diff --git "a/train_job_output.txt" "b/train_job_output.txt"
--- "a/train_job_output.txt"
+++ "b/train_job_output.txt"
@@ -568,4 +568,99 @@ command outputs:
 
 78%|███████▊  | 8650/11074 [1:13:44<19:59,  2.02it/s]
 78%|███████▊  | 8651/11074 [1:13:45<19:59,  2.02it/s]
 78%|███████▊  | 8652/11074 [1:13:45<19:59,  2.02it/s]
 78%|███████▊  | 8653/11074 [1:13:46<19:57,  2.02it/s]
 78%|███████▊  | 8654/11074 [1:13:46<19:56,  2.02it/s]
 78%|███████▊  | 8655/11074 [1:13:47<19:55,  2.02it/s]
 78%|███████▊  | 8656/11074 [1:13:47<19:56,  2.02it/s]
 78%|███████▊  | 8657/11074 [1:13:48<19:55,  2.02it/s]
 78%|███████▊  | 8658/11074 [1:13:48<19:56,  2.02it/s]
 78%|███████▊  | 8659/11074 [1:13:49<19:54,  2.02it/s]
 78%|███████▊  | 8660/11074 [1:13:49<19:55,  2.02it/s]
 78%|███████▊  | 8661/11074 [1:13:49<19:53,  2.02it/s]
 78%|███████▊  | 8662/11074 [1:13:50<19:52,  2.02it/s]
 78%|███████▊  | 8663/11074 [1:13:50<19:52,  2.02it/s]
 78%|███████▊  | 8664/11074 [1:13:51<19:50,  2.02it/s]
 78%|███████▊  | 8665/11074 [1:13:51<19:51,  2.02it/s]
 78%|███████▊  | 8666/11074 [1:13:52<19:51,  2.02it/s]
 78%|███████▊  | 8667/11074 [1:13:52<19:50,  2.02it/s]
 78%|███████▊  | 8668/11074 [1:13:53<19:48,  2.02it/s]
 78%|███████▊  | 8669/11074 [1:13:53<19:49,  2.02it/s]
 78%|███████▊  | 8670/11074 [1:13:54<19:48,  2.02it/s]
 78%|███████▊  | 8671/11074 [1:13:54<19:48,  2.02it/s]
 78%|███████▊  | 8672/11074 [1:13:55<19:48,  2.02it/s]
 78%|███████▊  | 8673/11074 [1:13:55<19:47,  2.02it/s]
 78%|███████▊  | 8674/11074 [1:13:56<19:47,  2.02it/s]
 78%|███████▊  | 8675/11074 [1:13:56<19:46,  2.02it/s]
                                                      
{'loss': 3.2442, 'grad_norm': 0.23779277503490448, 'learning_rate': 0.00013628914403109144, 'epoch': 10.96}
 
 78%|███████▊  | 8675/11074 [1:13:56<19:46,  2.02it/s]
 78%|███████▊  | 8676/11074 [1:13:57<19:46,  2.02it/s]
 78%|███████▊  | 8677/11074 [1:13:57<19:45,  2.02it/s]
 78%|███████▊  | 8678/11074 [1:13:58<19:45,  2.02it/s]
 78%|███████▊  | 8679/11074 [1:13:58<19:43,  2.02it/s]
 78%|███████▊  | 8680/11074 [1:13:59<19:44,  2.02it/s]
 78%|███████▊  | 8681/11074 [1:13:59<19:43,  2.02it/s]
 78%|███████▊  | 8682/11074 [1:14:00<19:42,  2.02it/s]
 78%|███████▊  | 8683/11074 [1:14:00<19:42,  2.02it/s]
 78%|███████▊  | 8684/11074 [1:14:01<19:41,  2.02it/s]
 78%|███████▊  | 8685/11074 [1:14:01<19:41,  2.02it/s]
 78%|███████▊  | 8686/11074 [1:14:02<19:42,  2.02it/s]
 78%|███████▊  | 8687/11074 [1:14:02<19:41,  2.02it/s]
 78%|███████▊  | 8688/11074 [1:14:03<19:40,  2.02it/s]
 78%|███████▊  | 8689/11074 [1:14:03<19:39,  2.02it/s]
 78%|███████▊  | 8690/11074 [1:14:04<19:40,  2.02it/s]
 78%|███████▊  | 8691/11074 [1:14:04<19:40,  2.02it/s]
 78%|███████▊  | 8692/11074 [1:14:05<19:38,  2.02it/s]
 78%|███████▊  | 8693/11074 [1:14:05<19:37,  2.02it/s]
 79%|███████▊  | 8694/11074 [1:14:06<19:36,  2.02it/s]
 79%|███████▊  | 8695/11074 [1:14:06<19:36,  2.02it/s]
 79%|███████▊  | 8696/11074 [1:14:07<19:35,  2.02it/s]
 79%|███████▊  | 8697/11074 [1:14:07<19:35,  2.02it/s]
 79%|███████▊  | 8698/11074 [1:14:08<19:35,  2.02it/s]
 79%|███████▊  | 8699/11074 [1:14:08<19:34,  2.02it/s]
 79%|███████▊  | 8700/11074 [1:14:09<19:33,  2.02it/s]{'loss': 3.2503, 'grad_norm': 0.24186570942401886, 'learning_rate': 0.000133596608763568, 'epoch': 10.99}
 
                                                      

 79%|███████▊  | 8700/11074 [1:14:09<19:33,  2.02it/s]
 79%|███████▊  | 8701/11074 [1:14:09<19:34,  2.02it/s]
 79%|███████▊  | 8702/11074 [1:14:10<19:34,  2.02it/s]
 79%|███████▊  | 8703/11074 [1:14:10<19:33,  2.02it/s]
 79%|███████▊  | 8704/11074 [1:14:11<19:33,  2.02it/s]
 79%|███████▊  | 8705/11074 [1:14:11<19:31,  2.02it/s]
 79%|███████▊  | 8706/11074 [1:14:12<20:00,  1.97it/s]
 79%|███████▊  | 8707/11074 [1:14:24<2:38:17,  4.01s/it]
 79%|███████▊  | 8708/11074 [1:14:24<1:56:41,  2.96s/it]
 79%|███████▊  | 8709/11074 [1:14:25<1:27:30,  2.22s/it]
 79%|███████▊  | 8710/11074 [1:14:25<1:07:12,  1.71s/it]
 79%|███████▊  | 8711/11074 [1:14:26<52:50,  1.34s/it]  
 79%|███████▊  | 8712/11074 [1:14:26<42:50,  1.09s/it]
 79%|███████▊  | 8713/11074 [1:14:27<35:46,  1.10it/s]
 79%|███████▊  | 8714/11074 [1:14:27<30:52,  1.27it/s]
 79%|███████▊  | 8715/11074 [1:14:28<27:25,  1.43it/s]
 79%|███████▊  | 8716/11074 [1:14:28<25:00,  1.57it/s]
 79%|███████▊  | 8717/11074 [1:14:29<23:20,  1.68it/s]
 79%|███████▊  | 8718/11074 [1:14:29<22:07,  1.77it/s]
 79%|███████▊  | 8719/11074 [1:14:30<21:26,  1.83it/s]
 79%|███████▊  | 8720/11074 [1:14:30<20:52,  1.88it/s]
 79%|███████▉  | 8721/11074 [1:14:31<20:27,  1.92it/s]
 79%|███████▉  | 8722/11074 [1:14:31<20:06,  1.95it/s]
 79%|███████▉  | 8723/11074 [1:14:32<19:52,  1.97it/s]
 79%|███████▉  | 8724/11074 [1:14:32<19:42,  1.99it/s]
 79%|███████▉  | 8725/11074 [1:14:33<19:37,  2.00it/s]
                                                      {'loss': 3.1825, 'grad_norm': 0.2393781691789627, 'learning_rate': 0.0001309268294598309, 'epoch': 11.02}
-

 79%|███████▉  | 8725/11074 [1:14:33<19:37,  2.00it/s]
 79%|███████▉  | 8726/11074 [1:14:33<19:32,  2.00it/s]
 79%|███████▉  | 8727/11074 [1:14:34<19:28,  2.01it/s]
 79%|███████▉  | 8728/11074 [1:14:34<19:27,  2.01it/s]
 79%|███████▉  | 8729/11074 [1:14:35<19:24,  2.01it/s]
 79%|███████▉  | 8730/11074 [1:14:35<19:22,  2.02it/s]
 79%|███████▉  | 8731/11074 [1:14:36<19:23,  2.01it/s]
 79%|███████▉  | 8732/11074 [1:14:36<19:22,  2.01it/s]
 79%|███████▉  | 8733/11074 [1:14:37<19:21,  2.01it/s]
 79%|███████▉  | 8734/11074 [1:14:37<19:20,  2.02it/s]
 79%|███████▉  | 8735/11074 [1:14:38<19:17,  2.02it/s]
 79%|███████▉  | 8736/11074 [1:14:38<19:15,  2.02it/s]
 79%|███████▉  | 8737/11074 [1:14:39<19:15,  2.02it/s]
\ No newline at end of file
+

 79%|███████▉  | 8725/11074 [1:14:33<19:37,  2.00it/s]
 79%|███████▉  | 8726/11074 [1:14:33<19:32,  2.00it/s]
 79%|███████▉  | 8727/11074 [1:14:34<19:28,  2.01it/s]
 79%|███████▉  | 8728/11074 [1:14:34<19:27,  2.01it/s]
 79%|███████▉  | 8729/11074 [1:14:35<19:24,  2.01it/s]
 79%|███████▉  | 8730/11074 [1:14:35<19:22,  2.02it/s]
 79%|███████▉  | 8731/11074 [1:14:36<19:23,  2.01it/s]
 79%|███████▉  | 8732/11074 [1:14:36<19:22,  2.01it/s]
 79%|███████▉  | 8733/11074 [1:14:37<19:21,  2.01it/s]
 79%|███████▉  | 8734/11074 [1:14:37<19:20,  2.02it/s]
 79%|███████▉  | 8735/11074 [1:14:38<19:17,  2.02it/s]
 79%|███████▉  | 8736/11074 [1:14:38<19:15,  2.02it/s]
 79%|███████▉  | 8737/11074 [1:14:39<19:15,  2.02it/s]
 79%|███████▉  | 8738/11074 [1:14:39<19:14,  2.02it/s]
 79%|███████▉  | 8739/11074 [1:14:40<19:13,  2.03it/s]
 79%|███████▉  | 8740/11074 [1:14:40<19:12,  2.03it/s]
 79%|███████▉  | 8741/11074 [1:14:41<19:12,  2.02it/s]
 79%|███████▉  | 8742/11074 [1:14:41<19:11,  2.03it/s]
 79%|███████▉  | 8743/11074 [1:14:42<19:10,  2.03it/s]
 79%|███████▉  | 8744/11074 [1:14:42<19:10,  2.02it/s]
 79%|███████▉  | 8745/11074 [1:14:43<19:10,  2.02it/s]
 79%|███████▉  | 8746/11074 [1:14:43<19:10,  2.02it/s]
 79%|███████▉  | 8747/11074 [1:14:44<19:08,  2.03it/s]
 79%|███████▉  | 8748/11074 [1:14:44<19:08,  2.02it/s]
 79%|███████▉  | 8749/11074 [1:14:45<19:07,  2.03it/s]
 79%|███████▉  | 8750/11074 [1:14:45<19:07,  2.03it/s]
                                                      
{'loss': 3.1575, 'grad_norm': 0.23903685808181763, 'learning_rate': 0.00012827997193002967, 'epoch': 11.05}
+
 79%|███████▉  | 8750/11074 [1:14:45<19:07,  2.03it/s]
 79%|███████▉  | 8751/11074 [1:14:46<19:07,  2.02it/s]
 79%|███████▉  | 8752/11074 [1:14:46<19:07,  2.02it/s]
 79%|███████▉  | 8753/11074 [1:14:47<19:07,  2.02it/s]
 79%|███████▉  | 8754/11074 [1:14:47<19:06,  2.02it/s]
 79%|███████▉  | 8755/11074 [1:14:48<19:05,  2.02it/s]
 79%|███████▉  | 8756/11074 [1:14:48<19:04,  2.03it/s]
 79%|███████▉  | 8757/11074 [1:14:49<19:05,  2.02it/s]
 79%|███████▉  | 8758/11074 [1:14:49<19:04,  2.02it/s]
 79%|███████▉  | 8759/11074 [1:14:50<19:05,  2.02it/s]
 79%|███████▉  | 8760/11074 [1:14:50<19:03,  2.02it/s]
 79%|███████▉  | 8761/11074 [1:14:51<19:04,  2.02it/s]
 79%|███████▉  | 8762/11074 [1:14:51<19:04,  2.02it/s]
 79%|███████▉  | 8763/11074 [1:14:52<19:03,  2.02it/s]
 79%|███████▉  | 8764/11074 [1:14:52<19:01,  2.02it/s]
 79%|███████▉  | 8765/11074 [1:14:53<19:01,  2.02it/s]
 79%|███████▉  | 8766/11074 [1:14:53<18:59,  2.02it/s]
 79%|███████▉  | 8767/11074 [1:14:54<18:59,  2.02it/s]
 79%|███████▉  | 8768/11074 [1:14:54<18:59,  2.02it/s]
 79%|███████▉  | 8769/11074 [1:14:55<19:00,  2.02it/s]
 79%|███████▉  | 8770/11074 [1:14:55<18:58,  2.02it/s]
 79%|███████▉  | 8771/11074 [1:14:56<18:58,  2.02it/s]
 79%|███████▉  | 8772/11074 [1:14:56<18:56,  2.03it/s]
 79%|███████▉  | 8773/11074 [1:14:57<18:55,  2.03it/s]
 79%|███████▉  | 8774/11074 [1:14:57<18:55,  2.03it/s]
 79%|███████▉  | 8775/11074 [1:14:58<18:53,  2.03it/s]
                                                      
{'loss': 3.1605, 'grad_norm': 0.24214917421340942, 'learning_rate': 0.00012565620056072718, 'epoch': 11.09}
+
 79%|███████▉  | 8775/11074 [1:14:58<18:53,  2.03it/s]
 79%|███████▉  | 8776/11074 [1:14:58<18:56,  2.02it/s]
 79%|███████▉  | 8777/11074 [1:14:59<18:54,  2.02it/s]
 79%|███████▉  | 8778/11074 [1:14:59<18:55,  2.02it/s]
 79%|███████▉  | 8779/11074 [1:15:00<18:53,  2.02it/s]
 79%|███████▉  | 8780/11074 [1:15:00<18:53,  2.02it/s]
 79%|███████▉  | 8781/11074 [1:15:01<18:52,  2.02it/s]
 79%|███████▉  | 8782/11074 [1:15:01<18:52,  2.02it/s]
 79%|███████▉  | 8783/11074 [1:15:02<18:51,  2.02it/s]
 79%|███████▉  | 8784/11074 [1:15:02<18:52,  2.02it/s]
 79%|███████▉  | 8785/11074 [1:15:03<18:51,  2.02it/s]
 79%|███████▉  | 8786/11074 [1:15:03<18:51,  2.02it/s]
 79%|███████▉  | 8787/11074 [1:15:04<18:50,  2.02it/s]
 79%|███████▉  | 8788/11074 [1:15:04<18:49,  2.02it/s]
 79%|███████▉  | 8789/11074 [1:15:05<18:49,  2.02it/s]
 79%|███████▉  | 8790/11074 [1:15:05<18:47,  2.02it/s]
 79%|███████▉  | 8791/11074 [1:15:06<18:48,  2.02it/s]
 79%|███████▉  | 8792/11074 [1:15:06<18:47,  2.02it/s]
 79%|███████▉  | 8793/11074 [1:15:07<18:46,  2.02it/s]
 79%|███████▉  | 8794/11074 [1:15:07<18:45,  2.03it/s]
 79%|███████▉  | 8795/11074 [1:15:08<18:45,  2.02it/s]
 79%|███████▉  | 8796/11074 [1:15:08<18:44,  2.03it/s]
 79%|███████▉  | 8797/11074 [1:15:09<18:43,  2.03it/s]
 79%|███████▉  | 8798/11074 [1:15:09<18:44,  2.02it/s]
 79%|███████▉  | 8799/11074 [1:15:09<18:42,  2.03it/s]
 79%|███████▉  | 8800/11074 [1:15:10<18:42,  2.03it/s]
                                                      
{'loss': 3.1775, 'grad_norm': 0.2367831915616989, 'learning_rate': 0.00012305567830469017, 'epoch': 11.12}
+
 79%|███████▉  | 8800/11074 [1:15:10<18:42,  2.03it/s]
 79%|███████▉  | 8801/11074 [1:15:10<18:49,  2.01it/s]
 79%|███████▉  | 8802/11074 [1:15:11<18:45,  2.02it/s]
 79%|███████▉  | 8803/11074 [1:15:11<18:46,  2.02it/s]
 80%|███████▉  | 8804/11074 [1:15:12<18:43,  2.02it/s]
 80%|███████▉  | 8805/11074 [1:15:12<18:42,  2.02it/s]
 80%|███████▉  | 8806/11074 [1:15:13<18:41,  2.02it/s]
 80%|███████▉  | 8807/11074 [1:15:13<18:41,  2.02it/s]
 80%|███████▉  | 8808/11074 [1:15:14<18:39,  2.02it/s]
 80%|███████▉  | 8809/11074 [1:15:14<18:39,  2.02it/s]
 80%|███████▉  | 8810/11074 [1:15:15<18:38,  2.02it/s]
 80%|███████▉  | 8811/11074 [1:15:15<18:38,  2.02it/s]
 80%|███████▉  | 8812/11074 [1:15:16<18:38,  2.02it/s]
 80%|███████▉  | 8813/11074 [1:15:16<18:36,  2.02it/s]
 80%|███████▉  | 8814/11074 [1:15:17<18:36,  2.02it/s]
 80%|███████▉  | 8815/11074 [1:15:17<18:37,  2.02it/s]
 80%|███████▉  | 8816/11074 [1:15:18<18:37,  2.02it/s]
 80%|███████▉  | 8817/11074 [1:15:18<18:35,  2.02it/s]
 80%|███████▉  | 8818/11074 [1:15:19<18:35,  2.02it/s]
 80%|███████▉  | 8819/11074 [1:15:19<18:33,  2.02it/s]
 80%|███████▉  | 8820/11074 [1:15:20<18:33,  2.02it/s]
 80%|███████▉  | 8821/11074 [1:15:20<18:32,  2.02it/s]
 80%|███████▉  | 8822/11074 [1:15:21<18:31,  2.03it/s]
 80%|███████▉  | 8823/11074 [1:15:21<18:31,  2.02it/s]
 80%|███████▉  | 8824/11074 [1:15:22<18:31,  2.02it/s]
 80%|███████▉  | 8825/11074 [1:15:22<18:31,  2.02it/s]
                                                      
{'loss': 3.1689, 'grad_norm': 0.2386203557252884, 'learning_rate': 0.00012047856667076834, 'epoch': 11.15}
+
 80%|███████▉  | 8825/11074 [1:15:22<18:31,  2.02it/s]
 80%|███████▉  | 8826/11074 [1:15:23<18:32,  2.02it/s]
 80%|███████▉  | 8827/11074 [1:15:23<18:32,  2.02it/s]
 80%|███████▉  | 8828/11074 [1:15:24<18:32,  2.02it/s]
 80%|███████▉  | 8829/11074 [1:15:24<18:31,  2.02it/s]
 80%|███████▉  | 8830/11074 [1:15:25<18:30,  2.02it/s]
 80%|███████▉  | 8831/11074 [1:15:25<18:29,  2.02it/s]
 80%|███████▉  | 8832/11074 [1:15:26<18:28,  2.02it/s]
 80%|███████▉  | 8833/11074 [1:15:26<18:28,  2.02it/s]
 80%|███████▉  | 8834/11074 [1:15:27<18:27,  2.02it/s]
 80%|███████▉  | 8835/11074 [1:15:27<18:28,  2.02it/s]
 80%|███████▉  | 8836/11074 [1:15:28<18:26,  2.02it/s]
 80%|███████▉  | 8837/11074 [1:15:28<18:26,  2.02it/s]
 80%|███████▉  | 8838/11074 [1:15:29<18:25,  2.02it/s]
 80%|███████▉  | 8839/11074 [1:15:29<18:23,  2.03it/s]
 80%|███████▉  | 8840/11074 [1:15:30<18:24,  2.02it/s]
 80%|███████▉  | 8841/11074 [1:15:30<18:22,  2.03it/s]
 80%|███████▉  | 8842/11074 [1:15:31<18:23,  2.02it/s]
 80%|███████▉  | 8843/11074 [1:15:31<18:22,  2.02it/s]
 80%|███████▉  | 8844/11074 [1:15:32<18:21,  2.02it/s]
 80%|███████▉  | 8845/11074 [1:15:32<18:22,  2.02it/s]
 80%|███████▉  | 8846/11074 [1:15:33<18:20,  2.02it/s]
 80%|███████▉  | 8847/11074 [1:15:33<18:20,  2.02it/s]
 80%|███████▉  | 8848/11074 [1:15:34<18:20,  2.02it/s]
 80%|███████▉  | 8849/11074 [1:15:34<18:19,  2.02it/s]
 80%|███████▉  | 8850/11074 [1:15:35<18:19,  2.02it/s]{'loss': 3.173, 'grad_norm': 0.24308112263679504, 'learning_rate': 0.00011792502571386382, 'epoch': 11.18}
                                                      
+
 80%|███████▉  | 8850/11074 [1:15:35<18:19,  2.02it/s]
 80%|███████▉  | 8851/11074 [1:15:35<18:19,  2.02it/s]
 80%|███████▉  | 8852/11074 [1:15:36<18:18,  2.02it/s]
 80%|███████▉  | 8853/11074 [1:15:36<18:17,  2.02it/s]
 80%|███████▉  | 8854/11074 [1:15:37<18:18,  2.02it/s]
 80%|███████▉  | 8855/11074 [1:15:37<18:16,  2.02it/s]
 80%|███████▉  | 8856/11074 [1:15:38<18:16,  2.02it/s]
 80%|███████▉  | 8857/11074 [1:15:38<18:15,  2.02it/s]
 80%|███████▉  | 8858/11074 [1:15:39<18:15,  2.02it/s]
 80%|███████▉  | 8859/11074 [1:15:39<18:15,  2.02it/s]
 80%|████████  | 8860/11074 [1:15:40<18:14,  2.02it/s]
 80%|████████  | 8861/11074 [1:15:40<18:13,  2.02it/s]
 80%|████████  | 8862/11074 [1:15:41<18:14,  2.02it/s]
 80%|████████  | 8863/11074 [1:15:41<18:13,  2.02it/s]
 80%|████████  | 8864/11074 [1:15:42<18:13,  2.02it/s]
 80%|████████  | 8865/11074 [1:15:42<18:12,  2.02it/s]
 80%|████████  | 8866/11074 [1:15:43<18:13,  2.02it/s]
 80%|████████  | 8867/11074 [1:15:43<18:11,  2.02it/s]
 80%|████████  | 8868/11074 [1:15:44<18:11,  2.02it/s]
 80%|████████  | 8869/11074 [1:15:44<18:10,  2.02it/s]
 80%|████████  | 8870/11074 [1:15:45<18:09,  2.02it/s]
 80%|████████  | 8871/11074 [1:15:45<18:09,  2.02it/s]
 80%|████████  | 8872/11074 [1:15:46<18:09,  2.02it/s]
 80%|████████  | 8873/11074 [1:15:46<18:08,  2.02it/s]
 80%|████████  | 8874/11074 [1:15:47<18:08,  2.02it/s]
 80%|████████  | 8875/11074 [1:15:47<18:06,  2.02it/s]
                                                      {'loss': 3.1753, 'grad_norm': 0.241014301776886, 'learning_rate': 0.00011539521402499143, 'epoch': 11.21}
+

 80%|████████  | 8875/11074 [1:15:47<18:06,  2.02it/s]
 80%|████████  | 8876/11074 [1:15:48<18:08,  2.02it/s]
 80%|████████  | 8877/11074 [1:15:48<18:06,  2.02it/s]
 80%|████████  | 8878/11074 [1:15:49<18:05,  2.02it/s]
 80%|████████  | 8879/11074 [1:15:49<18:04,  2.02it/s]
 80%|████████  | 8880/11074 [1:15:50<18:03,  2.02it/s]
 80%|████████  | 8881/11074 [1:15:50<18:03,  2.02it/s]
 80%|████████  | 8882/11074 [1:15:51<19:38,  1.86it/s]
 80%|████████  | 8883/11074 [1:15:51<19:08,  1.91it/s]
 80%|████████  | 8884/11074 [1:15:52<18:48,  1.94it/s]
 80%|████████  | 8885/11074 [1:15:52<18:35,  1.96it/s]
 80%|████████  | 8886/11074 [1:15:53<18:25,  1.98it/s]
 80%|████████  | 8887/11074 [1:15:53<18:16,  1.99it/s]
 80%|████████  | 8888/11074 [1:15:54<18:11,  2.00it/s]
 80%|████████  | 8889/11074 [1:15:54<18:07,  2.01it/s]
 80%|████████  | 8890/11074 [1:15:55<18:04,  2.01it/s]
 80%|████████  | 8891/11074 [1:15:55<18:02,  2.02it/s]
 80%|████████  | 8892/11074 [1:15:56<18:01,  2.02it/s]
 80%|████████  | 8893/11074 [1:15:56<17:59,  2.02it/s]
 80%|████████  | 8894/11074 [1:15:57<17:57,  2.02it/s]
 80%|████████  | 8895/11074 [1:15:57<17:56,  2.02it/s]
 80%|████████  | 8896/11074 [1:15:58<17:55,  2.02it/s]
 80%|████████  | 8897/11074 [1:15:58<17:55,  2.02it/s]
 80%|████████  | 8898/11074 [1:15:59<17:54,  2.03it/s]
 80%|████████  | 8899/11074 [1:15:59<17:54,  2.02it/s]
 80%|████████  | 8900/11074 [1:16:00<17:54,  2.02it/s]{'loss': 3.1751, 'grad_norm': 0.24310094118118286, 'learning_rate': 0.00011288928872142829, 'epoch': 11.24}
+
                                                      

 80%|████████  | 8900/11074 [1:16:00<17:54,  2.02it/s]
 80%|████████  | 8901/11074 [1:16:00<17:55,  2.02it/s]
 80%|████████  | 8902/11074 [1:16:01<17:54,  2.02it/s]
 80%|████████  | 8903/11074 [1:16:01<17:53,  2.02it/s]
 80%|████████  | 8904/11074 [1:16:02<17:51,  2.03it/s]
 80%|████████  | 8905/11074 [1:16:02<17:51,  2.02it/s]
 80%|████████  | 8906/11074 [1:16:03<17:50,  2.03it/s]
 80%|████████  | 8907/11074 [1:16:03<17:50,  2.03it/s]
 80%|████████  | 8908/11074 [1:16:04<17:49,  2.02it/s]
 80%|████████  | 8909/11074 [1:16:04<17:48,  2.03it/s]
 80%|████████  | 8910/11074 [1:16:05<17:48,  2.02it/s]
 80%|████████  | 8911/11074 [1:16:05<17:47,  2.03it/s]
 80%|████████  | 8912/11074 [1:16:05<17:47,  2.03it/s]
 80%|████████  | 8913/11074 [1:16:06<17:45,  2.03it/s]
 80%|████████  | 8914/11074 [1:16:06<17:45,  2.03it/s]
 81%|████████  | 8915/11074 [1:16:07<17:45,  2.03it/s]
 81%|████████  | 8916/11074 [1:16:07<17:44,  2.03it/s]
 81%|████████  | 8917/11074 [1:16:08<17:45,  2.02it/s]
 81%|████████  | 8918/11074 [1:16:08<17:43,  2.03it/s]
 81%|████████  | 8919/11074 [1:16:09<17:44,  2.03it/s]
 81%|████████  | 8920/11074 [1:16:09<17:42,  2.03it/s]
 81%|████████  | 8921/11074 [1:16:10<17:42,  2.03it/s]
 81%|████████  | 8922/11074 [1:16:10<17:41,  2.03it/s]
 81%|████████  | 8923/11074 [1:16:11<17:41,  2.03it/s]
 81%|████████  | 8924/11074 [1:16:11<17:41,  2.03it/s]
 81%|████████  | 8925/11074 [1:16:12<17:40,  2.03it/s]
                                                      {'loss': 3.1725, 'grad_norm': 0.2414514720439911, 'learning_rate': 0.00011040740543695604, 'epoch': 11.28}
+
 81%|████████  | 8925/11074 [1:16:12<17:40,  2.03it/s]
 81%|████████  | 8926/11074 [1:16:12<17:42,  2.02it/s]
 81%|████████  | 8927/11074 [1:16:13<17:40,  2.02it/s]
 81%|████████  | 8928/11074 [1:16:13<17:41,  2.02it/s]
 81%|████████  | 8929/11074 [1:16:14<17:39,  2.02it/s]
 81%|████████  | 8930/11074 [1:16:14<17:39,  2.02it/s]
 81%|████████  | 8931/11074 [1:16:15<17:39,  2.02it/s]
 81%|████████  | 8932/11074 [1:16:15<17:38,  2.02it/s]
 81%|████████  | 8933/11074 [1:16:16<17:38,  2.02it/s]
 81%|████████  | 8934/11074 [1:16:16<17:37,  2.02it/s]
 81%|████████  | 8935/11074 [1:16:17<17:36,  2.02it/s]
 81%|████████  | 8936/11074 [1:16:17<17:36,  2.02it/s]
 81%|████████  | 8937/11074 [1:16:18<17:34,  2.03it/s]
 81%|████████  | 8938/11074 [1:16:18<17:34,  2.03it/s]
 81%|████████  | 8939/11074 [1:16:19<17:33,  2.03it/s]
 81%|████████  | 8940/11074 [1:16:19<17:32,  2.03it/s]
 81%|████████  | 8941/11074 [1:16:20<17:32,  2.03it/s]
 81%|████████  | 8942/11074 [1:16:20<17:33,  2.02it/s]
 81%|████████  | 8943/11074 [1:16:21<17:33,  2.02it/s]
 81%|████████  | 8944/11074 [1:16:21<17:34,  2.02it/s]
 81%|████████  | 8945/11074 [1:16:22<17:33,  2.02it/s]
 81%|████████  | 8946/11074 [1:16:22<17:33,  2.02it/s]
 81%|████████  | 8947/11074 [1:16:23<17:31,  2.02it/s]
 81%|████████  | 8948/11074 [1:16:23<17:31,  2.02it/s]
 81%|████████  | 8949/11074 [1:16:24<17:30,  2.02it/s]
 81%|████████  | 8950/11074 [1:16:24<17:29,  2.02it/s]
                                                      
{'loss': 3.1773, 'grad_norm': 0.246285080909729, 'learning_rate': 0.00010794971831219563, 'epoch': 11.31}
+
 81%|████████  | 8950/11074 [1:16:24<17:29,  2.02it/s]
 81%|████████  | 8951/11074 [1:16:25<17:29,  2.02it/s]
 81%|████████  | 8952/11074 [1:16:25<17:29,  2.02it/s]
 81%|████████  | 8953/11074 [1:16:26<17:28,  2.02it/s]
 81%|████████  | 8954/11074 [1:16:26<17:28,  2.02it/s]
 81%|████████  | 8955/11074 [1:16:27<17:27,  2.02it/s]
 81%|████████  | 8956/11074 [1:16:27<17:27,  2.02it/s]
 81%|████████  | 8957/11074 [1:16:28<17:25,  2.02it/s]
 81%|████████  | 8958/11074 [1:16:28<17:24,  2.03it/s]
 81%|████████  | 8959/11074 [1:16:29<17:24,  2.02it/s]
 81%|████████  | 8960/11074 [1:16:29<17:25,  2.02it/s]
 81%|████████  | 8961/11074 [1:16:30<17:24,  2.02it/s]
 81%|████████  | 8962/11074 [1:16:30<17:23,  2.02it/s]
 81%|████████  | 8963/11074 [1:16:31<17:22,  2.02it/s]
 81%|████████  | 8964/11074 [1:16:31<17:22,  2.02it/s]
 81%|████████  | 8965/11074 [1:16:32<17:21,  2.02it/s]
 81%|████████  | 8966/11074 [1:16:32<17:20,  2.03it/s]
 81%|████████  | 8967/11074 [1:16:33<17:20,  2.02it/s]
 81%|████████  | 8968/11074 [1:16:33<18:51,  1.86it/s]
 81%|████████  | 8969/11074 [1:16:34<18:23,  1.91it/s]
 81%|████████  | 8970/11074 [1:16:34<18:03,  1.94it/s]
 81%|████████  | 8971/11074 [1:16:35<17:49,  1.97it/s]
 81%|████████  | 8972/11074 [1:16:35<17:39,  1.98it/s]
 81%|████████  | 8973/11074 [1:16:36<17:32,  2.00it/s]
 81%|████████  | 8974/11074 [1:16:36<17:27,  2.01it/s]
 81%|████████  | 8975/11074 [1:16:37<17:23,  2.01it/s]{'loss': 3.1745, 'grad_norm': 0.2431633323431015, 'learning_rate': 0.00010551637998503322, 'epoch': 11.34}
+
                                                      

 81%|████████  | 8975/11074 [1:16:37<17:23,  2.01it/s]
 81%|████████  | 8976/11074 [1:16:37<17:24,  2.01it/s]
 81%|████████  | 8977/11074 [1:16:38<17:20,  2.02it/s]
 81%|████████  | 8978/11074 [1:16:38<17:19,  2.02it/s]
 81%|████████  | 8979/11074 [1:16:39<17:16,  2.02it/s]
 81%|████████  | 8980/11074 [1:16:39<17:15,  2.02it/s]
 81%|████████  | 8981/11074 [1:16:40<17:14,  2.02it/s]
 81%|████████  | 8982/11074 [1:16:40<17:11,  2.03it/s]
 81%|████████  | 8983/11074 [1:16:41<17:12,  2.03it/s]
 81%|████████  | 8984/11074 [1:16:41<17:11,  2.03it/s]
 81%|████████  | 8985/11074 [1:16:42<17:11,  2.03it/s]
 81%|████████  | 8986/11074 [1:16:42<17:11,  2.02it/s]
 81%|████████  | 8987/11074 [1:16:43<17:10,  2.02it/s]
 81%|████████  | 8988/11074 [1:16:43<17:10,  2.02it/s]
 81%|████████  | 8989/11074 [1:16:44<17:10,  2.02it/s]
 81%|████████  | 8990/11074 [1:16:44<17:09,  2.02it/s]
 81%|████████  | 8991/11074 [1:16:45<17:08,  2.02it/s]
 81%|████████  | 8992/11074 [1:16:45<17:07,  2.03it/s]
 81%|████████  | 8993/11074 [1:16:46<17:07,  2.03it/s]
 81%|████████  | 8994/11074 [1:16:46<17:06,  2.03it/s]
 81%|████████  | 8995/11074 [1:16:47<17:05,  2.03it/s]
 81%|████████  | 8996/11074 [1:16:47<17:05,  2.03it/s]
 81%|████████  | 8997/11074 [1:16:48<17:05,  2.03it/s]
 81%|████████▏ | 8998/11074 [1:16:48<17:04,  2.03it/s]
 81%|████████▏ | 8999/11074 [1:16:49<17:03,  2.03it/s]
 81%|████████▏ | 9000/11074 [1:16:49<17:03,  2.03it/s]
                                                      
{'loss': 3.1841, 'grad_norm': 0.23753663897514343, 'learning_rate': 0.00010310754158114144, 'epoch': 11.37}
+
 81%|████████▏ | 9000/11074 [1:16:49<17:03,  2.03it/s]
 81%|████████▏ | 9001/11074 [1:16:50<17:10,  2.01it/s]
 81%|████████▏ | 9002/11074 [1:16:50<17:07,  2.02it/s]
 81%|████████▏ | 9003/11074 [1:16:51<17:06,  2.02it/s]
 81%|████████▏ | 9004/11074 [1:16:51<17:04,  2.02it/s]
 81%|████████▏ | 9005/11074 [1:16:52<17:04,  2.02it/s]
 81%|████████▏ | 9006/11074 [1:16:52<17:03,  2.02it/s]
 81%|████████▏ | 9007/11074 [1:16:53<17:02,  2.02it/s]
 81%|████████▏ | 9008/11074 [1:16:53<17:00,  2.02it/s]
 81%|████████▏ | 9009/11074 [1:16:54<17:00,  2.02it/s]
 81%|████████▏ | 9010/11074 [1:16:54<17:00,  2.02it/s]
 81%|████████▏ | 9011/11074 [1:16:55<17:00,  2.02it/s]
 81%|████████▏ | 9012/11074 [1:16:55<16:59,  2.02it/s]
 81%|████████▏ | 9013/11074 [1:16:56<16:59,  2.02it/s]
 81%|████████▏ | 9014/11074 [1:16:56<16:58,  2.02it/s]
 81%|████████▏ | 9015/11074 [1:16:57<16:57,  2.02it/s]
 81%|████████▏ | 9016/11074 [1:16:57<16:56,  2.02it/s]
 81%|████████▏ | 9017/11074 [1:16:58<16:56,  2.02it/s]
 81%|████████▏ | 9018/11074 [1:16:58<16:56,  2.02it/s]
 81%|████████▏ | 9019/11074 [1:16:59<16:55,  2.02it/s]
 81%|████████▏ | 9020/11074 [1:16:59<16:55,  2.02it/s]
 81%|████████▏ | 9021/11074 [1:16:59<16:53,  2.02it/s]
 81%|████████▏ | 9022/11074 [1:17:00<16:53,  2.02it/s]
 81%|████████▏ | 9023/11074 [1:17:00<16:52,  2.03it/s]
 81%|████████▏ | 9024/11074 [1:17:01<16:52,  2.03it/s]
 81%|████████▏ | 9025/11074 [1:17:01<16:51,  2.03it/s]
                                                      
{'loss': 3.1856, 'grad_norm': 0.24234531819820404, 'learning_rate': 0.00010072335270459255, 'epoch': 11.4}
+
 81%|████████▏ | 9025/11074 [1:17:01<16:51,  2.03it/s]
 82%|████████▏ | 9026/11074 [1:17:02<16:55,  2.02it/s]
 82%|████████▏ | 9027/11074 [1:17:02<16:52,  2.02it/s]
 82%|████████▏ | 9028/11074 [1:17:03<16:52,  2.02it/s]
 82%|████████▏ | 9029/11074 [1:17:03<16:50,  2.02it/s]
 82%|████████▏ | 9030/11074 [1:17:04<16:49,  2.02it/s]
 82%|████████▏ | 9031/11074 [1:17:04<16:48,  2.02it/s]
 82%|████████▏ | 9032/11074 [1:17:05<16:49,  2.02it/s]
 82%|████████▏ | 9033/11074 [1:17:05<16:48,  2.02it/s]
 82%|████████▏ | 9034/11074 [1:17:06<16:48,  2.02it/s]
 82%|████████▏ | 9035/11074 [1:17:06<16:47,  2.02it/s]
 82%|████████▏ | 9036/11074 [1:17:07<16:48,  2.02it/s]
 82%|████████▏ | 9037/11074 [1:17:07<16:46,  2.02it/s]
 82%|████████▏ | 9038/11074 [1:17:08<16:46,  2.02it/s]
 82%|████████▏ | 9039/11074 [1:17:08<16:44,  2.02it/s]
 82%|████████▏ | 9040/11074 [1:17:09<16:44,  2.03it/s]
 82%|████████▏ | 9041/11074 [1:17:09<16:43,  2.03it/s]
 82%|████████▏ | 9042/11074 [1:17:10<16:42,  2.03it/s]
 82%|████████▏ | 9043/11074 [1:17:10<16:42,  2.03it/s]
 82%|████████▏ | 9044/11074 [1:17:11<16:41,  2.03it/s]
 82%|████████▏ | 9045/11074 [1:17:11<16:40,  2.03it/s]
 82%|████████▏ | 9046/11074 [1:17:12<16:41,  2.03it/s]
 82%|████████▏ | 9047/11074 [1:17:12<16:39,  2.03it/s]
 82%|████████▏ | 9048/11074 [1:17:13<16:40,  2.02it/s]
 82%|████████▏ | 9049/11074 [1:17:13<16:39,  2.03it/s]
 82%|████████▏ | 9050/11074 [1:17:14<16:39,  2.03it/s]
                                                      
{'loss': 3.1763, 'grad_norm': 0.23996089398860931, 'learning_rate': 9.836396142856762e-05, 'epoch': 11.43}
+
 82%|████████▏ | 9050/11074 [1:17:14<16:39,  2.03it/s]
 82%|████████▏ | 9051/11074 [1:17:14<16:39,  2.02it/s]
 82%|████████▏ | 9052/11074 [1:17:15<16:38,  2.02it/s]
 82%|████████▏ | 9053/11074 [1:17:15<16:38,  2.02it/s]
 82%|████████▏ | 9054/11074 [1:17:16<16:37,  2.03it/s]
 82%|████████▏ | 9055/11074 [1:17:16<16:37,  2.02it/s]
 82%|████████▏ | 9056/11074 [1:17:17<16:36,  2.03it/s]
 82%|████████▏ | 9057/11074 [1:17:17<16:36,  2.02it/s]
 82%|████████▏ | 9058/11074 [1:17:18<16:34,  2.03it/s]
 82%|████████▏ | 9059/11074 [1:17:18<16:36,  2.02it/s]
 82%|████████▏ | 9060/11074 [1:17:19<16:35,  2.02it/s]
 82%|████████▏ | 9061/11074 [1:17:19<16:34,  2.02it/s]
 82%|████████▏ | 9062/11074 [1:17:20<16:33,  2.03it/s]
 82%|████████▏ | 9063/11074 [1:17:20<16:33,  2.02it/s]
 82%|████████▏ | 9064/11074 [1:17:21<16:32,  2.02it/s]
 82%|████████▏ | 9065/11074 [1:17:21<16:31,  2.03it/s]
 82%|████████▏ | 9066/11074 [1:17:22<16:31,  2.03it/s]
 82%|████████▏ | 9067/11074 [1:17:22<16:29,  2.03it/s]
 82%|████████▏ | 9068/11074 [1:17:23<16:30,  2.03it/s]
 82%|████████▏ | 9069/11074 [1:17:23<16:28,  2.03it/s]
 82%|████████▏ | 9070/11074 [1:17:24<16:29,  2.03it/s]
 82%|████████▏ | 9071/11074 [1:17:24<16:29,  2.03it/s]
 82%|████████▏ | 9072/11074 [1:17:25<16:28,  2.03it/s]
 82%|████████▏ | 9073/11074 [1:17:25<16:28,  2.02it/s]
 82%|████████▏ | 9074/11074 [1:17:26<16:28,  2.02it/s]
 82%|████████▏ | 9075/11074 [1:17:26<16:27,  2.02it/s]
                                                      
{'loss': 3.1841, 'grad_norm': 0.24090521037578583, 'learning_rate': 9.602951428616059e-05, 'epoch': 11.47}
+
 82%|████████▏ | 9075/11074 [1:17:26<16:27,  2.02it/s]
 82%|████████▏ | 9076/11074 [1:17:27<16:27,  2.02it/s]
 82%|████████▏ | 9077/11074 [1:17:27<16:27,  2.02it/s]
 82%|████████▏ | 9078/11074 [1:17:28<16:25,  2.03it/s]
 82%|████████▏ | 9079/11074 [1:17:28<16:25,  2.02it/s]
 82%|████████▏ | 9080/11074 [1:17:29<16:24,  2.03it/s]
 82%|████████▏ | 9081/11074 [1:17:29<16:24,  2.03it/s]
 82%|████████▏ | 9082/11074 [1:17:30<16:23,  2.02it/s]
 82%|████████▏ | 9083/11074 [1:17:30<16:22,  2.03it/s]
 82%|████████▏ | 9084/11074 [1:17:31<16:23,  2.02it/s]
 82%|████████▏ | 9085/11074 [1:17:31<16:22,  2.03it/s]
 82%|████████▏ | 9086/11074 [1:17:32<16:22,  2.02it/s]
 82%|████████▏ | 9087/11074 [1:17:32<16:20,  2.03it/s]
 82%|████████▏ | 9088/11074 [1:17:33<16:20,  2.03it/s]
 82%|████████▏ | 9089/11074 [1:17:33<16:19,  2.03it/s]
 82%|████████▏ | 9090/11074 [1:17:34<16:19,  2.03it/s]
 82%|████████▏ | 9091/11074 [1:17:34<16:18,  2.03it/s]
 82%|████████▏ | 9092/11074 [1:17:35<16:17,  2.03it/s]
 82%|████████▏ | 9093/11074 [1:17:35<16:17,  2.03it/s]
 82%|████████▏ | 9094/11074 [1:17:36<16:17,  2.03it/s]
 82%|████████▏ | 9095/11074 [1:17:36<16:18,  2.02it/s]
 82%|████████▏ | 9096/11074 [1:17:37<16:18,  2.02it/s]
 82%|████████▏ | 9097/11074 [1:17:37<16:17,  2.02it/s]
 82%|████████▏ | 9098/11074 [1:17:38<16:16,  2.02it/s]
 82%|████████▏ | 9099/11074 [1:17:38<16:15,  2.02it/s]
 82%|████████▏ | 9100/11074 [1:17:39<16:14,  2.03it/s]{'loss': 3.1876, 'grad_norm': 0.24239379167556763, 'learning_rate': 9.372015626127656e-05, 'epoch': 11.5}
                                                      
+
 82%|████████▏ | 9100/11074 [1:17:39<16:14,  2.03it/s]
 82%|████████▏ | 9101/11074 [1:17:39<16:15,  2.02it/s]
 82%|████████▏ | 9102/11074 [1:17:39<16:13,  2.03it/s]
 82%|████████▏ | 9103/11074 [1:17:40<16:13,  2.02it/s]
 82%|████████▏ | 9104/11074 [1:17:40<16:13,  2.02it/s]
 82%|████████▏ | 9105/11074 [1:17:41<16:13,  2.02it/s]
 82%|████████▏ | 9106/11074 [1:17:41<16:12,  2.02it/s]
 82%|████████▏ | 9107/11074 [1:17:42<16:12,  2.02it/s]
 82%|████████▏ | 9108/11074 [1:17:42<16:11,  2.02it/s]
 82%|████████▏ | 9109/11074 [1:17:43<16:11,  2.02it/s]
 82%|████████▏ | 9110/11074 [1:17:43<16:10,  2.02it/s]
 82%|████████▏ | 9111/11074 [1:17:44<16:10,  2.02it/s]
 82%|████████▏ | 9112/11074 [1:17:44<16:09,  2.02it/s]
 82%|████████▏ | 9113/11074 [1:17:45<16:09,  2.02it/s]
 82%|████████▏ | 9114/11074 [1:17:45<16:08,  2.02it/s]
 82%|████████▏ | 9115/11074 [1:17:46<16:08,  2.02it/s]
 82%|████████▏ | 9116/11074 [1:17:46<16:07,  2.02it/s]
 82%|████████▏ | 9117/11074 [1:17:47<16:07,  2.02it/s]
 82%|████████▏ | 9118/11074 [1:17:47<16:06,  2.02it/s]
 82%|████████▏ | 9119/11074 [1:17:48<16:07,  2.02it/s]
 82%|████████▏ | 9120/11074 [1:17:48<16:07,  2.02it/s]
 82%|████████▏ | 9121/11074 [1:17:49<16:06,  2.02it/s]
 82%|████████▏ | 9122/11074 [1:17:49<16:05,  2.02it/s]
 82%|████████▏ | 9123/11074 [1:17:50<16:04,  2.02it/s]
 82%|████████▏ | 9124/11074 [1:17:50<16:03,  2.02it/s]
 82%|████████▏ | 9125/11074 [1:17:51<16:03,  2.02it/s]
                                                      
{'loss': 3.1881, 'grad_norm': 0.2376471906900406, 'learning_rate': 9.143603077962858e-05, 'epoch': 11.53}
+
 82%|████████▏ | 9125/11074 [1:17:51<16:03,  2.02it/s]
 82%|████████▏ | 9126/11074 [1:17:51<16:04,  2.02it/s]
 82%|████████▏ | 9127/11074 [1:17:52<16:03,  2.02it/s]
 82%|████████▏ | 9128/11074 [1:17:52<16:01,  2.02it/s]
 82%|████████▏ | 9129/11074 [1:17:53<16:01,  2.02it/s]
 82%|████████▏ | 9130/11074 [1:17:53<15:59,  2.03it/s]
 82%|████████▏ | 9131/11074 [1:17:54<15:59,  2.02it/s]
 82%|████████▏ | 9132/11074 [1:17:54<15:58,  2.03it/s]
 82%|████████▏ | 9133/11074 [1:17:55<15:58,  2.02it/s]
 82%|████████▏ | 9134/11074 [1:17:55<15:58,  2.02it/s]
 82%|████████▏ | 9135/11074 [1:17:56<15:58,  2.02it/s]
 82%|████████▏ | 9136/11074 [1:17:56<15:57,  2.02it/s]
 83%|████████▎ | 9137/11074 [1:17:57<15:56,  2.02it/s]
 83%|████████▎ | 9138/11074 [1:17:57<15:56,  2.02it/s]
 83%|████████▎ | 9139/11074 [1:17:58<15:55,  2.02it/s]
 83%|████████▎ | 9140/11074 [1:17:58<15:55,  2.02it/s]
 83%|████████▎ | 9141/11074 [1:17:59<15:54,  2.03it/s]
 83%|████████▎ | 9142/11074 [1:17:59<15:54,  2.02it/s]
 83%|████████▎ | 9143/11074 [1:18:00<15:52,  2.03it/s]
 83%|████████▎ | 9144/11074 [1:18:00<15:52,  2.03it/s]
 83%|████████▎ | 9145/11074 [1:18:01<15:52,  2.03it/s]
 83%|████████▎ | 9146/11074 [1:18:01<15:53,  2.02it/s]
 83%|████████▎ | 9147/11074 [1:18:02<15:51,  2.03it/s]
 83%|████████▎ | 9148/11074 [1:18:02<15:50,  2.03it/s]
 83%|████████▎ | 9149/11074 [1:18:03<15:49,  2.03it/s]
 83%|████████▎ | 9150/11074 [1:18:03<15:49,  2.03it/s]
                                                      
{'loss': 3.1884, 'grad_norm': 0.2383396029472351, 'learning_rate': 8.917727969982942e-05, 'epoch': 11.56}
+
 83%|████████▎ | 9150/11074 [1:18:03<15:49,  2.03it/s]
 83%|████████▎ | 9151/11074 [1:18:04<15:50,  2.02it/s]
 83%|████████▎ | 9152/11074 [1:18:04<15:48,  2.03it/s]
 83%|████████▎ | 9153/11074 [1:18:05<15:48,  2.02it/s]
 83%|████████▎ | 9154/11074 [1:18:05<15:46,  2.03it/s]
 83%|████████▎ | 9155/11074 [1:18:06<15:47,  2.02it/s]
 83%|████████▎ | 9156/11074 [1:18:06<15:46,  2.03it/s]
 83%|████████▎ | 9157/11074 [1:18:07<15:46,  2.03it/s]
 83%|████████▎ | 9158/11074 [1:18:07<15:45,  2.03it/s]
 83%|████████▎ | 9159/11074 [1:18:08<15:45,  2.03it/s]
 83%|████████▎ | 9160/11074 [1:18:08<15:44,  2.03it/s]
 83%|████████▎ | 9161/11074 [1:18:09<15:43,  2.03it/s]
 83%|████████▎ | 9162/11074 [1:18:09<15:43,  2.03it/s]
 83%|████████▎ | 9163/11074 [1:18:10<15:42,  2.03it/s]
 83%|████████▎ | 9164/11074 [1:18:10<15:42,  2.03it/s]
 83%|████████▎ | 9165/11074 [1:18:11<15:41,  2.03it/s]
 83%|████████▎ | 9166/11074 [1:18:11<15:43,  2.02it/s]
 83%|████████▎ | 9167/11074 [1:18:12<15:42,  2.02it/s]
 83%|████████▎ | 9168/11074 [1:18:12<15:42,  2.02it/s]
 83%|████████▎ | 9169/11074 [1:18:13<15:41,  2.02it/s]
 83%|████████▎ | 9170/11074 [1:18:13<15:41,  2.02it/s]
 83%|████████▎ | 9171/11074 [1:18:14<15:40,  2.02it/s]
 83%|████████▎ | 9172/11074 [1:18:14<15:40,  2.02it/s]
 83%|████████▎ | 9173/11074 [1:18:15<15:39,  2.02it/s]
 83%|████████▎ | 9174/11074 [1:18:15<15:38,  2.02it/s]
 83%|████████▎ | 9175/11074 [1:18:16<15:38,  2.02it/s]
                                                      {'loss': 3.1853, 'grad_norm': 0.24211829900741577, 'learning_rate': 8.6944043304581e-05, 'epoch': 11.59}
+
 83%|████████▎ | 9175/11074 [1:18:16<15:38,  2.02it/s]
 83%|████████▎ | 9176/11074 [1:18:16<15:39,  2.02it/s]
 83%|████████▎ | 9177/11074 [1:18:17<15:38,  2.02it/s]
 83%|████████▎ | 9178/11074 [1:18:17<15:38,  2.02it/s]
 83%|████████▎ | 9179/11074 [1:18:18<15:37,  2.02it/s]
 83%|████████▎ | 9180/11074 [1:18:18<15:37,  2.02it/s]
 83%|████████▎ | 9181/11074 [1:18:19<15:36,  2.02it/s]
 83%|████████▎ | 9182/11074 [1:18:19<15:36,  2.02it/s]
 83%|████████▎ | 9183/11074 [1:18:20<15:35,  2.02it/s]
 83%|████████▎ | 9184/11074 [1:18:20<15:35,  2.02it/s]
 83%|████████▎ | 9185/11074 [1:18:21<15:34,  2.02it/s]
 83%|████████▎ | 9186/11074 [1:18:21<15:33,  2.02it/s]
 83%|████████▎ | 9187/11074 [1:18:22<15:33,  2.02it/s]
 83%|████████▎ | 9188/11074 [1:18:22<15:31,  2.02it/s]
 83%|████████▎ | 9189/11074 [1:18:22<15:30,  2.02it/s]
 83%|████████▎ | 9190/11074 [1:18:23<15:30,  2.02it/s]
 83%|████████▎ | 9191/11074 [1:18:23<15:30,  2.02it/s]
 83%|████████▎ | 9192/11074 [1:18:24<15:29,  2.02it/s]
 83%|████████▎ | 9193/11074 [1:18:24<15:30,  2.02it/s]
 83%|████████▎ | 9194/11074 [1:18:25<15:28,  2.02it/s]
 83%|████████▎ | 9195/11074 [1:18:25<15:27,  2.03it/s]
 83%|████████▎ | 9196/11074 [1:18:26<15:27,  2.02it/s]
 83%|████████▎ | 9197/11074 [1:18:26<15:26,  2.03it/s]
 83%|████████▎ | 9198/11074 [1:18:27<15:26,  2.02it/s]
 83%|████████▎ | 9199/11074 [1:18:27<15:25,  2.03it/s]
 83%|████████▎ | 9200/11074 [1:18:28<15:25,  2.02it/s]
                                                      
{'loss': 3.1916, 'grad_norm': 0.23818659782409668, 'learning_rate': 8.47364602919623e-05, 'epoch': 11.62}
+
 83%|████████▎ | 9200/11074 [1:18:28<15:25,  2.02it/s]
 83%|████████▎ | 9201/11074 [1:18:28<15:25,  2.02it/s]
 83%|████████▎ | 9202/11074 [1:18:29<15:25,  2.02it/s]
 83%|████████▎ | 9203/11074 [1:18:29<15:24,  2.02it/s]
 83%|████████▎ | 9204/11074 [1:18:30<15:24,  2.02it/s]
 83%|████████▎ | 9205/11074 [1:18:30<15:23,  2.02it/s]
 83%|████████▎ | 9206/11074 [1:18:31<15:23,  2.02it/s]
 83%|████████▎ | 9207/11074 [1:18:31<15:23,  2.02it/s]
 83%|████████▎ | 9208/11074 [1:18:32<15:21,  2.02it/s]
 83%|████████▎ | 9209/11074 [1:18:32<15:21,  2.02it/s]
 83%|████████▎ | 9210/11074 [1:18:33<15:20,  2.03it/s]
 83%|████████▎ | 9211/11074 [1:18:33<15:20,  2.02it/s]
 83%|████████▎ | 9212/11074 [1:18:34<15:19,  2.03it/s]
 83%|████████▎ | 9213/11074 [1:18:34<15:19,  2.02it/s]
 83%|████████▎ | 9214/11074 [1:18:35<15:18,  2.03it/s]
 83%|████████▎ | 9215/11074 [1:18:35<15:17,  2.03it/s]
 83%|████████▎ | 9216/11074 [1:18:36<15:16,  2.03it/s]
 83%|████████▎ | 9217/11074 [1:18:36<15:16,  2.03it/s]
 83%|████████▎ | 9218/11074 [1:18:37<15:15,  2.03it/s]
 83%|████████▎ | 9219/11074 [1:18:37<15:13,  2.03it/s]
 83%|████████▎ | 9220/11074 [1:18:38<15:14,  2.03it/s]
 83%|████████▎ | 9221/11074 [1:18:38<15:13,  2.03it/s]
 83%|████████▎ | 9222/11074 [1:18:39<15:13,  2.03it/s]
 83%|████████▎ | 9223/11074 [1:18:39<15:13,  2.03it/s]
 83%|████████▎ | 9224/11074 [1:18:40<15:13,  2.02it/s]
 83%|████████▎ | 9225/11074 [1:18:40<15:13,  2.02it/s]{'loss': 3.1904, 'grad_norm': 0.23970608413219452, 'learning_rate': 8.255466776681563e-05, 'epoch': 11.66}
                                                      
+
 83%|████████▎ | 9225/11074 [1:18:40<15:13,  2.02it/s]
 83%|████████▎ | 9226/11074 [1:18:41<15:14,  2.02it/s]
 83%|████████▎ | 9227/11074 [1:18:41<15:13,  2.02it/s]
 83%|████████▎ | 9228/11074 [1:18:42<15:12,  2.02it/s]
 83%|████████▎ | 9229/11074 [1:18:42<15:11,  2.02it/s]
 83%|████████▎ | 9230/11074 [1:18:43<15:11,  2.02it/s]
 83%|████████▎ | 9231/11074 [1:18:43<15:10,  2.02it/s]
 83%|████████▎ | 9232/11074 [1:18:44<15:10,  2.02it/s]
 83%|████████▎ | 9233/11074 [1:18:44<15:08,  2.03it/s]
 83%|████████▎ | 9234/11074 [1:18:45<15:09,  2.02it/s]
 83%|████████▎ | 9235/11074 [1:18:45<15:08,  2.03it/s]
 83%|████████▎ | 9236/11074 [1:18:46<15:08,  2.02it/s]
 83%|████████▎ | 9237/11074 [1:18:46<15:08,  2.02it/s]
 83%|████████▎ | 9238/11074 [1:18:47<15:08,  2.02it/s]
 83%|████████▎ | 9239/11074 [1:18:47<15:08,  2.02it/s]
 83%|████████▎ | 9240/11074 [1:18:48<15:07,  2.02it/s]
 83%|████████▎ | 9241/11074 [1:18:48<15:06,  2.02it/s]
 83%|████████▎ | 9242/11074 [1:18:49<15:06,  2.02it/s]
 83%|████████▎ | 9243/11074 [1:18:49<15:04,  2.02it/s]
 83%|████████▎ | 9244/11074 [1:18:50<15:04,  2.02it/s]
 83%|████████▎ | 9245/11074 [1:18:50<15:02,  2.03it/s]
 83%|████████▎ | 9246/11074 [1:18:51<15:02,  2.03it/s]
 84%|████████▎ | 9247/11074 [1:18:51<15:02,  2.03it/s]
 84%|████████▎ | 9248/11074 [1:18:52<15:02,  2.02it/s]
 84%|████████▎ | 9249/11074 [1:18:52<15:01,  2.02it/s]
 84%|████████▎ | 9250/11074 [1:18:53<15:00,  2.03it/s]{'loss': 3.1955, 'grad_norm': 0.2404358983039856, 'learning_rate': 8.039880123223076e-05, 'epoch': 11.69}
                                                      
+
 84%|████████▎ | 9250/11074 [1:18:53<15:00,  2.03it/s]
 84%|████████▎ | 9251/11074 [1:18:53<15:01,  2.02it/s]
 84%|████████▎ | 9252/11074 [1:18:54<15:01,  2.02it/s]
 84%|████████▎ | 9253/11074 [1:18:54<15:00,  2.02it/s]
 84%|████████▎ | 9254/11074 [1:18:55<15:00,  2.02it/s]
 84%|████████▎ | 9255/11074 [1:18:55<14:59,  2.02it/s]
 84%|████████▎ | 9256/11074 [1:18:56<14:58,  2.02it/s]
 84%|████████▎ | 9257/11074 [1:18:56<14:58,  2.02it/s]
 84%|████████▎ | 9258/11074 [1:18:57<14:57,  2.02it/s]
 84%|████████▎ | 9259/11074 [1:18:57<14:57,  2.02it/s]
 84%|████████▎ | 9260/11074 [1:18:58<14:57,  2.02it/s]
 84%|████████▎ | 9261/11074 [1:18:58<14:56,  2.02it/s]
 84%|████████▎ | 9262/11074 [1:18:59<14:55,  2.02it/s]
 84%|████████▎ | 9263/11074 [1:18:59<14:54,  2.02it/s]
 84%|████████▎ | 9264/11074 [1:19:00<14:53,  2.03it/s]
 84%|████████▎ | 9265/11074 [1:19:00<14:53,  2.02it/s]
 84%|████████▎ | 9266/11074 [1:19:01<14:52,  2.03it/s]
 84%|████████▎ | 9267/11074 [1:19:01<14:52,  2.03it/s]
 84%|████████▎ | 9268/11074 [1:19:02<14:50,  2.03it/s]
 84%|████████▎ | 9269/11074 [1:19:02<14:50,  2.03it/s]
 84%|████████▎ | 9270/11074 [1:19:03<14:50,  2.03it/s]
 84%|████████▎ | 9271/11074 [1:19:03<14:49,  2.03it/s]
 84%|████████▎ | 9272/11074 [1:19:03<14:49,  2.03it/s]
 84%|████████▎ | 9273/11074 [1:19:04<14:48,  2.03it/s]
 84%|████████▎ | 9274/11074 [1:19:04<14:48,  2.02it/s]
 84%|████████▍ | 9275/11074 [1:19:05<14:47,  2.03it/s]{'loss': 3.185, 'grad_norm': 0.23950828611850739, 'learning_rate': 7.826899458113019e-05, 'epoch': 11.72}
                                                      
+
 84%|████████▍ | 9275/11074 [1:19:05<14:47,  2.03it/s]
 84%|████████▍ | 9276/11074 [1:19:05<14:49,  2.02it/s]
 84%|████████▍ | 9277/11074 [1:19:06<14:47,  2.02it/s]
 84%|████████▍ | 9278/11074 [1:19:06<14:47,  2.02it/s]
 84%|████████▍ | 9279/11074 [1:19:07<14:46,  2.03it/s]
 84%|████████▍ | 9280/11074 [1:19:07<14:45,  2.03it/s]
 84%|████████▍ | 9281/11074 [1:19:08<14:44,  2.03it/s]
 84%|████████▍ | 9282/11074 [1:19:08<14:44,  2.03it/s]
 84%|████████▍ | 9283/11074 [1:19:09<14:44,  2.02it/s]
 84%|████████▍ | 9284/11074 [1:19:09<14:44,  2.02it/s]
 84%|████████▍ | 9285/11074 [1:19:10<14:44,  2.02it/s]
 84%|████████▍ | 9286/11074 [1:19:10<14:43,  2.02it/s]
 84%|████████▍ | 9287/11074 [1:19:11<14:42,  2.02it/s]
 84%|████████▍ | 9288/11074 [1:19:11<14:41,  2.03it/s]
 84%|████████▍ | 9289/11074 [1:19:12<14:41,  2.02it/s]
 84%|████████▍ | 9290/11074 [1:19:12<14:40,  2.03it/s]
 84%|████████▍ | 9291/11074 [1:19:13<14:39,  2.03it/s]
 84%|████████▍ | 9292/11074 [1:19:13<14:39,  2.03it/s]
 84%|████████▍ | 9293/11074 [1:19:14<14:38,  2.03it/s]
 84%|████████▍ | 9294/11074 [1:19:14<14:39,  2.02it/s]
 84%|████████▍ | 9295/11074 [1:19:15<14:39,  2.02it/s]
 84%|████████▍ | 9296/11074 [1:19:15<14:39,  2.02it/s]
 84%|████████▍ | 9297/11074 [1:19:16<14:40,  2.02it/s]
 84%|████████▍ | 9298/11074 [1:19:16<14:38,  2.02it/s]
 84%|████████▍ | 9299/11074 [1:19:17<14:37,  2.02it/s]
 84%|████████▍ | 9300/11074 [1:19:17<14:36,  2.02it/s]
                                                      
{'loss': 3.1879, 'grad_norm': 0.24105164408683777, 'learning_rate': 7.616538008795282e-05, 'epoch': 11.75}
+
 84%|████████▍ | 9300/11074 [1:19:17<14:36,  2.02it/s]
 84%|████████▍ | 9301/11074 [1:19:18<14:37,  2.02it/s]
 84%|████████▍ | 9302/11074 [1:19:18<14:36,  2.02it/s]
 84%|████████▍ | 9303/11074 [1:19:19<14:34,  2.02it/s]
 84%|████████▍ | 9304/11074 [1:19:19<14:34,  2.02it/s]
 84%|████████▍ | 9305/11074 [1:19:20<14:33,  2.02it/s]
 84%|████████▍ | 9306/11074 [1:19:20<14:33,  2.02it/s]
 84%|████████▍ | 9307/11074 [1:19:21<14:32,  2.03it/s]
 84%|████████▍ | 9308/11074 [1:19:21<14:32,  2.02it/s]
 84%|████████▍ | 9309/11074 [1:19:22<14:31,  2.03it/s]
 84%|████████▍ | 9310/11074 [1:19:22<14:30,  2.03it/s]
 84%|████████▍ | 9311/11074 [1:19:23<14:31,  2.02it/s]
 84%|████████▍ | 9312/11074 [1:19:23<14:29,  2.03it/s]
 84%|████████▍ | 9313/11074 [1:19:24<14:29,  2.02it/s]
 84%|████████▍ | 9314/11074 [1:19:24<14:28,  2.03it/s]
 84%|████████▍ | 9315/11074 [1:19:25<14:28,  2.02it/s]
 84%|████████▍ | 9316/11074 [1:19:25<14:27,  2.03it/s]
 84%|████████▍ | 9317/11074 [1:19:26<14:27,  2.03it/s]
 84%|████████▍ | 9318/11074 [1:19:26<14:27,  2.02it/s]
 84%|████████▍ | 9319/11074 [1:19:27<14:26,  2.03it/s]
 84%|████████▍ | 9320/11074 [1:19:27<14:26,  2.03it/s]
 84%|████████▍ | 9321/11074 [1:19:28<14:25,  2.03it/s]
 84%|████████▍ | 9322/11074 [1:19:28<14:26,  2.02it/s]
 84%|████████▍ | 9323/11074 [1:19:29<14:24,  2.03it/s]
 84%|████████▍ | 9324/11074 [1:19:29<14:23,  2.03it/s]
 84%|████████▍ | 9325/11074 [1:19:30<14:23,  2.03it/s]{'loss': 3.1905, 'grad_norm': 0.2403911054134369, 'learning_rate': 7.408808840043912e-05, 'epoch': 11.78}
+
                                                      

 84%|████████▍ | 9325/11074 [1:19:30<14:23,  2.03it/s]
 84%|████████▍ | 9326/11074 [1:19:30<14:24,  2.02it/s]
 84%|████████▍ | 9327/11074 [1:19:31<14:23,  2.02it/s]
 84%|████████▍ | 9328/11074 [1:19:31<14:23,  2.02it/s]
 84%|████████▍ | 9329/11074 [1:19:32<14:23,  2.02it/s]
 84%|��███████▍ | 9330/11074 [1:19:32<14:22,  2.02it/s]
 84%|████████▍ | 9331/11074 [1:19:33<14:22,  2.02it/s]
 84%|████████▍ | 9332/11074 [1:19:33<14:21,  2.02it/s]
 84%|████████▍ | 9333/11074 [1:19:34<14:20,  2.02it/s]
 84%|████████▍ | 9334/11074 [1:19:34<14:19,  2.02it/s]
 84%|████████▍ | 9335/11074 [1:19:35<14:19,  2.02it/s]
 84%|████████▍ | 9336/11074 [1:19:35<14:18,  2.02it/s]
 84%|████████▍ | 9337/11074 [1:19:36<14:17,  2.03it/s]
 84%|████████▍ | 9338/11074 [1:19:36<14:17,  2.03it/s]
 84%|████████▍ | 9339/11074 [1:19:37<14:16,  2.03it/s]
 84%|████████▍ | 9340/11074 [1:19:37<14:16,  2.02it/s]
 84%|████████▍ | 9341/11074 [1:19:38<14:16,  2.02it/s]
 84%|████████▍ | 9342/11074 [1:19:38<14:16,  2.02it/s]
 84%|████████▍ | 9343/11074 [1:19:39<14:15,  2.02it/s]
 84%|████████▍ | 9344/11074 [1:19:39<14:16,  2.02it/s]
 84%|████████▍ | 9345/11074 [1:19:40<14:15,  2.02it/s]
 84%|████████▍ | 9346/11074 [1:19:40<14:14,  2.02it/s]
 84%|████████▍ | 9347/11074 [1:19:41<14:14,  2.02it/s]
 84%|████████▍ | 9348/11074 [1:19:41<14:14,  2.02it/s]
 84%|████████▍ | 9349/11074 [1:19:42<14:13,  2.02it/s]
 84%|████████▍ | 9350/11074 [1:19:42<14:13,  2.02it/s]{'loss': 3.1936, 'grad_norm': 0.24034592509269714, 'learning_rate': 7.20372485315176e-05, 'epoch': 11.81}
+
                                                      

 84%|████████▍ | 9350/11074 [1:19:42<14:13,  2.02it/s]
 84%|████████▍ | 9351/11074 [1:19:43<14:13,  2.02it/s]
 84%|████████▍ | 9352/11074 [1:19:43<14:12,  2.02it/s]
 84%|████████▍ | 9353/11074 [1:19:44<14:12,  2.02it/s]
 84%|████████▍ | 9354/11074 [1:19:44<14:11,  2.02it/s]
 84%|████████▍ | 9355/11074 [1:19:45<14:10,  2.02it/s]
 84%|████████▍ | 9356/11074 [1:19:45<14:09,  2.02it/s]
 84%|████████▍ | 9357/11074 [1:19:45<14:08,  2.02it/s]
 85%|████████▍ | 9358/11074 [1:19:46<14:08,  2.02it/s]
 85%|████████▍ | 9359/11074 [1:19:46<14:06,  2.03it/s]
 85%|████████▍ | 9360/11074 [1:19:47<14:06,  2.03it/s]
 85%|████████▍ | 9361/11074 [1:19:47<14:05,  2.02it/s]
 85%|████████▍ | 9362/11074 [1:19:48<14:05,  2.03it/s]
 85%|████████▍ | 9363/11074 [1:19:48<14:05,  2.02it/s]
 85%|████████▍ | 9364/11074 [1:19:49<14:04,  2.02it/s]
 85%|████████▍ | 9365/11074 [1:19:49<14:04,  2.02it/s]
 85%|████████▍ | 9366/11074 [1:19:50<14:04,  2.02it/s]
 85%|████████▍ | 9367/11074 [1:19:50<14:03,  2.02it/s]
 85%|████████▍ | 9368/11074 [1:19:51<14:03,  2.02it/s]
 85%|████████▍ | 9369/11074 [1:19:51<14:02,  2.02it/s]
 85%|████████▍ | 9370/11074 [1:19:52<14:02,  2.02it/s]
 85%|████████▍ | 9371/11074 [1:19:52<14:01,  2.02it/s]
 85%|████████▍ | 9372/11074 [1:19:53<14:01,  2.02it/s]
 85%|████████▍ | 9373/11074 [1:19:53<14:00,  2.02it/s]
 85%|████████▍ | 9374/11074 [1:19:54<13:59,  2.02it/s]
 85%|████████▍ | 9375/11074 [1:19:54<13:58,  2.03it/s]{'loss': 3.1969, 'grad_norm': 0.23983672261238098, 'learning_rate': 7.00129878512914e-05, 'epoch': 11.84}
                                                      
+
 85%|████████▍ | 9375/11074 [1:19:54<13:58,  2.03it/s]
 85%|████████▍ | 9376/11074 [1:19:55<14:00,  2.02it/s]
 85%|████████▍ | 9377/11074 [1:19:55<14:00,  2.02it/s]
 85%|████████▍ | 9378/11074 [1:19:56<13:59,  2.02it/s]
 85%|████████▍ | 9379/11074 [1:19:56<13:58,  2.02it/s]
 85%|████████▍ | 9380/11074 [1:19:57<13:58,  2.02it/s]
 85%|████████▍ | 9381/11074 [1:19:57<13:57,  2.02it/s]
 85%|████████▍ | 9382/11074 [1:19:58<13:56,  2.02it/s]
 85%|████████▍ | 9383/11074 [1:19:58<13:55,  2.02it/s]
 85%|████████▍ | 9384/11074 [1:19:59<13:54,  2.02it/s]
 85%|████████▍ | 9385/11074 [1:19:59<13:54,  2.02it/s]
 85%|████████▍ | 9386/11074 [1:20:00<13:54,  2.02it/s]
 85%|████████▍ | 9387/11074 [1:20:00<13:53,  2.02it/s]
 85%|████████▍ | 9388/11074 [1:20:01<13:54,  2.02it/s]
 85%|████████▍ | 9389/11074 [1:20:01<13:52,  2.02it/s]
 85%|████████▍ | 9390/11074 [1:20:02<13:51,  2.02it/s]
 85%|████████▍ | 9391/11074 [1:20:02<13:51,  2.02it/s]
 85%|████████▍ | 9392/11074 [1:20:03<13:50,  2.02it/s]
 85%|████████▍ | 9393/11074 [1:20:03<13:50,  2.02it/s]
 85%|████████▍ | 9394/11074 [1:20:04<13:49,  2.03it/s]
 85%|████████▍ | 9395/11074 [1:20:04<13:49,  2.02it/s]
 85%|████████▍ | 9396/11074 [1:20:05<13:48,  2.02it/s]
 85%|████████▍ | 9397/11074 [1:20:05<13:48,  2.02it/s]
 85%|████████▍ | 9398/11074 [1:20:06<13:48,  2.02it/s]
 85%|████████▍ | 9399/11074 [1:20:06<13:47,  2.02it/s]
 85%|████████▍ | 9400/11074 [1:20:07<13:47,  2.02it/s]
                                                      
{'loss': 3.1908, 'grad_norm': 0.23928986489772797, 'learning_rate': 6.80154320791283e-05, 'epoch': 11.88}
+
 85%|████████▍ | 9400/11074 [1:20:07<13:47,  2.02it/s]
 85%|████████▍ | 9401/11074 [1:20:07<13:48,  2.02it/s]
 85%|████████▍ | 9402/11074 [1:20:08<13:47,  2.02it/s]
 85%|████████▍ | 9403/11074 [1:20:08<13:47,  2.02it/s]
 85%|████████▍ | 9404/11074 [1:20:09<13:46,  2.02it/s]
 85%|████████▍ | 9405/11074 [1:20:09<13:45,  2.02it/s]
 85%|████████▍ | 9406/11074 [1:20:10<13:44,  2.02it/s]
 85%|████████▍ | 9407/11074 [1:20:10<13:44,  2.02it/s]
 85%|████████▍ | 9408/11074 [1:20:11<13:43,  2.02it/s]
 85%|████████▍ | 9409/11074 [1:20:11<13:43,  2.02it/s]
 85%|████████▍ | 9410/11074 [1:20:12<13:42,  2.02it/s]
 85%|████████▍ | 9411/11074 [1:20:12<13:41,  2.02it/s]
 85%|████████▍ | 9412/11074 [1:20:13<13:42,  2.02it/s]
 85%|████████▌ | 9413/11074 [1:20:13<13:42,  2.02it/s]
 85%|████████▌ | 9414/11074 [1:20:14<13:41,  2.02it/s]
 85%|████████▌ | 9415/11074 [1:20:14<13:40,  2.02it/s]
 85%|████████▌ | 9416/11074 [1:20:15<13:40,  2.02it/s]
 85%|████████▌ | 9417/11074 [1:20:15<13:39,  2.02it/s]
 85%|████████▌ | 9418/11074 [1:20:16<13:38,  2.02it/s]
 85%|████████▌ | 9419/11074 [1:20:16<13:38,  2.02it/s]
 85%|████████▌ | 9420/11074 [1:20:17<13:37,  2.02it/s]
 85%|████████▌ | 9421/11074 [1:20:17<13:37,  2.02it/s]
 85%|████████▌ | 9422/11074 [1:20:18<13:36,  2.02it/s]
 85%|████████▌ | 9423/11074 [1:20:18<13:35,  2.02it/s]
 85%|████████▌ | 9424/11074 [1:20:19<13:35,  2.02it/s]
 85%|████████▌ | 9425/11074 [1:20:19<13:34,  2.02it/s]{'loss': 3.1896, 'grad_norm': 0.2401139885187149, 'learning_rate': 6.604470527585304e-05, 'epoch': 11.91}
                                                      
+
 85%|████████▌ | 9425/11074 [1:20:19<13:34,  2.02it/s]
 85%|████████▌ | 9426/11074 [1:20:20<13:35,  2.02it/s]
 85%|████████▌ | 9427/11074 [1:20:20<13:34,  2.02it/s]
 85%|████████▌ | 9428/11074 [1:20:21<13:34,  2.02it/s]
 85%|████████▌ | 9429/11074 [1:20:21<13:33,  2.02it/s]
 85%|████████▌ | 9430/11074 [1:20:22<13:33,  2.02it/s]
 85%|████████▌ | 9431/11074 [1:20:22<13:32,  2.02it/s]
 85%|████████▌ | 9432/11074 [1:20:23<13:31,  2.02it/s]
 85%|████████▌ | 9433/11074 [1:20:23<13:30,  2.02it/s]
 85%|████████▌ | 9434/11074 [1:20:24<13:30,  2.02it/s]
 85%|████████▌ | 9435/11074 [1:20:24<13:31,  2.02it/s]
 85%|████████▌ | 9436/11074 [1:20:25<13:30,  2.02it/s]
 85%|████████▌ | 9437/11074 [1:20:25<13:28,  2.02it/s]
 85%|████████▌ | 9438/11074 [1:20:26<13:28,  2.02it/s]
 85%|████████▌ | 9439/11074 [1:20:26<13:27,  2.03it/s]
 85%|████████▌ | 9440/11074 [1:20:27<13:26,  2.03it/s]
 85%|████████▌ | 9441/11074 [1:20:27<13:25,  2.03it/s]
 85%|████████▌ | 9442/11074 [1:20:28<13:25,  2.02it/s]
 85%|████████▌ | 9443/11074 [1:20:28<13:25,  2.02it/s]
 85%|████████▌ | 9444/11074 [1:20:29<13:25,  2.02it/s]
 85%|████████▌ | 9445/11074 [1:20:29<13:25,  2.02it/s]
 85%|████████▌ | 9446/11074 [1:20:29<13:24,  2.02it/s]
 85%|████████▌ | 9447/11074 [1:20:30<13:24,  2.02it/s]
 85%|████████▌ | 9448/11074 [1:20:30<13:23,  2.02it/s]
 85%|████████▌ | 9449/11074 [1:20:31<13:23,  2.02it/s]
 85%|████████▌ | 9450/11074 [1:20:31<13:21,  2.03it/s]
                                                      
{'loss': 3.1938, 'grad_norm': 0.24146758019924164, 'learning_rate': 6.410092983604148e-05, 'epoch': 11.94}
+
 85%|████████▌ | 9450/11074 [1:20:31<13:21,  2.03it/s]
 85%|████████▌ | 9451/11074 [1:20:32<13:23,  2.02it/s]
 85%|████████▌ | 9452/11074 [1:20:32<13:21,  2.02it/s]
 85%|████████▌ | 9453/11074 [1:20:33<13:21,  2.02it/s]
 85%|████████▌ | 9454/11074 [1:20:33<13:19,  2.03it/s]
 85%|████████▌ | 9455/11074 [1:20:34<13:19,  2.02it/s]
 85%|████████▌ | 9456/11074 [1:20:34<13:18,  2.03it/s]
 85%|████████▌ | 9457/11074 [1:20:35<13:18,  2.03it/s]
 85%|████████▌ | 9458/11074 [1:20:35<13:19,  2.02it/s]
 85%|████████▌ | 9459/11074 [1:20:36<13:17,  2.03it/s]
 85%|████████▌ | 9460/11074 [1:20:36<13:17,  2.02it/s]
 85%|████████▌ | 9461/11074 [1:20:37<13:15,  2.03it/s]
 85%|████████▌ | 9462/11074 [1:20:37<13:15,  2.03it/s]
 85%|████████▌ | 9463/11074 [1:20:38<13:15,  2.03it/s]
 85%|████████▌ | 9464/11074 [1:20:38<13:14,  2.03it/s]
 85%|████████▌ | 9465/11074 [1:20:39<13:14,  2.02it/s]
 85%|████████▌ | 9466/11074 [1:20:39<13:13,  2.03it/s]
 85%|████████▌ | 9467/11074 [1:20:40<13:13,  2.03it/s]
 85%|████████▌ | 9468/11074 [1:20:40<13:12,  2.03it/s]
 86%|████████▌ | 9469/11074 [1:20:41<13:12,  2.03it/s]
 86%|████████▌ | 9470/11074 [1:20:41<13:12,  2.02it/s]
 86%|████████▌ | 9471/11074 [1:20:42<13:12,  2.02it/s]
 86%|████████▌ | 9472/11074 [1:20:42<13:12,  2.02it/s]
 86%|████████▌ | 9473/11074 [1:20:43<13:11,  2.02it/s]
 86%|████████▌ | 9474/11074 [1:20:43<13:12,  2.02it/s]
 86%|████████▌ | 9475/11074 [1:20:44<13:10,  2.02it/s]
                                                      
{'loss': 3.1885, 'grad_norm': 0.23939968645572662, 'learning_rate': 6.218422648042027e-05, 'epoch': 11.97}
+
 86%|████████▌ | 9475/11074 [1:20:44<13:10,  2.02it/s]
 86%|████████▌ | 9476/11074 [1:20:44<13:11,  2.02it/s]
 86%|████████▌ | 9477/11074 [1:20:45<13:10,  2.02it/s]
 86%|████████▌ | 9478/11074 [1:20:45<13:09,  2.02it/s]
 86%|████████▌ | 9479/11074 [1:20:46<13:08,  2.02it/s]
 86%|████████▌ | 9480/11074 [1:20:46<13:07,  2.02it/s]
 86%|████████▌ | 9481/11074 [1:20:47<13:07,  2.02it/s]
 86%|████████▌ | 9482/11074 [1:20:47<13:06,  2.03it/s]
 86%|████████▌ | 9483/11074 [1:20:48<13:05,  2.02it/s]
 86%|████████▌ | 9484/11074 [1:20:48<13:04,  2.03it/s]
 86%|████████▌ | 9485/11074 [1:20:49<13:04,  2.02it/s]
 86%|████████▌ | 9486/11074 [1:20:49<13:03,  2.03it/s]
 86%|████████▌ | 9487/11074 [1:20:50<13:03,  2.02it/s]
 86%|████████▌ | 9488/11074 [1:20:50<13:03,  2.02it/s]
 86%|████████▌ | 9489/11074 [1:20:51<13:03,  2.02it/s]
 86%|████████▌ | 9490/11074 [1:20:51<13:02,  2.02it/s]
 86%|████████▌ | 9491/11074 [1:20:52<13:02,  2.02it/s]
 86%|████████▌ | 9492/11074 [1:20:52<13:02,  2.02it/s]
 86%|████████▌ | 9493/11074 [1:20:53<13:01,  2.02it/s]
 86%|████████▌ | 9494/11074 [1:20:53<13:01,  2.02it/s]
 86%|████████▌ | 9495/11074 [1:20:54<13:01,  2.02it/s]
 86%|████████▌ | 9496/11074 [1:20:54<13:00,  2.02it/s]
 86%|████████▌ | 9497/11074 [1:20:55<12:59,  2.02it/s]
 86%|████████▌ | 9498/11074 [1:20:55<12:51,  2.04it/s]
 86%|████████▌ | 9499/11074 [1:21:07<1:45:56,  4.04s/it]
 86%|████████▌ | 9500/11074 [1:21:08<1:18:04,  2.98s/it]
                                                        
{'loss': 3.1825, 'grad_norm': 0.23993708193302155, 'learning_rate': 6.0294714248368074e-05, 'epoch': 12.0}
+
 86%|████████▌ | 9500/11074 [1:21:08<1:18:04,  2.98s/it]
 86%|████████▌ | 9501/11074 [1:21:08<58:33,  2.23s/it]  
 86%|████████▌ | 9502/11074 [1:21:09<44:55,  1.71s/it]
 86%|████████▌ | 9503/11074 [1:21:09<35:17,  1.35s/it]
 86%|████████▌ | 9504/11074 [1:21:10<28:34,  1.09s/it]
 86%|████████▌ | 9505/11074 [1:21:10<23:51,  1.10it/s]
 86%|████████▌ | 9506/11074 [1:21:11<20:34,  1.27it/s]
 86%|████████▌ | 9507/11074 [1:21:11<18:15,  1.43it/s]
 86%|████████▌ | 9508/11074 [1:21:12<16:38,  1.57it/s]
 86%|████████▌ | 9509/11074 [1:21:12<15:30,  1.68it/s]
 86%|████████▌ | 9510/11074 [1:21:13<14:44,  1.77it/s]
 86%|████████▌ | 9511/11074 [1:21:13<14:10,  1.84it/s]
 86%|████████▌ | 9512/11074 [1:21:14<13:48,  1.89it/s]
 86%|████████▌ | 9513/11074 [1:21:14<13:29,  1.93it/s]
 86%|████████▌ | 9514/11074 [1:21:15<13:17,  1.96it/s]
 86%|████████▌ | 9515/11074 [1:21:15<13:09,  1.97it/s]
 86%|████████▌ | 9516/11074 [1:21:16<13:04,  1.99it/s]
 86%|████████▌ | 9517/11074 [1:21:16<12:59,  2.00it/s]
 86%|████████▌ | 9518/11074 [1:21:17<14:03,  1.85it/s]
 86%|████████▌ | 9519/11074 [1:21:18<13:39,  1.90it/s]
 86%|████████▌ | 9520/11074 [1:21:18<13:24,  1.93it/s]
 86%|████████▌ | 9521/11074 [1:21:19<13:12,  1.96it/s]
 86%|████████▌ | 9522/11074 [1:21:19<13:04,  1.98it/s]
 86%|████████▌ | 9523/11074 [1:21:20<13:00,  1.99it/s]
 86%|████████▌ | 9524/11074 [1:21:20<12:55,  2.00it/s]
 86%|████████▌ | 9525/11074 [1:21:21<12:52,  2.00it/s]
                                                      {'loss': 3.1309, 'grad_norm': 0.2406255155801773, 'learning_rate': 5.8432510490523564e-05, 'epoch': 12.03}
+
 86%|████████▌ | 9525/11074 [1:21:21<12:52,  2.00it/s]
 86%|████████▌ | 9526/11074 [1:21:21<12:51,  2.01it/s]
 86%|████████▌ | 9527/11074 [1:21:22<12:49,  2.01it/s]
 86%|████████▌ | 9528/11074 [1:21:22<12:48,  2.01it/s]
 86%|████████▌ | 9529/11074 [1:21:22<12:46,  2.02it/s]
 86%|████████▌ | 9530/11074 [1:21:23<12:45,  2.02it/s]
 86%|████████▌ | 9531/11074 [1:21:23<12:43,  2.02it/s]
 86%|████████▌ | 9532/11074 [1:21:24<12:42,  2.02it/s]
 86%|████████▌ | 9533/11074 [1:21:24<12:42,  2.02it/s]
 86%|████████▌ | 9534/11074 [1:21:25<12:40,  2.03it/s]
 86%|████████▌ | 9535/11074 [1:21:25<12:39,  2.03it/s]
 86%|████████▌ | 9536/11074 [1:21:26<12:38,  2.03it/s]
 86%|████████▌ | 9537/11074 [1:21:26<12:38,  2.03it/s]
 86%|████████▌ | 9538/11074 [1:21:27<12:38,  2.03it/s]
 86%|████████▌ | 9539/11074 [1:21:27<12:37,  2.03it/s]
 86%|████████▌ | 9540/11074 [1:21:28<12:36,  2.03it/s]
 86%|████████▌ | 9541/11074 [1:21:28<12:35,  2.03it/s]
 86%|████████▌ | 9542/11074 [1:21:29<12:35,  2.03it/s]
 86%|████████▌ | 9543/11074 [1:21:29<12:35,  2.03it/s]
 86%|████████▌ | 9544/11074 [1:21:30<12:34,  2.03it/s]
 86%|████████▌ | 9545/11074 [1:21:30<12:34,  2.03it/s]
 86%|████████▌ | 9546/11074 [1:21:31<12:33,  2.03it/s]
 86%|████████▌ | 9547/11074 [1:21:31<12:33,  2.03it/s]
 86%|████████▌ | 9548/11074 [1:21:32<12:32,  2.03it/s]
 86%|████████▌ | 9549/11074 [1:21:32<12:31,  2.03it/s]
 86%|████████▌ | 9550/11074 [1:21:33<12:31,  2.03it/s]
                                                      
{'loss': 3.1295, 'grad_norm': 0.23934835195541382, 'learning_rate': 5.659773086149672e-05, 'epoch': 12.07}
+
 86%|████████▌ | 9550/11074 [1:21:33<12:31,  2.03it/s]
 86%|████████▌ | 9551/11074 [1:21:33<12:32,  2.02it/s]
 86%|████████▋ | 9552/11074 [1:21:34<12:31,  2.02it/s]
 86%|████████▋ | 9553/11074 [1:21:34<12:32,  2.02it/s]
 86%|████████▋ | 9554/11074 [1:21:35<12:31,  2.02it/s]
 86%|████████▋ | 9555/11074 [1:21:35<12:31,  2.02it/s]
 86%|████████▋ | 9556/11074 [1:21:36<12:30,  2.02it/s]
 86%|████████▋ | 9557/11074 [1:21:36<12:30,  2.02it/s]
 86%|████████▋ | 9558/11074 [1:21:37<12:29,  2.02it/s]
 86%|████████▋ | 9559/11074 [1:21:37<12:30,  2.02it/s]
 86%|████████▋ | 9560/11074 [1:21:38<12:29,  2.02it/s]
 86%|████████▋ | 9561/11074 [1:21:38<12:28,  2.02it/s]
 86%|████████▋ | 9562/11074 [1:21:39<12:27,  2.02it/s]
 86%|████████▋ | 9563/11074 [1:21:39<12:26,  2.02it/s]
 86%|████████▋ | 9564/11074 [1:21:40<12:26,  2.02it/s]
 86%|████████▋ | 9565/11074 [1:21:40<12:25,  2.02it/s]
 86%|████████▋ | 9566/11074 [1:21:41<12:24,  2.02it/s]
 86%|████████▋ | 9567/11074 [1:21:41<12:24,  2.02it/s]
 86%|████████▋ | 9568/11074 [1:21:42<12:23,  2.03it/s]
 86%|████████▋ | 9569/11074 [1:21:42<12:23,  2.02it/s]
 86%|████████▋ | 9570/11074 [1:21:43<12:23,  2.02it/s]
 86%|████████▋ | 9571/11074 [1:21:43<12:23,  2.02it/s]
 86%|████████▋ | 9572/11074 [1:21:44<12:22,  2.02it/s]
 86%|████████▋ | 9573/11074 [1:21:44<12:21,  2.02it/s]
 86%|████████▋ | 9574/11074 [1:21:45<12:21,  2.02it/s]
 86%|████████▋ | 9575/11074 [1:21:45<12:20,  2.02it/s]{'loss': 3.1358, 'grad_norm': 0.24136708676815033, 'learning_rate': 5.479048931268582e-05, 'epoch': 12.1}
+
                                                      

 86%|████████▋ | 9575/11074 [1:21:45<12:20,  2.02it/s]
 86%|████████▋ | 9576/11074 [1:21:46<12:23,  2.01it/s]
 86%|████████▋ | 9577/11074 [1:21:46<12:21,  2.02it/s]
 86%|████████▋ | 9578/11074 [1:21:47<12:20,  2.02it/s]
 86%|████████▋ | 9579/11074 [1:21:47<12:19,  2.02it/s]
 87%|████████▋ | 9580/11074 [1:21:48<12:18,  2.02it/s]
 87%|████████▋ | 9581/11074 [1:21:48<12:17,  2.02it/s]
 87%|████████▋ | 9582/11074 [1:21:49<12:16,  2.02it/s]
 87%|████████▋ | 9583/11074 [1:21:49<12:16,  2.02it/s]
 87%|████████▋ | 9584/11074 [1:21:50<12:15,  2.03it/s]
 87%|████████▋ | 9585/11074 [1:21:50<12:15,  2.02it/s]
 87%|████████▋ | 9586/11074 [1:21:51<12:14,  2.03it/s]
 87%|████████▋ | 9587/11074 [1:21:51<12:14,  2.03it/s]
 87%|████████▋ | 9588/11074 [1:21:52<12:12,  2.03it/s]
 87%|████████▋ | 9589/11074 [1:21:52<12:12,  2.03it/s]
 87%|████████▋ | 9590/11074 [1:21:53<12:12,  2.03it/s]
 87%|████████▋ | 9591/11074 [1:21:53<12:11,  2.03it/s]
 87%|████████▋ | 9592/11074 [1:21:54<12:11,  2.03it/s]
 87%|████████▋ | 9593/11074 [1:21:54<12:10,  2.03it/s]
 87%|████████▋ | 9594/11074 [1:21:55<12:11,  2.02it/s]
 87%|████████▋ | 9595/11074 [1:21:55<12:09,  2.03it/s]
 87%|████████▋ | 9596/11074 [1:21:56<12:09,  2.03it/s]
 87%|████████▋ | 9597/11074 [1:21:56<12:08,  2.03it/s]
 87%|████████▋ | 9598/11074 [1:21:57<12:08,  2.03it/s]
 87%|████████▋ | 9599/11074 [1:21:57<12:08,  2.03it/s]
 87%|████████▋ | 9600/11074 [1:21:58<12:06,  2.03it/s]{'loss': 3.1398, 'grad_norm': 0.23873379826545715, 'learning_rate': 5.301089808520049e-05, 'epoch': 12.13}
+
                                                      

 87%|████████▋ | 9600/11074 [1:21:58<12:06,  2.03it/s]
 87%|████████▋ | 9601/11074 [1:21:58<12:08,  2.02it/s]
 87%|████████▋ | 9602/11074 [1:21:59<12:06,  2.03it/s]
 87%|████████▋ | 9603/11074 [1:21:59<12:06,  2.02it/s]
 87%|████████▋ | 9604/11074 [1:22:00<12:05,  2.03it/s]
 87%|████████▋ | 9605/11074 [1:22:00<12:05,  2.03it/s]
 87%|████████▋ | 9606/11074 [1:22:01<12:05,  2.02it/s]
 87%|████████▋ | 9607/11074 [1:22:01<12:05,  2.02it/s]
 87%|████████▋ | 9608/11074 [1:22:02<12:04,  2.02it/s]
 87%|████████▋ | 9609/11074 [1:22:02<12:04,  2.02it/s]
 87%|████████▋ | 9610/11074 [1:22:02<12:02,  2.03it/s]
 87%|████████▋ | 9611/11074 [1:22:03<13:04,  1.86it/s]
 87%|████████▋ | 9612/11074 [1:22:04<12:45,  1.91it/s]
 87%|████████▋ | 9613/11074 [1:22:04<12:32,  1.94it/s]
 87%|████████▋ | 9614/11074 [1:22:05<12:22,  1.97it/s]
 87%|████████▋ | 9615/11074 [1:22:05<12:15,  1.98it/s]
 87%|████████▋ | 9616/11074 [1:22:06<12:10,  1.99it/s]
 87%|████████▋ | 9617/11074 [1:22:06<12:06,  2.01it/s]
 87%|████████▋ | 9618/11074 [1:22:07<12:04,  2.01it/s]
 87%|████████▋ | 9619/11074 [1:22:07<12:02,  2.01it/s]
 87%|████████▋ | 9620/11074 [1:22:08<12:00,  2.02it/s]
 87%|████████▋ | 9621/11074 [1:22:08<11:59,  2.02it/s]
 87%|████████▋ | 9622/11074 [1:22:09<12:00,  2.02it/s]
 87%|████████▋ | 9623/11074 [1:22:09<11:58,  2.02it/s]
 87%|████████▋ | 9624/11074 [1:22:10<11:56,  2.02it/s]
 87%|████████▋ | 9625/11074 [1:22:10<11:56,  2.02it/s]{'loss': 3.1425, 'grad_norm': 0.24040587246418, 'learning_rate': 5.125906770289113e-05, 'epoch': 12.16}
                                                      
+
 87%|████████▋ | 9625/11074 [1:22:10<11:56,  2.02it/s]
 87%|████████▋ | 9626/11074 [1:22:11<11:55,  2.02it/s]
 87%|████████▋ | 9627/11074 [1:22:11<11:55,  2.02it/s]
 87%|████████▋ | 9628/11074 [1:22:12<11:54,  2.02it/s]
 87%|████████▋ | 9629/11074 [1:22:12<11:54,  2.02it/s]
 87%|████████▋ | 9630/11074 [1:22:13<11:53,  2.02it/s]
 87%|████████▋ | 9631/11074 [1:22:13<11:53,  2.02it/s]
 87%|████████▋ | 9632/11074 [1:22:14<11:52,  2.02it/s]
 87%|████████▋ | 9633/11074 [1:22:14<11:52,  2.02it/s]
 87%|████████▋ | 9634/11074 [1:22:15<11:52,  2.02it/s]
 87%|████████▋ | 9635/11074 [1:22:15<11:51,  2.02it/s]
 87%|████████▋ | 9636/11074 [1:22:15<11:50,  2.02it/s]
 87%|████████▋ | 9637/11074 [1:22:16<11:50,  2.02it/s]
 87%|████████▋ | 9638/11074 [1:22:16<11:50,  2.02it/s]
 87%|████████▋ | 9639/11074 [1:22:17<11:49,  2.02it/s]
 87%|████████▋ | 9640/11074 [1:22:17<11:48,  2.02it/s]
 87%|████████▋ | 9641/11074 [1:22:18<11:49,  2.02it/s]
 87%|████████▋ | 9642/11074 [1:22:18<11:47,  2.02it/s]
 87%|████████▋ | 9643/11074 [1:22:19<11:47,  2.02it/s]
 87%|████████▋ | 9644/11074 [1:22:19<11:46,  2.02it/s]
 87%|████████▋ | 9645/11074 [1:22:20<11:46,  2.02it/s]
 87%|████████▋ | 9646/11074 [1:22:20<11:45,  2.02it/s]
 87%|████████▋ | 9647/11074 [1:22:21<11:45,  2.02it/s]
 87%|████████▋ | 9648/11074 [1:22:21<11:44,  2.02it/s]
 87%|████████▋ | 9649/11074 [1:22:22<11:43,  2.02it/s]
 87%|████████▋ | 9650/11074 [1:22:22<11:44,  2.02it/s]{'loss': 3.1417, 'grad_norm': 0.24300391972064972, 'learning_rate': 4.953510696548413e-05, 'epoch': 12.19}
+
                                                      

 87%|████████▋ | 9650/11074 [1:22:22<11:44,  2.02it/s]
 87%|████████▋ | 9651/11074 [1:22:23<11:43,  2.02it/s]
 87%|████████▋ | 9652/11074 [1:22:23<11:43,  2.02it/s]
 87%|████████▋ | 9653/11074 [1:22:24<11:41,  2.03it/s]
 87%|████████▋ | 9654/11074 [1:22:24<11:41,  2.02it/s]
 87%|████████▋ | 9655/11074 [1:22:25<11:40,  2.03it/s]
 87%|████████▋ | 9656/11074 [1:22:25<11:40,  2.03it/s]
 87%|████████▋ | 9657/11074 [1:22:26<11:39,  2.03it/s]
 87%|████████▋ | 9658/11074 [1:22:26<11:39,  2.03it/s]
 87%|████████▋ | 9659/11074 [1:22:27<11:38,  2.03it/s]
 87%|████████▋ | 9660/11074 [1:22:27<11:38,  2.02it/s]
 87%|████████▋ | 9661/11074 [1:22:28<11:38,  2.02it/s]
 87%|████████▋ | 9662/11074 [1:22:28<11:37,  2.02it/s]
 87%|████████▋ | 9663/11074 [1:22:29<11:37,  2.02it/s]
 87%|████████▋ | 9664/11074 [1:22:29<11:36,  2.02it/s]
 87%|████████▋ | 9665/11074 [1:22:30<11:36,  2.02it/s]
 87%|████████▋ | 9666/11074 [1:22:30<11:35,  2.02it/s]
 87%|████████▋ | 9667/11074 [1:22:31<11:35,  2.02it/s]
 87%|████████▋ | 9668/11074 [1:22:31<11:34,  2.02it/s]
 87%|████████▋ | 9669/11074 [1:22:32<11:34,  2.02it/s]
 87%|████████▋ | 9670/11074 [1:22:32<11:33,  2.02it/s]
 87%|████████▋ | 9671/11074 [1:22:33<11:33,  2.02it/s]
 87%|████████▋ | 9672/11074 [1:22:33<11:32,  2.02it/s]
 87%|████████▋ | 9673/11074 [1:22:34<11:32,  2.02it/s]
 87%|████████▋ | 9674/11074 [1:22:34<11:31,  2.02it/s]
 87%|████████▋ | 9675/11074 [1:22:35<11:31,  2.02it/s]{'loss': 3.1396, 'grad_norm': 0.24217967689037323, 'learning_rate': 4.783912294182485e-05, 'epoch': 12.22}
                                                      
+
 87%|████████▋ | 9675/11074 [1:22:35<11:31,  2.02it/s]
 87%|████████▋ | 9676/11074 [1:22:35<11:35,  2.01it/s]
 87%|████████▋ | 9677/11074 [1:22:36<11:33,  2.02it/s]
 87%|████████▋ | 9678/11074 [1:22:36<11:32,  2.02it/s]
 87%|████████▋ | 9679/11074 [1:22:37<11:30,  2.02it/s]
 87%|████████▋ | 9680/11074 [1:22:37<11:29,  2.02it/s]
 87%|████████▋ | 9681/11074 [1:22:38<11:28,  2.02it/s]
 87%|████████▋ | 9682/11074 [1:22:38<11:27,  2.02it/s]
 87%|████████▋ | 9683/11074 [1:22:39<11:27,  2.02it/s]
 87%|████████▋ | 9684/11074 [1:22:39<11:26,  2.03it/s]
 87%|████████▋ | 9685/11074 [1:22:40<11:26,  2.02it/s]
 87%|████████▋ | 9686/11074 [1:22:40<11:25,  2.03it/s]
 87%|████████▋ | 9687/11074 [1:22:41<11:25,  2.02it/s]
 87%|████████▋ | 9688/11074 [1:22:41<11:24,  2.03it/s]
 87%|████████▋ | 9689/11074 [1:22:42<11:24,  2.02it/s]
 88%|████████▊ | 9690/11074 [1:22:42<11:23,  2.02it/s]
 88%|████████▊ | 9691/11074 [1:22:43<11:23,  2.02it/s]
 88%|████████▊ | 9692/11074 [1:22:43<11:22,  2.02it/s]
 88%|████████▊ | 9693/11074 [1:22:44<11:21,  2.03it/s]
 88%|████████▊ | 9694/11074 [1:22:44<11:21,  2.02it/s]
 88%|████████▊ | 9695/11074 [1:22:45<11:20,  2.03it/s]
 88%|████████▊ | 9696/11074 [1:22:45<11:20,  2.02it/s]
 88%|████████▊ | 9697/11074 [1:22:46<11:19,  2.03it/s]
 88%|████████▊ | 9698/11074 [1:22:46<11:19,  2.02it/s]
 88%|████████▊ | 9699/11074 [1:22:47<11:19,  2.02it/s]
 88%|████████▊ | 9700/11074 [1:22:47<11:18,  2.02it/s]{'loss': 3.1419, 'grad_norm': 0.2392173707485199, 'learning_rate': 4.617122096322862e-05, 'epoch': 12.26}
                                                      
+
 88%|████████▊ | 9700/11074 [1:22:47<11:18,  2.02it/s]
 88%|████████▊ | 9701/11074 [1:22:48<11:18,  2.02it/s]
 88%|███���████▊ | 9702/11074 [1:22:48<11:18,  2.02it/s]
 88%|████████▊ | 9703/11074 [1:22:49<11:17,  2.02it/s]
 88%|████████▊ | 9704/11074 [1:22:49<11:16,  2.02it/s]
 88%|████████▊ | 9705/11074 [1:22:50<11:16,  2.02it/s]
 88%|████████▊ | 9706/11074 [1:22:50<11:15,  2.03it/s]
 88%|████████▊ | 9707/11074 [1:22:51<11:15,  2.02it/s]
 88%|████████▊ | 9708/11074 [1:22:51<11:14,  2.03it/s]
 88%|████████▊ | 9709/11074 [1:22:52<11:14,  2.02it/s]
 88%|████████▊ | 9710/11074 [1:22:52<11:13,  2.02it/s]
 88%|████████▊ | 9711/11074 [1:22:53<11:13,  2.02it/s]
 88%|████████▊ | 9712/11074 [1:22:53<11:12,  2.02it/s]
 88%|████████▊ | 9713/11074 [1:22:54<11:12,  2.02it/s]
 88%|████████▊ | 9714/11074 [1:22:54<11:12,  2.02it/s]
 88%|████████▊ | 9715/11074 [1:22:55<11:11,  2.02it/s]
 88%|████████▊ | 9716/11074 [1:22:55<11:10,  2.02it/s]
 88%|████████▊ | 9717/11074 [1:22:56<11:10,  2.02it/s]
 88%|████████▊ | 9718/11074 [1:22:56<11:09,  2.02it/s]
 88%|████████▊ | 9719/11074 [1:22:57<11:08,  2.03it/s]
 88%|████████▊ | 9720/11074 [1:22:57<11:08,  2.02it/s]
 88%|████████▊ | 9721/11074 [1:22:57<11:08,  2.02it/s]
 88%|████████▊ | 9722/11074 [1:22:58<11:07,  2.02it/s]
 88%|████████▊ | 9723/11074 [1:22:58<11:07,  2.02it/s]
 88%|████████▊ | 9724/11074 [1:22:59<11:06,  2.02it/s]
 88%|████████▊ | 9725/11074 [1:22:59<11:05,  2.03it/s]{'loss': 3.1522, 'grad_norm': 0.24311304092407227, 'learning_rate': 4.453150461693789e-05, 'epoch': 12.29}
+
                                                      

 88%|████████▊ | 9725/11074 [1:22:59<11:05,  2.03it/s]
 88%|████████▊ | 9726/11074 [1:23:00<11:08,  2.02it/s]
 88%|████████▊ | 9727/11074 [1:23:00<11:07,  2.02it/s]
 88%|████████▊ | 9728/11074 [1:23:01<11:06,  2.02it/s]
 88%|████████▊ | 9729/11074 [1:23:01<11:05,  2.02it/s]
 88%|████████▊ | 9730/11074 [1:23:02<11:04,  2.02it/s]
 88%|████████▊ | 9731/11074 [1:23:02<11:04,  2.02it/s]
 88%|████████▊ | 9732/11074 [1:23:03<11:03,  2.02it/s]
 88%|████████▊ | 9733/11074 [1:23:03<11:02,  2.02it/s]
 88%|████████▊ | 9734/11074 [1:23:04<11:02,  2.02it/s]
 88%|████████▊ | 9735/11074 [1:23:04<11:02,  2.02it/s]
 88%|████████▊ | 9736/11074 [1:23:05<11:01,  2.02it/s]
 88%|████████▊ | 9737/11074 [1:23:05<11:01,  2.02it/s]
 88%|████████▊ | 9738/11074 [1:23:06<11:00,  2.02it/s]
 88%|████████▊ | 9739/11074 [1:23:06<11:00,  2.02it/s]
 88%|████████▊ | 9740/11074 [1:23:07<10:59,  2.02it/s]
 88%|████████▊ | 9741/11074 [1:23:07<10:58,  2.02it/s]
 88%|████████▊ | 9742/11074 [1:23:08<10:57,  2.03it/s]
 88%|████████▊ | 9743/11074 [1:23:08<10:57,  2.02it/s]
 88%|████████▊ | 9744/11074 [1:23:09<10:56,  2.02it/s]
 88%|████████▊ | 9745/11074 [1:23:09<10:56,  2.03it/s]
 88%|████████▊ | 9746/11074 [1:23:10<10:56,  2.02it/s]
 88%|████████▊ | 9747/11074 [1:23:10<10:54,  2.03it/s]
 88%|████████▊ | 9748/11074 [1:23:11<10:55,  2.02it/s]
 88%|████████▊ | 9749/11074 [1:23:11<10:54,  2.03it/s]
 88%|████████▊ | 9750/11074 [1:23:12<10:53,  2.03it/s]
                                                      
{'loss': 3.155, 'grad_norm': 0.2436213493347168, 'learning_rate': 4.2920075739689926e-05, 'epoch': 12.32}
+
 88%|████████▊ | 9750/11074 [1:23:12<10:53,  2.03it/s]
 88%|████████▊ | 9751/11074 [1:23:12<10:53,  2.02it/s]
 88%|████████▊ | 9752/11074 [1:23:13<10:53,  2.02it/s]
 88%|████████▊ | 9753/11074 [1:23:13<10:53,  2.02it/s]
 88%|████████▊ | 9754/11074 [1:23:14<10:52,  2.02it/s]
 88%|████████▊ | 9755/11074 [1:23:14<10:52,  2.02it/s]
 88%|████████▊ | 9756/11074 [1:23:15<10:51,  2.02it/s]
 88%|████████▊ | 9757/11074 [1:23:15<10:50,  2.02it/s]
 88%|████████▊ | 9758/11074 [1:23:16<10:51,  2.02it/s]
 88%|████████▊ | 9759/11074 [1:23:16<10:50,  2.02it/s]
 88%|████████▊ | 9760/11074 [1:23:17<10:50,  2.02it/s]
 88%|████████▊ | 9761/11074 [1:23:17<10:49,  2.02it/s]
 88%|████████▊ | 9762/11074 [1:23:18<10:48,  2.02it/s]
 88%|████████▊ | 9763/11074 [1:23:18<10:47,  2.02it/s]
 88%|████████▊ | 9764/11074 [1:23:19<10:46,  2.03it/s]
 88%|████████▊ | 9765/11074 [1:23:19<10:46,  2.03it/s]
 88%|████████▊ | 9766/11074 [1:23:20<10:45,  2.03it/s]
 88%|████████▊ | 9767/11074 [1:23:20<10:45,  2.02it/s]
 88%|████████▊ | 9768/11074 [1:23:21<10:45,  2.02it/s]
 88%|████████▊ | 9769/11074 [1:23:21<10:45,  2.02it/s]
 88%|████████▊ | 9770/11074 [1:23:22<10:44,  2.02it/s]
 88%|████████▊ | 9771/11074 [1:23:22<10:43,  2.03it/s]
 88%|████████▊ | 9772/11074 [1:23:23<10:43,  2.02it/s]
 88%|████████▊ | 9773/11074 [1:23:23<10:41,  2.03it/s]
 88%|████████▊ | 9774/11074 [1:23:24<10:41,  2.03it/s]
 88%|████████▊ | 9775/11074 [1:23:24<10:40,  2.03it/s]{'loss': 3.1382, 'grad_norm': 0.2395491898059845, 'learning_rate': 4.1337034411391136e-05, 'epoch': 12.35}
+
                                                      

 88%|████████▊ | 9775/11074 [1:23:24<10:40,  2.03it/s]
 88%|████████▊ | 9776/11074 [1:23:25<10:41,  2.02it/s]
 88%|████████▊ | 9777/11074 [1:23:25<10:40,  2.02it/s]
 88%|████████▊ | 9778/11074 [1:23:26<10:39,  2.03it/s]
 88%|████████▊ | 9779/11074 [1:23:26<10:39,  2.03it/s]
 88%|████████▊ | 9780/11074 [1:23:27<10:39,  2.02it/s]
 88%|████████▊ | 9781/11074 [1:23:27<10:38,  2.02it/s]
 88%|████████▊ | 9782/11074 [1:23:28<10:38,  2.02it/s]
 88%|████████▊ | 9783/11074 [1:23:28<10:38,  2.02it/s]
 88%|████████▊ | 9784/11074 [1:23:29<10:37,  2.02it/s]
 88%|████████▊ | 9785/11074 [1:23:29<10:37,  2.02it/s]
 88%|████████▊ | 9786/11074 [1:23:30<10:36,  2.02it/s]
 88%|████████▊ | 9787/11074 [1:23:30<10:36,  2.02it/s]
 88%|████████▊ | 9788/11074 [1:23:31<10:35,  2.02it/s]
 88%|████████▊ | 9789/11074 [1:23:31<10:35,  2.02it/s]
 88%|████████▊ | 9790/11074 [1:23:32<10:34,  2.02it/s]
 88%|████████▊ | 9791/11074 [1:23:32<10:33,  2.02it/s]
 88%|████████▊ | 9792/11074 [1:23:33<10:33,  2.02it/s]
 88%|████████▊ | 9793/11074 [1:23:33<10:33,  2.02it/s]
 88%|████████▊ | 9794/11074 [1:23:34<10:32,  2.02it/s]
 88%|████████▊ | 9795/11074 [1:23:34<10:32,  2.02it/s]
 88%|████████▊ | 9796/11074 [1:23:35<10:31,  2.02it/s]
 88%|████████▊ | 9797/11074 [1:23:35<10:30,  2.02it/s]
 88%|████████▊ | 9798/11074 [1:23:36<10:30,  2.03it/s]
 88%|████████▊ | 9799/11074 [1:23:36<10:29,  2.03it/s]
 88%|████████▊ | 9800/11074 [1:23:37<10:28,  2.03it/s]{'loss': 3.1497, 'grad_norm': 0.24011148512363434, 'learning_rate': 3.978247894890219e-05, 'epoch': 12.38}
                                                      
+
 88%|████████▊ | 9800/11074 [1:23:37<10:28,  2.03it/s]
 89%|████████▊ | 9801/11074 [1:23:37<10:29,  2.02it/s]
 89%|████████▊ | 9802/11074 [1:23:38<10:28,  2.02it/s]
 89%|████████▊ | 9803/11074 [1:23:38<10:28,  2.02it/s]
 89%|████████▊ | 9804/11074 [1:23:39<10:28,  2.02it/s]
 89%|████████▊ | 9805/11074 [1:23:39<10:27,  2.02it/s]
 89%|████████▊ | 9806/11074 [1:23:40<10:26,  2.02it/s]
 89%|████████▊ | 9807/11074 [1:23:40<10:26,  2.02it/s]
 89%|████████▊ | 9808/11074 [1:23:40<10:25,  2.02it/s]
 89%|████████▊ | 9809/11074 [1:23:41<10:25,  2.02it/s]
 89%|████████▊ | 9810/11074 [1:23:41<10:24,  2.02it/s]
 89%|████████▊ | 9811/11074 [1:23:42<10:24,  2.02it/s]
 89%|████████▊ | 9812/11074 [1:23:42<10:24,  2.02it/s]
 89%|████████▊ | 9813/11074 [1:23:43<10:22,  2.02it/s]
 89%|████████▊ | 9814/11074 [1:23:43<10:22,  2.02it/s]
 89%|████████▊ | 9815/11074 [1:23:44<10:21,  2.03it/s]
 89%|████████▊ | 9816/11074 [1:23:44<10:21,  2.02it/s]
 89%|████████▊ | 9817/11074 [1:23:45<10:20,  2.02it/s]
 89%|████████▊ | 9818/11074 [1:23:45<10:20,  2.03it/s]
 89%|████████▊ | 9819/11074 [1:23:46<10:19,  2.03it/s]
 89%|████████▊ | 9820/11074 [1:23:46<10:19,  2.02it/s]
 89%|████████▊ | 9821/11074 [1:23:47<10:19,  2.02it/s]
 89%|████████▊ | 9822/11074 [1:23:47<10:19,  2.02it/s]
 89%|████████▊ | 9823/11074 [1:23:48<10:18,  2.02it/s]
 89%|████████▊ | 9824/11074 [1:23:48<10:17,  2.02it/s]
 89%|████████▊ | 9825/11074 [1:23:49<10:17,  2.02it/s]{'loss': 3.1444, 'grad_norm': 0.24186104536056519, 'learning_rate': 3.825650589993163e-05, 'epoch': 12.41}
                                                      
+
 89%|████████▊ | 9825/11074 [1:23:49<10:17,  2.02it/s]
 89%|████��███▊ | 9826/11074 [1:23:49<10:18,  2.02it/s]
 89%|████████▊ | 9827/11074 [1:23:50<10:16,  2.02it/s]
 89%|████████▊ | 9828/11074 [1:23:50<10:16,  2.02it/s]
 89%|████████▉ | 9829/11074 [1:23:51<10:15,  2.02it/s]
 89%|████████▉ | 9830/11074 [1:23:51<10:15,  2.02it/s]
 89%|████████▉ | 9831/11074 [1:23:52<10:15,  2.02it/s]
 89%|████████▉ | 9832/11074 [1:23:52<10:14,  2.02it/s]
 89%|████████▉ | 9833/11074 [1:23:53<10:13,  2.02it/s]
 89%|████████▉ | 9834/11074 [1:23:53<10:12,  2.02it/s]
 89%|████████▉ | 9835/11074 [1:23:54<10:11,  2.03it/s]
 89%|████████▉ | 9836/11074 [1:23:54<10:11,  2.02it/s]
 89%|████████▉ | 9837/11074 [1:23:55<10:10,  2.02it/s]
 89%|████████▉ | 9838/11074 [1:23:55<10:10,  2.03it/s]
 89%|████████▉ | 9839/11074 [1:23:56<10:09,  2.02it/s]
 89%|████████▉ | 9840/11074 [1:23:56<10:09,  2.02it/s]
 89%|████████▉ | 9841/11074 [1:23:57<10:09,  2.02it/s]
 89%|████████▉ | 9842/11074 [1:23:57<10:08,  2.02it/s]
 89%|████████▉ | 9843/11074 [1:23:58<10:08,  2.02it/s]
 89%|████████▉ | 9844/11074 [1:23:58<10:06,  2.03it/s]
 89%|████████▉ | 9845/11074 [1:23:59<10:06,  2.03it/s]
 89%|████████▉ | 9846/11074 [1:23:59<10:06,  2.03it/s]
 89%|████████▉ | 9847/11074 [1:24:00<10:06,  2.02it/s]
 89%|████████▉ | 9848/11074 [1:24:00<10:06,  2.02it/s]
 89%|████████▉ | 9849/11074 [1:24:01<10:05,  2.02it/s]
 89%|████████▉ | 9850/11074 [1:24:01<10:05,  2.02it/s]
                                                      
{'loss': 3.148, 'grad_norm': 0.242679163813591, 'learning_rate': 3.675921003703958e-05, 'epoch': 12.44}
+
 89%|████████▉ | 9850/11074 [1:24:01<10:05,  2.02it/s]
 89%|████████▉ | 9851/11074 [1:24:02<10:04,  2.02it/s]
 89%|████████▉ | 9852/11074 [1:24:02<10:03,  2.02it/s]
 89%|████████▉ | 9853/11074 [1:24:03<10:03,  2.02it/s]
 89%|████████▉ | 9854/11074 [1:24:03<10:02,  2.02it/s]
 89%|████████▉ | 9855/11074 [1:24:04<10:02,  2.02it/s]
 89%|████████▉ | 9856/11074 [1:24:04<10:01,  2.02it/s]
 89%|████████▉ | 9857/11074 [1:24:05<10:01,  2.02it/s]
 89%|████████▉ | 9858/11074 [1:24:05<10:00,  2.03it/s]
 89%|████████▉ | 9859/11074 [1:24:06<10:00,  2.02it/s]
 89%|████████▉ | 9860/11074 [1:24:06<09:59,  2.02it/s]
 89%|████████▉ | 9861/11074 [1:24:07<09:59,  2.02it/s]
 89%|████████▉ | 9862/11074 [1:24:07<09:58,  2.03it/s]
 89%|████████▉ | 9863/11074 [1:24:08<09:58,  2.02it/s]
 89%|████████▉ | 9864/11074 [1:24:08<09:57,  2.03it/s]
 89%|████████▉ | 9865/11074 [1:24:09<09:56,  2.03it/s]
 89%|████████▉ | 9866/11074 [1:24:09<09:56,  2.03it/s]
 89%|████████▉ | 9867/11074 [1:24:10<09:55,  2.03it/s]
 89%|████████▉ | 9868/11074 [1:24:10<09:55,  2.03it/s]
 89%|████████▉ | 9869/11074 [1:24:11<09:54,  2.03it/s]
 89%|████████▉ | 9870/11074 [1:24:11<09:54,  2.03it/s]
 89%|████████▉ | 9871/11074 [1:24:12<09:53,  2.03it/s]
 89%|████████▉ | 9872/11074 [1:24:12<09:53,  2.03it/s]
 89%|████████▉ | 9873/11074 [1:24:13<09:52,  2.03it/s]
 89%|████████▉ | 9874/11074 [1:24:13<09:52,  2.03it/s]
 89%|████████▉ | 9875/11074 [1:24:14<09:52,  2.02it/s]{'loss': 3.148, 'grad_norm': 0.2414790391921997, 'learning_rate': 3.52906843517517e-05, 'epoch': 12.48}
+
                                                      

 89%|████████▉ | 9875/11074 [1:24:14<09:52,  2.02it/s]
 89%|████████▉ | 9876/11074 [1:24:14<09:53,  2.02it/s]
 89%|████████▉ | 9877/11074 [1:24:15<09:52,  2.02it/s]
 89%|████████▉ | 9878/11074 [1:24:15<09:51,  2.02it/s]
 89%|████████▉ | 9879/11074 [1:24:16<09:50,  2.02it/s]
 89%|████████▉ | 9880/11074 [1:24:16<09:50,  2.02it/s]
 89%|████████▉ | 9881/11074 [1:24:17<09:49,  2.02it/s]
 89%|████████▉ | 9882/11074 [1:24:17<09:48,  2.02it/s]
 89%|████████▉ | 9883/11074 [1:24:18<09:47,  2.03it/s]
 89%|████████▉ | 9884/11074 [1:24:18<09:47,  2.03it/s]
 89%|████████▉ | 9885/11074 [1:24:19<09:46,  2.03it/s]
 89%|████████▉ | 9886/11074 [1:24:19<09:45,  2.03it/s]
 89%|████████▉ | 9887/11074 [1:24:20<09:45,  2.03it/s]
 89%|████████▉ | 9888/11074 [1:24:20<09:45,  2.03it/s]
 89%|████████▉ | 9889/11074 [1:24:21<09:44,  2.03it/s]
 89%|████████▉ | 9890/11074 [1:24:21<09:44,  2.03it/s]
 89%|████████▉ | 9891/11074 [1:24:21<09:43,  2.03it/s]
 89%|████████▉ | 9892/11074 [1:24:22<09:43,  2.03it/s]
 89%|████████▉ | 9893/11074 [1:24:22<09:43,  2.02it/s]
 89%|████████▉ | 9894/11074 [1:24:23<09:42,  2.02it/s]
 89%|████████▉ | 9895/11074 [1:24:23<09:41,  2.03it/s]
 89%|████████▉ | 9896/11074 [1:24:24<09:42,  2.02it/s]
 89%|████████▉ | 9897/11074 [1:24:24<09:41,  2.03it/s]
 89%|████████▉ | 9898/11074 [1:24:25<09:41,  2.02it/s]
 89%|████████▉ | 9899/11074 [1:24:25<09:39,  2.03it/s]
 89%|████████▉ | 9900/11074 [1:24:26<09:39,  2.02it/s]
                                                      {'loss': 3.1459, 'grad_norm': 0.23893502354621887, 'learning_rate': 3.38510200487846e-05, 'epoch': 12.51}
+

 89%|████████▉ | 9900/11074 [1:24:26<09:39,  2.02it/s]
 89%|████████▉ | 9901/11074 [1:24:26<09:40,  2.02it/s]
 89%|████████▉ | 9902/11074 [1:24:27<09:38,  2.02it/s]
 89%|████████▉ | 9903/11074 [1:24:27<09:38,  2.02it/s]
 89%|████████▉ | 9904/11074 [1:24:28<09:37,  2.03it/s]
 89%|████████▉ | 9905/11074 [1:24:28<09:37,  2.03it/s]
 89%|████████▉ | 9906/11074 [1:24:29<09:36,  2.03it/s]
 89%|████████▉ | 9907/11074 [1:24:29<09:36,  2.03it/s]
 89%|████████▉ | 9908/11074 [1:24:30<09:35,  2.03it/s]
 89%|████████▉ | 9909/11074 [1:24:30<09:35,  2.02it/s]
 89%|████████▉ | 9910/11074 [1:24:31<09:34,  2.03it/s]
 89%|████████▉ | 9911/11074 [1:24:31<09:34,  2.02it/s]
 90%|████████▉ | 9912/11074 [1:24:32<09:33,  2.03it/s]
 90%|████████▉ | 9913/11074 [1:24:32<09:33,  2.03it/s]
 90%|████████▉ | 9914/11074 [1:24:33<09:32,  2.03it/s]
 90%|████████▉ | 9915/11074 [1:24:33<09:31,  2.03it/s]
 90%|████████▉ | 9916/11074 [1:24:34<09:31,  2.02it/s]
 90%|████████▉ | 9917/11074 [1:24:34<09:30,  2.03it/s]
 90%|████████▉ | 9918/11074 [1:24:35<09:30,  2.03it/s]
 90%|████████▉ | 9919/11074 [1:24:35<09:29,  2.03it/s]
 90%|████████▉ | 9920/11074 [1:24:36<09:29,  2.03it/s]
 90%|████████▉ | 9921/11074 [1:24:36<09:29,  2.03it/s]
 90%|████████▉ | 9922/11074 [1:24:37<09:28,  2.03it/s]
 90%|████████▉ | 9923/11074 [1:24:37<09:28,  2.02it/s]
 90%|████████▉ | 9924/11074 [1:24:38<09:27,  2.03it/s]
 90%|████████▉ | 9925/11074 [1:24:38<09:27,  2.03it/s]
                                                      
{'loss': 3.1475, 'grad_norm': 0.23960614204406738, 'learning_rate': 3.244030654038022e-05, 'epoch': 12.54}
+
 90%|████████▉ | 9925/11074 [1:24:38<09:27,  2.03it/s]
 90%|████████▉ | 9926/11074 [1:24:39<09:29,  2.02it/s]
 90%|████████▉ | 9927/11074 [1:24:39<09:28,  2.02it/s]
 90%|████████▉ | 9928/11074 [1:24:40<09:27,  2.02it/s]
 90%|████████▉ | 9929/11074 [1:24:40<09:26,  2.02it/s]
 90%|████████▉ | 9930/11074 [1:24:41<09:25,  2.02it/s]
 90%|████████▉ | 9931/11074 [1:24:41<09:25,  2.02it/s]
 90%|████████▉ | 9932/11074 [1:24:42<09:23,  2.03it/s]
 90%|████████▉ | 9933/11074 [1:24:42<09:24,  2.02it/s]
 90%|████████▉ | 9934/11074 [1:24:43<09:23,  2.02it/s]
 90%|████████▉ | 9935/11074 [1:24:43<09:22,  2.02it/s]
 90%|████████▉ | 9936/11074 [1:24:44<09:22,  2.02it/s]
 90%|████████▉ | 9937/11074 [1:24:44<09:21,  2.02it/s]
 90%|████████▉ | 9938/11074 [1:24:45<09:21,  2.02it/s]
 90%|████████▉ | 9939/11074 [1:24:45<09:20,  2.02it/s]
 90%|████████▉ | 9940/11074 [1:24:46<09:20,  2.02it/s]
 90%|████████▉ | 9941/11074 [1:24:46<09:20,  2.02it/s]
 90%|████████▉ | 9942/11074 [1:24:47<09:19,  2.02it/s]
 90%|████████▉ | 9943/11074 [1:24:47<09:19,  2.02it/s]
 90%|████████▉ | 9944/11074 [1:24:48<09:18,  2.02it/s]
 90%|████████▉ | 9945/11074 [1:24:48<09:18,  2.02it/s]
 90%|████████▉ | 9946/11074 [1:24:49<09:17,  2.02it/s]
 90%|████████▉ | 9947/11074 [1:24:49<09:17,  2.02it/s]
 90%|████████▉ | 9948/11074 [1:24:50<09:16,  2.02it/s]
 90%|████████▉ | 9949/11074 [1:24:50<09:16,  2.02it/s]
 90%|████████▉ | 9950/11074 [1:24:51<09:15,  2.02it/s]{'loss': 3.1544, 'grad_norm': 0.23955270648002625, 'learning_rate': 3.1058631440753546e-05, 'epoch': 12.57}
+
                                                      

 90%|███████��▉ | 9950/11074 [1:24:51<09:15,  2.02it/s]
 90%|████████▉ | 9951/11074 [1:24:51<09:15,  2.02it/s]
 90%|████████▉ | 9952/11074 [1:24:52<09:14,  2.02it/s]
 90%|████████▉ | 9953/11074 [1:24:52<09:13,  2.02it/s]
 90%|████████▉ | 9954/11074 [1:24:53<09:13,  2.02it/s]
 90%|████████▉ | 9955/11074 [1:24:53<09:13,  2.02it/s]
 90%|████████▉ | 9956/11074 [1:24:54<09:12,  2.02it/s]
 90%|████████▉ | 9957/11074 [1:24:54<09:11,  2.02it/s]
 90%|████████▉ | 9958/11074 [1:24:55<09:11,  2.02it/s]
 90%|████████▉ | 9959/11074 [1:24:55<09:10,  2.03it/s]
 90%|████████▉ | 9960/11074 [1:24:56<09:10,  2.02it/s]
 90%|████████▉ | 9961/11074 [1:24:56<09:09,  2.03it/s]
 90%|████████▉ | 9962/11074 [1:24:57<09:09,  2.02it/s]
 90%|████████▉ | 9963/11074 [1:24:57<09:08,  2.03it/s]
 90%|████████▉ | 9964/11074 [1:24:58<09:08,  2.02it/s]
 90%|████████▉ | 9965/11074 [1:24:58<09:07,  2.02it/s]
 90%|████████▉ | 9966/11074 [1:24:59<09:06,  2.03it/s]
 90%|█████████ | 9967/11074 [1:24:59<09:06,  2.03it/s]
 90%|█████████ | 9968/11074 [1:25:00<09:05,  2.03it/s]
 90%|█████████ | 9969/11074 [1:25:00<09:06,  2.02it/s]
 90%|█████████ | 9970/11074 [1:25:01<09:05,  2.03it/s]
 90%|█████████ | 9971/11074 [1:25:01<09:05,  2.02it/s]
 90%|█████████ | 9972/11074 [1:25:02<09:04,  2.02it/s]
 90%|█████████ | 9973/11074 [1:25:02<09:03,  2.02it/s]
 90%|█████████ | 9974/11074 [1:25:03<09:03,  2.02it/s]
 90%|█████████ | 9975/11074 [1:25:03<09:03,  2.02it/s]
                                                      {'loss': 3.1471, 'grad_norm': 0.23880618810653687, 'learning_rate': 2.9706080560651215e-05, 'epoch': 12.6}
+

 90%|█████████ | 9975/11074 [1:25:03<09:03,  2.02it/s]
 90%|█████████ | 9976/11074 [1:25:03<09:03,  2.02it/s]
 90%|█████████ | 9977/11074 [1:25:04<09:01,  2.02it/s]
 90%|█████████ | 9978/11074 [1:25:04<09:02,  2.02it/s]
 90%|█████████ | 9979/11074 [1:25:05<09:00,  2.02it/s]
 90%|█████████ | 9980/11074 [1:25:05<09:00,  2.02it/s]
 90%|█████████ | 9981/11074 [1:25:06<09:00,  2.02it/s]
 90%|█████████ | 9982/11074 [1:25:06<08:59,  2.02it/s]
 90%|█████████ | 9983/11074 [1:25:07<08:58,  2.02it/s]
 90%|█████████ | 9984/11074 [1:25:07<08:57,  2.03it/s]
 90%|█████████ | 9985/11074 [1:25:08<08:57,  2.03it/s]
 90%|█████████ | 9986/11074 [1:25:08<08:57,  2.02it/s]
 90%|█████████ | 9987/11074 [1:25:09<08:57,  2.02it/s]
 90%|█████████ | 9988/11074 [1:25:09<08:56,  2.02it/s]
 90%|█████████ | 9989/11074 [1:25:10<08:56,  2.02it/s]
 90%|█████████ | 9990/11074 [1:25:10<08:55,  2.02it/s]
 90%|█████████ | 9991/11074 [1:25:11<08:55,  2.02it/s]
 90%|█████████ | 9992/11074 [1:25:11<08:55,  2.02it/s]
 90%|█████████ | 9993/11074 [1:25:12<08:55,  2.02it/s]
 90%|█████████ | 9994/11074 [1:25:12<08:55,  2.02it/s]
 90%|█████████ | 9995/11074 [1:25:13<08:54,  2.02it/s]
 90%|█████████ | 9996/11074 [1:25:13<08:53,  2.02it/s]
 90%|█████████ | 9997/11074 [1:25:14<08:53,  2.02it/s]
 90%|█████████ | 9998/11074 [1:25:14<08:52,  2.02it/s]
 90%|█████████ | 9999/11074 [1:25:15<08:51,  2.02it/s]
 90%|█████████ | 10000/11074 [1:25:15<08:51,  2.02it/s]
                                                       {'loss': 3.1567, 'grad_norm': 0.24364879727363586, 'learning_rate': 2.8382737902021838e-05, 'epoch': 12.63}
+
 90%|█████████ | 10000/11074 [1:25:15<08:51,  2.02it/s]
 90%|█████████ | 10001/11074 [1:25:16<08:51,  2.02it/s]
 90%|█████████ | 10002/11074 [1:25:16<08:50,  2.02it/s]
 90%|█████████ | 10003/11074 [1:25:17<08:49,  2.02it/s]
 90%|█████████ | 10004/11074 [1:25:17<08:49,  2.02it/s]
 90%|█████████ | 10005/11074 [1:25:18<08:48,  2.02it/s]
 90%|█████████ | 10006/11074 [1:25:18<08:48,  2.02it/s]
 90%|█████████ | 10007/11074 [1:25:19<08:47,  2.02it/s]
 90%|█████████ | 10008/11074 [1:25:19<08:47,  2.02it/s]
 90%|█████████ | 10009/11074 [1:25:20<08:46,  2.02it/s]
 90%|█████████ | 10010/11074 [1:25:20<08:45,  2.02it/s]
 90%|█████████ | 10011/11074 [1:25:21<08:45,  2.02it/s]
 90%|█████████ | 10012/11074 [1:25:21<08:45,  2.02it/s]
 90%|█████████ | 10013/11074 [1:25:22<08:44,  2.02it/s]
 90%|█████████ | 10014/11074 [1:25:22<08:43,  2.02it/s]
 90%|█████████ | 10015/11074 [1:25:23<08:43,  2.02it/s]
 90%|█████████ | 10016/11074 [1:25:23<08:42,  2.03it/s]
 90%|█████████ | 10017/11074 [1:25:24<08:42,  2.02it/s]
 90%|█████████ | 10018/11074 [1:25:24<08:41,  2.02it/s]
 90%|█████████ | 10019/11074 [1:25:25<08:40,  2.03it/s]
 90%|█████████ | 10020/11074 [1:25:25<08:40,  2.02it/s]
 90%|█████████ | 10021/11074 [1:25:26<08:39,  2.03it/s]
 91%|█████████ | 10022/11074 [1:25:26<08:39,  2.02it/s]
 91%|█████████ | 10023/11074 [1:25:27<08:39,  2.02it/s]
 91%|█████████ | 10024/11074 [1:25:27<08:39,  2.02it/s]
 91%|█████████ | 10025/11074 [1:25:28<08:38,  2.02it/s]
                                                       {'loss': 3.1458, 'grad_norm': 0.23872315883636475, 'learning_rate': 2.7088685652798927e-05, 'epoch': 12.67}
+
 91%|█████████ | 10025/11074 [1:25:28<08:38,  2.02it/s]
 91%|█████████ | 10026/11074 [1:25:28<08:39,  2.02it/s]
 91%|█████████ | 10027/11074 [1:25:29<08:38,  2.02it/s]
 91%|█████████ | 10028/11074 [1:25:29<08:37,  2.02it/s]
 91%|█████████ | 10029/11074 [1:25:30<08:37,  2.02it/s]
 91%|█████████ | 10030/11074 [1:25:30<08:36,  2.02it/s]
 91%|█████████ | 10031/11074 [1:25:31<08:35,  2.02it/s]
 91%|█████████ | 10032/11074 [1:25:31<08:35,  2.02it/s]
 91%|█████████ | 10033/11074 [1:25:32<08:34,  2.02it/s]
 91%|█████████ | 10034/11074 [1:25:32<08:34,  2.02it/s]
 91%|█████████ | 10035/11074 [1:25:33<08:33,  2.02it/s]
 91%|█████████ | 10036/11074 [1:25:33<08:33,  2.02it/s]
 91%|█████████ | 10037/11074 [1:25:34<08:32,  2.02it/s]
 91%|█████████ | 10038/11074 [1:25:34<08:32,  2.02it/s]
 91%|█████████ | 10039/11074 [1:25:35<08:32,  2.02it/s]
 91%|█████████ | 10040/11074 [1:25:35<08:31,  2.02it/s]
 91%|█████████ | 10041/11074 [1:25:36<08:31,  2.02it/s]
 91%|█████████ | 10042/11074 [1:25:36<08:30,  2.02it/s]
 91%|█████████ | 10043/11074 [1:25:37<08:30,  2.02it/s]
 91%|█████████ | 10044/11074 [1:25:37<08:29,  2.02it/s]
 91%|█████████ | 10045/11074 [1:25:38<08:29,  2.02it/s]
 91%|█████████ | 10046/11074 [1:25:38<08:28,  2.02it/s]
 91%|█████████ | 10047/11074 [1:25:39<08:28,  2.02it/s]
 91%|█████████ | 10048/11074 [1:25:39<08:26,  2.02it/s]
 91%|█████████ | 10049/11074 [1:25:40<08:26,  2.02it/s]
 91%|█████████ | 10050/11074 [1:25:40<08:25,  2.02it/s]
                                                       {'loss': 3.1483, 'grad_norm': 0.24116002023220062, 'learning_rate': 2.5824004181797035e-05, 'epoch': 12.7}
+
 91%|█████████ | 10050/11074 [1:25:40<08:25,  2.02it/s]
 91%|█████████ | 10051/11074 [1:25:41<08:26,  2.02it/s]
 91%|█████████ | 10052/11074 [1:25:41<08:25,  2.02it/s]
 91%|█████████ | 10053/11074 [1:25:42<08:25,  2.02it/s]
 91%|█████████ | 10054/11074 [1:25:42<08:24,  2.02it/s]
 91%|█████████ | 10055/11074 [1:25:43<08:24,  2.02it/s]
 91%|█████████ | 10056/11074 [1:25:43<08:23,  2.02it/s]
 91%|█████████ | 10057/11074 [1:25:44<08:22,  2.02it/s]
 91%|█████████ | 10058/11074 [1:25:44<08:22,  2.02it/s]
 91%|█████████ | 10059/11074 [1:25:45<08:21,  2.02it/s]
 91%|█████████ | 10060/11074 [1:25:45<08:21,  2.02it/s]
 91%|█████████ | 10061/11074 [1:25:46<08:20,  2.02it/s]
 91%|█████████ | 10062/11074 [1:25:46<08:20,  2.02it/s]
 91%|█████████ | 10063/11074 [1:25:47<08:19,  2.02it/s]
 91%|█████████ | 10064/11074 [1:25:47<08:19,  2.02it/s]
 91%|█████████ | 10065/11074 [1:25:48<08:19,  2.02it/s]
 91%|█████████ | 10066/11074 [1:25:48<08:18,  2.02it/s]
 91%|█████████ | 10067/11074 [1:25:48<08:17,  2.02it/s]
 91%|█████████ | 10068/11074 [1:25:49<08:16,  2.03it/s]
 91%|█████████ | 10069/11074 [1:25:49<08:16,  2.02it/s]
 91%|█████████ | 10070/11074 [1:25:50<08:15,  2.03it/s]
 91%|█████████ | 10071/11074 [1:25:50<08:15,  2.02it/s]
 91%|█████████ | 10072/11074 [1:25:51<08:14,  2.03it/s]
 91%|█████████ | 10073/11074 [1:25:51<08:14,  2.02it/s]
 91%|█████████ | 10074/11074 [1:25:52<08:13,  2.02it/s]
 91%|█████████ | 10075/11074 [1:25:52<08:13,  2.02it/s]
                                                       {'loss': 3.1564, 'grad_norm': 0.2407723069190979, 'learning_rate': 2.4588772033719563e-05, 'epoch': 12.73}
+
 91%|█████████ | 10075/11074 [1:25:52<08:13,  2.02it/s]
 91%|█████████ | 10076/11074 [1:25:53<08:13,  2.02it/s]
 91%|█████████ | 10077/11074 [1:25:53<08:13,  2.02it/s]
 91%|█████████ | 10078/11074 [1:25:54<08:12,  2.02it/s]
 91%|█████████ | 10079/11074 [1:25:54<08:12,  2.02it/s]
 91%|█████████ | 10080/11074 [1:25:55<08:11,  2.02it/s]
 91%|█████████ | 10081/11074 [1:25:55<08:10,  2.02it/s]
 91%|█████████ | 10082/11074 [1:25:56<08:10,  2.02it/s]
 91%|█████████ | 10083/11074 [1:25:56<08:09,  2.02it/s]
 91%|█████████ | 10084/11074 [1:25:57<08:08,  2.02it/s]
 91%|█████████ | 10085/11074 [1:25:57<08:08,  2.03it/s]
 91%|█████████ | 10086/11074 [1:25:58<08:08,  2.02it/s]
 91%|█████████ | 10087/11074 [1:25:58<08:06,  2.03it/s]
 91%|█████████ | 10088/11074 [1:25:59<08:06,  2.03it/s]
 91%|█████████ | 10089/11074 [1:25:59<08:05,  2.03it/s]
 91%|█████████ | 10090/11074 [1:26:00<08:05,  2.03it/s]
 91%|█████████ | 10091/11074 [1:26:00<08:05,  2.03it/s]
 91%|█████████ | 10092/11074 [1:26:01<08:04,  2.03it/s]
 91%|█████████ | 10093/11074 [1:26:01<08:04,  2.02it/s]
 91%|█████████ | 10094/11074 [1:26:02<08:03,  2.03it/s]
 91%|█████████ | 10095/11074 [1:26:02<08:03,  2.02it/s]
 91%|█████████ | 10096/11074 [1:26:03<08:02,  2.03it/s]
 91%|█████████ | 10097/11074 [1:26:03<08:02,  2.02it/s]
 91%|█████████ | 10098/11074 [1:26:04<08:02,  2.02it/s]
 91%|█████████ | 10099/11074 [1:26:04<08:01,  2.02it/s]
 91%|█████████ | 10100/11074 [1:26:05<08:00,  2.03it/s]{'loss': 3.1479, 'grad_norm': 0.24200809001922607, 'learning_rate': 2.3383065924281353e-05, 'epoch': 12.76}
                                                       
+
 91%|█████████ | 10100/11074 [1:26:05<08:00,  2.03it/s]
 91%|█████████ | 10101/11074 [1:26:05<08:01,  2.02it/s]
 91%|█████████ | 10102/11074 [1:26:06<08:00,  2.02it/s]
 91%|█████████ | 10103/11074 [1:26:06<08:00,  2.02it/s]
 91%|█████████ | 10104/11074 [1:26:07<07:59,  2.02it/s]
 91%|█████████ | 10105/11074 [1:26:07<07:58,  2.02it/s]
 91%|█████████▏| 10106/11074 [1:26:08<07:58,  2.02it/s]
 91%|█████████▏| 10107/11074 [1:26:08<07:57,  2.02it/s]
 91%|█████████▏| 10108/11074 [1:26:09<07:57,  2.02it/s]
 91%|█████████▏| 10109/11074 [1:26:09<07:56,  2.03it/s]
 91%|█████████▏| 10110/11074 [1:26:10<07:56,  2.02it/s]
 91%|█████████▏| 10111/11074 [1:26:10<07:56,  2.02it/s]
 91%|█████████▏| 10112/11074 [1:26:11<07:55,  2.02it/s]
 91%|█████████▏| 10113/11074 [1:26:11<07:54,  2.02it/s]
 91%|█████████▏| 10114/11074 [1:26:12<07:54,  2.02it/s]
 91%|█████████▏| 10115/11074 [1:26:12<07:53,  2.02it/s]
 91%|█████████▏| 10116/11074 [1:26:13<07:53,  2.02it/s]
 91%|█████████▏| 10117/11074 [1:26:13<07:52,  2.02it/s]
 91%|█████████▏| 10118/11074 [1:26:14<07:51,  2.03it/s]
 91%|█████████▏| 10119/11074 [1:26:14<07:51,  2.02it/s]
 91%|█████████▏| 10120/11074 [1:26:15<07:51,  2.03it/s]
 91%|█████████▏| 10121/11074 [1:26:15<07:50,  2.02it/s]
 91%|█████████▏| 10122/11074 [1:26:16<07:49,  2.03it/s]
 91%|█████████▏| 10123/11074 [1:26:16<07:49,  2.02it/s]
 91%|█████████▏| 10124/11074 [1:26:17<07:48,  2.03it/s]
 91%|█████████▏| 10125/11074 [1:26:17<07:48,  2.02it/s]
                                                       {'loss': 3.1546, 'grad_norm': 0.23952969908714294, 'learning_rate': 2.2206960735443648e-05, 'epoch': 12.79}
+

 91%|█████████▏| 10125/11074 [1:26:17<07:48,  2.02it/s]
 91%|█████████▏| 10126/11074 [1:26:18<07:48,  2.02it/s]
 91%|█████████▏| 10127/11074 [1:26:18<07:48,  2.02it/s]
 91%|█████████▏| 10128/11074 [1:26:19<07:47,  2.02it/s]
 91%|█████████▏| 10129/11074 [1:26:19<07:46,  2.02it/s]
 91%|█████████▏| 10130/11074 [1:26:20<07:46,  2.02it/s]
 91%|█████████▏| 10131/11074 [1:26:20<07:45,  2.02it/s]
 91%|█████████▏| 10132/11074 [1:26:21<07:45,  2.02it/s]
 92%|█████████▏| 10133/11074 [1:26:21<07:45,  2.02it/s]
 92%|█████████▏| 10134/11074 [1:26:22<07:44,  2.02it/s]
 92%|██��██████▏| 10135/11074 [1:26:22<07:44,  2.02it/s]
 92%|█████████▏| 10136/11074 [1:26:23<07:43,  2.02it/s]
 92%|█████████▏| 10137/11074 [1:26:23<07:43,  2.02it/s]
 92%|█████████▏| 10138/11074 [1:26:24<07:42,  2.02it/s]
 92%|█████████▏| 10139/11074 [1:26:24<07:42,  2.02it/s]
 92%|█████████▏| 10140/11074 [1:26:25<07:41,  2.02it/s]
 92%|█████████▏| 10141/11074 [1:26:25<07:41,  2.02it/s]
 92%|█████████▏| 10142/11074 [1:26:26<07:40,  2.02it/s]
 92%|█████████▏| 10143/11074 [1:26:26<07:40,  2.02it/s]
 92%|█████████▏| 10144/11074 [1:26:27<07:39,  2.02it/s]
 92%|█████████▏| 10145/11074 [1:26:27<07:39,  2.02it/s]
 92%|█████████▏| 10146/11074 [1:26:28<07:38,  2.02it/s]
 92%|█████████▏| 10147/11074 [1:26:28<07:38,  2.02it/s]
 92%|█████████▏| 10148/11074 [1:26:29<07:37,  2.02it/s]
 92%|█████████▏| 10149/11074 [1:26:29<07:37,  2.02it/s]
 92%|█████████▏| 10150/11074 [1:26:30<07:36,  2.02it/s]{'loss': 3.1513, 'grad_norm': 0.237594336271286, 'learning_rate': 2.1060529510763648e-05, 'epoch': 12.82}
                                                       
+
 92%|█████████▏| 10150/11074 [1:26:30<07:36,  2.02it/s]
 92%|█████████▏| 10151/11074 [1:26:30<07:36,  2.02it/s]
 92%|█████████▏| 10152/11074 [1:26:30<07:36,  2.02it/s]
 92%|█████████▏| 10153/11074 [1:26:31<07:35,  2.02it/s]
 92%|█████████▏| 10154/11074 [1:26:31<07:34,  2.02it/s]
 92%|█████████▏| 10155/11074 [1:26:32<07:34,  2.02it/s]
 92%|█████████▏| 10156/11074 [1:26:32<07:33,  2.02it/s]
 92%|█████████▏| 10157/11074 [1:26:33<07:33,  2.02it/s]
 92%|█████████▏| 10158/11074 [1:26:33<07:32,  2.02it/s]
 92%|█████████▏| 10159/11074 [1:26:34<07:32,  2.02it/s]
 92%|█████████▏| 10160/11074 [1:26:34<07:31,  2.03it/s]
 92%|█████████▏| 10161/11074 [1:26:35<08:11,  1.86it/s]
 92%|█████████▏| 10162/11074 [1:26:36<07:58,  1.91it/s]
 92%|█████████▏| 10163/11074 [1:26:36<07:49,  1.94it/s]
 92%|█████████▏| 10164/11074 [1:26:37<07:42,  1.97it/s]
 92%|█████████▏| 10165/11074 [1:26:37<07:38,  1.98it/s]
 92%|█████████▏| 10166/11074 [1:26:38<07:35,  2.00it/s]
 92%|█████████▏| 10167/11074 [1:26:38<07:32,  2.00it/s]
 92%|█████████▏| 10168/11074 [1:26:39<07:31,  2.01it/s]
 92%|█████████▏| 10169/11074 [1:26:39<07:29,  2.01it/s]
 92%|█████████▏| 10170/11074 [1:26:40<07:28,  2.02it/s]
 92%|█████████▏| 10171/11074 [1:26:40<07:27,  2.02it/s]
 92%|█████████▏| 10172/11074 [1:26:41<07:26,  2.02it/s]
 92%|█████████▏| 10173/11074 [1:26:41<07:25,  2.02it/s]
 92%|█████████▏| 10174/11074 [1:26:42<07:24,  2.02it/s]
 92%|█████████▏| 10175/11074 [1:26:42<07:24,  2.02it/s]{'loss': 3.1437, 'grad_norm': 0.240619495511055, 'learning_rate': 1.99438434508582e-05, 'epoch': 12.86}
+
                                                       

 92%|█████████▏| 10175/11074 [1:26:42<07:24,  2.02it/s]
 92%|█████████▏| 10176/11074 [1:26:42<07:24,  2.02it/s]
 92%|█████████▏| 10177/11074 [1:26:43<07:23,  2.02it/s]
 92%|█████████▏| 10178/11074 [1:26:43<07:23,  2.02it/s]
 92%|█████████▏| 10179/11074 [1:26:44<07:22,  2.02it/s]
 92%|█████████▏| 10180/11074 [1:26:44<07:22,  2.02it/s]
 92%|█████████▏| 10181/11074 [1:26:45<07:21,  2.02it/s]
 92%|█████████▏| 10182/11074 [1:26:45<07:21,  2.02it/s]
 92%|█████████▏| 10183/11074 [1:26:46<07:20,  2.02it/s]
 92%|█████████▏| 10184/11074 [1:26:46<07:20,  2.02it/s]
 92%|█████████▏| 10185/11074 [1:26:47<07:19,  2.02it/s]
 92%|█████████▏| 10186/11074 [1:26:47<07:18,  2.02it/s]
 92%|█████████▏| 10187/11074 [1:26:48<07:18,  2.02it/s]
 92%|█████████▏| 10188/11074 [1:26:48<07:18,  2.02it/s]
 92%|█████████▏| 10189/11074 [1:26:49<07:17,  2.02it/s]
 92%|█████████▏| 10190/11074 [1:26:49<07:16,  2.02it/s]
 92%|█████████▏| 10191/11074 [1:26:50<07:16,  2.02it/s]
 92%|█████████▏| 10192/11074 [1:26:50<07:16,  2.02it/s]
 92%|█████████▏| 10193/11074 [1:26:51<07:15,  2.02it/s]
 92%|█████████▏| 10194/11074 [1:26:51<07:15,  2.02it/s]
 92%|█████████▏| 10195/11074 [1:26:52<07:15,  2.02it/s]
 92%|████��████▏| 10196/11074 [1:26:52<07:14,  2.02it/s]
 92%|█████████▏| 10197/11074 [1:26:53<07:13,  2.02it/s]
 92%|█████████▏| 10198/11074 [1:26:53<07:12,  2.02it/s]
 92%|█████████▏| 10199/11074 [1:26:54<07:12,  2.02it/s]
 92%|█████████▏| 10200/11074 [1:26:54<07:12,  2.02it/s]{'loss': 3.1512, 'grad_norm': 0.23940086364746094, 'learning_rate': 1.88569719089815e-05, 'epoch': 12.89}
+
                                                       

 92%|█████████▏| 10200/11074 [1:26:54<07:12,  2.02it/s]
 92%|█████████▏| 10201/11074 [1:26:55<07:12,  2.02it/s]
 92%|█████████▏| 10202/11074 [1:26:55<07:11,  2.02it/s]
 92%|█████████▏| 10203/11074 [1:26:56<07:11,  2.02it/s]
 92%|█████████▏| 10204/11074 [1:26:56<07:10,  2.02it/s]
 92%|█████████▏| 10205/11074 [1:26:57<07:09,  2.02it/s]
 92%|█████████▏| 10206/11074 [1:26:57<07:08,  2.02it/s]
 92%|█████████▏| 10207/11074 [1:26:58<07:08,  2.02it/s]
 92%|█████████▏| 10208/11074 [1:26:58<07:07,  2.02it/s]
 92%|█████████▏| 10209/11074 [1:26:59<07:07,  2.02it/s]
 92%|█████████▏| 10210/11074 [1:26:59<07:07,  2.02it/s]
 92%|█████████▏| 10211/11074 [1:27:00<07:07,  2.02it/s]
 92%|█████████▏| 10212/11074 [1:27:00<07:06,  2.02it/s]
 92%|█████████▏| 10213/11074 [1:27:01<07:06,  2.02it/s]
 92%|█████████▏| 10214/11074 [1:27:01<07:05,  2.02it/s]
 92%|█████████▏| 10215/11074 [1:27:02<07:05,  2.02it/s]
 92%|█████████▏| 10216/11074 [1:27:02<07:04,  2.02it/s]
 92%|█████████▏| 10217/11074 [1:27:03<07:03,  2.02it/s]
 92%|█████████▏| 10218/11074 [1:27:03<07:02,  2.02it/s]
 92%|█████████▏| 10219/11074 [1:27:04<07:02,  2.02it/s]
 92%|█████████▏| 10220/11074 [1:27:04<07:01,  2.02it/s]
 92%|█████████▏| 10221/11074 [1:27:05<07:00,  2.03it/s]
 92%|█████████▏| 10222/11074 [1:27:05<07:00,  2.03it/s]
 92%|█████████▏| 10223/11074 [1:27:06<06:59,  2.03it/s]
 92%|█████████▏| 10224/11074 [1:27:06<06:59,  2.03it/s]
 92%|█████████▏| 10225/11074 [1:27:07<06:58,  2.03it/s]
                                                       
{'loss': 3.1548, 'grad_norm': 0.24040085077285767, 'learning_rate': 1.7799982386717872e-05, 'epoch': 12.92}
+
 92%|█████████▏| 10225/11074 [1:27:07<06:58,  2.03it/s]
 92%|█████████▏| 10226/11074 [1:27:07<06:59,  2.02it/s]
 92%|█████████▏| 10227/11074 [1:27:08<06:58,  2.02it/s]
 92%|█████████▏| 10228/11074 [1:27:08<06:58,  2.02it/s]
 92%|█████████▏| 10229/11074 [1:27:09<06:57,  2.02it/s]
 92%|█████████▏| 10230/11074 [1:27:09<06:57,  2.02it/s]
 92%|█████████▏| 10231/11074 [1:27:10<06:56,  2.02it/s]
 92%|█████████▏| 10232/11074 [1:27:10<06:55,  2.02it/s]
 92%|█████████▏| 10233/11074 [1:27:11<06:55,  2.03it/s]
 92%|█████████▏| 10234/11074 [1:27:11<06:54,  2.02it/s]
 92%|█████████▏| 10235/11074 [1:27:12<06:54,  2.03it/s]
 92%|█████████▏| 10236/11074 [1:27:12<06:53,  2.03it/s]
 92%|█████████▏| 10237/11074 [1:27:13<06:53,  2.03it/s]
 92%|█████████▏| 10238/11074 [1:27:13<06:53,  2.02it/s]
 92%|█████████▏| 10239/11074 [1:27:14<06:52,  2.02it/s]
 92%|█████████▏| 10240/11074 [1:27:14<06:51,  2.02it/s]
 92%|█████████▏| 10241/11074 [1:27:15<06:51,  2.03it/s]
 92%|█████████▏| 10242/11074 [1:27:15<06:51,  2.02it/s]
 92%|█████████▏| 10243/11074 [1:27:16<06:50,  2.02it/s]
 93%|█████████▎| 10244/11074 [1:27:16<06:50,  2.02it/s]
 93%|█████████▎| 10245/11074 [1:27:17<06:49,  2.02it/s]
 93%|█████████▎| 10246/11074 [1:27:17<06:49,  2.02it/s]
 93%|█████████▎| 10247/11074 [1:27:18<06:48,  2.03it/s]
 93%|█████████▎| 10248/11074 [1:27:18<06:47,  2.03it/s]
 93%|█████████▎| 10249/11074 [1:27:19<06:47,  2.03it/s]
 93%|█████████▎| 10250/11074 [1:27:19<06:46,  2.03it/s]{'loss': 3.1554, 'grad_norm': 0.24188894033432007, 'learning_rate': 1.6772940529789784e-05, 'epoch': 12.95}
+
                                                       

 93%|█████████▎| 10250/11074 [1:27:19<06:46,  2.03it/s]
 93%|█████████▎| 10251/11074 [1:27:20<06:46,  2.02it/s]
 93%|█████████▎| 10252/11074 [1:27:20<06:46,  2.02it/s]
 93%|█████████▎| 10253/11074 [1:27:21<07:21,  1.86it/s]
 93%|��████████▎| 10254/11074 [1:27:21<07:10,  1.91it/s]
 93%|█████████▎| 10255/11074 [1:27:22<07:02,  1.94it/s]
 93%|█████████▎| 10256/11074 [1:27:22<06:56,  1.96it/s]
 93%|█████████▎| 10257/11074 [1:27:23<06:52,  1.98it/s]
 93%|█████████▎| 10258/11074 [1:27:23<06:48,  2.00it/s]
 93%|█████████▎| 10259/11074 [1:27:24<06:46,  2.00it/s]
 93%|█████████▎| 10260/11074 [1:27:24<06:44,  2.01it/s]
 93%|█████████▎| 10261/11074 [1:27:25<06:43,  2.01it/s]
 93%|█████████▎| 10262/11074 [1:27:25<06:42,  2.02it/s]
 93%|█████████▎| 10263/11074 [1:27:26<06:41,  2.02it/s]
 93%|█████████▎| 10264/11074 [1:27:26<06:40,  2.02it/s]
 93%|█████████▎| 10265/11074 [1:27:27<06:40,  2.02it/s]
 93%|█████████▎| 10266/11074 [1:27:27<06:39,  2.02it/s]
 93%|█████████▎| 10267/11074 [1:27:28<06:38,  2.03it/s]
 93%|█████████▎| 10268/11074 [1:27:28<06:38,  2.02it/s]
 93%|█████████▎| 10269/11074 [1:27:29<06:37,  2.03it/s]
 93%|█████████▎| 10270/11074 [1:27:29<06:36,  2.03it/s]
 93%|█████████▎| 10271/11074 [1:27:30<06:36,  2.03it/s]
 93%|█████████▎| 10272/11074 [1:27:30<06:35,  2.03it/s]
 93%|█████████▎| 10273/11074 [1:27:31<06:35,  2.03it/s]
 93%|█████████▎| 10274/11074 [1:27:31<06:34,  2.03it/s]
 93%|█████████▎| 10275/11074 [1:27:32<06:34,  2.03it/s]
                                                       {'loss': 3.1493, 'grad_norm': 0.23945145308971405, 'learning_rate': 1.5775910123980497e-05, 'epoch': 12.98}
+

 93%|█████████▎| 10275/11074 [1:27:32<06:34,  2.03it/s]
 93%|█████████▎| 10276/11074 [1:27:32<06:34,  2.02it/s]
 93%|█████████▎| 10277/11074 [1:27:33<06:34,  2.02it/s]
 93%|█████████▎| 10278/11074 [1:27:33<06:33,  2.02it/s]
 93%|█████████▎| 10279/11074 [1:27:34<06:32,  2.02it/s]
 93%|█████████▎| 10280/11074 [1:27:34<06:32,  2.03it/s]
 93%|█████████▎| 10281/11074 [1:27:35<06:31,  2.02it/s]
 93%|█████████▎| 10282/11074 [1:27:35<06:31,  2.02it/s]
 93%|█████████▎| 10283/11074 [1:27:36<06:30,  2.03it/s]
 93%|█████████▎| 10284/11074 [1:27:36<06:30,  2.02it/s]
 93%|█████████▎| 10285/11074 [1:27:37<06:29,  2.02it/s]
 93%|█████████▎| 10286/11074 [1:27:37<06:29,  2.02it/s]
 93%|█████████▎| 10287/11074 [1:27:37<06:28,  2.02it/s]
 93%|█████████▎| 10288/11074 [1:27:38<06:28,  2.03it/s]
 93%|█████████▎| 10289/11074 [1:27:39<06:37,  1.98it/s]
 93%|█████████▎| 10290/11074 [1:27:52<58:22,  4.47s/it]
 93%|█████████▎| 10291/11074 [1:27:53<42:44,  3.28s/it]
 93%|█████████▎| 10292/11074 [1:27:53<31:50,  2.44s/it]
 93%|█████████▎| 10293/11074 [1:27:54<24:11,  1.86s/it]
 93%|█████████▎| 10294/11074 [1:27:54<18:50,  1.45s/it]
 93%|█████████▎| 10295/11074 [1:27:55<15:08,  1.17s/it]
 93%|█████████▎| 10296/11074 [1:27:55<12:29,  1.04it/s]
 93%|█████████▎| 10297/11074 [1:27:56<10:39,  1.22it/s]
 93%|█████████▎| 10298/11074 [1:27:56<09:21,  1.38it/s]
 93%|█████████▎| 10299/11074 [1:27:57<08:27,  1.53it/s]
 93%|█████████▎| 10300/11074 [1:27:57<07:49,  1.65it/s]
                                                       
{'loss': 3.1388, 'grad_norm': 0.24089667201042175, 'learning_rate': 1.4808953091172639e-05, 'epoch': 13.01}
+
 93%|█████████▎| 10300/11074 [1:27:57<07:49,  1.65it/s]
 93%|█████████▎| 10301/11074 [1:27:58<07:23,  1.74it/s]
 93%|█████████▎| 10302/11074 [1:27:58<07:05,  1.82it/s]
 93%|█████████▎| 10303/11074 [1:27:59<06:51,  1.87it/s]
 93%|█████████▎| 10304/11074 [1:27:59<06:41,  1.92it/s]
 93%|█████████▎| 10305/11074 [1:28:00<06:35,  1.95it/s]
 93%|█████████▎| 10306/11074 [1:28:00<06:30,  1.97it/s]
 93%|█████████▎| 10307/11074 [1:28:01<06:26,  1.98it/s]
 93%|█████████▎| 10308/11074 [1:28:01<06:24,  1.99it/s]
 93%|█████████▎| 10309/11074 [1:28:02<06:22,  2.00it/s]
 93%|█████████▎| 10310/11074 [1:28:02<06:20,  2.01it/s]
 93%|█████████▎| 10311/11074 [1:28:03<06:19,  2.01it/s]
 93%|█████████▎| 10312/11074 [1:28:03<06:18,  2.01it/s]
 93%|█████████▎| 10313/11074 [1:28:04<06:17,  2.02it/s]
 93%|█████████▎| 10314/11074 [1:28:04<06:16,  2.02it/s]
 93%|��████████▎| 10315/11074 [1:28:05<06:15,  2.02it/s]
 93%|█████████▎| 10316/11074 [1:28:05<06:15,  2.02it/s]
 93%|█████████▎| 10317/11074 [1:28:06<06:15,  2.01it/s]
 93%|█████████▎| 10318/11074 [1:28:06<06:14,  2.02it/s]
 93%|█████████▎| 10319/11074 [1:28:07<06:13,  2.02it/s]
 93%|█████████▎| 10320/11074 [1:28:07<06:12,  2.02it/s]
 93%|█████████▎| 10321/11074 [1:28:08<06:12,  2.02it/s]
 93%|█████████▎| 10322/11074 [1:28:08<06:12,  2.02it/s]
 93%|█████████▎| 10323/11074 [1:28:09<06:11,  2.02it/s]
 93%|█████████▎| 10324/11074 [1:28:09<06:10,  2.02it/s]
 93%|█████████▎| 10325/11074 [1:28:10<06:10,  2.02it/s]{'loss': 3.1259, 'grad_norm': 0.24167923629283905, 'learning_rate': 1.387212948550265e-05, 'epoch': 13.04}
                                                       
+
 93%|█████████▎| 10325/11074 [1:28:10<06:10,  2.02it/s]
 93%|█████████▎| 10326/11074 [1:28:10<06:09,  2.02it/s]
 93%|█████████▎| 10327/11074 [1:28:11<06:09,  2.02it/s]
 93%|█████████▎| 10328/11074 [1:28:11<06:08,  2.02it/s]
 93%|█████████▎| 10329/11074 [1:28:12<06:08,  2.02it/s]
 93%|█████████▎| 10330/11074 [1:28:12<06:07,  2.02it/s]
 93%|█████████▎| 10331/11074 [1:28:13<06:06,  2.03it/s]
 93%|█████████▎| 10332/11074 [1:28:13<06:06,  2.03it/s]
 93%|█████████▎| 10333/11074 [1:28:14<06:05,  2.03it/s]
 93%|█████████▎| 10334/11074 [1:28:14<06:05,  2.02it/s]
 93%|█████████▎| 10335/11074 [1:28:14<06:04,  2.03it/s]
 93%|█████████▎| 10336/11074 [1:28:15<06:04,  2.03it/s]
 93%|█████████▎| 10337/11074 [1:28:15<06:03,  2.03it/s]
 93%|█████████▎| 10338/11074 [1:28:16<06:03,  2.02it/s]
 93%|█████████▎| 10339/11074 [1:28:16<06:03,  2.02it/s]
 93%|█████████▎| 10340/11074 [1:28:17<06:02,  2.03it/s]
 93%|█████████▎| 10341/11074 [1:28:17<06:02,  2.02it/s]
 93%|█████████▎| 10342/11074 [1:28:18<06:01,  2.03it/s]
 93%|█████████▎| 10343/11074 [1:28:18<06:01,  2.02it/s]
 93%|█████████▎| 10344/11074 [1:28:19<06:00,  2.03it/s]
 93%|█████████▎| 10345/11074 [1:28:19<06:00,  2.02it/s]
 93%|█████████▎| 10346/11074 [1:28:20<05:59,  2.03it/s]
 93%|█████████▎| 10347/11074 [1:28:20<05:59,  2.02it/s]
 93%|█████████▎| 10348/11074 [1:28:21<05:58,  2.02it/s]
 93%|█████████▎| 10349/11074 [1:28:21<05:58,  2.02it/s]
 93%|█████████▎| 10350/11074 [1:28:22<05:57,  2.02it/s]
                                                       {'loss': 3.1319, 'grad_norm': 0.238393634557724, 'learning_rate': 1.2965497489630717e-05, 'epoch': 13.08}
+

 93%|█████████▎| 10350/11074 [1:28:22<05:57,  2.02it/s]
 93%|█████████▎| 10351/11074 [1:28:22<05:57,  2.02it/s]
 93%|█████████▎| 10352/11074 [1:28:23<05:56,  2.02it/s]
 93%|█████████▎| 10353/11074 [1:28:23<05:56,  2.03it/s]
 93%|█████████▎| 10354/11074 [1:28:24<05:55,  2.02it/s]
 94%|█████████▎| 10355/11074 [1:28:24<05:54,  2.03it/s]
 94%|█████████▎| 10356/11074 [1:28:25<05:54,  2.03it/s]
 94%|█████████▎| 10357/11074 [1:28:25<05:53,  2.03it/s]
 94%|█████████▎| 10358/11074 [1:28:26<05:53,  2.03it/s]
 94%|█████████▎| 10359/11074 [1:28:26<05:53,  2.02it/s]
 94%|█████████▎| 10360/11074 [1:28:27<05:52,  2.02it/s]
 94%|█████████▎| 10361/11074 [1:28:27<05:52,  2.02it/s]
 94%|█████████▎| 10362/11074 [1:28:28<05:51,  2.03it/s]
 94%|█████████▎| 10363/11074 [1:28:28<05:51,  2.02it/s]
 94%|█████████▎| 10364/11074 [1:28:29<05:50,  2.03it/s]
 94%|█████████▎| 10365/11074 [1:28:29<05:50,  2.02it/s]
 94%|█████████▎| 10366/11074 [1:28:30<05:50,  2.02it/s]
 94%|█████████▎| 10367/11074 [1:28:30<05:50,  2.02it/s]
 94%|█████████▎| 10368/11074 [1:28:31<05:49,  2.02it/s]
 94%|█████████▎| 10369/11074 [1:28:31<05:48,  2.02it/s]
 94%|█████████▎| 10370/11074 [1:28:32<05:48,  2.02it/s]
 94%|█████████▎| 10371/11074 [1:28:32<05:47,  2.02it/s]
 94%|█████████▎| 10372/11074 [1:28:33<05:46,  2.02it/s]
 94%|█████████▎| 10373/11074 [1:28:33<05:46,  2.02it/s]
 94%|█████████▎| 10374/11074 [1:28:34<05:45,  2.02it/s]
 94%|█████████▎| 10375/11074 [1:28:34<05:45,  2.02it/s]
                                                       
{'loss': 3.1293, 'grad_norm': 0.23897463083267212, 'learning_rate': 1.2089113411127673e-05, 'epoch': 13.11}
+
 94%|█████████▎| 10375/11074 [1:28:34<05:45,  2.02it/s]
 94%|█████████▎| 10376/11074 [1:28:35<05:45,  2.02it/s]
 94%|█████████▎| 10377/11074 [1:28:35<05:45,  2.02it/s]
 94%|█████████▎| 10378/11074 [1:28:36<05:44,  2.02it/s]
 94%|█████████▎| 10379/11074 [1:28:36<05:43,  2.02it/s]
 94%|█████████▎| 10380/11074 [1:28:37<05:42,  2.02it/s]
 94%|█████████▎| 10381/11074 [1:28:37<05:42,  2.02it/s]
 94%|█████████▍| 10382/11074 [1:28:38<05:42,  2.02it/s]
 94%|█████████▍| 10383/11074 [1:28:38<05:41,  2.02it/s]
 94%|█████████▍| 10384/11074 [1:28:39<05:41,  2.02it/s]
 94%|█████████▍| 10385/11074 [1:28:39<05:40,  2.02it/s]
 94%|█████████▍| 10386/11074 [1:28:40<05:39,  2.02it/s]
 94%|█████████▍| 10387/11074 [1:28:40<05:39,  2.02it/s]
 94%|█████████▍| 10388/11074 [1:28:41<05:38,  2.02it/s]
 94%|█████████▍| 10389/11074 [1:28:41<05:38,  2.02it/s]
 94%|█████████▍| 10390/11074 [1:28:42<05:37,  2.03it/s]
 94%|█████████▍| 10391/11074 [1:28:42<05:37,  2.03it/s]
 94%|█████████▍| 10392/11074 [1:28:43<05:36,  2.03it/s]
 94%|█████████▍| 10393/11074 [1:28:43<05:36,  2.02it/s]
 94%|█████████▍| 10394/11074 [1:28:44<05:35,  2.02it/s]
 94%|█████████▍| 10395/11074 [1:28:44<05:35,  2.02it/s]
 94%|█████████▍| 10396/11074 [1:28:45<05:35,  2.02it/s]
 94%|█████████▍| 10397/11074 [1:28:45<05:34,  2.02it/s]
 94%|█████████▍| 10398/11074 [1:28:46<05:34,  2.02it/s]
 94%|█████████▍| 10399/11074 [1:28:46<05:33,  2.02it/s]
 94%|█████████▍| 10400/11074 [1:28:47<05:32,  2.02it/s]{'loss': 3.1223, 'grad_norm': 0.23806162178516388, 'learning_rate': 1.1243031678977611e-05, 'epoch': 13.14}
                                                       
+
 94%|█████████▍| 10400/11074 [1:28:47<05:32,  2.02it/s]
 94%|█████████▍| 10401/11074 [1:28:47<05:33,  2.02it/s]
 94%|█████████▍| 10402/11074 [1:28:48<05:32,  2.02it/s]
 94%|█████████▍| 10403/11074 [1:28:48<05:31,  2.02it/s]
 94%|█████████▍| 10404/11074 [1:28:49<05:31,  2.02it/s]
 94%|█████████▍| 10405/11074 [1:28:49<05:31,  2.02it/s]
 94%|█████████▍| 10406/11074 [1:28:50<05:30,  2.02it/s]
 94%|█████████▍| 10407/11074 [1:28:50<05:29,  2.02it/s]
 94%|█████████▍| 10408/11074 [1:28:51<05:28,  2.02it/s]
 94%|█████████▍| 10409/11074 [1:28:51<05:28,  2.02it/s]
 94%|█████████▍| 10410/11074 [1:28:52<05:27,  2.03it/s]
 94%|█████████▍| 10411/11074 [1:28:52<05:27,  2.03it/s]
 94%|█████████▍| 10412/11074 [1:28:53<05:26,  2.02it/s]
 94%|█████████▍| 10413/11074 [1:28:53<05:26,  2.03it/s]
 94%|█████████▍| 10414/11074 [1:28:54<05:25,  2.03it/s]
 94%|█████████▍| 10415/11074 [1:28:54<05:25,  2.02it/s]
 94%|█████████▍| 10416/11074 [1:28:55<05:25,  2.02it/s]
 94%|█████████▍| 10417/11074 [1:28:55<05:24,  2.03it/s]
 94%|█████████▍| 10418/11074 [1:28:56<05:24,  2.02it/s]
 94%|█████████▍| 10419/11074 [1:28:56<05:23,  2.03it/s]
 94%|█████████▍| 10420/11074 [1:28:57<05:22,  2.03it/s]
 94%|█████████▍| 10421/11074 [1:28:57<05:22,  2.03it/s]
 94%|█████████▍| 10422/11074 [1:28:57<05:21,  2.03it/s]
 94%|█████████▍| 10423/11074 [1:28:58<05:21,  2.03it/s]
 94%|█████████▍| 10424/11074 [1:28:58<05:20,  2.03it/s]
 94%|█████████▍| 10425/11074 [1:28:59<05:19,  2.03it/s]
                                                       {'loss': 3.1214, 'grad_norm': 0.24141210317611694, 'learning_rate': 1.0427304840197493e-05, 'epoch': 13.17}
+

 94%|█████████▍| 10425/11074 [1:28:59<05:19,  2.03it/s]
 94%|█████████▍| 10426/11074 [1:28:59<05:19,  2.03it/s]
 94%|█████████▍| 10427/11074 [1:29:00<05:19,  2.03it/s]
 94%|█████████▍| 10428/11074 [1:29:00<05:19,  2.02it/s]
 94%|█████████▍| 10429/11074 [1:29:01<05:18,  2.03it/s]
 94%|█████████▍| 10430/11074 [1:29:01<05:17,  2.03it/s]
 94%|█████████▍| 10431/11074 [1:29:02<05:17,  2.02it/s]
 94%|█████████▍| 10432/11074 [1:29:02<05:17,  2.02it/s]
 94%|█████████▍| 10433/11074 [1:29:03<05:16,  2.02it/s]
 94%|█████████▍| 10434/11074 [1:29:03<05:16,  2.02it/s]
 94%|█████████▍| 10435/11074 [1:29:04<05:15,  2.03it/s]
 94%|█████████▍| 10436/11074 [1:29:04<05:15,  2.02it/s]
 94%|█████████▍| 10437/11074 [1:29:05<05:14,  2.03it/s]
 94%|█████████▍| 10438/11074 [1:29:05<05:14,  2.03it/s]
 94%|█████████▍| 10439/11074 [1:29:06<05:13,  2.03it/s]
 94%|█████████▍| 10440/11074 [1:29:06<05:13,  2.03it/s]
 94%|█████████▍| 10441/11074 [1:29:07<05:12,  2.03it/s]
 94%|█████████▍| 10442/11074 [1:29:07<05:11,  2.03it/s]
 94%|█████████▍| 10443/11074 [1:29:08<05:11,  2.02it/s]
 94%|█████████▍| 10444/11074 [1:29:08<05:10,  2.03it/s]
 94%|█████████▍| 10445/11074 [1:29:09<05:10,  2.03it/s]
 94%|█████████▍| 10446/11074 [1:29:09<05:09,  2.03it/s]
 94%|█████████▍| 10447/11074 [1:29:10<05:09,  2.03it/s]
 94%|█████████▍| 10448/11074 [1:29:10<05:08,  2.03it/s]
 94%|█████████▍| 10449/11074 [1:29:11<05:08,  2.02it/s]
 94%|█████████▍| 10450/11074 [1:29:11<05:08,  2.02it/s]
                                                       
{'loss': 3.1255, 'grad_norm': 0.2402477115392685, 'learning_rate': 9.64198355657403e-06, 'epoch': 13.2}
+
 94%|█████████▍| 10450/11074 [1:29:11<05:08,  2.02it/s]
 94%|█████████▍| 10451/11074 [1:29:12<05:08,  2.02it/s]
 94%|█████████▍| 10452/11074 [1:29:12<05:07,  2.02it/s]
 94%|█████████▍| 10453/11074 [1:29:13<05:06,  2.02it/s]
 94%|█████████▍| 10454/11074 [1:29:13<05:06,  2.03it/s]
 94%|█████████▍| 10455/11074 [1:29:14<05:05,  2.03it/s]
 94%|█████████▍| 10456/11074 [1:29:14<05:05,  2.02it/s]
 94%|█████████▍| 10457/11074 [1:29:15<05:04,  2.02it/s]
 94%|█████████▍| 10458/11074 [1:29:15<05:04,  2.02it/s]
 94%|█████████▍| 10459/11074 [1:29:16<05:03,  2.02it/s]
 94%|█████████▍| 10460/11074 [1:29:16<05:03,  2.02it/s]
 94%|█████████▍| 10461/11074 [1:29:17<05:02,  2.03it/s]
 94%|█████████▍| 10462/11074 [1:29:17<05:02,  2.02it/s]
 94%|█████████▍| 10463/11074 [1:29:18<05:01,  2.03it/s]
 94%|█████████▍| 10464/11074 [1:29:18<05:01,  2.02it/s]
 95%|█████████▍| 10465/11074 [1:29:19<05:00,  2.03it/s]
 95%|█████████▍| 10466/11074 [1:29:19<05:00,  2.02it/s]
 95%|█████████▍| 10467/11074 [1:29:20<04:59,  2.02it/s]
 95%|█████████▍| 10468/11074 [1:29:20<04:59,  2.02it/s]
 95%|█████████▍| 10469/11074 [1:29:21<04:58,  2.03it/s]
 95%|█████████▍| 10470/11074 [1:29:21<04:57,  2.03it/s]
 95%|█████████▍| 10471/11074 [1:29:22<04:57,  2.03it/s]
 95%|█████████▍| 10472/11074 [1:29:22<04:56,  2.03it/s]
 95%|█████████▍| 10473/11074 [1:29:23<04:56,  2.03it/s]
 95%|█████████▍| 10474/11074 [1:29:23<04:56,  2.03it/s]
 95%|█████████▍| 10475/11074 [1:29:24<04:55,  2.03it/s]
                                                       {'loss': 3.1284, 'grad_norm': 0.24028468132019043, 'learning_rate': 8.887116601516753e-06, 'epoch': 13.23}
+

 95%|█████████▍| 10475/11074 [1:29:24<04:55,  2.03it/s]
 95%|█████████▍| 10476/11074 [1:29:24<04:55,  2.02it/s]
 95%|█████████▍| 10477/11074 [1:29:25<04:55,  2.02it/s]
 95%|█████████▍| 10478/11074 [1:29:25<04:54,  2.02it/s]
 95%|█████████▍| 10479/11074 [1:29:26<04:53,  2.02it/s]
 95%|█████████▍| 10480/11074 [1:29:26<04:53,  2.02it/s]
 95%|█████████▍| 10481/11074 [1:29:27<04:52,  2.03it/s]
 95%|█████████▍| 10482/11074 [1:29:27<04:52,  2.02it/s]
 95%|█████████▍| 10483/11074 [1:29:28<04:51,  2.03it/s]
 95%|█████████▍| 10484/11074 [1:29:28<04:51,  2.02it/s]
 95%|█████████▍| 10485/11074 [1:29:29<04:50,  2.03it/s]
 95%|█████████▍| 10486/11074 [1:29:29<04:50,  2.02it/s]
 95%|█████████▍| 10487/11074 [1:29:30<04:49,  2.03it/s]
 95%|█████████▍| 10488/11074 [1:29:30<04:49,  2.03it/s]
 95%|█████████▍| 10489/11074 [1:29:31<04:48,  2.03it/s]
 95%|█████████▍| 10490/11074 [1:29:31<04:48,  2.03it/s]
 95%|█████████▍| 10491/11074 [1:29:32<04:48,  2.02it/s]
 95%|█████████▍| 10492/11074 [1:29:32<04:47,  2.02it/s]
 95%|█████████▍| 10493/11074 [1:29:33<04:47,  2.02it/s]
 95%|█████████▍| 10494/11074 [1:29:33<04:46,  2.03it/s]
 95%|█████████▍| 10495/11074 [1:29:34<04:46,  2.02it/s]
 95%|█████████▍| 10496/11074 [1:29:34<04:45,  2.02it/s]
 95%|█████████▍| 10497/11074 [1:29:35<04:45,  2.02it/s]
 95%|█████████▍| 10498/11074 [1:29:35<04:44,  2.02it/s]
 95%|█████████▍| 10499/11074 [1:29:36<04:43,  2.02it/s]
 95%|█████████▍| 10500/11074 [1:29:36<04:43,  2.02it/s]
                                                       
{'loss': 3.1273, 'grad_norm': 0.24102579057216644, 'learning_rate': 8.162750857029223e-06, 'epoch': 13.27}
+
 95%|█████████▍| 10500/11074 [1:29:36<04:43,  2.02it/s]
 95%|█████████▍| 10501/11074 [1:29:37<04:43,  2.02it/s]
 95%|█████████▍| 10502/11074 [1:29:37<04:43,  2.02it/s]
 95%|█████████▍| 10503/11074 [1:29:37<04:42,  2.02it/s]
 95%|█████████▍| 10504/11074 [1:29:38<04:42,  2.02it/s]
 95%|█████████▍| 10505/11074 [1:29:38<04:41,  2.02it/s]
 95%|█████████▍| 10506/11074 [1:29:39<04:40,  2.02it/s]
 95%|█████████▍| 10507/11074 [1:29:39<04:39,  2.03it/s]
 95%|█████████▍| 10508/11074 [1:29:40<04:39,  2.02it/s]
 95%|█████████▍| 10509/11074 [1:29:40<04:39,  2.02it/s]
 95%|█████████▍| 10510/11074 [1:29:41<04:38,  2.02it/s]
 95%|█████████▍| 10511/11074 [1:29:41<04:38,  2.02it/s]
 95%|█████████▍| 10512/11074 [1:29:42<04:37,  2.02it/s]
 95%|█████████▍| 10513/11074 [1:29:42<04:37,  2.02it/s]
 95%|█████████▍| 10514/11074 [1:29:43<04:37,  2.02it/s]
 95%|█████████▍| 10515/11074 [1:29:43<04:36,  2.02it/s]
 95%|█████████▍| 10516/11074 [1:29:44<04:36,  2.02it/s]
 95%|█████████▍| 10517/11074 [1:29:44<04:35,  2.02it/s]
 95%|█████████▍| 10518/11074 [1:29:45<04:34,  2.02it/s]
 95%|█████████▍| 10519/11074 [1:29:45<04:34,  2.02it/s]
 95%|█████████▍| 10520/11074 [1:29:46<04:33,  2.02it/s]
 95%|█████████▌| 10521/11074 [1:29:46<04:33,  2.02it/s]
 95%|█████████▌| 10522/11074 [1:29:47<04:32,  2.02it/s]
 95%|█████████▌| 10523/11074 [1:29:47<04:32,  2.02it/s]
 95%|█████████▌| 10524/11074 [1:29:48<04:31,  2.02it/s]
 95%|█████████▌| 10525/11074 [1:29:48<04:31,  2.02it/s]
                                                       {'loss': 3.1204, 'grad_norm': 0.23936308920383453, 'learning_rate': 7.468931310797344e-06, 'epoch': 13.3}
+
 95%|█████████▌| 10525/11074 [1:29:48<04:31,  2.02it/s]
 95%|█████████▌| 10526/11074 [1:29:49<04:31,  2.02it/s]
 95%|█████████▌| 10527/11074 [1:29:49<04:30,  2.02it/s]
 95%|█████████▌| 10528/11074 [1:29:50<04:30,  2.02it/s]
 95%|█████████▌| 10529/11074 [1:29:50<04:29,  2.02it/s]
 95%|█████████▌| 10530/11074 [1:29:51<04:28,  2.02it/s]
 95%|█████████▌| 10531/11074 [1:29:51<04:28,  2.02it/s]
 95%|█████████▌| 10532/11074 [1:29:52<04:28,  2.02it/s]
 95%|█████████▌| 10533/11074 [1:29:52<04:27,  2.02it/s]
 95%|█████████▌| 10534/11074 [1:29:53<04:26,  2.02it/s]
 95%|█████████▌| 10535/11074 [1:29:53<04:26,  2.02it/s]
 95%|█████████▌| 10536/11074 [1:29:54<04:25,  2.02it/s]
 95%|█████████▌| 10537/11074 [1:29:54<04:25,  2.02it/s]
 95%|█████████▌| 10538/11074 [1:29:55<04:24,  2.02it/s]
 95%|█████████▌| 10539/11074 [1:29:55<04:24,  2.03it/s]
 95%|█████████▌| 10540/11074 [1:29:56<04:23,  2.02it/s]
 95%|█████████▌| 10541/11074 [1:29:56<04:23,  2.02it/s]
 95%|█████████▌| 10542/11074 [1:29:57<04:22,  2.02it/s]
 95%|█████████▌| 10543/11074 [1:29:57<04:21,  2.03it/s]
 95%|█████████▌| 10544/11074 [1:29:58<04:21,  2.03it/s]
 95%|█████████▌| 10545/11074 [1:29:58<04:21,  2.02it/s]
 95%|█████████▌| 10546/11074 [1:29:59<04:20,  2.02it/s]
 95%|█████████▌| 10547/11074 [1:29:59<04:20,  2.02it/s]
 95%|█████████▌| 10548/11074 [1:30:00<04:19,  2.03it/s]
 95%|█████████▌| 10549/11074 [1:30:00<04:19,  2.02it/s]
 95%|█████████▌| 10550/11074 [1:30:01<04:18,  2.02it/s]{'loss': 3.1293, 'grad_norm': 0.24025702476501465, 'learning_rate': 6.805701053395119e-06, 'epoch': 13.33}
                                                       
+
 95%|█████████▌| 10550/11074 [1:30:01<04:18,  2.02it/s]
 95%|█████████▌| 10551/11074 [1:30:01<04:18,  2.02it/s]
 95%|█████████▌| 10552/11074 [1:30:02<04:17,  2.02it/s]
 95%|█████████▌| 10553/11074 [1:30:02<04:17,  2.02it/s]
 95%|█████████▌| 10554/11074 [1:30:03<04:16,  2.02it/s]
 95%|█████████▌| 10555/11074 [1:30:03<04:16,  2.02it/s]
 95%|█████████▌| 10556/11074 [1:30:04<04:15,  2.02it/s]
 95%|█████████▌| 10557/11074 [1:30:04<04:15,  2.02it/s]
 95%|█████████▌| 10558/11074 [1:30:05<04:15,  2.02it/s]
 95%|█████████▌| 10559/11074 [1:30:05<04:14,  2.02it/s]
 95%|█████████▌| 10560/11074 [1:30:06<04:14,  2.02it/s]
 95%|█████████▌| 10561/11074 [1:30:06<04:13,  2.02it/s]
 95%|█████████▌| 10562/11074 [1:30:07<04:13,  2.02it/s]
 95%|█████████▌| 10563/11074 [1:30:07<04:12,  2.02it/s]
 95%|█████████▌| 10564/11074 [1:30:08<04:12,  2.02it/s]
 95%|█████████▌| 10565/11074 [1:30:08<04:11,  2.02it/s]
 95%|█████████▌| 10566/11074 [1:30:09<04:11,  2.02it/s]
 95%|█████████▌| 10567/11074 [1:30:09<04:10,  2.02it/s]
 95%|█████████▌| 10568/11074 [1:30:10<04:10,  2.02it/s]
 95%|█████████▌| 10569/11074 [1:30:10<04:10,  2.02it/s]
 95%|█████████▌| 10570/11074 [1:30:11<04:09,  2.02it/s]
 95%|█████████▌| 10571/11074 [1:30:11<04:08,  2.02it/s]
 95%|█████████▌| 10572/11074 [1:30:12<04:08,  2.02it/s]
 95%|█████████▌| 10573/11074 [1:30:12<04:08,  2.02it/s]
 95%|█████████▌| 10574/11074 [1:30:13<04:07,  2.02it/s]
 95%|█████████▌| 10575/11074 [1:30:13<04:06,  2.02it/s]{'loss': 3.1221, 'grad_norm': 0.24080324172973633, 'learning_rate': 6.173101275608662e-06, 'epoch': 13.36}
                                                       
+
 95%|█████████▌| 10575/11074 [1:30:13<04:06,  2.02it/s]
 96%|█████████▌| 10576/11074 [1:30:14<04:06,  2.02it/s]
 96%|█████████▌| 10577/11074 [1:30:14<04:06,  2.02it/s]
 96%|█████████▌| 10578/11074 [1:30:15<04:05,  2.02it/s]
 96%|█████████▌| 10579/11074 [1:30:15<04:04,  2.02it/s]
 96%|█████████▌| 10580/11074 [1:30:16<04:04,  2.02it/s]
 96%|█████████▌| 10581/11074 [1:30:16<04:03,  2.02it/s]
 96%|█████████▌| 10582/11074 [1:30:17<04:02,  2.02it/s]
 96%|█████████▌| 10583/11074 [1:30:17<04:02,  2.02it/s]
 96%|█████████▌| 10584/11074 [1:30:18<04:02,  2.02it/s]
 96%|█████████▌| 10585/11074 [1:30:18<04:01,  2.02it/s]
 96%|█████████▌| 10586/11074 [1:30:19<04:01,  2.02it/s]
 96%|█████████▌| 10587/11074 [1:30:19<04:00,  2.02it/s]
 96%|█████████▌| 10588/11074 [1:30:20<04:00,  2.02it/s]
 96%|█████████▌| 10589/11074 [1:30:20<03:59,  2.02it/s]
 96%|█████████▌| 10590/11074 [1:30:21<03:59,  2.02it/s]
 96%|█████████▌| 10591/11074 [1:30:21<03:58,  2.02it/s]
 96%|█████████▌| 10592/11074 [1:30:21<03:58,  2.02it/s]
 96%|█████████▌| 10593/11074 [1:30:22<03:57,  2.02it/s]
 96%|█████████▌| 10594/11074 [1:30:22<03:57,  2.02it/s]
 96%|█████████▌| 10595/11074 [1:30:23<03:56,  2.02it/s]
 96%|█████████▌| 10596/11074 [1:30:23<03:56,  2.02it/s]
 96%|█████████▌| 10597/11074 [1:30:24<03:55,  2.02it/s]
 96%|█████████▌| 10598/11074 [1:30:24<03:55,  2.02it/s]
 96%|█████████▌| 10599/11074 [1:30:25<03:54,  2.02it/s]
 96%|█████████▌| 10600/11074 [1:30:25<03:53,  2.03it/s]
                                                       
{'loss': 3.1263, 'grad_norm': 0.2395830899477005, 'learning_rate': 5.571171265877917e-06, 'epoch': 13.39}
+
 96%|█████████▌| 10600/11074 [1:30:25<03:53,  2.03it/s]
 96%|█████████▌| 10601/11074 [1:30:26<03:54,  2.02it/s]
 96%|█████████▌| 10602/11074 [1:30:26<03:53,  2.02it/s]
 96%|█████████▌| 10603/11074 [1:30:27<03:53,  2.02it/s]
 96%|█████████▌| 10604/11074 [1:30:27<03:52,  2.02it/s]
 96%|█████████▌| 10605/11074 [1:30:28<03:51,  2.02it/s]
 96%|█████████▌| 10606/11074 [1:30:28<03:51,  2.02it/s]
 96%|█████████▌| 10607/11074 [1:30:29<03:50,  2.02it/s]
 96%|█████████▌| 10608/11074 [1:30:29<03:50,  2.02it/s]
 96%|█████████▌| 10609/11074 [1:30:30<03:50,  2.02it/s]
 96%|█████████▌| 10610/11074 [1:30:30<03:49,  2.02it/s]
 96%|█████████▌| 10611/11074 [1:30:31<03:49,  2.02it/s]
 96%|█████████▌| 10612/11074 [1:30:31<03:48,  2.02it/s]
 96%|█████████▌| 10613/11074 [1:30:32<03:48,  2.02it/s]
 96%|█████████▌| 10614/11074 [1:30:32<03:47,  2.02it/s]
 96%|█████████▌| 10615/11074 [1:30:33<03:47,  2.02it/s]
 96%|█████████▌| 10616/11074 [1:30:33<03:46,  2.02it/s]
 96%|█████████▌| 10617/11074 [1:30:34<03:46,  2.02it/s]
 96%|█████████▌| 10618/11074 [1:30:34<03:45,  2.02it/s]
 96%|█████████▌| 10619/11074 [1:30:35<03:45,  2.02it/s]
 96%|█████████▌| 10620/11074 [1:30:35<03:44,  2.02it/s]
 96%|█████████▌| 10621/11074 [1:30:36<03:43,  2.02it/s]
 96%|█████████▌| 10622/11074 [1:30:36<03:43,  2.02it/s]
 96%|█████████▌| 10623/11074 [1:30:37<03:42,  2.02it/s]
 96%|█████████▌| 10624/11074 [1:30:37<03:42,  2.02it/s]
 96%|█████████▌| 10625/11074 [1:30:38<03:42,  2.02it/s]{'loss': 3.1287, 'grad_norm': 0.2416277527809143, 'learning_rate': 4.999948407856669e-06, 'epoch': 13.42}
                                                       
+
 96%|█████████▌| 10625/11074 [1:30:38<03:42,  2.02it/s]
 96%|█████████▌| 10626/11074 [1:30:38<03:41,  2.02it/s]
 96%|█████████▌| 10627/11074 [1:30:39<03:41,  2.02it/s]
 96%|█████████▌| 10628/11074 [1:30:39<03:40,  2.02it/s]
 96%|█████████▌| 10629/11074 [1:30:40<03:39,  2.02it/s]
 96%|█████████▌| 10630/11074 [1:30:40<03:39,  2.02it/s]
 96%|█████████▌| 10631/11074 [1:30:41<03:39,  2.02it/s]
 96%|█████████▌| 10632/11074 [1:30:41<03:38,  2.02it/s]
 96%|█████████▌| 10633/11074 [1:30:42<03:38,  2.02it/s]
 96%|█████████▌| 10634/11074 [1:30:42<03:37,  2.02it/s]
 96%|█████████▌| 10635/11074 [1:30:43<03:37,  2.02it/s]
 96%|█████████▌| 10636/11074 [1:30:43<03:36,  2.02it/s]
 96%|█████████▌| 10637/11074 [1:30:44<03:36,  2.02it/s]
 96%|█████████▌| 10638/11074 [1:30:44<03:35,  2.02it/s]
 96%|█████████▌| 10639/11074 [1:30:45<03:35,  2.02it/s]
 96%|█████████▌| 10640/11074 [1:30:45<03:34,  2.02it/s]
 96%|█████████▌| 10641/11074 [1:30:46<03:34,  2.02it/s]
 96%|█████████▌| 10642/11074 [1:30:46<03:33,  2.02it/s]
 96%|█████████▌| 10643/11074 [1:30:47<03:33,  2.02it/s]
 96%|█████████▌| 10644/11074 [1:30:47<03:32,  2.02it/s]
 96%|█████████▌| 10645/11074 [1:30:48<03:32,  2.02it/s]
 96%|█████████▌| 10646/11074 [1:30:48<03:31,  2.02it/s]
 96%|█████████▌| 10647/11074 [1:30:49<03:31,  2.02it/s]
 96%|█████████▌| 10648/11074 [1:30:49<03:30,  2.02it/s]
 96%|█████████▌| 10649/11074 [1:30:50<03:30,  2.02it/s]
 96%|█████████▌| 10650/11074 [1:30:50<03:29,  2.02it/s]
                                                       
{'loss': 3.1299, 'grad_norm': 0.2407858520746231, 'learning_rate': 4.4594681780905646e-06, 'epoch': 13.46}
+
 96%|█████████▌| 10650/11074 [1:30:50<03:29,  2.02it/s]
 96%|█████████▌| 10651/11074 [1:30:51<03:29,  2.02it/s]
 96%|█████████▌| 10652/11074 [1:30:51<03:28,  2.02it/s]
 96%|█████████▌| 10653/11074 [1:30:52<03:27,  2.03it/s]
 96%|█████████▌| 10654/11074 [1:30:52<03:27,  2.02it/s]
 96%|█████████▌| 10655/11074 [1:30:53<03:26,  2.02it/s]
 96%|█████████▌| 10656/11074 [1:30:53<03:26,  2.02it/s]
 96%|█████████▌| 10657/11074 [1:30:54<03:25,  2.03it/s]
 96%|█████████▌| 10658/11074 [1:30:54<03:25,  2.03it/s]
 96%|█████████▋| 10659/11074 [1:30:55<03:24,  2.03it/s]
 96%|█████████▋| 10660/11074 [1:30:55<03:24,  2.03it/s]
 96%|█████████▋| 10661/11074 [1:30:56<03:23,  2.03it/s]
 96%|█████████▋| 10662/11074 [1:30:56<03:23,  2.03it/s]
 96%|█████████▋| 10663/11074 [1:30:57<03:22,  2.03it/s]
 96%|█████████▋| 10664/11074 [1:30:57<03:22,  2.03it/s]
 96%|█████████▋| 10665/11074 [1:30:58<03:21,  2.03it/s]
 96%|█████████▋| 10666/11074 [1:30:58<03:21,  2.03it/s]
 96%|█████████▋| 10667/11074 [1:30:59<03:21,  2.02it/s]
 96%|█████████▋| 10668/11074 [1:30:59<03:20,  2.03it/s]
 96%|█████████▋| 10669/11074 [1:31:00<03:20,  2.02it/s]
 96%|█████████▋| 10670/11074 [1:31:00<03:19,  2.02it/s]
 96%|█████████▋| 10671/11074 [1:31:01<03:19,  2.02it/s]
 96%|█████████▋| 10672/11074 [1:31:01<03:18,  2.03it/s]
 96%|█████████▋| 10673/11074 [1:31:02<03:18,  2.03it/s]
 96%|█████████▋| 10674/11074 [1:31:02<03:17,  2.03it/s]
 96%|█████████▋| 10675/11074 [1:31:03<03:17,  2.02it/s]
                                                       
{'loss': 3.1243, 'grad_norm': 0.23897473514080048, 'learning_rate': 3.949764143814149e-06, 'epoch': 13.49}
+
 96%|█████████▋| 10675/11074 [1:31:03<03:17,  2.02it/s]
 96%|█████████▋| 10676/11074 [1:31:03<03:16,  2.02it/s]
 96%|█████████▋| 10677/11074 [1:31:04<03:15,  2.03it/s]
 96%|█████████▋| 10678/11074 [1:31:04<03:15,  2.02it/s]
 96%|█████████▋| 10679/11074 [1:31:05<03:15,  2.03it/s]
 96%|█████████▋| 10680/11074 [1:31:05<03:14,  2.02it/s]
 96%|█████████▋| 10681/11074 [1:31:05<03:14,  2.02it/s]
 96%|█████████▋| 10682/11074 [1:31:06<03:13,  2.02it/s]
 96%|█████████▋| 10683/11074 [1:31:06<03:13,  2.03it/s]
 96%|█████████▋| 10684/11074 [1:31:07<03:12,  2.02it/s]
 96%|█████████▋| 10685/11074 [1:31:07<03:12,  2.02it/s]
 96%|█████████▋| 10686/11074 [1:31:08<03:11,  2.02it/s]
 97%|█████████▋| 10687/11074 [1:31:08<03:11,  2.02it/s]
 97%|█████████▋| 10688/11074 [1:31:09<03:10,  2.03it/s]
 97%|█████████▋| 10689/11074 [1:31:09<03:10,  2.02it/s]
 97%|█████████▋| 10690/11074 [1:31:10<03:09,  2.02it/s]
 97%|█████████▋| 10691/11074 [1:31:10<03:09,  2.02it/s]
 97%|█████████▋| 10692/11074 [1:31:11<03:08,  2.02it/s]
 97%|█████████▋| 10693/11074 [1:31:11<03:08,  2.02it/s]
 97%|█████████▋| 10694/11074 [1:31:12<03:07,  2.02it/s]
 97%|█████████▋| 10695/11074 [1:31:12<03:07,  2.02it/s]
 97%|█████████▋| 10696/11074 [1:31:13<03:06,  2.02it/s]
 97%|█████████▋| 10697/11074 [1:31:13<03:06,  2.02it/s]
 97%|█████████▋| 10698/11074 [1:31:14<03:05,  2.02it/s]
 97%|█████████▋| 10699/11074 [1:31:14<03:05,  2.02it/s]
 97%|█████████▋| 10700/11074 [1:31:15<03:04,  2.03it/s]
                                                       
{'loss': 3.1207, 'grad_norm': 0.24008893966674805, 'learning_rate': 3.47086796086582e-06, 'epoch': 13.52}
+
 97%|█████████▋| 10700/11074 [1:31:15<03:04,  2.03it/s]
 97%|█████████▋| 10701/11074 [1:31:15<03:04,  2.02it/s]
 97%|█████████▋| 10702/11074 [1:31:16<03:03,  2.02it/s]
 97%|█████████▋| 10703/11074 [1:31:16<03:03,  2.02it/s]
 97%|█████████▋| 10704/11074 [1:31:17<03:02,  2.03it/s]
 97%|█████████▋| 10705/11074 [1:31:17<03:02,  2.03it/s]
 97%|█████████▋| 10706/11074 [1:31:18<03:01,  2.02it/s]
 97%|█████████▋| 10707/11074 [1:31:18<03:01,  2.02it/s]
 97%|█████████▋| 10708/11074 [1:31:19<03:00,  2.02it/s]
 97%|█████████▋| 10709/11074 [1:31:19<03:00,  2.03it/s]
 97%|█████████▋| 10710/11074 [1:31:20<02:59,  2.02it/s]
 97%|█████████▋| 10711/11074 [1:31:20<02:59,  2.03it/s]
 97%|█████████▋| 10712/11074 [1:31:21<02:58,  2.02it/s]
 97%|█████████▋| 10713/11074 [1:31:21<02:58,  2.02it/s]
 97%|█████████▋| 10714/11074 [1:31:22<02:57,  2.02it/s]
 97%|█████████▋| 10715/11074 [1:31:22<02:57,  2.02it/s]
 97%|█████████▋| 10716/11074 [1:31:23<02:57,  2.02it/s]
 97%|█████████▋| 10717/11074 [1:31:23<02:56,  2.02it/s]
 97%|█████████▋| 10718/11074 [1:31:24<02:56,  2.02it/s]
 97%|█████████▋| 10719/11074 [1:31:24<02:55,  2.02it/s]
 97%|█████████▋| 10720/11074 [1:31:25<02:55,  2.02it/s]
 97%|█████████▋| 10721/11074 [1:31:25<02:54,  2.02it/s]
 97%|█████████▋| 10722/11074 [1:31:26<02:53,  2.02it/s]
 97%|█████████▋| 10723/11074 [1:31:26<02:53,  2.02it/s]
 97%|█████████▋| 10724/11074 [1:31:27<02:52,  2.02it/s]
 97%|█████████▋| 10725/11074 [1:31:27<02:52,  2.02it/s]
                                                       {'loss': 3.1258, 'grad_norm': 0.2390448898077011, 'learning_rate': 3.022809371721891e-06, 'epoch': 13.55}
+

 97%|█████████▋| 10725/11074 [1:31:27<02:52,  2.02it/s]
 97%|█████████▋| 10726/11074 [1:31:28<02:52,  2.02it/s]
 97%|█████████▋| 10727/11074 [1:31:28<02:51,  2.02it/s]
 97%|█████████▋| 10728/11074 [1:31:29<02:51,  2.02it/s]
 97%|█████████▋| 10729/11074 [1:31:29<02:50,  2.02it/s]
 97%|█████████▋| 10730/11074 [1:31:30<02:50,  2.02it/s]
 97%|█████████▋| 10731/11074 [1:31:30<02:49,  2.02it/s]
 97%|█████████▋| 10732/11074 [1:31:31<02:49,  2.02it/s]
 97%|█████████▋| 10733/11074 [1:31:31<02:48,  2.02it/s]
 97%|█████████▋| 10734/11074 [1:31:32<02:47,  2.02it/s]
 97%|█████████▋| 10735/11074 [1:31:32<02:47,  2.02it/s]
 97%|█████████▋| 10736/11074 [1:31:33<02:47,  2.02it/s]
 97%|█████████▋| 10737/11074 [1:31:33<02:46,  2.02it/s]
 97%|█████████▋| 10738/11074 [1:31:34<02:46,  2.02it/s]
 97%|█████████▋| 10739/11074 [1:31:34<02:45,  2.02it/s]
 97%|█████████▋| 10740/11074 [1:31:35<02:45,  2.02it/s]
 97%|█████████▋| 10741/11074 [1:31:35<02:44,  2.02it/s]
 97%|█████████▋| 10742/11074 [1:31:36<02:44,  2.02it/s]
 97%|█████████▋| 10743/11074 [1:31:36<02:43,  2.02it/s]
 97%|█████████▋| 10744/11074 [1:31:37<02:43,  2.02it/s]
 97%|█████████▋| 10745/11074 [1:31:37<02:42,  2.02it/s]
 97%|█████████▋| 10746/11074 [1:31:38<02:42,  2.02it/s]
 97%|█████████▋| 10747/11074 [1:31:38<02:41,  2.02it/s]
 97%|█████████▋| 10748/11074 [1:31:39<02:41,  2.02it/s]
 97%|█████████▋| 10749/11074 [1:31:39<02:40,  2.02it/s]
 97%|█████████▋| 10750/11074 [1:31:40<02:40,  2.02it/s]
                                                       
{'loss': 3.1294, 'grad_norm': 0.241078719496727, 'learning_rate': 2.605616203649408e-06, 'epoch': 13.58}
+
 97%|█████████▋| 10750/11074 [1:31:40<02:40,  2.02it/s]
 97%|█████████▋| 10751/11074 [1:31:40<02:39,  2.02it/s]
 97%|█████████▋| 10752/11074 [1:31:41<02:39,  2.02it/s]
 97%|█████████▋| 10753/11074 [1:31:41<02:38,  2.02it/s]
 97%|█████████▋| 10754/11074 [1:31:42<02:38,  2.02it/s]
 97%|█████████▋| 10755/11074 [1:31:42<02:37,  2.02it/s]
 97%|█████████▋| 10756/11074 [1:31:43<02:37,  2.02it/s]
 97%|█████████▋| 10757/11074 [1:31:43<02:36,  2.02it/s]
 97%|█████████▋| 10758/11074 [1:31:44<02:36,  2.03it/s]
 97%|█████████▋| 10759/11074 [1:31:44<02:35,  2.02it/s]
 97%|█████████▋| 10760/11074 [1:31:45<02:35,  2.03it/s]
 97%|█████████▋| 10761/11074 [1:31:45<02:34,  2.02it/s]
 97%|█████████▋| 10762/11074 [1:31:46<02:34,  2.03it/s]
 97%|█████████▋| 10763/11074 [1:31:46<02:33,  2.02it/s]
 97%|█████████▋| 10764/11074 [1:31:47<02:33,  2.02it/s]
 97%|█████████▋| 10765/11074 [1:31:47<02:32,  2.02it/s]
 97%|█████████▋| 10766/11074 [1:31:48<02:32,  2.02it/s]
 97%|█████████▋| 10767/11074 [1:31:48<02:31,  2.02it/s]
 97%|█████████▋| 10768/11074 [1:31:48<02:31,  2.02it/s]
 97%|█████████▋| 10769/11074 [1:31:49<02:30,  2.02it/s]
 97%|█████████▋| 10770/11074 [1:31:49<02:30,  2.02it/s]
 97%|█████████▋| 10771/11074 [1:31:50<02:29,  2.02it/s]
 97%|█████████▋| 10772/11074 [1:31:50<02:29,  2.02it/s]
 97%|█████████▋| 10773/11074 [1:31:51<02:28,  2.02it/s]
 97%|█████████▋| 10774/11074 [1:31:51<02:28,  2.02it/s]
 97%|█████████▋| 10775/11074 [1:31:52<02:27,  2.02it/s]
                                                       {'loss': 3.1297, 'grad_norm': 0.2402792125940323, 'learning_rate': 2.2193143669780316e-06, 'epoch': 13.61}
+
 97%|█████████▋| 10775/11074 [1:31:52<02:27,  2.02it/s]
 97%|█████████▋| 10776/11074 [1:31:52<02:27,  2.02it/s]
 97%|█████████▋| 10777/11074 [1:31:53<02:26,  2.02it/s]
 97%|█████████▋| 10778/11074 [1:31:53<02:26,  2.02it/s]
 97%|█████████▋| 10779/11074 [1:31:54<02:25,  2.02it/s]
 97%|█████████▋| 10780/11074 [1:31:54<02:25,  2.02it/s]
 97%|█████████▋| 10781/11074 [1:31:55<02:24,  2.02it/s]
 97%|█████████▋| 10782/11074 [1:31:55<02:24,  2.02it/s]
 97%|█████████▋| 10783/11074 [1:31:56<02:23,  2.02it/s]
 97%|█████████▋| 10784/11074 [1:31:56<02:23,  2.02it/s]
 97%|█████████▋| 10785/11074 [1:31:57<02:22,  2.02it/s]
 97%|█████████▋| 10786/11074 [1:31:57<02:22,  2.02it/s]
 97%|█████████▋| 10787/11074 [1:31:58<02:21,  2.02it/s]
 97%|█████████▋| 10788/11074 [1:31:58<02:21,  2.03it/s]
 97%|█████████▋| 10789/11074 [1:31:59<02:20,  2.03it/s]
 97%|█████████▋| 10790/11074 [1:31:59<02:20,  2.02it/s]
 97%|█████████▋| 10791/11074 [1:32:00<02:19,  2.03it/s]
 97%|█████████▋| 10792/11074 [1:32:00<02:19,  2.03it/s]
 97%|█████████▋| 10793/11074 [1:32:01<02:18,  2.03it/s]
 97%|█████████▋| 10794/11074 [1:32:01<02:18,  2.02it/s]
 97%|█████████▋| 10795/11074 [1:32:02<02:17,  2.02it/s]
 97%|█████████▋| 10796/11074 [1:32:02<02:17,  2.02it/s]
 97%|█████████▋| 10797/11074 [1:32:03<02:29,  1.86it/s]
 98%|█████████▊| 10798/11074 [1:32:03<02:24,  1.90it/s]
 98%|█████████▊| 10799/11074 [1:32:04<02:21,  1.94it/s]
 98%|█████████▊| 10800/11074 [1:32:04<02:19,  1.96it/s]{'loss': 3.1268, 'grad_norm': 0.2377961426973343, 'learning_rate': 1.863927853490599e-06, 'epoch': 13.64}
                                                       
+
 98%|█████████▊| 10800/11074 [1:32:04<02:19,  1.96it/s]
 98%|█████████▊| 10801/11074 [1:32:05<02:17,  1.98it/s]
 98%|█████████▊| 10802/11074 [1:32:05<02:16,  1.99it/s]
 98%|█████████▊| 10803/11074 [1:32:06<02:15,  2.00it/s]
 98%|█████████▊| 10804/11074 [1:32:06<02:14,  2.01it/s]
 98%|█████████▊| 10805/11074 [1:32:07<02:13,  2.02it/s]
 98%|█████████▊| 10806/11074 [1:32:07<02:12,  2.02it/s]
 98%|█████████▊| 10807/11074 [1:32:08<02:12,  2.02it/s]
 98%|█████████▊| 10808/11074 [1:32:08<02:11,  2.02it/s]
 98%|█████████▊| 10809/11074 [1:32:09<02:11,  2.02it/s]
 98%|█████████▊| 10810/11074 [1:32:09<02:10,  2.02it/s]
 98%|█████████▊| 10811/11074 [1:32:10<02:10,  2.02it/s]
 98%|█████████▊| 10812/11074 [1:32:10<02:09,  2.02it/s]
 98%|█████████▊| 10813/11074 [1:32:11<02:08,  2.02it/s]
 98%|█████████▊| 10814/11074 [1:32:11<02:08,  2.03it/s]
 98%|█████████▊| 10815/11074 [1:32:12<02:07,  2.03it/s]
 98%|█████████▊| 10816/11074 [1:32:12<02:07,  2.03it/s]
 98%|█████████▊| 10817/11074 [1:32:13<02:06,  2.03it/s]
 98%|█████████▊| 10818/11074 [1:32:13<02:06,  2.03it/s]
 98%|█████████▊| 10819/11074 [1:32:14<02:05,  2.03it/s]
 98%|█████████▊| 10820/11074 [1:32:14<02:05,  2.03it/s]
 98%|█████████▊| 10821/11074 [1:32:15<02:04,  2.03it/s]
 98%|█████████▊| 10822/11074 [1:32:15<02:04,  2.03it/s]
 98%|█████████▊| 10823/11074 [1:32:16<02:03,  2.03it/s]
 98%|█████████▊| 10824/11074 [1:32:16<02:03,  2.03it/s]
 98%|█████████▊| 10825/11074 [1:32:17<02:02,  2.03it/s]{'loss': 3.125, 'grad_norm': 0.23915834724903107, 'learning_rate': 1.5394787349332084e-06, 'epoch': 13.68}
                                                       
+
 98%|█████████▊| 10825/11074 [1:32:17<02:02,  2.03it/s]
 98%|█████████▊| 10826/11074 [1:32:17<02:02,  2.02it/s]
 98%|█████████▊| 10827/11074 [1:32:18<02:01,  2.03it/s]
 98%|█████████▊| 10828/11074 [1:32:18<02:01,  2.03it/s]
 98%|█████████▊| 10829/11074 [1:32:19<02:00,  2.03it/s]
 98%|█████████▊| 10830/11074 [1:32:19<02:00,  2.02it/s]
 98%|█████████▊| 10831/11074 [1:32:20<01:59,  2.03it/s]
 98%|█████████▊| 10832/11074 [1:32:20<01:59,  2.02it/s]
 98%|█████████▊| 10833/11074 [1:32:21<01:59,  2.02it/s]
 98%|█████████▊| 10834/11074 [1:32:21<01:58,  2.02it/s]
 98%|█████████▊| 10835/11074 [1:32:22<01:58,  2.02it/s]
 98%|█████████▊| 10836/11074 [1:32:22<01:57,  2.02it/s]
 98%|█████████▊| 10837/11074 [1:32:23<01:57,  2.02it/s]
 98%|█████████▊| 10838/11074 [1:32:23<01:56,  2.02it/s]
 98%|█████████▊| 10839/11074 [1:32:24<01:56,  2.02it/s]
 98%|█████████▊| 10840/11074 [1:32:24<01:55,  2.02it/s]
 98%|█████████▊| 10841/11074 [1:32:25<01:55,  2.02it/s]
 98%|█████████▊| 10842/11074 [1:32:25<01:54,  2.02it/s]
 98%|█████████▊| 10843/11074 [1:32:26<01:54,  2.03it/s]
 98%|█████████▊| 10844/11074 [1:32:26<01:53,  2.02it/s]
 98%|█████████▊| 10845/11074 [1:32:27<01:52,  2.03it/s]
 98%|█████████▊| 10846/11074 [1:32:27<01:52,  2.03it/s]
 98%|█████████▊| 10847/11074 [1:32:28<01:52,  2.03it/s]
 98%|█████████▊| 10848/11074 [1:32:28<01:51,  2.03it/s]
 98%|█████████▊| 10849/11074 [1:32:29<01:50,  2.03it/s]
 98%|█████████▊| 10850/11074 [1:32:29<01:50,  2.03it/s]
                                                       {'loss': 3.1288, 'grad_norm': 0.23611889779567719, 'learning_rate': 1.2459871616444796e-06, 'epoch': 13.71}
+

 98%|█████████▊| 10850/11074 [1:32:29<01:50,  2.03it/s]
 98%|█████████▊| 10851/11074 [1:32:30<01:50,  2.02it/s]
 98%|█████████▊| 10852/11074 [1:32:30<01:49,  2.02it/s]
 98%|█████████▊| 10853/11074 [1:32:31<01:49,  2.02it/s]
 98%|█████████▊| 10854/11074 [1:32:31<01:48,  2.02it/s]
 98%|█████████▊| 10855/11074 [1:32:32<01:48,  2.02it/s]
 98%|█████████▊| 10856/11074 [1:32:32<01:47,  2.03it/s]
 98%|█████████▊| 10857/11074 [1:32:33<01:47,  2.03it/s]
 98%|█████████▊| 10858/11074 [1:32:33<01:46,  2.02it/s]
 98%|█████████▊| 10859/11074 [1:32:34<01:46,  2.02it/s]
 98%|█████████▊| 10860/11074 [1:32:34<01:45,  2.02it/s]
 98%|█████████▊| 10861/11074 [1:32:35<01:45,  2.02it/s]
 98%|█████████▊| 10862/11074 [1:32:35<01:44,  2.02it/s]
 98%|█████████▊| 10863/11074 [1:32:36<01:44,  2.02it/s]
 98%|█████████▊| 10864/11074 [1:32:36<01:43,  2.02it/s]
 98%|█████████▊| 10865/11074 [1:32:37<01:43,  2.02it/s]
 98%|█████████▊| 10866/11074 [1:32:37<01:42,  2.02it/s]
 98%|█████████▊| 10867/11074 [1:32:38<01:42,  2.03it/s]
 98%|█████████▊| 10868/11074 [1:32:38<01:41,  2.02it/s]
 98%|█████████▊| 10869/11074 [1:32:39<01:41,  2.03it/s]
 98%|█████████▊| 10870/11074 [1:32:39<01:40,  2.03it/s]
 98%|█████████▊| 10871/11074 [1:32:40<01:40,  2.02it/s]
 98%|█████████▊| 10872/11074 [1:32:40<01:39,  2.02it/s]
 98%|█████████▊| 10873/11074 [1:32:41<01:39,  2.02it/s]
 98%|█████████▊| 10874/11074 [1:32:41<01:38,  2.02it/s]
 98%|█████████▊| 10875/11074 [1:32:41<01:38,  2.02it/s]{'loss': 3.1334, 'grad_norm': 0.23865821957588196, 'learning_rate': 9.834713613040003e-07, 'epoch': 13.74}
+
                                                       

 98%|█████████▊| 10875/11074 [1:32:41<01:38,  2.02it/s]
 98%|█████████▊| 10876/11074 [1:32:42<01:38,  2.02it/s]
 98%|█████████▊| 10877/11074 [1:32:42<01:37,  2.02it/s]
 98%|█████████▊| 10878/11074 [1:32:43<01:36,  2.02it/s]
 98%|█████████▊| 10879/11074 [1:32:43<01:36,  2.02it/s]
 98%|█████████▊| 10880/11074 [1:32:44<01:35,  2.02it/s]
 98%|█████████▊| 10881/11074 [1:32:44<01:35,  2.02it/s]
 98%|█████████▊| 10882/11074 [1:32:45<01:34,  2.02it/s]
 98%|█████████▊| 10883/11074 [1:32:45<01:34,  2.02it/s]
 98%|█████████▊| 10884/11074 [1:32:46<01:33,  2.02it/s]
 98%|█████████▊| 10885/11074 [1:32:46<01:33,  2.02it/s]
 98%|█████████▊| 10886/11074 [1:32:47<01:32,  2.03it/s]
 98%|█████████▊| 10887/11074 [1:32:47<01:32,  2.02it/s]
 98%|█████████▊| 10888/11074 [1:32:48<01:31,  2.03it/s]
 98%|█████████▊| 10889/11074 [1:32:48<01:31,  2.02it/s]
 98%|█████████▊| 10890/11074 [1:32:49<01:30,  2.03it/s]
 98%|█████████▊| 10891/11074 [1:32:49<01:30,  2.03it/s]
 98%|█████████▊| 10892/11074 [1:32:50<01:29,  2.03it/s]
 98%|█████████▊| 10893/11074 [1:32:50<01:29,  2.03it/s]
 98%|█████████▊| 10894/11074 [1:32:51<01:28,  2.03it/s]
 98%|█████████▊| 10895/11074 [1:32:51<01:28,  2.03it/s]
 98%|█████████▊| 10896/11074 [1:32:52<01:35,  1.86it/s]
 98%|█████████▊| 10897/11074 [1:32:53<01:32,  1.91it/s]
 98%|█████████▊| 10898/11074 [1:32:53<01:30,  1.94it/s]
 98%|█████████▊| 10899/11074 [1:32:53<01:29,  1.97it/s]
 98%|█████████▊| 10900/11074 [1:32:54<01:27,  1.98it/s]
                                                       {'loss': 3.126, 'grad_norm': 0.23910360038280487, 'learning_rate': 7.519476378002321e-07, 'epoch': 13.77}
+

 98%|█████████▊| 10900/11074 [1:32:54<01:27,  1.98it/s]
 98%|█████████▊| 10901/11074 [1:32:54<01:26,  1.99it/s]
 98%|█████████▊| 10902/11074 [1:32:55<01:25,  2.00it/s]
 98%|█████████▊| 10903/11074 [1:32:55<01:25,  2.01it/s]
 98%|█████████▊| 10904/11074 [1:32:56<01:24,  2.01it/s]
 98%|█████████▊| 10905/11074 [1:32:56<01:23,  2.02it/s]
 98%|█████████▊| 10906/11074 [1:32:57<01:23,  2.02it/s]
 98%|█████████▊| 10907/11074 [1:32:57<01:22,  2.02it/s]
 99%|█████████▊| 10908/11074 [1:32:58<01:22,  2.02it/s]
 99%|█████████▊| 10909/11074 [1:32:58<01:21,  2.02it/s]
 99%|█████████▊| 10910/11074 [1:32:59<01:20,  2.02it/s]
 99%|█████████▊| 10911/11074 [1:32:59<01:20,  2.02it/s]
 99%|█████████▊| 10912/11074 [1:33:00<01:20,  2.02it/s]
 99%|█████████▊| 10913/11074 [1:33:00<01:19,  2.03it/s]
 99%|█████████▊| 10914/11074 [1:33:01<01:19,  2.02it/s]
 99%|█████████▊| 10915/11074 [1:33:01<01:18,  2.03it/s]
 99%|█████████▊| 10916/11074 [1:33:02<01:18,  2.03it/s]
 99%|█████████▊| 10917/11074 [1:33:02<01:17,  2.03it/s]
 99%|█████████▊| 10918/11074 [1:33:03<01:17,  2.02it/s]
 99%|█████████▊| 10919/11074 [1:33:03<01:16,  2.03it/s]
 99%|█████████▊| 10920/11074 [1:33:04<01:16,  2.03it/s]
 99%|█████████▊| 10921/11074 [1:33:04<01:15,  2.03it/s]
 99%|█████████▊| 10922/11074 [1:33:05<01:15,  2.03it/s]
 99%|█████████▊| 10923/11074 [1:33:05<01:14,  2.02it/s]
 99%|█████████▊| 10924/11074 [1:33:06<01:14,  2.02it/s]
 99%|█████████▊| 10925/11074 [1:33:06<01:13,  2.02it/s]
                                                       
{'loss': 3.1259, 'grad_norm': 0.23768281936645508, 'learning_rate': 5.514303702181533e-07, 'epoch': 13.8}
+
 99%|█████████▊| 10925/11074 [1:33:06<01:13,  2.02it/s]
 99%|█████████▊| 10926/11074 [1:33:07<01:13,  2.01it/s]
 99%|█████████▊| 10927/11074 [1:33:07<01:12,  2.02it/s]
 99%|█████████▊| 10928/11074 [1:33:08<01:12,  2.02it/s]
 99%|█████████▊| 10929/11074 [1:33:08<01:11,  2.02it/s]
 99%|█████████▊| 10930/11074 [1:33:09<01:11,  2.02it/s]
 99%|█████████▊| 10931/11074 [1:33:09<01:10,  2.02it/s]
 99%|█████████▊| 10932/11074 [1:33:10<01:10,  2.02it/s]
 99%|█████████▊| 10933/11074 [1:33:10<01:09,  2.02it/s]
 99%|█████████▊| 10934/11074 [1:33:11<01:09,  2.02it/s]
 99%|█████████▊| 10935/11074 [1:33:11<01:08,  2.02it/s]
 99%|█████████▉| 10936/11074 [1:33:12<01:08,  2.02it/s]
 99%|█████████▉| 10937/11074 [1:33:12<01:07,  2.02it/s]
 99%|█████████▉| 10938/11074 [1:33:13<01:07,  2.03it/s]
 99%|█████████▉| 10939/11074 [1:33:13<01:06,  2.02it/s]
 99%|█████████▉| 10940/11074 [1:33:14<01:06,  2.03it/s]
 99%|█████████▉| 10941/11074 [1:33:14<01:05,  2.02it/s]
 99%|█████████▉| 10942/11074 [1:33:15<01:05,  2.03it/s]
 99%|█████████▉| 10943/11074 [1:33:15<01:04,  2.02it/s]
 99%|█████████▉| 10944/11074 [1:33:16<01:04,  2.02it/s]
 99%|█████████▉| 10945/11074 [1:33:16<01:03,  2.02it/s]
 99%|█████████▉| 10946/11074 [1:33:17<01:03,  2.02it/s]
 99%|█████████▉| 10947/11074 [1:33:17<01:02,  2.02it/s]
 99%|█████████▉| 10948/11074 [1:33:18<01:02,  2.02it/s]
 99%|█████████▉| 10949/11074 [1:33:18<01:01,  2.02it/s]
 99%|█████████▉| 10950/11074 [1:33:19<01:01,  2.02it/s]
                                                       
{'loss': 3.1238, 'grad_norm': 0.24086894094944, 'learning_rate': 3.819320119459735e-07, 'epoch': 13.83}
+
 99%|█████████▉| 10950/11074 [1:33:19<01:01,  2.02it/s]
 99%|█████████▉| 10951/11074 [1:33:19<01:00,  2.02it/s]
 99%|█████████▉| 10952/11074 [1:33:20<01:00,  2.02it/s]
 99%|█████████▉| 10953/11074 [1:33:20<00:59,  2.02it/s]
 99%|█████████▉| 10954/11074 [1:33:21<00:59,  2.02it/s]
 99%|█████████▉| 10955/11074 [1:33:21<00:58,  2.02it/s]
 99%|█████████▉| 10956/11074 [1:33:22<00:58,  2.02it/s]
 99%|█████████▉| 10957/11074 [1:33:22<00:57,  2.03it/s]
 99%|█████████▉| 10958/11074 [1:33:23<00:57,  2.02it/s]
 99%|█████████▉| 10959/11074 [1:33:23<00:56,  2.02it/s]
 99%|█████████▉| 10960/11074 [1:33:24<00:56,  2.02it/s]
 99%|█████████▉| 10961/11074 [1:33:24<00:55,  2.02it/s]
 99%|█████████▉| 10962/11074 [1:33:25<00:55,  2.02it/s]
 99%|█████████▉| 10963/11074 [1:33:25<00:54,  2.02it/s]
 99%|█████████▉| 10964/11074 [1:33:26<00:54,  2.02it/s]
 99%|█████████▉| 10965/11074 [1:33:26<00:53,  2.02it/s]
 99%|█████████▉| 10966/11074 [1:33:27<00:53,  2.02it/s]
 99%|█████████▉| 10967/11074 [1:33:27<00:52,  2.03it/s]
 99%|█████████▉| 10968/11074 [1:33:28<00:52,  2.02it/s]
 99%|█████████▉| 10969/11074 [1:33:28<00:51,  2.03it/s]
 99%|█████████▉| 10970/11074 [1:33:29<00:51,  2.02it/s]
 99%|█████████▉| 10971/11074 [1:33:29<00:50,  2.03it/s]
 99%|█████████▉| 10972/11074 [1:33:30<00:50,  2.03it/s]
 99%|█████████▉| 10973/11074 [1:33:30<00:49,  2.02it/s]
 99%|█████████▉| 10974/11074 [1:33:31<00:49,  2.03it/s]
 99%|█████████▉| 10975/11074 [1:33:31<00:48,  2.02it/s]
                                                       
{'loss': 3.1229, 'grad_norm': 0.23922082781791687, 'learning_rate': 2.4346308990175204e-07, 'epoch': 13.87}
+
 99%|█████████▉| 10975/11074 [1:33:31<00:48,  2.02it/s]
 99%|█████████▉| 10976/11074 [1:33:32<00:48,  2.02it/s]
 99%|█████████▉| 10977/11074 [1:33:32<00:47,  2.02it/s]
 99%|█████████▉| 10978/11074 [1:33:33<00:47,  2.02it/s]
 99%|█████████▉| 10979/11074 [1:33:33<00:46,  2.02it/s]
 99%|█████████▉| 10980/11074 [1:33:34<00:46,  2.02it/s]
 99%|█████████▉| 10981/11074 [1:33:34<00:46,  2.02it/s]
 99%|█████████▉| 10982/11074 [1:33:35<00:45,  2.02it/s]
 99%|█████████▉| 10983/11074 [1:33:35<00:45,  2.02it/s]
 99%|█████████▉| 10984/11074 [1:33:35<00:44,  2.02it/s]
 99%|█████████▉| 10985/11074 [1:33:36<00:44,  2.02it/s]
 99%|█████████▉| 10986/11074 [1:33:36<00:43,  2.02it/s]
 99%|█████████▉| 10987/11074 [1:33:37<00:43,  2.02it/s]
 99%|█████████▉| 10988/11074 [1:33:37<00:42,  2.02it/s]
 99%|█████████▉| 10989/11074 [1:33:38<00:42,  2.02it/s]
 99%|█████████▉| 10990/11074 [1:33:38<00:41,  2.02it/s]
 99%|█████████▉| 10991/11074 [1:33:39<00:41,  2.02it/s]
 99%|█████████▉| 10992/11074 [1:33:39<00:40,  2.02it/s]
 99%|█████████▉| 10993/11074 [1:33:40<00:40,  2.02it/s]
 99%|█████████▉| 10994/11074 [1:33:40<00:39,  2.02it/s]
 99%|█████████▉| 10995/11074 [1:33:41<00:39,  2.02it/s]
 99%|█████████▉| 10996/11074 [1:33:41<00:38,  2.02it/s]
 99%|█████████▉| 10997/11074 [1:33:42<00:38,  2.02it/s]
 99%|█████████▉| 10998/11074 [1:33:42<00:37,  2.02it/s]
 99%|█████████▉| 10999/11074 [1:33:43<00:37,  2.02it/s]
 99%|█████████▉| 11000/11074 [1:33:43<00:36,  2.02it/s]
                                                       {'loss': 3.1286, 'grad_norm': 0.24025793373584747, 'learning_rate': 1.3603220387981008e-07, 'epoch': 13.9}
+
 99%|█████████▉| 11000/11074 [1:33:43<00:36,  2.02it/s]
 99%|█████████▉| 11001/11074 [1:33:44<00:36,  2.02it/s]
 99%|█████████▉| 11002/11074 [1:33:44<00:35,  2.02it/s]
 99%|█████████▉| 11003/11074 [1:33:45<00:35,  2.02it/s]
 99%|█████████▉| 11004/11074 [1:33:45<00:34,  2.02it/s]
 99%|█████████▉| 11005/11074 [1:33:46<00:34,  2.02it/s]
 99%|█████████▉| 11006/11074 [1:33:46<00:33,  2.02it/s]
 99%|█████████▉| 11007/11074 [1:33:47<00:33,  2.02it/s]
 99%|█████████▉| 11008/11074 [1:33:47<00:32,  2.03it/s]
 99%|█████████▉| 11009/11074 [1:33:48<00:32,  2.02it/s]
 99%|█████████▉| 11010/11074 [1:33:48<00:31,  2.03it/s]
 99%|█████████▉| 11011/11074 [1:33:49<00:31,  2.02it/s]
 99%|█████████▉| 11012/11074 [1:33:49<00:30,  2.03it/s]
 99%|█████████▉| 11013/11074 [1:33:50<00:30,  2.02it/s]
 99%|█████████▉| 11014/11074 [1:33:50<00:29,  2.03it/s]
 99%|█████████▉| 11015/11074 [1:33:51<00:29,  2.02it/s]
 99%|█████████▉| 11016/11074 [1:33:51<00:28,  2.02it/s]
 99%|█████████▉| 11017/11074 [1:33:52<00:28,  2.02it/s]
 99%|█████████▉| 11018/11074 [1:33:52<00:27,  2.02it/s]
100%|█████████▉| 11019/11074 [1:33:53<00:27,  2.03it/s]
100%|█████████▉| 11020/11074 [1:33:53<00:26,  2.03it/s]
100%|█████████▉| 11021/11074 [1:33:54<00:26,  2.03it/s]
100%|█████████▉| 11022/11074 [1:33:54<00:25,  2.02it/s]
100%|█████████▉| 11023/11074 [1:33:55<00:25,  2.03it/s]
100%|█████████▉| 11024/11074 [1:33:55<00:24,  2.03it/s]
100%|█████████▉| 11025/11074 [1:33:56<00:24,  2.03it/s]
                                                       
{'loss': 3.1304, 'grad_norm': 0.2407616823911667, 'learning_rate': 5.964602601643554e-08, 'epoch': 13.93}
+
100%|█████████▉| 11025/11074 [1:33:56<00:24,  2.03it/s]
100%|█████████▉| 11026/11074 [1:33:56<00:23,  2.02it/s]
100%|█████████▉| 11027/11074 [1:33:57<00:23,  2.02it/s]
100%|█████████▉| 11028/11074 [1:33:57<00:22,  2.02it/s]
100%|█████████▉| 11029/11074 [1:33:58<00:22,  2.02it/s]
100%|█████████▉| 11030/11074 [1:33:58<00:21,  2.02it/s]
100%|█████████▉| 11031/11074 [1:33:59<00:21,  2.02it/s]
100%|█████████▉| 11032/11074 [1:33:59<00:20,  2.02it/s]
100%|█████████▉| 11033/11074 [1:34:00<00:20,  2.02it/s]
100%|█████████▉| 11034/11074 [1:34:00<00:19,  2.02it/s]
100%|█████████▉| 11035/11074 [1:34:01<00:19,  2.02it/s]
100%|█████████▉| 11036/11074 [1:34:01<00:18,  2.02it/s]
100%|█████████▉| 11037/11074 [1:34:02<00:18,  2.02it/s]
100%|█████████▉| 11038/11074 [1:34:02<00:17,  2.02it/s]
100%|█████████▉| 11039/11074 [1:34:03<00:17,  2.02it/s]
100%|█████████▉| 11040/11074 [1:34:03<00:16,  2.02it/s]
100%|█████████▉| 11041/11074 [1:34:04<00:16,  2.02it/s]
100%|█████████▉| 11042/11074 [1:34:04<00:15,  2.02it/s]
100%|█████████▉| 11043/11074 [1:34:05<00:15,  2.02it/s]
100%|█████████▉| 11044/11074 [1:34:05<00:14,  2.02it/s]
100%|█████████▉| 11045/11074 [1:34:06<00:14,  2.02it/s]
100%|█████████▉| 11046/11074 [1:34:06<00:13,  2.02it/s]
100%|█████████▉| 11047/11074 [1:34:07<00:13,  2.03it/s]
100%|█████████▉| 11048/11074 [1:34:07<00:12,  2.02it/s]
100%|█████████▉| 11049/11074 [1:34:08<00:12,  2.02it/s]
100%|█████████▉| 11050/11074 [1:34:08<00:11,  2.02it/s]{'loss': 3.1269, 'grad_norm': 0.23971088230609894, 'learning_rate': 1.4309300375381362e-08, 'epoch': 13.96}
+
                                                       

100%|█████████▉| 11050/11074 [1:34:08<00:11,  2.02it/s]
100%|█████████▉| 11051/11074 [1:34:09<00:11,  2.02it/s]
100%|█████████▉| 11052/11074 [1:34:09<00:10,  2.02it/s]
100%|█████████▉| 11053/11074 [1:34:10<00:10,  2.02it/s]
100%|█████████▉| 11054/11074 [1:34:10<00:09,  2.02it/s]
100%|█████████▉| 11055/11074 [1:34:11<00:09,  2.02it/s]
100%|█████████▉| 11056/11074 [1:34:11<00:08,  2.02it/s]
100%|█████████▉| 11057/11074 [1:34:12<00:08,  2.02it/s]
100%|█████████▉| 11058/11074 [1:34:12<00:07,  2.02it/s]
100%|█████████▉| 11059/11074 [1:34:13<00:07,  2.02it/s]
100%|█████████▉| 11060/11074 [1:34:13<00:06,  2.02it/s]
100%|█████████▉| 11061/11074 [1:34:14<00:06,  2.02it/s]
100%|█████████▉| 11062/11074 [1:34:14<00:05,  2.02it/s]
100%|█████████▉| 11063/11074 [1:34:15<00:05,  2.02it/s]
100%|█████████▉| 11064/11074 [1:34:15<00:04,  2.02it/s]
100%|█████████▉| 11065/11074 [1:34:16<00:04,  2.02it/s]
100%|█████████▉| 11066/11074 [1:34:16<00:03,  2.03it/s]
100%|█████████▉| 11067/11074 [1:34:17<00:03,  2.03it/s]
100%|█████████▉| 11068/11074 [1:34:17<00:02,  2.03it/s]
100%|█████████▉| 11069/11074 [1:34:18<00:02,  2.03it/s]
100%|█████████▉| 11070/11074 [1:34:18<00:01,  2.03it/s]
100%|█████████▉| 11071/11074 [1:34:18<00:01,  2.03it/s]
100%|█████████▉| 11072/11074 [1:34:19<00:00,  2.03it/s]
100%|█████████▉| 11073/11074 [1:34:19<00:00,  2.03it/s]
100%|██████████| 11074/11074 [1:34:20<00:00,  2.03it/s]
                                                       
{'train_runtime': 5672.2282, 'train_samples_per_second': 2000.029, 'train_steps_per_second': 1.952, 'train_loss': 3.67646764955388, 'epoch': 13.99}
+
100%|██████████| 11074/11074 [1:34:32<00:00,  2.03it/s]
100%|██████████| 11074/11074 [1:34:32<00:00,  1.95it/s]
+Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.