
| "loss": 0.3982, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 6.384269684917676e-07, | |
| "loss": 0.4198, | |
| "step": 3795 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 5.891951445784583e-07, | |
| "loss": 0.3316, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 5.419328810947755e-07, | |
| "loss": 0.4272, | |
| "step": 3805 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 4.966411144166405e-07, | |
| "loss": 0.3763, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 4.533207418798147e-07, | |
| "loss": 0.3747, | |
| "step": 3815 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 4.119726217621578e-07, | |
| "loss": 0.4468, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 3.7259757326657543e-07, | |
| "loss": 0.3989, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 3.351963765048427e-07, | |
| "loss": 0.3995, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 2.997697724820725e-07, | |
| "loss": 0.4069, | |
| "step": 3835 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 2.663184630821158e-07, | |
| "loss": 0.397, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 2.3484311105360645e-07, | |
| "loss": 0.383, | |
| "step": 3845 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 2.053443399968602e-07, | |
| "loss": 0.3835, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 1.778227343514627e-07, | |
| "loss": 0.4508, | |
| "step": 3855 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 1.522788393847785e-07, | |
| "loss": 0.4136, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.375, | |
| "learning_rate": 1.287131611810599e-07, | |
| "loss": 0.3933, | |
| "step": 3865 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 1.0712616663149932e-07, | |
| "loss": 0.3702, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 8.751828342491442e-08, | |
| "loss": 0.3742, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 6.988990003929941e-08, | |
| "loss": 0.4224, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 5.4241365734153396e-08, | |
| "loss": 0.4584, | |
| "step": 3885 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 4.057299054350816e-08, | |
| "loss": 0.4101, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 2.8885045269833044e-08, | |
| "loss": 0.4152, | |
| "step": 3895 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 1.9177761478639255e-08, | |
| "loss": 0.4205, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 1.1451331493916862e-08, | |
| "loss": 0.3096, | |
| "step": 3905 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 5.7059083943045154e-09, | |
| "loss": 0.3654, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 1.941606010069652e-09, | |
| "loss": 0.4881, | |
| "step": 3915 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 1.5849892083252828e-10, | |
| "loss": 0.406, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.39284008741378784, | |
| "eval_runtime": 450.3981, | |
| "eval_samples_per_second": 3.71, | |
| "eval_steps_per_second": 0.464, | |
| "step": 3922 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 3922, | |
| "total_flos": 2.757410325375484e+18, | |
| "train_loss": 0.47386226463779874, | |
| "train_runtime": 32073.4099, | |
| "train_samples_per_second": 0.978, | |
| "train_steps_per_second": 0.122 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 3922, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 2.757410325375484e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |