diff --git "a/train_job_output.txt" "b/train_job_output.txt"
--- "a/train_job_output.txt"
+++ "b/train_job_output.txt"
@@ -391,4 +391,65 @@ Retrying in 1s [Retry 1/5].
 
 71%|███████   | 7575/10682 [1:04:22<25:35,  2.02it/s]
 71%|███████   | 7576/10682 [1:04:23<25:37,  2.02it/s]
 71%|███████   | 7577/10682 [1:04:23<25:35,  2.02it/s]
 71%|███████   | 7578/10682 [1:04:24<25:35,  2.02it/s]
 71%|███████   | 7579/10682 [1:04:24<25:34,  2.02it/s]
 71%|███████   | 7580/10682 [1:04:25<25:34,  2.02it/s]
 71%|███████   | 7581/10682 [1:04:25<25:32,  2.02it/s]
 71%|███████   | 7582/10682 [1:04:26<25:31,  2.02it/s]
 71%|███████   | 7583/10682 [1:04:26<25:31,  2.02it/s]
 71%|███████   | 7584/10682 [1:04:27<25:30,  2.02it/s]
 71%|███████   | 7585/10682 [1:04:27<25:30,  2.02it/s]
 71%|███████   | 7586/10682 [1:04:28<25:29,  2.02it/s]
 71%|███████   | 7587/10682 [1:04:28<25:31,  2.02it/s]
 71%|███████   | 7588/10682 [1:04:29<25:28,  2.02it/s]
 71%|███████   | 7589/10682 [1:04:29<25:29,  2.02it/s]
 71%|███████   | 7590/10682 [1:04:30<25:27,  2.02it/s]
 71%|███████   | 7591/10682 [1:04:30<25:27,  2.02it/s]
 71%|███████   | 7592/10682 [1:04:31<25:27,  2.02it/s]
 71%|███████   | 7593/10682 [1:04:31<25:27,  2.02it/s]
 71%|███████   | 7594/10682 [1:04:32<25:28,  2.02it/s]
 71%|███████   | 7595/10682 [1:04:32<25:27,  2.02it/s]
 71%|███████   | 7596/10682 [1:04:33<25:25,  2.02it/s]
 71%|███████   | 7597/10682 [1:04:33<25:26,  2.02it/s]
 71%|███████   | 7598/10682 [1:04:34<25:23,  2.02it/s]
 71%|███████   | 7599/10682 [1:04:34<25:24,  2.02it/s]
 71%|███████   | 7600/10682 [1:04:35<25:22,  2.02it/s]
                                                      
{'loss': 2.9468, 'grad_norm': 0.27557408809661865, 'learning_rate': 0.00023289281143540065, 'epoch': 9.95}
 
 71%|███████   | 7600/10682 [1:04:35<25:22,  2.02it/s]
 71%|███████   | 7601/10682 [1:04:35<25:25,  2.02it/s]
 71%|███████   | 7602/10682 [1:04:36<25:23,  2.02it/s]
 71%|███████   | 7603/10682 [1:04:36<25:22,  2.02it/s]
 71%|███████   | 7604/10682 [1:04:37<25:20,  2.02it/s]
 71%|███████   | 7605/10682 [1:04:37<25:21,  2.02it/s]
 71%|███████   | 7606/10682 [1:04:38<25:19,  2.02it/s]
 71%|███████   | 7607/10682 [1:04:38<25:20,  2.02it/s]
 71%|███████   | 7608/10682 [1:04:39<25:19,  2.02it/s]
 71%|███████   | 7609/10682 [1:04:39<25:18,  2.02it/s]
 71%|███████   | 7610/10682 [1:04:40<25:17,  2.02it/s]
 71%|███████▏  | 7611/10682 [1:04:40<25:18,  2.02it/s]
 71%|███████▏  | 7612/10682 [1:04:41<25:17,  2.02it/s]
 71%|███████▏  | 7613/10682 [1:04:41<25:17,  2.02it/s]
 71%|███████▏  | 7614/10682 [1:04:42<25:16,  2.02it/s]
 71%|███████▏  | 7615/10682 [1:04:42<25:18,  2.02it/s]
 71%|███████▏  | 7616/10682 [1:04:43<25:15,  2.02it/s]
 71%|███████▏  | 7617/10682 [1:04:43<25:15,  2.02it/s]
 71%|███████▏  | 7618/10682 [1:04:44<25:14,  2.02it/s]
 71%|███████▏  | 7619/10682 [1:04:44<25:13,  2.02it/s]
 71%|███████▏  | 7620/10682 [1:04:45<25:12,  2.03it/s]
 71%|███████▏  | 7621/10682 [1:04:45<25:12,  2.02it/s]
 71%|███████▏  | 7622/10682 [1:04:46<25:10,  2.03it/s]
 71%|███████▏  | 7623/10682 [1:04:46<25:11,  2.02it/s]
 71%|███████▏  | 7624/10682 [1:04:47<25:09,  2.03it/s]
 71%|███████▏  | 7625/10682 [1:04:47<25:08,  2.03it/s]{'loss': 2.9439, 'grad_norm': 0.2665178179740906, 'learning_rate': 0.00022944844326774121, 'epoch': 9.99}
                                                      
 
 71%|███████▏  | 7625/10682 [1:04:47<25:08,  2.03it/s]
 71%|███████▏  | 7626/10682 [1:04:48<25:09,  2.02it/s]
 71%|███████▏  | 7627/10682 [1:04:48<25:09,  2.02it/s]
 71%|███████▏  | 7628/10682 [1:04:49<25:09,  2.02it/s]
 71%|███████▏  | 7629/10682 [1:04:49<25:07,  2.02it/s]
 71%|███████▏  | 7630/10682 [1:04:50<25:08,  2.02it/s]
 71%|███████▏  | 7631/10682 [1:04:50<25:07,  2.02it/s]
 71%|███████▏  | 7632/10682 [1:04:51<25:07,  2.02it/s]
 71%|███████▏  | 7633/10682 [1:04:51<25:05,  2.02it/s]
 71%|███████▏  | 7634/10682 [1:04:52<25:05,  2.02it/s]
 71%|███████▏  | 7635/10682 [1:04:52<24:48,  2.05it/s]
 71%|███████▏  | 7636/10682 [1:05:04<3:21:36,  3.97s/it]
 71%|███████▏  | 7637/10682 [1:05:05<2:28:36,  2.93s/it]
 72%|███████▏  | 7638/10682 [1:05:05<1:51:28,  2.20s/it]
 72%|███████▏  | 7639/10682 [1:05:06<1:25:31,  1.69s/it]
 72%|███████▏  | 7640/10682 [1:05:06<1:07:26,  1.33s/it]
 72%|███████▏  | 7641/10682 [1:05:07<54:50,  1.08s/it]  
 72%|███████▏  | 7642/10682 [1:05:07<45:52,  1.10it/s]
 72%|███████▏  | 7643/10682 [1:05:08<39:41,  1.28it/s]
 72%|███████▏  | 7644/10682 [1:05:08<35:15,  1.44it/s]
 72%|███████▏  | 7645/10682 [1:05:09<32:09,  1.57it/s]
 72%|███████▏  | 7646/10682 [1:05:09<30:00,  1.69it/s]
 72%|███████▏  | 7647/10682 [1:05:10<28:28,  1.78it/s]
 72%|███████▏  | 7648/10682 [1:05:10<27:25,  1.84it/s]
 72%|███████▏  | 7649/10682 [1:05:11<26:42,  1.89it/s]
 72%|███████▏  | 7650/10682 [1:05:11<26:10,  1.93it/s]
                                                      
{'loss': 2.8611, 'grad_norm': 0.281035840511322, 'learning_rate': 0.00022602213475715589, 'epoch': 10.02}
-
 72%|███████▏  | 7650/10682 [1:05:11<26:10,  1.93it/s]
 72%|███████▏  | 7651/10682 [1:05:12<25:49,  1.96it/s]
 72%|███████▏  | 7652/10682 [1:05:12<25:34,  1.97it/s]
 72%|███████▏  | 7653/10682 [1:05:13<27:27,  1.84it/s]
 72%|███████▏  | 7654/10682 [1:05:13<26:42,  1.89it/s]
 72%|███████▏  | 7655/10682 [1:05:14<26:13,  1.92it/s]
 72%|███████▏  | 7656/10682 [1:05:14<25:50,  1.95it/s]
 72%|███████▏  | 7657/10682 [1:05:15<25:33,  1.97it/s]
 72%|███████▏  | 7658/10682 [1:05:15<25:24,  1.98it/s]
 72%|███████▏  | 7659/10682 [1:05:16<25:13,  2.00it/s]
\ No newline at end of file
+
 72%|███████▏  | 7650/10682 [1:05:11<26:10,  1.93it/s]
 72%|███████▏  | 7651/10682 [1:05:12<25:49,  1.96it/s]
 72%|███████▏  | 7652/10682 [1:05:12<25:34,  1.97it/s]
 72%|███████▏  | 7653/10682 [1:05:13<27:27,  1.84it/s]
 72%|███████▏  | 7654/10682 [1:05:13<26:42,  1.89it/s]
 72%|███████▏  | 7655/10682 [1:05:14<26:13,  1.92it/s]
 72%|███████▏  | 7656/10682 [1:05:14<25:50,  1.95it/s]
 72%|███████▏  | 7657/10682 [1:05:15<25:33,  1.97it/s]
 72%|███████▏  | 7658/10682 [1:05:15<25:24,  1.98it/s]
 72%|███████▏  | 7659/10682 [1:05:16<25:13,  2.00it/s]
 72%|███████▏  | 7660/10682 [1:05:16<25:15,  1.99it/s]
 72%|███████▏  | 7661/10682 [1:05:17<25:06,  2.01it/s]
 72%|███████▏  | 7662/10682 [1:05:17<25:01,  2.01it/s]
 72%|███████▏  | 7663/10682 [1:05:18<24:57,  2.02it/s]
 72%|███████▏  | 7664/10682 [1:05:18<24:54,  2.02it/s]
 72%|███████▏  | 7665/10682 [1:05:19<24:51,  2.02it/s]
 72%|███████▏  | 7666/10682 [1:05:19<24:48,  2.03it/s]
 72%|███████▏  | 7667/10682 [1:05:20<24:49,  2.02it/s]
 72%|███████▏  | 7668/10682 [1:05:20<24:47,  2.03it/s]
 72%|███████▏  | 7669/10682 [1:05:21<24:48,  2.02it/s]
 72%|███████▏  | 7670/10682 [1:05:21<24:46,  2.03it/s]
 72%|███████▏  | 7671/10682 [1:05:22<24:44,  2.03it/s]
 72%|███████▏  | 7672/10682 [1:05:22<24:44,  2.03it/s]
 72%|███████▏  | 7673/10682 [1:05:23<24:42,  2.03it/s]
 72%|███████▏  | 7674/10682 [1:05:23<24:43,  2.03it/s]
 72%|███████▏  | 7675/10682 [1:05:24<24:41,  2.03it/s]
                                                      
{'loss': 2.8005, 'grad_norm': 0.27186882495880127, 'learning_rate': 0.0002226141146140523, 'epoch': 10.05}
+
 72%|███████▏  | 7675/10682 [1:05:24<24:41,  2.03it/s]
 72%|███████▏  | 7676/10682 [1:05:24<24:44,  2.03it/s]
 72%|███████▏  | 7677/10682 [1:05:25<24:42,  2.03it/s]
 72%|███████▏  | 7678/10682 [1:05:25<24:43,  2.03it/s]
 72%|███████▏  | 7679/10682 [1:05:26<24:42,  2.03it/s]
 72%|███████▏  | 7680/10682 [1:05:26<24:41,  2.03it/s]
 72%|███████▏  | 7681/10682 [1:05:27<24:41,  2.03it/s]
 72%|███████▏  | 7682/10682 [1:05:27<24:41,  2.03it/s]
 72%|███████▏  | 7683/10682 [1:05:28<24:40,  2.03it/s]
 72%|███████▏  | 7684/10682 [1:05:28<24:39,  2.03it/s]
 72%|███████▏  | 7685/10682 [1:05:29<24:39,  2.03it/s]
 72%|███████▏  | 7686/10682 [1:05:29<24:40,  2.02it/s]
 72%|███████▏  | 7687/10682 [1:05:30<24:38,  2.03it/s]
 72%|███████▏  | 7688/10682 [1:05:30<24:38,  2.03it/s]
 72%|███████▏  | 7689/10682 [1:05:31<24:38,  2.02it/s]
 72%|███████▏  | 7690/10682 [1:05:31<24:36,  2.03it/s]
 72%|███████▏  | 7691/10682 [1:05:32<24:36,  2.03it/s]
 72%|███████▏  | 7692/10682 [1:05:32<24:35,  2.03it/s]
 72%|███████▏  | 7693/10682 [1:05:33<24:35,  2.03it/s]
 72%|███████▏  | 7694/10682 [1:05:33<24:35,  2.02it/s]
 72%|███████▏  | 7695/10682 [1:05:34<24:33,  2.03it/s]
 72%|███████▏  | 7696/10682 [1:05:34<24:33,  2.03it/s]
 72%|███████▏  | 7697/10682 [1:05:35<24:33,  2.03it/s]
 72%|███████▏  | 7698/10682 [1:05:35<24:33,  2.03it/s]
 72%|███████▏  | 7699/10682 [1:05:35<24:33,  2.02it/s]
 72%|███████▏  | 7700/10682 [1:05:36<24:32,  2.02it/s]
                                                      
{'loss': 2.8169, 'grad_norm': 0.2712630331516266, 'learning_rate': 0.00021922461032806601, 'epoch': 10.09}
+
 72%|███████▏  | 7700/10682 [1:05:36<24:32,  2.02it/s]
 72%|███████▏  | 7701/10682 [1:05:36<24:33,  2.02it/s]
 72%|███████▏  | 7702/10682 [1:05:37<24:30,  2.03it/s]
 72%|███████▏  | 7703/10682 [1:05:37<24:32,  2.02it/s]
 72%|███████▏  | 7704/10682 [1:05:38<24:30,  2.03it/s]
 72%|███████▏  | 7705/10682 [1:05:38<24:30,  2.02it/s]
 72%|███████▏  | 7706/10682 [1:05:39<24:29,  2.02it/s]
 72%|███████▏  | 7707/10682 [1:05:39<24:29,  2.02it/s]
 72%|███████▏  | 7708/10682 [1:05:40<24:30,  2.02it/s]
 72%|███████▏  | 7709/10682 [1:05:40<24:27,  2.03it/s]
 72%|███████▏  | 7710/10682 [1:05:41<24:26,  2.03it/s]
 72%|███████▏  | 7711/10682 [1:05:41<24:26,  2.03it/s]
 72%|███████▏  | 7712/10682 [1:05:42<24:26,  2.02it/s]
 72%|███████▏  | 7713/10682 [1:05:42<24:26,  2.02it/s]
 72%|███████▏  | 7714/10682 [1:05:43<24:26,  2.02it/s]
 72%|███████▏  | 7715/10682 [1:05:43<24:26,  2.02it/s]
 72%|███████▏  | 7716/10682 [1:05:44<24:24,  2.02it/s]
 72%|███████▏  | 7717/10682 [1:05:44<24:24,  2.02it/s]
 72%|███████▏  | 7718/10682 [1:05:45<24:23,  2.03it/s]
 72%|███████▏  | 7719/10682 [1:05:45<24:22,  2.03it/s]
 72%|███████▏  | 7720/10682 [1:05:46<24:21,  2.03it/s]
 72%|███████▏  | 7721/10682 [1:05:46<24:21,  2.03it/s]
 72%|███████▏  | 7722/10682 [1:05:47<24:22,  2.02it/s]
 72%|███████▏  | 7723/10682 [1:05:47<24:20,  2.03it/s]
 72%|███████▏  | 7724/10682 [1:05:48<24:21,  2.02it/s]
 72%|███████▏  | 7725/10682 [1:05:48<24:20,  2.03it/s]
                                                      
{'loss': 2.8163, 'grad_norm': 0.27432700991630554, 'learning_rate': 0.0002158538481528759, 'epoch': 10.12}
+
 72%|███████▏  | 7725/10682 [1:05:48<24:20,  2.03it/s]
 72%|███████▏  | 7726/10682 [1:05:49<24:21,  2.02it/s]
 72%|███████▏  | 7727/10682 [1:05:49<24:20,  2.02it/s]
 72%|███████▏  | 7728/10682 [1:05:50<24:19,  2.02it/s]
 72%|███████▏  | 7729/10682 [1:05:50<24:18,  2.02it/s]
 72%|███████▏  | 7730/10682 [1:05:51<24:19,  2.02it/s]
 72%|███████▏  | 7731/10682 [1:05:51<24:18,  2.02it/s]
 72%|███████▏  | 7732/10682 [1:05:52<24:18,  2.02it/s]
 72%|███████▏  | 7733/10682 [1:05:52<24:15,  2.03it/s]
 72%|███████▏  | 7734/10682 [1:05:53<24:16,  2.02it/s]
 72%|███████▏  | 7735/10682 [1:05:53<24:13,  2.03it/s]
 72%|███████▏  | 7736/10682 [1:05:54<24:14,  2.03it/s]
 72%|███████▏  | 7737/10682 [1:05:54<24:13,  2.03it/s]
 72%|███████▏  | 7738/10682 [1:05:55<24:13,  2.02it/s]
 72%|███████▏  | 7739/10682 [1:05:55<24:14,  2.02it/s]
 72%|███████▏  | 7740/10682 [1:05:56<24:12,  2.03it/s]
 72%|███████▏  | 7741/10682 [1:05:56<24:13,  2.02it/s]
 72%|███████▏  | 7742/10682 [1:05:57<24:11,  2.03it/s]
 72%|███████▏  | 7743/10682 [1:05:57<24:11,  2.02it/s]
 72%|███████▏  | 7744/10682 [1:05:58<24:10,  2.03it/s]
 73%|███████▎  | 7745/10682 [1:05:58<24:10,  2.02it/s]
 73%|███████▎  | 7746/10682 [1:05:59<24:09,  2.03it/s]
 73%|███████▎  | 7747/10682 [1:05:59<24:10,  2.02it/s]
 73%|███████▎  | 7748/10682 [1:06:00<24:09,  2.02it/s]
 73%|███████▎  | 7749/10682 [1:06:00<24:08,  2.02it/s]
 73%|███████▎  | 7750/10682 [1:06:01<24:07,  2.02it/s]{'loss': 2.8167, 'grad_norm': 0.2737395167350769, 'learning_rate': 0.00021250205309110155, 'epoch': 10.15}
                                                      
+
 73%|███████▎  | 7750/10682 [1:06:01<24:07,  2.02it/s]
 73%|███████▎  | 7751/10682 [1:06:01<24:08,  2.02it/s]
 73%|███████▎  | 7752/10682 [1:06:02<24:07,  2.02it/s]
 73%|███████▎  | 7753/10682 [1:06:02<24:06,  2.02it/s]
 73%|███████▎  | 7754/10682 [1:06:03<24:06,  2.02it/s]
 73%|███████▎  | 7755/10682 [1:06:03<24:05,  2.02it/s]
 73%|███████▎  | 7756/10682 [1:06:04<24:06,  2.02it/s]
 73%|███████▎  | 7757/10682 [1:06:04<24:05,  2.02it/s]
 73%|███████▎  | 7758/10682 [1:06:05<24:04,  2.02it/s]
 73%|███████▎  | 7759/10682 [1:06:05<24:02,  2.03it/s]
 73%|███████▎  | 7760/10682 [1:06:06<24:04,  2.02it/s]
 73%|███████▎  | 7761/10682 [1:06:06<24:02,  2.03it/s]
 73%|███████▎  | 7762/10682 [1:06:07<24:03,  2.02it/s]
 73%|███████▎  | 7763/10682 [1:06:07<24:01,  2.02it/s]
 73%|███████▎  | 7764/10682 [1:06:08<24:02,  2.02it/s]
 73%|███████▎  | 7765/10682 [1:06:08<24:00,  2.02it/s]
 73%|███████▎  | 7766/10682 [1:06:09<24:01,  2.02it/s]
 73%|███████▎  | 7767/10682 [1:06:09<24:00,  2.02it/s]
 73%|███████▎  | 7768/10682 [1:06:10<24:00,  2.02it/s]
 73%|███████▎  | 7769/10682 [1:06:10<23:59,  2.02it/s]
 73%|███████▎  | 7770/10682 [1:06:11<23:59,  2.02it/s]
 73%|███████▎  | 7771/10682 [1:06:11<23:56,  2.03it/s]
 73%|███████▎  | 7772/10682 [1:06:12<23:56,  2.03it/s]
 73%|███████▎  | 7773/10682 [1:06:12<23:57,  2.02it/s]
 73%|███████▎  | 7774/10682 [1:06:13<23:56,  2.02it/s]
 73%|███████▎  | 7775/10682 [1:06:13<23:55,  2.02it/s]
                                                      
{'loss': 2.8211, 'grad_norm': 0.2715729773044586, 'learning_rate': 0.00020916944887928359, 'epoch': 10.18}
+
 73%|███████▎  | 7775/10682 [1:06:13<23:55,  2.02it/s]
 73%|███████▎  | 7776/10682 [1:06:14<23:56,  2.02it/s]
 73%|███████▎  | 7777/10682 [1:06:14<23:55,  2.02it/s]
 73%|███████▎  | 7778/10682 [1:06:15<23:54,  2.03it/s]
 73%|███████▎  | 7779/10682 [1:06:15<23:54,  2.02it/s]
 73%|███████▎  | 7780/10682 [1:06:16<23:51,  2.03it/s]
 73%|███████▎  | 7781/10682 [1:06:16<23:52,  2.02it/s]
 73%|███████▎  | 7782/10682 [1:06:16<23:50,  2.03it/s]
 73%|███████▎  | 7783/10682 [1:06:17<23:51,  2.03it/s]
 73%|███████▎  | 7784/10682 [1:06:17<23:50,  2.03it/s]
 73%|███████▎  | 7785/10682 [1:06:18<23:49,  2.03it/s]
 73%|███████▎  | 7786/10682 [1:06:18<23:50,  2.02it/s]
 73%|███████▎  | 7787/10682 [1:06:19<23:49,  2.03it/s]
 73%|███████▎  | 7788/10682 [1:06:19<23:50,  2.02it/s]
 73%|███████▎  | 7789/10682 [1:06:20<23:47,  2.03it/s]
 73%|███████▎  | 7790/10682 [1:06:20<23:47,  2.03it/s]
 73%|███████▎  | 7791/10682 [1:06:21<23:47,  2.03it/s]
 73%|███████▎  | 7792/10682 [1:06:21<23:47,  2.02it/s]
 73%|███████▎  | 7793/10682 [1:06:22<23:46,  2.02it/s]
 73%|███████▎  | 7794/10682 [1:06:22<23:47,  2.02it/s]
 73%|███████▎  | 7795/10682 [1:06:23<23:45,  2.02it/s]
 73%|███████▎  | 7796/10682 [1:06:23<23:44,  2.03it/s]
 73%|███████▎  | 7797/10682 [1:06:24<23:44,  2.02it/s]
 73%|███████▎  | 7798/10682 [1:06:24<23:43,  2.03it/s]
 73%|███████▎  | 7799/10682 [1:06:25<23:43,  2.02it/s]
 73%|███████▎  | 7800/10682 [1:06:25<23:41,  2.03it/s]
                                                      
{'loss': 2.8289, 'grad_norm': 0.2772938907146454, 'learning_rate': 0.00020585625797294927, 'epoch': 10.22}
+
 73%|███████▎  | 7800/10682 [1:06:25<23:41,  2.03it/s]
 73%|███████▎  | 7801/10682 [1:06:26<23:43,  2.02it/s]
 73%|███████▎  | 7802/10682 [1:06:26<23:42,  2.02it/s]
 73%|███████▎  | 7803/10682 [1:06:27<23:42,  2.02it/s]
 73%|███████▎  | 7804/10682 [1:06:27<23:40,  2.03it/s]
 73%|███████▎  | 7805/10682 [1:06:28<23:40,  2.03it/s]
 73%|███████▎  | 7806/10682 [1:06:28<23:40,  2.03it/s]
 73%|███████▎  | 7807/10682 [1:06:29<23:39,  2.03it/s]
 73%|███████▎  | 7808/10682 [1:06:29<23:39,  2.02it/s]
 73%|███████▎  | 7809/10682 [1:06:30<23:38,  2.03it/s]
 73%|███████▎  | 7810/10682 [1:06:30<23:40,  2.02it/s]
 73%|███████▎  | 7811/10682 [1:06:31<23:38,  2.02it/s]
 73%|███████▎  | 7812/10682 [1:06:31<23:39,  2.02it/s]
 73%|███████▎  | 7813/10682 [1:06:32<23:37,  2.02it/s]
 73%|███████▎  | 7814/10682 [1:06:32<23:37,  2.02it/s]
 73%|███████▎  | 7815/10682 [1:06:33<23:36,  2.02it/s]
 73%|███████▎  | 7816/10682 [1:06:33<23:36,  2.02it/s]
 73%|███████▎  | 7817/10682 [1:06:34<23:35,  2.02it/s]
 73%|███████▎  | 7818/10682 [1:06:34<23:35,  2.02it/s]
 73%|███████▎  | 7819/10682 [1:06:35<23:34,  2.02it/s]
 73%|███████▎  | 7820/10682 [1:06:35<23:32,  2.03it/s]
 73%|███████▎  | 7821/10682 [1:06:36<23:33,  2.02it/s]
 73%|███████▎  | 7822/10682 [1:06:36<23:31,  2.03it/s]
 73%|███████▎  | 7823/10682 [1:06:37<23:32,  2.02it/s]
 73%|███████▎  | 7824/10682 [1:06:37<23:30,  2.03it/s]
 73%|███████▎  | 7825/10682 [1:06:38<23:30,  2.02it/s]
                                                      
{'loss': 2.8212, 'grad_norm': 0.2719649076461792, 'learning_rate': 0.00020256270153176371, 'epoch': 10.25}
+
 73%|███████▎  | 7825/10682 [1:06:38<23:30,  2.02it/s]
 73%|███████▎  | 7826/10682 [1:06:38<23:31,  2.02it/s]
 73%|███████▎  | 7827/10682 [1:06:39<23:31,  2.02it/s]
 73%|███████▎  | 7828/10682 [1:06:39<23:30,  2.02it/s]
 73%|███████▎  | 7829/10682 [1:06:40<23:30,  2.02it/s]
 73%|███████▎  | 7830/10682 [1:06:40<23:27,  2.03it/s]
 73%|███████▎  | 7831/10682 [1:06:41<23:26,  2.03it/s]
 73%|███████▎  | 7832/10682 [1:06:41<23:26,  2.03it/s]
 73%|███████▎  | 7833/10682 [1:06:42<23:27,  2.02it/s]
 73%|███████▎  | 7834/10682 [1:06:42<23:28,  2.02it/s]
 73%|███████▎  | 7835/10682 [1:06:43<23:26,  2.02it/s]
 73%|███████▎  | 7836/10682 [1:06:43<23:25,  2.02it/s]
 73%|███████▎  | 7837/10682 [1:06:44<23:25,  2.02it/s]
 73%|███████▎  | 7838/10682 [1:06:44<23:25,  2.02it/s]
 73%|███████▎  | 7839/10682 [1:06:45<23:24,  2.02it/s]
 73%|███████▎  | 7840/10682 [1:06:45<23:24,  2.02it/s]
 73%|███████▎  | 7841/10682 [1:06:46<23:23,  2.02it/s]
 73%|███████▎  | 7842/10682 [1:06:46<23:22,  2.03it/s]
 73%|███████▎  | 7843/10682 [1:06:47<23:20,  2.03it/s]
 73%|███████▎  | 7844/10682 [1:06:47<23:20,  2.03it/s]
 73%|███████▎  | 7845/10682 [1:06:48<23:20,  2.03it/s]
 73%|███████▎  | 7846/10682 [1:06:48<23:20,  2.03it/s]
 73%|███████▎  | 7847/10682 [1:06:49<23:19,  2.03it/s]
 73%|███████▎  | 7848/10682 [1:06:49<23:18,  2.03it/s]
 73%|███████▎  | 7849/10682 [1:06:50<23:19,  2.02it/s]
 73%|███████▎  | 7850/10682 [1:06:50<23:17,  2.03it/s]{'loss': 2.8496, 'grad_norm': 0.2761227786540985, 'learning_rate': 0.00019928899940476624, 'epoch': 10.28}
+
                                                      

 73%|███████▎  | 7850/10682 [1:06:50<23:17,  2.03it/s]
 73%|███████▎  | 7851/10682 [1:06:51<23:19,  2.02it/s]
 74%|███████▎  | 7852/10682 [1:06:51<23:18,  2.02it/s]
 74%|███████▎  | 7853/10682 [1:06:52<23:18,  2.02it/s]
 74%|███████▎  | 7854/10682 [1:06:52<23:16,  2.02it/s]
 74%|███████▎  | 7855/10682 [1:06:53<23:17,  2.02it/s]
 74%|███████▎  | 7856/10682 [1:06:53<23:16,  2.02it/s]
 74%|███████▎  | 7857/10682 [1:06:54<23:17,  2.02it/s]
 74%|███████▎  | 7858/10682 [1:06:54<23:15,  2.02it/s]
 74%|███████▎  | 7859/10682 [1:06:55<23:13,  2.03it/s]
 74%|███████▎  | 7860/10682 [1:06:55<23:12,  2.03it/s]
 74%|███████▎  | 7861/10682 [1:06:56<23:12,  2.03it/s]
 74%|███████▎  | 7862/10682 [1:06:56<23:12,  2.02it/s]
 74%|███████▎  | 7863/10682 [1:06:57<23:11,  2.03it/s]
 74%|███████▎  | 7864/10682 [1:06:57<23:11,  2.02it/s]
 74%|███████▎  | 7865/10682 [1:06:57<23:10,  2.03it/s]
 74%|███████▎  | 7866/10682 [1:06:58<23:10,  2.02it/s]
 74%|███████▎  | 7867/10682 [1:06:58<23:08,  2.03it/s]
 74%|███████▎  | 7868/10682 [1:06:59<23:09,  2.02it/s]
 74%|███████▎  | 7869/10682 [1:06:59<23:08,  2.03it/s]
 74%|███████▎  | 7870/10682 [1:07:00<23:08,  2.03it/s]
 74%|███████▎  | 7871/10682 [1:07:00<23:07,  2.03it/s]
 74%|███████▎  | 7872/10682 [1:07:01<23:07,  2.02it/s]
 74%|███████▎  | 7873/10682 [1:07:01<23:07,  2.02it/s]
 74%|███████▎  | 7874/10682 [1:07:02<23:05,  2.03it/s]
 74%|███████▎  | 7875/10682 [1:07:02<23:06,  2.02it/s]
                                                      
{'loss': 2.8414, 'grad_norm': 0.27529338002204895, 'learning_rate': 0.00019603537011569566, 'epoch': 10.31}
+
 74%|███████▎  | 7875/10682 [1:07:02<23:06,  2.02it/s]
 74%|███████▎  | 7876/10682 [1:07:03<23:06,  2.02it/s]
 74%|███████▎  | 7877/10682 [1:07:03<23:05,  2.02it/s]
 74%|███████▍  | 7878/10682 [1:07:04<23:03,  2.03it/s]
 74%|███████▍  | 7879/10682 [1:07:04<23:04,  2.03it/s]
 74%|███████▍  | 7880/10682 [1:07:05<23:04,  2.02it/s]
 74%|███████▍  | 7881/10682 [1:07:05<23:03,  2.02it/s]
 74%|███████▍  | 7882/10682 [1:07:06<23:03,  2.02it/s]
 74%|███████▍  | 7883/10682 [1:07:06<23:04,  2.02it/s]
 74%|███████▍  | 7884/10682 [1:07:07<23:02,  2.02it/s]
 74%|███████▍  | 7885/10682 [1:07:07<23:01,  2.02it/s]
 74%|███████▍  | 7886/10682 [1:07:08<23:00,  2.02it/s]
 74%|███████▍  | 7887/10682 [1:07:08<23:01,  2.02it/s]
 74%|███████▍  | 7888/10682 [1:07:09<23:00,  2.02it/s]
 74%|███████▍  | 7889/10682 [1:07:09<23:00,  2.02it/s]
 74%|███████▍  | 7890/10682 [1:07:10<22:59,  2.02it/s]
 74%|███████▍  | 7891/10682 [1:07:10<22:58,  2.02it/s]
 74%|███████▍  | 7892/10682 [1:07:11<22:59,  2.02it/s]
 74%|███████▍  | 7893/10682 [1:07:11<22:57,  2.03it/s]
 74%|███████▍  | 7894/10682 [1:07:12<22:58,  2.02it/s]
 74%|███████▍  | 7895/10682 [1:07:12<22:58,  2.02it/s]
 74%|███████▍  | 7896/10682 [1:07:13<22:57,  2.02it/s]
 74%|███████▍  | 7897/10682 [1:07:13<22:55,  2.02it/s]
 74%|███████▍  | 7898/10682 [1:07:14<22:55,  2.02it/s]
 74%|███████▍  | 7899/10682 [1:07:14<22:53,  2.03it/s]
 74%|███████▍  | 7900/10682 [1:07:15<22:53,  2.02it/s]{'loss': 2.8346, 'grad_norm': 0.27512723207473755, 'learning_rate': 0.0001928020308484042, 'epoch': 10.35}
                                                      
+
 74%|███████▍  | 7900/10682 [1:07:15<22:53,  2.02it/s]
 74%|███████▍  | 7901/10682 [1:07:15<22:54,  2.02it/s]
 74%|███████▍  | 7902/10682 [1:07:16<22:54,  2.02it/s]
 74%|███████▍  | 7903/10682 [1:07:16<22:53,  2.02it/s]
 74%|███████▍  | 7904/10682 [1:07:17<22:53,  2.02it/s]
 74%|███████▍  | 7905/10682 [1:07:17<22:51,  2.02it/s]
 74%|███████▍  | 7906/10682 [1:07:18<22:51,  2.02it/s]
 74%|███████▍  | 7907/10682 [1:07:18<22:49,  2.03it/s]
 74%|███████▍  | 7908/10682 [1:07:19<22:50,  2.02it/s]
 74%|███████▍  | 7909/10682 [1:07:19<22:49,  2.03it/s]
 74%|███████▍  | 7910/10682 [1:07:20<22:50,  2.02it/s]
 74%|███████▍  | 7911/10682 [1:07:20<22:48,  2.02it/s]
 74%|███████▍  | 7912/10682 [1:07:21<22:47,  2.03it/s]
 74%|███████▍  | 7913/10682 [1:07:21<22:47,  2.02it/s]
 74%|███████▍  | 7914/10682 [1:07:22<22:46,  2.03it/s]
 74%|███████▍  | 7915/10682 [1:07:22<22:47,  2.02it/s]
 74%|███████▍  | 7916/10682 [1:07:23<22:45,  2.02it/s]
 74%|███████▍  | 7917/10682 [1:07:23<22:46,  2.02it/s]
 74%|███████▍  | 7918/10682 [1:07:24<22:45,  2.02it/s]
 74%|███████▍  | 7919/10682 [1:07:24<22:46,  2.02it/s]
 74%|███████▍  | 7920/10682 [1:07:25<22:43,  2.03it/s]
 74%|███████▍  | 7921/10682 [1:07:25<22:44,  2.02it/s]
 74%|███████▍  | 7922/10682 [1:07:26<22:43,  2.02it/s]
 74%|███████▍  | 7923/10682 [1:07:26<22:43,  2.02it/s]
 74%|███████▍  | 7924/10682 [1:07:27<22:42,  2.02it/s]
 74%|███████▍  | 7925/10682 [1:07:27<22:42,  2.02it/s]{'loss': 2.849, 'grad_norm': 0.27163901925086975, 'learning_rate': 0.00018958919743235897, 'epoch': 10.38}
+
                                                      

 74%|███████▍  | 7925/10682 [1:07:27<22:42,  2.02it/s]
 74%|███████▍  | 7926/10682 [1:07:28<22:43,  2.02it/s]
 74%|███████▍  | 7927/10682 [1:07:28<22:42,  2.02it/s]
 74%|███████▍  | 7928/10682 [1:07:29<22:41,  2.02it/s]
 74%|███████▍  | 7929/10682 [1:07:29<22:41,  2.02it/s]
 74%|███████▍  | 7930/10682 [1:07:30<22:41,  2.02it/s]
 74%|███████▍  | 7931/10682 [1:07:30<22:40,  2.02it/s]
 74%|███████▍  | 7932/10682 [1:07:31<22:40,  2.02it/s]
 74%|███████▍  | 7933/10682 [1:07:31<22:39,  2.02it/s]
 74%|███████▍  | 7934/10682 [1:07:32<22:39,  2.02it/s]
 74%|███████▍  | 7935/10682 [1:07:32<22:38,  2.02it/s]
 74%|███████▍  | 7936/10682 [1:07:33<22:37,  2.02it/s]
 74%|███████▍  | 7937/10682 [1:07:33<22:36,  2.02it/s]
 74%|███████▍  | 7938/10682 [1:07:34<22:37,  2.02it/s]
 74%|███████▍  | 7939/10682 [1:07:34<22:36,  2.02it/s]
 74%|███████▍  | 7940/10682 [1:07:35<22:37,  2.02it/s]
 74%|███████▍  | 7941/10682 [1:07:35<22:34,  2.02it/s]
 74%|███████▍  | 7942/10682 [1:07:36<22:35,  2.02it/s]
 74%|███████▍  | 7943/10682 [1:07:36<22:33,  2.02it/s]
 74%|███████▍  | 7944/10682 [1:07:37<22:33,  2.02it/s]
 74%|███████▍  | 7945/10682 [1:07:37<22:32,  2.02it/s]
 74%|███████▍  | 7946/10682 [1:07:38<22:33,  2.02it/s]
 74%|███████▍  | 7947/10682 [1:07:38<22:31,  2.02it/s]
 74%|███████▍  | 7948/10682 [1:07:39<22:32,  2.02it/s]
 74%|███████▍  | 7949/10682 [1:07:39<22:30,  2.02it/s]
 74%|███████▍  | 7950/10682 [1:07:39<22:30,  2.02it/s]
                                                      
{'loss': 2.8396, 'grad_norm': 0.2769433259963989, 'learning_rate': 0.0001863970843282357, 'epoch': 10.41}
+
 74%|███████▍  | 7950/10682 [1:07:40<22:30,  2.02it/s]
 74%|███████▍  | 7951/10682 [1:07:40<22:29,  2.02it/s]
 74%|███████▍  | 7952/10682 [1:07:40<22:30,  2.02it/s]
 74%|███████▍  | 7953/10682 [1:07:41<22:28,  2.02it/s]
 74%|███████▍  | 7954/10682 [1:07:41<22:29,  2.02it/s]
 74%|███████▍  | 7955/10682 [1:07:42<22:29,  2.02it/s]
 74%|███████▍  | 7956/10682 [1:07:42<22:29,  2.02it/s]
 74%|███████▍  | 7957/10682 [1:07:43<22:26,  2.02it/s]
 74%|███████▍  | 7958/10682 [1:07:43<22:26,  2.02it/s]
 75%|███████▍  | 7959/10682 [1:07:44<22:25,  2.02it/s]
 75%|███████▍  | 7960/10682 [1:07:44<22:24,  2.03it/s]
 75%|███████▍  | 7961/10682 [1:07:45<22:24,  2.02it/s]
 75%|███████▍  | 7962/10682 [1:07:45<22:22,  2.03it/s]
 75%|███████▍  | 7963/10682 [1:07:46<22:22,  2.02it/s]
 75%|███████▍  | 7964/10682 [1:07:46<22:21,  2.03it/s]
 75%|███████▍  | 7965/10682 [1:07:47<22:22,  2.02it/s]
 75%|███████▍  | 7966/10682 [1:07:47<22:21,  2.02it/s]
 75%|███████▍  | 7967/10682 [1:07:48<22:22,  2.02it/s]
 75%|███████▍  | 7968/10682 [1:07:48<22:21,  2.02it/s]
 75%|███████▍  | 7969/10682 [1:07:49<22:21,  2.02it/s]
 75%|███████▍  | 7970/10682 [1:07:49<22:21,  2.02it/s]
 75%|███████▍  | 7971/10682 [1:07:50<22:20,  2.02it/s]
 75%|███████▍  | 7972/10682 [1:07:50<22:19,  2.02it/s]
 75%|███████▍  | 7973/10682 [1:07:51<22:19,  2.02it/s]
 75%|███████▍  | 7974/10682 [1:07:51<22:19,  2.02it/s]
 75%|███████▍  | 7975/10682 [1:07:52<22:18,  2.02it/s]{'loss': 2.8431, 'grad_norm': 0.28411322832107544, 'learning_rate': 0.00018322590461360383, 'epoch': 10.45}
+
                                                      

 75%|███████▍  | 7975/10682 [1:07:52<22:18,  2.02it/s]
 75%|███████▍  | 7976/10682 [1:07:52<22:19,  2.02it/s]
 75%|███████▍  | 7977/10682 [1:07:53<22:18,  2.02it/s]
 75%|███████▍  | 7978/10682 [1:07:53<22:17,  2.02it/s]
 75%|███████▍  | 7979/10682 [1:07:54<22:17,  2.02it/s]
 75%|███████▍  | 7980/10682 [1:07:54<22:15,  2.02it/s]
 75%|███████▍  | 7981/10682 [1:07:55<22:16,  2.02it/s]
 75%|███████▍  | 7982/10682 [1:07:55<22:14,  2.02it/s]
 75%|███████▍  | 7983/10682 [1:07:56<22:12,  2.02it/s]
 75%|███████▍  | 7984/10682 [1:07:56<22:13,  2.02it/s]
 75%|███████▍  | 7985/10682 [1:07:57<22:13,  2.02it/s]
 75%|███████▍  | 7986/10682 [1:07:57<22:13,  2.02it/s]
 75%|███████▍  | 7987/10682 [1:07:58<22:10,  2.02it/s]
 75%|███████▍  | 7988/10682 [1:07:58<22:11,  2.02it/s]
 75%|███████▍  | 7989/10682 [1:07:59<22:09,  2.03it/s]
 75%|███████▍  | 7990/10682 [1:07:59<22:08,  2.03it/s]
 75%|███████▍  | 7991/10682 [1:08:00<22:08,  2.02it/s]
 75%|███████▍  | 7992/10682 [1:08:00<22:07,  2.03it/s]
 75%|███████▍  | 7993/10682 [1:08:01<22:07,  2.02it/s]
 75%|███████▍  | 7994/10682 [1:08:01<22:06,  2.03it/s]
 75%|███████▍  | 7995/10682 [1:08:02<22:07,  2.02it/s]
 75%|███████▍  | 7996/10682 [1:08:02<22:05,  2.03it/s]
 75%|███████▍  | 7997/10682 [1:08:03<22:07,  2.02it/s]
 75%|███████▍  | 7998/10682 [1:08:03<22:06,  2.02it/s]
 75%|███████▍  | 7999/10682 [1:08:04<22:06,  2.02it/s]
 75%|███████▍  | 8000/10682 [1:08:04<22:06,  2.02it/s]{'loss': 2.856, 'grad_norm': 0.26959338784217834, 'learning_rate': 0.00018007586996870206, 'epoch': 10.48}
                                                      
+
 75%|███████▍  | 8000/10682 [1:08:04<22:06,  2.02it/s]
 75%|███████▍  | 8001/10682 [1:08:05<22:07,  2.02it/s]
 75%|███████▍  | 8002/10682 [1:08:05<22:05,  2.02it/s]
 75%|███████▍  | 8003/10682 [1:08:06<22:04,  2.02it/s]
 75%|███████▍  | 8004/10682 [1:08:06<22:02,  2.03it/s]
 75%|███████▍  | 8005/10682 [1:08:07<22:02,  2.02it/s]
 75%|███████▍  | 8006/10682 [1:08:07<22:01,  2.02it/s]
 75%|███████▍  | 8007/10682 [1:08:08<22:03,  2.02it/s]
 75%|███████▍  | 8008/10682 [1:08:08<22:01,  2.02it/s]
 75%|███████▍  | 8009/10682 [1:08:09<22:01,  2.02it/s]
 75%|███████▍  | 8010/10682 [1:08:09<21:59,  2.02it/s]
 75%|███████▍  | 8011/10682 [1:08:10<22:00,  2.02it/s]
 75%|███████▌  | 8012/10682 [1:08:10<21:58,  2.02it/s]
 75%|████��██▌  | 8013/10682 [1:08:11<21:58,  2.02it/s]
 75%|███████▌  | 8014/10682 [1:08:11<21:58,  2.02it/s]
 75%|███████▌  | 8015/10682 [1:08:12<21:56,  2.03it/s]
 75%|███████▌  | 8016/10682 [1:08:12<21:57,  2.02it/s]
 75%|███████▌  | 8017/10682 [1:08:13<21:56,  2.02it/s]
 75%|███████▌  | 8018/10682 [1:08:13<21:56,  2.02it/s]
 75%|███████▌  | 8019/10682 [1:08:14<21:55,  2.02it/s]
 75%|███████▌  | 8020/10682 [1:08:14<21:55,  2.02it/s]
 75%|███████▌  | 8021/10682 [1:08:15<21:53,  2.03it/s]
 75%|███████▌  | 8022/10682 [1:08:15<21:54,  2.02it/s]
 75%|███████▌  | 8023/10682 [1:08:16<21:52,  2.03it/s]
 75%|███████▌  | 8024/10682 [1:08:16<21:51,  2.03it/s]
 75%|███████▌  | 8025/10682 [1:08:17<21:51,  2.03it/s]{'loss': 2.859, 'grad_norm': 0.27145281434059143, 'learning_rate': 0.00017694719066230924, 'epoch': 10.51}
+
                                                      

 75%|███████▌  | 8025/10682 [1:08:17<21:51,  2.03it/s]
 75%|███████▌  | 8026/10682 [1:08:17<21:52,  2.02it/s]
 75%|███████▌  | 8027/10682 [1:08:18<21:52,  2.02it/s]
 75%|███████▌  | 8028/10682 [1:08:18<21:51,  2.02it/s]
 75%|███████▌  | 8029/10682 [1:08:19<21:50,  2.02it/s]
 75%|███████▌  | 8030/10682 [1:08:19<21:48,  2.03it/s]
 75%|███████▌  | 8031/10682 [1:08:20<21:49,  2.02it/s]
 75%|███████▌  | 8032/10682 [1:08:20<21:49,  2.02it/s]
 75%|███████▌  | 8033/10682 [1:08:21<21:49,  2.02it/s]
 75%|███████▌  | 8034/10682 [1:08:21<21:47,  2.03it/s]
 75%|███████▌  | 8035/10682 [1:08:21<21:48,  2.02it/s]
 75%|███████▌  | 8036/10682 [1:08:22<21:46,  2.02it/s]
 75%|███████▌  | 8037/10682 [1:08:22<21:47,  2.02it/s]
 75%|███████▌  | 8038/10682 [1:08:23<21:45,  2.03it/s]
 75%|███████▌  | 8039/10682 [1:08:23<21:44,  2.03it/s]
 75%|███████▌  | 8040/10682 [1:08:24<21:44,  2.03it/s]
 75%|███████▌  | 8041/10682 [1:08:24<21:44,  2.02it/s]
 75%|███████▌  | 8042/10682 [1:08:25<21:43,  2.03it/s]
 75%|███████▌  | 8043/10682 [1:08:25<21:44,  2.02it/s]
 75%|███████▌  | 8044/10682 [1:08:26<21:43,  2.02it/s]
 75%|███████▌  | 8045/10682 [1:08:26<21:42,  2.03it/s]
 75%|███████▌  | 8046/10682 [1:08:27<21:42,  2.02it/s]
 75%|███████▌  | 8047/10682 [1:08:27<21:40,  2.03it/s]
 75%|███████▌  | 8048/10682 [1:08:28<21:40,  2.03it/s]
 75%|███████▌  | 8049/10682 [1:08:28<21:39,  2.03it/s]
 75%|███████▌  | 8050/10682 [1:08:29<21:40,  2.02it/s]{'loss': 2.8619, 'grad_norm': 0.270614355802536, 'learning_rate': 0.00017384007553770858, 'epoch': 10.54}
                                                      
+
 75%|███████▌  | 8050/10682 [1:08:29<21:40,  2.02it/s]
 75%|███████▌  | 8051/10682 [1:08:29<21:39,  2.02it/s]
 75%|███████▌  | 8052/10682 [1:08:30<21:40,  2.02it/s]
 75%|███████▌  | 8053/10682 [1:08:30<21:38,  2.02it/s]
 75%|███████▌  | 8054/10682 [1:08:31<21:40,  2.02it/s]
 75%|███████▌  | 8055/10682 [1:08:31<21:40,  2.02it/s]
 75%|███████▌  | 8056/10682 [1:08:32<21:38,  2.02it/s]
 75%|███████▌  | 8057/10682 [1:08:32<21:37,  2.02it/s]
 75%|███████▌  | 8058/10682 [1:08:33<21:36,  2.02it/s]
 75%|███████▌  | 8059/10682 [1:08:33<21:36,  2.02it/s]
 75%|███████▌  | 8060/10682 [1:08:34<21:36,  2.02it/s]
 75%|███████▌  | 8061/10682 [1:08:34<21:36,  2.02it/s]
 75%|███████▌  | 8062/10682 [1:08:35<21:35,  2.02it/s]
 75%|███████▌  | 8063/10682 [1:08:35<21:35,  2.02it/s]
 75%|███████▌  | 8064/10682 [1:08:36<21:34,  2.02it/s]
 76%|███████▌  | 8065/10682 [1:08:36<21:34,  2.02it/s]
 76%|███████▌  | 8066/10682 [1:08:37<21:33,  2.02it/s]
 76%|███████▌  | 8067/10682 [1:08:37<21:33,  2.02it/s]
 76%|███████▌  | 8068/10682 [1:08:38<21:32,  2.02it/s]
 76%|███████▌  | 8069/10682 [1:08:38<21:31,  2.02it/s]
 76%|███████▌  | 8070/10682 [1:08:39<21:32,  2.02it/s]
 76%|███████▌  | 8071/10682 [1:08:39<21:31,  2.02it/s]
 76%|███████▌  | 8072/10682 [1:08:40<21:30,  2.02it/s]
 76%|███████▌  | 8073/10682 [1:08:40<21:30,  2.02it/s]
 76%|███████▌  | 8074/10682 [1:08:41<21:30,  2.02it/s]
 76%|███████▌  | 8075/10682 [1:08:41<21:30,  2.02it/s]
                                                      
{'loss': 2.852, 'grad_norm': 0.27252912521362305, 'learning_rate': 0.00017075473199874692, 'epoch': 10.58}
+
 76%|███████▌  | 8075/10682 [1:08:41<21:30,  2.02it/s]
 76%|███████▌  | 8076/10682 [1:08:42<21:32,  2.02it/s]
 76%|███████▌  | 8077/10682 [1:08:42<21:31,  2.02it/s]
 76%|███████▌  | 8078/10682 [1:08:43<21:30,  2.02it/s]
 76%|███████▌  | 8079/10682 [1:08:43<21:29,  2.02it/s]
 76%|███████▌  | 8080/10682 [1:08:44<21:27,  2.02it/s]
 76%|███████▌  | 8081/10682 [1:08:44<21:26,  2.02it/s]
 76%|███████▌  | 8082/10682 [1:08:45<21:25,  2.02it/s]
 76%|███████▌  | 8083/10682 [1:08:45<21:24,  2.02it/s]
 76%|███████▌  | 8084/10682 [1:08:46<21:24,  2.02it/s]
 76%|███████▌  | 8085/10682 [1:08:46<21:24,  2.02it/s]
 76%|███████▌  | 8086/10682 [1:08:47<21:23,  2.02it/s]
 76%|███████▌  | 8087/10682 [1:08:47<21:23,  2.02it/s]
 76%|███████▌  | 8088/10682 [1:08:48<21:22,  2.02it/s]
 76%|███████▌  | 8089/10682 [1:08:48<21:22,  2.02it/s]
 76%|███████▌  | 8090/10682 [1:08:49<21:21,  2.02it/s]
 76%|███████▌  | 8091/10682 [1:08:49<21:21,  2.02it/s]
 76%|███████▌  | 8092/10682 [1:08:50<21:20,  2.02it/s]
 76%|███████▌  | 8093/10682 [1:08:50<21:19,  2.02it/s]
 76%|███████▌  | 8094/10682 [1:08:51<21:18,  2.02it/s]
 76%|███████▌  | 8095/10682 [1:08:51<21:18,  2.02it/s]
 76%|███████▌  | 8096/10682 [1:08:52<21:17,  2.02it/s]
 76%|███████▌  | 8097/10682 [1:08:52<21:17,  2.02it/s]
 76%|███████▌  | 8098/10682 [1:08:53<21:16,  2.02it/s]
 76%|███████▌  | 8099/10682 [1:08:53<21:15,  2.02it/s]
 76%|███████▌  | 8100/10682 [1:08:54<21:15,  2.02it/s]
                                                      {'loss': 2.8546, 'grad_norm': 0.2722909450531006, 'learning_rate': 0.00016769136599599017, 'epoch': 10.61}
+
 76%|███████▌  | 8100/10682 [1:08:54<21:15,  2.02it/s]
 76%|███████▌  | 8101/10682 [1:08:54<21:17,  2.02it/s]
 76%|███████▌  | 8102/10682 [1:08:55<21:16,  2.02it/s]
 76%|███████▌  | 8103/10682 [1:08:55<21:15,  2.02it/s]
 76%|███████▌  | 8104/10682 [1:08:56<21:14,  2.02it/s]
 76%|███████▌  | 8105/10682 [1:08:56<21:14,  2.02it/s]
 76%|███████▌  | 8106/10682 [1:08:57<21:13,  2.02it/s]
 76%|███████▌  | 8107/10682 [1:08:57<21:13,  2.02it/s]
 76%|███████▌  | 8108/10682 [1:08:58<21:12,  2.02it/s]
 76%|███████▌  | 8109/10682 [1:08:58<21:11,  2.02it/s]
 76%|███████▌  | 8110/10682 [1:08:59<21:10,  2.02it/s]
 76%|███████▌  | 8111/10682 [1:08:59<21:10,  2.02it/s]
 76%|███████▌  | 8112/10682 [1:09:00<21:10,  2.02it/s]
 76%|███████▌  | 8113/10682 [1:09:00<21:11,  2.02it/s]
 76%|███████▌  | 8114/10682 [1:09:01<21:09,  2.02it/s]
 76%|███████▌  | 8115/10682 [1:09:01<21:09,  2.02it/s]
 76%|███████▌  | 8116/10682 [1:09:02<21:08,  2.02it/s]
 76%|███████▌  | 8117/10682 [1:09:02<21:09,  2.02it/s]
 76%|███████▌  | 8118/10682 [1:09:03<21:07,  2.02it/s]
 76%|███████▌  | 8119/10682 [1:09:03<21:07,  2.02it/s]
 76%|███████▌  | 8120/10682 [1:09:04<21:06,  2.02it/s]
 76%|███████▌  | 8121/10682 [1:09:04<21:05,  2.02it/s]
 76%|███████▌  | 8122/10682 [1:09:05<21:05,  2.02it/s]
 76%|███████▌  | 8123/10682 [1:09:05<21:03,  2.03it/s]
 76%|███████▌  | 8124/10682 [1:09:05<21:03,  2.02it/s]
 76%|███████▌  | 8125/10682 [1:09:06<21:03,  2.02it/s]{'loss': 2.8539, 'grad_norm': 0.27448949217796326, 'learning_rate': 0.0001646501820129766, 'epoch': 10.64}
                                                      
+
 76%|███████▌  | 8125/10682 [1:09:06<21:03,  2.02it/s]
 76%|███████▌  | 8126/10682 [1:09:06<21:04,  2.02it/s]
 76%|███████▌  | 8127/10682 [1:09:07<21:01,  2.02it/s]
 76%|███████▌  | 8128/10682 [1:09:07<21:02,  2.02it/s]
 76%|███████▌  | 8129/10682 [1:09:08<21:00,  2.03it/s]
 76%|███████▌  | 8130/10682 [1:09:08<21:02,  2.02it/s]
 76%|███████▌  | 8131/10682 [1:09:09<21:00,  2.02it/s]
 76%|███████▌  | 8132/10682 [1:09:09<21:00,  2.02it/s]
 76%|███████▌  | 8133/10682 [1:09:10<20:59,  2.02it/s]
 76%|███████▌  | 8134/10682 [1:09:10<20:59,  2.02it/s]
 76%|███████▌  | 8135/10682 [1:09:11<20:58,  2.02it/s]
 76%|███████▌  | 8136/10682 [1:09:11<20:58,  2.02it/s]
 76%|███████▌  | 8137/10682 [1:09:12<20:58,  2.02it/s]
 76%|███████▌  | 8138/10682 [1:09:12<20:58,  2.02it/s]
 76%|███████▌  | 8139/10682 [1:09:13<20:56,  2.02it/s]
 76%|███████▌  | 8140/10682 [1:09:13<20:56,  2.02it/s]
 76%|███████▌  | 8141/10682 [1:09:14<20:55,  2.02it/s]
 76%|███████▌  | 8142/10682 [1:09:14<20:54,  2.03it/s]
 76%|███████▌  | 8143/10682 [1:09:15<20:53,  2.02it/s]
 76%|███████▌  | 8144/10682 [1:09:15<20:52,  2.03it/s]
 76%|███████▌  | 8145/10682 [1:09:16<20:52,  2.02it/s]
 76%|███████▋  | 8146/10682 [1:09:16<20:51,  2.03it/s]
 76%|███████▋  | 8147/10682 [1:09:17<20:52,  2.02it/s]
 76%|███████▋  | 8148/10682 [1:09:17<20:52,  2.02it/s]
 76%|███████▋  | 8149/10682 [1:09:18<20:52,  2.02it/s]
 76%|███████▋  | 8150/10682 [1:09:18<20:49,  2.03it/s]{'loss': 2.8656, 'grad_norm': 0.2788846492767334, 'learning_rate': 0.00016163138305256598, 'epoch': 10.67}
+
                                                      

 76%|███████▋  | 8150/10682 [1:09:18<20:49,  2.03it/s]
 76%|███████▋  | 8151/10682 [1:09:19<20:51,  2.02it/s]
 76%|███████▋  | 8152/10682 [1:09:19<20:50,  2.02it/s]
 76%|███████▋  | 8153/10682 [1:09:20<20:50,  2.02it/s]
 76%|███████▋  | 8154/10682 [1:09:20<20:49,  2.02it/s]
 76%|███████▋  | 8155/10682 [1:09:21<20:49,  2.02it/s]
 76%|███████▋  | 8156/10682 [1:09:21<20:47,  2.02it/s]
 76%|███████▋  | 8157/10682 [1:09:22<20:48,  2.02it/s]
 76%|███████▋  | 8158/10682 [1:09:22<20:47,  2.02it/s]
 76%|███████▋  | 8159/10682 [1:09:23<20:48,  2.02it/s]
 76%|███████▋  | 8160/10682 [1:09:23<20:47,  2.02it/s]
 76%|███████▋  | 8161/10682 [1:09:24<20:46,  2.02it/s]
 76%|███████▋  | 8162/10682 [1:09:24<20:46,  2.02it/s]
 76%|███████▋  | 8163/10682 [1:09:25<20:47,  2.02it/s]
 76%|███████▋  | 8164/10682 [1:09:25<20:45,  2.02it/s]
 76%|███████▋  | 8165/10682 [1:09:26<20:45,  2.02it/s]
 76%|███████▋  | 8166/10682 [1:09:26<20:43,  2.02it/s]
 76%|███████▋  | 8167/10682 [1:09:27<20:43,  2.02it/s]
 76%|███████▋  | 8168/10682 [1:09:27<20:42,  2.02it/s]
 76%|███████▋  | 8169/10682 [1:09:28<20:42,  2.02it/s]
 76%|███████▋  | 8170/10682 [1:09:28<20:40,  2.02it/s]
 76%|███████▋  | 8171/10682 [1:09:29<20:40,  2.02it/s]
 77%|███████▋  | 8172/10682 [1:09:29<20:39,  2.03it/s]
 77%|███████▋  | 8173/10682 [1:09:30<20:39,  2.02it/s]
 77%|███████▋  | 8174/10682 [1:09:30<20:39,  2.02it/s]
 77%|███████▋  | 8175/10682 [1:09:31<20:38,  2.02it/s]
                                                      
{'loss': 2.8525, 'grad_norm': 0.27568942308425903, 'learning_rate': 0.00015863517062339038, 'epoch': 10.71}
+
 77%|███████▋  | 8175/10682 [1:09:31<20:38,  2.02it/s]
 77%|███████▋  | 8176/10682 [1:09:31<20:39,  2.02it/s]
 77%|███████▋  | 8177/10682 [1:09:32<20:39,  2.02it/s]
 77%|███████▋  | 8178/10682 [1:09:32<20:38,  2.02it/s]
 77%|███████▋  | 8179/10682 [1:09:33<20:37,  2.02it/s]
 77%|███████▋  | 8180/10682 [1:09:33<20:36,  2.02it/s]
 77%|███████▋  | 8181/10682 [1:09:34<20:36,  2.02it/s]
 77%|███████▋  | 8182/10682 [1:09:34<20:35,  2.02it/s]
 77%|███████▋  | 8183/10682 [1:09:35<20:35,  2.02it/s]
 77%|███████▋  | 8184/10682 [1:09:35<20:35,  2.02it/s]
 77%|███████▋  | 8185/10682 [1:09:36<20:34,  2.02it/s]
 77%|███████▋  | 8186/10682 [1:09:36<20:33,  2.02it/s]
 77%|███████▋  | 8187/10682 [1:09:37<20:33,  2.02it/s]
 77%|███████▋  | 8188/10682 [1:09:37<20:32,  2.02it/s]
 77%|███████▋  | 8189/10682 [1:09:38<20:31,  2.02it/s]
 77%|███████▋  | 8190/10682 [1:09:38<20:30,  2.02it/s]
 77%|███████▋  | 8191/10682 [1:09:39<20:31,  2.02it/s]
 77%|███████▋  | 8192/10682 [1:09:39<20:31,  2.02it/s]
 77%|███████▋  | 8193/10682 [1:09:40<22:16,  1.86it/s]
 77%|███████▋  | 8194/10682 [1:09:40<21:44,  1.91it/s]
 77%|███████▋  | 8195/10682 [1:09:41<21:22,  1.94it/s]
 77%|███████▋  | 8196/10682 [1:09:41<21:05,  1.96it/s]
 77%|███████▋  | 8197/10682 [1:09:42<20:55,  1.98it/s]
 77%|███████▋  | 8198/10682 [1:09:42<20:47,  1.99it/s]
 77%|███████▋  | 8199/10682 [1:09:43<20:41,  2.00it/s]
 77%|███████▋  | 8200/10682 [1:09:43<20:36,  2.01it/s]
                                                      
{'loss': 2.8585, 'grad_norm': 0.27291959524154663, 'learning_rate': 0.00015566174472640188, 'epoch': 10.74}
+
 77%|███████▋  | 8200/10682 [1:09:43<20:36,  2.01it/s]
 77%|███████▋  | 8201/10682 [1:09:44<20:35,  2.01it/s]
 77%|███████▋  | 8202/10682 [1:09:44<20:31,  2.01it/s]
 77%|███████▋  | 8203/10682 [1:09:45<20:30,  2.02it/s]
 77%|███████▋  | 8204/10682 [1:09:45<20:27,  2.02it/s]
 77%|███████▋  | 8205/10682 [1:09:46<20:26,  2.02it/s]
 77%|███████▋  | 8206/10682 [1:09:46<20:26,  2.02it/s]
 77%|███████▋  | 8207/10682 [1:09:47<20:25,  2.02it/s]
 77%|███████▋  | 8208/10682 [1:09:47<20:23,  2.02it/s]
 77%|███████▋  | 8209/10682 [1:09:48<20:22,  2.02it/s]
 77%|███████▋  | 8210/10682 [1:09:48<20:22,  2.02it/s]
 77%|███████▋  | 8211/10682 [1:09:49<20:23,  2.02it/s]
 77%|███████▋  | 8212/10682 [1:09:49<20:22,  2.02it/s]
 77%|███████▋  | 8213/10682 [1:09:50<20:20,  2.02it/s]
 77%|███████▋  | 8214/10682 [1:09:50<20:20,  2.02it/s]
 77%|███████▋  | 8215/10682 [1:09:51<20:19,  2.02it/s]
 77%|███████▋  | 8216/10682 [1:09:51<20:19,  2.02it/s]
 77%|███████▋  | 8217/10682 [1:09:52<20:18,  2.02it/s]
 77%|███████▋  | 8218/10682 [1:09:52<20:18,  2.02it/s]
 77%|███████▋  | 8219/10682 [1:09:53<20:17,  2.02it/s]
 77%|███████▋  | 8220/10682 [1:09:53<20:16,  2.02it/s]
 77%|███████▋  | 8221/10682 [1:09:54<20:16,  2.02it/s]
 77%|███████▋  | 8222/10682 [1:09:54<20:16,  2.02it/s]
 77%|███████▋  | 8223/10682 [1:09:55<20:16,  2.02it/s]
 77%|███████▋  | 8224/10682 [1:09:55<20:15,  2.02it/s]
 77%|███████▋  | 8225/10682 [1:09:56<20:14,  2.02it/s]
                                                      
{'loss': 2.8559, 'grad_norm': 0.27311572432518005, 'learning_rate': 0.0001527113038415231, 'epoch': 10.77}
+
 77%|███████▋  | 8225/10682 [1:09:56<20:14,  2.02it/s]
 77%|███████▋  | 8226/10682 [1:09:56<20:14,  2.02it/s]
 77%|███████▋  | 8227/10682 [1:09:57<20:13,  2.02it/s]
 77%|███████▋  | 8228/10682 [1:09:57<20:13,  2.02it/s]
 77%|███████▋  | 8229/10682 [1:09:58<20:12,  2.02it/s]
 77%|███████▋  | 8230/10682 [1:09:58<20:12,  2.02it/s]
 77%|███████▋  | 8231/10682 [1:09:59<20:11,  2.02it/s]
 77%|███████▋  | 8232/10682 [1:09:59<20:11,  2.02it/s]
 77%|███████▋  | 8233/10682 [1:10:00<20:12,  2.02it/s]
 77%|███████▋  | 8234/10682 [1:10:00<20:11,  2.02it/s]
 77%|███████▋  | 8235/10682 [1:10:01<20:10,  2.02it/s]
 77%|███████▋  | 8236/10682 [1:10:01<20:09,  2.02it/s]
 77%|███████▋  | 8237/10682 [1:10:02<20:08,  2.02it/s]
 77%|███████▋  | 8238/10682 [1:10:02<20:09,  2.02it/s]
 77%|███████▋  | 8239/10682 [1:10:03<20:09,  2.02it/s]
 77%|███████▋  | 8240/10682 [1:10:03<20:08,  2.02it/s]
 77%|███████▋  | 8241/10682 [1:10:03<20:07,  2.02it/s]
 77%|███████▋  | 8242/10682 [1:10:04<20:07,  2.02it/s]
 77%|███████▋  | 8243/10682 [1:10:04<20:05,  2.02it/s]
 77%|███████▋  | 8244/10682 [1:10:05<20:07,  2.02it/s]
 77%|███████▋  | 8245/10682 [1:10:05<20:05,  2.02it/s]
 77%|███████▋  | 8246/10682 [1:10:06<20:04,  2.02it/s]
 77%|███████▋  | 8247/10682 [1:10:06<20:04,  2.02it/s]
 77%|███████▋  | 8248/10682 [1:10:07<20:04,  2.02it/s]
 77%|███████▋  | 8249/10682 [1:10:07<20:04,  2.02it/s]
 77%|███████▋  | 8250/10682 [1:10:08<20:02,  2.02it/s]{'loss': 2.8655, 'grad_norm': 0.27253279089927673, 'learning_rate': 0.00014978404491439802, 'epoch': 10.81}
+
                                                      

 77%|███████▋  | 8250/10682 [1:10:08<20:02,  2.02it/s]
 77%|███████▋  | 8251/10682 [1:10:08<20:04,  2.02it/s]
 77%|███████▋  | 8252/10682 [1:10:09<20:02,  2.02it/s]
 77%|███████▋  | 8253/10682 [1:10:09<20:01,  2.02it/s]
 77%|███████▋  | 8254/10682 [1:10:10<20:00,  2.02it/s]
 77%|███████▋  | 8255/10682 [1:10:10<20:00,  2.02it/s]
 77%|███████▋  | 8256/10682 [1:10:11<19:58,  2.02it/s]
 77%|███████▋  | 8257/10682 [1:10:11<19:58,  2.02it/s]
 77%|███████▋  | 8258/10682 [1:10:12<19:58,  2.02it/s]
 77%|███████▋  | 8259/10682 [1:10:12<19:57,  2.02it/s]
 77%|███████▋  | 8260/10682 [1:10:13<19:57,  2.02it/s]
 77%|███████▋  | 8261/10682 [1:10:13<19:56,  2.02it/s]
 77%|███████▋  | 8262/10682 [1:10:14<19:56,  2.02it/s]
 77%|███████▋  | 8263/10682 [1:10:14<19:54,  2.02it/s]
 77%|███████▋  | 8264/10682 [1:10:15<19:56,  2.02it/s]
 77%|███████▋  | 8265/10682 [1:10:15<19:54,  2.02it/s]
 77%|███████▋  | 8266/10682 [1:10:16<19:54,  2.02it/s]
 77%|███████▋  | 8267/10682 [1:10:16<19:53,  2.02it/s]
 77%|███████▋  | 8268/10682 [1:10:17<19:53,  2.02it/s]
 77%|███████▋  | 8269/10682 [1:10:17<19:53,  2.02it/s]
 77%|███████▋  | 8270/10682 [1:10:18<19:52,  2.02it/s]
 77%|███████▋  | 8271/10682 [1:10:18<19:52,  2.02it/s]
 77%|███████▋  | 8272/10682 [1:10:19<19:52,  2.02it/s]
 77%|███████▋  | 8273/10682 [1:10:19<19:50,  2.02it/s]
 77%|███████▋  | 8274/10682 [1:10:20<19:50,  2.02it/s]
 77%|███████▋  | 8275/10682 [1:10:20<19:49,  2.02it/s]{'loss': 2.8654, 'grad_norm': 0.28846028447151184, 'learning_rate': 0.00014688016334324605, 'epoch': 10.84}
                                                      
+
 77%|███████▋  | 8275/10682 [1:10:20<19:49,  2.02it/s]
 77%|███████▋  | 8276/10682 [1:10:21<19:50,  2.02it/s]
 77%|███████▋  | 8277/10682 [1:10:21<19:49,  2.02it/s]
 77%|███████▋  | 8278/10682 [1:10:22<19:49,  2.02it/s]
 78%|███████▊  | 8279/10682 [1:10:22<19:47,  2.02it/s]
 78%|███████▊  | 8280/10682 [1:10:23<19:47,  2.02it/s]
 78%|███████▊  | 8281/10682 [1:10:23<19:46,  2.02it/s]
 78%|███████▊  | 8282/10682 [1:10:24<19:46,  2.02it/s]
 78%|███████▊  | 8283/10682 [1:10:24<19:45,  2.02it/s]
 78%|███████▊  | 8284/10682 [1:10:25<19:46,  2.02it/s]
 78%|███████▊  | 8285/10682 [1:10:25<19:44,  2.02it/s]
 78%|███████▊  | 8286/10682 [1:10:26<19:45,  2.02it/s]
 78%|███████▊  | 8287/10682 [1:10:26<19:44,  2.02it/s]
 78%|███████▊  | 8288/10682 [1:10:27<19:44,  2.02it/s]
 78%|███████▊  | 8289/10682 [1:10:27<19:43,  2.02it/s]
 78%|███████▊  | 8290/10682 [1:10:28<21:25,  1.86it/s]
 78%|███████▊  | 8291/10682 [1:10:28<20:52,  1.91it/s]
 78%|███████▊  | 8292/10682 [1:10:29<20:31,  1.94it/s]
 78%|███████▊  | 8293/10682 [1:10:29<20:14,  1.97it/s]
 78%|███████▊  | 8294/10682 [1:10:30<20:03,  1.98it/s]
 78%|███████▊  | 8295/10682 [1:10:30<19:56,  2.00it/s]
 78%|███████▊  | 8296/10682 [1:10:31<19:49,  2.01it/s]
 78%|███████▊  | 8297/10682 [1:10:31<19:46,  2.01it/s]
 78%|███████▊  | 8298/10682 [1:10:32<19:41,  2.02it/s]
 78%|███████▊  | 8299/10682 [1:10:32<19:40,  2.02it/s]
 78%|███████▊  | 8300/10682 [1:10:33<19:38,  2.02it/s]{'loss': 2.8666, 'grad_norm': 0.27803298830986023, 'learning_rate': 0.00014399985296581835, 'epoch': 10.87}
+
                                                      

 78%|███████▊  | 8300/10682 [1:10:33<19:38,  2.02it/s]
 78%|███████▊  | 8301/10682 [1:10:33<19:38,  2.02it/s]
 78%|███████▊  | 8302/10682 [1:10:34<19:37,  2.02it/s]
 78%|███████▊  | 8303/10682 [1:10:34<19:37,  2.02it/s]
 78%|███████▊  | 8304/10682 [1:10:35<19:35,  2.02it/s]
 78%|███████▊  | 8305/10682 [1:10:35<19:35,  2.02it/s]
 78%|███████▊  | 8306/10682 [1:10:36<19:33,  2.02it/s]
 78%|███████▊  | 8307/10682 [1:10:36<19:33,  2.02it/s]
 78%|███████▊  | 8308/10682 [1:10:37<19:32,  2.02it/s]
 78%|███████▊  | 8309/10682 [1:10:37<19:31,  2.03it/s]
 78%|███████▊  | 8310/10682 [1:10:38<19:31,  2.03it/s]
 78%|███████▊  | 8311/10682 [1:10:38<19:29,  2.03it/s]
 78%|███████▊  | 8312/10682 [1:10:39<19:30,  2.03it/s]
 78%|███████▊  | 8313/10682 [1:10:39<19:28,  2.03it/s]
 78%|███████▊  | 8314/10682 [1:10:40<19:28,  2.03it/s]
 78%|███████▊  | 8315/10682 [1:10:40<19:28,  2.03it/s]
 78%|███████▊  | 8316/10682 [1:10:41<19:28,  2.02it/s]
 78%|███████▊  | 8317/10682 [1:10:41<19:27,  2.03it/s]
 78%|███████▊  | 8318/10682 [1:10:42<19:28,  2.02it/s]
 78%|███████▊  | 8319/10682 [1:10:42<19:27,  2.02it/s]
 78%|███████▊  | 8320/10682 [1:10:43<19:26,  2.02it/s]
 78%|███████▊  | 8321/10682 [1:10:43<19:26,  2.02it/s]
 78%|███████▊  | 8322/10682 [1:10:44<19:28,  2.02it/s]
 78%|███████▊  | 8323/10682 [1:10:44<19:27,  2.02it/s]
 78%|███████▊  | 8324/10682 [1:10:45<19:26,  2.02it/s]
 78%|███████▊  | 8325/10682 [1:10:45<19:24,  2.02it/s]
                                                      {'loss': 2.8571, 'grad_norm': 0.2763269543647766, 'learning_rate': 0.00014114330604645943, 'epoch': 10.9}
+
 78%|███████▊  | 8325/10682 [1:10:45<19:24,  2.02it/s]
 78%|███████▊  | 8326/10682 [1:10:46<19:25,  2.02it/s]
 78%|███████▊  | 8327/10682 [1:10:46<19:24,  2.02it/s]
 78%|███████▊  | 8328/10682 [1:10:47<19:24,  2.02it/s]
 78%|███████▊  | 8329/10682 [1:10:47<19:23,  2.02it/s]
 78%|███████▊  | 8330/10682 [1:10:48<19:22,  2.02it/s]
 78%|███████▊  | 8331/10682 [1:10:48<19:21,  2.02it/s]
 78%|███████▊  | 8332/10682 [1:10:49<19:20,  2.02it/s]
 78%|███████▊  | 8333/10682 [1:10:49<19:21,  2.02it/s]
 78%|███████▊  | 8334/10682 [1:10:50<19:20,  2.02it/s]
 78%|███████▊  | 8335/10682 [1:10:50<19:21,  2.02it/s]
 78%|███████▊  | 8336/10682 [1:10:51<19:19,  2.02it/s]
 78%|███████▊  | 8337/10682 [1:10:51<19:20,  2.02it/s]
 78%|███████▊  | 8338/10682 [1:10:52<19:19,  2.02it/s]
 78%|███████▊  | 8339/10682 [1:10:52<19:19,  2.02it/s]
 78%|███████▊  | 8340/10682 [1:10:53<19:16,  2.02it/s]
 78%|███████▊  | 8341/10682 [1:10:53<19:16,  2.03it/s]
 78%|███████▊  | 8342/10682 [1:10:54<19:15,  2.03it/s]
 78%|███████▊  | 8343/10682 [1:10:54<19:15,  2.03it/s]
 78%|███████▊  | 8344/10682 [1:10:55<19:15,  2.02it/s]
 78%|███████▊  | 8345/10682 [1:10:55<19:13,  2.03it/s]
 78%|███████▊  | 8346/10682 [1:10:56<19:14,  2.02it/s]
 78%|███████▊  | 8347/10682 [1:10:56<19:12,  2.03it/s]
 78%|███████▊  | 8348/10682 [1:10:57<19:13,  2.02it/s]
 78%|███████▊  | 8349/10682 [1:10:57<19:11,  2.03it/s]
 78%|███████▊  | 8350/10682 [1:10:58<19:12,  2.02it/s]
                                                      
{'loss': 2.8635, 'grad_norm': 0.2721956968307495, 'learning_rate': 0.00013831071326327282, 'epoch': 10.94}
+
 78%|███████▊  | 8350/10682 [1:10:58<19:12,  2.02it/s]
 78%|███████▊  | 8351/10682 [1:10:58<19:11,  2.02it/s]
 78%|███████▊  | 8352/10682 [1:10:58<19:11,  2.02it/s]
 78%|███████▊  | 8353/10682 [1:10:59<19:10,  2.03it/s]
 78%|███████▊  | 8354/10682 [1:10:59<19:10,  2.02it/s]
 78%|███████▊  | 8355/10682 [1:11:00<19:09,  2.02it/s]
 78%|███████▊  | 8356/10682 [1:11:00<19:09,  2.02it/s]
 78%|███████▊  | 8357/10682 [1:11:01<19:08,  2.03it/s]
 78%|███████▊  | 8358/10682 [1:11:01<19:06,  2.03it/s]
 78%|███████▊  | 8359/10682 [1:11:02<19:07,  2.02it/s]
 78%|███████▊  | 8360/10682 [1:11:02<19:06,  2.03it/s]
 78%|███████▊  | 8361/10682 [1:11:03<19:07,  2.02it/s]
 78%|███████▊  | 8362/10682 [1:11:03<19:04,  2.03it/s]
 78%|███████▊  | 8363/10682 [1:11:04<19:05,  2.02it/s]
 78%|███████▊  | 8364/10682 [1:11:04<19:03,  2.03it/s]
 78%|███████▊  | 8365/10682 [1:11:05<19:03,  2.03it/s]
 78%|███████▊  | 8366/10682 [1:11:05<19:03,  2.03it/s]
 78%|███████▊  | 8367/10682 [1:11:06<19:02,  2.03it/s]
 78%|███████▊  | 8368/10682 [1:11:06<19:03,  2.02it/s]
 78%|███████▊  | 8369/10682 [1:11:07<19:01,  2.03it/s]
 78%|███████▊  | 8370/10682 [1:11:07<19:02,  2.02it/s]
 78%|███████▊  | 8371/10682 [1:11:08<19:00,  2.03it/s]
 78%|███████▊  | 8372/10682 [1:11:08<19:01,  2.02it/s]
 78%|███████▊  | 8373/10682 [1:11:09<19:00,  2.03it/s]
 78%|███████▊  | 8374/10682 [1:11:09<19:00,  2.02it/s]
 78%|███████▊  | 8375/10682 [1:11:10<18:58,  2.03it/s]{'loss': 2.8677, 'grad_norm': 0.2699893116950989, 'learning_rate': 0.0001355022636953933, 'epoch': 10.97}
                                                      
+
 78%|███████▊  | 8375/10682 [1:11:10<18:58,  2.03it/s]
 78%|███████▊  | 8376/10682 [1:11:10<19:00,  2.02it/s]
 78%|███████▊  | 8377/10682 [1:11:11<18:58,  2.02it/s]
 78%|███████▊  | 8378/10682 [1:11:11<18:58,  2.02it/s]
 78%|███████▊  | 8379/10682 [1:11:12<18:59,  2.02it/s]
 78%|███████▊  | 8380/10682 [1:11:12<18:57,  2.02it/s]
 78%|███████▊  | 8381/10682 [1:11:13<18:57,  2.02it/s]
 78%|███████▊  | 8382/10682 [1:11:13<18:56,  2.02it/s]
 78%|███████▊  | 8383/10682 [1:11:14<18:57,  2.02it/s]
 78%|███████▊  | 8384/10682 [1:11:14<18:56,  2.02it/s]
 78%|███████▊  | 8385/10682 [1:11:15<18:56,  2.02it/s]
 79%|███████▊  | 8386/10682 [1:11:15<18:55,  2.02it/s]
 79%|███████▊  | 8387/10682 [1:11:16<18:54,  2.02it/s]
 79%|███████▊  | 8388/10682 [1:11:16<18:54,  2.02it/s]
 79%|███████▊  | 8389/10682 [1:11:17<18:53,  2.02it/s]
 79%|███████▊  | 8390/10682 [1:11:17<18:52,  2.02it/s]
 79%|███████▊  | 8391/10682 [1:11:18<18:51,  2.02it/s]
 79%|███████▊  | 8392/10682 [1:11:18<18:52,  2.02it/s]
 79%|███████▊  | 8393/10682 [1:11:19<18:51,  2.02it/s]
 79%|███████▊  | 8394/10682 [1:11:19<18:49,  2.03it/s]
 79%|███████▊  | 8395/10682 [1:11:20<18:49,  2.02it/s]
 79%|████��██▊  | 8396/10682 [1:11:20<18:48,  2.03it/s]
 79%|███████▊  | 8397/10682 [1:11:21<18:48,  2.02it/s]
 79%|███████▊  | 8398/10682 [1:11:21<19:13,  1.98it/s]
 79%|███████▊  | 8399/10682 [1:11:34<2:35:02,  4.07s/it]
 79%|███████▊  | 8400/10682 [1:11:34<1:54:08,  3.00s/it]{'loss': 2.8614, 'grad_norm': 0.2776219844818115, 'learning_rate': 0.0001327181448103661, 'epoch': 11.0}
+
                                                        

 79%|███████▊  | 8400/10682 [1:11:34<1:54:08,  3.00s/it]
 79%|███████▊  | 8401/10682 [1:11:35<1:25:30,  2.25s/it]
 79%|███████▊  | 8402/10682 [1:11:35<1:05:32,  1.72s/it]
 79%|███████▊  | 8403/10682 [1:11:36<51:29,  1.36s/it]  
 79%|███████▊  | 8404/10682 [1:11:36<41:45,  1.10s/it]
 79%|███████▊  | 8405/10682 [1:11:37<34:50,  1.09it/s]
 79%|███████▊  | 8406/10682 [1:11:37<30:05,  1.26it/s]
 79%|███████▊  | 8407/10682 [1:11:38<26:39,  1.42it/s]
 79%|███████▊  | 8408/10682 [1:11:38<24:16,  1.56it/s]
 79%|███████▊  | 8409/10682 [1:11:39<22:34,  1.68it/s]
 79%|███████▊  | 8410/10682 [1:11:39<21:27,  1.76it/s]
 79%|███████▊  | 8411/10682 [1:11:40<20:38,  1.83it/s]
 79%|███████▊  | 8412/10682 [1:11:40<20:03,  1.89it/s]
 79%|███████▉  | 8413/10682 [1:11:41<19:39,  1.92it/s]
 79%|███████▉  | 8414/10682 [1:11:41<19:22,  1.95it/s]
 79%|███████▉  | 8415/10682 [1:11:42<19:11,  1.97it/s]
 79%|███████▉  | 8416/10682 [1:11:42<19:02,  1.98it/s]
 79%|███████▉  | 8417/10682 [1:11:43<18:56,  1.99it/s]
 79%|███████▉  | 8418/10682 [1:11:43<18:51,  2.00it/s]
 79%|███████▉  | 8419/10682 [1:11:44<18:47,  2.01it/s]
 79%|███████▉  | 8420/10682 [1:11:44<18:43,  2.01it/s]
 79%|███████▉  | 8421/10682 [1:11:45<18:41,  2.02it/s]
 79%|███████▉  | 8422/10682 [1:11:45<18:39,  2.02it/s]
 79%|███████▉  | 8423/10682 [1:11:46<18:40,  2.02it/s]
 79%|███████▉  | 8424/10682 [1:11:46<18:38,  2.02it/s]
 79%|███████▉  | 8425/10682 [1:11:47<18:38,  2.02it/s]{'loss': 2.7578, 'grad_norm': 0.2850797176361084, 'learning_rate': 0.00012995854245163207, 'epoch': 11.03}
                                                      
+
 79%|███████▉  | 8425/10682 [1:11:47<18:38,  2.02it/s]
 79%|███████▉  | 8426/10682 [1:11:47<18:42,  2.01it/s]
 79%|███████▉  | 8427/10682 [1:11:48<18:39,  2.01it/s]
 79%|███████▉  | 8428/10682 [1:11:48<18:37,  2.02it/s]
 79%|███████▉  | 8429/10682 [1:11:49<18:36,  2.02it/s]
 79%|███████▉  | 8430/10682 [1:11:49<18:36,  2.02it/s]
 79%|███████▉  | 8431/10682 [1:11:50<18:33,  2.02it/s]
 79%|███████▉  | 8432/10682 [1:11:50<18:33,  2.02it/s]
 79%|███████▉  | 8433/10682 [1:11:51<18:31,  2.02it/s]
 79%|███████▉  | 8434/10682 [1:11:51<18:32,  2.02it/s]
 79%|███████▉  | 8435/10682 [1:11:52<18:31,  2.02it/s]
 79%|███████▉  | 8436/10682 [1:11:52<18:31,  2.02it/s]
 79%|███████▉  | 8437/10682 [1:11:52<18:30,  2.02it/s]
 79%|███████▉  | 8438/10682 [1:11:53<18:29,  2.02it/s]
 79%|███████▉  | 8439/10682 [1:11:53<18:28,  2.02it/s]
 79%|███████▉  | 8440/10682 [1:11:54<18:28,  2.02it/s]
 79%|███████▉  | 8441/10682 [1:11:54<18:27,  2.02it/s]
 79%|███████▉  | 8442/10682 [1:11:55<18:26,  2.02it/s]
 79%|███████▉  | 8443/10682 [1:11:55<18:25,  2.03it/s]
 79%|███████▉  | 8444/10682 [1:11:56<18:25,  2.02it/s]
 79%|███████▉  | 8445/10682 [1:11:56<18:24,  2.02it/s]
 79%|███████▉  | 8446/10682 [1:11:57<18:24,  2.02it/s]
 79%|███████▉  | 8447/10682 [1:11:57<18:24,  2.02it/s]
 79%|███████▉  | 8448/10682 [1:11:58<18:22,  2.03it/s]
 79%|███████▉  | 8449/10682 [1:11:58<18:23,  2.02it/s]
 79%|███████▉  | 8450/10682 [1:11:59<18:21,  2.03it/s]
                                                      
{'loss': 2.7557, 'grad_norm': 0.27638164162635803, 'learning_rate': 0.0001272236408261237, 'epoch': 11.07}
+
 79%|███████▉  | 8450/10682 [1:11:59<18:21,  2.03it/s]
 79%|███████▉  | 8451/10682 [1:11:59<18:24,  2.02it/s]
 79%|███████▉  | 8452/10682 [1:12:00<18:23,  2.02it/s]
 79%|███████▉  | 8453/10682 [1:12:00<18:22,  2.02it/s]
 79%|███████▉  | 8454/10682 [1:12:01<18:21,  2.02it/s]
 79%|███████▉  | 8455/10682 [1:12:01<18:21,  2.02it/s]
 79%|███████▉  | 8456/10682 [1:12:02<18:20,  2.02it/s]
 79%|███████▉  | 8457/10682 [1:12:02<18:19,  2.02it/s]
 79%|███████▉  | 8458/10682 [1:12:03<18:18,  2.02it/s]
 79%|███████▉  | 8459/10682 [1:12:03<18:18,  2.02it/s]
 79%|███████▉  | 8460/10682 [1:12:04<18:17,  2.02it/s]
 79%|███████▉  | 8461/10682 [1:12:04<18:17,  2.02it/s]
 79%|███████▉  | 8462/10682 [1:12:05<18:17,  2.02it/s]
 79%|███████▉  | 8463/10682 [1:12:05<18:16,  2.02it/s]
 79%|███████▉  | 8464/10682 [1:12:06<18:16,  2.02it/s]
 79%|███████▉  | 8465/10682 [1:12:06<18:16,  2.02it/s]
 79%|███████▉  | 8466/10682 [1:12:07<18:15,  2.02it/s]
 79%|███████▉  | 8467/10682 [1:12:07<18:15,  2.02it/s]
 79%|███████▉  | 8468/10682 [1:12:08<18:14,  2.02it/s]
 79%|███████▉  | 8469/10682 [1:12:08<18:13,  2.02it/s]
 79%|███████▉  | 8470/10682 [1:12:09<18:13,  2.02it/s]
 79%|███████▉  | 8471/10682 [1:12:09<18:14,  2.02it/s]
 79%|███████▉  | 8472/10682 [1:12:10<18:13,  2.02it/s]
 79%|███████▉  | 8473/10682 [1:12:10<18:12,  2.02it/s]
 79%|███████▉  | 8474/10682 [1:12:11<18:12,  2.02it/s]
 79%|███████▉  | 8475/10682 [1:12:11<18:11,  2.02it/s]{'loss': 2.7631, 'grad_norm': 0.2775728106498718, 'learning_rate': 0.00012451362249196797, 'epoch': 11.1}
+
                                                      

 79%|███████▉  | 8475/10682 [1:12:11<18:11,  2.02it/s]
 79%|███████▉  | 8476/10682 [1:12:12<18:12,  2.02it/s]
 79%|███████▉  | 8477/10682 [1:12:12<18:12,  2.02it/s]
 79%|███████▉  | 8478/10682 [1:12:13<18:11,  2.02it/s]
 79%|███████▉  | 8479/10682 [1:12:13<18:10,  2.02it/s]
 79%|███████▉  | 8480/10682 [1:12:14<18:09,  2.02it/s]
 79%|███████▉  | 8481/10682 [1:12:14<18:11,  2.02it/s]
 79%|███████▉  | 8482/10682 [1:12:15<18:09,  2.02it/s]
 79%|███████▉  | 8483/10682 [1:12:15<18:10,  2.02it/s]
 79%|███████▉  | 8484/10682 [1:12:16<18:07,  2.02it/s]
 79%|███████▉  | 8485/10682 [1:12:16<18:08,  2.02it/s]
 79%|███████▉  | 8486/10682 [1:12:17<18:07,  2.02it/s]
 79%|███████▉  | 8487/10682 [1:12:17<18:07,  2.02it/s]
 79%|███████▉  | 8488/10682 [1:12:18<18:05,  2.02it/s]
 79%|███████▉  | 8489/10682 [1:12:18<18:05,  2.02it/s]
 79%|███████▉  | 8490/10682 [1:12:19<18:04,  2.02it/s]
 79%|███████▉  | 8491/10682 [1:12:19<18:04,  2.02it/s]
 79%|███████▉  | 8492/10682 [1:12:20<18:03,  2.02it/s]
 80%|███████▉  | 8493/10682 [1:12:20<18:02,  2.02it/s]
 80%|███████▉  | 8494/10682 [1:12:21<18:02,  2.02it/s]
 80%|███████▉  | 8495/10682 [1:12:21<18:01,  2.02it/s]
 80%|███████▉  | 8496/10682 [1:12:22<18:02,  2.02it/s]
 80%|███████▉  | 8497/10682 [1:12:22<18:02,  2.02it/s]
 80%|███████▉  | 8498/10682 [1:12:23<18:01,  2.02it/s]
 80%|███████▉  | 8499/10682 [1:12:23<18:00,  2.02it/s]
 80%|███████▉  | 8500/10682 [1:12:24<17:59,  2.02it/s]{'loss': 2.7798, 'grad_norm': 0.28207170963287354, 'learning_rate': 0.00012182866834630096, 'epoch': 11.13}
+
                                                      

 80%|███████▉  | 8500/10682 [1:12:24<17:59,  2.02it/s]
 80%|███████▉  | 8501/10682 [1:12:24<17:59,  2.02it/s]
 80%|███████▉  | 8502/10682 [1:12:25<17:59,  2.02it/s]
 80%|███████▉  | 8503/10682 [1:12:25<17:58,  2.02it/s]
 80%|███████▉  | 8504/10682 [1:12:26<17:59,  2.02it/s]
 80%|███████▉  | 8505/10682 [1:12:26<17:57,  2.02it/s]
 80%|███████▉  | 8506/10682 [1:12:27<17:57,  2.02it/s]
 80%|███████▉  | 8507/10682 [1:12:27<17:56,  2.02it/s]
 80%|███████▉  | 8508/10682 [1:12:28<17:56,  2.02it/s]
 80%|███████▉  | 8509/10682 [1:12:28<17:55,  2.02it/s]
 80%|███████▉  | 8510/10682 [1:12:29<17:54,  2.02it/s]
 80%|███████▉  | 8511/10682 [1:12:29<17:54,  2.02it/s]
 80%|███████▉  | 8512/10682 [1:12:30<17:52,  2.02it/s]
 80%|███████▉  | 8513/10682 [1:12:30<17:52,  2.02it/s]
 80%|███████▉  | 8514/10682 [1:12:31<17:52,  2.02it/s]
 80%|███████▉  | 8515/10682 [1:12:31<17:51,  2.02it/s]
 80%|███████▉  | 8516/10682 [1:12:32<17:51,  2.02it/s]
 80%|███████▉  | 8517/10682 [1:12:32<17:52,  2.02it/s]
 80%|███████▉  | 8518/10682 [1:12:33<17:52,  2.02it/s]
 80%|███████▉  | 8519/10682 [1:12:33<17:50,  2.02it/s]
 80%|███████▉  | 8520/10682 [1:12:34<17:51,  2.02it/s]
 80%|███████▉  | 8521/10682 [1:12:34<17:49,  2.02it/s]
 80%|███████▉  | 8522/10682 [1:12:35<17:49,  2.02it/s]
 80%|███████▉  | 8523/10682 [1:12:35<17:47,  2.02it/s]
 80%|███████▉  | 8524/10682 [1:12:36<17:48,  2.02it/s]
 80%|███████▉  | 8525/10682 [1:12:36<17:47,  2.02it/s]{'loss': 2.7749, 'grad_norm': 0.2733232080936432, 'learning_rate': 0.00011916895761319264, 'epoch': 11.17}
+
                                                      

 80%|███████▉  | 8525/10682 [1:12:36<17:47,  2.02it/s]
 80%|███████▉  | 8526/10682 [1:12:37<17:47,  2.02it/s]
 80%|███████▉  | 8527/10682 [1:12:37<17:46,  2.02it/s]
 80%|███████▉  | 8528/10682 [1:12:38<17:47,  2.02it/s]
 80%|███████▉  | 8529/10682 [1:12:38<17:46,  2.02it/s]
 80%|███████▉  | 8530/10682 [1:12:39<17:46,  2.02it/s]
 80%|███████▉  | 8531/10682 [1:12:39<17:44,  2.02it/s]
 80%|███████▉  | 8532/10682 [1:12:39<17:44,  2.02it/s]
 80%|███████▉  | 8533/10682 [1:12:40<17:43,  2.02it/s]
 80%|███████▉  | 8534/10682 [1:12:40<17:43,  2.02it/s]
 80%|███████▉  | 8535/10682 [1:12:41<17:42,  2.02it/s]
 80%|███████▉  | 8536/10682 [1:12:41<17:42,  2.02it/s]
 80%|███████▉  | 8537/10682 [1:12:42<17:44,  2.01it/s]
 80%|███████▉  | 8538/10682 [1:12:42<17:42,  2.02it/s]
 80%|███████▉  | 8539/10682 [1:12:43<17:42,  2.02it/s]
 80%|███████▉  | 8540/10682 [1:12:43<17:40,  2.02it/s]
 80%|███████▉  | 8541/10682 [1:12:44<17:40,  2.02it/s]
 80%|███████▉  | 8542/10682 [1:12:44<17:38,  2.02it/s]
 80%|███████▉  | 8543/10682 [1:12:45<17:39,  2.02it/s]
 80%|███████▉  | 8544/10682 [1:12:45<17:38,  2.02it/s]
 80%|███████▉  | 8545/10682 [1:12:46<17:38,  2.02it/s]
 80%|████████  | 8546/10682 [1:12:46<17:36,  2.02it/s]
 80%|████████  | 8547/10682 [1:12:47<17:36,  2.02it/s]
 80%|████████  | 8548/10682 [1:12:47<17:36,  2.02it/s]
 80%|████████  | 8549/10682 [1:12:48<17:35,  2.02it/s]
 80%|████████  | 8550/10682 [1:12:48<17:35,  2.02it/s]
                                                      {'loss': 2.7808, 'grad_norm': 0.2801077663898468, 'learning_rate': 0.0001165346678316832, 'epoch': 11.2}
+
 80%|████████  | 8550/10682 [1:12:48<17:35,  2.02it/s]
 80%|████████  | 8551/10682 [1:12:49<17:35,  2.02it/s]
 80%|████████  | 8552/10682 [1:12:49<17:34,  2.02it/s]
 80%|████████  | 8553/10682 [1:12:50<17:34,  2.02it/s]
 80%|████████  | 8554/10682 [1:12:50<17:34,  2.02it/s]
 80%|████████  | 8555/10682 [1:12:51<17:33,  2.02it/s]
 80%|████████  | 8556/10682 [1:12:51<17:34,  2.02it/s]
 80%|████████  | 8557/10682 [1:12:52<17:33,  2.02it/s]
 80%|████████  | 8558/10682 [1:12:52<17:33,  2.02it/s]
 80%|████████  | 8559/10682 [1:12:53<17:33,  2.02it/s]
 80%|████████  | 8560/10682 [1:12:53<17:32,  2.02it/s]
 80%|████████  | 8561/10682 [1:12:54<17:32,  2.02it/s]
 80%|████████  | 8562/10682 [1:12:54<17:31,  2.02it/s]
 80%|████████  | 8563/10682 [1:12:55<17:30,  2.02it/s]
 80%|████████  | 8564/10682 [1:12:55<17:30,  2.02it/s]
 80%|████████  | 8565/10682 [1:12:56<17:29,  2.02it/s]
 80%|████████  | 8566/10682 [1:12:56<17:27,  2.02it/s]
 80%|████████  | 8567/10682 [1:12:57<17:27,  2.02it/s]
 80%|████████  | 8568/10682 [1:12:57<17:25,  2.02it/s]
 80%|████████  | 8569/10682 [1:12:58<17:25,  2.02it/s]
 80%|████████  | 8570/10682 [1:12:58<17:23,  2.02it/s]
 80%|████████  | 8571/10682 [1:12:59<17:23,  2.02it/s]
 80%|████████  | 8572/10682 [1:12:59<17:21,  2.03it/s]
 80%|████████  | 8573/10682 [1:13:00<17:22,  2.02it/s]
 80%|████████  | 8574/10682 [1:13:00<17:20,  2.03it/s]
 80%|████████  | 8575/10682 [1:13:01<17:20,  2.03it/s]
                                                      
{'loss': 2.7783, 'grad_norm': 0.28348976373672485, 'learning_rate': 0.00011392597484393285, 'epoch': 11.23}
+
 80%|████████  | 8575/10682 [1:13:01<17:20,  2.03it/s]
 80%|████████  | 8576/10682 [1:13:01<17:23,  2.02it/s]
 80%|████████  | 8577/10682 [1:13:02<17:22,  2.02it/s]
 80%|████████  | 8578/10682 [1:13:02<17:20,  2.02it/s]
 80%|████████  | 8579/10682 [1:13:03<17:19,  2.02it/s]
 80%|████████  | 8580/10682 [1:13:03<17:18,  2.02it/s]
 80%|████████  | 8581/10682 [1:13:04<17:17,  2.03it/s]
 80%|████████  | 8582/10682 [1:13:04<17:17,  2.02it/s]
 80%|████████  | 8583/10682 [1:13:05<17:16,  2.03it/s]
 80%|████████  | 8584/10682 [1:13:05<17:16,  2.02it/s]
 80%|████████  | 8585/10682 [1:13:06<17:15,  2.03it/s]
 80%|████████  | 8586/10682 [1:13:06<17:14,  2.03it/s]
 80%|████████  | 8587/10682 [1:13:07<17:14,  2.03it/s]
 80%|████████  | 8588/10682 [1:13:07<17:14,  2.02it/s]
 80%|████████  | 8589/10682 [1:13:08<17:14,  2.02it/s]
 80%|████████  | 8590/10682 [1:13:08<17:13,  2.03it/s]
 80%|████████  | 8591/10682 [1:13:09<17:13,  2.02it/s]
 80%|████████  | 8592/10682 [1:13:09<17:12,  2.02it/s]
 80%|████████  | 8593/10682 [1:13:10<17:11,  2.03it/s]
 80%|████████  | 8594/10682 [1:13:10<17:09,  2.03it/s]
 80%|████████  | 8595/10682 [1:13:11<17:08,  2.03it/s]
 80%|████████  | 8596/10682 [1:13:11<17:09,  2.03it/s]
 80%|████████  | 8597/10682 [1:13:12<17:09,  2.03it/s]
 80%|████████  | 8598/10682 [1:13:12<17:09,  2.02it/s]
 80%|████████  | 8599/10682 [1:13:13<17:07,  2.03it/s]
 81%|████████  | 8600/10682 [1:13:13<17:08,  2.02it/s]
                                                      
{'loss': 2.7817, 'grad_norm': 0.28865718841552734, 'learning_rate': 0.00011134305278348312, 'epoch': 11.26}
+
 81%|████████  | 8600/10682 [1:13:13<17:08,  2.02it/s]
 81%|████████  | 8601/10682 [1:13:14<17:08,  2.02it/s]
 81%|████████  | 8602/10682 [1:13:14<17:08,  2.02it/s]
 81%|████████  | 8603/10682 [1:13:15<17:07,  2.02it/s]
 81%|████████  | 8604/10682 [1:13:15<17:06,  2.02it/s]
 81%|████████  | 8605/10682 [1:13:16<17:04,  2.03it/s]
 81%|████████  | 8606/10682 [1:13:16<17:05,  2.02it/s]
 81%|████████  | 8607/10682 [1:13:17<17:04,  2.03it/s]
 81%|████████  | 8608/10682 [1:13:17<17:03,  2.03it/s]
 81%|████████  | 8609/10682 [1:13:18<17:04,  2.02it/s]
 81%|████████  | 8610/10682 [1:13:18<17:02,  2.03it/s]
 81%|████████  | 8611/10682 [1:13:19<17:03,  2.02it/s]
 81%|████████  | 8612/10682 [1:13:19<17:01,  2.03it/s]
 81%|████████  | 8613/10682 [1:13:20<17:00,  2.03it/s]
 81%|████████  | 8614/10682 [1:13:20<17:00,  2.03it/s]
 81%|████████  | 8615/10682 [1:13:21<17:00,  2.03it/s]
 81%|████████  | 8616/10682 [1:13:21<17:00,  2.02it/s]
 81%|████████  | 8617/10682 [1:13:22<17:00,  2.02it/s]
 81%|████████  | 8618/10682 [1:13:22<16:59,  2.02it/s]
 81%|████████  | 8619/10682 [1:13:23<16:57,  2.03it/s]
 81%|████████  | 8620/10682 [1:13:23<16:58,  2.02it/s]
 81%|████████  | 8621/10682 [1:13:24<16:57,  2.03it/s]
 81%|████████  | 8622/10682 [1:13:24<16:58,  2.02it/s]
 81%|████████  | 8623/10682 [1:13:24<16:56,  2.02it/s]
 81%|████████  | 8624/10682 [1:13:25<16:56,  2.02it/s]
 81%|████████  | 8625/10682 [1:13:25<16:55,  2.02it/s]
                                                      {'loss': 2.78, 'grad_norm': 0.2784369885921478, 'learning_rate': 0.00010878607406363367, 'epoch': 11.3}
+

 81%|████████  | 8625/10682 [1:13:25<16:55,  2.02it/s]
 81%|████████  | 8626/10682 [1:13:26<16:57,  2.02it/s]
 81%|████████  | 8627/10682 [1:13:26<16:56,  2.02it/s]
 81%|████████  | 8628/10682 [1:13:27<16:55,  2.02it/s]
 81%|████████  | 8629/10682 [1:13:27<16:54,  2.02it/s]
 81%|████████  | 8630/10682 [1:13:28<16:54,  2.02it/s]
 81%|████████  | 8631/10682 [1:13:28<16:53,  2.02it/s]
 81%|████████  | 8632/10682 [1:13:29<16:52,  2.02it/s]
 81%|████████  | 8633/10682 [1:13:29<16:51,  2.03it/s]
 81%|████████  | 8634/10682 [1:13:30<16:51,  2.02it/s]
 81%|████████  | 8635/10682 [1:13:30<16:50,  2.03it/s]
 81%|████████  | 8636/10682 [1:13:31<16:51,  2.02it/s]
 81%|████████  | 8637/10682 [1:13:31<16:50,  2.02it/s]
 81%|████████  | 8638/10682 [1:13:32<16:50,  2.02it/s]
 81%|████████  | 8639/10682 [1:13:32<16:49,  2.02it/s]
 81%|████████  | 8640/10682 [1:13:33<16:48,  2.03it/s]
 81%|████████  | 8641/10682 [1:13:33<16:48,  2.02it/s]
 81%|████████  | 8642/10682 [1:13:34<16:47,  2.03it/s]
 81%|████████  | 8643/10682 [1:13:34<16:46,  2.03it/s]
 81%|████████  | 8644/10682 [1:13:35<16:45,  2.03it/s]
 81%|████████  | 8645/10682 [1:13:35<16:45,  2.03it/s]
 81%|████████  | 8646/10682 [1:13:36<16:47,  2.02it/s]
 81%|████████  | 8647/10682 [1:13:36<16:45,  2.02it/s]
 81%|████████  | 8648/10682 [1:13:37<16:45,  2.02it/s]
 81%|████████  | 8649/10682 [1:13:37<16:44,  2.02it/s]
 81%|████████  | 8650/10682 [1:13:38<16:44,  2.02it/s]
                                                      
{'loss': 2.7789, 'grad_norm': 0.27582377195358276, 'learning_rate': 0.00010625520936593375, 'epoch': 11.33}
+
 81%|████████  | 8650/10682 [1:13:38<16:44,  2.02it/s]
 81%|████████  | 8651/10682 [1:13:38<16:43,  2.02it/s]
 81%|████████  | 8652/10682 [1:13:39<16:43,  2.02it/s]
 81%|████████  | 8653/10682 [1:13:39<16:42,  2.02it/s]
 81%|████████  | 8654/10682 [1:13:40<16:42,  2.02it/s]
 81%|████████  | 8655/10682 [1:13:40<16:41,  2.02it/s]
 81%|████████  | 8656/10682 [1:13:41<16:41,  2.02it/s]
 81%|████████  | 8657/10682 [1:13:41<16:40,  2.02it/s]
 81%|████████  | 8658/10682 [1:13:42<16:41,  2.02it/s]
 81%|████████  | 8659/10682 [1:13:42<16:41,  2.02it/s]
 81%|████████  | 8660/10682 [1:13:43<16:40,  2.02it/s]
 81%|████████  | 8661/10682 [1:13:43<16:38,  2.02it/s]
 81%|████████  | 8662/10682 [1:13:44<16:38,  2.02it/s]
 81%|████████  | 8663/10682 [1:13:44<16:37,  2.02it/s]
 81%|████████  | 8664/10682 [1:13:45<16:37,  2.02it/s]
 81%|████████  | 8665/10682 [1:13:45<16:35,  2.03it/s]
 81%|████████  | 8666/10682 [1:13:46<16:36,  2.02it/s]
 81%|████████  | 8667/10682 [1:13:46<16:34,  2.03it/s]
 81%|████████  | 8668/10682 [1:13:47<16:34,  2.03it/s]
 81%|████████  | 8669/10682 [1:13:47<16:34,  2.03it/s]
 81%|████████  | 8670/10682 [1:13:48<16:34,  2.02it/s]
 81%|████████  | 8671/10682 [1:13:48<16:33,  2.02it/s]
 81%|████████  | 8672/10682 [1:13:49<16:35,  2.02it/s]
 81%|████████  | 8673/10682 [1:13:49<16:33,  2.02it/s]
 81%|████████  | 8674/10682 [1:13:50<16:32,  2.02it/s]
 81%|████████  | 8675/10682 [1:13:50<16:31,  2.02it/s]{'loss': 2.7822, 'grad_norm': 0.2749269902706146, 'learning_rate': 0.0001037506276287885, 'epoch': 11.36}
+
                                                      

 81%|████████  | 8675/10682 [1:13:50<16:31,  2.02it/s]
 81%|████████  | 8676/10682 [1:13:51<16:37,  2.01it/s]
 81%|████████  | 8677/10682 [1:13:51<16:34,  2.02it/s]
 81%|████████  | 8678/10682 [1:13:52<16:34,  2.01it/s]
 81%|████████  | 8679/10682 [1:13:52<16:33,  2.02it/s]
 81%|████████▏ | 8680/10682 [1:13:53<16:31,  2.02it/s]
 81%|████████▏ | 8681/10682 [1:13:53<16:30,  2.02it/s]
 81%|████████▏ | 8682/10682 [1:13:54<16:29,  2.02it/s]
 81%|████████▏ | 8683/10682 [1:13:54<16:28,  2.02it/s]
 81%|████████▏ | 8684/10682 [1:13:55<16:28,  2.02it/s]
 81%|████████▏ | 8685/10682 [1:13:55<16:27,  2.02it/s]
 81%|████████▏ | 8686/10682 [1:13:56<16:26,  2.02it/s]
 81%|████████▏ | 8687/10682 [1:13:56<16:25,  2.02it/s]
 81%|████████▏ | 8688/10682 [1:13:57<16:26,  2.02it/s]
 81%|████████▏ | 8689/10682 [1:13:57<16:26,  2.02it/s]
 81%|████████▏ | 8690/10682 [1:13:58<16:26,  2.02it/s]
 81%|████████▏ | 8691/10682 [1:13:58<16:26,  2.02it/s]
 81%|████████▏ | 8692/10682 [1:13:59<16:25,  2.02it/s]
 81%|████████▏ | 8693/10682 [1:13:59<16:24,  2.02it/s]
 81%|████████▏ | 8694/10682 [1:14:00<16:23,  2.02it/s]
 81%|████████▏ | 8695/10682 [1:14:00<16:24,  2.02it/s]
 81%|████████▏ | 8696/10682 [1:14:01<16:23,  2.02it/s]
 81%|████████▏ | 8697/10682 [1:14:01<16:22,  2.02it/s]
 81%|████████▏ | 8698/10682 [1:14:02<16:21,  2.02it/s]
 81%|████████▏ | 8699/10682 [1:14:02<16:22,  2.02it/s]
 81%|████████▏ | 8700/10682 [1:14:03<16:20,  2.02it/s]{'loss': 2.7891, 'grad_norm': 0.27783605456352234, 'learning_rate': 0.0001012724960361826, 'epoch': 11.39}
+
                                                      

 81%|████████▏ | 8700/10682 [1:14:03<16:20,  2.02it/s]
 81%|████████▏ | 8701/10682 [1:14:03<16:22,  2.02it/s]
 81%|████████▏ | 8702/10682 [1:14:04<16:20,  2.02it/s]
 81%|████████▏ | 8703/10682 [1:14:04<16:21,  2.02it/s]
 81%|████████▏ | 8704/10682 [1:14:05<16:19,  2.02it/s]
 81%|████████▏ | 8705/10682 [1:14:05<16:19,  2.02it/s]
 82%|████████▏ | 8706/10682 [1:14:06<16:17,  2.02it/s]
 82%|████████▏ | 8707/10682 [1:14:06<16:17,  2.02it/s]
 82%|████████▏ | 8708/10682 [1:14:07<16:15,  2.02it/s]
 82%|████████▏ | 8709/10682 [1:14:07<16:17,  2.02it/s]
 82%|████████▏ | 8710/10682 [1:14:08<16:16,  2.02it/s]
 82%|████████▏ | 8711/10682 [1:14:08<16:16,  2.02it/s]
 82%|████████▏ | 8712/10682 [1:14:09<16:16,  2.02it/s]
 82%|████████▏ | 8713/10682 [1:14:09<16:15,  2.02it/s]
 82%|████████▏ | 8714/10682 [1:14:10<16:15,  2.02it/s]
 82%|████████▏ | 8715/10682 [1:14:10<16:13,  2.02it/s]
 82%|████████▏ | 8716/10682 [1:14:10<16:12,  2.02it/s]
 82%|████████▏ | 8717/10682 [1:14:11<16:12,  2.02it/s]
 82%|████████▏ | 8718/10682 [1:14:11<16:13,  2.02it/s]
 82%|████████▏ | 8719/10682 [1:14:12<16:13,  2.02it/s]
 82%|████████▏ | 8720/10682 [1:14:12<16:13,  2.02it/s]
 82%|████████▏ | 8721/10682 [1:14:13<16:11,  2.02it/s]
 82%|████████▏ | 8722/10682 [1:14:13<16:11,  2.02it/s]
 82%|████████▏ | 8723/10682 [1:14:14<16:10,  2.02it/s]
 82%|████████▏ | 8724/10682 [1:14:14<16:10,  2.02it/s]
 82%|████████▏ | 8725/10682 [1:14:15<16:09,  2.02it/s]{'loss': 2.7834, 'grad_norm': 0.27818813920021057, 'learning_rate': 9.882098000652034e-05, 'epoch': 11.43}
+
                                                      

 82%|████████▏ | 8725/10682 [1:14:15<16:09,  2.02it/s]
 82%|████████▏ | 8726/10682 [1:14:15<16:09,  2.02it/s]
 82%|████████▏ | 8727/10682 [1:14:16<16:09,  2.02it/s]
 82%|████████▏ | 8728/10682 [1:14:16<16:09,  2.02it/s]
 82%|████████▏ | 8729/10682 [1:14:17<16:08,  2.02it/s]
 82%|████████▏ | 8730/10682 [1:14:17<16:08,  2.02it/s]
 82%|████████▏ | 8731/10682 [1:14:18<16:07,  2.02it/s]
 82%|████████▏ | 8732/10682 [1:14:18<16:05,  2.02it/s]
 82%|████████▏ | 8733/10682 [1:14:19<16:05,  2.02it/s]
 82%|████████▏ | 8734/10682 [1:14:19<16:03,  2.02it/s]
 82%|████████▏ | 8735/10682 [1:14:20<16:04,  2.02it/s]
 82%|████████▏ | 8736/10682 [1:14:20<16:03,  2.02it/s]
 82%|████████▏ | 8737/10682 [1:14:21<16:03,  2.02it/s]
 82%|████████▏ | 8738/10682 [1:14:21<16:02,  2.02it/s]
 82%|████████▏ | 8739/10682 [1:14:22<16:02,  2.02it/s]
 82%|████████▏ | 8740/10682 [1:14:22<16:01,  2.02it/s]
 82%|████████▏ | 8741/10682 [1:14:23<16:01,  2.02it/s]
 82%|████████▏ | 8742/10682 [1:14:23<15:59,  2.02it/s]
 82%|████████▏ | 8743/10682 [1:14:24<15:59,  2.02it/s]
 82%|████████▏ | 8744/10682 [1:14:24<15:58,  2.02it/s]
 82%|████████▏ | 8745/10682 [1:14:25<15:58,  2.02it/s]
 82%|████████▏ | 8746/10682 [1:14:25<15:57,  2.02it/s]
 82%|████████▏ | 8747/10682 [1:14:26<15:57,  2.02it/s]
 82%|████████▏ | 8748/10682 [1:14:26<15:56,  2.02it/s]
 82%|████████▏ | 8749/10682 [1:14:27<15:56,  2.02it/s]
 82%|████████▏ | 8750/10682 [1:14:27<15:56,  2.02it/s]{'loss': 2.796, 'grad_norm': 0.27351343631744385, 'learning_rate': 9.639624318158335e-05, 'epoch': 11.46}
+
                                                      

 82%|████████▏ | 8750/10682 [1:14:27<15:56,  2.02it/s]
 82%|████████▏ | 8751/10682 [1:14:28<16:00,  2.01it/s]
 82%|████████▏ | 8752/10682 [1:14:28<15:59,  2.01it/s]
 82%|████████▏ | 8753/10682 [1:14:29<15:58,  2.01it/s]
 82%|████████▏ | 8754/10682 [1:14:29<15:57,  2.01it/s]
 82%|████████▏ | 8755/10682 [1:14:30<15:54,  2.02it/s]
 82%|████████▏ | 8756/10682 [1:14:30<15:54,  2.02it/s]
 82%|████████▏ | 8757/10682 [1:14:31<15:54,  2.02it/s]
 82%|████████▏ | 8758/10682 [1:14:31<15:52,  2.02it/s]
 82%|████████▏ | 8759/10682 [1:14:32<15:52,  2.02it/s]
 82%|████████▏ | 8760/10682 [1:14:32<15:51,  2.02it/s]
 82%|████████▏ | 8761/10682 [1:14:33<15:50,  2.02it/s]
 82%|████████▏ | 8762/10682 [1:14:33<15:50,  2.02it/s]
 82%|████████▏ | 8763/10682 [1:14:34<15:51,  2.02it/s]
 82%|████████▏ | 8764/10682 [1:14:34<15:49,  2.02it/s]
 82%|████████▏ | 8765/10682 [1:14:35<15:49,  2.02it/s]
 82%|████████▏ | 8766/10682 [1:14:35<15:48,  2.02it/s]
 82%|████████▏ | 8767/10682 [1:14:36<15:48,  2.02it/s]
 82%|████████▏ | 8768/10682 [1:14:36<15:48,  2.02it/s]
 82%|████████▏ | 8769/10682 [1:14:37<15:48,  2.02it/s]
 82%|████████▏ | 8770/10682 [1:14:37<15:46,  2.02it/s]
 82%|████████▏ | 8771/10682 [1:14:38<15:46,  2.02it/s]
 82%|████████▏ | 8772/10682 [1:14:38<15:45,  2.02it/s]
 82%|████████▏ | 8773/10682 [1:14:39<15:44,  2.02it/s]
 82%|████████▏ | 8774/10682 [1:14:39<15:45,  2.02it/s]
 82%|████████▏ | 8775/10682 [1:14:40<15:44,  2.02it/s]{'loss': 2.7862, 'grad_norm': 0.2807120978832245, 'learning_rate': 9.399844741560781e-05, 'epoch': 11.49}
+
                                                      

