| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.99922546665634, | |
| "eval_steps": 500, | |
| "global_step": 64550, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 0.00029699999999999996, | |
| "loss": 13.8243, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 0.00029768149882903976, | |
| "loss": 3.4778, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 0.0002953395784543325, | |
| "loss": 1.7573, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 0.00029299765807962527, | |
| "loss": 1.4528, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 0.000290655737704918, | |
| "loss": 1.325, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 0.0002883138173302108, | |
| "loss": 1.2209, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 0.0002859718969555035, | |
| "loss": 1.1738, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 0.00028362997658079624, | |
| "loss": 1.1056, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 0.0002812927400468384, | |
| "loss": 1.0772, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 0.00027895550351288053, | |
| "loss": 1.0548, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 0.00027661358313817326, | |
| "loss": 1.0193, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 0.00027427166276346605, | |
| "loss": 1.0066, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 0.0002719297423887588, | |
| "loss": 0.9963, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 0.0002695878220140515, | |
| "loss": 0.9402, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 0.0002672505854800937, | |
| "loss": 0.9338, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 0.0002649086651053864, | |
| "loss": 0.9176, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 0.00026256674473067914, | |
| "loss": 0.9084, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 0.00026022482435597187, | |
| "loss": 0.8971, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 0.00025788758782201404, | |
| "loss": 0.9126, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 0.00025554566744730677, | |
| "loss": 0.8965, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 0.0002532037470725995, | |
| "loss": 0.891, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 0.0002508618266978922, | |
| "loss": 0.8817, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 0.00024851990632318496, | |
| "loss": 0.8858, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 0.00024617798594847774, | |
| "loss": 0.8546, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 0.00024383606557377047, | |
| "loss": 0.849, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 0.00024149414519906322, | |
| "loss": 0.8345, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 0.00023915690866510534, | |
| "loss": 0.7928, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 0.00023681498829039813, | |
| "loss": 0.7886, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 0.00023447306791569086, | |
| "loss": 0.7955, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 0.00023213114754098358, | |
| "loss": 0.7896, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 0.00022979391100702576, | |
| "loss": 0.7932, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 0.0002274519906323185, | |
| "loss": 0.7742, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 0.00022511007025761124, | |
| "loss": 0.7631, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "learning_rate": 0.00022276814988290397, | |
| "loss": 0.7689, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "learning_rate": 0.0002204262295081967, | |
| "loss": 0.7624, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 0.00021808430913348943, | |
| "loss": 0.7571, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 0.0002157423887587822, | |
| "loss": 0.7493, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "learning_rate": 0.00021340046838407492, | |
| "loss": 0.7466, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 0.00021105854800936765, | |
| "loss": 0.7267, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "learning_rate": 0.0002087166276346604, | |
| "loss": 0.6861, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "learning_rate": 0.00020637939110070255, | |
| "loss": 0.6996, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "learning_rate": 0.00020403747072599528, | |
| "loss": 0.6981, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "learning_rate": 0.00020169555035128803, | |
| "loss": 0.6879, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 0.00019935362997658076, | |
| "loss": 0.6903, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "learning_rate": 0.00019701170960187355, | |
| "loss": 0.6872, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "learning_rate": 0.00019466978922716628, | |
| "loss": 0.691, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "learning_rate": 0.000192327868852459, | |
| "loss": 0.6859, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "learning_rate": 0.00018998594847775176, | |
| "loss": 0.6735, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "learning_rate": 0.0001876440281030445, | |
