Training in progress, epoch 8

logs/events.out.tfevents.1716666516.sphinx2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:945421138e7b912d08d777fb7d869ac412c05b0157327b96fde5c102da685bcc
-size 61143
+oid sha256:416c97712b799604d8d1dbc257c7d59e1375e58d1c1d9979ffd97a311fdd2953
+size 68528

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:757708ce1edb8b67a3acd15d5f06ba367f961637ebc5567ec589e81bd58505fa
+oid sha256:a93ccbf9e7de1c05d13972e93f7e2089ab9d11136586cc04dc3036dbd059f0cc
 size 281715176

train_job_output.txt

@@ -756,4 +756,38 @@ command outputs:
 
 57%|█████▋    | 6625/11704 [56:18<41:55,  2.02it/s]
 57%|█████▋    | 6626/11704 [56:18<41:53,  2.02it/s]
 57%|█████▋    | 6627/11704 [56:19<41:54,  2.02it/s]
 57%|█████▋    | 6628/11704 [56:19<41:51,  2.02it/s]
 57%|█████▋    | 6629/11704 [56:20<41:52,  2.02it/s]
 57%|█████▋    | 6630/11704 [56:20<41:50,  2.02it/s]
 57%|█████▋    | 6631/11704 [56:21<41:50,  2.02it/s]
 57%|█████▋    | 6632/11704 [56:21<41:47,  2.02it/s]
 57%|█████▋    | 6633/11704 [56:22<41:49,  2.02it/s]
 57%|█████▋    | 6634/11704 [56:22<41:48,  2.02it/s]
 57%|█████▋    | 6635/11704 [56:23<41:49,  2.02it/s]
 57%|█████▋    | 6636/11704 [56:23<41:47,  2.02it/s]
 57%|█████▋    | 6637/11704 [56:24<41:48,  2.02it/s]
 57%|█████▋    | 6638/11704 [56:24<41:45,  2.02it/s]
 57%|█████▋    | 6639/11704 [56:25<41:45,  2.02it/s]
 57%|█████▋    | 6640/11704 [56:25<41:43,  2.02it/s]
 57%|█████▋    | 6641/11704 [56:26<41:47,  2.02it/s]
 57%|█████▋    | 6642/11704 [56:26<41:46,  2.02it/s]
 57%|█████▋    | 6643/11704 [56:27<41:49,  2.02it/s]
 57%|█████▋    | 6644/11704 [56:27<41:44,  2.02it/s]
 57%|█████▋    | 6645/11704 [56:28<41:44,  2.02it/s]
 57%|█████▋    | 6646/11704 [56:28<41:42,  2.02it/s]
 57%|█████▋    | 6647/11704 [56:29<41:41,  2.02it/s]
 57%|█████▋    | 6648/11704 [56:29<41:40,  2.02it/s]
 57%|█████▋    | 6649/11704 [56:30<41:39,  2.02it/s]
 57%|█████▋    | 6650/11704 [56:30<41:40,  2.02it/s]
                                                    
 
 57%|█████▋    | 6650/11704 [56:30<41:40,  2.02it/s]
 57%|█████▋    | 6651/11704 [56:31<41:44,  2.02it/s]
 57%|█████▋    | 6652/11704 [56:31<41:43,  2.02it/s]
 57%|█████▋    | 6653/11704 [56:32<41:43,  2.02it/s]
 57%|█████▋    | 6654/11704 [56:32<41:42,  2.02it/s]
 57%|█████▋    | 6655/11704 [56:33<41:39,  2.02it/s]
 57%|█████▋    | 6656/11704 [56:33<41:41,  2.02it/s]
 57%|█████▋    | 6657/11704 [56:34<41:38,  2.02it/s]
 57%|█████▋    | 6658/11704 [56:34<41:39,  2.02it/s]
 57%|█████▋    | 6659/11704 [56:35<41:37,  2.02it/s]
 57%|█████▋    | 6660/11704 [56:35<41:37,  2.02it/s]
 57%|█████▋    | 6661/11704 [56:36<41:35,  2.02it/s]
 57%|█████▋    | 6662/11704 [56:36<41:36,  2.02it/s]
 57%|█████▋    | 6663/11704 [56:37<41:34,  2.02it/s]
 57%|█████▋    | 6664/11704 [56:37<41:35,  2.02it/s]
 57%|█████▋    | 6665/11704 [56:38<41:34,  2.02it/s]
 57%|█████▋    | 6666/11704 [56:38<41:35,  2.02it/s]
 57%|█████▋    | 6667/11704 [56:39<41:34,  2.02it/s]
 57%|█████▋    | 6668/11704 [56:39<41:32,  2.02it/s]
 57%|█████▋    | 6669/11704 [56:40<41:30,  2.02it/s]
 57%|█████▋    | 6670/11704 [56:40<41:31,  2.02it/s]
 57%|█████▋    | 6671/11704 [56:41<41:29,  2.02it/s]
 57%|█████▋    | 6672/11704 [56:41<41:30,  2.02it/s]
 57%|█████▋    | 6673/11704 [56:42<41:29,  2.02it/s]
 57%|█████▋    | 6674/11704 [56:42<41:30,  2.02it/s]
 57%|█████▋    | 6675/11704 [56:43<41:28,  2.02it/s]
                                                    
 
 57%|█████▋    | 6675/11704 [56:43<41:28,  2.02it/s]
 57%|█████▋    | 6676/11704 [56:43<41:29,  2.02it/s]
 57%|█████▋    | 6677/11704 [56:44<41:28,  2.02it/s]
 57%|█████▋    | 6678/11704 [56:44<41:28,  2.02it/s]
 57%|█████▋    | 6679/11704 [56:45<41:28,  2.02it/s]
 57%|█████▋    | 6680/11704 [56:45<41:27,  2.02it/s]
 57%|█████▋    | 6681/11704 [56:46<41:27,  2.02it/s]
 57%|█████▋    | 6682/11704 [56:46<41:26,  2.02it/s]
 57%|█████▋    | 6683/11704 [56:47<41:28,  2.02it/s]
 57%|█████▋    | 6684/11704 [56:47<41:26,  2.02it/s]
 57%|█████▋    | 6685/11704 [56:48<41:26,  2.02it/s]
 57%|█████▋    | 6686/11704 [56:48<41:26,  2.02it/s]
 57%|█████▋    | 6687/11704 [56:49<41:24,  2.02it/s]
 57%|█████▋    | 6688/11704 [56:49<41:23,  2.02it/s]
 57%|█████▋    | 6689/11704 [56:50<41:24,  2.02it/s]
 57%|█████▋    | 6690/11704 [56:50<41:21,  2.02it/s]
 57%|█████▋    | 6691/11704 [56:50<41:22,  2.02it/s]
 57%|█████▋    | 6692/11704 [56:51<41:03,  2.03it/s]
 57%|█████▋    | 6693/11704 [57:03<5:33:00,  3.99s/it]
 57%|█████▋    | 6694/11704 [57:04<4:05:30,  2.94s/it]
 57%|█████▋    | 6695/11704 [57:04<3:04:12,  2.21s/it]
 57%|█���███▋    | 6696/11704 [57:05<2:21:21,  1.69s/it]
 57%|█████▋    | 6697/11704 [57:05<1:51:19,  1.33s/it]
 57%|█████▋    | 6698/11704 [57:06<1:30:20,  1.08s/it]
 57%|█████▋    | 6699/11704 [57:06<1:15:36,  1.10it/s]
 57%|█████▋    | 6700/11704 [57:07<1:05:58,  1.26it/s]{'loss': 3.4591, 'grad_norm': 0.23688948154449463, 'learning_rate': 0.0004608931070888296, 'epoch': 8.01}
-
                                                      
 57%|█████▋    | 6700/11704 [57:07<1:05:58,  1.26it/s]
 57%|█████▋    | 6701/11704 [57:07<58:38,  1.42it/s]  
 57%|█████▋    | 6702/11704 [57:08<53:25,  1.56it/s]
 57%|█████▋    | 6703/11704 [57:08<49:46,  1.67it/s]
 57%|█████▋    | 6704/11704 [57:09<47:11,  1.77it/s]
 57%|█████▋    | 6705/11704 [57:09<45:25,  1.83it/s]
 57%|█████▋    | 6706/11704 [57:10<44:10,  1.89it/s]
 57%|█████▋    | 6707/11704 [57:10<43:15,  1.93it/s]
 57%|█████▋    | 6708/11704 [57:11<42:36,  1.95it/s]
 57%|█████▋    | 6709/11704 [57:11<42:14,  1.97it/s]
 57%|█████▋    | 6710/11704 [57:12<42:01,  1.98it/s]
 57%|█████▋    | 6711/11704 [57:12<41:45,  1.99it/s]
 57%|█████▋    | 6712/11704 [57:13<41:40,  2.00it/s]
 57%|█████▋    | 6713/11704 [57:13<41:30,  2.00it/s]
 57%|█████▋    | 6714/11704 [57:14<41:25,  2.01it/s]
+
                                                      
 57%|█████▋    | 6700/11704 [57:07<1:05:58,  1.26it/s]
 57%|█████▋    | 6701/11704 [57:07<58:38,  1.42it/s]  
 57%|█████▋    | 6702/11704 [57:08<53:25,  1.56it/s]
 57%|█████▋    | 6703/11704 [57:08<49:46,  1.67it/s]
 57%|█████▋    | 6704/11704 [57:09<47:11,  1.77it/s]
 57%|█████▋    | 6705/11704 [57:09<45:25,  1.83it/s]
 57%|█████▋    | 6706/11704 [57:10<44:10,  1.89it/s]
 57%|█████▋    | 6707/11704 [57:10<43:15,  1.93it/s]
 57%|█████▋    | 6708/11704 [57:11<42:36,  1.95it/s]
 57%|█████▋    | 6709/11704 [57:11<42:14,  1.97it/s]
 57%|█████▋    | 6710/11704 [57:12<42:01,  1.98it/s]
 57%|█████▋    | 6711/11704 [57:12<41:45,  1.99it/s]
 57%|█████▋    | 6712/11704 [57:13<41:40,  2.00it/s]
 57%|█████▋    | 6713/11704 [57:13<41:30,  2.00it/s]
 57%|█████▋    | 6714/11704 [57:14<41:25,  2.01it/s]
 57%|█████▋    | 6715/11704 [57:14<41:19,  2.01it/s]
 57%|█████▋    | 6716/11704 [57:15<41:16,  2.01it/s]
 57%|█████▋    | 6717/11704 [57:15<41:13,  2.02it/s]
 57%|█████▋    | 6718/11704 [57:16<41:10,  2.02it/s]
 57%|█████▋    | 6719/11704 [57:16<41:12,  2.02it/s]
 57%|█████▋    | 6720/11704 [57:17<41:08,  2.02it/s]
 57%|█████▋    | 6721/11704 [57:17<41:10,  2.02it/s]
 57%|█████▋    | 6722/11704 [57:18<41:08,  2.02it/s]
 57%|█████▋    | 6723/11704 [57:18<41:10,  2.02it/s]
 57%|█████▋    | 6724/11704 [57:19<41:08,  2.02it/s]
 57%|█████▋    | 6725/11704 [57:19<41:08,  2.02it/s]{'loss': 3.3641, 'grad_norm': 0.22972285747528076, 'learning_rate': 0.0004571773760359283, 'epoch': 8.04}
+
                                                    
 57%|█████▋    | 6725/11704 [57:19<41:08,  2.02it/s]
 57%|█████▋    | 6726/11704 [57:20<41:07,  2.02it/s]
 57%|█████▋    | 6727/11704 [57:20<41:09,  2.02it/s]
 57%|█████▋    | 6728/11704 [57:21<41:06,  2.02it/s]
 57%|█████▋    | 6729/11704 [57:21<41:05,  2.02it/s]
 58%|█████▊    | 6730/11704 [57:21<41:04,  2.02it/s]
 58%|█████▊    | 6731/11704 [57:22<41:05,  2.02it/s]
 58%|█████▊    | 6732/11704 [57:22<41:04,  2.02it/s]
 58%|█████▊    | 6733/11704 [57:23<41:03,  2.02it/s]
 58%|█████▊    | 6734/11704 [57:23<41:05,  2.02it/s]
 58%|█████▊    | 6735/11704 [57:24<41:03,  2.02it/s]
 58%|█████▊    | 6736/11704 [57:24<41:02,  2.02it/s]
 58%|█████▊    | 6737/11704 [57:25<41:01,  2.02it/s]
 58%|█████▊    | 6738/11704 [57:25<41:01,  2.02it/s]
 58%|█████▊    | 6739/11704 [57:26<40:58,  2.02it/s]
 58%|█████▊    | 6740/11704 [57:26<41:02,  2.02it/s]
 58%|█████▊    | 6741/11704 [57:27<41:00,  2.02it/s]
 58%|█████▊    | 6742/11704 [57:27<40:59,  2.02it/s]
 58%|█████▊    | 6743/11704 [57:28<40:57,  2.02it/s]
 58%|█████▊    | 6744/11704 [57:28<40:59,  2.02it/s]
 58%|█████▊    | 6745/11704 [57:29<41:02,  2.01it/s]
 58%|█████▊    | 6746/11704 [57:29<40:59,  2.02it/s]
 58%|█████▊    | 6747/11704 [57:30<40:58,  2.02it/s]
 58%|█████▊    | 6748/11704 [57:30<40:56,  2.02it/s]
 58%|█████▊    | 6749/11704 [57:31<40:58,  2.02it/s]
 58%|█████▊    | 6750/11704 [57:31<41:01,  2.01it/s]{'loss': 3.3755, 'grad_norm': 0.227346271276474, 'learning_rate': 0.00045346402591428883, 'epoch': 8.07}
+
                                                    
 58%|█████▊    | 6750/11704 [57:31<41:01,  2.01it/s]
 58%|█████▊    | 6751/11704 [57:32<41:01,  2.01it/s]
 58%|█████▊    | 6752/11704 [57:32<41:02,  2.01it/s]
 58%|█████▊    | 6753/11704 [57:33<41:00,  2.01it/s]
 58%|█████▊    | 6754/11704 [57:33<41:01,  2.01it/s]
 58%|█████▊    | 6755/11704 [57:34<41:00,  2.01it/s]
 58%|█████▊    | 6756/11704 [57:34<40:56,  2.01it/s]
 58%|█████▊    | 6757/11704 [57:35<40:57,  2.01it/s]
 58%|█████▊    | 6758/11704 [57:35<40:54,  2.02it/s]
 58%|█████▊    | 6759/11704 [57:36<40:54,  2.01it/s]
 58%|█████▊    | 6760/11704 [57:36<40:49,  2.02it/s]
 58%|█████▊    | 6761/11704 [57:37<40:54,  2.01it/s]
 58%|█████▊    | 6762/11704 [57:37<40:51,  2.02it/s]
 58%|█████▊    | 6763/11704 [57:38<40:52,  2.01it/s]
 58%|█████▊    | 6764/11704 [57:38<40:51,  2.01it/s]
 58%|█████▊    | 6765/11704 [57:39<40:54,  2.01it/s]
 58%|█████▊    | 6766/11704 [57:39<40:57,  2.01it/s]
 58%|█████▊    | 6767/11704 [57:40<40:52,  2.01it/s]
 58%|█████▊    | 6768/11704 [57:40<40:51,  2.01it/s]
 58%|█████▊    | 6769/11704 [57:41<40:48,  2.02it/s]
 58%|█████▊    | 6770/11704 [57:41<40:48,  2.02it/s]
 58%|█████▊    | 6771/11704 [57:42<40:45,  2.02it/s]
 58%|█████▊    | 6772/11704 [57:42<40:45,  2.02it/s]
 58%|█████▊    | 6773/11704 [57:43<40:43,  2.02it/s]
 58%|█████▊    | 6774/11704 [57:43<40:44,  2.02it/s]
 58%|█████▊    | 6775/11704 [57:44<40:42,  2.02it/s]{'loss': 3.383, 'grad_norm': 0.2244180291891098, 'learning_rate': 0.00044975326318560303, 'epoch': 8.1}
+
                                                    
 58%|█████▊    | 6775/11704 [57:44<40:42,  2.02it/s]
 58%|█████▊    | 6776/11704 [57:44<40:44,  2.02it/s]
 58%|█████▊    | 6777/11704 [57:45<40:43,  2.02it/s]
 58%|█████▊    | 6778/11704 [57:45<40:41,  2.02it/s]
 58%|█████▊    | 6779/11704 [57:46<40:43,  2.02it/s]
 58%|█████▊    | 6780/11704 [57:46<40:39,  2.02it/s]
 58%|█████▊    | 6781/11704 [57:47<40:40,  2.02it/s]
 58%|█████▊    | 6782/11704 [57:47<40:35,  2.02it/s]
 58%|█████▊    | 6783/11704 [57:48<40:37,  2.02it/s]
 58%|█████▊    | 6784/11704 [57:48<40:36,  2.02it/s]
 58%|█████▊    | 6785/11704 [57:49<40:37,  2.02it/s]
 58%|█████▊    | 6786/11704 [57:49<40:35,  2.02it/s]
 58%|█████▊    | 6787/11704 [57:50<40:35,  2.02it/s]
 58%|█████▊    | 6788/11704 [57:50<40:32,  2.02it/s]
 58%|█████▊    | 6789/11704 [57:51<40:33,  2.02it/s]
 58%|█████▊    | 6790/11704 [57:51<40:30,  2.02it/s]
 58%|█████▊    | 6791/11704 [57:52<40:30,  2.02it/s]
 58%|█████▊    | 6792/11704 [57:52<40:29,  2.02it/s]
 58%|█████▊    | 6793/11704 [57:53<40:30,  2.02it/s]
 58%|█████▊    | 6794/11704 [57:53<40:28,  2.02it/s]
 58%|█████▊    | 6795/11704 [57:54<40:29,  2.02it/s]
 58%|█████▊    | 6796/11704 [57:54<40:29,  2.02it/s]
 58%|█████▊    | 6797/11704 [57:55<40:30,  2.02it/s]
 58%|█████▊    | 6798/11704 [57:55<40:32,  2.02it/s]
 58%|█████▊    | 6799/11704 [57:56<40:27,  2.02it/s]
 58%|█████▊    | 6800/11704 [57:56<40:30,  2.02it/s]{'loss': 3.3888, 'grad_norm': 0.22012494504451752, 'learning_rate': 0.0004460452941677038, 'epoch': 8.13}
+
                                                    
 58%|█████▊    | 6800/11704 [57:56<40:30,  2.02it/s]
 58%|█████▊    | 6801/11704 [57:57<40:29,  2.02it/s]
 58%|█████▊    | 6802/11704 [57:57<40:29,  2.02it/s]
 58%|█████▊    | 6803/11704 [57:58<40:24,  2.02it/s]
 58%|█████▊    | 6804/11704 [57:58<40:25,  2.02it/s]
 58%|█████▊    | 6805/11704 [57:59<40:26,  2.02it/s]
 58%|█████▊    | 6806/11704 [57:59<40:27,  2.02it/s]
 58%|█████▊    | 6807/11704 [58:00<40:23,  2.02it/s]
 58%|█████▊    | 6808/11704 [58:00<40:25,  2.02it/s]
 58%|█████▊    | 6809/11704 [58:01<40:24,  2.02it/s]
 58%|█████▊    | 6810/11704 [58:01<40:23,  2.02it/s]
 58%|█████▊    | 6811/11704 [58:02<40:19,  2.02it/s]
 58%|█████▊    | 6812/11704 [58:02<40:19,  2.02it/s]
 58%|█████▊    | 6813/11704 [58:03<40:20,  2.02it/s]
 58%|█████▊    | 6814/11704 [58:03<40:22,  2.02it/s]
 58%|█████▊    | 6815/11704 [58:04<40:22,  2.02it/s]
 58%|█████▊    | 6816/11704 [58:04<40:19,  2.02it/s]
 58%|█████▊    | 6817/11704 [58:05<40:23,  2.02it/s]
 58%|█████▊    | 6818/11704 [58:05<40:20,  2.02it/s]
 58%|█████▊    | 6819/11704 [58:06<40:19,  2.02it/s]
 58%|█████▊    | 6820/11704 [58:06<40:17,  2.02it/s]
 58%|█████▊    | 6821/11704 [58:07<40:16,  2.02it/s]
 58%|█████▊    | 6822/11704 [58:07<40:12,  2.02it/s]
 58%|█████▊    | 6823/11704 [58:08<40:14,  2.02it/s]
 58%|█████▊    | 6824/11704 [58:08<40:11,  2.02it/s]
 58%|█████▊    | 6825/11704 [58:09<40:13,  2.02it/s]{'loss': 3.3934, 'grad_norm': 0.23053328692913055, 'learning_rate': 0.00044234032502309464, 'epoch': 8.16}
                                                    
+
 58%|█████▊    | 6825/11704 [58:09<40:13,  2.02it/s]
 58%|█████▊    | 6826/11704 [58:09<40:18,  2.02it/s]
 58%|█████▊    | 6827/11704 [58:10<40:15,  2.02it/s]
 58%|█████▊    | 6828/11704 [58:10<40:11,  2.02it/s]
 58%|█████▊    | 6829/11704 [58:11<40:11,  2.02it/s]
 58%|█████▊    | 6830/11704 [58:11<40:09,  2.02it/s]
 58%|█████▊    | 6831/11704 [58:12<40:10,  2.02it/s]
 58%|█████▊    | 6832/11704 [58:12<40:07,  2.02it/s]
 58%|█████▊    | 6833/11704 [58:13<40:08,  2.02it/s]
 58%|█████▊    | 6834/11704 [58:13<40:06,  2.02it/s]
 58%|█████▊    | 6835/11704 [58:14<40:08,  2.02it/s]
 58%|█████▊    | 6836/11704 [58:14<40:05,  2.02it/s]
 58%|█████▊    | 6837/11704 [58:15<40:04,  2.02it/s]
 58%|█████▊    | 6838/11704 [58:15<40:04,  2.02it/s]
 58%|█████▊    | 6839/11704 [58:16<40:06,  2.02it/s]
 58%|█████▊    | 6840/11704 [58:16<40:05,  2.02it/s]
 58%|█████▊    | 6841/11704 [58:16<40:06,  2.02it/s]
 58%|█████▊    | 6842/11704 [58:17<40:02,  2.02it/s]
 58%|█████▊    | 6843/11704 [58:17<40:02,  2.02it/s]
 58%|█████▊    | 6844/11704 [58:18<40:02,  2.02it/s]
 58%|█████▊    | 6845/11704 [58:18<40:01,  2.02it/s]
 58%|█████▊    | 6846/11704 [58:19<40:00,  2.02it/s]
 59%|█████▊    | 6847/11704 [58:19<39:59,  2.02it/s]
 59%|█████▊    | 6848/11704 [58:20<40:00,  2.02it/s]
 59%|█████▊    | 6849/11704 [58:20<39:59,  2.02it/s]
 59%|█████▊    | 6850/11704 [58:21<39:59,  2.02it/s]{'loss': 3.3891, 'grad_norm': 0.23988386988639832, 'learning_rate': 0.0004386385617474864, 'epoch': 8.19}
+
                                                    
 59%|█████▊    | 6850/11704 [58:21<39:59,  2.02it/s]
 59%|█████▊    | 6851/11704 [58:21<40:02,  2.02it/s]
 59%|█████▊    | 6852/11704 [58:22<40:04,  2.02it/s]
 59%|█████▊    | 6853/11704 [58:22<40:02,  2.02it/s]
 59%|█████▊    | 6854/11704 [58:23<40:00,  2.02it/s]
 59%|█████▊    | 6855/11704 [58:23<39:57,  2.02it/s]
 59%|█████▊    | 6856/11704 [58:24<39:59,  2.02it/s]
 59%|█████▊    | 6857/11704 [58:24<39:56,  2.02it/s]
 59%|█████▊    | 6858/11704 [58:25<39:56,  2.02it/s]
 59%|█████▊    | 6859/11704 [58:25<39:54,  2.02it/s]
 59%|█████▊    | 6860/11704 [58:26<39:53,  2.02it/s]
 59%|█████▊    | 6861/11704 [58:26<39:54,  2.02it/s]
 59%|█████▊    | 6862/11704 [58:27<39:55,  2.02it/s]
 59%|█████▊    | 6863/11704 [58:27<39:52,  2.02it/s]
 59%|█████▊    | 6864/11704 [58:28<39:52,  2.02it/s]
 59%|█████▊    | 6865/11704 [58:28<39:50,  2.02it/s]
 59%|█████▊    | 6866/11704 [58:29<39:48,  2.03it/s]
 59%|█████▊    | 6867/11704 [58:29<39:49,  2.02it/s]
 59%|█████▊    | 6868/11704 [58:30<39:47,  2.03it/s]
 59%|█████▊    | 6869/11704 [58:30<39:48,  2.02it/s]
 59%|█████▊    | 6870/11704 [58:31<39:45,  2.03it/s]
 59%|█████▊    | 6871/11704 [58:31<39:48,  2.02it/s]
 59%|█████▊    | 6872/11704 [58:32<39:49,  2.02it/s]
 59%|█████▊    | 6873/11704 [58:32<39:51,  2.02it/s]
 59%|█████▊    | 6874/11704 [58:33<39:48,  2.02it/s]
 59%|█████▊    | 6875/11704 [58:33<39:49,  2.02it/s]{'loss': 3.3966, 'grad_norm': 0.23840314149856567, 'learning_rate': 0.0004349402101583436, 'epoch': 8.22}
+
                                                    
 59%|█████▊    | 6875/11704 [58:33<39:49,  2.02it/s]
 59%|█████▊    | 6876/11704 [58:34<39:49,  2.02it/s]
 59%|█████▉    | 6877/11704 [58:34<39:48,  2.02it/s]
 59%|█████▉    | 6878/11704 [58:35<39:45,  2.02it/s]
 59%|█████▉    | 6879/11704 [58:35<39:45,  2.02it/s]
 59%|█████▉    | 6880/11704 [58:36<39:43,  2.02it/s]
 59%|█████▉    | 6881/11704 [58:36<39:44,  2.02it/s]
 59%|█████▉    | 6882/11704 [58:37<39:41,  2.02it/s]
 59%|█████▉    | 6883/11704 [58:37<39:46,  2.02it/s]
 59%|█████▉    | 6884/11704 [58:38<39:43,  2.02it/s]
 59%|█████▉    | 6885/11704 [58:38<39:44,  2.02it/s]
 59%|█████▉    | 6886/11704 [58:39<39:43,  2.02it/s]
 59%|█████▉    | 6887/11704 [58:39<39:41,  2.02it/s]
 59%|█████▉    | 6888/11704 [58:40<39:38,  2.02it/s]
 59%|█████▉    | 6889/11704 [58:40<39:41,  2.02it/s]
 59%|█████▉    | 6890/11704 [58:41<39:39,  2.02it/s]
 59%|█████▉    | 6891/11704 [58:41<39:37,  2.02it/s]
 59%|█████▉    | 6892/11704 [58:42<39:38,  2.02it/s]
 59%|█████▉    | 6893/11704 [58:42<39:36,  2.02it/s]
 59%|█████▉    | 6894/11704 [58:43<39:38,  2.02it/s]
 59%|█████▉    | 6895/11704 [58:43<39:37,  2.02it/s]
 59%|█████▉    | 6896/11704 [58:44<39:38,  2.02it/s]
 59%|█████▉    | 6897/11704 [58:44<39:35,  2.02it/s]
 59%|█████▉    | 6898/11704 [58:45<39:36,  2.02it/s]
 59%|█████▉    | 6899/11704 [58:45<39:35,  2.02it/s]
 59%|█████▉    | 6900/11704 [58:46<39:35,  2.02it/s]{'loss': 3.4031, 'grad_norm': 0.23334480822086334, 'learning_rate': 0.0004312454758834423, 'epoch': 8.25}
+
                                                    
 59%|█████▉    | 6900/11704 [58:46<39:35,  2.02it/s]
 59%|█████▉    | 6901/11704 [58:46<39:38,  2.02it/s]
 59%|█████▉    | 6902/11704 [58:47<39:37,  2.02it/s]
 59%|█████▉    | 6903/11704 [58:47<39:36,  2.02it/s]
 59%|█████▉    | 6904/11704 [58:48<39:35,  2.02it/s]
 59%|█████▉    | 6905/11704 [58:48<39:36,  2.02it/s]
 59%|█████▉    | 6906/11704 [58:49<39:34,  2.02it/s]
 59%|█████▉    | 6907/11704 [58:49<39:33,  2.02it/s]
 59%|█████▉    | 6908/11704 [58:50<39:32,  2.02it/s]
 59%|█████▉    | 6909/11704 [58:50<39:34,  2.02it/s]
 59%|█████▉    | 6910/11704 [58:51<39:32,  2.02it/s]
 59%|█████▉    | 6911/11704 [58:51<39:33,  2.02it/s]
 59%|█████▉    | 6912/11704 [58:52<39:30,  2.02it/s]
 59%|█████▉    | 6913/11704 [58:52<39:32,  2.02it/s]
 59%|█████▉    | 6914/11704 [58:53<39:30,  2.02it/s]
 59%|█████▉    | 6915/11704 [58:53<39:31,  2.02it/s]
 59%|█████▉    | 6916/11704 [58:54<39:28,  2.02it/s]
 59%|█████▉    | 6917/11704 [58:54<39:29,  2.02it/s]
 59%|█████▉    | 6918/11704 [58:55<39:27,  2.02it/s]
 59%|█████▉    | 6919/11704 [58:55<39:27,  2.02it/s]
 59%|█████▉    | 6920/11704 [58:56<39:26,  2.02it/s]
 59%|█████▉    | 6921/11704 [58:56<39:26,  2.02it/s]
 59%|█████▉    | 6922/11704 [58:57<39:25,  2.02it/s]
 59%|█████▉    | 6923/11704 [58:57<39:25,  2.02it/s]
 59%|█████▉    | 6924/11704 [58:58<39:23,  2.02it/s]
 59%|█████▉    | 6925/11704 [58:58<39:26,  2.02it/s]{'loss': 3.4072, 'grad_norm': 0.23080752789974213, 'learning_rate': 0.0004275545643494362, 'epoch': 8.28}
+
                                                    
 59%|█████▉    | 6925/11704 [58:58<39:26,  2.02it/s]
 59%|█████▉    | 6926/11704 [58:59<39:28,  2.02it/s]
 59%|█████▉    | 6927/11704 [58:59<39:25,  2.02it/s]
 59%|█████▉    | 6928/11704 [59:00<39:24,  2.02it/s]
 59%|█████▉    | 6929/11704 [59:00<39:24,  2.02it/s]
 59%|█████▉    | 6930/11704 [59:01<39:22,  2.02it/s]
 59%|█████▉    | 6931/11704 [59:01<39:22,  2.02it/s]
 59%|█████▉    | 6932/11704 [59:01<39:19,  2.02it/s]
 59%|█████▉    | 6933/11704 [59:02<39:19,  2.02it/s]
 59%|█████▉    | 6934/11704 [59:02<39:19,  2.02it/s]
 59%|█████▉    | 6935/11704 [59:03<39:19,  2.02it/s]
 59%|█████▉    | 6936/11704 [59:03<39:19,  2.02it/s]
 59%|█████▉    | 6937/11704 [59:04<39:19,  2.02it/s]
 59%|█████▉    | 6938/11704 [59:04<39:17,  2.02it/s]
 59%|█████▉    | 6939/11704 [59:05<39:17,  2.02it/s]
 59%|█████▉    | 6940/11704 [59:05<39:14,  2.02it/s]
 59%|█████▉    | 6941/11704 [59:06<39:16,  2.02it/s]
 59%|█████▉    | 6942/11704 [59:06<39:13,  2.02it/s]
 59%|█████▉    | 6943/11704 [59:07<39:14,  2.02it/s]
 59%|█████▉    | 6944/11704 [59:07<39:11,  2.02it/s]
 59%|█████▉    | 6945/11704 [59:08<39:11,  2.02it/s]
 59%|█████▉    | 6946/11704 [59:08<39:09,  2.02it/s]
 59%|█████▉    | 6947/11704 [59:09<39:12,  2.02it/s]
 59%|█████▉    | 6948/11704 [59:09<39:12,  2.02it/s]
 59%|█████▉    | 6949/11704 [59:10<39:12,  2.02it/s]
 59%|█████▉    | 6950/11704 [59:10<39:11,  2.02it/s]{'loss': 3.405, 'grad_norm': 0.2382190078496933, 'learning_rate': 0.0004238676807704347, 'epoch': 8.31}
+
                                                    
 59%|█████▉    | 6950/11704 [59:10<39:11,  2.02it/s]
 59%|█████▉    | 6951/11704 [59:11<39:17,  2.02it/s]
 59%|█████▉    | 6952/11704 [59:11<39:17,  2.02it/s]
 59%|█████▉    | 6953/11704 [59:12<39:13,  2.02it/s]
 59%|█████▉    | 6954/11704 [59:12<39:13,  2.02it/s]
 59%|█████▉    | 6955/11704 [59:13<39:09,  2.02it/s]
 59%|█████▉    | 6956/11704 [59:13<39:09,  2.02it/s]
 59%|█████▉    | 6957/11704 [59:14<39:07,  2.02it/s]
 59%|█████▉    | 6958/11704 [59:14<39:09,  2.02it/s]
 59%|█████▉    | 6959/11704 [59:15<39:06,  2.02it/s]
 59%|█████▉    | 6960/11704 [59:15<39:05,  2.02it/s]
 59%|█████▉    | 6961/11704 [59:16<39:04,  2.02it/s]
 59%|█████▉    | 6962/11704 [59:16<39:04,  2.02it/s]
 59%|█████▉    | 6963/11704 [59:17<39:03,  2.02it/s]
 60%|█████▉    | 6964/11704 [59:17<39:01,  2.02it/s]
 60%|█████▉    | 6965/11704 [59:18<39:01,  2.02it/s]
 60%|█████▉    | 6966/11704 [59:18<39:01,  2.02it/s]
 60%|█████▉    | 6967/11704 [59:19<39:00,  2.02it/s]
 60%|█████▉    | 6968/11704 [59:19<39:03,  2.02it/s]
 60%|█████▉    | 6969/11704 [59:20<39:02,  2.02it/s]
 60%|█████▉    | 6970/11704 [59:20<38:59,  2.02it/s]
 60%|█████▉    | 6971/11704 [59:21<39:01,  2.02it/s]
 60%|█████▉    | 6972/11704 [59:21<38:59,  2.02it/s]
 60%|█████▉    | 6973/11704 [59:22<39:00,  2.02it/s]
 60%|█████▉    | 6974/11704 [59:22<38:57,  2.02it/s]
 60%|█████▉    | 6975/11704 [59:23<42:26,  1.86it/s]{'loss': 3.409, 'grad_norm': 0.22229425609111786, 'learning_rate': 0.0004201850301365941, 'epoch': 8.34}
+
                                                    
 60%|█████▉    | 6975/11704 [59:23<42:26,  1.86it/s]
 60%|█████▉    | 6976/11704 [59:23<41:27,  1.90it/s]
 60%|█████▉    | 6977/11704 [59:24<40:40,  1.94it/s]
 60%|█████▉    | 6978/11704 [59:24<40:10,  1.96it/s]
 60%|█████▉    | 6979/11704 [59:25<39:48,  1.98it/s]
 60%|█████▉    | 6980/11704 [59:25<39:31,  1.99it/s]
 60%|█████▉    | 6981/11704 [59:26<39:17,  2.00it/s]
 60%|█████▉    | 6982/11704 [59:26<39:10,  2.01it/s]
 60%|█████▉    | 6983/11704 [59:27<39:06,  2.01it/s]
 60%|█████▉    | 6984/11704 [59:27<39:00,  2.02it/s]
 60%|█████▉    | 6985/11704 [59:28<38:58,  2.02it/s]
 60%|█████▉    | 6986/11704 [59:28<38:52,  2.02it/s]
 60%|█████▉    | 6987/11704 [59:29<38:53,  2.02it/s]
 60%|█████▉    | 6988/11704 [59:29<38:49,  2.02it/s]
 60%|█████▉    | 6989/11704 [59:30<38:49,  2.02it/s]
 60%|█████▉    | 6990/11704 [59:30<38:48,  2.02it/s]
 60%|█████▉    | 6991/11704 [59:31<38:49,  2.02it/s]
 60%|█████▉    | 6992/11704 [59:31<38:48,  2.02it/s]
 60%|█████▉    | 6993/11704 [59:32<38:48,  2.02it/s]
 60%|█████▉    | 6994/11704 [59:32<38:48,  2.02it/s]
 60%|█████▉    | 6995/11704 [59:33<38:47,  2.02it/s]
 60%|█████▉    | 6996/11704 [59:33<38:46,  2.02it/s]
 60%|█████▉    | 6997/11704 [59:34<38:43,  2.03it/s]
 60%|█████▉    | 6998/11704 [59:34<38:43,  2.03it/s]
 60%|█████▉    | 6999/11704 [59:35<38:42,  2.03it/s]
 60%|█████▉    | 7000/11704 [59:35<38:44,  2.02it/s]
                                                    
+
 60%|█████▉    | 7000/11704 [59:35<38:44,  2.02it/s]
 60%|█████▉    | 7001/11704 [59:36<38:48,  2.02it/s]
 60%|█████▉    | 7002/11704 [59:36<38:48,  2.02it/s]
 60%|█████▉    | 7003/11704 [59:37<38:44,  2.02it/s]
 60%|█████▉    | 7004/11704 [59:37<38:43,  2.02it/s]
 60%|█████▉    | 7005/11704 [59:38<38:42,  2.02it/s]
 60%|█████▉    | 7006/11704 [59:38<38:40,  2.02it/s]
 60%|█████▉    | 7007/11704 [59:39<38:40,  2.02it/s]
 60%|█████▉    | 7008/11704 [59:39<38:39,  2.02it/s]
 60%|█████▉    | 7009/11704 [59:40<38:38,  2.02it/s]
 60%|█████▉    | 7010/11704 [59:40<38:39,  2.02it/s]
 60%|█████▉    | 7011/11704 [59:41<38:42,  2.02it/s]
 60%|█████▉    | 7012/11704 [59:41<38:38,  2.02it/s]
 60%|█████▉    | 7013/11704 [59:42<38:39,  2.02it/s]
 60%|█████▉    | 7014/11704 [59:42<38:38,  2.02it/s]
 60%|█████▉    | 7015/11704 [59:43<38:37,  2.02it/s]
 60%|█████▉    | 7016/11704 [59:43<38:35,  2.02it/s]
 60%|█████▉    | 7017/11704 [59:44<38:35,  2.02it/s]
 60%|█████▉    | 7018/11704 [59:44<38:36,  2.02it/s]
 60%|█████▉    | 7019/11704 [59:45<38:36,  2.02it/s]
 60%|█████▉    | 7020/11704 [59:45<38:37,  2.02it/s]
 60%|█████▉    | 7021/11704 [59:46<38:38,  2.02it/s]
 60%|█████▉    | 7022/11704 [59:46<38:38,  2.02it/s]
 60%|██████    | 7023/11704 [59:47<38:36,  2.02it/s]
 60%|██████    | 7024/11704 [59:47<38:33,  2.02it/s]
 60%|██████    | 7025/11704 [59:48<38:32,  2.02it/s]{'loss': 3.4131, 'grad_norm': 0.22735567390918732, 'learning_rate': 0.0004128332464768806, 'epoch': 8.4}
                                                    
+
 60%|██████    | 7025/11704 [59:48<38:32,  2.02it/s]
 60%|██████    | 7026/11704 [59:48<38:33,  2.02it/s]
 60%|██████    | 7027/11704 [59:49<38:35,  2.02it/s]
 60%|██████    | 7028/11704 [59:49<38:33,  2.02it/s]
 60%|██████    | 7029/11704 [59:50<38:31,  2.02it/s]
 60%|██████    | 7030/11704 [59:50<38:31,  2.02it/s]
 60%|██████    | 7031/11704 [59:51<38:31,  2.02it/s]
 60%|██████    | 7032/11704 [59:51<38:33,  2.02it/s]
 60%|██████    | 7033/11704 [59:52<38:29,  2.02it/s]
 60%|██████    | 7034/11704 [59:52<38:30,  2.02it/s]
 60%|██████    | 7035/11704 [59:53<38:27,  2.02it/s]
 60%|██████    | 7036/11704 [59:53<38:26,  2.02it/s]
 60%|██████    | 7037/11704 [59:54<38:25,  2.02it/s]
 60%|██████    | 7038/11704 [59:54<38:25,  2.02it/s]
 60%|██████    | 7039/11704 [59:55<38:26,  2.02it/s]
 60%|██████    | 7040/11704 [59:55<38:23,  2.02it/s]
 60%|██████    | 7041/11704 [59:56<41:42,  1.86it/s]
 60%|██████    | 7042/11704 [59:56<40:41,  1.91it/s]
 60%|██████    | 7043/11704 [59:57<40:00,  1.94it/s]
 60%|██████    | 7044/11704 [59:57<39:29,  1.97it/s]
 60%|██████    | 7045/11704 [59:58<39:09,  1.98it/s]
 60%|██████    | 7046/11704 [59:58<38:54,  1.99it/s]
 60%|██████    | 7047/11704 [59:59<38:44,  2.00it/s]
 60%|██████    | 7048/11704 [59:59<38:37,  2.01it/s]
 60%|██████    | 7049/11704 [1:00:00<38:29,  2.02it/s]
 60%|██████    | 7050/11704 [1:00:00<38:29,  2.02it/s]{'loss': 3.4196, 'grad_norm': 0.22713714838027954, 'learning_rate': 0.00040916452220904066, 'epoch': 8.43}
+
                                                      
 60%|██████    | 7050/11704 [1:00:00<38:29,  2.02it/s]
 60%|██████    | 7051/11704 [1:00:01<38:28,  2.02it/s]
 60%|██████    | 7052/11704 [1:00:01<38:25,  2.02it/s]
 60%|██████    | 7053/11704 [1:00:02<38:25,  2.02it/s]
 60%|██████    | 7054/11704 [1:00:02<38:22,  2.02it/s]
 60%|██████    | 7055/11704 [1:00:03<38:20,  2.02it/s]
 60%|██████    | 7056/11704 [1:00:03<38:20,  2.02it/s]
 60%|██████    | 7057/11704 [1:00:04<38:21,  2.02it/s]
 60%|██████    | 7058/11704 [1:00:04<38:20,  2.02it/s]
 60%|██████    | 7059/11704 [1:00:05<38:18,  2.02it/s]
 60%|██████    | 7060/11704 [1:00:05<38:19,  2.02it/s]
 60%|██████    | 7061/11704 [1:00:06<38:18,  2.02it/s]
 60%|██████    | 7062/11704 [1:00:06<38:16,  2.02it/s]
 60%|██████    | 7063/11704 [1:00:07<38:14,  2.02it/s]
 60%|██████    | 7064/11704 [1:00:07<38:16,  2.02it/s]
 60%|██████    | 7065/11704 [1:00:08<38:13,  2.02it/s]
 60%|██████    | 7066/11704 [1:00:08<38:13,  2.02it/s]
 60%|██████    | 7067/11704 [1:00:09<38:13,  2.02it/s]
 60%|██████    | 7068/11704 [1:00:09<38:14,  2.02it/s]
 60%|██████    | 7069/11704 [1:00:10<38:12,  2.02it/s]
 60%|██████    | 7070/11704 [1:00:10<38:11,  2.02it/s]
 60%|██████    | 7071/11704 [1:00:11<38:11,  2.02it/s]
 60%|██████    | 7072/11704 [1:00:11<38:10,  2.02it/s]
 60%|██████    | 7073/11704 [1:00:12<38:08,  2.02it/s]
 60%|██████    | 7074/11704 [1:00:12<38:10,  2.02it/s]
 60%|██████    | 7075/11704 [1:00:12<38:09,  2.02it/s]{'loss': 3.42, 'grad_norm': 0.2443581074476242, 'learning_rate': 0.0004055008483797011, 'epoch': 8.46}
+
                                                      
 60%|██████    | 7075/11704 [1:00:13<38:09,  2.02it/s]
 60%|██████    | 7076/11704 [1:00:13<38:09,  2.02it/s]
 60%|██████    | 7077/11704 [1:00:13<38:06,  2.02it/s]
 60%|██████    | 7078/11704 [1:00:14<38:07,  2.02it/s]
 60%|██████    | 7079/11704 [1:00:14<38:05,  2.02it/s]
 60%|██████    | 7080/11704 [1:00:15<38:06,  2.02it/s]
 61%|██████    | 7081/11704 [1:00:15<38:04,  2.02it/s]
 61%|██████    | 7082/11704 [1:00:16<38:07,  2.02it/s]
 61%|██████    | 7083/11704 [1:00:16<38:05,  2.02it/s]
 61%|██████    | 7084/11704 [1:00:17<38:05,  2.02it/s]
 61%|██████    | 7085/11704 [1:00:17<38:02,  2.02it/s]
 61%|██████    | 7086/11704 [1:00:18<38:05,  2.02it/s]
 61%|██████    | 7087/11704 [1:00:18<38:03,  2.02it/s]
 61%|██████    | 7088/11704 [1:00:19<38:02,  2.02it/s]
 61%|██████    | 7089/11704 [1:00:19<38:00,  2.02it/s]
 61%|██████    | 7090/11704 [1:00:20<38:01,  2.02it/s]
 61%|██████    | 7091/11704 [1:00:20<37:59,  2.02it/s]
 61%|██████    | 7092/11704 [1:00:21<38:00,  2.02it/s]
 61%|██████    | 7093/11704 [1:00:21<37:56,  2.03it/s]
 61%|██████    | 7094/11704 [1:00:22<37:59,  2.02it/s]
 61%|���█████    | 7095/11704 [1:00:22<37:58,  2.02it/s]
 61%|██████    | 7096/11704 [1:00:23<38:00,  2.02it/s]
 61%|██████    | 7097/11704 [1:00:23<37:56,  2.02it/s]
 61%|██████    | 7098/11704 [1:00:24<37:58,  2.02it/s]
 61%|██████    | 7099/11704 [1:00:24<37:55,  2.02it/s]
 61%|██████    | 7100/11704 [1:00:25<37:57,  2.02it/s]{'loss': 3.4159, 'grad_norm': 0.22324584424495697, 'learning_rate': 0.00040184242868855944, 'epoch': 8.49}
+
                                                      
 61%|██████    | 7100/11704 [1:00:25<37:57,  2.02it/s]
 61%|██████    | 7101/11704 [1:00:25<37:58,  2.02it/s]
 61%|██████    | 7102/11704 [1:00:26<37:59,  2.02it/s]
 61%|██████    | 7103/11704 [1:00:26<37:55,  2.02it/s]
 61%|██████    | 7104/11704 [1:00:27<37:56,  2.02it/s]
 61%|██████    | 7105/11704 [1:00:27<37:54,  2.02it/s]
 61%|██████    | 7106/11704 [1:00:28<37:55,  2.02it/s]
 61%|██████    | 7107/11704 [1:00:28<37:52,  2.02it/s]
 61%|██████    | 7108/11704 [1:00:29<37:53,  2.02it/s]
 61%|██████    | 7109/11704 [1:00:29<37:52,  2.02it/s]
 61%|██████    | 7110/11704 [1:00:30<37:53,  2.02it/s]
 61%|██████    | 7111/11704 [1:00:30<37:48,  2.02it/s]
 61%|██████    | 7112/11704 [1:00:31<37:51,  2.02it/s]
 61%|██████    | 7113/11704 [1:00:31<37:49,  2.02it/s]
 61%|██████    | 7114/11704 [1:00:32<37:50,  2.02it/s]
 61%|██████    | 7115/11704 [1:00:32<37:47,  2.02it/s]
 61%|██████    | 7116/11704 [1:00:33<37:48,  2.02it/s]
 61%|██████    | 7117/11704 [1:00:33<37:49,  2.02it/s]
 61%|██████    | 7118/11704 [1:00:34<37:48,  2.02it/s]
 61%|██████    | 7119/11704 [1:00:34<37:47,  2.02it/s]
 61%|██████    | 7120/11704 [1:00:35<37:47,  2.02it/s]
 61%|██████    | 7121/11704 [1:00:35<37:46,  2.02it/s]
 61%|██████    | 7122/11704 [1:00:36<37:47,  2.02it/s]
 61%|██████    | 7123/11704 [1:00:36<37:45,  2.02it/s]
 61%|██████    | 7124/11704 [1:00:37<37:43,  2.02it/s]
 61%|██████    | 7125/11704 [1:00:37<37:43,  2.02it/s]{'loss': 3.4211, 'grad_norm': 0.23106643557548523, 'learning_rate': 0.00039818946654318404, 'epoch': 8.52}
+
                                                      
 61%|██████    | 7125/11704 [1:00:37<37:43,  2.02it/s]
 61%|██████    | 7126/11704 [1:00:38<37:45,  2.02it/s]
 61%|██████    | 7127/11704 [1:00:38<37:43,  2.02it/s]
 61%|██████    | 7128/11704 [1:00:39<37:48,  2.02it/s]
 61%|██████    | 7129/11704 [1:00:39<37:45,  2.02it/s]
 61%|██████    | 7130/11704 [1:00:40<37:43,  2.02it/s]
 61%|██████    | 7131/11704 [1:00:40<37:42,  2.02it/s]
 61%|██████    | 7132/11704 [1:00:41<37:42,  2.02it/s]
 61%|██████    | 7133/11704 [1:00:41<37:40,  2.02it/s]
 61%|██████    | 7134/11704 [1:00:42<37:40,  2.02it/s]
 61%|██████    | 7135/11704 [1:00:42<37:41,  2.02it/s]
 61%|██████    | 7136/11704 [1:00:43<37:39,  2.02it/s]
 61%|██████    | 7137/11704 [1:00:43<37:39,  2.02it/s]
 61%|██████    | 7138/11704 [1:00:44<37:40,  2.02it/s]
 61%|██████    | 7139/11704 [1:00:44<37:38,  2.02it/s]
 61%|██████    | 7140/11704 [1:00:45<37:37,  2.02it/s]
 61%|██████    | 7141/11704 [1:00:45<37:36,  2.02it/s]
 61%|██████    | 7142/11704 [1:00:46<37:35,  2.02it/s]
 61%|██████    | 7143/11704 [1:00:46<37:35,  2.02it/s]
 61%|██████    | 7144/11704 [1:00:47<37:34,  2.02it/s]
 61%|██████    | 7145/11704 [1:00:47<37:32,  2.02it/s]
 61%|██████    | 7146/11704 [1:00:48<37:33,  2.02it/s]
 61%|██████    | 7147/11704 [1:00:48<37:33,  2.02it/s]
 61%|██████    | 7148/11704 [1:00:49<37:32,  2.02it/s]
 61%|██████    | 7149/11704 [1:00:49<37:34,  2.02it/s]
 61%|██████    | 7150/11704 [1:00:50<37:33,  2.02it/s]{'loss': 3.4232, 'grad_norm': 0.22898118197917938, 'learning_rate': 0.00039454216504770484, 'epoch': 8.55}
+
                                                      
 61%|██████    | 7150/11704 [1:00:50<37:33,  2.02it/s]
 61%|██████    | 7151/11704 [1:00:50<37:33,  2.02it/s]
 61%|██████    | 7152/11704 [1:00:51<37:32,  2.02it/s]
 61%|██████    | 7153/11704 [1:00:51<37:31,  2.02it/s]
 61%|██████    | 7154/11704 [1:00:52<37:28,  2.02it/s]
 61%|██████    | 7155/11704 [1:00:52<37:30,  2.02it/s]
 61%|██████    | 7156/11704 [1:00:53<37:29,  2.02it/s]
 61%|██████    | 7157/11704 [1:00:53<37:31,  2.02it/s]
 61%|██████    | 7158/11704 [1:00:54<37:28,  2.02it/s]
 61%|██████    | 7159/11704 [1:00:54<37:28,  2.02it/s]
 61%|██████    | 7160/11704 [1:00:55<37:28,  2.02it/s]
 61%|██████    | 7161/11704 [1:00:55<37:28,  2.02it/s]
 61%|██████    | 7162/11704 [1:00:56<37:27,  2.02it/s]
 61%|██████    | 7163/11704 [1:00:56<37:27,  2.02it/s]
 61%|██████    | 7164/11704 [1:00:57<37:26,  2.02it/s]
 61%|██████    | 7165/11704 [1:00:57<37:27,  2.02it/s]
 61%|██████    | 7166/11704 [1:00:58<37:25,  2.02it/s]
 61%|██████    | 7167/11704 [1:00:58<37:24,  2.02it/s]
 61%|██████    | 7168/11704 [1:00:58<37:24,  2.02it/s]
 61%|██████▏   | 7169/11704 [1:00:59<37:21,  2.02it/s]
 61%|██████▏   | 7170/11704 [1:00:59<37:22,  2.02it/s]
 61%|██████▏   | 7171/11704 [1:01:00<37:19,  2.02it/s]
 61%|██████▏   | 7172/11704 [1:01:00<37:21,  2.02it/s]
 61%|██████▏   | 7173/11704 [1:01:01<37:19,  2.02it/s]
 61%|██████▏   | 7174/11704 [1:01:01<37:19,  2.02it/s]
 61%|██████▏   | 7175/11704 [1:01:02<37:18,  2.02it/s]{'loss': 3.4352, 'grad_norm': 0.23332154750823975, 'learning_rate': 0.0003909007269915198, 'epoch': 8.58}
+
                                                      
 61%|██████▏   | 7175/11704 [1:01:02<37:18,  2.02it/s]
 61%|██████▏   | 7176/11704 [1:01:02<37:19,  2.02it/s]
 61%|██████▏   | 7177/11704 [1:01:03<37:18,  2.02it/s]
 61%|██████▏   | 7178/11704 [1:01:03<37:17,  2.02it/s]
 61%|██████▏   | 7179/11704 [1:01:04<37:19,  2.02it/s]
 61%|██████▏   | 7180/11704 [1:01:04<37:15,  2.02it/s]
 61%|██████▏   | 7181/11704 [1:01:05<37:18,  2.02it/s]
 61%|██████▏   | 7182/11704 [1:01:05<37:16,  2.02it/s]
 61%|██████▏   | 7183/11704 [1:01:06<37:16,  2.02it/s]
 61%|██████▏   | 7184/11704 [1:01:06<37:14,  2.02it/s]
 61%|██████▏   | 7185/11704 [1:01:07<37:14,  2.02it/s]
 61%|██████▏   | 7186/11704 [1:01:07<37:12,  2.02it/s]
 61%|██████▏   | 7187/11704 [1:01:08<37:12,  2.02it/s]
 61%|██████▏   | 7188/11704 [1:01:08<37:10,  2.03it/s]
 61%|██████▏   | 7189/11704 [1:01:09<37:11,  2.02it/s]
 61%|██████▏   | 7190/11704 [1:01:09<37:10,  2.02it/s]
 61%|██████▏   | 7191/11704 [1:01:10<37:10,  2.02it/s]
 61%|██████▏   | 7192/11704 [1:01:10<37:11,  2.02it/s]
 61%|██████▏   | 7193/11704 [1:01:11<37:13,  2.02it/s]
 61%|██████▏   | 7194/11704 [1:01:11<37:10,  2.02it/s]
 61%|██████▏   | 7195/11704 [1:01:12<37:09,  2.02it/s]
 61%|██████▏   | 7196/11704 [1:01:12<37:08,  2.02it/s]
 61%|██████▏   | 7197/11704 [1:01:13<37:07,  2.02it/s]
 62%|██████▏   | 7198/11704 [1:01:13<37:07,  2.02it/s]
 62%|██████▏   | 7199/11704 [1:01:14<37:07,  2.02it/s]
 62%|██████▏   | 7200/11704 [1:01:14<37:07,  2.02it/s]{'loss': 3.4373, 'grad_norm': 0.22517715394496918, 'learning_rate': 0.00038726535483802124, 'epoch': 8.61}
+
                                                      
 62%|██████▏   | 7200/11704 [1:01:14<37:07,  2.02it/s]
 62%|██████▏   | 7201/11704 [1:01:15<37:11,  2.02it/s]
 62%|██████▏   | 7202/11704 [1:01:15<37:09,  2.02it/s]
 62%|██████▏   | 7203/11704 [1:01:16<37:08,  2.02it/s]
 62%|██████▏   | 7204/11704 [1:01:16<37:06,  2.02it/s]
 62%|██████▏   | 7205/11704 [1:01:17<37:04,  2.02it/s]
 62%|██████▏   | 7206/11704 [1:01:17<37:03,  2.02it/s]
 62%|██████▏   | 7207/11704 [1:01:18<37:02,  2.02it/s]
 62%|██████▏   | 7208/11704 [1:01:18<37:03,  2.02it/s]
 62%|██████▏   | 7209/11704 [1:01:19<37:00,  2.02it/s]
 62%|██████▏   | 7210/11704 [1:01:19<37:00,  2.02it/s]
 62%|██████▏   | 7211/11704 [1:01:20<36:59,  2.02it/s]
 62%|██████▏   | 7212/11704 [1:01:20<37:00,  2.02it/s]
 62%|██████▏   | 7213/11704 [1:01:21<37:01,  2.02it/s]
 62%|██████▏   | 7214/11704 [1:01:21<37:01,  2.02it/s]
 62%|██████▏   | 7215/11704 [1:01:22<36:58,  2.02it/s]
 62%|██████▏   | 7216/11704 [1:01:22<36:59,  2.02it/s]
 62%|██████▏   | 7217/11704 [1:01:23<36:57,  2.02it/s]
 62%|██████▏   | 7218/11704 [1:01:23<36:58,  2.02it/s]
 62%|██████▏   | 7219/11704 [1:01:24<36:56,  2.02it/s]
 62%|██████▏   | 7220/11704 [1:01:24<36:57,  2.02it/s]
 62%|██████▏   | 7221/11704 [1:01:25<36:55,  2.02it/s]
 62%|██████▏   | 7222/11704 [1:01:25<36:55,  2.02it/s]
 62%|██████▏   | 7223/11704 [1:01:26<36:54,  2.02it/s]
 62%|██████▏   | 7224/11704 [1:01:26<36:56,  2.02it/s]
 62%|██████▏   | 7225/11704 [1:01:27<36:55,  2.02it/s]{'loss': 3.4331, 'grad_norm': 0.22490887343883514, 'learning_rate': 0.0003836362507133381, 'epoch': 8.64}
+
                                                      
 62%|██████▏   | 7225/11704 [1:01:27<36:55,  2.02it/s]
 62%|██████▏   | 7226/11704 [1:01:27<36:57,  2.02it/s]
 62%|██████▏   | 7227/11704 [1:01:28<36:56,  2.02it/s]
 62%|██████▏   | 7228/11704 [1:01:28<36:54,  2.02it/s]
 62%|██████▏   | 7229/11704 [1:01:29<36:53,  2.02it/s]
 62%|██████▏   | 7230/11704 [1:01:29<36:52,  2.02it/s]
 62%|██████▏   | 7231/11704 [1:01:30<36:49,  2.02it/s]
 62%|██████▏   | 7232/11704 [1:01:30<36:51,  2.02it/s]
 62%|██████▏   | 7233/11704 [1:01:31<36:47,  2.02it/s]
 62%|██████▏   | 7234/11704 [1:01:31<36:52,  2.02it/s]
 62%|██████▏   | 7235/11704 [1:01:32<36:48,  2.02it/s]
 62%|██████▏   | 7236/11704 [1:01:32<36:50,  2.02it/s]
 62%|██████▏   | 7237/11704 [1:01:33<36:48,  2.02it/s]
 62%|██████▏   | 7238/11704 [1:01:33<36:48,  2.02it/s]
 62%|██████▏   | 7239/11704 [1:01:34<36:46,  2.02it/s]
 62%|██████▏   | 7240/11704 [1:01:34<36:49,  2.02it/s]
 62%|██████▏   | 7241/11704 [1:01:35<36:46,  2.02it/s]
 62%|██████▏   | 7242/11704 [1:01:35<36:47,  2.02it/s]
 62%|██████▏   | 7243/11704 [1:01:36<36:43,  2.02it/s]
 62%|██████▏   | 7244/11704 [1:01:36<36:45,  2.02it/s]
 62%|██████▏   | 7245/11704 [1:01:37<36:45,  2.02it/s]
 62%|██████▏   | 7246/11704 [1:01:37<36:44,  2.02it/s]
 62%|██████▏   | 7247/11704 [1:01:38<36:41,  2.02it/s]
 62%|██████▏   | 7248/11704 [1:01:38<36:43,  2.02it/s]
 62%|██████▏   | 7249/11704 [1:01:39<36:42,  2.02it/s]
 62%|██████▏   | 7250/11704 [1:01:39<36:45,  2.02it/s]{'loss': 3.4319, 'grad_norm': 0.22441811859607697, 'learning_rate': 0.0003800136163950977, 'epoch': 8.67}
+
                                                      
 62%|██████▏   | 7250/11704 [1:01:39<36:45,  2.02it/s]
 62%|██████▏   | 7251/11704 [1:01:40<36:46,  2.02it/s]
 62%|██████▏   | 7252/11704 [1:01:40<36:46,  2.02it/s]
 62%|██████▏   | 7253/11704 [1:01:41<36:42,  2.02it/s]
 62%|██████▏   | 7254/11704 [1:01:41<36:42,  2.02it/s]
 62%|██████▏   | 7255/11704 [1:01:42<36:40,  2.02it/s]
 62%|██████▏   | 7256/11704 [1:01:42<36:42,  2.02it/s]
 62%|██████▏   | 7257/11704 [1:01:43<36:38,  2.02it/s]
 62%|██████▏   | 7258/11704 [1:01:43<36:39,  2.02it/s]
 62%|██████▏   | 7259/11704 [1:01:43<36:38,  2.02it/s]
 62%|██████▏   | 7260/11704 [1:01:44<36:38,  2.02it/s]
 62%|██████▏   | 7261/11704 [1:01:44<36:37,  2.02it/s]
 62%|██████▏   | 7262/11704 [1:01:45<36:38,  2.02it/s]
 62%|██████▏   | 7263/11704 [1:01:45<36:35,  2.02it/s]
 62%|██████▏   | 7264/11704 [1:01:46<36:37,  2.02it/s]
 62%|██████▏   | 7265/11704 [1:01:46<36:35,  2.02it/s]
 62%|██████▏   | 7266/11704 [1:01:47<36:40,  2.02it/s]
 62%|██████▏   | 7267/11704 [1:01:47<36:37,  2.02it/s]
 62%|██████▏   | 7268/11704 [1:01:48<36:39,  2.02it/s]
 62%|██████▏   | 7269/11704 [1:01:48<36:37,  2.02it/s]
 62%|██████▏   | 7270/11704 [1:01:49<36:36,  2.02it/s]
 62%|██████▏   | 7271/11704 [1:01:49<36:35,  2.02it/s]
 62%|██████▏   | 7272/11704 [1:01:50<36:33,  2.02it/s]
 62%|██████▏   | 7273/11704 [1:01:50<36:31,  2.02it/s]
 62%|██████▏   | 7274/11704 [1:01:51<36:30,  2.02it/s]
 62%|██████▏   | 7275/11704 [1:01:51<36:29,  2.02it/s]{'loss': 3.431, 'grad_norm': 0.23661942780017853, 'learning_rate': 0.0003763976533012071, 'epoch': 8.7}
+
                                                      
 62%|██████▏   | 7275/11704 [1:01:51<36:29,  2.02it/s]
 62%|██████▏   | 7276/11704 [1:01:52<36:31,  2.02it/s]
 62%|██████▏   | 7277/11704 [1:01:52<36:33,  2.02it/s]
 62%|██████▏   | 7278/11704 [1:01:53<36:30,  2.02it/s]
 62%|██████▏   | 7279/11704 [1:01:53<36:29,  2.02it/s]
 62%|██████▏   | 7280/11704 [1:01:54<36:27,  2.02it/s]
 62%|██████▏   | 7281/11704 [1:01:54<36:25,  2.02it/s]
 62%|██████▏   | 7282/11704 [1:01:55<36:24,  2.02it/s]
 62%|██████▏   | 7283/11704 [1:01:55<36:22,  2.03it/s]
 62%|██████▏   | 7284/11704 [1:01:56<36:22,  2.03it/s]
 62%|██████▏   | 7285/11704 [1:01:56<36:21,  2.03it/s]
 62%|██████▏   | 7286/11704 [1:01:57<36:24,  2.02it/s]
 62%|██████▏   | 7287/11704 [1:01:57<36:22,  2.02it/s]
 62%|██████▏   | 7288/11704 [1:01:58<36:25,  2.02it/s]
 62%|██████▏   | 7289/11704 [1:01:58<36:22,  2.02it/s]
 62%|██████▏   | 7290/11704 [1:01:59<36:22,  2.02it/s]
 62%|██████▏   | 7291/11704 [1:01:59<36:20,  2.02it/s]
 62%|██████▏   | 7292/11704 [1:02:00<36:21,  2.02it/s]
 62%|██████▏   | 7293/11704 [1:02:00<36:17,  2.03it/s]
 62%|██████▏   | 7294/11704 [1:02:01<36:18,  2.02it/s]
 62%|██████▏   | 7295/11704 [1:02:01<36:17,  2.02it/s]
 62%|██████▏   | 7296/11704 [1:02:02<36:18,  2.02it/s]
 62%|██████▏   | 7297/11704 [1:02:02<36:18,  2.02it/s]
 62%|██████▏   | 7298/11704 [1:02:03<36:17,  2.02it/s]
 62%|██████▏   | 7299/11704 [1:02:03<36:19,  2.02it/s]
 62%|██████▏   | 7300/11704 [1:02:04<36:18,  2.02it/s]{'loss': 3.4237, 'grad_norm': 0.23303642868995667, 'learning_rate': 0.00037278856247865446, 'epoch': 8.73}
+
                                                      
 62%|██████▏   | 7300/11704 [1:02:04<36:18,  2.02it/s]
 62%|██████▏   | 7301/11704 [1:02:04<36:21,  2.02it/s]
 62%|██████▏   | 7302/11704 [1:02:05<36:18,  2.02it/s]
 62%|██████▏   | 7303/11704 [1:02:05<36:18,  2.02it/s]
 62%|██████▏   | 7304/11704 [1:02:06<36:17,  2.02it/s]
 62%|██████▏   | 7305/11704 [1:02:06<36:18,  2.02it/s]
 62%|██████▏   | 7306/11704 [1:02:07<36:17,  2.02it/s]
 62%|██████▏   | 7307/11704 [1:02:07<36:16,  2.02it/s]
 62%|██████▏   | 7308/11704 [1:02:08<36:14,  2.02it/s]
 62%|██████▏   | 7309/11704 [1:02:08<36:15,  2.02it/s]
 62%|██████▏   | 7310/11704 [1:02:09<36:17,  2.02it/s]
 62%|██████▏   | 7311/11704 [1:02:09<36:14,  2.02it/s]
 62%|██████▏   | 7312/11704 [1:02:10<36:14,  2.02it/s]
 62%|██████▏   | 7313/11704 [1:02:10<36:11,  2.02it/s]
 62%|██████▏   | 7314/11704 [1:02:11<36:12,  2.02it/s]
 62%|██████▎   | 7315/11704 [1:02:11<36:10,  2.02it/s]
 63%|██████▎   | 7316/11704 [1:02:12<36:11,  2.02it/s]
 63%|██████▎   | 7317/11704 [1:02:12<36:10,  2.02it/s]
 63%|██████▎   | 7318/11704 [1:02:13<36:09,  2.02it/s]
 63%|██████▎   | 7319/11704 [1:02:13<36:08,  2.02it/s]
 63%|██████▎   | 7320/11704 [1:02:14<36:11,  2.02it/s]
 63%|██████▎   | 7321/11704 [1:02:14<36:08,  2.02it/s]
 63%|██████▎   | 7322/11704 [1:02:15<36:08,  2.02it/s]
 63%|██████▎   | 7323/11704 [1:02:15<36:06,  2.02it/s]
 63%|██████▎   | 7324/11704 [1:02:16<36:05,  2.02it/s]
 63%|██████▎   | 7325/11704 [1:02:16<36:03,  2.02it/s]{'loss': 3.4348, 'grad_norm': 0.22636759281158447, 'learning_rate': 0.000369186544592331, 'epoch': 8.76}
+
                                                      
 63%|██████▎   | 7325/11704 [1:02:16<36:03,  2.02it/s]
 63%|██████▎   | 7326/11704 [1:02:17<36:07,  2.02it/s]
 63%|██████▎   | 7327/11704 [1:02:17<36:05,  2.02it/s]
 63%|██████▎   | 7328/11704 [1:02:18<36:04,  2.02it/s]
 63%|██████▎   | 7329/11704 [1:02:18<36:02,  2.02it/s]
 63%|██████▎   | 7330/11704 [1:02:19<36:05,  2.02it/s]
 63%|██████▎   | 7331/11704 [1:02:19<36:02,  2.02it/s]
 63%|██████▎   | 7332/11704 [1:02:20<36:02,  2.02it/s]
 63%|██████▎   | 7333/11704 [1:02:20<36:01,  2.02it/s]
 63%|██████▎   | 7334/11704 [1:02:21<36:00,  2.02it/s]
 63%|██████▎   | 7335/11704 [1:02:21<35:59,  2.02it/s]
 63%|██████▎   | 7336/11704 [1:02:22<35:59,  2.02it/s]
 63%|██████▎   | 7337/11704 [1:02:22<35:58,  2.02it/s]
 63%|██████▎   | 7338/11704 [1:02:23<35:58,  2.02it/s]
 63%|██████▎   | 7339/11704 [1:02:23<35:58,  2.02it/s]
 63%|██████▎   | 7340/11704 [1:02:24<35:58,  2.02it/s]
 63%|██████▎   | 7341/11704 [1:02:24<35:58,  2.02it/s]
 63%|██████▎   | 7342/11704 [1:02:25<35:57,  2.02it/s]
 63%|██████▎   | 7343/11704 [1:02:25<35:57,  2.02it/s]
 63%|██████▎   | 7344/11704 [1:02:26<35:55,  2.02it/s]
 63%|██████▎   | 7345/11704 [1:02:26<35:54,  2.02it/s]
 63%|██████▎   | 7346/11704 [1:02:27<35:54,  2.02it/s]
 63%|██████▎   | 7347/11704 [1:02:27<35:53,  2.02it/s]
 63%|██████▎   | 7348/11704 [1:02:28<35:53,  2.02it/s]
 63%|██████▎   | 7349/11704 [1:02:28<35:52,  2.02it/s]
 63%|██████▎   | 7350/11704 [1:02:29<35:52,  2.02it/s]{'loss': 3.4373, 'grad_norm': 0.22749438881874084, 'learning_rate': 0.00036559179991387295, 'epoch': 8.79}
+
                                                      
 63%|██████▎   | 7350/11704 [1:02:29<35:52,  2.02it/s]
 63%|██████▎   | 7351/11704 [1:02:29<35:52,  2.02it/s]
 63%|██████▎   | 7352/11704 [1:02:29<35:53,  2.02it/s]
 63%|██████▎   | 7353/11704 [1:02:30<35:52,  2.02it/s]
 63%|██████▎   | 7354/11704 [1:02:30<35:50,  2.02it/s]
 63%|██████▎   | 7355/11704 [1:02:31<35:50,  2.02it/s]
 63%|██████▎   | 7356/11704 [1:02:31<35:48,  2.02it/s]
 63%|��█████▎   | 7357/11704 [1:02:32<35:50,  2.02it/s]
 63%|██████▎   | 7358/11704 [1:02:32<35:49,  2.02it/s]
 63%|██████▎   | 7359/11704 [1:02:33<35:48,  2.02it/s]
 63%|██████▎   | 7360/11704 [1:02:33<35:46,  2.02it/s]
 63%|██████▎   | 7361/11704 [1:02:34<35:45,  2.02it/s]
 63%|██████▎   | 7362/11704 [1:02:34<35:46,  2.02it/s]
 63%|██████▎   | 7363/11704 [1:02:35<35:47,  2.02it/s]
 63%|██████▎   | 7364/11704 [1:02:35<35:47,  2.02it/s]
 63%|██████▎   | 7365/11704 [1:02:36<35:46,  2.02it/s]
 63%|██████▎   | 7366/11704 [1:02:36<35:45,  2.02it/s]
 63%|██████▎   | 7367/11704 [1:02:37<35:44,  2.02it/s]
 63%|██████▎   | 7368/11704 [1:02:37<35:42,  2.02it/s]
 63%|██████▎   | 7369/11704 [1:02:38<35:43,  2.02it/s]
 63%|██████▎   | 7370/11704 [1:02:38<35:42,  2.02it/s]
 63%|██████▎   | 7371/11704 [1:02:39<35:44,  2.02it/s]
 63%|██████▎   | 7372/11704 [1:02:39<35:42,  2.02it/s]
 63%|██████▎   | 7373/11704 [1:02:40<35:43,  2.02it/s]
 63%|██████▎   | 7374/11704 [1:02:40<35:43,  2.02it/s]
 63%|██████▎   | 7375/11704 [1:02:41<35:41,  2.02it/s]{'loss': 3.4323, 'grad_norm': 0.22370922565460205, 'learning_rate': 0.0003620045283105281, 'epoch': 8.82}
                                                      
+
 63%|██████▎   | 7375/11704 [1:02:41<35:41,  2.02it/s]
 63%|██████▎   | 7376/11704 [1:02:41<35:42,  2.02it/s]
 63%|██████▎   | 7377/11704 [1:02:42<35:41,  2.02it/s]
 63%|██████▎   | 7378/11704 [1:02:42<35:40,  2.02it/s]
 63%|██████▎   | 7379/11704 [1:02:43<35:39,  2.02it/s]
 63%|██████▎   | 7380/11704 [1:02:43<35:39,  2.02it/s]
 63%|██████▎   | 7381/11704 [1:02:44<35:39,  2.02it/s]
 63%|██████▎   | 7382/11704 [1:02:44<35:38,  2.02it/s]
 63%|██████▎   | 7383/11704 [1:02:45<35:38,  2.02it/s]
 63%|██████▎   | 7384/11704 [1:02:45<35:37,  2.02it/s]
 63%|██████▎   | 7385/11704 [1:02:46<35:36,  2.02it/s]
 63%|██████▎   | 7386/11704 [1:02:46<35:35,  2.02it/s]
 63%|██████▎   | 7387/11704 [1:02:47<35:33,  2.02it/s]
 63%|██████▎   | 7388/11704 [1:02:47<35:32,  2.02it/s]
 63%|██████▎   | 7389/11704 [1:02:48<35:32,  2.02it/s]
 63%|██████▎   | 7390/11704 [1:02:48<35:32,  2.02it/s]
 63%|██████▎   | 7391/11704 [1:02:49<35:30,  2.02it/s]
 63%|██████▎   | 7392/11704 [1:02:49<35:31,  2.02it/s]
 63%|██████▎   | 7393/11704 [1:02:50<35:29,  2.02it/s]
 63%|██████▎   | 7394/11704 [1:02:50<35:30,  2.02it/s]
 63%|██████▎   | 7395/11704 [1:02:51<35:28,  2.02it/s]
 63%|██████▎   | 7396/11704 [1:02:51<35:29,  2.02it/s]
 63%|██████▎   | 7397/11704 [1:02:52<35:28,  2.02it/s]
 63%|██████▎   | 7398/11704 [1:02:52<35:29,  2.02it/s]
 63%|██████▎   | 7399/11704 [1:02:53<35:26,  2.02it/s]
 63%|██████▎   | 7400/11704 [1:02:53<35:26,  2.02it/s]{'loss': 3.4399, 'grad_norm': 0.22380071878433228, 'learning_rate': 0.0003584249292340423, 'epoch': 8.85}
+
                                                      
 63%|██████▎   | 7400/11704 [1:02:53<35:26,  2.02it/s]
 63%|██████▎   | 7401/11704 [1:02:54<35:27,  2.02it/s]
 63%|██████▎   | 7402/11704 [1:02:54<35:28,  2.02it/s]
 63%|██████▎   | 7403/11704 [1:02:55<35:25,  2.02it/s]
 63%|██████▎   | 7404/11704 [1:02:55<35:25,  2.02it/s]
 63%|██████▎   | 7405/11704 [1:02:56<35:23,  2.02it/s]
 63%|██████▎   | 7406/11704 [1:02:56<35:24,  2.02it/s]
 63%|██████▎   | 7407/11704 [1:02:57<35:23,  2.02it/s]
 63%|██████▎   | 7408/11704 [1:02:57<35:24,  2.02it/s]
 63%|██████▎   | 7409/11704 [1:02:58<35:23,  2.02it/s]
 63%|██████▎   | 7410/11704 [1:02:58<35:22,  2.02it/s]
 63%|██████▎   | 7411/11704 [1:02:59<35:22,  2.02it/s]
 63%|██████▎   | 7412/11704 [1:02:59<35:22,  2.02it/s]
 63%|██████▎   | 7413/11704 [1:03:00<35:21,  2.02it/s]
 63%|██████▎   | 7414/11704 [1:03:00<35:23,  2.02it/s]
 63%|██████▎   | 7415/11704 [1:03:01<35:21,  2.02it/s]
 63%|██████▎   | 7416/11704 [1:03:01<35:24,  2.02it/s]
 63%|██████▎   | 7417/11704 [1:03:02<35:22,  2.02it/s]
 63%|██████▎   | 7418/11704 [1:03:02<35:22,  2.02it/s]
 63%|██████▎   | 7419/11704 [1:03:03<35:20,  2.02it/s]
 63%|██████▎   | 7420/11704 [1:03:03<35:19,  2.02it/s]
 63%|██████▎   | 7421/11704 [1:03:04<35:17,  2.02it/s]
 63%|██████▎   | 7422/11704 [1:03:04<35:17,  2.02it/s]
 63%|██████▎   | 7423/11704 [1:03:05<35:15,  2.02it/s]
 63%|██████▎   | 7424/11704 [1:03:05<35:17,  2.02it/s]
 63%|██████▎   | 7425/11704 [1:03:06<35:15,  2.02it/s]{'loss': 3.4316, 'grad_norm': 0.2450966238975525, 'learning_rate': 0.0003548532017095696, 'epoch': 8.88}
+
                                                      
 63%|██████▎   | 7425/11704 [1:03:06<35:15,  2.02it/s]
 63%|██████▎   | 7426/11704 [1:03:06<35:18,  2.02it/s]
 63%|██████▎   | 7427/11704 [1:03:07<35:15,  2.02it/s]
 63%|██████▎   | 7428/11704 [1:03:07<35:15,  2.02it/s]
 63%|██████▎   | 7429/11704 [1:03:08<35:12,  2.02it/s]
 63%|██████▎   | 7430/11704 [1:03:08<35:12,  2.02it/s]
 63%|██████▎   | 7431/11704 [1:03:09<35:12,  2.02it/s]
 63%|██████▎   | 7432/11704 [1:03:09<35:15,  2.02it/s]
 64%|██████▎   | 7433/11704 [1:03:10<35:12,  2.02it/s]
 64%|██████▎   | 7434/11704 [1:03:10<35:10,  2.02it/s]
 64%|██████▎   | 7435/11704 [1:03:11<35:10,  2.02it/s]
 64%|██████▎   | 7436/11704 [1:03:11<35:10,  2.02it/s]
 64%|██████▎   | 7437/11704 [1:03:12<35:12,  2.02it/s]
 64%|██████▎   | 7438/11704 [1:03:12<35:10,  2.02it/s]
 64%|██████▎   | 7439/11704 [1:03:13<35:11,  2.02it/s]
 64%|██████▎   | 7440/11704 [1:03:13<35:10,  2.02it/s]
 64%|██████▎   | 7441/11704 [1:03:14<35:10,  2.02it/s]
 64%|██████▎   | 7442/11704 [1:03:14<35:10,  2.02it/s]
 64%|██████▎   | 7443/11704 [1:03:15<35:09,  2.02it/s]
 64%|██████▎   | 7444/11704 [1:03:15<35:08,  2.02it/s]
 64%|██████▎   | 7445/11704 [1:03:15<35:08,  2.02it/s]
 64%|██████▎   | 7446/11704 [1:03:16<35:06,  2.02it/s]
 64%|██████▎   | 7447/11704 [1:03:16<35:04,  2.02it/s]
 64%|██████▎   | 7448/11704 [1:03:17<35:06,  2.02it/s]
 64%|██████▎   | 7449/11704 [1:03:17<35:05,  2.02it/s]
 64%|██████▎   | 7450/11704 [1:03:18<35:05,  2.02it/s]{'loss': 3.4375, 'grad_norm': 0.22761301696300507, 'learning_rate': 0.0003512895443246075, 'epoch': 8.91}
+
                                                      
 64%|██████▎   | 7450/11704 [1:03:18<35:05,  2.02it/s]
 64%|██████▎   | 7451/11704 [1:03:18<35:05,  2.02it/s]
 64%|██████▎   | 7452/11704 [1:03:19<35:05,  2.02it/s]
 64%|██████▎   | 7453/11704 [1:03:19<35:01,  2.02it/s]
 64%|██████▎   | 7454/11704 [1:03:20<35:01,  2.02it/s]
 64%|██████▎   | 7455/11704 [1:03:20<34:59,  2.02it/s]
 64%|██████▎   | 7456/11704 [1:03:21<35:00,  2.02it/s]
 64%|██████▎   | 7457/11704 [1:03:21<34:58,  2.02it/s]
 64%|██████▎   | 7458/11704 [1:03:22<34:58,  2.02it/s]
 64%|██████▎   | 7459/11704 [1:03:22<34:57,  2.02it/s]
 64%|██████▎   | 7460/11704 [1:03:23<35:00,  2.02it/s]
 64%|██████▎   | 7461/11704 [1:03:23<34:57,  2.02it/s]
 64%|██████▍   | 7462/11704 [1:03:24<34:58,  2.02it/s]
 64%|██████▍   | 7463/11704 [1:03:24<34:56,  2.02it/s]
 64%|██████▍   | 7464/11704 [1:03:25<34:54,  2.02it/s]
 64%|██████▍   | 7465/11704 [1:03:25<34:54,  2.02it/s]
 64%|██████▍   | 7466/11704 [1:03:26<34:53,  2.02it/s]
 64%|██████▍   | 7467/11704 [1:03:26<34:54,  2.02it/s]
 64%|██████▍   | 7468/11704 [1:03:27<34:55,  2.02it/s]
 64%|██████▍   | 7469/11704 [1:03:27<34:55,  2.02it/s]
 64%|██████▍   | 7470/11704 [1:03:28<34:58,  2.02it/s]
 64%|██████▍   | 7471/11704 [1:03:28<34:54,  2.02it/s]
 64%|██████▍   | 7472/11704 [1:03:29<34:55,  2.02it/s]
 64%|██████▍   | 7473/11704 [1:03:29<34:50,  2.02it/s]
 64%|██████▍   | 7474/11704 [1:03:30<34:50,  2.02it/s]
 64%|██████▍   | 7475/11704 [1:03:30<34:50,  2.02it/s]{'loss': 3.4432, 'grad_norm': 0.22205999493598938, 'learning_rate': 0.000347734155217955, 'epoch': 8.94}
+
                                                      
 64%|██████▍   | 7475/11704 [1:03:30<34:50,  2.02it/s]
 64%|██████▍   | 7476/11704 [1:03:31<34:52,  2.02it/s]
 64%|██████▍   | 7477/11704 [1:03:31<34:50,  2.02it/s]
 64%|██████▍   | 7478/11704 [1:03:32<34:50,  2.02it/s]
 64%|██████▍   | 7479/11704 [1:03:32<34:48,  2.02it/s]
 64%|██████▍   | 7480/11704 [1:03:33<34:50,  2.02it/s]
 64%|██████▍   | 7481/11704 [1:03:33<34:48,  2.02it/s]
 64%|██████▍   | 7482/11704 [1:03:34<34:46,  2.02it/s]
 64%|██████▍   | 7483/11704 [1:03:34<34:44,  2.03it/s]
 64%|██████▍   | 7484/11704 [1:03:35<34:42,  2.03it/s]
 64%|██████▍   | 7485/11704 [1:03:35<34:43,  2.02it/s]
 64%|██████▍   | 7486/11704 [1:03:36<34:42,  2.03it/s]
 64%|██████▍   | 7487/11704 [1:03:36<34:43,  2.02it/s]
 64%|██████▍   | 7488/11704 [1:03:37<34:41,  2.03it/s]
 64%|██████▍   | 7489/11704 [1:03:37<34:41,  2.02it/s]
 64%|██████▍   | 7490/11704 [1:03:38<34:41,  2.02it/s]
 64%|██████▍   | 7491/11704 [1:03:38<34:41,  2.02it/s]
 64%|██████▍   | 7492/11704 [1:03:39<34:45,  2.02it/s]
 64%|██████▍   | 7493/11704 [1:03:39<34:42,  2.02it/s]
 64%|██████▍   | 7494/11704 [1:03:40<34:42,  2.02it/s]
 64%|██████▍   | 7495/11704 [1:03:40<34:41,  2.02it/s]
 64%|██████▍   | 7496/11704 [1:03:41<34:38,  2.02it/s]
 64%|██████▍   | 7497/11704 [1:03:41<34:38,  2.02it/s]
 64%|██████▍   | 7498/11704 [1:03:42<34:37,  2.02it/s]
 64%|██████▍   | 7499/11704 [1:03:42<34:37,  2.02it/s]
 64%|██████▍   | 7500/11704 [1:03:43<34:36,  2.02it/s]{'loss': 3.4408, 'grad_norm': 0.2216581404209137, 'learning_rate': 0.0003441872320686955, 'epoch': 8.97}
+
                                                      
 64%|██████▍   | 7500/11704 [1:03:43<34:36,  2.02it/s]
 64%|██████▍   | 7501/11704 [1:03:43<34:40,  2.02it/s]
 64%|██████▍   | 7502/11704 [1:03:44<34:37,  2.02it/s]
 64%|██████▍   | 7503/11704 [1:03:44<34:39,  2.02it/s]
 64%|██████▍   | 7504/11704 [1:03:45<34:36,  2.02it/s]
 64%|██████▍   | 7505/11704 [1:03:45<34:37,  2.02it/s]
 64%|██████▍   | 7506/11704 [1:03:46<34:35,  2.02it/s]
 64%|██████▍   | 7507/11704 [1:03:46<34:35,  2.02it/s]
 64%|██████▍   | 7508/11704 [1:03:47<34:32,  2.02it/s]
 64%|██████▍   | 7509/11704 [1:03:47<34:32,  2.02it/s]
 64%|██████▍   | 7510/11704 [1:03:48<34:30,  2.03it/s]
 64%|██████▍   | 7511/11704 [1:03:48<34:32,  2.02it/s]
 64%|██████▍   | 7512/11704 [1:03:49<34:33,  2.02it/s]
 64%|██████▍   | 7513/11704 [1:03:49<34:34,  2.02it/s]
 64%|██████▍   | 7514/11704 [1:03:50<34:32,  2.02it/s]
 64%|██████▍   | 7515/11704 [1:03:50<34:33,  2.02it/s]
 64%|██████▍   | 7516/11704 [1:03:51<34:31,  2.02it/s]
 64%|██████▍   | 7517/11704 [1:03:51<34:30,  2.02it/s]
 64%|██████▍   | 7518/11704 [1:03:52<34:28,  2.02it/s]
 64%|██████▍   | 7519/11704 [1:03:52<34:25,  2.03it/s]
 64%|██████▍   | 7520/11704 [1:03:53<34:25,  2.03it/s]
 64%|██████▍   | 7521/11704 [1:03:53<34:24,  2.03it/s]
 64%|██████▍   | 7522/11704 [1:03:54<34:26,  2.02it/s]
 64%|██████▍   | 7523/11704 [1:03:54<34:25,  2.02it/s]
 64%|██████▍   | 7524/11704 [1:03:55<34:27,  2.02it/s]
 64%|██████▍   | 7525/11704 [1:03:55<34:26,  2.02it/s]{'loss': 3.4361, 'grad_norm': 0.22303152084350586, 'learning_rate': 0.0003406489720852072, 'epoch': 9.0}
+
                                                      
 64%|██████▍   | 7525/11704 [1:03:55<34:26,  2.02it/s]
 64%|██████▍   | 7526/11704 [1:03:56<34:29,  2.02it/s]
 64%|██████▍   | 7527/11704 [1:03:56<34:25,  2.02it/s]
 64%|██████▍   | 7528/11704 [1:03:57<35:19,  1.97it/s]
 64%|██████▍   | 7529/11704 [1:04:09<4:46:55,  4.12s/it]
 64%|██████▍   | 7530/11704 [1:04:10<3:31:10,  3.04s/it]
 64%|██████▍   | 7531/11704 [1:04:10<2:38:07,  2.27s/it]
 64%|██████▍   | 7532/11704 [1:04:11<2:01:09,  1.74s/it]
 64%|██████▍   | 7533/11704 [1:04:11<1:35:16,  1.37s/it]
 64%|██████▍   | 7534/11704 [1:04:12<1:17:02,  1.11s/it]
 64%|██████▍   | 7535/11704 [1:04:12<1:04:13,  1.08it/s]
 64%|██████▍   | 7536/11704 [1:04:13<55:18,  1.26it/s]  
 64%|██████▍   | 7537/11704 [1:04:13<49:01,  1.42it/s]
 64%|██████▍   | 7538/11704 [1:04:14<44:41,  1.55it/s]
 64%|██████▍   | 7539/11704 [1:04:14<41:38,  1.67it/s]
 64%|██████▍   | 7540/11704 [1:04:15<39:25,  1.76it/s]
 64%|██████▍   | 7541/11704 [1:04:15<37:53,  1.83it/s]
 64%|██████▍   | 7542/11704 [1:04:16<36:50,  1.88it/s]
 64%|██████▍   | 7543/11704 [1:04:16<36:04,  1.92it/s]
 64%|██████▍   | 7544/11704 [1:04:17<35:30,  1.95it/s]
 64%|██████▍   | 7545/11704 [1:04:17<35:06,  1.97it/s]
 64%|██████▍   | 7546/11704 [1:04:18<34:49,  1.99it/s]
 64%|██████▍   | 7547/11704 [1:04:18<34:38,  2.00it/s]
 64%|██████▍   | 7548/11704 [1:04:19<34:29,  2.01it/s]
 64%|██████▍   | 7549/11704 [1:04:19<34:22,  2.01it/s]
 65%|██████▍   | 7550/11704 [1:04:20<34:21,  2.02it/s]
                                                      {'loss': 3.3287, 'grad_norm': 0.23561598360538483, 'learning_rate': 0.0003371195719941973, 'epoch': 9.03}
+
 65%|██████▍   | 7550/11704 [1:04:20<34:21,  2.02it/s]
 65%|██████▍   | 7551/11704 [1:04:20<34:20,  2.02it/s]
 65%|██████▍   | 7552/11704 [1:04:21<34:15,  2.02it/s]
 65%|██████▍   | 7553/11704 [1:04:21<34:14,  2.02it/s]
 65%|██████▍   | 7554/11704 [1:04:22<34:13,  2.02it/s]
 65%|██████▍   | 7555/11704 [1:04:22<34:14,  2.02it/s]
 65%|██████▍   | 7556/11704 [1:04:23<34:12,  2.02it/s]
 65%|██████▍   | 7557/11704 [1:04:23<34:09,  2.02it/s]