 82%|████████▏ | 8775/10682 [1:14:40<15:44,  2.02it/s]
 82%|████████▏ | 8776/10682 [1:14:40<15:44,  2.02it/s]
 82%|████████▏ | 8777/10682 [1:14:41<15:43,  2.02it/s]
 82%|████████▏ | 8778/10682 [1:14:41<15:43,  2.02it/s]
 82%|████████▏ | 8779/10682 [1:14:42<15:43,  2.02it/s]
 82%|████████▏ | 8780/10682 [1:14:42<15:43,  2.02it/s]
 82%|████████▏ | 8781/10682 [1:14:43<15:43,  2.02it/s]
 82%|████████▏ | 8782/10682 [1:14:43<15:41,  2.02it/s]
 82%|████████▏ | 8783/10682 [1:14:44<15:41,  2.02it/s]
 82%|████████▏ | 8784/10682 [1:14:44<15:41,  2.02it/s]
 82%|████████▏ | 8785/10682 [1:14:45<15:40,  2.02it/s]
 82%|████████▏ | 8786/10682 [1:14:45<15:39,  2.02it/s]
 82%|████████▏ | 8787/10682 [1:14:46<15:39,  2.02it/s]
 82%|████████▏ | 8788/10682 [1:14:46<15:38,  2.02it/s]
 82%|████████▏ | 8789/10682 [1:14:47<15:37,  2.02it/s]
 82%|████████▏ | 8790/10682 [1:14:47<15:37,  2.02it/s]
 82%|████████▏ | 8791/10682 [1:14:48<15:35,  2.02it/s]
 82%|████████▏ | 8792/10682 [1:14:48<15:34,  2.02it/s]
 82%|████████▏ | 8793/10682 [1:14:49<15:34,  2.02it/s]
 82%|████████▏ | 8794/10682 [1:14:49<15:33,  2.02it/s]
 82%|████████▏ | 8795/10682 [1:14:50<15:34,  2.02it/s]
 82%|████████▏ | 8796/10682 [1:14:50<15:33,  2.02it/s]
 82%|████████▏ | 8797/10682 [1:14:51<15:33,  2.02it/s]
 82%|████████▏ | 8798/10682 [1:14:51<15:32,  2.02it/s]
 82%|████████▏ | 8799/10682 [1:14:52<15:32,  2.02it/s]
 82%|████████▏ | 8800/10682 [1:14:52<15:31,  2.02it/s]{'loss': 2.8095, 'grad_norm': 0.27686017751693726, 'learning_rate': 9.162775276448015e-05, 'epoch': 11.53}
+
                                                      

 82%|████████▏ | 8800/10682 [1:14:52<15:31,  2.02it/s]
 82%|████████▏ | 8801/10682 [1:14:53<15:32,  2.02it/s]
 82%|████████▏ | 8802/10682 [1:14:53<15:31,  2.02it/s]
 82%|████████▏ | 8803/10682 [1:14:54<15:30,  2.02it/s]
 82%|████████▏ | 8804/10682 [1:14:54<15:30,  2.02it/s]
 82%|████████▏ | 8805/10682 [1:14:55<15:28,  2.02it/s]
 82%|████████▏ | 8806/10682 [1:14:55<15:28,  2.02it/s]
 82%|████████▏ | 8807/10682 [1:14:56<15:28,  2.02it/s]
 82%|████████▏ | 8808/10682 [1:14:56<15:27,  2.02it/s]
 82%|████████▏ | 8809/10682 [1:14:57<15:26,  2.02it/s]
 82%|████████▏ | 8810/10682 [1:14:57<15:26,  2.02it/s]
 82%|████████▏ | 8811/10682 [1:14:58<15:26,  2.02it/s]
 82%|████████▏ | 8812/10682 [1:14:58<15:25,  2.02it/s]
 83%|████████▎ | 8813/10682 [1:14:59<15:24,  2.02it/s]
 83%|████████▎ | 8814/10682 [1:14:59<15:23,  2.02it/s]
 83%|████████▎ | 8815/10682 [1:15:00<15:23,  2.02it/s]
 83%|████████▎ | 8816/10682 [1:15:00<15:23,  2.02it/s]
 83%|████████▎ | 8817/10682 [1:15:01<15:23,  2.02it/s]
 83%|████████▎ | 8818/10682 [1:15:01<15:23,  2.02it/s]
 83%|████████▎ | 8819/10682 [1:15:02<15:21,  2.02it/s]
 83%|████████▎ | 8820/10682 [1:15:02<15:20,  2.02it/s]
 83%|████████▎ | 8821/10682 [1:15:02<15:19,  2.02it/s]
 83%|████████▎ | 8822/10682 [1:15:03<15:19,  2.02it/s]
 83%|████████▎ | 8823/10682 [1:15:04<16:39,  1.86it/s]
 83%|████████▎ | 8824/10682 [1:15:04<16:13,  1.91it/s]
 83%|████████▎ | 8825/10682 [1:15:05<15:56,  1.94it/s]
                                                      
{'loss': 2.7909, 'grad_norm': 0.28055277466773987, 'learning_rate': 8.928431747505355e-05, 'epoch': 11.56}
+
 83%|████████▎ | 8825/10682 [1:15:05<15:56,  1.94it/s]
 83%|████████▎ | 8826/10682 [1:15:05<15:48,  1.96it/s]
 83%|████████▎ | 8827/10682 [1:15:06<15:39,  1.97it/s]
 83%|████████▎ | 8828/10682 [1:15:06<15:31,  1.99it/s]
 83%|████████▎ | 8829/10682 [1:15:07<15:26,  2.00it/s]
 83%|████████▎ | 8830/10682 [1:15:07<15:23,  2.01it/s]
 83%|████████▎ | 8831/10682 [1:15:08<15:21,  2.01it/s]
 83%|████████▎ | 8832/10682 [1:15:08<15:19,  2.01it/s]
 83%|████████▎ | 8833/10682 [1:15:09<15:17,  2.02it/s]
 83%|████████▎ | 8834/10682 [1:15:09<15:15,  2.02it/s]
 83%|████████▎ | 8835/10682 [1:15:10<15:14,  2.02it/s]
 83%|████████▎ | 8836/10682 [1:15:10<15:14,  2.02it/s]
 83%|████████▎ | 8837/10682 [1:15:11<15:13,  2.02it/s]
 83%|████████▎ | 8838/10682 [1:15:11<15:12,  2.02it/s]
 83%|████████▎ | 8839/10682 [1:15:12<15:11,  2.02it/s]
 83%|████████▎ | 8840/10682 [1:15:12<15:11,  2.02it/s]
 83%|████████▎ | 8841/10682 [1:15:13<15:10,  2.02it/s]
 83%|████████▎ | 8842/10682 [1:15:13<15:09,  2.02it/s]
 83%|████████▎ | 8843/10682 [1:15:14<15:09,  2.02it/s]
 83%|████████▎ | 8844/10682 [1:15:14<15:09,  2.02it/s]
 83%|████████▎ | 8845/10682 [1:15:15<15:08,  2.02it/s]
 83%|████████▎ | 8846/10682 [1:15:15<15:07,  2.02it/s]
 83%|████████▎ | 8847/10682 [1:15:16<15:06,  2.02it/s]
 83%|████████▎ | 8848/10682 [1:15:16<15:05,  2.02it/s]
 83%|████████▎ | 8849/10682 [1:15:16<15:04,  2.03it/s]
 83%|████████▎ | 8850/10682 [1:15:17<15:05,  2.02it/s]
                                                      
{'loss': 2.7797, 'grad_norm': 0.27887654304504395, 'learning_rate': 8.6968297974584e-05, 'epoch': 11.59}
+
 83%|████████▎ | 8850/10682 [1:15:17<15:05,  2.02it/s]
 83%|████████▎ | 8851/10682 [1:15:17<15:08,  2.02it/s]
 83%|████████▎ | 8852/10682 [1:15:18<15:06,  2.02it/s]
 83%|████████▎ | 8853/10682 [1:15:18<15:04,  2.02it/s]
 83%|████████▎ | 8854/10682 [1:15:19<15:04,  2.02it/s]
 83%|████████▎ | 8855/10682 [1:15:19<15:02,  2.02it/s]
 83%|████████▎ | 8856/10682 [1:15:20<15:02,  2.02it/s]
 83%|████████▎ | 8857/10682 [1:15:20<15:01,  2.02it/s]
 83%|████████▎ | 8858/10682 [1:15:21<15:01,  2.02it/s]
 83%|████████▎ | 8859/10682 [1:15:21<15:00,  2.03it/s]
 83%|████████▎ | 8860/10682 [1:15:22<15:00,  2.02it/s]
 83%|████████▎ | 8861/10682 [1:15:22<14:59,  2.02it/s]
 83%|████████▎ | 8862/10682 [1:15:23<14:58,  2.03it/s]
 83%|████████▎ | 8863/10682 [1:15:23<14:58,  2.02it/s]
 83%|████████▎ | 8864/10682 [1:15:24<14:56,  2.03it/s]
 83%|████████▎ | 8865/10682 [1:15:24<14:56,  2.03it/s]
 83%|████████▎ | 8866/10682 [1:15:25<14:55,  2.03it/s]
 83%|████████▎ | 8867/10682 [1:15:25<14:55,  2.03it/s]
 83%|████████▎ | 8868/10682 [1:15:26<14:55,  2.03it/s]
 83%|████████▎ | 8869/10682 [1:15:26<14:55,  2.03it/s]
 83%|████████▎ | 8870/10682 [1:15:27<14:55,  2.02it/s]
 83%|████████▎ | 8871/10682 [1:15:27<14:53,  2.03it/s]
 83%|████████▎ | 8872/10682 [1:15:28<14:53,  2.02it/s]
 83%|████████▎ | 8873/10682 [1:15:28<14:53,  2.03it/s]
 83%|████████▎ | 8874/10682 [1:15:29<14:53,  2.02it/s]
 83%|████████▎ | 8875/10682 [1:15:29<14:52,  2.03it/s]{'loss': 2.7982, 'grad_norm': 0.2797921597957611, 'learning_rate': 8.467984886028967e-05, 'epoch': 11.62}
+
                                                      

 83%|████████▎ | 8875/10682 [1:15:29<14:52,  2.03it/s]
 83%|████████▎ | 8876/10682 [1:15:30<14:53,  2.02it/s]
 83%|████████▎ | 8877/10682 [1:15:30<14:52,  2.02it/s]
 83%|████████▎ | 8878/10682 [1:15:31<14:52,  2.02it/s]
 83%|████████▎ | 8879/10682 [1:15:31<14:50,  2.02it/s]
 83%|████████▎ | 8880/10682 [1:15:32<14:50,  2.02it/s]
 83%|████████▎ | 8881/10682 [1:15:32<14:50,  2.02it/s]
 83%|████████▎ | 8882/10682 [1:15:33<14:50,  2.02it/s]
 83%|████████▎ | 8883/10682 [1:15:33<14:49,  2.02it/s]
 83%|████████▎ | 8884/10682 [1:15:34<14:49,  2.02it/s]
 83%|████████▎ | 8885/10682 [1:15:34<14:48,  2.02it/s]
 83%|████████▎ | 8886/10682 [1:15:35<14:48,  2.02it/s]
 83%|████████▎ | 8887/10682 [1:15:35<14:47,  2.02it/s]
 83%|████████▎ | 8888/10682 [1:15:36<14:47,  2.02it/s]
 83%|████████▎ | 8889/10682 [1:15:36<14:46,  2.02it/s]
 83%|████████▎ | 8890/10682 [1:15:37<14:46,  2.02it/s]
 83%|████████▎ | 8891/10682 [1:15:37<14:45,  2.02it/s]
 83%|████████▎ | 8892/10682 [1:15:38<14:45,  2.02it/s]
 83%|████████▎ | 8893/10682 [1:15:38<14:44,  2.02it/s]
 83%|████████▎ | 8894/10682 [1:15:39<14:43,  2.02it/s]
 83%|████████▎ | 8895/10682 [1:15:39<14:43,  2.02it/s]
 83%|████████▎ | 8896/10682 [1:15:40<14:43,  2.02it/s]
 83%|████████▎ | 8897/10682 [1:15:40<14:43,  2.02it/s]
 83%|████████▎ | 8898/10682 [1:15:41<14:43,  2.02it/s]
 83%|████████▎ | 8899/10682 [1:15:41<14:41,  2.02it/s]
 83%|████████▎ | 8900/10682 [1:15:42<14:42,  2.02it/s]{'loss': 2.8015, 'grad_norm': 0.27951231598854065, 'learning_rate': 8.24191228890303e-05, 'epoch': 11.66}
+
                                                      

 83%|████████▎ | 8900/10682 [1:15:42<14:42,  2.02it/s]
 83%|████████▎ | 8901/10682 [1:15:42<14:42,  2.02it/s]
 83%|████████▎ | 8902/10682 [1:15:43<14:41,  2.02it/s]
 83%|████████▎ | 8903/10682 [1:15:43<14:40,  2.02it/s]
 83%|████████▎ | 8904/10682 [1:15:44<14:40,  2.02it/s]
 83%|████████▎ | 8905/10682 [1:15:44<14:39,  2.02it/s]
 83%|████████▎ | 8906/10682 [1:15:45<14:38,  2.02it/s]
 83%|████████▎ | 8907/10682 [1:15:45<14:38,  2.02it/s]
 83%|████████▎ | 8908/10682 [1:15:46<14:37,  2.02it/s]
 83%|████████▎ | 8909/10682 [1:15:46<14:37,  2.02it/s]
 83%|████████▎ | 8910/10682 [1:15:47<14:36,  2.02it/s]
 83%|████████▎ | 8911/10682 [1:15:47<14:34,  2.02it/s]
 83%|████████▎ | 8912/10682 [1:15:48<14:34,  2.02it/s]
 83%|████████▎ | 8913/10682 [1:15:48<14:33,  2.03it/s]
 83%|████████▎ | 8914/10682 [1:15:49<14:33,  2.02it/s]
 83%|████████▎ | 8915/10682 [1:15:49<14:33,  2.02it/s]
 83%|████████▎ | 8916/10682 [1:15:50<14:32,  2.02it/s]
 83%|████████▎ | 8917/10682 [1:15:50<14:32,  2.02it/s]
 83%|████████▎ | 8918/10682 [1:15:51<14:32,  2.02it/s]
 83%|████████▎ | 8919/10682 [1:15:51<14:32,  2.02it/s]
 84%|████████▎ | 8920/10682 [1:15:52<14:32,  2.02it/s]
 84%|████████▎ | 8921/10682 [1:15:52<14:31,  2.02it/s]
 84%|████████▎ | 8922/10682 [1:15:53<14:30,  2.02it/s]
 84%|████████▎ | 8923/10682 [1:15:53<14:29,  2.02it/s]
 84%|████████▎ | 8924/10682 [1:15:54<14:29,  2.02it/s]
 84%|████████▎ | 8925/10682 [1:15:54<14:28,  2.02it/s]{'loss': 2.8067, 'grad_norm': 0.27667585015296936, 'learning_rate': 8.018627096711106e-05, 'epoch': 11.69}
+
                                                      

 84%|████████▎ | 8925/10682 [1:15:54<14:28,  2.02it/s]
 84%|████████▎ | 8926/10682 [1:15:55<14:29,  2.02it/s]
 84%|████████▎ | 8927/10682 [1:15:55<14:28,  2.02it/s]
 84%|████████▎ | 8928/10682 [1:15:56<14:28,  2.02it/s]
 84%|████████▎ | 8929/10682 [1:15:56<14:26,  2.02it/s]
 84%|████████▎ | 8930/10682 [1:15:57<14:26,  2.02it/s]
 84%|████████▎ | 8931/10682 [1:15:57<14:25,  2.02it/s]
 84%|████████▎ | 8932/10682 [1:15:58<15:39,  1.86it/s]
 84%|████████▎ | 8933/10682 [1:15:58<15:16,  1.91it/s]
 84%|████████▎ | 8934/10682 [1:15:59<15:00,  1.94it/s]
 84%|████████▎ | 8935/10682 [1:15:59<14:49,  1.96it/s]
 84%|████████▎ | 8936/10682 [1:16:00<14:41,  1.98it/s]
 84%|████████▎ | 8937/10682 [1:16:00<14:35,  1.99it/s]
 84%|████████▎ | 8938/10682 [1:16:01<14:31,  2.00it/s]
 84%|████████▎ | 8939/10682 [1:16:01<14:28,  2.01it/s]
 84%|████████▎ | 8940/10682 [1:16:02<14:26,  2.01it/s]
 84%|████████▎ | 8941/10682 [1:16:02<14:24,  2.01it/s]
 84%|████████▎ | 8942/10682 [1:16:03<14:22,  2.02it/s]
 84%|████████▎ | 8943/10682 [1:16:03<14:21,  2.02it/s]
 84%|████████▎ | 8944/10682 [1:16:04<14:20,  2.02it/s]
 84%|████████▎ | 8945/10682 [1:16:04<14:19,  2.02it/s]
 84%|████████▎ | 8946/10682 [1:16:05<14:18,  2.02it/s]
 84%|████████▍ | 8947/10682 [1:16:05<14:17,  2.02it/s]
 84%|████████▍ | 8948/10682 [1:16:06<14:18,  2.02it/s]
 84%|████████▍ | 8949/10682 [1:16:06<14:17,  2.02it/s]
 84%|████████▍ | 8950/10682 [1:16:07<14:16,  2.02it/s]
                                                      
{'loss': 2.7831, 'grad_norm': 0.2778463661670685, 'learning_rate': 7.798144214020909e-05, 'epoch': 11.72}
+
 84%|████████▍ | 8950/10682 [1:16:07<14:16,  2.02it/s]
 84%|████████▍ | 8951/10682 [1:16:07<14:16,  2.02it/s]
 84%|████████▍ | 8952/10682 [1:16:08<14:16,  2.02it/s]
 84%|████████▍ | 8953/10682 [1:16:08<14:14,  2.02it/s]
 84%|████████▍ | 8954/10682 [1:16:09<14:15,  2.02it/s]
 84%|████████▍ | 8955/10682 [1:16:09<14:14,  2.02it/s]
 84%|████████▍ | 8956/10682 [1:16:10<14:13,  2.02it/s]
 84%|████████▍ | 8957/10682 [1:16:10<14:12,  2.02it/s]
 84%|████████▍ | 8958/10682 [1:16:11<14:12,  2.02it/s]
 84%|████████▍ | 8959/10682 [1:16:11<14:11,  2.02it/s]
 84%|████████▍ | 8960/10682 [1:16:12<14:11,  2.02it/s]
 84%|████████▍ | 8961/10682 [1:16:12<14:11,  2.02it/s]
 84%|████████▍ | 8962/10682 [1:16:13<14:11,  2.02it/s]
 84%|████████▍ | 8963/10682 [1:16:13<14:09,  2.02it/s]
 84%|████████▍ | 8964/10682 [1:16:13<14:09,  2.02it/s]
 84%|████████▍ | 8965/10682 [1:16:14<14:08,  2.02it/s]
 84%|████████▍ | 8966/10682 [1:16:14<14:07,  2.02it/s]
 84%|████████▍ | 8967/10682 [1:16:15<14:07,  2.02it/s]
 84%|████████▍ | 8968/10682 [1:16:15<14:07,  2.02it/s]
 84%|████████▍ | 8969/10682 [1:16:16<14:06,  2.02it/s]
 84%|████████▍ | 8970/10682 [1:16:16<14:05,  2.03it/s]
 84%|████████▍ | 8971/10682 [1:16:17<14:04,  2.03it/s]
 84%|████████▍ | 8972/10682 [1:16:17<14:04,  2.02it/s]
 84%|████████▍ | 8973/10682 [1:16:18<14:04,  2.02it/s]
 84%|████████▍ | 8974/10682 [1:16:18<14:03,  2.02it/s]
 84%|████████▍ | 8975/10682 [1:16:19<14:03,  2.02it/s]{'loss': 2.8038, 'grad_norm': 0.2781047523021698, 'learning_rate': 7.58047835834249e-05, 'epoch': 11.76}
                                                      