| "loss": 0.6719, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "learning_rate": 0.00018530210772833722, | |
| "loss": 0.6697, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "learning_rate": 0.00018296018735362995, | |
| "loss": 0.6702, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "learning_rate": 0.0001806182669789227, | |
| "loss": 0.6591, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "learning_rate": 0.00017828103044496485, | |
| "loss": 0.6266, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "learning_rate": 0.0001759391100702576, | |
| "loss": 0.6277, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "learning_rate": 0.00017359718969555034, | |
| "loss": 0.6308, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "learning_rate": 0.00017125995316159248, | |
| "loss": 0.6219, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "learning_rate": 0.00016891803278688524, | |
| "loss": 0.6252, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "learning_rate": 0.00016657611241217797, | |
| "loss": 0.6199, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "learning_rate": 0.0001642341920374707, | |
| "loss": 0.6253, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "learning_rate": 0.00016189227166276345, | |
| "loss": 0.6301, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "learning_rate": 0.00015955035128805618, | |
| "loss": 0.6084, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 0.00015720843091334894, | |
| "loss": 0.6077, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "learning_rate": 0.0001548665105386417, | |
| "loss": 0.609, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "learning_rate": 0.00015252927400468384, | |
| "loss": 0.5997, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "learning_rate": 0.00015018735362997657, | |
| "loss": 0.5903, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "learning_rate": 0.0001478454332552693, | |
| "loss": 0.5746, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 5.19, | |
| "learning_rate": 0.00014550819672131147, | |
| "loss": 0.5701, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 5.27, | |
| "learning_rate": 0.0001431662763466042, | |
| "loss": 0.5719, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 5.34, | |
| "learning_rate": 0.00014082435597189693, | |
| "loss": 0.5685, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 5.42, | |
| "learning_rate": 0.0001384824355971897, | |
| "loss": 0.561, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "learning_rate": 0.00013614051522248242, | |
| "loss": 0.5649, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 5.58, | |
| "learning_rate": 0.00013379859484777517, | |
| "loss": 0.5616, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 5.65, | |
| "learning_rate": 0.0001314566744730679, | |
| "loss": 0.562, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "learning_rate": 0.00012911475409836063, | |
| "loss": 0.5695, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 5.81, | |
| "learning_rate": 0.0001267728337236534, | |
| "loss": 0.5601, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 5.89, | |
| "learning_rate": 0.00012443091334894614, | |
| "loss": 0.5588, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 5.96, | |
| "learning_rate": 0.00012208899297423887, | |
| "loss": 0.5518, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 6.04, | |
| "learning_rate": 0.00011975175644028103, | |
| "loss": 0.5397, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "learning_rate": 0.00011741451990632316, | |
| "loss": 0.5217, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "learning_rate": 0.00011507728337236532, | |
| "loss": 0.5257, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 6.27, | |
| "learning_rate": 0.00011273536299765807, | |
| "loss": 0.5198, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 6.35, | |
| "learning_rate": 0.00011039344262295082, | |
| "loss": 0.5202, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "learning_rate": 0.00010805152224824355, | |
| "loss": 0.5175, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 6.51, | |
| "learning_rate": 0.0001057096018735363, | |
| "loss": 0.5165, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 6.58, | |
| "learning_rate": 0.00010336768149882904, | |
| "loss": 0.5189, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 6.66, | |
| "learning_rate": 0.00010102576112412177, | |
| "loss": 0.5185, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 6.74, | |
| "learning_rate": 9.868384074941451e-05, | |
| "loss": 0.5169, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "learning_rate": 9.634192037470724e-05, | |
| "loss": 0.5042, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 6.89, | |
| "learning_rate": 9.4e-05, | |
| "loss": 0.5121, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 6.97, | |
| "learning_rate": 9.165807962529274e-05, | |
| "loss": 0.5015, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 7.05, | |
| "learning_rate": 8.931615925058547e-05, | |
| "loss": 0.4854, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 7.13, | |
