Training in progress, epoch 12

logs/events.out.tfevents.1715298064.sphinx2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:65813779c9fe27f7778799ad879a8bb87edc140f565a7b7c71644ab13dad170d
-size 82052
+oid sha256:d227e76941721f0ac3578b2d44031298dc76b7e911b4e35d50eaaa757e9fa05c
+size 88382

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bcec737ae4fec6771c0fb82a11f811bb3faaceb71d66d7d7d217e5cdf027ed26
+oid sha256:cf1c832644896eb09d95b6184f4c813d541a78091236581a707b2a7d9fdc96cd
 size 281715176

train_job_output.txt

@@ -452,4 +452,34 @@ Retrying in 1s [Retry 1/5].
 
 85%|████████▌ | 9100/10682 [1:17:21<13:01,  2.02it/s]
 85%|████████▌ | 9101/10682 [1:17:21<13:02,  2.02it/s]
 85%|████████▌ | 9102/10682 [1:17:22<13:01,  2.02it/s]
 85%|████████▌ | 9103/10682 [1:17:22<13:00,  2.02it/s]
 85%|████████▌ | 9104/10682 [1:17:23<13:00,  2.02it/s]
 85%|████████▌ | 9105/10682 [1:17:23<13:00,  2.02it/s]
 85%|████████▌ | 9106/10682 [1:17:24<13:00,  2.02it/s]
 85%|████████▌ | 9107/10682 [1:17:24<12:59,  2.02it/s]
 85%|████████▌ | 9108/10682 [1:17:25<12:58,  2.02it/s]
 85%|████████▌ | 9109/10682 [1:17:25<12:57,  2.02it/s]
 85%|████████▌ | 9110/10682 [1:17:26<12:57,  2.02it/s]
 85%|████████▌ | 9111/10682 [1:17:26<12:56,  2.02it/s]
 85%|████████▌ | 9112/10682 [1:17:27<12:56,  2.02it/s]
 85%|████████▌ | 9113/10682 [1:17:27<12:56,  2.02it/s]
 85%|████████▌ | 9114/10682 [1:17:28<12:54,  2.02it/s]
 85%|████████▌ | 9115/10682 [1:17:28<12:54,  2.02it/s]
 85%|████████▌ | 9116/10682 [1:17:29<12:54,  2.02it/s]
 85%|████████▌ | 9117/10682 [1:17:29<12:54,  2.02it/s]
 85%|████████▌ | 9118/10682 [1:17:30<12:53,  2.02it/s]
 85%|████████▌ | 9119/10682 [1:17:30<12:52,  2.02it/s]
 85%|████████▌ | 9120/10682 [1:17:31<12:52,  2.02it/s]
 85%|████████▌ | 9121/10682 [1:17:31<12:51,  2.02it/s]
 85%|████████▌ | 9122/10682 [1:17:32<12:51,  2.02it/s]
 85%|████████▌ | 9123/10682 [1:17:32<12:51,  2.02it/s]
 85%|████████▌ | 9124/10682 [1:17:33<12:51,  2.02it/s]
 85%|████████▌ | 9125/10682 [1:17:33<12:50,  2.02it/s]{'loss': 2.7976, 'grad_norm': 0.275244802236557, 'learning_rate': 6.334441157888504e-05, 'epoch': 11.95}
                                                      
 
 85%|████████▌ | 9125/10682 [1:17:33<12:50,  2.02it/s]
 85%|████████▌ | 9126/10682 [1:17:34<12:52,  2.01it/s]
 85%|████████▌ | 9127/10682 [1:17:34<12:51,  2.02it/s]
 85%|████████▌ | 9128/10682 [1:17:35<12:49,  2.02it/s]
 85%|████████▌ | 9129/10682 [1:17:35<12:49,  2.02it/s]
 85%|████████▌ | 9130/10682 [1:17:36<12:47,  2.02it/s]
 85%|████████▌ | 9131/10682 [1:17:36<12:46,  2.02it/s]
 85%|████████▌ | 9132/10682 [1:17:37<12:46,  2.02it/s]
 85%|████████▌ | 9133/10682 [1:17:37<12:45,  2.02it/s]
 86%|████████▌ | 9134/10682 [1:17:38<12:45,  2.02it/s]
 86%|████████▌ | 9135/10682 [1:17:38<12:44,  2.02it/s]
 86%|████████▌ | 9136/10682 [1:17:39<12:44,  2.02it/s]
 86%|████████▌ | 9137/10682 [1:17:39<12:44,  2.02it/s]
 86%|████████▌ | 9138/10682 [1:17:40<12:43,  2.02it/s]
 86%|████████▌ | 9139/10682 [1:17:40<12:42,  2.02it/s]
 86%|████████▌ | 9140/10682 [1:17:41<12:42,  2.02it/s]
 86%|████████▌ | 9141/10682 [1:17:41<12:41,  2.02it/s]
 86%|████████▌ | 9142/10682 [1:17:42<12:41,  2.02it/s]
 86%|████████▌ | 9143/10682 [1:17:42<12:41,  2.02it/s]
 86%|████████▌ | 9144/10682 [1:17:42<12:41,  2.02it/s]
 86%|████████▌ | 9145/10682 [1:17:43<12:40,  2.02it/s]
 86%|████████▌ | 9146/10682 [1:17:43<12:39,  2.02it/s]
 86%|████████▌ | 9147/10682 [1:17:44<12:38,  2.02it/s]
 86%|████████▌ | 9148/10682 [1:17:44<12:38,  2.02it/s]
 86%|████████▌ | 9149/10682 [1:17:45<12:36,  2.03it/s]
 86%|████████▌ | 9150/10682 [1:17:45<12:36,  2.02it/s]
                                                      
 
 86%|████████▌ | 9150/10682 [1:17:45<12:36,  2.02it/s]
 86%|████████▌ | 9151/10682 [1:17:46<12:36,  2.02it/s]
 86%|████████▌ | 9152/10682 [1:17:46<12:36,  2.02it/s]
 86%|████████▌ | 9153/10682 [1:17:47<12:35,  2.02it/s]
 86%|████████▌ | 9154/10682 [1:17:47<12:34,  2.02it/s]
 86%|████████▌ | 9155/10682 [1:17:48<12:34,  2.02it/s]
 86%|████████▌ | 9156/10682 [1:17:48<12:34,  2.02it/s]
 86%|████████▌ | 9157/10682 [1:17:49<12:34,  2.02it/s]
 86%|████████▌ | 9158/10682 [1:17:49<12:34,  2.02it/s]
 86%|████████▌ | 9159/10682 [1:17:50<12:33,  2.02it/s]
 86%|████████▌ | 9160/10682 [1:17:50<12:32,  2.02it/s]
 86%|████████▌ | 9161/10682 [1:17:51<12:32,  2.02it/s]
 86%|████████▌ | 9162/10682 [1:17:51<12:23,  2.04it/s]
 86%|████████▌ | 9163/10682 [1:18:03<1:39:32,  3.93s/it]
 86%|████████▌ | 9164/10682 [1:18:04<1:13:27,  2.90s/it]
 86%|████████▌ | 9165/10682 [1:18:04<55:07,  2.18s/it]  
 86%|████████▌ | 9166/10682 [1:18:05<42:23,  1.68s/it]
 86%|████████▌ | 9167/10682 [1:18:05<33:24,  1.32s/it]
 86%|████████▌ | 9168/10682 [1:18:06<27:07,  1.07s/it]
 86%|████████▌ | 9169/10682 [1:18:06<22:43,  1.11it/s]
 86%|████████▌ | 9170/10682 [1:18:07<19:41,  1.28it/s]
 86%|████████▌ | 9171/10682 [1:18:07<17:30,  1.44it/s]
 86%|████████▌ | 9172/10682 [1:18:08<15:58,  1.57it/s]
 86%|████████▌ | 9173/10682 [1:18:08<14:55,  1.69it/s]
 86%|████████▌ | 9174/10682 [1:18:09<14:09,  1.78it/s]
 86%|████████▌ | 9175/10682 [1:18:09<13:39,  1.84it/s]
                                                      
-
 86%|████████▌ | 9175/10682 [1:18:09<13:39,  1.84it/s]
 86%|████████▌ | 9176/10682 [1:18:10<13:16,  1.89it/s]
 86%|████████▌ | 9177/10682 [1:18:10<13:00,  1.93it/s]
 86%|████████▌ | 9178/10682 [1:18:11<12:49,  1.95it/s]
 86%|████████▌ | 9179/10682 [1:18:11<12:41,  1.97it/s]
 86%|████████▌ | 9180/10682 [1:18:12<12:35,  1.99it/s]
 86%|████████▌ | 9181/10682 [1:18:12<12:32,  1.99it/s]
 86%|████████▌ | 9182/10682 [1:18:13<12:28,  2.00it/s]
 86%|████████▌ | 9183/10682 [1:18:13<12:26,  2.01it/s]
 86%|████████▌ | 9184/10682 [1:18:14<12:24,  2.01it/s]
 86%|████████▌ | 9185/10682 [1:18:14<12:23,  2.01it/s]
 86%|████████▌ | 9186/10682 [1:18:15<12:21,  2.02it/s]
+
 86%|████████▌ | 9175/10682 [1:18:09<13:39,  1.84it/s]
 86%|████████▌ | 9176/10682 [1:18:10<13:16,  1.89it/s]
 86%|████████▌ | 9177/10682 [1:18:10<13:00,  1.93it/s]
 86%|████████▌ | 9178/10682 [1:18:11<12:49,  1.95it/s]
 86%|████████▌ | 9179/10682 [1:18:11<12:41,  1.97it/s]
 86%|████████▌ | 9180/10682 [1:18:12<12:35,  1.99it/s]
 86%|████████▌ | 9181/10682 [1:18:12<12:32,  1.99it/s]
 86%|████████▌ | 9182/10682 [1:18:13<12:28,  2.00it/s]
 86%|████████▌ | 9183/10682 [1:18:13<12:26,  2.01it/s]
 86%|████████▌ | 9184/10682 [1:18:14<12:24,  2.01it/s]
 86%|████████▌ | 9185/10682 [1:18:14<12:23,  2.01it/s]
 86%|████████▌ | 9186/10682 [1:18:15<12:21,  2.02it/s]
 86%|████████▌ | 9187/10682 [1:18:15<12:25,  2.01it/s]
 86%|████████▌ | 9188/10682 [1:18:16<12:23,  2.01it/s]
 86%|████████▌ | 9189/10682 [1:18:16<12:21,  2.01it/s]
 86%|████████▌ | 9190/10682 [1:18:17<12:20,  2.02it/s]
 86%|████████▌ | 9191/10682 [1:18:17<12:18,  2.02it/s]
 86%|████████▌ | 9192/10682 [1:18:18<12:18,  2.02it/s]
 86%|████████▌ | 9193/10682 [1:18:18<12:16,  2.02it/s]
 86%|████████▌ | 9194/10682 [1:18:19<12:15,  2.02it/s]
 86%|████████▌ | 9195/10682 [1:18:19<12:14,  2.02it/s]
 86%|████████▌ | 9196/10682 [1:18:20<12:14,  2.02it/s]
 86%|████████▌ | 9197/10682 [1:18:20<12:13,  2.02it/s]
 86%|████████▌ | 9198/10682 [1:18:21<12:12,  2.02it/s]
 86%|████████▌ | 9199/10682 [1:18:21<12:12,  2.02it/s]
 86%|████████▌ | 9200/10682 [1:18:22<12:12,  2.02it/s]{'loss': 2.7368, 'grad_norm': 0.28075122833251953, 'learning_rate': 5.7505864256519716e-05, 'epoch': 12.05}
                                                      
+
 86%|████████▌ | 9200/10682 [1:18:22<12:12,  2.02it/s]
 86%|████████▌ | 9201/10682 [1:18:22<12:12,  2.02it/s]
 86%|████████▌ | 9202/10682 [1:18:23<12:11,  2.02it/s]
 86%|████████▌ | 9203/10682 [1:18:23<12:10,  2.02it/s]
 86%|████████▌ | 9204/10682 [1:18:24<12:10,  2.02it/s]
 86%|████████▌ | 9205/10682 [1:18:24<12:09,  2.02it/s]
 86%|████████▌ | 9206/10682 [1:18:25<12:09,  2.02it/s]
 86%|████████▌ | 9207/10682 [1:18:25<12:08,  2.02it/s]
 86%|████████▌ | 9208/10682 [1:18:26<12:07,  2.03it/s]
 86%|████████▌ | 9209/10682 [1:18:26<12:07,  2.02it/s]
 86%|████████▌ | 9210/10682 [1:18:27<12:06,  2.03it/s]
 86%|████████▌ | 9211/10682 [1:18:27<12:06,  2.02it/s]
 86%|████████▌ | 9212/10682 [1:18:28<12:05,  2.02it/s]
 86%|████████▌ | 9213/10682 [1:18:28<12:05,  2.02it/s]
 86%|████████▋ | 9214/10682 [1:18:29<12:04,  2.03it/s]
 86%|████████▋ | 9215/10682 [1:18:29<12:05,  2.02it/s]
 86%|████████▋ | 9216/10682 [1:18:30<12:04,  2.02it/s]
 86%|████████▋ | 9217/10682 [1:18:30<12:04,  2.02it/s]
 86%|████████▋ | 9218/10682 [1:18:31<12:03,  2.02it/s]
 86%|████████▋ | 9219/10682 [1:18:31<12:03,  2.02it/s]
 86%|████████▋ | 9220/10682 [1:18:32<12:03,  2.02it/s]
 86%|████████▋ | 9221/10682 [1:18:32<12:03,  2.02it/s]
 86%|████████▋ | 9222/10682 [1:18:33<12:01,  2.02it/s]
 86%|████████▋ | 9223/10682 [1:18:33<12:01,  2.02it/s]
 86%|████████▋ | 9224/10682 [1:18:34<12:00,  2.02it/s]
 86%|████████▋ | 9225/10682 [1:18:34<11:59,  2.02it/s]{'loss': 2.7433, 'grad_norm': 0.2763462960720062, 'learning_rate': 5.561858464291258e-05, 'epoch': 12.08}
+
                                                      
 86%|████████▋ | 9225/10682 [1:18:34<11:59,  2.02it/s]
 86%|████████▋ | 9226/10682 [1:18:35<11:59,  2.02it/s]
 86%|████████▋ | 9227/10682 [1:18:35<12:00,  2.02it/s]
 86%|████████▋ | 9228/10682 [1:18:36<11:58,  2.02it/s]
 86%|████████▋ | 9229/10682 [1:18:36<11:58,  2.02it/s]
 86%|████████▋ | 9230/10682 [1:18:37<11:57,  2.02it/s]
 86%|████████▋ | 9231/10682 [1:18:37<11:57,  2.02it/s]
 86%|████████▋ | 9232/10682 [1:18:37<11:56,  2.02it/s]
 86%|████████▋ | 9233/10682 [1:18:38<11:56,  2.02it/s]
 86%|████████▋ | 9234/10682 [1:18:38<11:55,  2.02it/s]
 86%|████████▋ | 9235/10682 [1:18:39<11:55,  2.02it/s]
 86%|████████▋ | 9236/10682 [1:18:39<11:54,  2.02it/s]
 86%|████████▋ | 9237/10682 [1:18:40<11:54,  2.02it/s]
 86%|████████▋ | 9238/10682 [1:18:40<11:54,  2.02it/s]
 86%|████████▋ | 9239/10682 [1:18:41<11:53,  2.02it/s]
 87%|████████▋ | 9240/10682 [1:18:41<11:52,  2.02it/s]
 87%|████████▋ | 9241/10682 [1:18:42<11:53,  2.02it/s]
 87%|████████▋ | 9242/10682 [1:18:42<11:52,  2.02it/s]
 87%|████████▋ | 9243/10682 [1:18:43<11:51,  2.02it/s]
 87%|████████▋ | 9244/10682 [1:18:43<11:51,  2.02it/s]
 87%|████████▋ | 9245/10682 [1:18:44<11:50,  2.02it/s]
 87%|████████▋ | 9246/10682 [1:18:44<11:50,  2.02it/s]
 87%|████████▋ | 9247/10682 [1:18:45<11:49,  2.02it/s]
 87%|████████▋ | 9248/10682 [1:18:45<11:49,  2.02it/s]
 87%|████████▋ | 9249/10682 [1:18:46<11:47,  2.02it/s]
 87%|████████▋ | 9250/10682 [1:18:46<11:47,  2.02it/s]
                                                      
+
 87%|████████▋ | 9250/10682 [1:18:46<11:47,  2.02it/s]
 87%|████████▋ | 9251/10682 [1:18:47<11:47,  2.02it/s]
 87%|████████▋ | 9252/10682 [1:18:47<11:46,  2.02it/s]
 87%|████████▋ | 9253/10682 [1:18:48<11:45,  2.03it/s]
 87%|████████▋ | 9254/10682 [1:18:48<11:45,  2.02it/s]
 87%|████████▋ | 9255/10682 [1:18:49<11:44,  2.02it/s]
 87%|████████▋ | 9256/10682 [1:18:49<11:45,  2.02it/s]
 87%|████████▋ | 9257/10682 [1:18:50<11:44,  2.02it/s]
 87%|████████▋ | 9258/10682 [1:18:50<11:44,  2.02it/s]
 87%|████████▋ | 9259/10682 [1:18:51<11:43,  2.02it/s]
 87%|████████▋ | 9260/10682 [1:18:51<11:43,  2.02it/s]
 87%|████████▋ | 9261/10682 [1:18:52<11:42,  2.02it/s]
 87%|████████▋ | 9262/10682 [1:18:52<11:41,  2.02it/s]
 87%|████████▋ | 9263/10682 [1:18:53<11:41,  2.02it/s]
 87%|████████▋ | 9264/10682 [1:18:53<11:40,  2.02it/s]
 87%|████████▋ | 9265/10682 [1:18:54<11:40,  2.02it/s]
 87%|████████▋ | 9266/10682 [1:18:54<11:39,  2.02it/s]
 87%|████████▋ | 9267/10682 [1:18:55<11:39,  2.02it/s]
 87%|████████▋ | 9268/10682 [1:18:55<11:38,  2.02it/s]
 87%|████████▋ | 9269/10682 [1:18:56<11:38,  2.02it/s]
 87%|████████▋ | 9270/10682 [1:18:56<11:37,  2.02it/s]
 87%|████████▋ | 9271/10682 [1:18:57<11:37,  2.02it/s]
 87%|████████▋ | 9272/10682 [1:18:57<11:36,  2.02it/s]
 87%|████████▋ | 9273/10682 [1:18:58<11:36,  2.02it/s]
 87%|████████▋ | 9274/10682 [1:18:58<11:35,  2.02it/s]
 87%|████████▋ | 9275/10682 [1:18:59<11:35,  2.02it/s]
                                                      
+
 87%|████████▋ | 9275/10682 [1:18:59<11:35,  2.02it/s]
 87%|████████▋ | 9276/10682 [1:18:59<11:35,  2.02it/s]
 87%|████████▋ | 9277/10682 [1:19:00<11:35,  2.02it/s]
 87%|████████▋ | 9278/10682 [1:19:00<11:34,  2.02it/s]
 87%|████████▋ | 9279/10682 [1:19:01<11:34,  2.02it/s]
 87%|████████▋ | 9280/10682 [1:19:01<11:32,  2.02it/s]
 87%|████████▋ | 9281/10682 [1:19:02<11:32,  2.02it/s]
 87%|████████▋ | 9282/10682 [1:19:02<11:31,  2.02it/s]
 87%|████████▋ | 9283/10682 [1:19:03<11:31,  2.02it/s]
 87%|████████▋ | 9284/10682 [1:19:03<11:30,  2.02it/s]
 87%|████████▋ | 9285/10682 [1:19:04<11:31,  2.02it/s]
 87%|████████▋ | 9286/10682 [1:19:04<11:30,  2.02it/s]
 87%|████████▋ | 9287/10682 [1:19:05<11:30,  2.02it/s]
 87%|████████▋ | 9288/10682 [1:19:05<11:29,  2.02it/s]
 87%|████████▋ | 9289/10682 [1:19:06<11:29,  2.02it/s]
 87%|████████▋ | 9290/10682 [1:19:06<11:28,  2.02it/s]
 87%|████████▋ | 9291/10682 [1:19:07<11:28,  2.02it/s]
 87%|████████▋ | 9292/10682 [1:19:07<11:28,  2.02it/s]
 87%|████████▋ | 9293/10682 [1:19:08<11:27,  2.02it/s]
 87%|████████▋ | 9294/10682 [1:19:08<11:27,  2.02it/s]
 87%|████████▋ | 9295/10682 [1:19:09<11:26,  2.02it/s]
 87%|████████▋ | 9296/10682 [1:19:09<11:25,  2.02it/s]
 87%|████████▋ | 9297/10682 [1:19:10<11:25,  2.02it/s]
 87%|████████▋ | 9298/10682 [1:19:10<11:24,  2.02it/s]
 87%|████████▋ | 9299/10682 [1:19:11<11:24,  2.02it/s]
 87%|████████▋ | 9300/10682 [1:19:11<11:24,  2.02it/s]{'loss': 2.7346, 'grad_norm': 0.27623340487480164, 'learning_rate': 5.0135217920839137e-05, 'epoch': 12.18}
+
                                                      
 87%|████████▋ | 9300/10682 [1:19:11<11:24,  2.02it/s]
 87%|████████▋ | 9301/10682 [1:19:12<11:24,  2.02it/s]
 87%|████████▋ | 9302/10682 [1:19:12<11:22,  2.02it/s]
 87%|████████▋ | 9303/10682 [1:19:13<11:22,  2.02it/s]
 87%|████████▋ | 9304/10682 [1:19:13<11:21,  2.02it/s]
 87%|████████▋ | 9305/10682 [1:19:14<11:21,  2.02it/s]
 87%|████████▋ | 9306/10682 [1:19:14<11:19,  2.02it/s]
 87%|████████▋ | 9307/10682 [1:19:15<11:19,  2.02it/s]
 87%|████████▋ | 9308/10682 [1:19:15<11:18,  2.02it/s]
 87%|████████▋ | 9309/10682 [1:19:16<11:18,  2.02it/s]
 87%|████████▋ | 9310/10682 [1:19:16<11:17,  2.02it/s]
 87%|████████▋ | 9311/10682 [1:19:17<11:17,  2.02it/s]
 87%|████████▋ | 9312/10682 [1:19:17<11:16,  2.03it/s]
 87%|████████▋ | 9313/10682 [1:19:18<11:16,  2.02it/s]
 87%|████████▋ | 9314/10682 [1:19:18<11:15,  2.02it/s]
 87%|████████▋ | 9315/10682 [1:19:19<11:15,  2.02it/s]
 87%|████████▋ | 9316/10682 [1:19:19<11:14,  2.02it/s]
 87%|████████▋ | 9317/10682 [1:19:20<11:14,  2.02it/s]
 87%|████████▋ | 9318/10682 [1:19:20<11:13,  2.02it/s]
 87%|████████▋ | 9319/10682 [1:19:21<11:15,  2.02it/s]
 87%|████████▋ | 9320/10682 [1:19:21<11:14,  2.02it/s]
 87%|████████▋ | 9321/10682 [1:19:22<11:13,  2.02it/s]
 87%|████████▋ | 9322/10682 [1:19:22<11:12,  2.02it/s]
 87%|████████▋ | 9323/10682 [1:19:22<11:11,  2.02it/s]
 87%|████████▋ | 9324/10682 [1:19:23<11:11,  2.02it/s]
 87%|████████▋ | 9325/10682 [1:19:23<11:11,  2.02it/s]{'loss': 2.7366, 'grad_norm': 0.2855311632156372, 'learning_rate': 4.836732641133895e-05, 'epoch': 12.21}
                                                      
+
 87%|████████▋ | 9325/10682 [1:19:23<11:11,  2.02it/s]
 87%|████████▋ | 9326/10682 [1:19:24<11:11,  2.02it/s]
 87%|████████▋ | 9327/10682 [1:19:24<11:10,  2.02it/s]
 87%|████████▋ | 9328/10682 [1:19:25<11:09,  2.02it/s]
 87%|████████▋ | 9329/10682 [1:19:25<11:08,  2.02it/s]
 87%|████████▋ | 9330/10682 [1:19:26<11:08,  2.02it/s]
 87%|████████▋ | 9331/10682 [1:19:26<11:08,  2.02it/s]
 87%|████████▋ | 9332/10682 [1:19:27<11:07,  2.02it/s]
 87%|████████▋ | 9333/10682 [1:19:27<11:06,  2.02it/s]
 87%|████████▋ | 9334/10682 [1:19:28<11:07,  2.02it/s]
 87%|████████▋ | 9335/10682 [1:19:28<11:06,  2.02it/s]
 87%|████████▋ | 9336/10682 [1:19:29<11:05,  2.02it/s]
 87%|████████▋ | 9337/10682 [1:19:29<11:05,  2.02it/s]
 87%|███��████▋ | 9338/10682 [1:19:30<11:04,  2.02it/s]
 87%|████████▋ | 9339/10682 [1:19:30<11:04,  2.02it/s]
 87%|████████▋ | 9340/10682 [1:19:31<11:03,  2.02it/s]
 87%|████████▋ | 9341/10682 [1:19:31<11:02,  2.02it/s]
 87%|████████▋ | 9342/10682 [1:19:32<11:02,  2.02it/s]
 87%|████████▋ | 9343/10682 [1:19:32<11:01,  2.02it/s]
 87%|████████▋ | 9344/10682 [1:19:33<11:01,  2.02it/s]
 87%|████████▋ | 9345/10682 [1:19:33<11:00,  2.02it/s]
 87%|████████▋ | 9346/10682 [1:19:34<11:00,  2.02it/s]
 88%|████████▊ | 9347/10682 [1:19:34<11:00,  2.02it/s]
 88%|████████▊ | 9348/10682 [1:19:35<10:59,  2.02it/s]
 88%|████████▊ | 9349/10682 [1:19:35<10:59,  2.02it/s]
 88%|████████▊ | 9350/10682 [1:19:36<10:58,  2.02it/s]{'loss': 2.7481, 'grad_norm': 0.28073567152023315, 'learning_rate': 4.662958195146971e-05, 'epoch': 12.25}
+
                                                      
 88%|████████▊ | 9350/10682 [1:19:36<10:58,  2.02it/s]
 88%|████████▊ | 9351/10682 [1:19:36<10:58,  2.02it/s]
 88%|████████▊ | 9352/10682 [1:19:37<10:57,  2.02it/s]
 88%|████████▊ | 9353/10682 [1:19:37<10:56,  2.02it/s]
 88%|████████▊ | 9354/10682 [1:19:38<10:56,  2.02it/s]
 88%|████████▊ | 9355/10682 [1:19:38<10:55,  2.02it/s]
 88%|████████▊ | 9356/10682 [1:19:39<10:55,  2.02it/s]
 88%|████████▊ | 9357/10682 [1:19:39<10:55,  2.02it/s]
 88%|████████▊ | 9358/10682 [1:19:40<10:54,  2.02it/s]
 88%|████████▊ | 9359/10682 [1:19:40<10:54,  2.02it/s]
 88%|████████▊ | 9360/10682 [1:19:41<10:54,  2.02it/s]
 88%|████████▊ | 9361/10682 [1:19:41<10:54,  2.02it/s]
 88%|████████▊ | 9362/10682 [1:19:42<10:53,  2.02it/s]
 88%|████████▊ | 9363/10682 [1:19:42<10:53,  2.02it/s]
 88%|████████▊ | 9364/10682 [1:19:43<10:52,  2.02it/s]
 88%|████████▊ | 9365/10682 [1:19:43<10:51,  2.02it/s]
 88%|████████▊ | 9366/10682 [1:19:44<10:50,  2.02it/s]
 88%|████████▊ | 9367/10682 [1:19:44<10:50,  2.02it/s]
 88%|████████▊ | 9368/10682 [1:19:45<10:49,  2.02it/s]
 88%|████████▊ | 9369/10682 [1:19:45<10:49,  2.02it/s]
 88%|████████▊ | 9370/10682 [1:19:46<10:48,  2.02it/s]
 88%|████████▊ | 9371/10682 [1:19:46<10:48,  2.02it/s]
 88%|████████▊ | 9372/10682 [1:19:47<10:47,  2.02it/s]
 88%|████████▊ | 9373/10682 [1:19:47<10:47,  2.02it/s]
 88%|████████▊ | 9374/10682 [1:19:48<10:46,  2.02it/s]
 88%|████████▊ | 9375/10682 [1:19:48<10:45,  2.03it/s]
                                                      
+
 88%|████████▊ | 9375/10682 [1:19:48<10:45,  2.03it/s]
 88%|████████▊ | 9376/10682 [1:19:49<10:45,  2.02it/s]
 88%|████████▊ | 9377/10682 [1:19:49<10:44,  2.02it/s]
 88%|████████▊ | 9378/10682 [1:19:50<10:44,  2.02it/s]
 88%|████████▊ | 9379/10682 [1:19:50<10:43,  2.02it/s]
 88%|████████▊ | 9380/10682 [1:19:51<10:44,  2.02it/s]
 88%|████████▊ | 9381/10682 [1:19:51<10:42,  2.02it/s]
 88%|████████▊ | 9382/10682 [1:19:52<10:42,  2.02it/s]
 88%|████████▊ | 9383/10682 [1:19:52<10:41,  2.03it/s]
 88%|████████▊ | 9384/10682 [1:19:53<10:41,  2.02it/s]
 88%|████████▊ | 9385/10682 [1:19:53<10:40,  2.03it/s]
 88%|████████▊ | 9386/10682 [1:19:54<10:40,  2.02it/s]
 88%|████████▊ | 9387/10682 [1:19:54<10:39,  2.02it/s]
 88%|████████▊ | 9388/10682 [1:19:55<10:39,  2.02it/s]
 88%|████████▊ | 9389/10682 [1:19:55<10:39,  2.02it/s]
 88%|████████▊ | 9390/10682 [1:19:56<10:37,  2.03it/s]
 88%|████████▊ | 9391/10682 [1:19:56<10:38,  2.02it/s]
 88%|████████▊ | 9392/10682 [1:19:57<10:37,  2.02it/s]
 88%|████████▊ | 9393/10682 [1:19:57<10:36,  2.02it/s]
 88%|████████▊ | 9394/10682 [1:19:58<10:35,  2.03it/s]
 88%|████████▊ | 9395/10682 [1:19:58<10:35,  2.02it/s]
 88%|████████▊ | 9396/10682 [1:19:59<10:35,  2.02it/s]
 88%|████████▊ | 9397/10682 [1:19:59<10:35,  2.02it/s]
 88%|████████▊ | 9398/10682 [1:20:00<10:34,  2.02it/s]
 88%|████████▊ | 9399/10682 [1:20:00<10:34,  2.02it/s]
 88%|████████▊ | 9400/10682 [1:20:01<10:33,  2.02it/s]{'loss': 2.7415, 'grad_norm': 0.27810999751091003, 'learning_rate': 4.3244996147050855e-05, 'epoch': 12.31}
                                                      
+
 88%|████████▊ | 9400/10682 [1:20:01<10:33,  2.02it/s]
 88%|████████▊ | 9401/10682 [1:20:01<10:34,  2.02it/s]
 88%|████████▊ | 9402/10682 [1:20:02<10:33,  2.02it/s]
 88%|████████▊ | 9403/10682 [1:20:02<10:32,  2.02it/s]
 88%|████████▊ | 9404/10682 [1:20:03<10:32,  2.02it/s]
 88%|████████▊ | 9405/10682 [1:20:03<10:32,  2.02it/s]
 88%|████████▊ | 9406/10682 [1:20:04<10:31,  2.02it/s]
 88%|████████▊ | 9407/10682 [1:20:04<10:30,  2.02it/s]
 88%|████████▊ | 9408/10682 [1:20:05<10:30,  2.02it/s]
 88%|████████▊ | 9409/10682 [1:20:05<10:29,  2.02it/s]
 88%|████████▊ | 9410/10682 [1:20:06<10:29,  2.02it/s]
 88%|████████▊ | 9411/10682 [1:20:06<10:28,  2.02it/s]
 88%|████████▊ | 9412/10682 [1:20:07<10:28,  2.02it/s]
 88%|████████▊ | 9413/10682 [1:20:07<10:27,  2.02it/s]
 88%|████████▊ | 9414/10682 [1:20:07<10:27,  2.02it/s]
 88%|████████▊ | 9415/10682 [1:20:08<10:26,  2.02it/s]
 88%|████████▊ | 9416/10682 [1:20:08<10:26,  2.02it/s]
 88%|████████▊ | 9417/10682 [1:20:09<10:26,  2.02it/s]
 88%|████████▊ | 9418/10682 [1:20:09<10:25,  2.02it/s]
 88%|████████▊ | 9419/10682 [1:20:10<10:25,  2.02it/s]
 88%|████████▊ | 9420/10682 [1:20:10<10:24,  2.02it/s]
 88%|████████▊ | 9421/10682 [1:20:11<10:23,  2.02it/s]
 88%|████████▊ | 9422/10682 [1:20:11<10:23,  2.02it/s]
 88%|████████▊ | 9423/10682 [1:20:12<10:23,  2.02it/s]
 88%|████████▊ | 9424/10682 [1:20:12<10:22,  2.02it/s]
 88%|████████▊ | 9425/10682 [1:20:13<10:21,  2.02it/s]
                                                      
+
 88%|████████▊ | 9425/10682 [1:20:13<10:21,  2.02it/s]
 88%|████████▊ | 9426/10682 [1:20:13<10:21,  2.02it/s]
 88%|████████▊ | 9427/10682 [1:20:14<10:21,  2.02it/s]
 88%|████████▊ | 9428/10682 [1:20:14<10:20,  2.02it/s]
 88%|████████▊ | 9429/10682 [1:20:15<10:19,  2.02it/s]
 88%|████████▊ | 9430/10682 [1:20:15<10:18,  2.02it/s]
 88%|████████▊ | 9431/10682 [1:20:16<10:18,  2.02it/s]
 88%|████████▊ | 9432/10682 [1:20:16<10:17,  2.02it/s]
 88%|████████▊ | 9433/10682 [1:20:17<10:17,  2.02it/s]
 88%|████████▊ | 9434/10682 [1:20:17<10:16,  2.02it/s]
 88%|████████▊ | 9435/10682 [1:20:18<10:16,  2.02it/s]
 88%|████████▊ | 9436/10682 [1:20:18<10:15,  2.02it/s]
 88%|████████▊ | 9437/10682 [1:20:19<10:15,  2.02it/s]
 88%|████████▊ | 9438/10682 [1:20:19<10:14,  2.02it/s]
 88%|████████▊ | 9439/10682 [1:20:20<10:14,  2.02it/s]
 88%|████████▊ | 9440/10682 [1:20:20<10:13,  2.02it/s]
 88%|████████▊ | 9441/10682 [1:20:21<10:13,  2.02it/s]
 88%|████████▊ | 9442/10682 [1:20:21<10:12,  2.02it/s]
 88%|████████▊ | 9443/10682 [1:20:22<10:12,  2.02it/s]
 88%|████████▊ | 9444/10682 [1:20:22<10:11,  2.02it/s]
 88%|████████▊ | 9445/10682 [1:20:23<10:12,  2.02it/s]
 88%|████████▊ | 9446/10682 [1:20:23<10:11,  2.02it/s]
 88%|████████▊ | 9447/10682 [1:20:24<10:10,  2.02it/s]
 88%|████████▊ | 9448/10682 [1:20:24<10:09,  2.02it/s]
 88%|████████▊ | 9449/10682 [1:20:25<10:08,  2.03it/s]
 88%|████████▊ | 9450/10682 [1:20:25<10:08,  2.02it/s]
                                                      {'loss': 2.7417, 'grad_norm': 0.2766771912574768, 'learning_rate': 3.998236419395806e-05, 'epoch': 12.38}
+
 88%|████████▊ | 9450/10682 [1:20:25<10:08,  2.02it/s]
 88%|████████▊ | 9451/10682 [1:20:26<10:08,  2.02it/s]
 88%|████████▊ | 9452/10682 [1:20:26<10:07,  2.02it/s]
 88%|████████▊ | 9453/10682 [1:20:27<10:07,  2.02it/s]
 89%|████████▊ | 9454/10682 [1:20:27<10:06,  2.02it/s]
 89%|████████▊ | 9455/10682 [1:20:28<10:06,  2.02it/s]
 89%|████████▊ | 9456/10682 [1:20:28<10:05,  2.02it/s]
 89%|████████▊ | 9457/10682 [1:20:29<10:05,  2.02it/s]
 89%|████████▊ | 9458/10682 [1:20:29<10:05,  2.02it/s]
 89%|████████▊ | 9459/10682 [1:20:30<10:57,  1.86it/s]
 89%|████████▊ | 9460/10682 [1:20:30<10:40,  1.91it/s]
 89%|████████▊ | 9461/10682 [1:20:31<10:29,  1.94it/s]
 89%|████��███▊ | 9462/10682 [1:20:31<10:21,  1.96it/s]
 89%|████████▊ | 9463/10682 [1:20:32<10:15,  1.98it/s]
 89%|████████▊ | 9464/10682 [1:20:32<10:10,  1.99it/s]
 89%|████████▊ | 9465/10682 [1:20:33<10:07,  2.00it/s]
 89%|████████▊ | 9466/10682 [1:20:33<10:05,  2.01it/s]
 89%|████████▊ | 9467/10682 [1:20:34<10:02,  2.02it/s]
 89%|████████▊ | 9468/10682 [1:20:34<10:01,  2.02it/s]
 89%|████████▊ | 9469/10682 [1:20:35<10:00,  2.02it/s]
 89%|████████▊ | 9470/10682 [1:20:35<09:59,  2.02it/s]
 89%|████████▊ | 9471/10682 [1:20:36<09:58,  2.02it/s]
 89%|████████▊ | 9472/10682 [1:20:36<09:58,  2.02it/s]
 89%|████████▊ | 9473/10682 [1:20:37<09:57,  2.02it/s]
 89%|████████▊ | 9474/10682 [1:20:37<09:56,  2.02it/s]
 89%|████████▊ | 9475/10682 [1:20:38<09:55,  2.03it/s]{'loss': 2.7382, 'grad_norm': 0.27711209654808044, 'learning_rate': 3.839705441646779e-05, 'epoch': 12.41}
                                                      
+
 89%|████████▊ | 9475/10682 [1:20:38<09:55,  2.03it/s]
 89%|████████▊ | 9476/10682 [1:20:38<09:58,  2.02it/s]
 89%|████████▊ | 9477/10682 [1:20:39<09:56,  2.02it/s]
 89%|████████▊ | 9478/10682 [1:20:39<09:56,  2.02it/s]
 89%|████████▊ | 9479/10682 [1:20:40<09:55,  2.02it/s]
 89%|████████▊ | 9480/10682 [1:20:40<09:54,  2.02it/s]
 89%|████████▉ | 9481/10682 [1:20:41<09:53,  2.02it/s]
 89%|████████▉ | 9482/10682 [1:20:41<09:53,  2.02it/s]
 89%|████████▉ | 9483/10682 [1:20:42<09:53,  2.02it/s]
 89%|████████▉ | 9484/10682 [1:20:42<09:52,  2.02it/s]
 89%|████████▉ | 9485/10682 [1:20:43<09:51,  2.02it/s]
 89%|████████▉ | 9486/10682 [1:20:43<09:51,  2.02it/s]
 89%|████████▉ | 9487/10682 [1:20:44<09:50,  2.02it/s]
 89%|████████▉ | 9488/10682 [1:20:44<09:50,  2.02it/s]
 89%|████████▉ | 9489/10682 [1:20:45<09:49,  2.02it/s]
 89%|████████▉ | 9490/10682 [1:20:45<09:49,  2.02it/s]
 89%|████████▉ | 9491/10682 [1:20:46<09:48,  2.02it/s]
 89%|████████▉ | 9492/10682 [1:20:46<09:47,  2.02it/s]
 89%|████████▉ | 9493/10682 [1:20:47<09:46,  2.03it/s]
 89%|████████▉ | 9494/10682 [1:20:47<09:46,  2.02it/s]
 89%|████████▉ | 9495/10682 [1:20:48<09:46,  2.02it/s]
 89%|████████▉ | 9496/10682 [1:20:48<09:46,  2.02it/s]
 89%|████████▉ | 9497/10682 [1:20:49<09:45,  2.02it/s]
 89%|████████▉ | 9498/10682 [1:20:49<09:45,  2.02it/s]
 89%|████████▉ | 9499/10682 [1:20:50<09:44,  2.02it/s]
 89%|████████▉ | 9500/10682 [1:20:50<09:43,  2.02it/s]
                                                      
+
 89%|████████▉ | 9500/10682 [1:20:50<09:43,  2.02it/s]
 89%|████████▉ | 9501/10682 [1:20:51<09:44,  2.02it/s]
 89%|████████▉ | 9502/10682 [1:20:51<09:43,  2.02it/s]
 89%|████████▉ | 9503/10682 [1:20:52<09:43,  2.02it/s]
 89%|████████▉ | 9504/10682 [1:20:52<09:42,  2.02it/s]
 89%|████████▉ | 9505/10682 [1:20:53<09:42,  2.02it/s]
 89%|████████▉ | 9506/10682 [1:20:53<09:41,  2.02it/s]
 89%|████████▉ | 9507/10682 [1:20:54<09:40,  2.02it/s]
 89%|████████▉ | 9508/10682 [1:20:54<09:39,  2.03it/s]
 89%|████████▉ | 9509/10682 [1:20:55<09:39,  2.02it/s]
 89%|████████▉ | 9510/10682 [1:20:55<09:38,  2.03it/s]
 89%|████████▉ | 9511/10682 [1:20:56<09:38,  2.02it/s]
 89%|████████▉ | 9512/10682 [1:20:56<09:37,  2.02it/s]
 89%|████████▉ | 9513/10682 [1:20:57<09:37,  2.02it/s]
 89%|████████▉ | 9514/10682 [1:20:57<09:37,  2.02it/s]
 89%|████████▉ | 9515/10682 [1:20:58<09:36,  2.02it/s]
 89%|████████▉ | 9516/10682 [1:20:58<09:36,  2.02it/s]
 89%|████████▉ | 9517/10682 [1:20:59<09:35,  2.02it/s]
 89%|████████▉ | 9518/10682 [1:20:59<09:35,  2.02it/s]
 89%|████████▉ | 9519/10682 [1:21:00<09:34,  2.03it/s]
 89%|████████▉ | 9520/10682 [1:21:00<09:34,  2.02it/s]
 89%|████████▉ | 9521/10682 [1:21:01<09:33,  2.02it/s]
 89%|████████▉ | 9522/10682 [1:21:01<09:33,  2.02it/s]
 89%|████████▉ | 9523/10682 [1:21:02<09:32,  2.02it/s]
 89%|████████▉ | 9524/10682 [1:21:02<09:32,  2.02it/s]
 89%|████████▉ | 9525/10682 [1:21:03<09:31,  2.02it/s]
                                                      
+
 89%|████████▉ | 9525/10682 [1:21:03<09:31,  2.02it/s]
 89%|████████▉ | 9526/10682 [1:21:03<09:32,  2.02it/s]
 89%|████████▉ | 9527/10682 [1:21:03<09:31,  2.02it/s]
 89%|████████▉ | 9528/10682 [1:21:04<09:30,  2.02it/s]
 89%|████████▉ | 9529/10682 [1:21:04<09:29,  2.02it/s]
 89%|████████▉ | 9530/10682 [1:21:05<09:28,  2.02it/s]
 89%|████████▉ | 9531/10682 [1:21:05<09:28,  2.02it/s]
 89%|████████▉ | 9532/10682 [1:21:06<09:27,  2.03it/s]
 89%|████████▉ | 9533/10682 [1:21:06<09:27,  2.02it/s]
 89%|████████▉ | 9534/10682 [1:21:07<09:26,  2.02it/s]
 89%|████████▉ | 9535/10682 [1:21:07<09:26,  2.02it/s]
 89%|████████▉ | 9536/10682 [1:21:08<09:26,  2.02it/s]
 89%|████████▉ | 9537/10682 [1:21:08<09:26,  2.02it/s]
 89%|████████▉ | 9538/10682 [1:21:09<09:25,  2.02it/s]
 89%|████████▉ | 9539/10682 [1:21:09<09:25,  2.02it/s]
 89%|████████▉ | 9540/10682 [1:21:10<09:24,  2.02it/s]
 89%|████████▉ | 9541/10682 [1:21:10<09:24,  2.02it/s]
 89%|████████▉ | 9542/10682 [1:21:11<09:23,  2.02it/s]
 89%|████████▉ | 9543/10682 [1:21:11<09:23,  2.02it/s]
 89%|████████▉ | 9544/10682 [1:21:12<09:22,  2.02it/s]
 89%|████████▉ | 9545/10682 [1:21:12<09:21,  2.02it/s]
 89%|████████▉ | 9546/10682 [1:21:13<09:21,  2.02it/s]
 89%|████████▉ | 9547/10682 [1:21:13<09:20,  2.02it/s]
 89%|████████▉ | 9548/10682 [1:21:14<09:20,  2.02it/s]
 89%|████████▉ | 9549/10682 [1:21:14<09:20,  2.02it/s]
 89%|████████▉ | 9550/10682 [1:21:15<09:19,  2.02it/s]
                                                      
+
 89%|████████▉ | 9550/10682 [1:21:15<09:19,  2.02it/s]
 89%|████████▉ | 9551/10682 [1:21:15<09:19,  2.02it/s]
 89%|████████▉ | 9552/10682 [1:21:16<09:18,  2.02it/s]
 89%|████████▉ | 9553/10682 [1:21:16<09:18,  2.02it/s]
 89%|████████▉ | 9554/10682 [1:21:17<09:17,  2.02it/s]
 89%|████████▉ | 9555/10682 [1:21:17<09:17,  2.02it/s]
 89%|████████▉ | 9556/10682 [1:21:18<09:16,  2.02it/s]
 89%|████████▉ | 9557/10682 [1:21:18<09:15,  2.03it/s]
 89%|████████▉ | 9558/10682 [1:21:19<09:15,  2.02it/s]
 89%|████████▉ | 9559/10682 [1:21:19<09:14,  2.03it/s]
 89%|████████▉ | 9560/10682 [1:21:20<09:13,  2.03it/s]
 90%|████████▉ | 9561/10682 [1:21:20<09:13,  2.02it/s]
 90%|████████▉ | 9562/10682 [1:21:21<09:12,  2.03it/s]
 90%|████████▉ | 9563/10682 [1:21:21<09:13,  2.02it/s]
 90%|████████▉ | 9564/10682 [1:21:22<09:12,  2.02it/s]
 90%|████████▉ | 9565/10682 [1:21:22<09:12,  2.02it/s]
 90%|████████▉ | 9566/10682 [1:21:23<09:11,  2.02it/s]
 90%|████████▉ | 9567/10682 [1:21:23<09:11,  2.02it/s]
 90%|████████▉ | 9568/10682 [1:21:24<09:10,  2.02it/s]
 90%|████████▉ | 9569/10682 [1:21:24<09:57,  1.86it/s]
 90%|████████▉ | 9570/10682 [1:21:25<09:43,  1.91it/s]
 90%|████████▉ | 9571/10682 [1:21:25<09:32,  1.94it/s]
 90%|████████▉ | 9572/10682 [1:21:26<09:25,  1.96it/s]
 90%|████████▉ | 9573/10682 [1:21:26<09:20,  1.98it/s]
 90%|████████▉ | 9574/10682 [1:21:27<09:15,  1.99it/s]
 90%|████████▉ | 9575/10682 [1:21:27<09:13,  2.00it/s]
                                                      
+
 90%|████████▉ | 9575/10682 [1:21:27<09:13,  2.00it/s]
 90%|████████▉ | 9576/10682 [1:21:28<09:13,  2.00it/s]
 90%|████████▉ | 9577/10682 [1:21:28<09:10,  2.01it/s]
 90%|████████▉ | 9578/10682 [1:21:29<09:08,  2.01it/s]
 90%|████████▉ | 9579/10682 [1:21:29<09:07,  2.02it/s]
 90%|████████▉ | 9580/10682 [1:21:30<09:06,  2.02it/s]
 90%|████████▉ | 9581/10682 [1:21:30<09:05,  2.02it/s]
 90%|████████▉ | 9582/10682 [1:21:31<09:05,  2.02it/s]
 90%|████████▉ | 9583/10682 [1:21:31<09:03,  2.02it/s]
 90%|████████▉ | 9584/10682 [1:21:32<09:03,  2.02it/s]
 90%|████████▉ | 9585/10682 [1:21:32<09:01,  2.02it/s]
 90%|████████▉ | 9586/10682 [1:21:33<09:01,  2.02it/s]
 90%|████████▉ | 9587/10682 [1:21:33<09:00,  2.03it/s]
 90%|████████▉ | 9588/10682 [1:21:34<09:00,  2.02it/s]
 90%|████████▉ | 9589/10682 [1:21:34<08:59,  2.02it/s]
 90%|████████▉ | 9590/10682 [1:21:35<08:59,  2.02it/s]
 90%|████████▉ | 9591/10682 [1:21:35<08:58,  2.03it/s]
 90%|████████▉ | 9592/10682 [1:21:36<08:58,  2.03it/s]
 90%|████████▉ | 9593/10682 [1:21:36<08:57,  2.03it/s]
 90%|████████▉ | 9594/10682 [1:21:37<08:57,  2.03it/s]
 90%|████████▉ | 9595/10682 [1:21:37<08:57,  2.02it/s]
 90%|████████▉ | 9596/10682 [1:21:38<08:56,  2.02it/s]
 90%|████████▉ | 9597/10682 [1:21:38<08:56,  2.02it/s]
 90%|████████▉ | 9598/10682 [1:21:39<08:55,  2.02it/s]
 90%|████████▉ | 9599/10682 [1:21:39<08:55,  2.02it/s]
 90%|████████▉ | 9600/10682 [1:21:40<08:55,  2.02it/s]
                                                      
+
 90%|████████▉ | 9600/10682 [1:21:40<08:55,  2.02it/s]
 90%|████████▉ | 9601/10682 [1:21:40<08:54,  2.02it/s]
 90%|████████▉ | 9602/10682 [1:21:41<08:54,  2.02it/s]
 90%|████████▉ | 9603/10682 [1:21:41<08:53,  2.02it/s]
 90%|████████▉ | 9604/10682 [1:21:42<08:52,  2.02it/s]
 90%|████████▉ | 9605/10682 [1:21:42<08:52,  2.02it/s]
 90%|████████▉ | 9606/10682 [1:21:43<08:52,  2.02it/s]
 90%|████████▉ | 9607/10682 [1:21:43<08:51,  2.02it/s]
 90%|████████▉ | 9608/10682 [1:21:44<08:51,  2.02it/s]
 90%|████████▉ | 9609/10682 [1:21:44<08:50,  2.02it/s]
 90%|████████▉ | 9610/10682 [1:21:45<08:49,  2.03it/s]
 90%|████████▉ | 9611/10682 [1:21:45<08:49,  2.02it/s]
 90%|████████▉ | 9612/10682 [1:21:46<08:48,  2.03it/s]
 90%|████████▉ | 9613/10682 [1:21:46<08:48,  2.02it/s]
 90%|█████████ | 9614/10682 [1:21:47<08:46,  2.03it/s]
 90%|█████████ | 9615/10682 [1:21:47<08:46,  2.02it/s]
 90%|█████████ | 9616/10682 [1:21:48<08:46,  2.03it/s]
 90%|█████████ | 9617/10682 [1:21:48<08:46,  2.02it/s]
 90%|█████████ | 9618/10682 [1:21:49<08:45,  2.02it/s]
 90%|█████████ | 9619/10682 [1:21:49<08:45,  2.02it/s]
 90%|█████████ | 9620/10682 [1:21:50<08:44,  2.02it/s]
 90%|█████████ | 9621/10682 [1:21:50<08:44,  2.02it/s]
 90%|█████████ | 9622/10682 [1:21:51<08:43,  2.02it/s]
 90%|█████████ | 9623/10682 [1:21:51<08:43,  2.02it/s]
 90%|█████████ | 9624/10682 [1:21:52<08:43,  2.02it/s]
 90%|█████████ | 9625/10682 [1:21:52<08:42,  2.02it/s]{'loss': 2.7489, 'grad_norm': 0.2770747244358063, 'learning_rate': 2.9535819277506203e-05, 'epoch': 12.61}
                                                      
+
 90%|█████████ | 9625/10682 [1:21:52<08:42,  2.02it/s]
 90%|█████████ | 9626/10682 [1:21:53<08:42,  2.02it/s]
 90%|█████████ | 9627/10682 [1:21:53<08:42,  2.02it/s]
 90%|█████████ | 9628/10682 [1:21:54<08:41,  2.02it/s]
 90%|█████████ | 9629/10682 [1:21:54<08:40,  2.02it/s]
 90%|█████████ | 9630/10682 [1:21:55<08:39,  2.02it/s]
 90%|█████████ | 9631/10682 [1:21:55<08:39,  2.02it/s]
 90%|█████████ | 9632/10682 [1:21:56<08:38,  2.02it/s]
 90%|█████████ | 9633/10682 [1:21:56<08:38,  2.02it/s]
 90%|█████████ | 9634/10682 [1:21:57<08:37,  2.02it/s]
 90%|█████████ | 9635/10682 [1:21:57<08:37,  2.02it/s]
 90%|█████████ | 9636/10682 [1:21:58<08:36,  2.02it/s]
 90%|█████████ | 9637/10682 [1:21:58<08:36,  2.02it/s]
 90%|█████████ | 9638/10682 [1:21:59<08:35,  2.02it/s]
 90%|█████████ | 9639/10682 [1:21:59<08:35,  2.02it/s]
 90%|█████████ | 9640/10682 [1:21:59<08:34,  2.02it/s]
 90%|█████████ | 9641/10682 [1:22:00<08:34,  2.02it/s]
 90%|█████████ | 9642/10682 [1:22:00<08:34,  2.02it/s]
 90%|█████████ | 9643/10682 [1:22:01<08:33,  2.02it/s]
 90%|█████████ | 9644/10682 [1:22:01<08:33,  2.02it/s]
 90%|█████████ | 9645/10682 [1:22:02<08:32,  2.02it/s]
 90%|█████████ | 9646/10682 [1:22:02<08:32,  2.02it/s]
 90%|█████████ | 9647/10682 [1:22:03<08:31,  2.02it/s]
 90%|█████████ | 9648/10682 [1:22:03<08:31,  2.02it/s]
 90%|█████████ | 9649/10682 [1:22:04<08:30,  2.02it/s]
 90%|█████████ | 9650/10682 [1:22:04<08:29,  2.02it/s]{'loss': 2.7507, 'grad_norm': 0.2786267101764679, 'learning_rate': 2.8168304138088295e-05, 'epoch': 12.64}
                                                      
+
 90%|█████████ | 9650/10682 [1:22:04<08:29,  2.02it/s]
 90%|█████████ | 9651/10682 [1:22:05<08:29,  2.02it/s]
 90%|█████████ | 9652/10682 [1:22:05<08:29,  2.02it/s]
 90%|█████████ | 9653/10682 [1:22:06<08:29,  2.02it/s]
 90%|█████████ | 9654/10682 [1:22:06<08:28,  2.02it/s]
 90%|█████████ | 9655/10682 [1:22:07<08:28,  2.02it/s]
 90%|█████████ | 9656/10682 [1:22:07<08:27,  2.02it/s]
 90%|█████████ | 9657/10682 [1:22:08<08:27,  2.02it/s]
 90%|█████████ | 9658/10682 [1:22:08<08:26,  2.02it/s]
 90%|█████████ | 9659/10682 [1:22:09<08:26,  2.02it/s]
 90%|█████████ | 9660/10682 [1:22:09<08:25,  2.02it/s]
 90%|█████████ | 9661/10682 [1:22:10<08:25,  2.02it/s]
 90%|█████████ | 9662/10682 [1:22:10<08:24,  2.02it/s]
 90%|█████████ | 9663/10682 [1:22:11<08:24,  2.02it/s]
 90%|█████████ | 9664/10682 [1:22:11<08:23,  2.02it/s]
 90%|█████████ | 9665/10682 [1:22:12<08:23,  2.02it/s]
 90%|█████████ | 9666/10682 [1:22:12<08:23,  2.02it/s]
 90%|█████████ | 9667/10682 [1:22:13<08:22,  2.02it/s]
 91%|█████████ | 9668/10682 [1:22:13<08:21,  2.02it/s]
 91%|█████████ | 9669/10682 [1:22:14<08:20,  2.02it/s]
 91%|█████████ | 9670/10682 [1:22:14<08:19,  2.02it/s]
 91%|█████████ | 9671/10682 [1:22:15<08:19,  2.02it/s]
 91%|█████████ | 9672/10682 [1:22:15<08:18,  2.03it/s]
 91%|█████████ | 9673/10682 [1:22:16<08:18,  2.02it/s]
 91%|█████████ | 9674/10682 [1:22:16<08:17,  2.03it/s]
 91%|█████████ | 9675/10682 [1:22:17<08:17,  2.02it/s]
                                                      {'loss': 2.7574, 'grad_norm': 0.27770131826400757, 'learning_rate': 2.6832284358471516e-05, 'epoch': 12.67}
+
 91%|█████████ | 9675/10682 [1:22:17<08:17,  2.02it/s]
 91%|█████████ | 9676/10682 [1:22:17<08:16,  2.02it/s]
 91%|█████████ | 9677/10682 [1:22:18<08:16,  2.02it/s]
 91%|█████████ | 9678/10682 [1:22:18<08:16,  2.02it/s]
 91%|█████████ | 9679/10682 [1:22:19<08:15,  2.02it/s]
 91%|█████████ | 9680/10682 [1:22:19<08:15,  2.02it/s]
 91%|█████████ | 9681/10682 [1:22:20<08:14,  2.02it/s]
 91%|█████████ | 9682/10682 [1:22:20<08:14,  2.02it/s]
 91%|█████████ | 9683/10682 [1:22:21<08:13,  2.02it/s]
 91%|█████████ | 9684/10682 [1:22:21<08:13,  2.02it/s]
 91%|█████████ | 9685/10682 [1:22:22<08:12,  2.02it/s]
 91%|█████████ | 9686/10682 [1:22:22<08:12,  2.02it/s]
 91%|█████████ | 9687/10682 [1:22:23<08:11,  2.03it/s]
 91%|█████████ | 9688/10682 [1:22:23<08:11,  2.02it/s]
 91%|█████████ | 9689/10682 [1:22:24<08:10,  2.02it/s]
 91%|█████████ | 9690/10682 [1:22:24<08:10,  2.02it/s]
 91%|█████████ | 9691/10682 [1:22:25<08:09,  2.02it/s]
 91%|█████████ | 9692/10682 [1:22:25<08:09,  2.02it/s]
 91%|█████████ | 9693/10682 [1:22:26<08:08,  2.02it/s]
 91%|█████████ | 9694/10682 [1:22:26<08:08,  2.02it/s]
 91%|█████████ | 9695/10682 [1:22:27<08:08,  2.02it/s]
 91%|█████████ | 9696/10682 [1:22:27<08:07,  2.02it/s]
 91%|█████████ | 9697/10682 [1:22:28<08:06,  2.02it/s]
 91%|█████████ | 9698/10682 [1:22:28<08:06,  2.02it/s]
 91%|█████████ | 9699/10682 [1:22:29<08:05,  2.02it/s]
 91%|█████████ | 9700/10682 [1:22:29<08:05,  2.02it/s]
                                                      {'loss': 2.749, 'grad_norm': 0.27603113651275635, 'learning_rate': 2.5527849119658387e-05, 'epoch': 12.7}
+
 91%|█████████ | 9700/10682 [1:22:29<08:05,  2.02it/s]
 91%|█████████ | 9701/10682 [1:22:30<08:05,  2.02it/s]
 91%|█████████ | 9702/10682 [1:22:30<08:04,  2.02it/s]
 91%|█████████ | 9703/10682 [1:22:31<08:04,  2.02it/s]
 91%|█████████ | 9704/10682 [1:22:31<08:03,  2.02it/s]
 91%|█████████ | 9705/10682 [1:22:32<08:03,  2.02it/s]
 91%|█████████ | 9706/10682 [1:22:32<08:02,  2.02it/s]
 91%|█████████ | 9707/10682 [1:22:33<08:02,  2.02it/s]
 91%|█████████ | 9708/10682 [1:22:33<08:01,  2.02it/s]
 91%|█████████ | 9709/10682 [1:22:34<08:01,  2.02it/s]
 91%|███████���█ | 9710/10682 [1:22:34<08:00,  2.02it/s]
 91%|█████████ | 9711/10682 [1:22:35<08:00,  2.02it/s]
 91%|█████████ | 9712/10682 [1:22:35<07:59,  2.02it/s]
 91%|█████████ | 9713/10682 [1:22:36<07:58,  2.02it/s]
 91%|█████████ | 9714/10682 [1:22:36<07:58,  2.02it/s]
 91%|█████████ | 9715/10682 [1:22:37<07:57,  2.02it/s]
 91%|█████████ | 9716/10682 [1:22:37<07:57,  2.02it/s]
 91%|█████████ | 9717/10682 [1:22:38<07:56,  2.02it/s]
 91%|█████████ | 9718/10682 [1:22:38<07:56,  2.02it/s]
 91%|█████████ | 9719/10682 [1:22:39<07:55,  2.02it/s]
 91%|█████████ | 9720/10682 [1:22:39<07:54,  2.03it/s]
 91%|█████████ | 9721/10682 [1:22:40<07:54,  2.03it/s]
 91%|█████████ | 9722/10682 [1:22:40<07:54,  2.02it/s]
 91%|█████████ | 9723/10682 [1:22:41<07:53,  2.02it/s]
 91%|█████████ | 9724/10682 [1:22:41<07:53,  2.02it/s]
 91%|█████████ | 9725/10682 [1:22:42<07:52,  2.03it/s]
                                                      {'loss': 2.756, 'grad_norm': 0.27851659059524536, 'learning_rate': 2.4255085494343522e-05, 'epoch': 12.74}
+
 91%|█████████ | 9725/10682 [1:22:42<07:52,  2.03it/s]
 91%|█████████ | 9726/10682 [1:22:42<07:53,  2.02it/s]
 91%|█████████ | 9727/10682 [1:22:42<07:51,  2.02it/s]
 91%|█████████ | 9728/10682 [1:22:43<07:51,  2.02it/s]
 91%|█████████ | 9729/10682 [1:22:43<07:51,  2.02it/s]
 91%|█████████ | 9730/10682 [1:22:44<07:50,  2.02it/s]
 91%|█████████ | 9731/10682 [1:22:44<07:50,  2.02it/s]
 91%|█████████ | 9732/10682 [1:22:45<07:49,  2.02it/s]
 91%|█████████ | 9733/10682 [1:22:45<07:48,  2.02it/s]
 91%|█████████ | 9734/10682 [1:22:46<07:48,  2.02it/s]
 91%|█████████ | 9735/10682 [1:22:46<07:48,  2.02it/s]
 91%|█████████ | 9736/10682 [1:22:47<07:47,  2.02it/s]
 91%|█████████ | 9737/10682 [1:22:47<07:47,  2.02it/s]
 91%|█████████ | 9738/10682 [1:22:48<07:46,  2.02it/s]
 91%|█████████ | 9739/10682 [1:22:48<07:46,  2.02it/s]
 91%|█████████ | 9740/10682 [1:22:49<07:45,  2.02it/s]
 91%|█████████ | 9741/10682 [1:22:49<07:45,  2.02it/s]
 91%|█████████ | 9742/10682 [1:22:50<07:44,  2.02it/s]
 91%|█████████ | 9743/10682 [1:22:50<07:44,  2.02it/s]
 91%|█████████ | 9744/10682 [1:22:51<07:44,  2.02it/s]
 91%|█████████ | 9745/10682 [1:22:51<07:43,  2.02it/s]
 91%|█████████ | 9746/10682 [1:22:52<07:42,  2.02it/s]
 91%|█████████ | 9747/10682 [1:22:52<07:42,  2.02it/s]
 91%|█████████▏| 9748/10682 [1:22:53<07:42,  2.02it/s]
 91%|█████████▏| 9749/10682 [1:22:53<07:41,  2.02it/s]
 91%|█████████▏| 9750/10682 [1:22:54<07:40,  2.02it/s]
                                                      {'loss': 2.7534, 'grad_norm': 0.2752934694290161, 'learning_rate': 2.301407844110154e-05, 'epoch': 12.77}
+
 91%|█████████▏| 9750/10682 [1:22:54<07:40,  2.02it/s]
 91%|█████████▏| 9751/10682 [1:22:54<07:40,  2.02it/s]
 91%|█████████▏| 9752/10682 [1:22:55<07:39,  2.02it/s]
 91%|█████████▏| 9753/10682 [1:22:55<07:39,  2.02it/s]
 91%|█████████▏| 9754/10682 [1:22:56<07:39,  2.02it/s]
 91%|█████████▏| 9755/10682 [1:22:56<07:38,  2.02it/s]
 91%|█████████▏| 9756/10682 [1:22:57<07:38,  2.02it/s]
 91%|█████████▏| 9757/10682 [1:22:57<07:37,  2.02it/s]
 91%|█████████▏| 9758/10682 [1:22:58<07:37,  2.02it/s]
 91%|█████████▏| 9759/10682 [1:22:58<07:36,  2.02it/s]
 91%|█████████▏| 9760/10682 [1:22:59<07:36,  2.02it/s]
 91%|█████████▏| 9761/10682 [1:22:59<07:35,  2.02it/s]
 91%|█████████▏| 9762/10682 [1:23:00<07:34,  2.02it/s]
 91%|█████████▏| 9763/10682 [1:23:00<07:34,  2.02it/s]
 91%|█████████▏| 9764/10682 [1:23:01<07:33,  2.02it/s]
 91%|█████████▏| 9765/10682 [1:23:01<07:33,  2.02it/s]
 91%|█████████▏| 9766/10682 [1:23:02<07:32,  2.02it/s]
 91%|█████████▏| 9767/10682 [1:23:02<07:32,  2.02it/s]
 91%|█████████▏| 9768/10682 [1:23:03<07:32,  2.02it/s]
 91%|█████████▏| 9769/10682 [1:23:03<07:31,  2.02it/s]
 91%|█████████▏| 9770/10682 [1:23:04<07:31,  2.02it/s]
 91%|█████████▏| 9771/10682 [1:23:04<07:30,  2.02it/s]
 91%|█████████▏| 9772/10682 [1:23:05<07:30,  2.02it/s]
 91%|█████████▏| 9773/10682 [1:23:05<07:29,  2.02it/s]
 91%|█████████▏| 9774/10682 [1:23:06<07:28,  2.02it/s]
 92%|█████████▏| 9775/10682 [1:23:06<07:28,  2.02it/s]
                                                      {'loss': 2.7569, 'grad_norm': 0.27959272265434265, 'learning_rate': 2.1804910798715826e-05, 'epoch': 12.8}
+
 92%|█████████▏| 9775/10682 [1:23:06<07:28,  2.02it/s]
 92%|█████████▏| 9776/10682 [1:23:07<07:28,  2.02it/s]
 92%|█████████▏| 9777/10682 [1:23:07<07:27,  2.02it/s]
 92%|█████████▏| 9778/10682 [1:23:08<07:27,  2.02it/s]
 92%|█████████▏| 9779/10682 [1:23:08<07:26,  2.02it/s]
 92%|█████████▏| 9780/10682 [1:23:09<07:25,  2.02it/s]
 92%|█████████▏| 9781/10682 [1:23:09<07:25,  2.02it/s]
 92%|█████████▏| 9782/10682 [1:23:10<07:25,  2.02it/s]
 92%|█████████▏| 9783/10682 [1:23:10<07:24,  2.02it/s]
 92%|█████████▏| 9784/10682 [1:23:11<07:24,  2.02it/s]
 92%|█████████▏| 9785/10682 [1:23:11<07:23,  2.02it/s]
 92%|█████████▏| 9786/10682 [1:23:12<07:23,  2.02it/s]
 92%|█████████▏| 9787/10682 [1:23:12<07:22,  2.02it/s]
 92%|█████████▏| 9788/10682 [1:23:13<07:21,  2.02it/s]
 92%|█████████▏| 9789/10682 [1:23:13<07:21,  2.02it/s]
 92%|█████████▏| 9790/10682 [1:23:14<07:20,  2.03it/s]
 92%|█████████▏| 9791/10682 [1:23:14<07:20,  2.02it/s]
 92%|█████████▏| 9792/10682 [1:23:15<07:19,  2.02it/s]
 92%|█████████▏| 9793/10682 [1:23:15<07:19,  2.02it/s]
 92%|█████████▏| 9794/10682 [1:23:16<07:18,  2.02it/s]
 92%|█████████▏| 9795/10682 [1:23:16<07:17,  2.03it/s]
 92%|█████████▏| 9796/10682 [1:23:17<07:17,  2.02it/s]
 92%|█████████▏| 9797/10682 [1:23:17<07:16,  2.03it/s]
 92%|█████████▏| 9798/10682 [1:23:18<07:16,  2.03it/s]
 92%|█████████▏| 9799/10682 [1:23:18<07:15,  2.03it/s]
 92%|█████████▏| 9800/10682 [1:23:19<07:15,  2.02it/s]
                                                      
+
 92%|█████████▏| 9800/10682 [1:23:19<07:15,  2.02it/s]
 92%|█████████▏| 9801/10682 [1:23:19<07:17,  2.01it/s]
 92%|█████████▏| 9802/10682 [1:23:20<07:16,  2.02it/s]
 92%|█████████▏| 9803/10682 [1:23:20<07:15,  2.02it/s]
 92%|█████████▏| 9804/10682 [1:23:21<07:14,  2.02it/s]
 92%|█████████▏| 9805/10682 [1:23:21<07:14,  2.02it/s]
 92%|█████████▏| 9806/10682 [1:23:22<07:13,  2.02it/s]
 92%|█████████▏| 9807/10682 [1:23:22<07:12,  2.02it/s]
 92%|█████████▏| 9808/10682 [1:23:23<07:12,  2.02it/s]
 92%|█████████▏| 9809/10682 [1:23:23<07:11,  2.02it/s]
 92%|█████████▏| 9810/10682 [1:23:24<07:11,  2.02it/s]
 92%|█████████▏| 9811/10682 [1:23:24<07:11,  2.02it/s]
 92%|█████████▏| 9812/10682 [1:23:25<07:10,  2.02it/s]
 92%|█████████▏| 9813/10682 [1:23:25<07:09,  2.02it/s]
 92%|█████████▏| 9814/10682 [1:23:26<07:09,  2.02it/s]
 92%|█████████▏| 9815/10682 [1:23:26<07:08,  2.02it/s]
 92%|█████████▏| 9816/10682 [1:23:27<07:08,  2.02it/s]
 92%|█████████▏| 9817/10682 [1:23:27<07:07,  2.02it/s]
 92%|█████████▏| 9818/10682 [1:23:28<07:07,  2.02it/s]
 92%|█████████▏| 9819/10682 [1:23:28<07:06,  2.02it/s]
 92%|█████████▏| 9820/10682 [1:23:28<07:06,  2.02it/s]
 92%|█████████▏| 9821/10682 [1:23:29<07:05,  2.02it/s]
 92%|█████████▏| 9822/10682 [1:23:29<07:05,  2.02it/s]
 92%|█████████▏| 9823/10682 [1:23:30<07:04,  2.02it/s]
 92%|█████████▏| 9824/10682 [1:23:30<07:04,  2.02it/s]
 92%|█████████▏| 9825/10682 [1:23:31<07:03,  2.02it/s]
                                                      
+
 92%|█████████▏| 9825/10682 [1:23:31<07:03,  2.02it/s]
 92%|█████████▏| 9826/10682 [1:23:31<07:03,  2.02it/s]
 92%|█████████▏| 9827/10682 [1:23:32<07:02,  2.02it/s]
 92%|█████████▏| 9828/10682 [1:23:32<07:02,  2.02it/s]
 92%|█████████▏| 9829/10682 [1:23:33<07:01,  2.02it/s]
 92%|█████████▏| 9830/10682 [1:23:33<07:01,  2.02it/s]
 92%|█████████▏| 9831/10682 [1:23:34<07:00,  2.02it/s]
 92%|█████████▏| 9832/10682 [1:23:34<06:59,  2.02it/s]
 92%|█████████▏| 9833/10682 [1:23:35<06:59,  2.02it/s]
 92%|█████████▏| 9834/10682 [1:23:35<06:58,  2.03it/s]
 92%|█████████▏| 9835/10682 [1:23:36<06:58,  2.02it/s]
 92%|█████████▏| 9836/10682 [1:23:36<06:57,  2.02it/s]
 92%|█████████▏| 9837/10682 [1:23:37<06:57,  2.02it/s]
 92%|█████████▏| 9838/10682 [1:23:37<06:57,  2.02it/s]
 92%|█████████▏| 9839/10682 [1:23:38<06:56,  2.02it/s]
 92%|█████████▏| 9840/10682 [1:23:38<06:56,  2.02it/s]
 92%|█████████▏| 9841/10682 [1:23:39<06:55,  2.02it/s]
 92%|█████████▏| 9842/10682 [1:23:39<06:55,  2.02it/s]
 92%|█████████▏| 9843/10682 [1:23:40<06:54,  2.02it/s]
 92%|█████████▏| 9844/10682 [1:23:40<06:54,  2.02it/s]
 92%|█████████▏| 9845/10682 [1:23:41<06:54,  2.02it/s]
 92%|█████████▏| 9846/10682 [1:23:41<06:54,  2.02it/s]
 92%|█████████▏| 9847/10682 [1:23:42<06:53,  2.02it/s]
 92%|█████████▏| 9848/10682 [1:23:42<06:52,  2.02it/s]
 92%|█████████▏| 9849/10682 [1:23:43<06:52,  2.02it/s]
 92%|█████████▏| 9850/10682 [1:23:43<06:51,  2.02it/s]{'loss': 2.7634, 'grad_norm': 0.27715760469436646, 'learning_rate': 1.8369240812535104e-05, 'epoch': 12.9}
                                                      
+
 92%|█████████▏| 9850/10682 [1:23:43<06:51,  2.02it/s]
 92%|█████████▏| 9851/10682 [1:23:44<06:51,  2.02it/s]
 92%|█████████▏| 9852/10682 [1:23:44<06:51,  2.02it/s]
 92%|█████████▏| 9853/10682 [1:23:45<06:50,  2.02it/s]
 92%|█████████▏| 9854/10682 [1:23:45<06:50,  2.02it/s]
 92%|█████████▏| 9855/10682 [1:23:46<06:49,  2.02it/s]
 92%|█████████▏| 9856/10682 [1:23:46<06:49,  2.02it/s]
 92%|█████████▏| 9857/10682 [1:23:47<06:47,  2.02it/s]
 92%|█████████▏| 9858/10682 [1:23:47<06:47,  2.02it/s]
 92%|█████████▏| 9859/10682 [1:23:48<06:46,  2.02it/s]
 92%|█████████▏| 9860/10682 [1:23:48<06:46,  2.02it/s]
 92%|█████████▏| 9861/10682 [1:23:49<06:45,  2.02it/s]
 92%|█████████▏| 9862/10682 [1:23:49<06:45,  2.02it/s]
 92%|█████████▏| 9863/10682 [1:23:50<06:44,  2.02it/s]
 92%|█████████▏| 9864/10682 [1:23:50<06:44,  2.02it/s]
 92%|█████████▏| 9865/10682 [1:23:51<06:43,  2.02it/s]
 92%|█████████▏| 9866/10682 [1:23:51<06:43,  2.02it/s]
 92%|█████████▏| 9867/10682 [1:23:52<06:43,  2.02it/s]
 92%|█████████▏| 9868/10682 [1:23:52<06:42,  2.02it/s]
 92%|█████████▏| 9869/10682 [1:23:53<06:42,  2.02it/s]
 92%|█████████▏| 9870/10682 [1:23:53<06:41,  2.02it/s]
 92%|█████████▏| 9871/10682 [1:23:54<06:40,  2.02it/s]
 92%|█████████▏| 9872/10682 [1:23:54<06:40,  2.02it/s]
 92%|█████████▏| 9873/10682 [1:23:55<06:39,  2.02it/s]
 92%|█████████▏| 9874/10682 [1:23:55<06:39,  2.02it/s]
 92%|█████████▏| 9875/10682 [1:23:56<06:38,  2.02it/s]
                                                      
+
 92%|█████████▏| 9875/10682 [1:23:56<06:38,  2.02it/s]
 92%|█████████▏| 9876/10682 [1:23:56<06:38,  2.02it/s]
 92%|█████████▏| 9877/10682 [1:23:57<06:38,  2.02it/s]
 92%|█████████▏| 9878/10682 [1:23:57<06:38,  2.02it/s]
 92%|█████████▏| 9879/10682 [1:23:58<06:37,  2.02it/s]
 92%|█████████▏| 9880/10682 [1:23:58<06:36,  2.02it/s]
 93%|█████████▎| 9881/10682 [1:23:59<06:35,  2.02it/s]
 93%|█████████▎| 9882/10682 [1:23:59<06:35,  2.02it/s]
 93%|█████████▎| 9883/10682 [1:24:00<06:35,  2.02it/s]
 93%|█████████▎| 9884/10682 [1:24:00<06:34,  2.02it/s]
 93%|█████████▎| 9885/10682 [1:24:01<06:34,  2.02it/s]
 93%|█████████▎| 9886/10682 [1:24:01<06:33,  2.02it/s]
 93%|█████████▎| 9887/10682 [1:24:02<06:32,  2.02it/s]
 93%|█████████▎| 9888/10682 [1:24:02<06:32,  2.02it/s]
 93%|█████████▎| 9889/10682 [1:24:03<06:32,  2.02it/s]
 93%|█████████▎| 9890/10682 [1:24:03<06:31,  2.02it/s]
 93%|█████████▎| 9891/10682 [1:24:04<06:31,  2.02it/s]
 93%|█████████▎| 9892/10682 [1:24:04<06:30,  2.02it/s]
 93%|█████████▎| 9893/10682 [1:24:05<06:29,  2.02it/s]
 93%|█████████▎| 9894/10682 [1:24:05<06:29,  2.02it/s]
 93%|█████████▎| 9895/10682 [1:24:06<06:28,  2.03it/s]
 93%|█████████▎| 9896/10682 [1:24:06<06:28,  2.02it/s]
 93%|█████████▎| 9897/10682 [1:24:07<06:27,  2.02it/s]
 93%|█████████▎| 9898/10682 [1:24:07<06:27,  2.02it/s]
 93%|█████████▎| 9899/10682 [1:24:08<06:26,  2.02it/s]
 93%|█████████▎| 9900/10682 [1:24:08<06:26,  2.02it/s]{'loss': 2.7483, 'grad_norm': 0.2770647406578064, 'learning_rate': 1.6239414036870183e-05, 'epoch': 12.97}
                                                      
+
 93%|█████████▎| 9900/10682 [1:24:08<06:26,  2.02it/s]
 93%|█████████▎| 9901/10682 [1:24:09<06:27,  2.02it/s]
 93%|█████████▎| 9902/10682 [1:24:09<06:26,  2.02it/s]
 93%|█████████▎| 9903/10682 [1:24:10<06:25,  2.02it/s]
 93%|█████████▎| 9904/10682 [1:24:10<06:25,  2.02it/s]
 93%|█████████▎| 9905/10682 [1:24:11<06:24,  2.02it/s]
 93%|█████████▎| 9906/10682 [1:24:11<06:23,  2.02it/s]
 93%|█████████▎| 9907/10682 [1:24:12<06:23,  2.02it/s]
 93%|█████████▎| 9908/10682 [1:24:12<06:23,  2.02it/s]
 93%|█████████▎| 9909/10682 [1:24:13<06:22,  2.02it/s]
 93%|█████████▎| 9910/10682 [1:24:13<06:21,  2.02it/s]
 93%|█████████▎| 9911/10682 [1:24:13<06:21,  2.02it/s]
 93%|█████████▎| 9912/10682 [1:24:14<06:20,  2.02it/s]
 93%|█████████▎| 9913/10682 [1:24:14<06:20,  2.02it/s]
 93%|█████████▎| 9914/10682 [1:24:15<06:19,  2.02it/s]
 93%|█████████▎| 9915/10682 [1:24:15<06:19,  2.02it/s]
 93%|█████████▎| 9916/10682 [1:24:16<06:18,  2.02it/s]
 93%|█████████▎| 9917/10682 [1:24:16<06:18,  2.02it/s]
 93%|█████████▎| 9918/10682 [1:24:17<06:17,  2.02it/s]
 93%|█████████▎| 9919/10682 [1:24:17<06:17,  2.02it/s]
 93%|█████████▎| 9920/10682 [1:24:18<06:16,  2.02it/s]
 93%|█████████▎| 9921/10682 [1:24:18<06:16,  2.02it/s]
 93%|█████████▎| 9922/10682 [1:24:19<06:15,  2.02it/s]
 93%|█████████▎| 9923/10682 [1:24:19<06:15,  2.02it/s]
 93%|█████████▎| 9924/10682 [1:24:20<06:15,  2.02it/s]
 93%|█████████▎| 9925/10682 [1:24:20<06:23,  1.98it/s]
                                                      
+
 93%|█████████▎| 9925/10682 [1:24:20<06:23,  1.98it/s]
 93%|█████████▎| 9926/10682 [1:24:32<49:46,  3.95s/it]
 93%|█████████▎| 9927/10682 [1:24:33<36:40,  2.91s/it]
 93%|█████████▎| 9928/10682 [1:24:33<27:31,  2.19s/it]
 93%|█████████▎| 9929/10682 [1:24:34<21:07,  1.68s/it]
 93%|█████████▎| 9930/10682 [1:24:34<16:38,  1.33s/it]
 93%|█████████▎| 9931/10682 [1:24:35<13:29,  1.08s/it]
 93%|█████████▎| 9932/10682 [1:24:35<11:17,  1.11it/s]
 93%|█████████▎| 9933/10682 [1:24:36<09:47,  1.28it/s]
 93%|█████████▎| 9934/10682 [1:24:36<08:42,  1.43it/s]
 93%|█████████▎| 9935/10682 [1:24:37<07:55,  1.57it/s]
 93%|█████████▎| 9936/10682 [1:24:37<07:22,  1.68it/s]
 93%|█████████▎| 9937/10682 [1:24:38<06:59,  1.78it/s]
 93%|█████████▎| 9938/10682 [1:24:38<06:43,  1.84it/s]
 93%|█████████▎| 9939/10682 [1:24:39<06:32,  1.89it/s]
 93%|█████████▎| 9940/10682 [1:24:39<06:24,  1.93it/s]
 93%|█████████▎| 9941/10682 [1:24:40<06:19,  1.95it/s]
 93%|█████████▎| 9942/10682 [1:24:40<06:15,  1.97it/s]
 93%|█████████▎| 9943/10682 [1:24:41<06:12,  1.99it/s]
 93%|█████████▎| 9944/10682 [1:24:41<06:10,  1.99it/s]
 93%|█████████▎| 9945/10682 [1:24:42<06:08,  2.00it/s]
 93%|█████████▎| 9946/10682 [1:24:42<06:06,  2.01it/s]
 93%|█████████▎| 9947/10682 [1:24:43<06:05,  2.01it/s]
 93%|█████████▎| 9948/10682 [1:24:43<06:04,  2.01it/s]
 93%|█████████▎| 9949/10682 [1:24:44<06:04,  2.01it/s]