+
 84%|████████▍ | 8975/10682 [1:16:19<14:03,  2.02it/s]
 84%|████████▍ | 8976/10682 [1:16:19<14:07,  2.01it/s]
 84%|████████▍ | 8977/10682 [1:16:20<14:05,  2.02it/s]
 84%|████████▍ | 8978/10682 [1:16:20<14:04,  2.02it/s]
 84%|████████▍ | 8979/10682 [1:16:21<14:02,  2.02it/s]
 84%|████████▍ | 8980/10682 [1:16:21<14:02,  2.02it/s]
 84%|████████▍ | 8981/10682 [1:16:22<14:01,  2.02it/s]
 84%|████████▍ | 8982/10682 [1:16:22<14:00,  2.02it/s]
 84%|████████▍ | 8983/10682 [1:16:23<14:00,  2.02it/s]
 84%|████████▍ | 8984/10682 [1:16:23<13:59,  2.02it/s]
 84%|████████▍ | 8985/10682 [1:16:24<13:59,  2.02it/s]
 84%|████████▍ | 8986/10682 [1:16:24<13:58,  2.02it/s]
 84%|████████▍ | 8987/10682 [1:16:25<13:57,  2.02it/s]
 84%|████████▍ | 8988/10682 [1:16:25<13:56,  2.02it/s]
 84%|████████▍ | 8989/10682 [1:16:26<13:56,  2.02it/s]
 84%|████████▍ | 8990/10682 [1:16:26<13:55,  2.02it/s]
 84%|████████▍ | 8991/10682 [1:16:27<13:55,  2.02it/s]
 84%|████████▍ | 8992/10682 [1:16:27<13:54,  2.02it/s]
 84%|████████▍ | 8993/10682 [1:16:28<13:55,  2.02it/s]
 84%|████████▍ | 8994/10682 [1:16:28<13:54,  2.02it/s]
 84%|████████▍ | 8995/10682 [1:16:29<13:53,  2.02it/s]
 84%|████████▍ | 8996/10682 [1:16:29<13:53,  2.02it/s]
 84%|████████▍ | 8997/10682 [1:16:30<13:51,  2.03it/s]
 84%|████████▍ | 8998/10682 [1:16:30<13:52,  2.02it/s]
 84%|████████▍ | 8999/10682 [1:16:31<13:51,  2.02it/s]
 84%|████████▍ | 9000/10682 [1:16:31<13:51,  2.02it/s]
                                                      
{'loss': 2.8144, 'grad_norm': 0.27680227160453796, 'learning_rate': 7.365644059145782e-05, 'epoch': 11.79}
+
 84%|████████▍ | 9000/10682 [1:16:31<13:51,  2.02it/s]
 84%|████████▍ | 9001/10682 [1:16:32<13:50,  2.02it/s]
 84%|████████▍ | 9002/10682 [1:16:32<13:50,  2.02it/s]
 84%|████████▍ | 9003/10682 [1:16:33<13:49,  2.02it/s]
 84%|████████▍ | 9004/10682 [1:16:33<13:49,  2.02it/s]
 84%|████████▍ | 9005/10682 [1:16:34<13:48,  2.02it/s]
 84%|████████▍ | 9006/10682 [1:16:34<13:48,  2.02it/s]
 84%|████████▍ | 9007/10682 [1:16:35<13:47,  2.02it/s]
 84%|████████▍ | 9008/10682 [1:16:35<13:46,  2.02it/s]
 84%|████████▍ | 9009/10682 [1:16:36<13:46,  2.03it/s]
 84%|████████▍ | 9010/10682 [1:16:36<13:44,  2.03it/s]
 84%|████████▍ | 9011/10682 [1:16:37<13:44,  2.03it/s]
 84%|████████▍ | 9012/10682 [1:16:37<13:43,  2.03it/s]
 84%|████████▍ | 9013/10682 [1:16:38<13:42,  2.03it/s]
 84%|████████▍ | 9014/10682 [1:16:38<13:42,  2.03it/s]
 84%|████████▍ | 9015/10682 [1:16:39<13:42,  2.03it/s]
 84%|████████▍ | 9016/10682 [1:16:39<13:42,  2.03it/s]
 84%|████████▍ | 9017/10682 [1:16:40<13:41,  2.03it/s]
 84%|████████▍ | 9018/10682 [1:16:40<13:41,  2.03it/s]
 84%|████████▍ | 9019/10682 [1:16:41<13:40,  2.03it/s]
 84%|████████▍ | 9020/10682 [1:16:41<13:41,  2.02it/s]
 84%|████████▍ | 9021/10682 [1:16:42<13:40,  2.02it/s]
 84%|████████▍ | 9022/10682 [1:16:42<13:40,  2.02it/s]
 84%|████████▍ | 9023/10682 [1:16:43<13:39,  2.02it/s]
 84%|████████▍ | 9024/10682 [1:16:43<13:39,  2.02it/s]
 84%|████████▍ | 9025/10682 [1:16:44<13:38,  2.02it/s]
                                                      
{'loss': 2.8089, 'grad_norm': 0.27644211053848267, 'learning_rate': 7.153655656890773e-05, 'epoch': 11.82}
+
 84%|████████▍ | 9025/10682 [1:16:44<13:38,  2.02it/s]
 84%|████████▍ | 9026/10682 [1:16:44<13:43,  2.01it/s]
 85%|████████▍ | 9027/10682 [1:16:45<13:41,  2.02it/s]
 85%|████████▍ | 9028/10682 [1:16:45<13:39,  2.02it/s]
 85%|████████▍ | 9029/10682 [1:16:46<13:37,  2.02it/s]
 85%|████████▍ | 9030/10682 [1:16:46<13:37,  2.02it/s]
 85%|████████▍ | 9031/10682 [1:16:47<13:36,  2.02it/s]
 85%|████████▍ | 9032/10682 [1:16:47<13:35,  2.02it/s]
 85%|████████▍ | 9033/10682 [1:16:48<13:34,  2.02it/s]
 85%|████████▍ | 9034/10682 [1:16:48<13:34,  2.02it/s]
 85%|████████▍ | 9035/10682 [1:16:49<13:33,  2.02it/s]
 85%|████████▍ | 9036/10682 [1:16:49<13:32,  2.03it/s]
 85%|████████▍ | 9037/10682 [1:16:50<13:32,  2.02it/s]
 85%|████████▍ | 9038/10682 [1:16:50<13:31,  2.03it/s]
 85%|████████▍ | 9039/10682 [1:16:51<13:32,  2.02it/s]
 85%|████████▍ | 9040/10682 [1:16:51<13:30,  2.02it/s]
 85%|████████▍ | 9041/10682 [1:16:52<13:31,  2.02it/s]
 85%|████████▍ | 9042/10682 [1:16:52<13:31,  2.02it/s]
 85%|████████▍ | 9043/10682 [1:16:53<13:30,  2.02it/s]
 85%|████████▍ | 9044/10682 [1:16:53<13:30,  2.02it/s]
 85%|████████▍ | 9045/10682 [1:16:54<13:29,  2.02it/s]
 85%|████████▍ | 9046/10682 [1:16:54<13:28,  2.02it/s]
 85%|████████▍ | 9047/10682 [1:16:55<13:28,  2.02it/s]
 85%|████████▍ | 9048/10682 [1:16:55<13:28,  2.02it/s]
 85%|████████▍ | 9049/10682 [1:16:56<13:27,  2.02it/s]
 85%|████████▍ | 9050/10682 [1:16:56<13:27,  2.02it/s]
                                                      {'loss': 2.7967, 'grad_norm': 0.2740754783153534, 'learning_rate': 6.94452730207023e-05, 'epoch': 11.85}
+

 85%|████████▍ | 9050/10682 [1:16:56<13:27,  2.02it/s]
 85%|████████▍ | 9051/10682 [1:16:56<13:27,  2.02it/s]
 85%|████████▍ | 9052/10682 [1:16:57<13:26,  2.02it/s]
 85%|████████▍ | 9053/10682 [1:16:57<13:26,  2.02it/s]
 85%|████████▍ | 9054/10682 [1:16:58<13:25,  2.02it/s]
 85%|████████▍ | 9055/10682 [1:16:58<13:25,  2.02it/s]
 85%|████████▍ | 9056/10682 [1:16:59<13:24,  2.02it/s]
 85%|████████▍ | 9057/10682 [1:16:59<13:23,  2.02it/s]
 85%|████████▍ | 9058/10682 [1:17:00<13:23,  2.02it/s]
 85%|████████▍ | 9059/10682 [1:17:00<13:21,  2.02it/s]
 85%|████████▍ | 9060/10682 [1:17:01<13:21,  2.02it/s]
 85%|████████▍ | 9061/10682 [1:17:01<13:20,  2.02it/s]
 85%|████████▍ | 9062/10682 [1:17:02<13:20,  2.02it/s]
 85%|████████▍ | 9063/10682 [1:17:02<13:20,  2.02it/s]
 85%|████████▍ | 9064/10682 [1:17:03<13:19,  2.02it/s]
 85%|████████▍ | 9065/10682 [1:17:03<13:19,  2.02it/s]
 85%|████████▍ | 9066/10682 [1:17:04<13:18,  2.02it/s]
 85%|████████▍ | 9067/10682 [1:17:04<13:17,  2.03it/s]
 85%|████████▍ | 9068/10682 [1:17:05<13:17,  2.02it/s]
 85%|████████▍ | 9069/10682 [1:17:05<13:17,  2.02it/s]
 85%|████████▍ | 9070/10682 [1:17:06<13:17,  2.02it/s]
 85%|████████▍ | 9071/10682 [1:17:06<13:17,  2.02it/s]
 85%|████████▍ | 9072/10682 [1:17:07<13:16,  2.02it/s]
 85%|████████▍ | 9073/10682 [1:17:07<13:15,  2.02it/s]
 85%|████████▍ | 9074/10682 [1:17:08<13:15,  2.02it/s]
 85%|████████▍ | 9075/10682 [1:17:08<13:15,  2.02it/s]{'loss': 2.7982, 'grad_norm': 0.27378374338150024, 'learning_rate': 6.738272954265156e-05, 'epoch': 11.89}
+
                                                      

 85%|████████▍ | 9075/10682 [1:17:08<13:15,  2.02it/s]
 85%|████████▍ | 9076/10682 [1:17:09<13:15,  2.02it/s]
 85%|████████▍ | 9077/10682 [1:17:09<13:14,  2.02it/s]
 85%|████████▍ | 9078/10682 [1:17:10<13:14,  2.02it/s]
 85%|████████▍ | 9079/10682 [1:17:10<13:13,  2.02it/s]
 85%|████████▌ | 9080/10682 [1:17:11<13:12,  2.02it/s]
 85%|████████▌ | 9081/10682 [1:17:11<13:12,  2.02it/s]
 85%|████████▌ | 9082/10682 [1:17:12<13:11,  2.02it/s]
 85%|████████▌ | 9083/10682 [1:17:12<13:10,  2.02it/s]
 85%|████████▌ | 9084/10682 [1:17:13<13:10,  2.02it/s]
 85%|████████▌ | 9085/10682 [1:17:13<13:09,  2.02it/s]
 85%|████████▌ | 9086/10682 [1:17:14<13:08,  2.02it/s]
 85%|████████▌ | 9087/10682 [1:17:14<13:08,  2.02it/s]
 85%|████████▌ | 9088/10682 [1:17:15<13:08,  2.02it/s]
 85%|████████▌ | 9089/10682 [1:17:15<13:07,  2.02it/s]
 85%|████████▌ | 9090/10682 [1:17:16<13:06,  2.02it/s]
 85%|████████▌ | 9091/10682 [1:17:16<13:06,  2.02it/s]
 85%|████████▌ | 9092/10682 [1:17:17<13:06,  2.02it/s]
 85%|████████▌ | 9093/10682 [1:17:17<13:05,  2.02it/s]
 85%|████████▌ | 9094/10682 [1:17:18<13:04,  2.02it/s]
 85%|████████▌ | 9095/10682 [1:17:18<13:04,  2.02it/s]
 85%|████████▌ | 9096/10682 [1:17:19<13:03,  2.02it/s]
 85%|████████▌ | 9097/10682 [1:17:19<13:03,  2.02it/s]
 85%|████████▌ | 9098/10682 [1:17:20<13:02,  2.02it/s]
 85%|████████▌ | 9099/10682 [1:17:20<13:02,  2.02it/s]
 85%|████████▌ | 9100/10682 [1:17:21<13:01,  2.02it/s]{'loss': 2.7966, 'grad_norm': 0.27451440691947937, 'learning_rate': 6.534906381212979e-05, 'epoch': 11.92}
                                                      
+
 85%|████████▌ | 9100/10682 [1:17:21<13:01,  2.02it/s]
 85%|████████▌ | 9101/10682 [1:17:21<13:02,  2.02it/s]
 85%|████████▌ | 9102/10682 [1:17:22<13:01,  2.02it/s]
 85%|████████▌ | 9103/10682 [1:17:22<13:00,  2.02it/s]
 85%|████████▌ | 9104/10682 [1:17:23<13:00,  2.02it/s]
 85%|████████▌ | 9105/10682 [1:17:23<13:00,  2.02it/s]
 85%|████████▌ | 9106/10682 [1:17:24<13:00,  2.02it/s]
 85%|████████▌ | 9107/10682 [1:17:24<12:59,  2.02it/s]
 85%|████████▌ | 9108/10682 [1:17:25<12:58,  2.02it/s]
 85%|████████▌ | 9109/10682 [1:17:25<12:57,  2.02it/s]
 85%|████████▌ | 9110/10682 [1:17:26<12:57,  2.02it/s]
 85%|████████▌ | 9111/10682 [1:17:26<12:56,  2.02it/s]
 85%|████████▌ | 9112/10682 [1:17:27<12:56,  2.02it/s]
 85%|████████▌ | 9113/10682 [1:17:27<12:56,  2.02it/s]
 85%|████████▌ | 9114/10682 [1:17:28<12:54,  2.02it/s]
 85%|████████▌ | 9115/10682 [1:17:28<12:54,  2.02it/s]
 85%|████████▌ | 9116/10682 [1:17:29<12:54,  2.02it/s]
 85%|████████▌ | 9117/10682 [1:17:29<12:54,  2.02it/s]
 85%|████████▌ | 9118/10682 [1:17:30<12:53,  2.02it/s]
 85%|████████▌ | 9119/10682 [1:17:30<12:52,  2.02it/s]
 85%|████████▌ | 9120/10682 [1:17:31<12:52,  2.02it/s]
 85%|████████▌ | 9121/10682 [1:17:31<12:51,  2.02it/s]
 85%|████████▌ | 9122/10682 [1:17:32<12:51,  2.02it/s]
 85%|████████▌ | 9123/10682 [1:17:32<12:51,  2.02it/s]
 85%|████████▌ | 9124/10682 [1:17:33<12:51,  2.02it/s]
 85%|████████▌ | 9125/10682 [1:17:33<12:50,  2.02it/s]{'loss': 2.7976, 'grad_norm': 0.275244802236557, 'learning_rate': 6.334441157888504e-05, 'epoch': 11.95}
                                                      
+
 85%|████████▌ | 9125/10682 [1:17:33<12:50,  2.02it/s]
 85%|████████▌ | 9126/10682 [1:17:34<12:52,  2.01it/s]
 85%|████████▌ | 9127/10682 [1:17:34<12:51,  2.02it/s]
 85%|████████▌ | 9128/10682 [1:17:35<12:49,  2.02it/s]
 85%|████████▌ | 9129/10682 [1:17:35<12:49,  2.02it/s]
 85%|████████▌ | 9130/10682 [1:17:36<12:47,  2.02it/s]
 85%|████████▌ | 9131/10682 [1:17:36<12:46,  2.02it/s]
 85%|████████▌ | 9132/10682 [1:17:37<12:46,  2.02it/s]
 85%|████████▌ | 9133/10682 [1:17:37<12:45,  2.02it/s]
 86%|████████▌ | 9134/10682 [1:17:38<12:45,  2.02it/s]
 86%|████████▌ | 9135/10682 [1:17:38<12:44,  2.02it/s]
 86%|████████▌ | 9136/10682 [1:17:39<12:44,  2.02it/s]
 86%|████████▌ | 9137/10682 [1:17:39<12:44,  2.02it/s]
 86%|████████▌ | 9138/10682 [1:17:40<12:43,  2.02it/s]
 86%|████████▌ | 9139/10682 [1:17:40<12:42,  2.02it/s]
 86%|████████▌ | 9140/10682 [1:17:41<12:42,  2.02it/s]
 86%|████████▌ | 9141/10682 [1:17:41<12:41,  2.02it/s]
 86%|████████▌ | 9142/10682 [1:17:42<12:41,  2.02it/s]
 86%|████████▌ | 9143/10682 [1:17:42<12:41,  2.02it/s]
 86%|████████▌ | 9144/10682 [1:17:42<12:41,  2.02it/s]
 86%|████████▌ | 9145/10682 [1:17:43<12:40,  2.02it/s]
 86%|████████▌ | 9146/10682 [1:17:43<12:39,  2.02it/s]
 86%|████████▌ | 9147/10682 [1:17:44<12:38,  2.02it/s]
 86%|████████▌ | 9148/10682 [1:17:44<12:38,  2.02it/s]
 86%|████████▌ | 9149/10682 [1:17:45<12:36,  2.03it/s]
 86%|████████▌ | 9150/10682 [1:17:45<12:36,  2.02it/s]
                                                      
{'loss': 2.7975, 'grad_norm': 0.27586477994918823, 'learning_rate': 6.1368906655978e-05, 'epoch': 11.98}
+
 86%|████████▌ | 9150/10682 [1:17:45<12:36,  2.02it/s]
 86%|████████▌ | 9151/10682 [1:17:46<12:36,  2.02it/s]
 86%|████████▌ | 9152/10682 [1:17:46<12:36,  2.02it/s]
 86%|████████▌ | 9153/10682 [1:17:47<12:35,  2.02it/s]
 86%|████████▌ | 9154/10682 [1:17:47<12:34,  2.02it/s]
 86%|████████▌ | 9155/10682 [1:17:48<12:34,  2.02it/s]
 86%|████████▌ | 9156/10682 [1:17:48<12:34,  2.02it/s]
 86%|████████▌ | 9157/10682 [1:17:49<12:34,  2.02it/s]
 86%|████████▌ | 9158/10682 [1:17:49<12:34,  2.02it/s]
 86%|████████▌ | 9159/10682 [1:17:50<12:33,  2.02it/s]
 86%|████████▌ | 9160/10682 [1:17:50<12:32,  2.02it/s]
 86%|████████▌ | 9161/10682 [1:17:51<12:32,  2.02it/s]
 86%|████████▌ | 9162/10682 [1:17:51<12:23,  2.04it/s]
 86%|████████▌ | 9163/10682 [1:18:03<1:39:32,  3.93s/it]
 86%|████████▌ | 9164/10682 [1:18:04<1:13:27,  2.90s/it]
 86%|████████▌ | 9165/10682 [1:18:04<55:07,  2.18s/it]  
 86%|████████▌ | 9166/10682 [1:18:05<42:23,  1.68s/it]
 86%|████████▌ | 9167/10682 [1:18:05<33:24,  1.32s/it]
 86%|████████▌ | 9168/10682 [1:18:06<27:07,  1.07s/it]
 86%|████████▌ | 9169/10682 [1:18:06<22:43,  1.11it/s]
 86%|████████▌ | 9170/10682 [1:18:07<19:41,  1.28it/s]
 86%|████████▌ | 9171/10682 [1:18:07<17:30,  1.44it/s]
 86%|████████▌ | 9172/10682 [1:18:08<15:58,  1.57it/s]
 86%|████████▌ | 9173/10682 [1:18:08<14:55,  1.69it/s]
 86%|████████▌ | 9174/10682 [1:18:09<14:09,  1.78it/s]
 86%|████████▌ | 9175/10682 [1:18:09<13:39,  1.84it/s]
                                                      
{'loss': 2.7676, 'grad_norm': 0.27505698800086975, 'learning_rate': 5.94226809108499e-05, 'epoch': 12.02}
+
 86%|████████▌ | 9175/10682 [1:18:09<13:39,  1.84it/s]
 86%|████████▌ | 9176/10682 [1:18:10<13:16,  1.89it/s]
 86%|████████▌ | 9177/10682 [1:18:10<13:00,  1.93it/s]
 86%|████████▌ | 9178/10682 [1:18:11<12:49,  1.95it/s]
 86%|████████▌ | 9179/10682 [1:18:11<12:41,  1.97it/s]
 86%|████████▌ | 9180/10682 [1:18:12<12:35,  1.99it/s]
 86%|████████▌ | 9181/10682 [1:18:12<12:32,  1.99it/s]
 86%|████████▌ | 9182/10682 [1:18:13<12:28,  2.00it/s]
 86%|████████▌ | 9183/10682 [1:18:13<12:26,  2.01it/s]
 86%|████████▌ | 9184/10682 [1:18:14<12:24,  2.01it/s]
 86%|████████▌ | 9185/10682 [1:18:14<12:23,  2.01it/s]
 86%|████████▌ | 9186/10682 [1:18:15<12:21,  2.02it/s]
\ No newline at end of file