diff --git "a/train_job_output.txt" "b/train_job_output.txt"
--- "a/train_job_output.txt"
+++ "b/train_job_output.txt"
@@ -429,4 +429,66 @@ command outputs:
 
                                                      

 85%|████████▌ | 9100/10682 [1:23:27<12:57,  2.03it/s]
 85%|████████▌ | 9101/10682 [1:23:27<12:58,  2.03it/s]
 85%|████████▌ | 9102/10682 [1:23:28<12:57,  2.03it/s]
 85%|████████▌ | 9103/10682 [1:23:28<12:57,  2.03it/s]
 85%|████████▌ | 9104/10682 [1:23:29<12:56,  2.03it/s]
 85%|████████▌ | 9105/10682 [1:23:29<12:56,  2.03it/s]
 85%|████████▌ | 9106/10682 [1:23:30<12:56,  2.03it/s]
 85%|████████▌ | 9107/10682 [1:23:30<12:54,  2.03it/s]
 85%|████████▌ | 9108/10682 [1:23:31<12:55,  2.03it/s]
 85%|████████▌ | 9109/10682 [1:23:31<12:55,  2.03it/s]
 85%|████████▌ | 9110/10682 [1:23:32<12:55,  2.03it/s]
 85%|████████▌ | 9111/10682 [1:23:32<12:54,  2.03it/s]
 85%|████████▌ | 9112/10682 [1:23:32<12:53,  2.03it/s]
 85%|████████▌ | 9113/10682 [1:23:33<12:52,  2.03it/s]
 85%|████████▌ | 9114/10682 [1:23:33<12:51,  2.03it/s]
 85%|████████▌ | 9115/10682 [1:23:34<12:50,  2.03it/s]
 85%|████████▌ | 9116/10682 [1:23:34<12:50,  2.03it/s]
 85%|████████▌ | 9117/10682 [1:23:35<12:50,  2.03it/s]
 85%|████████▌ | 9118/10682 [1:23:35<12:49,  2.03it/s]
 85%|████████▌ | 9119/10682 [1:23:36<12:49,  2.03it/s]
 85%|████████▌ | 9120/10682 [1:23:36<12:48,  2.03it/s]
 85%|████████▌ | 9121/10682 [1:23:37<12:48,  2.03it/s]
 85%|████████▌ | 9122/10682 [1:23:37<12:47,  2.03it/s]
 85%|████████▌ | 9123/10682 [1:23:38<12:46,  2.03it/s]
 85%|████████▌ | 9124/10682 [1:23:38<12:45,  2.03it/s]
 85%|████████▌ | 9125/10682 [1:23:39<12:46,  2.03it/s]
                                                      {'loss': 2.8591, 'grad_norm': 0.25908002257347107, 'learning_rate': 6.334441157888504e-05, 'epoch': 11.96}
 

 85%|████████▌ | 9125/10682 [1:23:39<12:46,  2.03it/s]
 85%|████████▌ | 9126/10682 [1:23:39<12:46,  2.03it/s]
 85%|████████▌ | 9127/10682 [1:23:40<12:45,  2.03it/s]
 85%|████████▌ | 9128/10682 [1:23:40<12:45,  2.03it/s]
 85%|████████▌ | 9129/10682 [1:23:41<12:43,  2.03it/s]
 85%|████████▌ | 9130/10682 [1:23:41<12:44,  2.03it/s]
 85%|████████▌ | 9131/10682 [1:23:42<12:43,  2.03it/s]
 85%|████████▌ | 9132/10682 [1:23:42<12:42,  2.03it/s]
 85%|████████▌ | 9133/10682 [1:23:43<12:43,  2.03it/s]
 86%|████████▌ | 9134/10682 [1:23:43<12:41,  2.03it/s]
 86%|████████▌ | 9135/10682 [1:23:44<12:41,  2.03it/s]
 86%|████████▌ | 9136/10682 [1:23:44<12:41,  2.03it/s]
 86%|████████▌ | 9137/10682 [1:23:45<12:39,  2.03it/s]
 86%|████████▌ | 9138/10682 [1:23:45<12:39,  2.03it/s]
 86%|████████▌ | 9139/10682 [1:23:46<12:39,  2.03it/s]
 86%|████████▌ | 9140/10682 [1:23:46<12:38,  2.03it/s]
 86%|████████▌ | 9141/10682 [1:23:47<12:37,  2.03it/s]
 86%|████████▌ | 9142/10682 [1:23:47<12:37,  2.03it/s]
 86%|████████▌ | 9143/10682 [1:23:48<12:36,  2.03it/s]
 86%|████████▌ | 9144/10682 [1:23:48<12:36,  2.03it/s]
 86%|████████▌ | 9145/10682 [1:23:49<12:36,  2.03it/s]
 86%|████████▌ | 9146/10682 [1:23:49<12:35,  2.03it/s]
 86%|████████▌ | 9147/10682 [1:23:50<12:35,  2.03it/s]
 86%|████████▌ | 9148/10682 [1:23:50<12:35,  2.03it/s]
 86%|████████▌ | 9149/10682 [1:23:51<12:34,  2.03it/s]
 86%|████████▌ | 9150/10682 [1:23:51<12:33,  2.03it/s]
                                                      
{'loss': 2.868, 'grad_norm': 0.26225173473358154, 'learning_rate': 6.1368906655978e-05, 'epoch': 11.99}
 
 86%|████████▌ | 9150/10682 [1:23:51<12:33,  2.03it/s]
 86%|████████▌ | 9151/10682 [1:23:52<12:34,  2.03it/s]
 86%|████████▌ | 9152/10682 [1:23:52<12:34,  2.03it/s]
 86%|████████▌ | 9153/10682 [1:23:53<12:33,  2.03it/s]
 86%|████████▌ | 9154/10682 [1:23:53<12:33,  2.03it/s]
 86%|████████▌ | 9155/10682 [1:23:54<12:31,  2.03it/s]
 86%|████████▌ | 9156/10682 [1:23:54<12:22,  2.05it/s]
 86%|████████▌ | 9157/10682 [1:24:48<7:00:18, 16.54s/it]
 86%|████████▌ | 9158/10682 [1:24:49<4:57:50, 11.73s/it]
 86%|████████▌ | 9159/10682 [1:24:49<3:32:06,  8.36s/it]
 86%|████████▌ | 9160/10682 [1:24:50<2:32:13,  6.00s/it]
 86%|████████▌ | 9161/10682 [1:24:50<1:50:13,  4.35s/it]
 86%|████████▌ | 9162/10682 [1:24:51<1:20:50,  3.19s/it]
 86%|████████▌ | 9163/10682 [1:24:51<1:00:18,  2.38s/it]
 86%|████████▌ | 9164/10682 [1:24:52<45:55,  1.82s/it]  
 86%|████████▌ | 9165/10682 [1:24:52<35:52,  1.42s/it]
 86%|████████▌ | 9166/10682 [1:24:53<28:51,  1.14s/it]
 86%|████████▌ | 9167/10682 [1:24:53<23:53,  1.06it/s]
 86%|████████▌ | 9168/10682 [1:24:54<20:27,  1.23it/s]
 86%|████████▌ | 9169/10682 [1:24:54<18:02,  1.40it/s]
 86%|████████▌ | 9170/10682 [1:24:55<16:21,  1.54it/s]
 86%|████████▌ | 9171/10682 [1:24:55<15:09,  1.66it/s]
 86%|████████▌ | 9172/10682 [1:24:56<14:19,  1.76it/s]
 86%|████████▌ | 9173/10682 [1:24:56<13:44,  1.83it/s]
 86%|████████▌ | 9174/10682 [1:24:57<13:20,  1.88it/s]
 86%|████████▌ | 9175/10682 [1:24:57<13:02,  1.92it/s]
                                                      
{'loss': 2.8099, 'grad_norm': 0.2629103660583496, 'learning_rate': 5.94226809108499e-05, 'epoch': 12.02}
-
 86%|████████▌ | 9175/10682 [1:24:57<13:02,  1.92it/s]
 86%|████████▌ | 9176/10682 [1:24:58<12:52,  1.95it/s]
 86%|████████▌ | 9177/10682 [1:24:58<12:43,  1.97it/s]
 86%|████████▌ | 9178/10682 [1:24:59<12:35,  1.99it/s]
 86%|████████▌ | 9179/10682 [1:24:59<12:32,  2.00it/s]
 86%|████████▌ | 9180/10682 [1:24:59<12:28,  2.01it/s]
 86%|████████▌ | 9181/10682 [1:25:00<12:24,  2.02it/s]
 86%|████████▌ | 9182/10682 [1:25:00<12:22,  2.02it/s]
 86%|████████▌ | 9183/10682 [1:25:01<12:21,  2.02it/s]
 86%|████████▌ | 9184/10682 [1:25:01<12:20,  2.02it/s]
 86%|████████▌ | 9185/10682 [1:25:02<12:19,  2.02it/s]
 86%|████████▌ | 9186/10682 [1:25:02<12:18,  2.03it/s]
 86%|████████▌ | 9187/10682 [1:25:03<12:17,  2.03it/s]
 86%|████████▌ | 9188/10682 [1:25:03<12:16,  2.03it/s]
 86%|████████▌ | 9189/10682 [1:25:04<12:16,  2.03it/s]
 86%|████████▌ | 9190/10682 [1:25:04<12:16,  2.03it/s]
 86%|████████▌ | 9191/10682 [1:25:05<12:16,  2.02it/s]
 86%|████████▌ | 9192/10682 [1:25:05<12:15,  2.03it/s]
 86%|████████▌ | 9193/10682 [1:25:06<12:14,  2.03it/s]
\ No newline at end of file
+
 86%|████████▌ | 9175/10682 [1:24:57<13:02,  1.92it/s]
 86%|████████▌ | 9176/10682 [1:24:58<12:52,  1.95it/s]
 86%|████████▌ | 9177/10682 [1:24:58<12:43,  1.97it/s]
 86%|████████▌ | 9178/10682 [1:24:59<12:35,  1.99it/s]
 86%|████████▌ | 9179/10682 [1:24:59<12:32,  2.00it/s]
 86%|████████▌ | 9180/10682 [1:24:59<12:28,  2.01it/s]
 86%|████████▌ | 9181/10682 [1:25:00<12:24,  2.02it/s]
 86%|████████▌ | 9182/10682 [1:25:00<12:22,  2.02it/s]
 86%|████████▌ | 9183/10682 [1:25:01<12:21,  2.02it/s]
 86%|████████▌ | 9184/10682 [1:25:01<12:20,  2.02it/s]
 86%|████████▌ | 9185/10682 [1:25:02<12:19,  2.02it/s]
 86%|████████▌ | 9186/10682 [1:25:02<12:18,  2.03it/s]
 86%|████████▌ | 9187/10682 [1:25:03<12:17,  2.03it/s]
 86%|████████▌ | 9188/10682 [1:25:03<12:16,  2.03it/s]
 86%|████████▌ | 9189/10682 [1:25:04<12:16,  2.03it/s]
 86%|████████▌ | 9190/10682 [1:25:04<12:16,  2.03it/s]
 86%|████████▌ | 9191/10682 [1:25:05<12:16,  2.02it/s]
 86%|████████▌ | 9192/10682 [1:25:05<12:15,  2.03it/s]
 86%|████████▌ | 9193/10682 [1:25:06<12:14,  2.03it/s]
 86%|████████▌ | 9194/10682 [1:25:06<12:14,  2.03it/s]
 86%|████████▌ | 9195/10682 [1:25:07<12:12,  2.03it/s]
 86%|████████▌ | 9196/10682 [1:25:07<12:12,  2.03it/s]
 86%|████████▌ | 9197/10682 [1:25:08<12:12,  2.03it/s]
 86%|████████▌ | 9198/10682 [1:25:08<12:10,  2.03it/s]
 86%|████████▌ | 9199/10682 [1:25:09<12:10,  2.03it/s]
 86%|████████▌ | 9200/10682 [1:25:09<12:10,  2.03it/s]
                                                      
{'loss': 2.7927, 'grad_norm': 0.26096153259277344, 'learning_rate': 5.7505864256519716e-05, 'epoch': 12.06}
+
 86%|████████▌ | 9200/10682 [1:25:09<12:10,  2.03it/s]
 86%|████████▌ | 9201/10682 [1:25:10<12:09,  2.03it/s]
 86%|████████▌ | 9202/10682 [1:25:10<12:09,  2.03it/s]
 86%|████████▌ | 9203/10682 [1:25:11<12:09,  2.03it/s]
 86%|████████▌ | 9204/10682 [1:25:11<12:10,  2.02it/s]
 86%|████████▌ | 9205/10682 [1:25:12<12:09,  2.03it/s]
 86%|████████▌ | 9206/10682 [1:25:12<12:09,  2.02it/s]
 86%|████████▌ | 9207/10682 [1:25:13<12:08,  2.03it/s]
 86%|████████▌ | 9208/10682 [1:25:13<12:07,  2.03it/s]
 86%|████████▌ | 9209/10682 [1:25:14<12:06,  2.03it/s]
 86%|████████▌ | 9210/10682 [1:25:14<12:05,  2.03it/s]
 86%|████████▌ | 9211/10682 [1:25:15<12:04,  2.03it/s]
 86%|████████▌ | 9212/10682 [1:25:15<12:03,  2.03it/s]
 86%|████████▌ | 9213/10682 [1:25:16<12:03,  2.03it/s]
 86%|████████▋ | 9214/10682 [1:25:16<12:02,  2.03it/s]
 86%|████████▋ | 9215/10682 [1:25:17<12:02,  2.03it/s]
 86%|████████▋ | 9216/10682 [1:25:17<12:03,  2.03it/s]
 86%|████████▋ | 9217/10682 [1:25:18<12:01,  2.03it/s]
 86%|████████▋ | 9218/10682 [1:25:18<12:02,  2.03it/s]
 86%|████████▋ | 9219/10682 [1:25:19<12:01,  2.03it/s]
 86%|████████▋ | 9220/10682 [1:25:19<12:00,  2.03it/s]
 86%|████████▋ | 9221/10682 [1:25:20<11:59,  2.03it/s]
 86%|████████▋ | 9222/10682 [1:25:20<11:58,  2.03it/s]
 86%|████████▋ | 9223/10682 [1:25:21<11:58,  2.03it/s]
 86%|████████▋ | 9224/10682 [1:25:21<11:58,  2.03it/s]
 86%|████████▋ | 9225/10682 [1:25:22<11:57,  2.03it/s]{'loss': 2.7828, 'grad_norm': 0.26214075088500977, 'learning_rate': 5.561858464291258e-05, 'epoch': 12.09}
+
                                                      

 86%|████████▋ | 9225/10682 [1:25:22<11:57,  2.03it/s]
 86%|████████▋ | 9226/10682 [1:25:22<11:58,  2.03it/s]
 86%|████████▋ | 9227/10682 [1:25:23<11:56,  2.03it/s]
 86%|████████▋ | 9228/10682 [1:25:23<11:57,  2.03it/s]
 86%|████████▋ | 9229/10682 [1:25:24<11:56,  2.03it/s]
 86%|████████▋ | 9230/10682 [1:25:24<11:56,  2.03it/s]
 86%|████████▋ | 9231/10682 [1:25:25<11:55,  2.03it/s]
 86%|████████▋ | 9232/10682 [1:25:25<11:54,  2.03it/s]
 86%|████████▋ | 9233/10682 [1:25:26<11:54,  2.03it/s]
 86%|████████▋ | 9234/10682 [1:25:26<11:52,  2.03it/s]
 86%|████████▋ | 9235/10682 [1:25:27<11:52,  2.03it/s]
 86%|████████▋ | 9236/10682 [1:25:27<11:52,  2.03it/s]
 86%|████████▋ | 9237/10682 [1:25:28<11:52,  2.03it/s]
 86%|████████▋ | 9238/10682 [1:25:28<11:51,  2.03it/s]
 86%|████████▋ | 9239/10682 [1:25:29<11:50,  2.03it/s]
 87%|████████▋ | 9240/10682 [1:25:29<11:50,  2.03it/s]
 87%|████████▋ | 9241/10682 [1:25:30<11:49,  2.03it/s]
 87%|████████▋ | 9242/10682 [1:25:30<11:49,  2.03it/s]
 87%|████████▋ | 9243/10682 [1:25:31<11:49,  2.03it/s]
 87%|████████▋ | 9244/10682 [1:25:31<11:48,  2.03it/s]
 87%|████████▋ | 9245/10682 [1:25:32<11:48,  2.03it/s]
 87%|████████▋ | 9246/10682 [1:25:32<11:47,  2.03it/s]
 87%|████████▋ | 9247/10682 [1:25:33<11:47,  2.03it/s]
 87%|████████▋ | 9248/10682 [1:25:33<11:46,  2.03it/s]
 87%|████████▋ | 9249/10682 [1:25:34<11:46,  2.03it/s]
 87%|████████▋ | 9250/10682 [1:25:34<11:45,  2.03it/s]
                                                      {'loss': 2.7998, 'grad_norm': 0.2620319128036499, 'learning_rate': 5.3760968048319145e-05, 'epoch': 12.12}
+
 87%|████████▋ | 9250/10682 [1:25:34<11:45,  2.03it/s]
 87%|████████▋ | 9251/10682 [1:25:34<11:45,  2.03it/s]
 87%|████████▋ | 9252/10682 [1:25:35<11:44,  2.03it/s]
 87%|████████▋ | 9253/10682 [1:25:35<11:44,  2.03it/s]
 87%|████████▋ | 9254/10682 [1:25:36<11:43,  2.03it/s]
 87%|████████▋ | 9255/10682 [1:25:36<11:43,  2.03it/s]
 87%|████████▋ | 9256/10682 [1:25:37<11:42,  2.03it/s]
 87%|████████▋ | 9257/10682 [1:25:37<11:42,  2.03it/s]
 87%|████████▋ | 9258/10682 [1:25:38<11:41,  2.03it/s]
 87%|████████▋ | 9259/10682 [1:25:38<11:41,  2.03it/s]
 87%|████████▋ | 9260/10682 [1:25:39<11:41,  2.03it/s]
 87%|████████▋ | 9261/10682 [1:25:39<11:39,  2.03it/s]
 87%|████████▋ | 9262/10682 [1:25:40<11:40,  2.03it/s]
 87%|████████▋ | 9263/10682 [1:25:40<11:40,  2.03it/s]
 87%|████████▋ | 9264/10682 [1:25:41<11:39,  2.03it/s]
 87%|████████▋ | 9265/10682 [1:25:41<11:38,  2.03it/s]
 87%|████████▋ | 9266/10682 [1:25:42<11:37,  2.03it/s]
 87%|████████▋ | 9267/10682 [1:25:42<11:38,  2.03it/s]
 87%|████████▋ | 9268/10682 [1:25:43<11:37,  2.03it/s]
 87%|████████▋ | 9269/10682 [1:25:43<11:37,  2.03it/s]
 87%|████████▋ | 9270/10682 [1:25:44<11:36,  2.03it/s]
 87%|████████▋ | 9271/10682 [1:25:44<11:36,  2.03it/s]
 87%|████████▋ | 9272/10682 [1:25:45<11:35,  2.03it/s]
 87%|████████▋ | 9273/10682 [1:25:45<11:34,  2.03it/s]
 87%|████████▋ | 9274/10682 [1:25:46<11:34,  2.03it/s]
 87%|████████▋ | 9275/10682 [1:25:46<11:33,  2.03it/s]{'loss': 2.7993, 'grad_norm': 0.261679470539093, 'learning_rate': 5.193313847098613e-05, 'epoch': 12.16}
+
                                                      

 87%|████████▋ | 9275/10682 [1:25:46<11:33,  2.03it/s]
 87%|████████▋ | 9276/10682 [1:25:47<11:34,  2.02it/s]
 87%|████████▋ | 9277/10682 [1:25:47<11:33,  2.03it/s]
 87%|████████▋ | 9278/10682 [1:25:48<11:32,  2.03it/s]
 87%|████████▋ | 9279/10682 [1:25:48<11:31,  2.03it/s]
 87%|████████▋ | 9280/10682 [1:25:49<11:30,  2.03it/s]
 87%|████████▋ | 9281/10682 [1:25:49<11:30,  2.03it/s]
 87%|████████▋ | 9282/10682 [1:25:50<11:29,  2.03it/s]
 87%|████████▋ | 9283/10682 [1:25:50<11:29,  2.03it/s]
 87%|████████▋ | 9284/10682 [1:25:51<11:29,  2.03it/s]
 87%|████████▋ | 9285/10682 [1:25:51<11:27,  2.03it/s]
 87%|████████▋ | 9286/10682 [1:25:52<11:28,  2.03it/s]
 87%|████████▋ | 9287/10682 [1:25:52<11:27,  2.03it/s]
 87%|████████▋ | 9288/10682 [1:25:53<11:26,  2.03it/s]
 87%|████████▋ | 9289/10682 [1:25:53<11:26,  2.03it/s]
 87%|████████▋ | 9290/10682 [1:25:54<11:26,  2.03it/s]
 87%|████████▋ | 9291/10682 [1:25:54<11:26,  2.03it/s]
 87%|████████▋ | 9292/10682 [1:25:55<11:24,  2.03it/s]
 87%|████████▋ | 9293/10682 [1:25:55<11:23,  2.03it/s]
 87%|████████▋ | 9294/10682 [1:25:56<11:23,  2.03it/s]
 87%|████████▋ | 9295/10682 [1:25:56<11:22,  2.03it/s]
 87%|████████▋ | 9296/10682 [1:25:57<11:22,  2.03it/s]
 87%|████████▋ | 9297/10682 [1:25:57<11:22,  2.03it/s]
 87%|████████▋ | 9298/10682 [1:25:58<11:20,  2.03it/s]
 87%|████████▋ | 9299/10682 [1:25:58<11:20,  2.03it/s]
 87%|████████▋ | 9300/10682 [1:25:59<11:20,  2.03it/s]{'loss': 2.7991, 'grad_norm': 0.2591189742088318, 'learning_rate': 5.0135217920839137e-05, 'epoch': 12.19}
+
                                                      

 87%|████████▋ | 9300/10682 [1:25:59<11:20,  2.03it/s]
 87%|████████▋ | 9301/10682 [1:25:59<11:19,  2.03it/s]
 87%|████████▋ | 9302/10682 [1:26:00<11:20,  2.03it/s]
 87%|████████▋ | 9303/10682 [1:26:00<11:19,  2.03it/s]
 87%|████████▋ | 9304/10682 [1:26:01<11:18,  2.03it/s]
 87%|████████▋ | 9305/10682 [1:26:01<11:17,  2.03it/s]
 87%|████████▋ | 9306/10682 [1:26:02<11:16,  2.03it/s]
 87%|████████▋ | 9307/10682 [1:26:02<11:16,  2.03it/s]
 87%|████████▋ | 9308/10682 [1:26:03<11:16,  2.03it/s]
 87%|████████▋ | 9309/10682 [1:26:03<11:15,  2.03it/s]
 87%|████████▋ | 9310/10682 [1:26:04<11:15,  2.03it/s]
 87%|████████▋ | 9311/10682 [1:26:04<11:14,  2.03it/s]
 87%|████████▋ | 9312/10682 [1:26:05<11:14,  2.03it/s]
 87%|████████▋ | 9313/10682 [1:26:05<11:14,  2.03it/s]
 87%|████████▋ | 9314/10682 [1:26:06<11:13,  2.03it/s]
 87%|████████▋ | 9315/10682 [1:26:06<11:13,  2.03it/s]
 87%|████████▋ | 9316/10682 [1:26:07<11:12,  2.03it/s]
 87%|████████▋ | 9317/10682 [1:26:07<11:12,  2.03it/s]
 87%|████████▋ | 9318/10682 [1:26:07<11:11,  2.03it/s]
 87%|████████▋ | 9319/10682 [1:26:08<11:10,  2.03it/s]
 87%|████████▋ | 9320/10682 [1:26:08<11:09,  2.03it/s]
 87%|████████▋ | 9321/10682 [1:26:09<11:10,  2.03it/s]
 87%|████████▋ | 9322/10682 [1:26:09<11:09,  2.03it/s]
 87%|████████▋ | 9323/10682 [1:26:10<11:09,  2.03it/s]
 87%|████████▋ | 9324/10682 [1:26:10<11:08,  2.03it/s]
 87%|████████▋ | 9325/10682 [1:26:11<11:08,  2.03it/s]
                                                      {'loss': 2.8032, 'grad_norm': 0.25974202156066895, 'learning_rate': 4.836732641133895e-05, 'epoch': 12.22}
+
 87%|████████▋ | 9325/10682 [1:26:11<11:08,  2.03it/s]
 87%|████████▋ | 9326/10682 [1:26:11<11:08,  2.03it/s]
 87%|████████▋ | 9327/10682 [1:26:12<11:07,  2.03it/s]
 87%|████████▋ | 9328/10682 [1:26:12<11:07,  2.03it/s]
 87%|████████▋ | 9329/10682 [1:26:13<11:06,  2.03it/s]
 87%|████████▋ | 9330/10682 [1:26:13<11:06,  2.03it/s]
 87%|████████▋ | 9331/10682 [1:26:14<11:05,  2.03it/s]
 87%|████████▋ | 9332/10682 [1:26:14<11:04,  2.03it/s]
 87%|████████▋ | 9333/10682 [1:26:15<11:04,  2.03it/s]
 87%|████████▋ | 9334/10682 [1:26:15<11:03,  2.03it/s]
 87%|████████▋ | 9335/10682 [1:26:16<11:03,  2.03it/s]
 87%|████████▋ | 9336/10682 [1:26:16<11:02,  2.03it/s]
 87%|████████▋ | 9337/10682 [1:26:17<11:01,  2.03it/s]
 87%|████████▋ | 9338/10682 [1:26:17<11:02,  2.03it/s]
 87%|████████▋ | 9339/10682 [1:26:18<11:01,  2.03it/s]
 87%|████████▋ | 9340/10682 [1:26:18<10:59,  2.03it/s]
 87%|████████▋ | 9341/10682 [1:26:19<11:00,  2.03it/s]
 87%|████████▋ | 9342/10682 [1:26:19<10:59,  2.03it/s]
 87%|████████▋ | 9343/10682 [1:26:20<10:59,  2.03it/s]
 87%|████████▋ | 9344/10682 [1:26:20<10:58,  2.03it/s]
 87%|████████▋ | 9345/10682 [1:26:21<10:58,  2.03it/s]
 87%|████████▋ | 9346/10682 [1:26:21<10:58,  2.03it/s]
 88%|████████▊ | 9347/10682 [1:26:22<10:57,  2.03it/s]
 88%|████████▊ | 9348/10682 [1:26:22<10:57,  2.03it/s]
 88%|████████▊ | 9349/10682 [1:26:23<10:56,  2.03it/s]
 88%|████████▊ | 9350/10682 [1:26:23<10:56,  2.03it/s]{'loss': 2.7992, 'grad_norm': 0.26328521966934204, 'learning_rate': 4.662958195146971e-05, 'epoch': 12.25}
+
                                                      

 88%|████████▊ | 9350/10682 [1:26:23<10:56,  2.03it/s]
 88%|████████▊ | 9351/10682 [1:26:24<10:57,  2.03it/s]
 88%|████████▊ | 9352/10682 [1:26:24<10:55,  2.03it/s]
 88%|████████▊ | 9353/10682 [1:26:25<10:55,  2.03it/s]
 88%|████████▊ | 9354/10682 [1:26:25<10:54,  2.03it/s]
 88%|████████▊ | 9355/10682 [1:26:26<10:54,  2.03it/s]
 88%|████████▊ | 9356/10682 [1:26:26<10:53,  2.03it/s]
 88%|████████▊ | 9357/10682 [1:26:27<10:52,  2.03it/s]
 88%|████████▊ | 9358/10682 [1:26:27<10:52,  2.03it/s]
 88%|████████▊ | 9359/10682 [1:26:28<10:51,  2.03it/s]
 88%|████████▊ | 9360/10682 [1:26:28<10:51,  2.03it/s]
 88%|████████▊ | 9361/10682 [1:26:29<10:50,  2.03it/s]
 88%|████████▊ | 9362/10682 [1:26:29<10:49,  2.03it/s]
 88%|████████▊ | 9363/10682 [1:26:30<10:49,  2.03it/s]
 88%|████████▊ | 9364/10682 [1:26:30<10:48,  2.03it/s]
 88%|████████▊ | 9365/10682 [1:26:31<10:47,  2.03it/s]
 88%|████████▊ | 9366/10682 [1:26:31<10:47,  2.03it/s]
 88%|████████▊ | 9367/10682 [1:26:32<10:46,  2.03it/s]
 88%|████████▊ | 9368/10682 [1:26:32<10:46,  2.03it/s]
 88%|████████▊ | 9369/10682 [1:26:33<10:46,  2.03it/s]
 88%|████████▊ | 9370/10682 [1:26:33<10:45,  2.03it/s]
 88%|████████▊ | 9371/10682 [1:26:34<10:45,  2.03it/s]
 88%|████████▊ | 9372/10682 [1:26:34<10:44,  2.03it/s]
 88%|████████▊ | 9373/10682 [1:26:35<10:43,  2.03it/s]
 88%|████████▊ | 9374/10682 [1:26:35<10:44,  2.03it/s]
 88%|████████▊ | 9375/10682 [1:26:36<10:43,  2.03it/s]
                                                      
{'loss': 2.7952, 'grad_norm': 0.260550320148468, 'learning_rate': 4.492210053786228e-05, 'epoch': 12.29}
+
 88%|████████▊ | 9375/10682 [1:26:36<10:43,  2.03it/s]
 88%|████████▊ | 9376/10682 [1:26:36<10:44,  2.03it/s]
 88%|████████▊ | 9377/10682 [1:26:37<10:43,  2.03it/s]
 88%|████████▊ | 9378/10682 [1:26:37<10:42,  2.03it/s]
 88%|████████▊ | 9379/10682 [1:26:38<10:41,  2.03it/s]
 88%|████████▊ | 9380/10682 [1:26:38<10:41,  2.03it/s]
 88%|████████▊ | 9381/10682 [1:26:39<10:40,  2.03it/s]
 88%|████████▊ | 9382/10682 [1:26:39<10:40,  2.03it/s]
 88%|████████▊ | 9383/10682 [1:26:40<10:39,  2.03it/s]
 88%|████████▊ | 9384/10682 [1:26:40<10:39,  2.03it/s]
 88%|████████▊ | 9385/10682 [1:26:40<10:39,  2.03it/s]
 88%|████████▊ | 9386/10682 [1:26:41<10:38,  2.03it/s]
 88%|████████▊ | 9387/10682 [1:26:41<10:38,  2.03it/s]
 88%|████████▊ | 9388/10682 [1:26:42<10:37,  2.03it/s]
 88%|████████▊ | 9389/10682 [1:26:42<10:36,  2.03it/s]
 88%|████████▊ | 9390/10682 [1:26:43<10:36,  2.03it/s]
 88%|████████▊ | 9391/10682 [1:26:43<10:35,  2.03it/s]
 88%|████████▊ | 9392/10682 [1:26:44<10:35,  2.03it/s]
 88%|████████▊ | 9393/10682 [1:26:44<10:34,  2.03it/s]
 88%|█��██████▊ | 9394/10682 [1:26:45<10:33,  2.03it/s]
 88%|████████▊ | 9395/10682 [1:26:45<10:33,  2.03it/s]
 88%|████████▊ | 9396/10682 [1:26:46<10:32,  2.03it/s]
 88%|████████▊ | 9397/10682 [1:26:46<10:32,  2.03it/s]
 88%|████████▊ | 9398/10682 [1:26:47<10:31,  2.03it/s]
 88%|████████▊ | 9399/10682 [1:26:47<10:31,  2.03it/s]
 88%|████████▊ | 9400/10682 [1:26:48<10:30,  2.03it/s]
                                                      {'loss': 2.81, 'grad_norm': 0.26355794072151184, 'learning_rate': 4.3244996147050855e-05, 'epoch': 12.32}
+
 88%|████████▊ | 9400/10682 [1:26:48<10:30,  2.03it/s]
 88%|████████▊ | 9401/10682 [1:26:48<10:31,  2.03it/s]
 88%|████████▊ | 9402/10682 [1:26:49<10:30,  2.03it/s]
 88%|████████▊ | 9403/10682 [1:26:49<10:31,  2.03it/s]
 88%|████████▊ | 9404/10682 [1:26:50<10:30,  2.03it/s]
 88%|████████▊ | 9405/10682 [1:26:50<10:30,  2.03it/s]
 88%|████████▊ | 9406/10682 [1:26:51<10:29,  2.03it/s]
 88%|████████▊ | 9407/10682 [1:26:51<10:28,  2.03it/s]
 88%|████████▊ | 9408/10682 [1:26:52<10:27,  2.03it/s]
 88%|████████▊ | 9409/10682 [1:26:52<10:27,  2.03it/s]
 88%|████████▊ | 9410/10682 [1:26:53<10:26,  2.03it/s]
 88%|████████▊ | 9411/10682 [1:26:53<10:25,  2.03it/s]
 88%|████████▊ | 9412/10682 [1:26:54<10:25,  2.03it/s]
 88%|████████▊ | 9413/10682 [1:26:54<10:24,  2.03it/s]
 88%|████████▊ | 9414/10682 [1:26:55<10:24,  2.03it/s]
 88%|████████▊ | 9415/10682 [1:26:55<10:23,  2.03it/s]
 88%|████████▊ | 9416/10682 [1:26:56<10:23,  2.03it/s]
 88%|████████▊ | 9417/10682 [1:26:56<10:22,  2.03it/s]
 88%|████████▊ | 9418/10682 [1:26:57<10:22,  2.03it/s]
 88%|████████▊ | 9419/10682 [1:26:57<10:21,  2.03it/s]
 88%|████████▊ | 9420/10682 [1:26:58<10:20,  2.03it/s]
 88%|████████▊ | 9421/10682 [1:26:58<10:21,  2.03it/s]
 88%|████████▊ | 9422/10682 [1:26:59<10:20,  2.03it/s]
 88%|████████▊ | 9423/10682 [1:26:59<10:19,  2.03it/s]
 88%|████████▊ | 9424/10682 [1:27:00<10:19,  2.03it/s]
 88%|████████▊ | 9425/10682 [1:27:00<10:18,  2.03it/s]
                                                      {'loss': 2.8002, 'grad_norm': 0.2632872462272644, 'learning_rate': 4.1598380727865315e-05, 'epoch': 12.35}
+
 88%|████████▊ | 9425/10682 [1:27:00<10:18,  2.03it/s]
 88%|████████▊ | 9426/10682 [1:27:01<10:19,  2.03it/s]
 88%|████████▊ | 9427/10682 [1:27:01<10:18,  2.03it/s]
 88%|████████▊ | 9428/10682 [1:27:02<10:17,  2.03it/s]
 88%|████████▊ | 9429/10682 [1:27:02<10:17,  2.03it/s]
 88%|████████▊ | 9430/10682 [1:27:03<10:16,  2.03it/s]
 88%|████████▊ | 9431/10682 [1:27:03<10:16,  2.03it/s]
 88%|████████▊ | 9432/10682 [1:27:04<10:16,  2.03it/s]
 88%|████████▊ | 9433/10682 [1:27:04<10:15,  2.03it/s]
 88%|████████▊ | 9434/10682 [1:27:05<10:15,  2.03it/s]
 88%|████████▊ | 9435/10682 [1:27:05<10:14,  2.03it/s]
 88%|████████▊ | 9436/10682 [1:27:06<10:13,  2.03it/s]
 88%|████████▊ | 9437/10682 [1:27:06<10:13,  2.03it/s]
 88%|████████▊ | 9438/10682 [1:27:07<10:12,  2.03it/s]
 88%|████████▊ | 9439/10682 [1:27:07<10:11,  2.03it/s]
 88%|████████▊ | 9440/10682 [1:27:08<10:11,  2.03it/s]
 88%|████████▊ | 9441/10682 [1:27:08<10:10,  2.03it/s]
 88%|████████▊ | 9442/10682 [1:27:09<10:10,  2.03it/s]
 88%|████████▊ | 9443/10682 [1:27:09<10:09,  2.03it/s]
 88%|████████▊ | 9444/10682 [1:27:10<10:09,  2.03it/s]
 88%|████████▊ | 9445/10682 [1:27:10<10:09,  2.03it/s]
 88%|████████▊ | 9446/10682 [1:27:11<10:08,  2.03it/s]
 88%|████████▊ | 9447/10682 [1:27:11<10:07,  2.03it/s]
 88%|████████▊ | 9448/10682 [1:27:12<10:07,  2.03it/s]
 88%|████████▊ | 9449/10682 [1:27:12<10:06,  2.03it/s]
 88%|████████▊ | 9450/10682 [1:27:13<10:06,  2.03it/s]{'loss': 2.7965, 'grad_norm': 0.2634727954864502, 'learning_rate': 3.998236419395806e-05, 'epoch': 12.39}
+
                                                      

 88%|████████▊ | 9450/10682 [1:27:13<10:06,  2.03it/s]
 88%|████████▊ | 9451/10682 [1:27:13<10:06,  2.03it/s]
 88%|████████▊ | 9452/10682 [1:27:13<10:06,  2.03it/s]
 88%|████████▊ | 9453/10682 [1:27:14<10:58,  1.87it/s]
 89%|████████▊ | 9454/10682 [1:27:15<10:42,  1.91it/s]
 89%|████████▊ | 9455/10682 [1:27:15<10:30,  1.94it/s]
 89%|████████▊ | 9456/10682 [1:27:16<10:22,  1.97it/s]
 89%|████████▊ | 9457/10682 [1:27:16<10:16,  1.99it/s]
 89%|████████▊ | 9458/10682 [1:27:17<10:11,  2.00it/s]
 89%|████████▊ | 9459/10682 [1:27:17<10:08,  2.01it/s]
 89%|████████▊ | 9460/10682 [1:27:18<10:06,  2.02it/s]
 89%|████████▊ | 9461/10682 [1:27:18<10:03,  2.02it/s]
 89%|████████▊ | 9462/10682 [1:27:19<10:03,  2.02it/s]
 89%|████████▊ | 9463/10682 [1:27:19<10:02,  2.02it/s]
 89%|████████▊ | 9464/10682 [1:27:20<10:01,  2.02it/s]
 89%|████████▊ | 9465/10682 [1:27:20<10:00,  2.03it/s]
 89%|████████▊ | 9466/10682 [1:27:21<09:59,  2.03it/s]
 89%|████████▊ | 9467/10682 [1:27:21<09:58,  2.03it/s]
 89%|████████▊ | 9468/10682 [1:27:22<09:57,  2.03it/s]
 89%|████████▊ | 9469/10682 [1:27:22<09:57,  2.03it/s]
 89%|████████▊ | 9470/10682 [1:27:22<09:56,  2.03it/s]
 89%|████████▊ | 9471/10682 [1:27:23<09:55,  2.03it/s]
 89%|████████▊ | 9472/10682 [1:27:23<09:55,  2.03it/s]
 89%|████████▊ | 9473/10682 [1:27:24<09:55,  2.03it/s]
 89%|████████▊ | 9474/10682 [1:27:24<09:54,  2.03it/s]
 89%|████████▊ | 9475/10682 [1:27:25<09:53,  2.03it/s]
                                                      
{'loss': 2.8, 'grad_norm': 0.2608312964439392, 'learning_rate': 3.839705441646779e-05, 'epoch': 12.42}
+
 89%|████████▊ | 9475/10682 [1:27:25<09:53,  2.03it/s]
 89%|████████▊ | 9476/10682 [1:27:25<09:54,  2.03it/s]
 89%|████████▊ | 9477/10682 [1:27:26<09:53,  2.03it/s]
 89%|████████▊ | 9478/10682 [1:27:26<09:52,  2.03it/s]
 89%|████████▊ | 9479/10682 [1:27:27<09:52,  2.03it/s]
 89%|████████▊ | 9480/10682 [1:27:27<09:51,  2.03it/s]
 89%|████████▉ | 9481/10682 [1:27:28<09:51,  2.03it/s]
 89%|████████▉ | 9482/10682 [1:27:28<09:51,  2.03it/s]
 89%|████████▉ | 9483/10682 [1:27:29<09:50,  2.03it/s]
 89%|████████▉ | 9484/10682 [1:27:29<09:49,  2.03it/s]
 89%|████████▉ | 9485/10682 [1:27:30<09:49,  2.03it/s]
 89%|████████▉ | 9486/10682 [1:27:30<09:48,  2.03it/s]
 89%|████████▉ | 9487/10682 [1:27:31<09:48,  2.03it/s]
 89%|████████▉ | 9488/10682 [1:27:31<09:47,  2.03it/s]
 89%|████████▉ | 9489/10682 [1:27:32<09:47,  2.03it/s]
 89%|████████▉ | 9490/10682 [1:27:32<09:46,  2.03it/s]
 89%|████████▉ | 9491/10682 [1:27:33<09:46,  2.03it/s]
 89%|████████▉ | 9492/10682 [1:27:33<09:46,  2.03it/s]
 89%|████████▉ | 9493/10682 [1:27:34<09:44,  2.03it/s]
 89%|████████▉ | 9494/10682 [1:27:34<09:44,  2.03it/s]
 89%|████████▉ | 9495/10682 [1:27:35<09:44,  2.03it/s]
 89%|████████▉ | 9496/10682 [1:27:35<09:43,  2.03it/s]
 89%|████████▉ | 9497/10682 [1:27:36<09:43,  2.03it/s]
 89%|████████▉ | 9498/10682 [1:27:36<09:42,  2.03it/s]
 89%|████████▉ | 9499/10682 [1:27:37<09:41,  2.03it/s]
 89%|████████▉ | 9500/10682 [1:27:37<09:41,  2.03it/s]{'loss': 2.8173, 'grad_norm': 0.2660999596118927, 'learning_rate': 3.6842557216818006e-05, 'epoch': 12.45}
+
                                                      

 89%|████████▉ | 9500/10682 [1:27:37<09:41,  2.03it/s]
 89%|████████▉ | 9501/10682 [1:27:38<09:41,  2.03it/s]
 89%|████████▉ | 9502/10682 [1:27:38<09:40,  2.03it/s]
 89%|████████▉ | 9503/10682 [1:27:39<09:40,  2.03it/s]
 89%|████████▉ | 9504/10682 [1:27:39<09:39,  2.03it/s]
 89%|████████▉ | 9505/10682 [1:27:40<09:38,  2.03it/s]
 89%|████████▉ | 9506/10682 [1:27:40<09:38,  2.03it/s]
 89%|████████▉ | 9507/10682 [1:27:41<09:37,  2.03it/s]
 89%|████████▉ | 9508/10682 [1:27:41<09:38,  2.03it/s]
 89%|████████▉ | 9509/10682 [1:27:42<09:37,  2.03it/s]
 89%|████████▉ | 9510/10682 [1:27:42<09:37,  2.03it/s]
 89%|████████▉ | 9511/10682 [1:27:43<09:36,  2.03it/s]
 89%|████████▉ | 9512/10682 [1:27:43<09:35,  2.03it/s]
 89%|████████▉ | 9513/10682 [1:27:44<09:35,  2.03it/s]
 89%|████████▉ | 9514/10682 [1:27:44<09:35,  2.03it/s]
 89%|████████▉ | 9515/10682 [1:27:45<09:34,  2.03it/s]
 89%|████████▉ | 9516/10682 [1:27:45<09:34,  2.03it/s]
 89%|████████▉ | 9517/10682 [1:27:46<09:33,  2.03it/s]
 89%|████��███▉ | 9518/10682 [1:27:46<09:32,  2.03it/s]
 89%|████████▉ | 9519/10682 [1:27:47<09:32,  2.03it/s]
 89%|████████▉ | 9520/10682 [1:27:47<09:31,  2.03it/s]
 89%|████████▉ | 9521/10682 [1:27:48<09:31,  2.03it/s]
 89%|████████▉ | 9522/10682 [1:27:48<09:30,  2.03it/s]
 89%|████████▉ | 9523/10682 [1:27:49<09:29,  2.03it/s]
 89%|████████▉ | 9524/10682 [1:27:49<09:30,  2.03it/s]
 89%|████████▉ | 9525/10682 [1:27:50<09:29,  2.03it/s]
                                                      {'loss': 2.8097, 'grad_norm': 0.2629110813140869, 'learning_rate': 3.531897635965431e-05, 'epoch': 12.48}
+
 89%|████████▉ | 9525/10682 [1:27:50<09:29,  2.03it/s]
 89%|████████▉ | 9526/10682 [1:27:50<09:30,  2.03it/s]
 89%|████████▉ | 9527/10682 [1:27:51<09:29,  2.03it/s]
 89%|████████▉ | 9528/10682 [1:27:51<09:28,  2.03it/s]
 89%|████████▉ | 9529/10682 [1:27:52<09:28,  2.03it/s]
 89%|████████▉ | 9530/10682 [1:27:52<09:27,  2.03it/s]
 89%|████████▉ | 9531/10682 [1:27:53<09:27,  2.03it/s]
 89%|████████▉ | 9532/10682 [1:27:53<09:26,  2.03it/s]
 89%|████████▉ | 9533/10682 [1:27:54<09:25,  2.03it/s]
 89%|████████▉ | 9534/10682 [1:27:54<09:25,  2.03it/s]
 89%|████████▉ | 9535/10682 [1:27:55<09:25,  2.03it/s]
 89%|████████▉ | 9536/10682 [1:27:55<09:25,  2.03it/s]
 89%|████████▉ | 9537/10682 [1:27:55<09:24,  2.03it/s]
 89%|████████▉ | 9538/10682 [1:27:56<09:23,  2.03it/s]
 89%|████████▉ | 9539/10682 [1:27:56<09:23,  2.03it/s]
 89%|████████▉ | 9540/10682 [1:27:57<09:22,  2.03it/s]
 89%|████████▉ | 9541/10682 [1:27:57<09:22,  2.03it/s]
 89%|████████▉ | 9542/10682 [1:27:58<09:22,  2.03it/s]
 89%|████████▉ | 9543/10682 [1:27:58<09:21,  2.03it/s]
 89%|████████▉ | 9544/10682 [1:27:59<09:20,  2.03it/s]
 89%|████████▉ | 9545/10682 [1:27:59<09:20,  2.03it/s]
 89%|████████▉ | 9546/10682 [1:28:00<09:20,  2.03it/s]
 89%|████████▉ | 9547/10682 [1:28:00<09:19,  2.03it/s]
 89%|████████▉ | 9548/10682 [1:28:01<09:18,  2.03it/s]
 89%|████████▉ | 9549/10682 [1:28:01<09:18,  2.03it/s]
 89%|████████▉ | 9550/10682 [1:28:02<09:17,  2.03it/s]
                                                      {'loss': 2.8177, 'grad_norm': 0.26221880316734314, 'learning_rate': 3.382641354591731e-05, 'epoch': 12.52}
+
 89%|████████▉ | 9550/10682 [1:28:02<09:17,  2.03it/s]
 89%|████████▉ | 9551/10682 [1:28:02<09:17,  2.03it/s]
 89%|████████▉ | 9552/10682 [1:28:03<09:17,  2.03it/s]
 89%|████████▉ | 9553/10682 [1:28:03<09:16,  2.03it/s]
 89%|████████▉ | 9554/10682 [1:28:04<09:15,  2.03it/s]
 89%|████████▉ | 9555/10682 [1:28:04<09:15,  2.03it/s]
 89%|████████▉ | 9556/10682 [1:28:05<09:14,  2.03it/s]
 89%|████████▉ | 9557/10682 [1:28:05<09:14,  2.03it/s]
 89%|████████▉ | 9558/10682 [1:28:06<09:12,  2.03it/s]
 89%|████████▉ | 9559/10682 [1:28:06<09:13,  2.03it/s]
 89%|████████▉ | 9560/10682 [1:28:07<09:12,  2.03it/s]
 90%|████████▉ | 9561/10682 [1:28:07<09:11,  2.03it/s]
 90%|████████▉ | 9562/10682 [1:28:08<09:11,  2.03it/s]
 90%|████████▉ | 9563/10682 [1:28:08<09:10,  2.03it/s]
 90%|████████▉ | 9564/10682 [1:28:09<09:09,  2.03it/s]
 90%|████████▉ | 9565/10682 [1:28:09<09:09,  2.03it/s]
 90%|████████▉ | 9566/10682 [1:28:10<09:08,  2.03it/s]
 90%|████████▉ | 9567/10682 [1:28:10<09:08,  2.03it/s]
 90%|████████▉ | 9568/10682 [1:28:11<09:08,  2.03it/s]
 90%|████████▉ | 9569/10682 [1:28:11<09:54,  1.87it/s]
 90%|████████▉ | 9570/10682 [1:28:12<09:40,  1.92it/s]
 90%|████████▉ | 9571/10682 [1:28:12<09:30,  1.95it/s]
 90%|████████▉ | 9572/10682 [1:28:13<09:22,  1.97it/s]
 90%|████████▉ | 9573/10682 [1:28:13<09:17,  1.99it/s]
 90%|████████▉ | 9574/10682 [1:28:14<09:13,  2.00it/s]
 90%|████████▉ | 9575/10682 [1:28:14<09:10,  2.01it/s]{'loss': 2.809, 'grad_norm': 0.26215025782585144, 'learning_rate': 3.2364968406054075e-05, 'epoch': 12.55}
                                                      
+
 90%|████████▉ | 9575/10682 [1:28:14<09:10,  2.01it/s]
 90%|████████▉ | 9576/10682 [1:28:15<09:09,  2.01it/s]
 90%|████████▉ | 9577/10682 [1:28:15<09:06,  2.02it/s]
 90%|████████▉ | 9578/10682 [1:28:16<09:06,  2.02it/s]
 90%|████████▉ | 9579/10682 [1:28:16<09:05,  2.02it/s]
 90%|████████▉ | 9580/10682 [1:28:17<09:03,  2.03it/s]
 90%|████████▉ | 9581/10682 [1:28:17<09:03,  2.03it/s]
 90%|████████▉ | 9582/10682 [1:28:18<09:02,  2.03it/s]
 90%|████████▉ | 9583/10682 [1:28:18<09:01,  2.03it/s]
 90%|████████▉ | 9584/10682 [1:28:19<09:01,  2.03it/s]
 90%|████████▉ | 9585/10682 [1:28:19<09:00,  2.03it/s]
 90%|████████▉ | 9586/10682 [1:28:20<09:00,  2.03it/s]
 90%|████████▉ | 9587/10682 [1:28:20<08:59,  2.03it/s]
 90%|████████▉ | 9588/10682 [1:28:21<08:59,  2.03it/s]
 90%|████████▉ | 9589/10682 [1:28:21<08:58,  2.03it/s]
 90%|████████▉ | 9590/10682 [1:28:22<08:58,  2.03it/s]
 90%|████████▉ | 9591/10682 [1:28:22<08:57,  2.03it/s]
 90%|████████▉ | 9592/10682 [1:28:23<08:57,  2.03it/s]
 90%|████████▉ | 9593/10682 [1:28:23<08:56,  2.03it/s]
 90%|████████▉ | 9594/10682 [1:28:24<08:56,  2.03it/s]
 90%|████████▉ | 9595/10682 [1:28:24<08:55,  2.03it/s]
 90%|████████▉ | 9596/10682 [1:28:25<08:55,  2.03it/s]
 90%|████████▉ | 9597/10682 [1:28:25<08:54,  2.03it/s]
 90%|████████▉ | 9598/10682 [1:28:26<08:54,  2.03it/s]
 90%|████████▉ | 9599/10682 [1:28:26<08:53,  2.03it/s]
 90%|████████▉ | 9600/10682 [1:28:27<08:53,  2.03it/s]
                                                      
{'loss': 2.8089, 'grad_norm': 0.26071804761886597, 'learning_rate': 3.093473849336781e-05, 'epoch': 12.58}
+
 90%|████████▉ | 9600/10682 [1:28:27<08:53,  2.03it/s]
 90%|████████▉ | 9601/10682 [1:28:27<08:53,  2.03it/s]
 90%|████████▉ | 9602/10682 [1:28:28<08:53,  2.03it/s]
 90%|████████▉ | 9603/10682 [1:28:28<08:52,  2.03it/s]
 90%|████████▉ | 9604/10682 [1:28:29<08:51,  2.03it/s]
 90%|████████▉ | 9605/10682 [1:28:29<08:50,  2.03it/s]
 90%|████████▉ | 9606/10682 [1:28:30<08:50,  2.03it/s]
 90%|████████▉ | 9607/10682 [1:28:30<08:49,  2.03it/s]
 90%|████████▉ | 9608/10682 [1:28:31<08:48,  2.03it/s]
 90%|████████▉ | 9609/10682 [1:28:31<08:48,  2.03it/s]
 90%|████████▉ | 9610/10682 [1:28:32<08:48,  2.03it/s]
 90%|████████▉ | 9611/10682 [1:28:32<08:47,  2.03it/s]
 90%|████████▉ | 9612/10682 [1:28:33<08:46,  2.03it/s]
 90%|████████▉ | 9613/10682 [1:28:33<08:46,  2.03it/s]
 90%|█████████ | 9614/10682 [1:28:34<08:46,  2.03it/s]
 90%|█████████ | 9615/10682 [1:28:34<08:46,  2.03it/s]
 90%|█████████ | 9616/10682 [1:28:35<08:46,  2.03it/s]
 90%|█████████ | 9617/10682 [1:28:35<08:45,  2.03it/s]
 90%|█████████ | 9618/10682 [1:28:36<08:45,  2.03it/s]
 90%|█████████ | 9619/10682 [1:28:36<08:43,  2.03it/s]
 90%|█████████ | 9620/10682 [1:28:37<08:43,  2.03it/s]
 90%|█████████ | 9621/10682 [1:28:37<08:42,  2.03it/s]
 90%|█████████ | 9622/10682 [1:28:38<08:41,  2.03it/s]
 90%|█████████ | 9623/10682 [1:28:38<08:41,  2.03it/s]
 90%|█████████ | 9624/10682 [1:28:38<08:40,  2.03it/s]
 90%|█████████ | 9625/10682 [1:28:39<08:39,  2.03it/s]
                                                      
{'loss': 2.8101, 'grad_norm': 0.2636934220790863, 'learning_rate': 2.9535819277506203e-05, 'epoch': 12.61}
+
 90%|█████████ | 9625/10682 [1:28:39<08:39,  2.03it/s]
 90%|█████████ | 9626/10682 [1:28:39<08:40,  2.03it/s]
 90%|█████████ | 9627/10682 [1:28:40<08:39,  2.03it/s]
 90%|█████████ | 9628/10682 [1:28:40<08:38,  2.03it/s]
 90%|█████████ | 9629/10682 [1:28:41<08:38,  2.03it/s]
 90%|█████████ | 9630/10682 [1:28:41<08:37,  2.03it/s]
 90%|█████████ | 9631/10682 [1:28:42<08:37,  2.03it/s]
 90%|█████████ | 9632/10682 [1:28:42<08:36,  2.03it/s]
 90%|█████████ | 9633/10682 [1:28:43<08:36,  2.03it/s]
 90%|█████████ | 9634/10682 [1:28:43<08:35,  2.03it/s]
 90%|█████████ | 9635/10682 [1:28:44<08:35,  2.03it/s]
 90%|█████████ | 9636/10682 [1:28:44<08:35,  2.03it/s]
 90%|█████████ | 9637/10682 [1:28:45<08:34,  2.03it/s]
 90%|█████████ | 9638/10682 [1:28:45<08:34,  2.03it/s]
 90%|█████████ | 9639/10682 [1:28:46<08:34,  2.03it/s]
 90%|█████████ | 9640/10682 [1:28:46<08:32,  2.03it/s]
 90%|█████████ | 9641/10682 [1:28:47<08:33,  2.03it/s]
 90%|█████████ | 9642/10682 [1:28:47<08:31,  2.03it/s]
 90%|█████████ | 9643/10682 [1:28:48<08:30,  2.03it/s]
 90%|█████████ | 9644/10682 [1:28:48<08:30,  2.03it/s]
 90%|█████████ | 9645/10682 [1:28:49<08:30,  2.03it/s]
 90%|█████████ | 9646/10682 [1:28:49<08:29,  2.03it/s]
 90%|█████████ | 9647/10682 [1:28:50<08:29,  2.03it/s]
 90%|█████████ | 9648/10682 [1:28:50<08:28,  2.03it/s]
 90%|█████████ | 9649/10682 [1:28:51<08:27,  2.03it/s]
 90%|█████████ | 9650/10682 [1:28:51<08:27,  2.03it/s]
                                                      {'loss': 2.8074, 'grad_norm': 0.2595975995063782, 'learning_rate': 2.8168304138088295e-05, 'epoch': 12.65}
+

 90%|█████████ | 9650/10682 [1:28:51<08:27,  2.03it/s]
 90%|█████████ | 9651/10682 [1:28:52<08:27,  2.03it/s]
 90%|█████████ | 9652/10682 [1:28:52<08:27,  2.03it/s]
 90%|█████████ | 9653/10682 [1:28:53<08:26,  2.03it/s]
 90%|█████████ | 9654/10682 [1:28:53<08:26,  2.03it/s]
 90%|█████████ | 9655/10682 [1:28:54<08:25,  2.03it/s]
 90%|█████████ | 9656/10682 [1:28:54<08:24,  2.03it/s]
 90%|█████████ | 9657/10682 [1:28:55<08:24,  2.03it/s]
 90%|█████████ | 9658/10682 [1:28:55<08:24,  2.03it/s]
 90%|█████████ | 9659/10682 [1:28:56<08:23,  2.03it/s]
 90%|█████████ | 9660/10682 [1:28:56<08:23,  2.03it/s]
 90%|█████████ | 9661/10682 [1:28:57<08:22,  2.03it/s]
 90%|█████████ | 9662/10682 [1:28:57<08:21,  2.03it/s]
 90%|█████████ | 9663/10682 [1:28:58<08:21,  2.03it/s]
 90%|█████████ | 9664/10682 [1:28:58<08:21,  2.03it/s]
 90%|█████████ | 9665/10682 [1:28:59<08:21,  2.03it/s]
 90%|█████████ | 9666/10682 [1:28:59<08:20,  2.03it/s]
 90%|█████████ | 9667/10682 [1:29:00<08:20,  2.03it/s]
 91%|█████████ | 9668/10682 [1:29:00<08:19,  2.03it/s]
 91%|█████████ | 9669/10682 [1:29:01<08:18,  2.03it/s]
 91%|█████████ | 9670/10682 [1:29:01<08:18,  2.03it/s]
 91%|█████████ | 9671/10682 [1:29:02<08:17,  2.03it/s]
 91%|█████████ | 9672/10682 [1:29:02<08:17,  2.03it/s]
 91%|█████████ | 9673/10682 [1:29:03<08:17,  2.03it/s]
 91%|█████████ | 9674/10682 [1:29:03<08:16,  2.03it/s]
 91%|█████████ | 9675/10682 [1:29:04<08:16,  2.03it/s]{'loss': 2.8054, 'grad_norm': 0.2622336149215698, 'learning_rate': 2.6832284358471516e-05, 'epoch': 12.68}
+
                                                      

 91%|█████████ | 9675/10682 [1:29:04<08:16,  2.03it/s]
 91%|█████████ | 9676/10682 [1:29:04<08:16,  2.03it/s]
 91%|█████████ | 9677/10682 [1:29:05<08:16,  2.03it/s]
 91%|█████████ | 9678/10682 [1:29:05<08:15,  2.03it/s]
 91%|█████████ | 9679/10682 [1:29:06<08:14,  2.03it/s]
 91%|█████████ | 9680/10682 [1:29:06<08:14,  2.03it/s]
 91%|█████████ | 9681/10682 [1:29:07<08:13,  2.03it/s]
 91%|█████████ | 9682/10682 [1:29:07<08:12,  2.03it/s]
 91%|█████████ | 9683/10682 [1:29:08<08:12,  2.03it/s]
 91%|█████████ | 9684/10682 [1:29:08<08:11,  2.03it/s]
 91%|█████████ | 9685/10682 [1:29:09<08:11,  2.03it/s]
 91%|█████████ | 9686/10682 [1:29:09<08:10,  2.03it/s]
 91%|█████████ | 9687/10682 [1:29:10<08:10,  2.03it/s]
 91%|█████████ | 9688/10682 [1:29:10<08:09,  2.03it/s]
 91%|█████████ | 9689/10682 [1:29:11<08:09,  2.03it/s]
 91%|█████████ | 9690/10682 [1:29:11<08:08,  2.03it/s]
 91%|█████████ | 9691/10682 [1:29:11<08:08,  2.03it/s]
 91%|█████████ | 9692/10682 [1:29:12<08:08,  2.03it/s]
 91%|█████████ | 9693/10682 [1:29:12<08:07,  2.03it/s]
 91%|█████████ | 9694/10682 [1:29:13<08:06,  2.03it/s]
 91%|█████████ | 9695/10682 [1:29:13<08:06,  2.03it/s]
 91%|█████████ | 9696/10682 [1:29:14<08:06,  2.03it/s]
 91%|█████████ | 9697/10682 [1:29:14<08:05,  2.03it/s]
 91%|█████████ | 9698/10682 [1:29:15<08:05,  2.03it/s]
 91%|█████████ | 9699/10682 [1:29:15<08:04,  2.03it/s]
 91%|█████████ | 9700/10682 [1:29:16<08:03,  2.03it/s]{'loss': 2.7967, 'grad_norm': 0.26076486706733704, 'learning_rate': 2.5527849119658387e-05, 'epoch': 12.71}
+
                                                      

 91%|█████████ | 9700/10682 [1:29:16<08:03,  2.03it/s]
 91%|█████████ | 9701/10682 [1:29:16<08:03,  2.03it/s]
 91%|█████████ | 9702/10682 [1:29:17<08:03,  2.03it/s]
 91%|█████████ | 9703/10682 [1:29:17<08:02,  2.03it/s]
 91%|█████████ | 9704/10682 [1:29:18<08:02,  2.03it/s]
 91%|█████████ | 9705/10682 [1:29:18<08:01,  2.03it/s]
 91%|█████████ | 9706/10682 [1:29:19<08:00,  2.03it/s]
 91%|█████████ | 9707/10682 [1:29:19<08:00,  2.03it/s]
 91%|█████████ | 9708/10682 [1:29:20<07:59,  2.03it/s]
 91%|█████████ | 9709/10682 [1:29:20<07:58,  2.03it/s]
 91%|█████████ | 9710/10682 [1:29:21<07:58,  2.03it/s]
 91%|█████████ | 9711/10682 [1:29:21<07:58,  2.03it/s]
 91%|█████████ | 9712/10682 [1:29:22<07:58,  2.03it/s]
 91%|█████████ | 9713/10682 [1:29:22<07:58,  2.03it/s]
 91%|█████████ | 9714/10682 [1:29:23<07:56,  2.03it/s]
 91%|█████████ | 9715/10682 [1:29:23<07:56,  2.03it/s]
 91%|█████████ | 9716/10682 [1:29:24<07:56,  2.03it/s]
 91%|█████████ | 9717/10682 [1:29:24<07:55,  2.03it/s]
 91%|█████████ | 9718/10682 [1:29:25<07:54,  2.03it/s]
 91%|█████████ | 9719/10682 [1:29:25<07:54,  2.03it/s]
 91%|█████████ | 9720/10682 [1:29:26<07:54,  2.03it/s]
 91%|█████████ | 9721/10682 [1:29:26<07:53,  2.03it/s]
 91%|█████████ | 9722/10682 [1:29:27<07:53,  2.03it/s]
 91%|█████████ | 9723/10682 [1:29:27<07:52,  2.03it/s]
 91%|█████████ | 9724/10682 [1:29:28<07:51,  2.03it/s]
 91%|█████████ | 9725/10682 [1:29:28<07:51,  2.03it/s]
                                                      
{'loss': 2.8041, 'grad_norm': 0.25956180691719055, 'learning_rate': 2.4255085494343522e-05, 'epoch': 12.75}
+
 91%|█████████ | 9725/10682 [1:29:28<07:51,  2.03it/s]
 91%|█████████ | 9726/10682 [1:29:29<07:51,  2.03it/s]
 91%|█████████ | 9727/10682 [1:29:29<07:50,  2.03it/s]
 91%|█████████ | 9728/10682 [1:29:30<07:50,  2.03it/s]
 91%|█████████ | 9729/10682 [1:29:30<07:48,  2.03it/s]
 91%|█████████ | 9730/10682 [1:29:31<07:48,  2.03it/s]
 91%|█████████ | 9731/10682 [1:29:31<07:48,  2.03it/s]
 91%|█████████ | 9732/10682 [1:29:32<07:47,  2.03it/s]
 91%|█████████ | 9733/10682 [1:29:32<07:47,  2.03it/s]
 91%|█████████ | 9734/10682 [1:29:33<07:46,  2.03it/s]
 91%|█████████ | 9735/10682 [1:29:33<07:46,  2.03it/s]
 91%|█████████ | 9736/10682 [1:29:34<07:46,  2.03it/s]
 91%|█████████ | 9737/10682 [1:29:34<07:45,  2.03it/s]
 91%|█████████ | 9738/10682 [1:29:35<07:44,  2.03it/s]
 91%|█████████ | 9739/10682 [1:29:35<07:44,  2.03it/s]
 91%|█████████ | 9740/10682 [1:29:36<07:43,  2.03it/s]
 91%|█████████ | 9741/10682 [1:29:36<07:43,  2.03it/s]
 91%|█████████ | 9742/10682 [1:29:37<07:43,  2.03it/s]
 91%|█████████ | 9743/10682 [1:29:37<07:42,  2.03it/s]
 91%|█████████ | 9744/10682 [1:29:38<07:41,  2.03it/s]
 91%|█████████ | 9745/10682 [1:29:38<07:41,  2.03it/s]
 91%|█████████ | 9746/10682 [1:29:39<07:41,  2.03it/s]
 91%|█████████ | 9747/10682 [1:29:39<07:40,  2.03it/s]
 91%|█████████▏| 9748/10682 [1:29:40<07:40,  2.03it/s]
 91%|█████████▏| 9749/10682 [1:29:40<07:39,  2.03it/s]
 91%|█████████▏| 9750/10682 [1:29:41<07:39,  2.03it/s]
                                                      
{'loss': 2.8088, 'grad_norm': 0.26079362630844116, 'learning_rate': 2.301407844110154e-05, 'epoch': 12.78}
+
 91%|█████████▏| 9750/10682 [1:29:41<07:39,  2.03it/s]
 91%|█████████▏| 9751/10682 [1:29:41<07:39,  2.03it/s]
 91%|█████████▏| 9752/10682 [1:29:42<07:38,  2.03it/s]
 91%|█████████▏| 9753/10682 [1:29:42<07:37,  2.03it/s]
 91%|█████████▏| 9754/10682 [1:29:43<07:36,  2.03it/s]
 91%|█████████▏| 9755/10682 [1:29:43<07:36,  2.03it/s]
 91%|█████████▏| 9756/10682 [1:29:44<07:36,  2.03it/s]
 91%|█████████▏| 9757/10682 [1:29:44<07:35,  2.03it/s]
 91%|█████████▏| 9758/10682 [1:29:44<07:34,  2.03it/s]
 91%|█████████▏| 9759/10682 [1:29:45<07:34,  2.03it/s]
 91%|█████████▏| 9760/10682 [1:29:45<07:33,  2.03it/s]
 91%|█████████▏| 9761/10682 [1:29:46<07:33,  2.03it/s]
 91%|█████████▏| 9762/10682 [1:29:46<07:32,  2.03it/s]
 91%|█████████▏| 9763/10682 [1:29:47<07:32,  2.03it/s]
 91%|█████████▏| 9764/10682 [1:29:47<07:32,  2.03it/s]
 91%|█████████▏| 9765/10682 [1:29:48<07:31,  2.03it/s]
 91%|█████████▏| 9766/10682 [1:29:48<07:31,  2.03it/s]
 91%|█████████▏| 9767/10682 [1:29:49<07:30,  2.03it/s]
 91%|█████████▏| 9768/10682 [1:29:49<07:30,  2.03it/s]
 91%|█████████▏| 9769/10682 [1:29:50<07:30,  2.03it/s]
 91%|█████████▏| 9770/10682 [1:29:50<07:28,  2.03it/s]
 91%|█████████▏| 9771/10682 [1:29:51<07:28,  2.03it/s]
 91%|█████████▏| 9772/10682 [1:29:51<07:27,  2.03it/s]
 91%|█████████▏| 9773/10682 [1:29:52<07:26,  2.03it/s]
 91%|█████████▏| 9774/10682 [1:29:52<07:26,  2.03it/s]
 92%|█████████▏| 9775/10682 [1:29:53<07:26,  2.03it/s]
                                                      {'loss': 2.8186, 'grad_norm': 0.2622942328453064, 'learning_rate': 2.1804910798715826e-05, 'epoch': 12.81}
+
 92%|█████████▏| 9775/10682 [1:29:53<07:26,  2.03it/s]
 92%|█████████▏| 9776/10682 [1:29:53<07:26,  2.03it/s]
 92%|█████████▏| 9777/10682 [1:29:54<07:25,  2.03it/s]
 92%|█████████▏| 9778/10682 [1:29:54<07:24,  2.03it/s]
 92%|█████████▏| 9779/10682 [1:29:55<07:24,  2.03it/s]
 92%|█████████▏| 9780/10682 [1:29:55<07:23,  2.03it/s]
 92%|█████████▏| 9781/10682 [1:29:56<07:23,  2.03it/s]
 92%|█████████▏| 9782/10682 [1:29:56<07:23,  2.03it/s]
 92%|█████████▏| 9783/10682 [1:29:57<07:22,  2.03it/s]
 92%|█████████▏| 9784/10682 [1:29:57<07:22,  2.03it/s]
 92%|█████████▏| 9785/10682 [1:29:58<07:21,  2.03it/s]
 92%|█████████▏| 9786/10682 [1:29:58<07:21,  2.03it/s]
 92%|█████████▏| 9787/10682 [1:29:59<07:20,  2.03it/s]
 92%|█████████▏| 9788/10682 [1:29:59<07:19,  2.03it/s]
 92%|█████████▏| 9789/10682 [1:30:00<07:19,  2.03it/s]
 92%|█████████▏| 9790/10682 [1:30:00<07:19,  2.03it/s]
 92%|█████████▏| 9791/10682 [1:30:01<07:18,  2.03it/s]
 92%|█████████▏| 9792/10682 [1:30:01<07:18,  2.03it/s]
 92%|█████████▏| 9793/10682 [1:30:02<07:17,  2.03it/s]
 92%|█████████▏| 9794/10682 [1:30:02<07:17,  2.03it/s]
 92%|█████████▏| 9795/10682 [1:30:03<07:16,  2.03it/s]
 92%|█████████▏| 9796/10682 [1:30:03<07:16,  2.03it/s]
 92%|█████████▏| 9797/10682 [1:30:04<07:15,  2.03it/s]
 92%|█████████▏| 9798/10682 [1:30:04<07:15,  2.03it/s]
 92%|█████████▏| 9799/10682 [1:30:05<07:15,  2.03it/s]
 92%|█████████▏| 9800/10682 [1:30:05<07:14,  2.03it/s]
                                                      
{'loss': 2.8062, 'grad_norm': 0.2626664638519287, 'learning_rate': 2.0627663280649135e-05, 'epoch': 12.84}
+
 92%|█████████▏| 9800/10682 [1:30:05<07:14,  2.03it/s]
 92%|█████████▏| 9801/10682 [1:30:06<07:14,  2.03it/s]
 92%|█████████▏| 9802/10682 [1:30:06<07:14,  2.03it/s]
 92%|█████████▏| 9803/10682 [1:30:07<07:13,  2.03it/s]
 92%|█████████▏| 9804/10682 [1:30:07<07:12,  2.03it/s]
 92%|█████████▏| 9805/10682 [1:30:08<07:12,  2.03it/s]
 92%|█████████▏| 9806/10682 [1:30:08<07:11,  2.03it/s]
 92%|█████████▏| 9807/10682 [1:30:09<07:10,  2.03it/s]
 92%|█████████▏| 9808/10682 [1:30:09<07:10,  2.03it/s]
 92%|█████████▏| 9809/10682 [1:30:10<07:10,  2.03it/s]
 92%|█████████▏| 9810/10682 [1:30:10<07:09,  2.03it/s]
 92%|█████████▏| 9811/10682 [1:30:11<07:09,  2.03it/s]
 92%|█████████▏| 9812/10682 [1:30:11<07:09,  2.03it/s]
 92%|█████████▏| 9813/10682 [1:30:12<07:08,  2.03it/s]
 92%|█████████▏| 9814/10682 [1:30:12<07:07,  2.03it/s]
 92%|█████████▏| 9815/10682 [1:30:13<07:07,  2.03it/s]
 92%|█████████▏| 9816/10682 [1:30:13<07:06,  2.03it/s]
 92%|█████████▏| 9817/10682 [1:30:14<07:06,  2.03it/s]
 92%|█████████▏| 9818/10682 [1:30:14<07:05,  2.03it/s]
 92%|█████████▏| 9819/10682 [1:30:15<07:05,  2.03it/s]
 92%|█████████▏| 9820/10682 [1:30:15<07:05,  2.03it/s]
 92%|█████████▏| 9821/10682 [1:30:16<07:04,  2.03it/s]
 92%|█████████▏| 9822/10682 [1:30:16<07:04,  2.03it/s]
 92%|█████████▏| 9823/10682 [1:30:17<07:03,  2.03it/s]
 92%|█████████▏| 9824/10682 [1:30:17<07:02,  2.03it/s]
 92%|█████████▏| 9825/10682 [1:30:17<07:02,  2.03it/s]{'loss': 2.8132, 'grad_norm': 0.26109743118286133, 'learning_rate': 1.9482414469655486e-05, 'epoch': 12.88}
+
                                                      

 92%|█████████▏| 9825/10682 [1:30:17<07:02,  2.03it/s]
 92%|█████████▏| 9826/10682 [1:30:18<07:01,  2.03it/s]
 92%|█████████▏| 9827/10682 [1:30:18<07:01,  2.03it/s]
 92%|█████████▏| 9828/10682 [1:30:19<07:00,  2.03it/s]
 92%|█████████▏| 9829/10682 [1:30:19<07:00,  2.03it/s]
 92%|█████████▏| 9830/10682 [1:30:20<06:59,  2.03it/s]
 92%|█████████▏| 9831/10682 [1:30:20<06:59,  2.03it/s]
 92%|█████████▏| 9832/10682 [1:30:21<06:58,  2.03it/s]
 92%|█████████▏| 9833/10682 [1:30:21<06:58,  2.03it/s]
 92%|█████████▏| 9834/10682 [1:30:22<06:57,  2.03it/s]
 92%|█████████▏| 9835/10682 [1:30:22<06:57,  2.03it/s]
 92%|█████████▏| 9836/10682 [1:30:23<06:56,  2.03it/s]
 92%|█████████▏| 9837/10682 [1:30:23<06:56,  2.03it/s]
 92%|█████████▏| 9838/10682 [1:30:24<06:56,  2.03it/s]
 92%|█████████▏| 9839/10682 [1:30:24<06:55,  2.03it/s]
 92%|█████████▏| 9840/10682 [1:30:25<06:54,  2.03it/s]
 92%|█████████▏| 9841/10682 [1:30:25<06:54,  2.03it/s]
 92%|█████████▏| 9842/10682 [1:30:26<06:54,  2.03it/s]
 92%|█████████▏| 9843/10682 [1:30:26<06:54,  2.03it/s]
 92%|█████████▏| 9844/10682 [1:30:27<06:53,  2.03it/s]
 92%|█████████▏| 9845/10682 [1:30:27<06:52,  2.03it/s]
 92%|█████████▏| 9846/10682 [1:30:28<06:51,  2.03it/s]
 92%|█████████▏| 9847/10682 [1:30:28<06:51,  2.03it/s]
 92%|█████████▏| 9848/10682 [1:30:29<06:51,  2.03it/s]
 92%|█████████▏| 9849/10682 [1:30:29<06:50,  2.03it/s]
 92%|█████████▏| 9850/10682 [1:30:30<06:49,  2.03it/s]{'loss': 2.7987, 'grad_norm': 0.2595779001712799, 'learning_rate': 1.8369240812535104e-05, 'epoch': 12.91}
+
                                                      

 92%|█████████▏| 9850/10682 [1:30:30<06:49,  2.03it/s]
 92%|█████████▏| 9851/10682 [1:30:30<06:49,  2.03it/s]
 92%|█████████▏| 9852/10682 [1:30:31<06:49,  2.03it/s]
 92%|█████████▏| 9853/10682 [1:30:31<06:48,  2.03it/s]
 92%|█████████▏| 9854/10682 [1:30:32<06:47,  2.03it/s]
 92%|█████████▏| 9855/10682 [1:30:32<06:47,  2.03it/s]
 92%|█████████▏| 9856/10682 [1:30:33<06:47,  2.03it/s]
 92%|█████████▏| 9857/10682 [1:30:33<06:46,  2.03it/s]
 92%|█████████▏| 9858/10682 [1:30:34<06:46,  2.03it/s]
 92%|█████████▏| 9859/10682 [1:30:34<06:45,  2.03it/s]
 92%|█████████▏| 9860/10682 [1:30:35<06:45,  2.03it/s]
 92%|█████████▏| 9861/10682 [1:30:35<06:44,  2.03it/s]
 92%|█████████▏| 9862/10682 [1:30:36<06:43,  2.03it/s]
 92%|█████████▏| 9863/10682 [1:30:36<06:43,  2.03it/s]
 92%|█████████▏| 9864/10682 [1:30:37<06:43,  2.03it/s]
 92%|█████████▏| 9865/10682 [1:30:37<06:43,  2.03it/s]
 92%|█████████▏| 9866/10682 [1:30:38<06:42,  2.03it/s]
 92%|█████████▏| 9867/10682 [1:30:38<06:42,  2.03it/s]
 92%|█████████▏| 9868/10682 [1:30:39<06:41,  2.03it/s]
 92%|█████████▏| 9869/10682 [1:30:39<06:41,  2.03it/s]
 92%|█████████▏| 9870/10682 [1:30:40<06:40,  2.03it/s]
 92%|█████████▏| 9871/10682 [1:30:40<06:41,  2.02it/s]
 92%|█████████▏| 9872/10682 [1:30:41<06:40,  2.02it/s]
 92%|█████████▏| 9873/10682 [1:30:41<06:39,  2.02it/s]
 92%|█████████▏| 9874/10682 [1:30:42<06:39,  2.02it/s]
 92%|█████████▏| 9875/10682 [1:30:42<06:38,  2.03it/s]
                                                      
{'loss': 2.8171, 'grad_norm': 0.2623179852962494, 'learning_rate': 1.7288216615031272e-05, 'epoch': 12.94}
+
 92%|█████████▏| 9875/10682 [1:30:42<06:38,  2.03it/s]
 92%|█████████▏| 9876/10682 [1:30:43<06:37,  2.03it/s]
 92%|█████████▏| 9877/10682 [1:30:43<06:37,  2.03it/s]
 92%|█████████▏| 9878/10682 [1:30:44<06:36,  2.03it/s]
 92%|█████████▏| 9879/10682 [1:30:44<06:36,  2.03it/s]
 92%|█████████▏| 9880/10682 [1:30:45<06:35,  2.03it/s]
 93%|█████████▎| 9881/10682 [1:30:45<06:34,  2.03it/s]
 93%|█████████▎| 9882/10682 [1:30:46<06:34,  2.03it/s]
 93%|█████████▎| 9883/10682 [1:30:46<06:33,  2.03it/s]
 93%|█████████▎| 9884/10682 [1:30:47<06:33,  2.03it/s]
 93%|█████████▎| 9885/10682 [1:30:47<06:33,  2.03it/s]
 93%|████████���▎| 9886/10682 [1:30:48<06:32,  2.03it/s]
 93%|█████████▎| 9887/10682 [1:30:48<06:31,  2.03it/s]
 93%|█████████▎| 9888/10682 [1:30:49<06:31,  2.03it/s]
 93%|█████████▎| 9889/10682 [1:30:49<06:31,  2.03it/s]
 93%|█████████▎| 9890/10682 [1:30:50<06:30,  2.03it/s]
 93%|█████████▎| 9891/10682 [1:30:50<06:30,  2.03it/s]
 93%|█████████▎| 9892/10682 [1:30:51<06:29,  2.03it/s]
 93%|█████████▎| 9893/10682 [1:30:51<06:28,  2.03it/s]
 93%|█████████▎| 9894/10682 [1:30:52<06:28,  2.03it/s]
 93%|█████████▎| 9895/10682 [1:30:52<06:27,  2.03it/s]
 93%|█████████▎| 9896/10682 [1:30:53<06:27,  2.03it/s]
 93%|█████████▎| 9897/10682 [1:30:53<06:26,  2.03it/s]
 93%|█████████▎| 9898/10682 [1:30:53<06:26,  2.03it/s]
 93%|█████████▎| 9899/10682 [1:30:54<06:25,  2.03it/s]
 93%|█████████▎| 9900/10682 [1:30:54<06:25,  2.03it/s]{'loss': 2.8168, 'grad_norm': 0.26172664761543274, 'learning_rate': 1.6239414036870183e-05, 'epoch': 12.98}
                                                      
+
 93%|█████████▎| 9900/10682 [1:30:54<06:25,  2.03it/s]
 93%|█████████▎| 9901/10682 [1:30:55<06:25,  2.03it/s]
 93%|█████████▎| 9902/10682 [1:30:55<06:24,  2.03it/s]
 93%|█████████▎| 9903/10682 [1:30:56<06:24,  2.03it/s]
 93%|█████████▎| 9904/10682 [1:30:56<06:23,  2.03it/s]
 93%|█████████▎| 9905/10682 [1:30:57<06:22,  2.03it/s]
 93%|█████████▎| 9906/10682 [1:30:57<06:22,  2.03it/s]
 93%|█████████▎| 9907/10682 [1:30:58<06:22,  2.03it/s]
 93%|█████████▎| 9908/10682 [1:30:58<06:21,  2.03it/s]
 93%|█████████▎| 9909/10682 [1:30:59<06:21,  2.03it/s]
 93%|█████████▎| 9910/10682 [1:30:59<06:20,  2.03it/s]
 93%|█████████▎| 9911/10682 [1:31:00<06:20,  2.03it/s]
 93%|█████████▎| 9912/10682 [1:31:00<06:19,  2.03it/s]
 93%|█████████▎| 9913/10682 [1:31:01<06:18,  2.03it/s]
 93%|█████████▎| 9914/10682 [1:31:01<06:18,  2.03it/s]
 93%|█████████▎| 9915/10682 [1:31:02<06:18,  2.03it/s]
 93%|█████████▎| 9916/10682 [1:31:02<06:17,  2.03it/s]
 93%|█████████▎| 9917/10682 [1:31:03<06:16,  2.03it/s]
 93%|█████████▎| 9918/10682 [1:31:03<06:16,  2.03it/s]
 93%|█████████▎| 9919/10682 [1:31:04<06:11,  2.05it/s]
 93%|█████████▎| 9920/10682 [1:31:50<3:00:39, 14.23s/it]
 93%|█████████▎| 9921/10682 [1:31:51<2:08:11, 10.11s/it]
 93%|█████████▎| 9922/10682 [1:31:51<1:31:28,  7.22s/it]
 93%|█████████▎| 9923/10682 [1:31:52<1:05:51,  5.21s/it]
 93%|█████████▎| 9924/10682 [1:31:52<47:54,  3.79s/it]  
 93%|█████████▎| 9925/10682 [1:31:53<35:21,  2.80s/it]
                                                      
{'loss': 2.8012, 'grad_norm': 0.25953999161720276, 'learning_rate': 1.5222903086944684e-05, 'epoch': 13.01}
+
 93%|█████████▎| 9925/10682 [1:31:53<35:21,  2.80s/it]
 93%|█████████▎| 9926/10682 [1:31:53<26:34,  2.11s/it]
 93%|█████████▎| 9927/10682 [1:31:54<20:26,  1.62s/it]
 93%|█████████▎| 9928/10682 [1:31:54<16:08,  1.28s/it]
 93%|█████████▎| 9929/10682 [1:31:55<13:08,  1.05s/it]
 93%|█████████▎| 9930/10682 [1:31:55<11:02,  1.14it/s]
 93%|█████████▎| 9931/10682 [1:31:56<09:33,  1.31it/s]
 93%|█████████▎| 9932/10682 [1:31:56<08:34,  1.46it/s]
 93%|█████████▎| 9933/10682 [1:31:57<07:51,  1.59it/s]
 93%|█████████▎| 9934/10682 [1:31:57<07:20,  1.70it/s]
 93%|█████████▎| 9935/10682 [1:31:58<06:58,  1.78it/s]
 93%|█████████▎| 9936/10682 [1:31:58<06:42,  1.85it/s]
 93%|█████████▎| 9937/10682 [1:31:58<06:31,  1.90it/s]
 93%|█████████▎| 9938/10682 [1:31:59<06:23,  1.94it/s]
 93%|█████████▎| 9939/10682 [1:31:59<06:18,  1.96it/s]
 93%|█████████▎| 9940/10682 [1:32:00<06:14,  1.98it/s]
 93%|█████████▎| 9941/10682 [1:32:00<06:10,  2.00it/s]
 93%|█████████▎| 9942/10682 [1:32:01<06:08,  2.01it/s]
 93%|█████████▎| 9943/10682 [1:32:01<06:07,  2.01it/s]
 93%|█████████▎| 9944/10682 [1:32:02<06:06,  2.01it/s]
 93%|█████████▎| 9945/10682 [1:32:02<06:05,  2.02it/s]
 93%|█████████▎| 9946/10682 [1:32:03<06:04,  2.02it/s]
 93%|█████████▎| 9947/10682 [1:32:03<06:03,  2.02it/s]
 93%|█��███████▎| 9948/10682 [1:32:04<06:03,  2.02it/s]
 93%|█████████▎| 9949/10682 [1:32:04<06:02,  2.02it/s]
 93%|█████████▎| 9950/10682 [1:32:05<06:01,  2.03it/s]
                                                      {'loss': 2.7863, 'grad_norm': 0.2605855166912079, 'learning_rate': 1.4238751618640577e-05, 'epoch': 13.04}
+

 93%|█████████▎| 9950/10682 [1:32:05<06:01,  2.03it/s]
 93%|█████████▎| 9951/10682 [1:32:05<06:01,  2.02it/s]
 93%|█████████▎| 9952/10682 [1:32:06<06:00,  2.03it/s]
 93%|█████████▎| 9953/10682 [1:32:06<05:59,  2.03it/s]
 93%|█████████▎| 9954/10682 [1:32:07<05:58,  2.03it/s]
 93%|█████████▎| 9955/10682 [1:32:07<05:58,  2.03it/s]
 93%|█████████▎| 9956/10682 [1:32:08<05:57,  2.03it/s]
 93%|█████████▎| 9957/10682 [1:32:08<05:57,  2.03it/s]
 93%|█████████▎| 9958/10682 [1:32:09<05:56,  2.03it/s]
 93%|█████████▎| 9959/10682 [1:32:09<05:55,  2.03it/s]
 93%|█████████▎| 9960/10682 [1:32:10<05:55,  2.03it/s]
 93%|█████████▎| 9961/10682 [1:32:10<05:54,  2.03it/s]
 93%|█████████▎| 9962/10682 [1:32:11<05:54,  2.03it/s]
 93%|█████████▎| 9963/10682 [1:32:11<05:54,  2.03it/s]
 93%|█████████▎| 9964/10682 [1:32:12<05:53,  2.03it/s]
 93%|█████████▎| 9965/10682 [1:32:12<05:52,  2.03it/s]
 93%|█████████▎| 9966/10682 [1:32:13<05:52,  2.03it/s]
 93%|█████████▎| 9967/10682 [1:32:13<05:51,  2.03it/s]
 93%|█████████▎| 9968/10682 [1:32:14<05:51,  2.03it/s]
 93%|█████████▎| 9969/10682 [1:32:14<05:51,  2.03it/s]
 93%|█████████▎| 9970/10682 [1:32:15<05:50,  2.03it/s]
 93%|█████████▎| 9971/10682 [1:32:15<05:50,  2.03it/s]
 93%|█████████▎| 9972/10682 [1:32:16<05:49,  2.03it/s]
 93%|█████████▎| 9973/10682 [1:32:16<05:49,  2.03it/s]
 93%|█████████▎| 9974/10682 [1:32:17<05:49,  2.03it/s]
 93%|█████████▎| 9975/10682 [1:32:17<05:48,  2.03it/s]
                                                      {'loss': 2.7725, 'grad_norm': 0.258526474237442, 'learning_rate': 1.3287025325307511e-05, 'epoch': 13.07}
+
 93%|█████████▎| 9975/10682 [1:32:17<05:48,  2.03it/s]
 93%|█████████▎| 9976/10682 [1:32:18<05:48,  2.03it/s]
 93%|█████████▎| 9977/10682 [1:32:18<05:47,  2.03it/s]
 93%|█████████▎| 9978/10682 [1:32:19<05:46,  2.03it/s]
 93%|█████████▎| 9979/10682 [1:32:19<05:46,  2.03it/s]
 93%|█████████▎| 9980/10682 [1:32:20<05:45,  2.03it/s]
 93%|█████████▎| 9981/10682 [1:32:20<05:45,  2.03it/s]
 93%|█████████▎| 9982/10682 [1:32:21<05:44,  2.03it/s]
 93%|█████████▎| 9983/10682 [1:32:21<05:43,  2.03it/s]
 93%|█████████▎| 9984/10682 [1:32:22<05:43,  2.03it/s]
 93%|█████████▎| 9985/10682 [1:32:22<05:43,  2.03it/s]
 93%|█████████▎| 9986/10682 [1:32:23<05:42,  2.03it/s]
 93%|█████████▎| 9987/10682 [1:32:23<05:42,  2.03it/s]
 94%|█████████▎| 9988/10682 [1:32:24<05:42,  2.03it/s]
 94%|█████████▎| 9989/10682 [1:32:24<05:41,  2.03it/s]
 94%|█████████▎| 9990/10682 [1:32:25<05:40,  2.03it/s]
 94%|█████████▎| 9991/10682 [1:32:25<05:40,  2.03it/s]
 94%|█████████▎| 9992/10682 [1:32:26<05:40,  2.03it/s]
 94%|█████████▎| 9993/10682 [1:32:26<05:39,  2.03it/s]
 94%|█████████▎| 9994/10682 [1:32:27<05:39,  2.03it/s]
 94%|█████████▎| 9995/10682 [1:32:27<05:38,  2.03it/s]
 94%|█████████▎| 9996/10682 [1:32:28<05:37,  2.03it/s]
 94%|█████████▎| 9997/10682 [1:32:28<05:37,  2.03it/s]
 94%|█████████▎| 9998/10682 [1:32:29<05:36,  2.03it/s]
 94%|█████████▎| 9999/10682 [1:32:29<05:36,  2.03it/s]
 94%|█████████▎| 10000/10682 [1:32:30<05:36,  2.03it/s]{'loss': 2.7817, 'grad_norm': 0.26026517152786255, 'learning_rate': 1.2367787735873993e-05, 'epoch': 13.11}
                                                       
+
 94%|█████████▎| 10000/10682 [1:32:30<05:36,  2.03it/s]
 94%|█████████▎| 10001/10682 [1:32:30<05:36,  2.03it/s]
 94%|█████████▎| 10002/10682 [1:32:31<05:35,  2.03it/s]
 94%|█████████▎| 10003/10682 [1:32:31<05:34,  2.03it/s]
 94%|█████████▎| 10004/10682 [1:32:32<05:34,  2.03it/s]
 94%|█████████▎| 10005/10682 [1:32:32<05:33,  2.03it/s]
 94%|█████████▎| 10006/10682 [1:32:33<05:33,  2.03it/s]
 94%|█████████▎| 10007/10682 [1:32:33<05:33,  2.03it/s]
 94%|█████████▎| 10008/10682 [1:32:33<05:32,  2.03it/s]
 94%|█████████▎| 10009/10682 [1:32:34<05:31,  2.03it/s]
 94%|█████████▎| 10010/10682 [1:32:34<05:31,  2.03it/s]
 94%|█████████▎| 10011/10682 [1:32:35<05:30,  2.03it/s]
 94%|█████████▎| 10012/10682 [1:32:35<05:30,  2.03it/s]
 94%|█████████▎| 10013/10682 [1:32:36<05:29,  2.03it/s]
 94%|█████████▎| 10014/10682 [1:32:36<05:29,  2.03it/s]
 94%|█████████▍| 10015/10682 [1:32:37<05:28,  2.03it/s]
 94%|█████████▍| 10016/10682 [1:32:37<05:28,  2.03it/s]
 94%|█████████▍| 10017/10682 [1:32:38<05:27,  2.03it/s]
 94%|█████████▍| 10018/10682 [1:32:38<05:27,  2.03it/s]
 94%|█████████▍| 10019/10682 [1:32:39<05:26,  2.03it/s]
 94%|█████████▍| 10020/10682 [1:32:39<05:26,  2.03it/s]
 94%|█████████▍| 10021/10682 [1:32:40<05:25,  2.03it/s]
 94%|█████████▍| 10022/10682 [1:32:40<05:25,  2.03it/s]
 94%|█████████▍| 10023/10682 [1:32:41<05:24,  2.03it/s]
 94%|█████████▍| 10024/10682 [1:32:41<05:24,  2.03it/s]
 94%|█████████▍| 10025/10682 [1:32:42<05:23,  2.03it/s]{'loss': 2.7818, 'grad_norm': 0.2585986256599426, 'learning_rate': 1.1481100210606388e-05, 'epoch': 13.14}
+
                                                       

 94%|█████████▍| 10025/10682 [1:32:42<05:23,  2.03it/s]
 94%|█████████▍| 10026/10682 [1:32:42<05:23,  2.03it/s]
 94%|█████████▍| 10027/10682 [1:32:43<05:24,  2.02it/s]
 94%|█████████▍| 10028/10682 [1:32:43<05:23,  2.02it/s]
 94%|█████████▍| 10029/10682 [1:32:44<05:22,  2.02it/s]
 94%|█████████▍| 10030/10682 [1:32:44<05:21,  2.03it/s]
 94%|█████████▍| 10031/10682 [1:32:45<05:21,  2.03it/s]
 94%|█████████▍| 10032/10682 [1:32:45<05:20,  2.03it/s]
 94%|█████████▍| 10033/10682 [1:32:46<05:19,  2.03it/s]
 94%|█████████▍| 10034/10682 [1:32:46<05:19,  2.03it/s]
 94%|█████████▍| 10035/10682 [1:32:47<05:19,  2.03it/s]
 94%|█████████▍| 10036/10682 [1:32:47<05:18,  2.03it/s]
 94%|█████████▍| 10037/10682 [1:32:48<05:17,  2.03it/s]
 94%|█████████▍| 10038/10682 [1:32:48<05:17,  2.03it/s]
 94%|█████████▍| 10039/10682 [1:32:49<05:17,  2.03it/s]
 94%|█████████▍| 10040/10682 [1:32:49<05:16,  2.03it/s]
 94%|█████████▍| 10041/10682 [1:32:50<05:16,  2.02it/s]
 94%|█████████▍| 10042/10682 [1:32:50<05:15,  2.03it/s]
 94%|█████████▍| 10043/10682 [1:32:51<05:15,  2.03it/s]
 94%|█████████▍| 10044/10682 [1:32:51<05:14,  2.03it/s]
 94%|█████████▍| 10045/10682 [1:32:52<05:14,  2.03it/s]
 94%|█████████▍| 10046/10682 [1:32:52<05:13,  2.03it/s]
 94%|█████████▍| 10047/10682 [1:32:53<05:12,  2.03it/s]
 94%|█████████▍| 10048/10682 [1:32:53<05:12,  2.03it/s]
 94%|█████████▍| 10049/10682 [1:32:54<05:11,  2.03it/s]
 94%|█████████▍| 10050/10682 [1:32:54<05:10,  2.03it/s]
                                                       {'loss': 2.7821, 'grad_norm': 0.2601000666618347, 'learning_rate': 1.0627021937013704e-05, 'epoch': 13.17}
+

 94%|█████████▍| 10050/10682 [1:32:54<05:10,  2.03it/s]
 94%|█████████▍| 10051/10682 [1:32:55<05:11,  2.03it/s]
 94%|█████████▍| 10052/10682 [1:32:55<05:10,  2.03it/s]
 94%|█████████▍| 10053/10682 [1:32:56<05:09,  2.03it/s]
 94%|█████████▍| 10054/10682 [1:32:56<05:09,  2.03it/s]
 94%|█████████▍| 10055/10682 [1:32:57<05:08,  2.03it/s]
 94%|█████████▍| 10056/10682 [1:32:57<05:08,  2.03it/s]
 94%|█████████▍| 10057/10682 [1:32:58<05:08,  2.03it/s]
 94%|█████████▍| 10058/10682 [1:32:58<05:07,  2.03it/s]
 94%|█████████▍| 10059/10682 [1:32:59<05:06,  2.03it/s]
 94%|█████████▍| 10060/10682 [1:32:59<05:06,  2.03it/s]
 94%|█████████▍| 10061/10682 [1:33:00<05:06,  2.03it/s]
 94%|█████████▍| 10062/10682 [1:33:00<05:05,  2.03it/s]
 94%|█████████▍| 10063/10682 [1:33:01<05:05,  2.03it/s]
 94%|█████████▍| 10064/10682 [1:33:01<05:04,  2.03it/s]
 94%|█████████▍| 10065/10682 [1:33:02<05:03,  2.03it/s]
 94%|█████████▍| 10066/10682 [1:33:02<05:03,  2.03it/s]
 94%|█████████▍| 10067/10682 [1:33:03<05:02,  2.03it/s]
 94%|█████████▍| 10068/10682 [1:33:03<05:02,  2.03it/s]
 94%|█████████▍| 10069/10682 [1:33:04<05:02,  2.03it/s]
 94%|█████████▍| 10070/10682 [1:33:04<05:01,  2.03it/s]
 94%|█████████▍| 10071/10682 [1:33:05<05:01,  2.03it/s]
 94%|█████████▍| 10072/10682 [1:33:05<05:00,  2.03it/s]
 94%|█████████▍| 10073/10682 [1:33:06<05:00,  2.03it/s]
 94%|█████████▍| 10074/10682 [1:33:06<04:59,  2.03it/s]
 94%|█████████▍| 10075/10682 [1:33:07<04:59,  2.03it/s]
                                                       {'loss': 2.7772, 'grad_norm': 0.2605122923851013, 'learning_rate': 9.805609925895964e-06, 'epoch': 13.2}
+
 94%|█████████▍| 10075/10682 [1:33:07<04:59,  2.03it/s]
 94%|█████████▍| 10076/10682 [1:33:07<04:59,  2.03it/s]
 94%|█████████▍| 10077/10682 [1:33:08<04:58,  2.03it/s]
 94%|█████████▍| 10078/10682 [1:33:08<04:57,  2.03it/s]
 94%|█████████▍| 10079/10682 [1:33:08<04:57,  2.03it/s]
 94%|█████████▍| 10080/10682 [1:33:09<04:56,  2.03it/s]
 94%|█████████▍| 10081/10682 [1:33:09<04:56,  2.03it/s]
 94%|█████████▍| 10082/10682 [1:33:10<04:55,  2.03it/s]
 94%|█████████▍| 10083/10682 [1:33:10<04:55,  2.03it/s]
 94%|█████████▍| 10084/10682 [1:33:11<04:54,  2.03it/s]
 94%|█████████▍| 10085/10682 [1:33:11<04:54,  2.03it/s]
 94%|█████████▍| 10086/10682 [1:33:12<04:53,  2.03it/s]
 94%|█████████▍| 10087/10682 [1:33:12<04:52,  2.03it/s]
 94%|█████████▍| 10088/10682 [1:33:13<04:52,  2.03it/s]
 94%|█████████▍| 10089/10682 [1:33:14<05:17,  1.87it/s]
 94%|█████████▍| 10090/10682 [1:33:14<05:09,  1.92it/s]
 94%|█████████▍| 10091/10682 [1:33:15<05:03,  1.95it/s]
 94%|█████████▍| 10092/10682 [1:33:15<04:58,  1.98it/s]
 94%|█████████▍| 10093/10682 [1:33:16<04:56,  1.99it/s]
 94%|█████████▍| 10094/10682 [1:33:16<04:53,  2.00it/s]
 95%|█████████▍| 10095/10682 [1:33:17<04:51,  2.01it/s]
 95%|█████████▍| 10096/10682 [1:33:17<04:50,  2.02it/s]
 95%|█████████▍| 10097/10682 [1:33:17<04:49,  2.02it/s]
 95%|█████████▍| 10098/10682 [1:33:18<04:48,  2.02it/s]
 95%|█████████▍| 10099/10682 [1:33:18<04:47,  2.03it/s]
 95%|█████████▍| 10100/10682 [1:33:19<04:47,  2.03it/s]{'loss': 2.7837, 'grad_norm': 0.26048970222473145, 'learning_rate': 9.01691900753926e-06, 'epoch': 13.24}
+
                                                       

 95%|█████████▍| 10100/10682 [1:33:19<04:47,  2.03it/s]
 95%|█████████▍| 10101/10682 [1:33:19<04:46,  2.03it/s]
 95%|█████████▍| 10102/10682 [1:33:20<04:45,  2.03it/s]
 95%|█████████▍| 10103/10682 [1:33:20<04:45,  2.03it/s]
 95%|█████████▍| 10104/10682 [1:33:21<04:44,  2.03it/s]
 95%|█████████▍| 10105/10682 [1:33:21<04:44,  2.03it/s]
 95%|█████████▍| 10106/10682 [1:33:22<04:43,  2.03it/s]
 95%|█████████▍| 10107/10682 [1:33:22<04:43,  2.03it/s]
 95%|█████████▍| 10108/10682 [1:33:23<04:42,  2.03it/s]
 95%|█████████▍| 10109/10682 [1:33:23<04:42,  2.03it/s]
 95%|█████████▍| 10110/10682 [1:33:24<04:41,  2.03it/s]
 95%|█████████▍| 10111/10682 [1:33:24<04:41,  2.03it/s]
 95%|█████████▍| 10112/10682 [1:33:25<04:40,  2.03it/s]
 95%|█████████▍| 10113/10682 [1:33:25<04:40,  2.03it/s]
 95%|█████████▍| 10114/10682 [1:33:26<04:39,  2.03it/s]
 95%|█████████▍| 10115/10682 [1:33:26<04:39,  2.03it/s]
 95%|█████████▍| 10116/10682 [1:33:27<04:38,  2.03it/s]
 95%|█████████▍| 10117/10682 [1:33:27<04:38,  2.03it/s]
 95%|█████████▍| 10118/10682 [1:33:28<04:37,  2.03it/s]
 95%|█████████▍| 10119/10682 [1:33:28<04:37,  2.03it/s]
 95%|█████████▍| 10120/10682 [1:33:29<04:36,  2.03it/s]
 95%|█████████▍| 10121/10682 [1:33:29<04:36,  2.03it/s]
 95%|█████████▍| 10122/10682 [1:33:30<04:35,  2.03it/s]
 95%|█████████▍| 10123/10682 [1:33:30<04:35,  2.03it/s]
 95%|█████████▍| 10124/10682 [1:33:31<04:34,  2.03it/s]
 95%|█████████▍| 10125/10682 [1:33:31<04:34,  2.03it/s]
                                                       
{'loss': 2.7744, 'grad_norm': 0.26445287466049194, 'learning_rate': 8.261001828055447e-06, 'epoch': 13.27}
+
 95%|█████████▍| 10125/10682 [1:33:31<04:34,  2.03it/s]
 95%|█████████▍| 10126/10682 [1:33:32<04:33,  2.03it/s]
 95%|█████████▍| 10127/10682 [1:33:32<04:33,  2.03it/s]
 95%|█████████▍| 10128/10682 [1:33:33<04:32,  2.03it/s]
 95%|█████████▍| 10129/10682 [1:33:33<04:31,  2.03it/s]
 95%|█████████▍| 10130/10682 [1:33:34<04:31,  2.03it/s]
 95%|█████████▍| 10131/10682 [1:33:34<04:31,  2.03it/s]
 95%|█████████▍| 10132/10682 [1:33:35<04:30,  2.03it/s]
 95%|█████████▍| 10133/10682 [1:33:35<04:30,  2.03it/s]
 95%|█████████▍| 10134/10682 [1:33:36<04:29,  2.03it/s]
 95%|█████████▍| 10135/10682 [1:33:36<04:29,  2.03it/s]
 95%|█████████▍| 10136/10682 [1:33:37<04:28,  2.03it/s]
 95%|█████████▍| 10137/10682 [1:33:37<04:28,  2.03it/s]
 95%|█████████▍| 10138/10682 [1:33:38<04:27,  2.03it/s]
 95%|█████████▍| 10139/10682 [1:33:38<04:27,  2.03it/s]
 95%|█████████▍| 10140/10682 [1:33:39<04:26,  2.03it/s]
 95%|█████████▍| 10141/10682 [1:33:39<04:26,  2.03it/s]
 95%|█████████▍| 10142/10682 [1:33:40<04:25,  2.03it/s]
 95%|█████████▍| 10143/10682 [1:33:40<04:25,  2.03it/s]
 95%|█████████▍| 10144/10682 [1:33:41<04:24,  2.03it/s]
 95%|█████████▍| 10145/10682 [1:33:41<04:24,  2.03it/s]
 95%|█████████▍| 10146/10682 [1:33:42<04:23,  2.03it/s]
 95%|█████████▍| 10147/10682 [1:33:42<04:23,  2.03it/s]
 95%|█████████▌| 10148/10682 [1:33:43<04:22,  2.03it/s]
 95%|█████████▌| 10149/10682 [1:33:43<04:22,  2.03it/s]
 95%|█████████▌| 10150/10682 [1:33:44<04:21,  2.03it/s]{'loss': 2.7789, 'grad_norm': 0.26000910997390747, 'learning_rate': 7.537908845868024e-06, 'epoch': 13.3}
                                                       
+
 95%|█████████▌| 10150/10682 [1:33:44<04:21,  2.03it/s]
 95%|█████████▌| 10151/10682 [1:33:44<04:21,  2.03it/s]
 95%|█████████▌| 10152/10682 [1:33:45<04:21,  2.03it/s]
 95%|█████████▌| 10153/10682 [1:33:45<04:20,  2.03it/s]
 95%|█████████▌| 10154/10682 [1:33:46<04:20,  2.03it/s]
 95%|█████████▌| 10155/10682 [1:33:46<04:19,  2.03it/s]
 95%|█████████▌| 10156/10682 [1:33:47<04:19,  2.03it/s]
 95%|█████████▌| 10157/10682 [1:33:47<04:18,  2.03it/s]
 95%|█████████▌| 10158/10682 [1:33:48<04:18,  2.03it/s]
 95%|█████████▌| 10159/10682 [1:33:48<04:17,  2.03it/s]
 95%|█████████▌| 10160/10682 [1:33:49<04:17,  2.03it/s]
 95%|█████████▌| 10161/10682 [1:33:49<04:16,  2.03it/s]
 95%|█████████▌| 10162/10682 [1:33:50<04:16,  2.03it/s]
 95%|█████████▌| 10163/10682 [1:33:50<04:15,  2.03it/s]
 95%|█████████▌| 10164/10682 [1:33:50<04:15,  2.03it/s]
 95%|█████████▌| 10165/10682 [1:33:51<04:14,  2.03it/s]
 95%|█████████▌| 10166/10682 [1:33:51<04:14,  2.03it/s]
 95%|█████████▌| 10167/10682 [1:33:52<04:14,  2.03it/s]
 95%|█████████▌| 10168/10682 [1:33:52<04:13,  2.03it/s]
 95%|█████████▌| 10169/10682 [1:33:53<04:12,  2.03it/s]
 95%|█████████▌| 10170/10682 [1:33:53<04:12,  2.03it/s]
 95%|█████████▌| 10171/10682 [1:33:54<04:12,  2.03it/s]
 95%|█████████▌| 10172/10682 [1:33:54<04:11,  2.03it/s]
 95%|█████████▌| 10173/10682 [1:33:55<04:10,  2.03it/s]
 95%|█████████▌| 10174/10682 [1:33:55<04:10,  2.03it/s]
 95%|█████████▌| 10175/10682 [1:33:56<04:10,  2.03it/s]{'loss': 2.7813, 'grad_norm': 0.26239922642707825, 'learning_rate': 6.847688328344037e-06, 'epoch': 13.34}
+
                                                       

 95%|█████████▌| 10175/10682 [1:33:56<04:10,  2.03it/s]
 95%|█████████▌| 10176/10682 [1:33:56<04:09,  2.03it/s]
 95%|█████████▌| 10177/10682 [1:33:57<04:09,  2.03it/s]
 95%|█████████▌| 10178/10682 [1:33:57<04:08,  2.03it/s]
 95%|█████████▌| 10179/10682 [1:33:58<04:07,  2.03it/s]
 95%|█████████▌| 10180/10682 [1:33:58<04:07,  2.03it/s]
 95%|█████████▌| 10181/10682 [1:33:59<04:06,  2.03it/s]
 95%|█████████▌| 10182/10682 [1:33:59<04:06,  2.03it/s]
 95%|█████████▌| 10183/10682 [1:34:00<04:05,  2.03it/s]
 95%|█████████▌| 10184/10682 [1:34:00<04:05,  2.03it/s]
 95%|█████████▌| 10185/10682 [1:34:01<04:04,  2.03it/s]
 95%|█████████▌| 10186/10682 [1:34:01<04:04,  2.03it/s]
 95%|█████████▌| 10187/10682 [1:34:02<04:03,  2.03it/s]
 95%|█████████▌| 10188/10682 [1:34:02<04:03,  2.03it/s]
 95%|█████████▌| 10189/10682 [1:34:03<04:02,  2.03it/s]
 95%|█████████▌| 10190/10682 [1:34:03<04:02,  2.03it/s]
 95%|█████████▌| 10191/10682 [1:34:04<04:01,  2.03it/s]
 95%|█████████▌| 10192/10682 [1:34:04<04:01,  2.03it/s]
 95%|█████████▌| 10193/10682 [1:34:05<04:00,  2.03it/s]
 95%|█████████▌| 10194/10682 [1:34:05<04:00,  2.03it/s]
 95%|█████████▌| 10195/10682 [1:34:06<03:59,  2.03it/s]
 95%|█████████▌| 10196/10682 [1:34:06<03:59,  2.03it/s]
 95%|█████████▌| 10197/10682 [1:34:07<03:58,  2.03it/s]
 95%|█████████▌| 10198/10682 [1:34:07<03:58,  2.03it/s]
 95%|█████████▌| 10199/10682 [1:34:08<03:57,  2.03it/s]
 95%|█████████▌| 10200/10682 [1:34:08<03:57,  2.03it/s]
                                                       
{'loss': 2.7754, 'grad_norm': 0.261570543050766, 'learning_rate': 6.190386348572108e-06, 'epoch': 13.37}
+
 95%|█████████▌| 10200/10682 [1:34:08<03:57,  2.03it/s]
 95%|█████████▌| 10201/10682 [1:34:09<03:57,  2.03it/s]
 96%|█████████▌| 10202/10682 [1:34:09<03:56,  2.03it/s]
 96%|█████████▌| 10203/10682 [1:34:10<03:56,  2.03it/s]
 96%|█████████▌| 10204/10682 [1:34:10<03:55,  2.03it/s]
 96%|█████████▌| 10205/10682 [1:34:11<03:54,  2.03it/s]
 96%|█████████▌| 10206/10682 [1:34:11<03:54,  2.03it/s]
 96%|█████████▌| 10207/10682 [1:34:12<03:54,  2.03it/s]
 96%|█████████▌| 10208/10682 [1:34:12<03:53,  2.03it/s]
 96%|█████████▌| 10209/10682 [1:34:13<03:52,  2.03it/s]
 96%|█████████▌| 10210/10682 [1:34:13<03:52,  2.03it/s]
 96%|█████████▌| 10211/10682 [1:34:14<04:12,  1.87it/s]
 96%|█████████▌| 10212/10682 [1:34:14<04:05,  1.92it/s]
 96%|█████████▌| 10213/10682 [1:34:15<04:00,  1.95it/s]
 96%|█████████▌| 10214/10682 [1:34:15<03:57,  1.97it/s]
 96%|█████████▌| 10215/10682 [1:34:16<03:54,  1.99it/s]
 96%|█████████▌| 10216/10682 [1:34:16<03:53,  2.00it/s]
 96%|█████████▌| 10217/10682 [1:34:17<03:51,  2.01it/s]
 96%|█████████▌| 10218/10682 [1:34:17<03:50,  2.01it/s]
 96%|█████████▌| 10219/10682 [1:34:18<03:49,  2.02it/s]
 96%|█████████▌| 10220/10682 [1:34:18<03:48,  2.02it/s]
 96%|█████████▌| 10221/10682 [1:34:19<03:47,  2.03it/s]
 96%|█████████▌| 10222/10682 [1:34:19<03:46,  2.03it/s]
 96%|█████████▌| 10223/10682 [1:34:20<03:46,  2.03it/s]
 96%|█████████▌| 10224/10682 [1:34:20<03:45,  2.03it/s]
 96%|█████████▌| 10225/10682 [1:34:21<03:45,  2.03it/s]
                                                       
{'loss': 2.7881, 'grad_norm': 0.26293087005615234, 'learning_rate': 5.56604678228706e-06, 'epoch': 13.4}
+
 96%|█████████▌| 10225/10682 [1:34:21<03:45,  2.03it/s]
 96%|█████████▌| 10226/10682 [1:34:21<03:44,  2.03it/s]
 96%|█████████▌| 10227/10682 [1:34:22<03:44,  2.03it/s]
 96%|█████████▌| 10228/10682 [1:34:22<03:44,  2.03it/s]
 96%|█████████▌| 10229/10682 [1:34:23<03:43,  2.03it/s]
 96%|█████████▌| 10230/10682 [1:34:23<03:43,  2.02it/s]
 96%|█████████▌| 10231/10682 [1:34:24<03:42,  2.03it/s]
 96%|█████████▌| 10232/10682 [1:34:24<03:42,  2.03it/s]
 96%|█████████▌| 10233/10682 [1:34:25<03:41,  2.03it/s]
 96%|█████████▌| 10234/10682 [1:34:25<03:40,  2.03it/s]
 96%|█████████▌| 10235/10682 [1:34:26<03:40,  2.03it/s]
 96%|█████████▌| 10236/10682 [1:34:26<03:39,  2.03it/s]
 96%|█████████▌| 10237/10682 [1:34:27<03:39,  2.03it/s]
 96%|█████████▌| 10238/10682 [1:34:27<03:38,  2.03it/s]
 96%|█████████▌| 10239/10682 [1:34:28<03:38,  2.03it/s]
 96%|█████████▌| 10240/10682 [1:34:28<03:38,  2.03it/s]
 96%|█████████▌| 10241/10682 [1:34:29<03:37,  2.03it/s]
 96%|█████████▌| 10242/10682 [1:34:29<03:37,  2.03it/s]
 96%|█████████▌| 10243/10682 [1:34:30<03:36,  2.03it/s]
 96%|█████████▌| 10244/10682 [1:34:30<03:35,  2.03it/s]
 96%|█████████▌| 10245/10682 [1:34:31<03:35,  2.03it/s]
 96%|█████████▌| 10246/10682 [1:34:31<03:34,  2.03it/s]
 96%|█████████▌| 10247/10682 [1:34:32<03:34,  2.03it/s]
 96%|█████████▌| 10248/10682 [1:34:32<03:33,  2.03it/s]
 96%|█████████▌| 10249/10682 [1:34:33<03:33,  2.03it/s]
 96%|█████████▌| 10250/10682 [1:34:33<03:32,  2.03it/s]{'loss': 2.7874, 'grad_norm': 0.26110267639160156, 'learning_rate': 4.974711304941093e-06, 'epoch': 13.43}
+
                                                       

 96%|█████████▌| 10250/10682 [1:34:33<03:32,  2.03it/s]
 96%|█████████▌| 10251/10682 [1:34:34<03:32,  2.03it/s]
 96%|█████████▌| 10252/10682 [1:34:34<03:32,  2.03it/s]
 96%|█████████▌| 10253/10682 [1:34:34<03:31,  2.03it/s]
 96%|█████████▌| 10254/10682 [1:34:35<03:31,  2.03it/s]
 96%|█████████▌| 10255/10682 [1:34:35<03:30,  2.03it/s]
 96%|█████████▌| 10256/10682 [1:34:36<03:30,  2.03it/s]
 96%|█████████▌| 10257/10682 [1:34:36<03:29,  2.03it/s]
 96%|█████████▌| 10258/10682 [1:34:37<03:28,  2.03it/s]
 96%|█████████▌| 10259/10682 [1:34:37<03:28,  2.03it/s]
 96%|█████████▌| 10260/10682 [1:34:38<03:27,  2.03it/s]
 96%|█████████▌| 10261/10682 [1:34:38<03:27,  2.03it/s]
 96%|█████████▌| 10262/10682 [1:34:39<03:27,  2.03it/s]
 96%|█████████▌| 10263/10682 [1:34:39<03:26,  2.02it/s]
 96%|█████████▌| 10264/10682 [1:34:40<03:26,  2.02it/s]
 96%|█████████▌| 10265/10682 [1:34:40<03:25,  2.02it/s]
 96%|█████████▌| 10266/10682 [1:34:41<03:25,  2.03it/s]
 96%|█████████▌| 10267/10682 [1:34:41<03:24,  2.03it/s]
 96%|█████████▌| 10268/10682 [1:34:42<03:24,  2.03it/s]
 96%|█████████▌| 10269/10682 [1:34:42<03:23,  2.03it/s]
 96%|█████████▌| 10270/10682 [1:34:43<03:23,  2.03it/s]
 96%|█████████▌| 10271/10682 [1:34:43<03:22,  2.03it/s]
 96%|█████████▌| 10272/10682 [1:34:44<03:22,  2.03it/s]
 96%|█████████▌| 10273/10682 [1:34:44<03:21,  2.03it/s]
 96%|█████████▌| 10274/10682 [1:34:45<03:21,  2.03it/s]
 96%|█████████▌| 10275/10682 [1:34:45<03:20,  2.03it/s]{'loss': 2.7836, 'grad_norm': 0.2616285979747772, 'learning_rate': 4.416419388921844e-06, 'epoch': 13.47}
                                                       
+
 96%|█████████▌| 10275/10682 [1:34:45<03:20,  2.03it/s]
 96%|█████████▌| 10276/10682 [1:34:46<03:20,  2.03it/s]
 96%|█████████▌| 10277/10682 [1:34:46<03:19,  2.03it/s]
 96%|█████████▌| 10278/10682 [1:34:47<03:19,  2.03it/s]
 96%|█████████▌| 10279/10682 [1:34:47<03:18,  2.03it/s]
 96%|█████████▌| 10280/10682 [1:34:48<03:18,  2.03it/s]
 96%|█████████▌| 10281/10682 [1:34:48<03:17,  2.03it/s]
 96%|█████████▋| 10282/10682 [1:34:49<03:17,  2.03it/s]
 96%|█████████▋| 10283/10682 [1:34:49<03:16,  2.03it/s]
 96%|█████████▋| 10284/10682 [1:34:50<03:15,  2.03it/s]
 96%|█████████▋| 10285/10682 [1:34:50<03:15,  2.03it/s]
 96%|█████████▋| 10286/10682 [1:34:51<03:15,  2.03it/s]
 96%|█████████▋| 10287/10682 [1:34:51<03:15,  2.02it/s]
 96%|█████████▋| 10288/10682 [1:34:52<03:14,  2.03it/s]
 96%|█████████▋| 10289/10682 [1:34:52<03:13,  2.03it/s]
 96%|█████████▋| 10290/10682 [1:34:53<03:13,  2.03it/s]
 96%|█████████▋| 10291/10682 [1:34:53<03:12,  2.03it/s]
 96%|█████████▋| 10292/10682 [1:34:54<03:12,  2.03it/s]
 96%|█████████▋| 10293/10682 [1:34:54<03:11,  2.03it/s]
 96%|█████████▋| 10294/10682 [1:34:55<03:11,  2.03it/s]
 96%|█████████▋| 10295/10682 [1:34:55<03:10,  2.03it/s]
 96%|█████████▋| 10296/10682 [1:34:56<03:10,  2.03it/s]
 96%|█████████▋| 10297/10682 [1:34:56<03:09,  2.03it/s]
 96%|█████████▋| 10298/10682 [1:34:57<03:09,  2.03it/s]
 96%|█████████▋| 10299/10682 [1:34:57<03:08,  2.03it/s]
 96%|█████████▋| 10300/10682 [1:34:58<03:08,  2.03it/s]
                                                       {'loss': 2.7732, 'grad_norm': 0.25958919525146484, 'learning_rate': 3.891208300917604e-06, 'epoch': 13.5}
+
 96%|█████████▋| 10300/10682 [1:34:58<03:08,  2.03it/s]
 96%|█████████▋| 10301/10682 [1:34:58<03:07,  2.03it/s]
 96%|█████████▋| 10302/10682 [1:34:59<03:07,  2.03it/s]
 96%|█████████▋| 10303/10682 [1:34:59<03:06,  2.03it/s]
 96%|█████████▋| 10304/10682 [1:35:00<03:06,  2.03it/s]
 96%|█████████▋| 10305/10682 [1:35:00<03:06,  2.03it/s]
 96%|█████████▋| 10306/10682 [1:35:01<03:05,  2.03it/s]
 96%|█████████▋| 10307/10682 [1:35:01<03:04,  2.03it/s]
 96%|█████████▋| 10308/10682 [1:35:02<03:04,  2.03it/s]
 97%|█████████▋| 10309/10682 [1:35:02<03:03,  2.03it/s]
 97%|█████████▋| 10310/10682 [1:35:03<03:03,  2.03it/s]
 97%|█████████▋| 10311/10682 [1:35:03<03:03,  2.03it/s]
 97%|█████████▋| 10312/10682 [1:35:04<03:02,  2.03it/s]
 97%|█████████▋| 10313/10682 [1:35:04<03:01,  2.03it/s]
 97%|█████████▋| 10314/10682 [1:35:05<03:01,  2.03it/s]
 97%|█████████▋| 10315/10682 [1:35:05<03:00,  2.03it/s]
 97%|█████████▋| 10316/10682 [1:35:06<03:00,  2.03it/s]
 97%|█████████▋| 10317/10682 [1:35:06<02:59,  2.03it/s]
 97%|█████████▋| 10318/10682 [1:35:07<02:59,  2.03it/s]
 97%|█████████▋| 10319/10682 [1:35:07<02:58,  2.03it/s]
 97%|█████████▋| 10320/10682 [1:35:08<02:58,  2.03it/s]
 97%|█████████▋| 10321/10682 [1:35:08<02:57,  2.03it/s]
 97%|█████████▋| 10322/10682 [1:35:09<02:57,  2.03it/s]
 97%|█████████▋| 10323/10682 [1:35:09<02:56,  2.03it/s]
 97%|█████████▋| 10324/10682 [1:35:10<02:56,  2.03it/s]
 97%|█████████▋| 10325/10682 [1:35:10<02:55,  2.03it/s]{'loss': 2.772, 'grad_norm': 0.26351845264434814, 'learning_rate': 3.3991130994299734e-06, 'epoch': 13.53}
                                                       
+
 97%|█████████▋| 10325/10682 [1:35:10<02:55,  2.03it/s]
 97%|█████████▋| 10326/10682 [1:35:10<02:55,  2.03it/s]
 97%|█████████▋| 10327/10682 [1:35:11<02:55,  2.03it/s]
 97%|█████████▋| 10328/10682 [1:35:11<02:54,  2.03it/s]
 97%|█████████▋| 10329/10682 [1:35:12<02:54,  2.03it/s]
 97%|█████████▋| 10330/10682 [1:35:12<02:53,  2.03it/s]
 97%|█████████▋| 10331/10682 [1:35:13<02:53,  2.03it/s]
 97%|█████████▋| 10332/10682 [1:35:13<02:52,  2.03it/s]
 97%|█████████▋| 10333/10682 [1:35:14<02:52,  2.03it/s]
 97%|█████████▋| 10334/10682 [1:35:14<02:51,  2.03it/s]
 97%|█████████▋| 10335/10682 [1:35:15<02:51,  2.03it/s]
 97%|█████████▋| 10336/10682 [1:35:15<02:50,  2.02it/s]
 97%|█████████▋| 10337/10682 [1:35:16<02:50,  2.02it/s]
 97%|█████████▋| 10338/10682 [1:35:16<02:49,  2.03it/s]
 97%|█████████▋| 10339/10682 [1:35:17<02:49,  2.03it/s]
 97%|█████████▋| 10340/10682 [1:35:17<02:48,  2.03it/s]
 97%|█████████▋| 10341/10682 [1:35:18<02:48,  2.03it/s]
 97%|█████████▋| 10342/10682 [1:35:18<02:47,  2.03it/s]
 97%|█████████▋| 10343/10682 [1:35:19<02:47,  2.03it/s]
 97%|█████████▋| 10344/10682 [1:35:19<02:46,  2.03it/s]
 97%|█████████▋| 10345/10682 [1:35:20<02:46,  2.03it/s]
 97%|█████████▋| 10346/10682 [1:35:20<02:45,  2.03it/s]
 97%|█████████▋| 10347/10682 [1:35:21<02:45,  2.03it/s]
 97%|█████████▋| 10348/10682 [1:35:21<02:44,  2.03it/s]
 97%|█████████▋| 10349/10682 [1:35:22<02:44,  2.03it/s]
 97%|█████████▋| 10350/10682 [1:35:22<02:43,  2.03it/s]{'loss': 2.786, 'grad_norm': 0.260882169008255, 'learning_rate': 2.940166632433183e-06, 'epoch': 13.56}
                                                       
+
 97%|█████████▋| 10350/10682 [1:35:22<02:43,  2.03it/s]
 97%|█████████▋| 10351/10682 [1:35:23<02:43,  2.03it/s]
 97%|█████████▋| 10352/10682 [1:35:23<02:42,  2.03it/s]
 97%|█████████▋| 10353/10682 [1:35:24<02:42,  2.03it/s]
 97%|█████████▋| 10354/10682 [1:35:24<02:41,  2.03it/s]
 97%|█████████▋| 10355/10682 [1:35:25<02:41,  2.03it/s]
 97%|█████████▋| 10356/10682 [1:35:25<02:40,  2.03it/s]
 97%|█████████▋| 10357/10682 [1:35:26<02:40,  2.03it/s]
 97%|█████████▋| 10358/10682 [1:35:26<02:39,  2.03it/s]
 97%|█████████▋| 10359/10682 [1:35:27<02:39,  2.03it/s]
 97%|█████████▋| 10360/10682 [1:35:27<02:38,  2.03it/s]
 97%|█████████▋| 10361/10682 [1:35:28<02:38,  2.03it/s]
 97%|█████████▋| 10362/10682 [1:35:28<02:37,  2.03it/s]
 97%|█████████▋| 10363/10682 [1:35:29<02:37,  2.03it/s]
 97%|█████████▋| 10364/10682 [1:35:29<02:36,  2.03it/s]
 97%|█████████▋| 10365/10682 [1:35:30<02:36,  2.03it/s]
 97%|█████████▋| 10366/10682 [1:35:30<02:35,  2.03it/s]
 97%|█████████▋| 10367/10682 [1:35:31<02:35,  2.03it/s]
 97%|█████████▋| 10368/10682 [1:35:31<02:34,  2.03it/s]
 97%|█████████▋| 10369/10682 [1:35:32<02:34,  2.03it/s]
 97%|█████████▋| 10370/10682 [1:35:32<02:33,  2.03it/s]
 97%|█████████▋| 10371/10682 [1:35:33<02:33,  2.03it/s]
 97%|█████████▋| 10372/10682 [1:35:33<02:32,  2.03it/s]
 97%|█████████▋| 10373/10682 [1:35:34<02:32,  2.03it/s]
 97%|█████████▋| 10374/10682 [1:35:34<02:31,  2.03it/s]
 97%|█████████▋| 10375/10682 [1:35:35<02:31,  2.03it/s]
                                                       
{'loss': 2.7855, 'grad_norm': 0.26043546199798584, 'learning_rate': 2.5143995351817882e-06, 'epoch': 13.6}
+
 97%|█████████▋| 10375/10682 [1:35:35<02:31,  2.03it/s]
 97%|█████████▋| 10376/10682 [1:35:35<02:31,  2.03it/s]
 97%|█████████▋| 10377/10682 [1:35:36<02:30,  2.02it/s]
 97%|█████████▋| 10378/10682 [1:35:36<02:29,  2.03it/s]
 97%|█████████▋| 10379/10682 [1:35:37<02:29,  2.03it/s]
 97%|█████████▋| 10380/10682 [1:35:37<02:28,  2.03it/s]
 97%|█████████▋| 10381/10682 [1:35:38<02:28,  2.03it/s]
 97%|█████████▋| 10382/10682 [1:35:38<02:28,  2.03it/s]
 97%|█████████▋| 10383/10682 [1:35:39<02:27,  2.03it/s]
 97%|█████████▋| 10384/10682 [1:35:39<02:27,  2.02it/s]
 97%|█████████▋| 10385/10682 [1:35:40<02:26,  2.03it/s]
 97%|█████████▋| 10386/10682 [1:35:40<02:26,  2.03it/s]
 97%|█████████▋| 10387/10682 [1:35:41<02:25,  2.03it/s]
 97%|█████████▋| 10388/10682 [1:35:41<02:24,  2.03it/s]
 97%|█████████▋| 10389/10682 [1:35:42<02:24,  2.03it/s]
 97%|█████████▋| 10390/10682 [1:35:42<02:23,  2.03it/s]
 97%|█████████▋| 10391/10682 [1:35:43<02:23,  2.03it/s]
 97%|█████████▋| 10392/10682 [1:35:43<02:23,  2.03it/s]
 97%|█████████▋| 10393/10682 [1:35:44<02:22,  2.03it/s]
 97%|█████████▋| 10394/10682 [1:35:44<02:22,  2.03it/s]
 97%|█████████▋| 10395/10682 [1:35:45<02:21,  2.03it/s]
 97%|█████████▋| 10396/10682 [1:35:45<02:21,  2.03it/s]
 97%|█████████▋| 10397/10682 [1:35:46<02:20,  2.03it/s]
 97%|█████████▋| 10398/10682 [1:35:46<02:19,  2.03it/s]
 97%|█████████▋| 10399/10682 [1:35:46<02:19,  2.03it/s]
 97%|█████████▋| 10400/10682 [1:35:47<02:19,  2.03it/s]
                                                       {'loss': 2.7878, 'grad_norm': 0.25981855392456055, 'learning_rate': 2.1218402281655835e-06, 'epoch': 13.63}
+
 97%|█████████▋| 10400/10682 [1:35:47<02:19,  2.03it/s]
 97%|█████████▋| 10401/10682 [1:35:47<02:18,  2.02it/s]
 97%|█████████▋| 10402/10682 [1:35:48<02:18,  2.03it/s]
 97%|█████████▋| 10403/10682 [1:35:48<02:17,  2.03it/s]
 97%|█████████▋| 10404/10682 [1:35:49<02:17,  2.03it/s]
 97%|█████████▋| 10405/10682 [1:35:49<02:16,  2.03it/s]
 97%|█████████▋| 10406/10682 [1:35:50<02:16,  2.03it/s]
 97%|█████████▋| 10407/10682 [1:35:50<02:15,  2.03it/s]
 97%|█████████▋| 10408/10682 [1:35:51<02:15,  2.03it/s]
 97%|█████████▋| 10409/10682 [1:35:51<02:14,  2.03it/s]
 97%|█████████▋| 10410/10682 [1:35:52<02:14,  2.03it/s]
 97%|█████████▋| 10411/10682 [1:35:52<02:13,  2.03it/s]
 97%|█████████▋| 10412/10682 [1:35:53<02:13,  2.03it/s]
 97%|█████████▋| 10413/10682 [1:35:53<02:12,  2.03it/s]
 97%|█████████▋| 10414/10682 [1:35:54<02:12,  2.03it/s]
 98%|█████████▊| 10415/10682 [1:35:54<02:11,  2.03it/s]
 98%|█████████▊| 10416/10682 [1:35:55<02:11,  2.03it/s]
 98%|█████████▊| 10417/10682 [1:35:55<02:10,  2.03it/s]
 98%|█████████▊| 10418/10682 [1:35:56<02:10,  2.03it/s]
 98%|█████████▊| 10419/10682 [1:35:56<02:09,  2.03it/s]
 98%|█████████▊| 10420/10682 [1:35:57<02:09,  2.03it/s]
 98%|█████████▊| 10421/10682 [1:35:57<02:08,  2.03it/s]
 98%|█████████▊| 10422/10682 [1:35:58<02:08,  2.03it/s]
 98%|█████████▊| 10423/10682 [1:35:58<02:07,  2.03it/s]
 98%|█████████▊| 10424/10682 [1:35:59<02:07,  2.03it/s]
 98%|█████████▊| 10425/10682 [1:35:59<02:06,  2.03it/s]{'loss': 2.7842, 'grad_norm': 0.2610546350479126, 'learning_rate': 1.7625149152127318e-06, 'epoch': 13.66}
+
                                                       

 98%|█████████▊| 10425/10682 [1:35:59<02:06,  2.03it/s]
 98%|█████████▊| 10426/10682 [1:36:00<02:06,  2.03it/s]
 98%|█████████▊| 10427/10682 [1:36:00<02:05,  2.03it/s]
 98%|█████████▊| 10428/10682 [1:36:01<02:05,  2.03it/s]
 98%|█████████▊| 10429/10682 [1:36:01<02:04,  2.03it/s]
 98%|█████████▊| 10430/10682 [1:36:02<02:04,  2.03it/s]
 98%|█████████▊| 10431/10682 [1:36:02<02:03,  2.03it/s]
 98%|█████████▊| 10432/10682 [1:36:03<02:03,  2.03it/s]
 98%|█████████▊| 10433/10682 [1:36:03<02:02,  2.03it/s]
 98%|█████████▊| 10434/10682 [1:36:04<02:02,  2.03it/s]
 98%|█████████▊| 10435/10682 [1:36:04<02:01,  2.03it/s]
 98%|█████████▊| 10436/10682 [1:36:05<02:01,  2.03it/s]
 98%|█████████▊| 10437/10682 [1:36:05<02:00,  2.03it/s]
 98%|█████████▊| 10438/10682 [1:36:06<02:00,  2.03it/s]
 98%|█████████▊| 10439/10682 [1:36:06<01:59,  2.03it/s]
 98%|█████████▊| 10440/10682 [1:36:07<01:59,  2.03it/s]
 98%|█████████▊| 10441/10682 [1:36:07<01:58,  2.03it/s]
 98%|█████████▊| 10442/10682 [1:36:08<01:58,  2.03it/s]
 98%|█████████▊| 10443/10682 [1:36:08<01:57,  2.03it/s]
 98%|█████████▊| 10444/10682 [1:36:09<01:57,  2.03it/s]
 98%|█████████▊| 10445/10682 [1:36:09<01:56,  2.03it/s]
 98%|█████████▊| 10446/10682 [1:36:10<01:56,  2.03it/s]
 98%|█████████▊| 10447/10682 [1:36:10<01:55,  2.03it/s]
 98%|█████████▊| 10448/10682 [1:36:11<01:55,  2.03it/s]
 98%|█████████▊| 10449/10682 [1:36:11<01:54,  2.03it/s]
 98%|█████████▊| 10450/10682 [1:36:12<01:54,  2.03it/s]
                                                       {'loss': 2.784, 'grad_norm': 0.25831174850463867, 'learning_rate': 1.4364475817401635e-06, 'epoch': 13.7}
+

 98%|█████████▊| 10450/10682 [1:36:12<01:54,  2.03it/s]
 98%|█████████▊| 10451/10682 [1:36:12<01:54,  2.02it/s]
 98%|█████████▊| 10452/10682 [1:36:13<01:53,  2.03it/s]
 98%|█████████▊| 10453/10682 [1:36:13<01:53,  2.03it/s]
 98%|█████████▊| 10454/10682 [1:36:14<01:52,  2.03it/s]
 98%|█████████▊| 10455/10682 [1:36:14<01:52,  2.03it/s]
 98%|█████████▊| 10456/10682 [1:36:15<01:51,  2.03it/s]
 98%|█████████▊| 10457/10682 [1:36:15<01:50,  2.03it/s]
 98%|█████████▊| 10458/10682 [1:36:16<01:50,  2.03it/s]
 98%|█████████▊| 10459/10682 [1:36:16<01:49,  2.03it/s]
 98%|█████████▊| 10460/10682 [1:36:17<01:49,  2.03it/s]
 98%|█████████▊| 10461/10682 [1:36:17<01:49,  2.03it/s]
 98%|█████████▊| 10462/10682 [1:36:18<01:48,  2.03it/s]
 98%|█████████▊| 10463/10682 [1:36:18<01:47,  2.03it/s]
 98%|█████████▊| 10464/10682 [1:36:19<01:47,  2.03it/s]
 98%|█████████▊| 10465/10682 [1:36:19<01:46,  2.03it/s]
 98%|█████████▊| 10466/10682 [1:36:20<01:46,  2.03it/s]
 98%|█████████▊| 10467/10682 [1:36:20<01:45,  2.03it/s]
 98%|█████████▊| 10468/10682 [1:36:21<01:45,  2.03it/s]
 98%|█████████▊| 10469/10682 [1:36:21<01:45,  2.03it/s]
 98%|█████████▊| 10470/10682 [1:36:22<01:44,  2.03it/s]
 98%|█████████▊| 10471/10682 [1:36:22<01:44,  2.03it/s]
 98%|█████████▊| 10472/10682 [1:36:22<01:43,  2.03it/s]
 98%|█████████▊| 10473/10682 [1:36:23<01:43,  2.03it/s]
 98%|█████████▊| 10474/10682 [1:36:23<01:42,  2.03it/s]
 98%|█████████▊| 10475/10682 [1:36:24<01:42,  2.03it/s]{'loss': 2.7868, 'grad_norm': 0.2604929506778717, 'learning_rate': 1.143659993153079e-06, 'epoch': 13.73}
                                                       
+
 98%|█████████▊| 10475/10682 [1:36:24<01:42,  2.03it/s]
 98%|█████████▊| 10476/10682 [1:36:24<01:41,  2.03it/s]
 98%|█████████▊| 10477/10682 [1:36:25<01:41,  2.02it/s]
 98%|█████████▊| 10478/10682 [1:36:25<01:40,  2.03it/s]
 98%|█████████▊| 10479/10682 [1:36:26<01:40,  2.03it/s]
 98%|█████████▊| 10480/10682 [1:36:26<01:39,  2.03it/s]
 98%|█████████▊| 10481/10682 [1:36:27<01:39,  2.03it/s]
 98%|█████████▊| 10482/10682 [1:36:27<01:38,  2.03it/s]
 98%|█████████▊| 10483/10682 [1:36:28<01:38,  2.03it/s]
 98%|█████████▊| 10484/10682 [1:36:28<01:37,  2.03it/s]
 98%|█████████▊| 10485/10682 [1:36:29<01:37,  2.03it/s]
 98%|█████████▊| 10486/10682 [1:36:29<01:36,  2.03it/s]
 98%|█████████▊| 10487/10682 [1:36:30<01:36,  2.03it/s]
 98%|█████████▊| 10488/10682 [1:36:30<01:35,  2.03it/s]
 98%|█████████▊| 10489/10682 [1:36:31<01:35,  2.03it/s]
 98%|█████████▊| 10490/10682 [1:36:31<01:34,  2.03it/s]
 98%|█████████▊| 10491/10682 [1:36:32<01:34,  2.03it/s]
 98%|█████████▊| 10492/10682 [1:36:32<01:33,  2.03it/s]
 98%|█████████▊| 10493/10682 [1:36:33<01:33,  2.03it/s]
 98%|█████████▊| 10494/10682 [1:36:33<01:32,  2.03it/s]
 98%|█████████▊| 10495/10682 [1:36:34<01:32,  2.03it/s]
 98%|█████████▊| 10496/10682 [1:36:34<01:31,  2.03it/s]
 98%|█████████▊| 10497/10682 [1:36:35<01:31,  2.03it/s]
 98%|█████████▊| 10498/10682 [1:36:35<01:30,  2.03it/s]
 98%|█████████▊| 10499/10682 [1:36:36<01:30,  2.03it/s]
 98%|█████████▊| 10500/10682 [1:36:36<01:29,  2.03it/s]
                                                       {'loss': 2.7768, 'grad_norm': 0.26020318269729614, 'learning_rate': 8.841716933915555e-07, 'epoch': 13.76}
+

 98%|█████████▊| 10500/10682 [1:36:36<01:29,  2.03it/s]
 98%|█████████▊| 10501/10682 [1:36:37<01:29,  2.02it/s]
 98%|█████████▊| 10502/10682 [1:36:37<01:28,  2.03it/s]
 98%|█████████▊| 10503/10682 [1:36:38<01:28,  2.03it/s]
 98%|█████████▊| 10504/10682 [1:36:38<01:27,  2.03it/s]
 98%|█████████▊| 10505/10682 [1:36:39<01:27,  2.03it/s]
 98%|█████████▊| 10506/10682 [1:36:39<01:26,  2.03it/s]
 98%|█████████▊| 10507/10682 [1:36:40<01:26,  2.03it/s]
 98%|█████████▊| 10508/10682 [1:36:40<01:25,  2.03it/s]
 98%|█████████▊| 10509/10682 [1:36:41<01:25,  2.03it/s]
 98%|█████████▊| 10510/10682 [1:36:41<01:25,  2.02it/s]
 98%|█████████▊| 10511/10682 [1:36:42<01:24,  2.03it/s]
 98%|█████████▊| 10512/10682 [1:36:42<01:23,  2.02it/s]
 98%|█████████▊| 10513/10682 [1:36:43<01:23,  2.03it/s]
 98%|█████████▊| 10514/10682 [1:36:43<01:22,  2.03it/s]
 98%|█████████▊| 10515/10682 [1:36:44<01:22,  2.03it/s]
 98%|█████████▊| 10516/10682 [1:36:44<01:21,  2.03it/s]
 98%|█████████▊| 10517/10682 [1:36:45<01:21,  2.03it/s]
 98%|█████████▊| 10518/10682 [1:36:45<01:20,  2.03it/s]
 98%|█████████▊| 10519/10682 [1:36:46<01:20,  2.03it/s]
 98%|█████████▊| 10520/10682 [1:36:46<01:20,  2.02it/s]
 98%|█████████▊| 10521/10682 [1:36:47<01:19,  2.03it/s]
 99%|█████████▊| 10522/10682 [1:36:47<01:18,  2.03it/s]
 99%|█████████▊| 10523/10682 [1:36:48<01:18,  2.03it/s]
 99%|█████████▊| 10524/10682 [1:36:48<01:17,  2.03it/s]
 99%|█████████▊| 10525/10682 [1:36:49<01:17,  2.03it/s]{'loss': 2.7826, 'grad_norm': 0.2594583332538605, 'learning_rate': 6.580000036264244e-07, 'epoch': 13.79}
                                                       
+
 99%|█████████▊| 10525/10682 [1:36:49<01:17,  2.03it/s]
 99%|█████████▊| 10526/10682 [1:36:49<01:17,  2.03it/s]
 99%|█████████▊| 10527/10682 [1:36:50<01:16,  2.03it/s]
 99%|█████████▊| 10528/10682 [1:36:50<01:16,  2.03it/s]
 99%|█████████▊| 10529/10682 [1:36:51<01:15,  2.03it/s]
 99%|█████████▊| 10530/10682 [1:36:51<01:15,  2.03it/s]
 99%|█████████▊| 10531/10682 [1:36:52<01:14,  2.03it/s]
 99%|█████████▊| 10532/10682 [1:36:52<01:14,  2.02it/s]
 99%|█████████▊| 10533/10682 [1:36:53<01:13,  2.02it/s]
 99%|█████████▊| 10534/10682 [1:36:53<01:13,  2.02it/s]
 99%|█████████▊| 10535/10682 [1:36:54<01:12,  2.02it/s]
 99%|█████████▊| 10536/10682 [1:36:54<01:12,  2.02it/s]
 99%|█████████▊| 10537/10682 [1:36:55<01:11,  2.02it/s]
 99%|█████████▊| 10538/10682 [1:36:55<01:11,  2.02it/s]
 99%|█████████▊| 10539/10682 [1:36:56<01:10,  2.03it/s]
 99%|█████████▊| 10540/10682 [1:36:56<01:10,  2.03it/s]
 99%|█████████▊| 10541/10682 [1:36:57<01:09,  2.03it/s]
 99%|█████████▊| 10542/10682 [1:36:57<01:09,  2.03it/s]
 99%|█████████▊| 10543/10682 [1:36:58<01:08,  2.03it/s]
 99%|█████████▊| 10544/10682 [1:36:58<01:08,  2.02it/s]
 99%|█████████▊| 10545/10682 [1:36:59<01:07,  2.03it/s]
 99%|█████████▊| 10546/10682 [1:36:59<01:07,  2.03it/s]
 99%|█████████▊| 10547/10682 [1:37:00<01:06,  2.03it/s]
 99%|█████████▊| 10548/10682 [1:37:00<01:06,  2.03it/s]
 99%|█████████▉| 10549/10682 [1:37:00<01:05,  2.03it/s]
 99%|█████████▉| 10550/10682 [1:37:01<01:05,  2.03it/s]
                                                       {'loss': 2.7837, 'grad_norm': 0.26167574524879456, 'learning_rate': 4.651600211027507e-07, 'epoch': 13.83}
+

 99%|█████████▉| 10550/10682 [1:37:01<01:05,  2.03it/s]
 99%|█████████▉| 10551/10682 [1:37:01<01:04,  2.03it/s]
 99%|█████████▉| 10552/10682 [1:37:02<01:04,  2.03it/s]
 99%|█████████▉| 10553/10682 [1:37:02<01:03,  2.03it/s]
 99%|█████████▉| 10554/10682 [1:37:03<01:03,  2.03it/s]
 99%|█████████▉| 10555/10682 [1:37:03<01:02,  2.03it/s]
 99%|█████████▉| 10556/10682 [1:37:04<01:02,  2.03it/s]
 99%|█████████▉| 10557/10682 [1:37:04<01:01,  2.03it/s]
 99%|█████████▉| 10558/10682 [1:37:05<01:01,  2.03it/s]
 99%|█████████▉| 10559/10682 [1:37:05<01:00,  2.03it/s]
 99%|█████████▉| 10560/10682 [1:37:06<01:00,  2.03it/s]
 99%|█████████▉| 10561/10682 [1:37:06<00:59,  2.03it/s]
 99%|█████████▉| 10562/10682 [1:37:07<00:59,  2.03it/s]
 99%|█████████▉| 10563/10682 [1:37:07<00:58,  2.03it/s]
 99%|█████████▉| 10564/10682 [1:37:08<00:58,  2.03it/s]
 99%|█████████▉| 10565/10682 [1:37:08<00:57,  2.03it/s]
 99%|█████████▉| 10566/10682 [1:37:09<00:57,  2.03it/s]
 99%|█████████▉| 10567/10682 [1:37:09<00:56,  2.03it/s]
 99%|█████████▉| 10568/10682 [1:37:10<00:56,  2.03it/s]
 99%|█████████▉| 10569/10682 [1:37:10<00:55,  2.03it/s]
 99%|█████████▉| 10570/10682 [1:37:11<00:55,  2.03it/s]
 99%|█████████▉| 10571/10682 [1:37:11<00:54,  2.03it/s]
 99%|█████████▉| 10572/10682 [1:37:12<00:54,  2.03it/s]
 99%|█████████▉| 10573/10682 [1:37:12<00:53,  2.03it/s]
 99%|█████████▉| 10574/10682 [1:37:13<00:53,  2.03it/s]
 99%|█████████▉| 10575/10682 [1:37:13<00:52,  2.03it/s]{'loss': 2.7867, 'grad_norm': 0.25890398025512695, 'learning_rate': 3.0566461813213986e-07, 'epoch': 13.86}
+
                                                       

 99%|█████████▉| 10575/10682 [1:37:13<00:52,  2.03it/s]
 99%|█████████▉| 10576/10682 [1:37:14<00:52,  2.02it/s]
 99%|█████████▉| 10577/10682 [1:37:14<00:51,  2.02it/s]
 99%|█████████▉| 10578/10682 [1:37:15<00:51,  2.02it/s]
 99%|█████████▉| 10579/10682 [1:37:15<00:50,  2.03it/s]
 99%|█████████▉| 10580/10682 [1:37:16<00:50,  2.03it/s]
 99%|█████████▉| 10581/10682 [1:37:16<00:49,  2.03it/s]
 99%|█████████▉| 10582/10682 [1:37:17<00:49,  2.03it/s]
 99%|█████████▉| 10583/10682 [1:37:17<00:48,  2.03it/s]
 99%|█████████▉| 10584/10682 [1:37:18<00:48,  2.03it/s]
 99%|█████████▉| 10585/10682 [1:37:18<00:47,  2.03it/s]
 99%|█████████▉| 10586/10682 [1:37:19<00:47,  2.03it/s]
 99%|█████████▉| 10587/10682 [1:37:19<00:46,  2.03it/s]
 99%|█████████▉| 10588/10682 [1:37:20<00:46,  2.03it/s]
 99%|█████████▉| 10589/10682 [1:37:20<00:45,  2.03it/s]
 99%|█████████▉| 10590/10682 [1:37:21<00:45,  2.03it/s]
 99%|█████████▉| 10591/10682 [1:37:21<00:44,  2.03it/s]
 99%|█████████▉| 10592/10682 [1:37:22<00:44,  2.03it/s]
 99%|█████████▉| 10593/10682 [1:37:22<00:43,  2.03it/s]
 99%|█████████▉| 10594/10682 [1:37:23<00:43,  2.03it/s]
 99%|█████████▉| 10595/10682 [1:37:23<00:42,  2.03it/s]
 99%|█████████▉| 10596/10682 [1:37:24<00:42,  2.03it/s]
 99%|█████████▉| 10597/10682 [1:37:24<00:41,  2.03it/s]
 99%|█████████▉| 10598/10682 [1:37:25<00:41,  2.03it/s]
 99%|█████████▉| 10599/10682 [1:37:25<00:40,  2.03it/s]
 99%|█████████▉| 10600/10682 [1:37:26<00:40,  2.03it/s]{'loss': 2.7834, 'grad_norm': 0.2615819573402405, 'learning_rate': 1.7952444123359167e-07, 'epoch': 13.89}
+
                                                       

 99%|█████████▉| 10600/10682 [1:37:26<00:40,  2.03it/s]
 99%|█████████▉| 10601/10682 [1:37:26<00:39,  2.03it/s]
 99%|█████████▉| 10602/10682 [1:37:27<00:39,  2.03it/s]
 99%|█████████▉| 10603/10682 [1:37:27<00:38,  2.03it/s]
 99%|█████████▉| 10604/10682 [1:37:28<00:38,  2.03it/s]
 99%|█████████▉| 10605/10682 [1:37:28<00:37,  2.03it/s]
 99%|█████████▉| 10606/10682 [1:37:29<00:37,  2.03it/s]
 99%|█████████▉| 10607/10682 [1:37:29<00:36,  2.03it/s]
 99%|█████████▉| 10608/10682 [1:37:30<00:36,  2.03it/s]
 99%|█████████▉| 10609/10682 [1:37:30<00:36,  2.03it/s]
 99%|█████████▉| 10610/10682 [1:37:31<00:35,  2.03it/s]
 99%|█████████▉| 10611/10682 [1:37:31<00:35,  2.03it/s]
 99%|█████████▉| 10612/10682 [1:37:32<00:34,  2.03it/s]
 99%|█████████▉| 10613/10682 [1:37:32<00:34,  2.03it/s]
 99%|█████████▉| 10614/10682 [1:37:33<00:33,  2.03it/s]
 99%|█████████▉| 10615/10682 [1:37:33<00:33,  2.03it/s]
 99%|█████████▉| 10616/10682 [1:37:34<00:32,  2.03it/s]
 99%|█████████▉| 10617/10682 [1:37:34<00:32,  2.03it/s]
 99%|█████████▉| 10618/10682 [1:37:35<00:31,  2.03it/s]
 99%|█████████▉| 10619/10682 [1:37:35<00:31,  2.03it/s]
 99%|█████████▉| 10620/10682 [1:37:35<00:30,  2.03it/s]
 99%|█████████▉| 10621/10682 [1:37:36<00:30,  2.03it/s]
 99%|█████████▉| 10622/10682 [1:37:36<00:29,  2.03it/s]
 99%|█████████▉| 10623/10682 [1:37:37<00:29,  2.03it/s]
 99%|█████████▉| 10624/10682 [1:37:37<00:28,  2.03it/s]
 99%|█████████▉| 10625/10682 [1:37:38<00:28,  2.03it/s]
                                                       {'loss': 2.788, 'grad_norm': 0.26222410798072815, 'learning_rate': 8.674791042273533e-08, 'epoch': 13.93}
+

 99%|█████████▉| 10625/10682 [1:37:38<00:28,  2.03it/s]
 99%|█████████▉| 10626/10682 [1:37:38<00:27,  2.03it/s]
 99%|█████████▉| 10627/10682 [1:37:39<00:27,  2.03it/s]
 99%|█████████▉| 10628/10682 [1:37:39<00:26,  2.03it/s]
100%|█████████▉| 10629/10682 [1:37:40<00:26,  2.03it/s]
100%|█████████▉| 10630/10682 [1:37:40<00:25,  2.03it/s]
100%|█████████▉| 10631/10682 [1:37:41<00:25,  2.03it/s]
100%|█████████▉| 10632/10682 [1:37:41<00:24,  2.03it/s]
100%|█████████▉| 10633/10682 [1:37:42<00:24,  2.03it/s]
100%|█████████▉| 10634/10682 [1:37:42<00:23,  2.03it/s]
100%|█████████▉| 10635/10682 [1:37:43<00:23,  2.03it/s]
100%|█████████▉| 10636/10682 [1:37:43<00:22,  2.03it/s]
100%|█████████▉| 10637/10682 [1:37:44<00:22,  2.03it/s]
100%|█████████▉| 10638/10682 [1:37:44<00:21,  2.03it/s]
100%|█████████▉| 10639/10682 [1:37:45<00:21,  2.03it/s]
100%|█████████▉| 10640/10682 [1:37:45<00:20,  2.03it/s]
100%|█████████▉| 10641/10682 [1:37:46<00:20,  2.03it/s]
100%|█████████▉| 10642/10682 [1:37:46<00:19,  2.03it/s]
100%|█████████▉| 10643/10682 [1:37:47<00:19,  2.03it/s]
100%|█████████▉| 10644/10682 [1:37:47<00:18,  2.03it/s]
100%|█████████▉| 10645/10682 [1:37:48<00:18,  2.03it/s]
100%|█████████▉| 10646/10682 [1:37:48<00:17,  2.03it/s]
100%|█████████▉| 10647/10682 [1:37:49<00:17,  2.03it/s]
100%|█████████▉| 10648/10682 [1:37:49<00:16,  2.03it/s]
100%|█████████▉| 10649/10682 [1:37:50<00:16,  2.03it/s]
100%|█████████▉| 10650/10682 [1:37:50<00:15,  2.03it/s]{'loss': 2.7858, 'grad_norm': 0.2616738975048065, 'learning_rate': 2.7341218649834522e-08, 'epoch': 13.96}
+
                                                       

100%|█████████▉| 10650/10682 [1:37:50<00:15,  2.03it/s]
100%|█████████▉| 10651/10682 [1:37:51<00:15,  2.02it/s]
100%|█████████▉| 10652/10682 [1:37:51<00:14,  2.03it/s]
100%|█████████▉| 10653/10682 [1:37:52<00:14,  2.02it/s]
100%|█████████▉| 10654/10682 [1:37:52<00:13,  2.03it/s]
100%|█████████▉| 10655/10682 [1:37:53<00:13,  2.03it/s]
100%|█████████▉| 10656/10682 [1:37:53<00:12,  2.03it/s]
100%|█████████▉| 10657/10682 [1:37:54<00:12,  2.03it/s]
100%|█████████▉| 10658/10682 [1:37:54<00:11,  2.03it/s]
100%|█████████▉| 10659/10682 [1:37:55<00:11,  2.03it/s]
100%|█████████▉| 10660/10682 [1:37:55<00:10,  2.03it/s]
100%|█████████▉| 10661/10682 [1:37:56<00:10,  2.03it/s]
100%|█████████▉| 10662/10682 [1:37:56<00:09,  2.03it/s]
100%|█████████▉| 10663/10682 [1:37:57<00:09,  2.03it/s]
100%|█████████▉| 10664/10682 [1:37:57<00:08,  2.03it/s]
100%|█████████▉| 10665/10682 [1:37:58<00:08,  2.03it/s]
100%|█████████▉| 10666/10682 [1:37:58<00:07,  2.03it/s]
100%|█████████▉| 10667/10682 [1:37:59<00:07,  2.03it/s]
100%|█████████▉| 10668/10682 [1:37:59<00:06,  2.03it/s]
100%|█████████▉| 10669/10682 [1:38:00<00:06,  2.03it/s]
100%|█████████▉| 10670/10682 [1:38:00<00:05,  2.03it/s]
100%|█████████▉| 10671/10682 [1:38:01<00:05,  2.03it/s]
100%|█████████▉| 10672/10682 [1:38:01<00:04,  2.03it/s]
100%|█████████▉| 10673/10682 [1:38:02<00:04,  2.03it/s]
100%|█████████▉| 10674/10682 [1:38:02<00:03,  2.03it/s]
100%|█████████▉| 10675/10682 [1:38:03<00:03,  2.03it/s]
                                                       {'loss': 2.7837, 'grad_norm': 0.2600068151950836, 'learning_rate': 1.3083313863404555e-09, 'epoch': 13.99}
+

100%|█████████▉| 10675/10682 [1:38:03<00:03,  2.03it/s]
100%|█████████▉| 10676/10682 [1:38:03<00:02,  2.03it/s]
100%|█████████▉| 10677/10682 [1:38:04<00:02,  2.03it/s]
100%|█████████▉| 10678/10682 [1:38:04<00:01,  2.03it/s]
100%|█████████▉| 10679/10682 [1:38:05<00:01,  2.03it/s]
100%|█████████▉| 10680/10682 [1:38:05<00:00,  2.03it/s]
100%|█████████▉| 10681/10682 [1:38:06<00:00,  2.03it/s]
100%|██████████| 10682/10682 [1:38:06<00:00,  2.05it/s]
                                                       
{'train_runtime': 5941.9145, 'train_samples_per_second': 1840.826, 'train_steps_per_second': 1.798, 'train_loss': 3.3962442974454117, 'epoch': 14.0}
+
100%|██████████| 10682/10682 [1:39:01<00:00,  2.05it/s]
100%|██████████| 10682/10682 [1:39:01<00:00,  1.80it/s]
+Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.