| "learning_rate": 8.697423887587821e-05, | |
| "loss": 0.474, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 8.463231850117095e-05, | |
| "loss": 0.4757, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 7.28, | |
| "learning_rate": 8.229039812646368e-05, | |
| "loss": 0.4804, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 7.36, | |
| "learning_rate": 7.995316159250584e-05, | |
| "loss": 0.4678, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 7.44, | |
| "learning_rate": 7.761124121779858e-05, | |
| "loss": 0.4692, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 7.51, | |
| "learning_rate": 7.527400468384074e-05, | |
| "loss": 0.4687, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 7.59, | |
| "learning_rate": 7.293208430913349e-05, | |
| "loss": 0.4704, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 7.67, | |
| "learning_rate": 7.059016393442623e-05, | |
| "loss": 0.4645, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 7.75, | |
| "learning_rate": 6.824824355971896e-05, | |
| "loss": 0.469, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 7.82, | |
| "learning_rate": 6.59063231850117e-05, | |
| "loss": 0.4712, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "learning_rate": 6.356440281030444e-05, | |
| "loss": 0.4652, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 7.98, | |
| "learning_rate": 6.122248243559719e-05, | |
| "loss": 0.4652, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 8.06, | |
| "learning_rate": 5.888056206088993e-05, | |
| "loss": 0.4445, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 8.13, | |
| "learning_rate": 5.6538641686182664e-05, | |
| "loss": 0.4403, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 8.21, | |
| "learning_rate": 5.4201405152224824e-05, | |
| "loss": 0.4421, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 8.29, | |
| "learning_rate": 5.185948477751756e-05, | |
| "loss": 0.4427, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 8.36, | |
| "learning_rate": 4.9517564402810296e-05, | |
| "loss": 0.4346, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 8.44, | |
| "learning_rate": 4.7175644028103045e-05, | |
| "loss": 0.4388, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 8.52, | |
| "learning_rate": 4.483372365339578e-05, | |
| "loss": 0.4294, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "learning_rate": 4.249180327868852e-05, | |
| "loss": 0.4298, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 8.67, | |
| "learning_rate": 4.014988290398127e-05, | |
| "loss": 0.431, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 8.75, | |
| "learning_rate": 3.781264637002341e-05, | |
| "loss": 0.4356, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 8.83, | |
| "learning_rate": 3.547540983606557e-05, | |
| "loss": 0.4311, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 8.91, | |
| "learning_rate": 3.313348946135831e-05, | |
| "loss": 0.4316, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 8.98, | |
| "learning_rate": 3.079156908665105e-05, | |
| "loss": 0.4254, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 9.06, | |
| "learning_rate": 2.8449648711943793e-05, | |
| "loss": 0.4119, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 9.14, | |
| "learning_rate": 2.610772833723653e-05, | |
| "loss": 0.4092, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 9.22, | |
| "learning_rate": 2.3765807962529272e-05, | |
| "loss": 0.4139, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 9.29, | |
| "learning_rate": 2.1423887587822015e-05, | |
| "loss": 0.4071, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 9.37, | |
| "learning_rate": 1.9086651053864168e-05, | |
| "loss": 0.4101, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 9.45, | |
| "learning_rate": 1.6744730679156907e-05, | |
| "loss": 0.4053, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 9.53, | |
| "learning_rate": 1.4402810304449646e-05, | |
| "loss": 0.4103, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 1.2060889929742389e-05, | |
| "loss": 0.4061, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 9.68, | |
| "learning_rate": 9.718969555035128e-06, | |
| "loss": 0.4024, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 9.76, | |
| "learning_rate": 7.377049180327868e-06, | |
| "loss": 0.3976, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 9.84, | |
| "learning_rate": 5.035128805620609e-06, | |
| "loss": 0.4042, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 9.91, | |
| "learning_rate": 2.6932084309133486e-06, | |
| "loss": 0.4074, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 9.99, | |
| "learning_rate": 3.5128805620608895e-07, | |
| "loss": 0.4046, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 64550, | |
| "total_flos": 2.6998692478577122e+20, | |
| "train_loss": 0.7716339270593952, | |
| "train_runtime": 45884.6476, | |
| "train_samples_per_second": 45.019, | |
| "train_steps_per_second": 1.407 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 64550, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 2.6998692478577122e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |