Training in progress, epoch 12

logs/events.out.tfevents.1715485378.sphinx2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3f3a58335fd5eb725030fc039f7e11e3127a98f54da6c218a392bfbb19039e11
-size 75300
+oid sha256:8f6eda382822333d3e0fb58eda7d6c08443a03c9086314112a65e1983ec9de9d
+size 81841

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:82c68f8f9738a0043334b029e1542f22c45c7c608a6c6e59e97e2f67e7892f7a
+oid sha256:93f05a8c0ba3bedaf26af6077768cd7c73e13a720bc5c026755d8230719bee8d
 size 281715176

train_job_output.txt

@@ -557,4 +557,35 @@ command outputs:
 
 78%|███████▊  | 8325/10682 [1:10:33<19:19,  2.03it/s]
 78%|███████▊  | 8326/10682 [1:10:33<19:21,  2.03it/s]
 78%|███████▊  | 8327/10682 [1:10:34<19:21,  2.03it/s]
 78%|███████▊  | 8328/10682 [1:10:34<19:20,  2.03it/s]
 78%|███████▊  | 8329/10682 [1:10:35<19:18,  2.03it/s]
 78%|███████▊  | 8330/10682 [1:10:35<19:17,  2.03it/s]
 78%|███████▊  | 8331/10682 [1:10:36<19:18,  2.03it/s]
 78%|███████▊  | 8332/10682 [1:10:36<19:16,  2.03it/s]
 78%|███████▊  | 8333/10682 [1:10:37<19:16,  2.03it/s]
 78%|███████▊  | 8334/10682 [1:10:37<19:16,  2.03it/s]
 78%|███████▊  | 8335/10682 [1:10:38<19:15,  2.03it/s]
 78%|███████▊  | 8336/10682 [1:10:38<19:15,  2.03it/s]
 78%|███████▊  | 8337/10682 [1:10:39<19:13,  2.03it/s]
 78%|███████▊  | 8338/10682 [1:10:39<19:13,  2.03it/s]
 78%|███████▊  | 8339/10682 [1:10:40<19:13,  2.03it/s]
 78%|███████▊  | 8340/10682 [1:10:40<19:12,  2.03it/s]
 78%|███████▊  | 8341/10682 [1:10:41<19:11,  2.03it/s]
 78%|███████▊  | 8342/10682 [1:10:41<19:12,  2.03it/s]
 78%|███████▊  | 8343/10682 [1:10:42<19:13,  2.03it/s]
 78%|███████▊  | 8344/10682 [1:10:42<19:12,  2.03it/s]
 78%|███████▊  | 8345/10682 [1:10:43<19:10,  2.03it/s]
 78%|███████▊  | 8346/10682 [1:10:43<19:09,  2.03it/s]
 78%|███████▊  | 8347/10682 [1:10:44<19:10,  2.03it/s]
 78%|███████▊  | 8348/10682 [1:10:44<19:09,  2.03it/s]
 78%|███████▊  | 8349/10682 [1:10:45<19:07,  2.03it/s]
 78%|███████▊  | 8350/10682 [1:10:45<19:07,  2.03it/s]{'loss': 2.9119, 'grad_norm': 0.256185919046402, 'learning_rate': 0.00013831071326327282, 'epoch': 10.94}
                                                      
 
 78%|███████▊  | 8350/10682 [1:10:45<19:07,  2.03it/s]
 78%|███████▊  | 8351/10682 [1:10:46<19:08,  2.03it/s]
 78%|███████▊  | 8352/10682 [1:10:46<19:09,  2.03it/s]
 78%|███████▊  | 8353/10682 [1:10:47<19:07,  2.03it/s]
 78%|███████▊  | 8354/10682 [1:10:47<19:08,  2.03it/s]
 78%|███████▊  | 8355/10682 [1:10:48<19:07,  2.03it/s]
 78%|███████▊  | 8356/10682 [1:10:48<19:06,  2.03it/s]
 78%|███████▊  | 8357/10682 [1:10:49<19:06,  2.03it/s]
 78%|███████▊  | 8358/10682 [1:10:49<19:04,  2.03it/s]
 78%|███████▊  | 8359/10682 [1:10:50<19:04,  2.03it/s]
 78%|███████▊  | 8360/10682 [1:10:50<19:03,  2.03it/s]
 78%|███████▊  | 8361/10682 [1:10:51<19:02,  2.03it/s]
 78%|███████▊  | 8362/10682 [1:10:51<19:03,  2.03it/s]
 78%|███████▊  | 8363/10682 [1:10:52<19:01,  2.03it/s]
 78%|███████▊  | 8364/10682 [1:10:52<19:00,  2.03it/s]
 78%|███████▊  | 8365/10682 [1:10:53<19:00,  2.03it/s]
 78%|███████▊  | 8366/10682 [1:10:53<18:59,  2.03it/s]
 78%|███████▊  | 8367/10682 [1:10:54<18:58,  2.03it/s]
 78%|███████▊  | 8368/10682 [1:10:54<18:58,  2.03it/s]
 78%|███████▊  | 8369/10682 [1:10:55<18:57,  2.03it/s]
 78%|███████▊  | 8370/10682 [1:10:55<18:59,  2.03it/s]
 78%|███████▊  | 8371/10682 [1:10:56<18:57,  2.03it/s]
 78%|███████▊  | 8372/10682 [1:10:56<18:58,  2.03it/s]
 78%|███████▊  | 8373/10682 [1:10:57<18:57,  2.03it/s]
 78%|███████▊  | 8374/10682 [1:10:57<18:56,  2.03it/s]
 78%|███████▊  | 8375/10682 [1:10:58<18:57,  2.03it/s]{'loss': 2.9219, 'grad_norm': 0.2577957808971405, 'learning_rate': 0.0001355022636953933, 'epoch': 10.98}
                                                      
 
 78%|███████▊  | 8375/10682 [1:10:58<18:57,  2.03it/s]
 78%|███████▊  | 8376/10682 [1:10:58<18:57,  2.03it/s]
 78%|███████▊  | 8377/10682 [1:10:59<18:58,  2.03it/s]
 78%|███████▊  | 8378/10682 [1:10:59<18:57,  2.03it/s]
 78%|███████▊  | 8379/10682 [1:11:00<18:56,  2.03it/s]
 78%|███████▊  | 8380/10682 [1:11:00<18:55,  2.03it/s]
 78%|███████▊  | 8381/10682 [1:11:01<18:54,  2.03it/s]
 78%|███████▊  | 8382/10682 [1:11:01<18:54,  2.03it/s]
 78%|███████▊  | 8383/10682 [1:11:02<18:53,  2.03it/s]
 78%|███████▊  | 8384/10682 [1:11:02<18:52,  2.03it/s]
 78%|███████▊  | 8385/10682 [1:11:03<18:51,  2.03it/s]
 79%|███████▊  | 8386/10682 [1:11:03<18:51,  2.03it/s]
 79%|███████▊  | 8387/10682 [1:11:04<18:51,  2.03it/s]
 79%|███████▊  | 8388/10682 [1:11:04<18:49,  2.03it/s]
 79%|███████▊  | 8389/10682 [1:11:05<18:48,  2.03it/s]
 79%|███████▊  | 8390/10682 [1:11:05<18:48,  2.03it/s]
 79%|███████▊  | 8391/10682 [1:11:06<18:47,  2.03it/s]
 79%|███████▊  | 8392/10682 [1:11:06<18:49,  2.03it/s]
 79%|███████▊  | 8393/10682 [1:11:06<18:36,  2.05it/s]
 79%|███████▊  | 8394/10682 [1:11:21<3:01:51,  4.77s/it]
 79%|███████▊  | 8395/10682 [1:11:22<2:12:51,  3.49s/it]
 79%|███████▊  | 8396/10682 [1:11:22<1:38:43,  2.59s/it]
 79%|███████▊  | 8397/10682 [1:11:23<1:14:42,  1.96s/it]
 79%|███████▊  | 8398/10682 [1:11:23<57:59,  1.52s/it]  
 79%|███████▊  | 8399/10682 [1:11:24<46:11,  1.21s/it]
 79%|███████▊  | 8400/10682 [1:11:24<37:57,  1.00it/s]
                                                      
-
 79%|███████▊  | 8400/10682 [1:11:24<37:57,  1.00it/s]
 79%|███████▊  | 8401/10682 [1:11:25<32:22,  1.17it/s]
 79%|███████▊  | 8402/10682 [1:11:25<28:16,  1.34it/s]
 79%|███████▊  | 8403/10682 [1:11:26<25:23,  1.50it/s]
 79%|███████▊  | 8404/10682 [1:11:26<23:24,  1.62it/s]
 79%|███████▊  | 8405/10682 [1:11:27<21:59,  1.73it/s]
 79%|███████▊  | 8406/10682 [1:11:27<20:58,  1.81it/s]
 79%|███████▊  | 8407/10682 [1:11:28<20:16,  1.87it/s]
 79%|███████▊  | 8408/10682 [1:11:28<19:46,  1.92it/s]
 79%|███████▊  | 8409/10682 [1:11:29<19:26,  1.95it/s]
 79%|███████▊  | 8410/10682 [1:11:29<19:11,  1.97it/s]
 79%|███████▊  | 8411/10682 [1:11:30<19:04,  1.98it/s]
 79%|███████▊  | 8412/10682 [1:11:30<18:54,  2.00it/s]
 79%|███████▉  | 8413/10682 [1:11:31<18:48,  2.01it/s]
 79%|███████▉  | 8414/10682 [1:11:31<18:46,  2.01it/s]
 79%|███████▉  | 8415/10682 [1:11:32<18:43,  2.02it/s]
 79%|███████▉  | 8416/10682 [1:11:32<18:41,  2.02it/s]
 79%|███████▉  | 8417/10682 [1:11:33<18:38,  2.02it/s]
 79%|███████▉  | 8418/10682 [1:11:33<18:38,  2.02it/s]
+
 79%|███████▊  | 8400/10682 [1:11:24<37:57,  1.00it/s]
 79%|███████▊  | 8401/10682 [1:11:25<32:22,  1.17it/s]
 79%|███████▊  | 8402/10682 [1:11:25<28:16,  1.34it/s]
 79%|███████▊  | 8403/10682 [1:11:26<25:23,  1.50it/s]
 79%|███████▊  | 8404/10682 [1:11:26<23:24,  1.62it/s]
 79%|███████▊  | 8405/10682 [1:11:27<21:59,  1.73it/s]
 79%|███████▊  | 8406/10682 [1:11:27<20:58,  1.81it/s]
 79%|███████▊  | 8407/10682 [1:11:28<20:16,  1.87it/s]
 79%|███████▊  | 8408/10682 [1:11:28<19:46,  1.92it/s]
 79%|███████▊  | 8409/10682 [1:11:29<19:26,  1.95it/s]
 79%|███████▊  | 8410/10682 [1:11:29<19:11,  1.97it/s]
 79%|███████▊  | 8411/10682 [1:11:30<19:04,  1.98it/s]
 79%|███████▊  | 8412/10682 [1:11:30<18:54,  2.00it/s]
 79%|███████▉  | 8413/10682 [1:11:31<18:48,  2.01it/s]
 79%|███████▉  | 8414/10682 [1:11:31<18:46,  2.01it/s]
 79%|███████▉  | 8415/10682 [1:11:32<18:43,  2.02it/s]
 79%|███████▉  | 8416/10682 [1:11:32<18:41,  2.02it/s]
 79%|███████▉  | 8417/10682 [1:11:33<18:38,  2.02it/s]
 79%|███████▉  | 8418/10682 [1:11:33<18:38,  2.02it/s]
 79%|███████▉  | 8419/10682 [1:11:34<18:36,  2.03it/s]
 79%|███████▉  | 8420/10682 [1:11:34<18:36,  2.03it/s]
 79%|███████▉  | 8421/10682 [1:11:35<18:35,  2.03it/s]
 79%|███████▉  | 8422/10682 [1:11:35<18:33,  2.03it/s]
 79%|███████▉  | 8423/10682 [1:11:36<18:33,  2.03it/s]
 79%|███████▉  | 8424/10682 [1:11:36<18:32,  2.03it/s]
 79%|███████▉  | 8425/10682 [1:11:37<18:32,  2.03it/s]
                                                      
+
 79%|███████▉  | 8425/10682 [1:11:37<18:32,  2.03it/s]
 79%|███████▉  | 8426/10682 [1:11:37<18:32,  2.03it/s]
 79%|███████▉  | 8427/10682 [1:11:38<18:30,  2.03it/s]
 79%|███████▉  | 8428/10682 [1:11:38<18:30,  2.03it/s]
 79%|███████▉  | 8429/10682 [1:11:39<18:29,  2.03it/s]
 79%|███████▉  | 8430/10682 [1:11:39<18:30,  2.03it/s]
 79%|███████▉  | 8431/10682 [1:11:40<18:29,  2.03it/s]
 79%|███████▉  | 8432/10682 [1:11:40<18:29,  2.03it/s]
 79%|███████▉  | 8433/10682 [1:11:40<18:28,  2.03it/s]
 79%|███████▉  | 8434/10682 [1:11:41<18:27,  2.03it/s]
 79%|███████▉  | 8435/10682 [1:11:41<18:28,  2.03it/s]
 79%|███████▉  | 8436/10682 [1:11:42<18:27,  2.03it/s]
 79%|███████▉  | 8437/10682 [1:11:42<18:27,  2.03it/s]
 79%|███████▉  | 8438/10682 [1:11:43<18:26,  2.03it/s]
 79%|███████▉  | 8439/10682 [1:11:43<18:23,  2.03it/s]
 79%|███████▉  | 8440/10682 [1:11:44<18:24,  2.03it/s]
 79%|███████▉  | 8441/10682 [1:11:44<18:23,  2.03it/s]
 79%|███████▉  | 8442/10682 [1:11:45<18:21,  2.03it/s]
 79%|███████▉  | 8443/10682 [1:11:45<18:22,  2.03it/s]
 79%|███████▉  | 8444/10682 [1:11:46<18:20,  2.03it/s]
 79%|███████▉  | 8445/10682 [1:11:46<18:20,  2.03it/s]
 79%|███████▉  | 8446/10682 [1:11:47<18:20,  2.03it/s]
 79%|███████▉  | 8447/10682 [1:11:47<18:19,  2.03it/s]
 79%|███████▉  | 8448/10682 [1:11:48<18:21,  2.03it/s]
 79%|███████▉  | 8449/10682 [1:11:48<18:21,  2.03it/s]
 79%|███████▉  | 8450/10682 [1:11:49<18:20,  2.03it/s]
                                                      
+
 79%|███████▉  | 8450/10682 [1:11:49<18:20,  2.03it/s]
 79%|███████▉  | 8451/10682 [1:11:49<18:21,  2.03it/s]
 79%|███████▉  | 8452/10682 [1:11:50<18:20,  2.03it/s]
 79%|███████▉  | 8453/10682 [1:11:50<18:19,  2.03it/s]
 79%|███████▉  | 8454/10682 [1:11:51<18:18,  2.03it/s]
 79%|███████▉  | 8455/10682 [1:11:51<18:17,  2.03it/s]
 79%|███████▉  | 8456/10682 [1:11:52<18:17,  2.03it/s]
 79%|███████▉  | 8457/10682 [1:11:52<18:15,  2.03it/s]
 79%|███████▉  | 8458/10682 [1:11:53<18:15,  2.03it/s]
 79%|███████▉  | 8459/10682 [1:11:53<18:15,  2.03it/s]
 79%|███████▉  | 8460/10682 [1:11:54<18:15,  2.03it/s]
 79%|███████▉  | 8461/10682 [1:11:54<18:14,  2.03it/s]
 79%|███████▉  | 8462/10682 [1:11:55<18:13,  2.03it/s]
 79%|███████▉  | 8463/10682 [1:11:55<18:13,  2.03it/s]
 79%|███████▉  | 8464/10682 [1:11:56<18:12,  2.03it/s]
 79%|███████▉  | 8465/10682 [1:11:56<18:11,  2.03it/s]
 79%|███████▉  | 8466/10682 [1:11:57<18:12,  2.03it/s]
 79%|███████▉  | 8467/10682 [1:11:57<18:11,  2.03it/s]
 79%|███████▉  | 8468/10682 [1:11:58<18:10,  2.03it/s]
 79%|███████▉  | 8469/10682 [1:11:58<18:10,  2.03it/s]
 79%|███████▉  | 8470/10682 [1:11:59<18:08,  2.03it/s]
 79%|███████▉  | 8471/10682 [1:11:59<18:08,  2.03it/s]
 79%|███████▉  | 8472/10682 [1:12:00<18:06,  2.03it/s]
 79%|███████▉  | 8473/10682 [1:12:00<18:06,  2.03it/s]
 79%|███████▉  | 8474/10682 [1:12:01<18:06,  2.03it/s]
 79%|███████▉  | 8475/10682 [1:12:01<18:05,  2.03it/s]
                                                      {'loss': 2.819, 'grad_norm': 0.2613165080547333, 'learning_rate': 0.00012451362249196797, 'epoch': 11.11}
+
 79%|███████▉  | 8475/10682 [1:12:01<18:05,  2.03it/s]
 79%|███████▉  | 8476/10682 [1:12:02<18:06,  2.03it/s]
 79%|███████▉  | 8477/10682 [1:12:02<18:05,  2.03it/s]
 79%|███████▉  | 8478/10682 [1:12:03<18:05,  2.03it/s]
 79%|███████▉  | 8479/10682 [1:12:03<18:05,  2.03it/s]
 79%|███████▉  | 8480/10682 [1:12:04<18:04,  2.03it/s]
 79%|███████▉  | 8481/10682 [1:12:04<18:04,  2.03it/s]
 79%|███████▉  | 8482/10682 [1:12:05<18:02,  2.03it/s]
 79%|███████▉  | 8483/10682 [1:12:05<18:02,  2.03it/s]
 79%|███████▉  | 8484/10682 [1:12:06<18:01,  2.03it/s]
 79%|███████▉  | 8485/10682 [1:12:06<18:00,  2.03it/s]
 79%|███████▉  | 8486/10682 [1:12:07<18:00,  2.03it/s]
 79%|███████▉  | 8487/10682 [1:12:07<18:00,  2.03it/s]
 79%|███████▉  | 8488/10682 [1:12:08<18:00,  2.03it/s]
 79%|███████▉  | 8489/10682 [1:12:08<18:00,  2.03it/s]
 79%|███████▉  | 8490/10682 [1:12:09<17:59,  2.03it/s]
 79%|███████▉  | 8491/10682 [1:12:09<18:00,  2.03it/s]
 79%|███████▉  | 8492/10682 [1:12:10<17:59,  2.03it/s]
 80%|███████▉  | 8493/10682 [1:12:10<17:59,  2.03it/s]
 80%|███████▉  | 8494/10682 [1:12:11<17:57,  2.03it/s]
 80%|███████▉  | 8495/10682 [1:12:11<17:57,  2.03it/s]
 80%|███████▉  | 8496/10682 [1:12:12<17:56,  2.03it/s]
 80%|███████▉  | 8497/10682 [1:12:12<17:56,  2.03it/s]
 80%|███████▉  | 8498/10682 [1:12:13<17:55,  2.03it/s]
 80%|███████▉  | 8499/10682 [1:12:13<17:55,  2.03it/s]
 80%|███████▉  | 8500/10682 [1:12:13<17:53,  2.03it/s]
                                                      
+
 80%|███████▉  | 8500/10682 [1:12:13<17:53,  2.03it/s]
 80%|███████▉  | 8501/10682 [1:12:14<17:56,  2.03it/s]
 80%|███████▉  | 8502/10682 [1:12:14<17:54,  2.03it/s]
 80%|███████▉  | 8503/10682 [1:12:15<17:53,  2.03it/s]
 80%|███████▉  | 8504/10682 [1:12:15<17:54,  2.03it/s]
 80%|███████▉  | 8505/10682 [1:12:16<17:52,  2.03it/s]
 80%|███████▉  | 8506/10682 [1:12:16<17:51,  2.03it/s]
 80%|███████▉  | 8507/10682 [1:12:17<17:51,  2.03it/s]
 80%|███████▉  | 8508/10682 [1:12:17<17:49,  2.03it/s]
 80%|███████▉  | 8509/10682 [1:12:18<17:50,  2.03it/s]
 80%|███████▉  | 8510/10682 [1:12:18<17:48,  2.03it/s]
 80%|███████▉  | 8511/10682 [1:12:19<17:48,  2.03it/s]
 80%|███████▉  | 8512/10682 [1:12:19<17:47,  2.03it/s]
 80%|███████▉  | 8513/10682 [1:12:20<17:47,  2.03it/s]
 80%|███████▉  | 8514/10682 [1:12:20<17:46,  2.03it/s]
 80%|███████▉  | 8515/10682 [1:12:21<17:46,  2.03it/s]
 80%|███████▉  | 8516/10682 [1:12:21<17:46,  2.03it/s]
 80%|███████▉  | 8517/10682 [1:12:22<17:45,  2.03it/s]
 80%|███████▉  | 8518/10682 [1:12:22<17:44,  2.03it/s]
 80%|███████▉  | 8519/10682 [1:12:23<17:44,  2.03it/s]
 80%|███████▉  | 8520/10682 [1:12:23<17:43,  2.03it/s]
 80%|███████▉  | 8521/10682 [1:12:24<17:43,  2.03it/s]
 80%|███████▉  | 8522/10682 [1:12:24<17:43,  2.03it/s]
 80%|███████▉  | 8523/10682 [1:12:25<17:41,  2.03it/s]
 80%|███████▉  | 8524/10682 [1:12:25<17:41,  2.03it/s]
 80%|███████▉  | 8525/10682 [1:12:26<17:41,  2.03it/s]
                                                      
+
 80%|███████▉  | 8525/10682 [1:12:26<17:41,  2.03it/s]
 80%|███████▉  | 8526/10682 [1:12:26<17:42,  2.03it/s]
 80%|███████▉  | 8527/10682 [1:12:27<17:42,  2.03it/s]
 80%|███████▉  | 8528/10682 [1:12:27<17:40,  2.03it/s]
 80%|███████▉  | 8529/10682 [1:12:28<17:40,  2.03it/s]
 80%|███████▉  | 8530/10682 [1:12:28<17:39,  2.03it/s]
 80%|███████▉  | 8531/10682 [1:12:29<17:38,  2.03it/s]
 80%|███████▉  | 8532/10682 [1:12:29<17:39,  2.03it/s]
 80%|███████▉  | 8533/10682 [1:12:30<17:37,  2.03it/s]
 80%|███████▉  | 8534/10682 [1:12:30<17:36,  2.03it/s]
 80%|███████▉  | 8535/10682 [1:12:31<17:36,  2.03it/s]
 80%|███████▉  | 8536/10682 [1:12:31<17:36,  2.03it/s]
 80%|███████▉  | 8537/10682 [1:12:32<17:35,  2.03it/s]
 80%|███████▉  | 8538/10682 [1:12:32<17:35,  2.03it/s]
 80%|███████▉  | 8539/10682 [1:12:33<17:34,  2.03it/s]
 80%|███████▉  | 8540/10682 [1:12:33<17:34,  2.03it/s]
 80%|███████▉  | 8541/10682 [1:12:34<17:32,  2.03it/s]
 80%|███████▉  | 8542/10682 [1:12:34<17:32,  2.03it/s]
 80%|███████▉  | 8543/10682 [1:12:35<17:32,  2.03it/s]
 80%|███████▉  | 8544/10682 [1:12:35<17:31,  2.03it/s]
 80%|███████▉  | 8545/10682 [1:12:36<17:32,  2.03it/s]
 80%|████████  | 8546/10682 [1:12:36<17:31,  2.03it/s]
 80%|████████  | 8547/10682 [1:12:37<17:31,  2.03it/s]
 80%|████████  | 8548/10682 [1:12:37<17:31,  2.03it/s]
 80%|████████  | 8549/10682 [1:12:38<17:29,  2.03it/s]
 80%|████████  | 8550/10682 [1:12:38<17:31,  2.03it/s]{'loss': 2.8274, 'grad_norm': 0.271226167678833, 'learning_rate': 0.0001165346678316832, 'epoch': 11.21}
+
                                                      
 80%|████████  | 8550/10682 [1:12:38<17:31,  2.03it/s]
 80%|████████  | 8551/10682 [1:12:39<17:32,  2.03it/s]
 80%|████████  | 8552/10682 [1:12:39<17:31,  2.03it/s]
 80%|████████  | 8553/10682 [1:12:40<17:30,  2.03it/s]
 80%|████████  | 8554/10682 [1:12:40<17:30,  2.03it/s]
 80%|████████  | 8555/10682 [1:12:41<17:29,  2.03it/s]
 80%|████████  | 8556/10682 [1:12:41<17:28,  2.03it/s]
 80%|████████  | 8557/10682 [1:12:42<17:27,  2.03it/s]
 80%|████████  | 8558/10682 [1:12:42<17:26,  2.03it/s]
 80%|████████  | 8559/10682 [1:12:43<17:27,  2.03it/s]
 80%|████████  | 8560/10682 [1:12:43<17:25,  2.03it/s]
 80%|████████  | 8561/10682 [1:12:44<17:25,  2.03it/s]
 80%|████████  | 8562/10682 [1:12:44<17:25,  2.03it/s]
 80%|████████  | 8563/10682 [1:12:45<17:23,  2.03it/s]
 80%|████████  | 8564/10682 [1:12:45<17:23,  2.03it/s]
 80%|████████  | 8565/10682 [1:12:46<17:22,  2.03it/s]
 80%|████████  | 8566/10682 [1:12:46<17:21,  2.03it/s]
 80%|████████  | 8567/10682 [1:12:46<17:21,  2.03it/s]
 80%|████████  | 8568/10682 [1:12:47<17:20,  2.03it/s]
 80%|████████  | 8569/10682 [1:12:47<17:20,  2.03it/s]
 80%|████████  | 8570/10682 [1:12:48<17:19,  2.03it/s]
 80%|████████  | 8571/10682 [1:12:48<17:18,  2.03it/s]
 80%|████████  | 8572/10682 [1:12:49<17:18,  2.03it/s]
 80%|████████  | 8573/10682 [1:12:49<17:17,  2.03it/s]
 80%|████████  | 8574/10682 [1:12:50<17:16,  2.03it/s]
 80%|████████  | 8575/10682 [1:12:50<17:16,  2.03it/s]{'loss': 2.8339, 'grad_norm': 0.2710798978805542, 'learning_rate': 0.00011392597484393285, 'epoch': 11.24}
                                                      
+
 80%|████████  | 8575/10682 [1:12:50<17:16,  2.03it/s]
 80%|████████  | 8576/10682 [1:12:51<17:16,  2.03it/s]
 80%|████████  | 8577/10682 [1:12:51<17:16,  2.03it/s]
 80%|████████  | 8578/10682 [1:12:52<17:16,  2.03it/s]
 80%|████████  | 8579/10682 [1:12:52<17:15,  2.03it/s]
 80%|████████  | 8580/10682 [1:12:53<17:15,  2.03it/s]
 80%|████████  | 8581/10682 [1:12:53<17:13,  2.03it/s]
 80%|████████  | 8582/10682 [1:12:54<17:13,  2.03it/s]
 80%|████████  | 8583/10682 [1:12:54<17:13,  2.03it/s]
 80%|████████  | 8584/10682 [1:12:55<17:12,  2.03it/s]
 80%|████████  | 8585/10682 [1:12:55<17:13,  2.03it/s]
 80%|████████  | 8586/10682 [1:12:56<17:12,  2.03it/s]
 80%|████████  | 8587/10682 [1:12:56<17:10,  2.03it/s]
 80%|████████  | 8588/10682 [1:12:57<17:11,  2.03it/s]
 80%|████████  | 8589/10682 [1:12:57<17:09,  2.03it/s]
 80%|████████  | 8590/10682 [1:12:58<17:09,  2.03it/s]
 80%|████████  | 8591/10682 [1:12:58<17:10,  2.03it/s]
 80%|████████  | 8592/10682 [1:12:59<17:08,  2.03it/s]
 80%|████████  | 8593/10682 [1:12:59<17:09,  2.03it/s]
 80%|████████  | 8594/10682 [1:13:00<17:08,  2.03it/s]
 80%|████████  | 8595/10682 [1:13:00<17:08,  2.03it/s]
 80%|████████  | 8596/10682 [1:13:01<17:08,  2.03it/s]
 80%|████████  | 8597/10682 [1:13:01<17:06,  2.03it/s]
 80%|████████  | 8598/10682 [1:13:02<17:08,  2.03it/s]
 80%|████████  | 8599/10682 [1:13:02<17:06,  2.03it/s]
 81%|████████  | 8600/10682 [1:13:03<17:04,  2.03it/s]
                                                      
+
 81%|████████  | 8600/10682 [1:13:03<17:04,  2.03it/s]
 81%|████████  | 8601/10682 [1:13:03<17:06,  2.03it/s]
 81%|████████  | 8602/10682 [1:13:04<17:05,  2.03it/s]
 81%|████████  | 8603/10682 [1:13:04<17:05,  2.03it/s]
 81%|████████  | 8604/10682 [1:13:05<17:04,  2.03it/s]
 81%|████████  | 8605/10682 [1:13:05<17:04,  2.03it/s]
 81%|████████  | 8606/10682 [1:13:06<17:03,  2.03it/s]
 81%|████████  | 8607/10682 [1:13:06<17:03,  2.03it/s]
 81%|████████  | 8608/10682 [1:13:07<17:02,  2.03it/s]
 81%|████████  | 8609/10682 [1:13:07<17:01,  2.03it/s]
 81%|████████  | 8610/10682 [1:13:08<17:02,  2.03it/s]
 81%|████████  | 8611/10682 [1:13:08<17:01,  2.03it/s]
 81%|████████  | 8612/10682 [1:13:09<17:01,  2.03it/s]
 81%|████████  | 8613/10682 [1:13:09<16:59,  2.03it/s]
 81%|████████  | 8614/10682 [1:13:10<16:59,  2.03it/s]
 81%|████████  | 8615/10682 [1:13:10<16:58,  2.03it/s]
 81%|████████  | 8616/10682 [1:13:11<16:57,  2.03it/s]
 81%|████████  | 8617/10682 [1:13:11<16:58,  2.03it/s]
 81%|████████  | 8618/10682 [1:13:12<16:56,  2.03it/s]
 81%|████████  | 8619/10682 [1:13:12<16:55,  2.03it/s]
 81%|████████  | 8620/10682 [1:13:13<16:55,  2.03it/s]
 81%|████████  | 8621/10682 [1:13:13<16:54,  2.03it/s]
 81%|████████  | 8622/10682 [1:13:14<16:54,  2.03it/s]
 81%|████████  | 8623/10682 [1:13:14<16:54,  2.03it/s]
 81%|████████  | 8624/10682 [1:13:15<16:53,  2.03it/s]
 81%|████████  | 8625/10682 [1:13:15<16:53,  2.03it/s]
                                                      
+
 81%|████████  | 8625/10682 [1:13:15<16:53,  2.03it/s]
 81%|████████  | 8626/10682 [1:13:16<16:53,  2.03it/s]
 81%|████████  | 8627/10682 [1:13:16<16:53,  2.03it/s]
 81%|████████  | 8628/10682 [1:13:17<16:52,  2.03it/s]
 81%|████████  | 8629/10682 [1:13:17<16:52,  2.03it/s]
 81%|████████  | 8630/10682 [1:13:18<16:50,  2.03it/s]
 81%|████████  | 8631/10682 [1:13:18<16:50,  2.03it/s]
 81%|████████  | 8632/10682 [1:13:19<16:50,  2.03it/s]
 81%|████████  | 8633/10682 [1:13:19<16:49,  2.03it/s]
 81%|████████  | 8634/10682 [1:13:19<16:48,  2.03it/s]
 81%|████████  | 8635/10682 [1:13:20<16:48,  2.03it/s]
 81%|████████  | 8636/10682 [1:13:20<16:47,  2.03it/s]
 81%|████████  | 8637/10682 [1:13:21<16:47,  2.03it/s]
 81%|████████  | 8638/10682 [1:13:21<16:46,  2.03it/s]
 81%|████████  | 8639/10682 [1:13:22<16:45,  2.03it/s]
 81%|████████  | 8640/10682 [1:13:22<16:45,  2.03it/s]
 81%|████████  | 8641/10682 [1:13:23<16:44,  2.03it/s]
 81%|████████  | 8642/10682 [1:13:23<16:44,  2.03it/s]
 81%|████████  | 8643/10682 [1:13:24<16:43,  2.03it/s]
 81%|████████  | 8644/10682 [1:13:24<16:43,  2.03it/s]
 81%|████████  | 8645/10682 [1:13:25<16:44,  2.03it/s]
 81%|████████  | 8646/10682 [1:13:25<16:43,  2.03it/s]
 81%|████████  | 8647/10682 [1:13:26<16:44,  2.03it/s]
 81%|████████  | 8648/10682 [1:13:26<16:43,  2.03it/s]
 81%|████████  | 8649/10682 [1:13:27<16:42,  2.03it/s]
 81%|████████  | 8650/10682 [1:13:27<16:42,  2.03it/s]
                                                      {'loss': 2.8442, 'grad_norm': 0.2641572654247284, 'learning_rate': 0.00010625520936593375, 'epoch': 11.34}
+
 81%|████████  | 8650/10682 [1:13:27<16:42,  2.03it/s]
 81%|████████  | 8651/10682 [1:13:28<16:42,  2.03it/s]
 81%|████████  | 8652/10682 [1:13:28<16:41,  2.03it/s]
 81%|████████  | 8653/10682 [1:13:29<16:41,  2.03it/s]
 81%|████████  | 8654/10682 [1:13:29<16:40,  2.03it/s]
 81%|████████  | 8655/10682 [1:13:30<16:39,  2.03it/s]
 81%|████████  | 8656/10682 [1:13:30<16:38,  2.03it/s]
 81%|████████  | 8657/10682 [1:13:31<16:37,  2.03it/s]
 81%|████████  | 8658/10682 [1:13:31<16:36,  2.03it/s]
 81%|████████  | 8659/10682 [1:13:32<16:35,  2.03it/s]
 81%|████████  | 8660/10682 [1:13:32<16:35,  2.03it/s]
 81%|████████  | 8661/10682 [1:13:33<16:35,  2.03it/s]
 81%|████████  | 8662/10682 [1:13:33<16:34,  2.03it/s]
 81%|████████  | 8663/10682 [1:13:34<16:33,  2.03it/s]
 81%|████████  | 8664/10682 [1:13:34<16:33,  2.03it/s]
 81%|████████  | 8665/10682 [1:13:35<16:33,  2.03it/s]
 81%|████████  | 8666/10682 [1:13:35<16:33,  2.03it/s]
 81%|████████  | 8667/10682 [1:13:36<16:32,  2.03it/s]
 81%|████████  | 8668/10682 [1:13:36<16:31,  2.03it/s]
 81%|████████  | 8669/10682 [1:13:37<16:31,  2.03it/s]
 81%|████████  | 8670/10682 [1:13:37<16:31,  2.03it/s]
 81%|████████  | 8671/10682 [1:13:38<16:31,  2.03it/s]
 81%|████████  | 8672/10682 [1:13:38<16:30,  2.03it/s]
 81%|████████  | 8673/10682 [1:13:39<16:28,  2.03it/s]
 81%|████████  | 8674/10682 [1:13:39<16:29,  2.03it/s]
 81%|████████  | 8675/10682 [1:13:40<16:28,  2.03it/s]{'loss': 2.8423, 'grad_norm': 0.2601149380207062, 'learning_rate': 0.0001037506276287885, 'epoch': 11.37}
+
                                                      
 81%|████████  | 8675/10682 [1:13:40<16:28,  2.03it/s]
 81%|████████  | 8676/10682 [1:13:40<16:29,  2.03it/s]
 81%|████████  | 8677/10682 [1:13:41<16:28,  2.03it/s]
 81%|████████  | 8678/10682 [1:13:41<16:28,  2.03it/s]
 81%|████████  | 8679/10682 [1:13:42<16:27,  2.03it/s]
 81%|████████▏ | 8680/10682 [1:13:42<16:26,  2.03it/s]
 81%|████████▏ | 8681/10682 [1:13:43<16:26,  2.03it/s]
 81%|████████▏ | 8682/10682 [1:13:43<16:24,  2.03it/s]
 81%|████████▏ | 8683/10682 [1:13:44<16:24,  2.03it/s]
 81%|████████▏ | 8684/10682 [1:13:44<16:23,  2.03it/s]
 81%|████████▏ | 8685/10682 [1:13:45<16:23,  2.03it/s]
 81%|████████▏ | 8686/10682 [1:13:45<16:23,  2.03it/s]
 81%|████████▏ | 8687/10682 [1:13:46<16:21,  2.03it/s]
 81%|████████▏ | 8688/10682 [1:13:46<16:21,  2.03it/s]
 81%|████████▏ | 8689/10682 [1:13:47<16:21,  2.03it/s]
 81%|████████▏ | 8690/10682 [1:13:47<16:19,  2.03it/s]
 81%|████████▏ | 8691/10682 [1:13:48<16:19,  2.03it/s]
 81%|████████▏ | 8692/10682 [1:13:48<16:18,  2.03it/s]
 81%|████████▏ | 8693/10682 [1:13:49<16:18,  2.03it/s]
 81%|████████▏ | 8694/10682 [1:13:49<16:18,  2.03it/s]
 81%|████████▏ | 8695/10682 [1:13:50<16:17,  2.03it/s]
 81%|████████▏ | 8696/10682 [1:13:50<16:18,  2.03it/s]
 81%|████████▏ | 8697/10682 [1:13:51<16:17,  2.03it/s]
 81%|████████▏ | 8698/10682 [1:13:51<16:16,  2.03it/s]
 81%|████████▏ | 8699/10682 [1:13:52<16:16,  2.03it/s]
 81%|████████▏ | 8700/10682 [1:13:52<16:15,  2.03it/s]{'loss': 2.8562, 'grad_norm': 0.2651384770870209, 'learning_rate': 0.0001012724960361826, 'epoch': 11.4}
+
                                                      
 81%|████████▏ | 8700/10682 [1:13:52<16:15,  2.03it/s]
 81%|████████▏ | 8701/10682 [1:13:53<16:16,  2.03it/s]
 81%|████████▏ | 8702/10682 [1:13:53<16:14,  2.03it/s]
 81%|████████▏ | 8703/10682 [1:13:53<16:14,  2.03it/s]
 81%|████████▏ | 8704/10682 [1:13:54<16:14,  2.03it/s]
 81%|████████▏ | 8705/10682 [1:13:54<16:13,  2.03it/s]
 82%|████████▏ | 8706/10682 [1:13:55<16:13,  2.03it/s]
 82%|████████▏ | 8707/10682 [1:13:55<16:12,  2.03it/s]
 82%|████████▏ | 8708/10682 [1:13:56<16:10,  2.03it/s]
 82%|████████▏ | 8709/10682 [1:13:56<16:11,  2.03it/s]
 82%|████████▏ | 8710/10682 [1:13:57<16:10,  2.03it/s]
 82%|████████▏ | 8711/10682 [1:13:57<16:10,  2.03it/s]
 82%|████████▏ | 8712/10682 [1:13:58<16:09,  2.03it/s]
 82%|████████▏ | 8713/10682 [1:13:58<16:09,  2.03it/s]
 82%|████████▏ | 8714/10682 [1:13:59<16:09,  2.03it/s]
 82%|████████▏ | 8715/10682 [1:13:59<16:08,  2.03it/s]
 82%|████████▏ | 8716/10682 [1:14:00<16:08,  2.03it/s]
 82%|████████▏ | 8717/10682 [1:14:00<16:08,  2.03it/s]
 82%|████████▏ | 8718/10682 [1:14:01<16:06,  2.03it/s]
 82%|████████▏ | 8719/10682 [1:14:01<16:06,  2.03it/s]
 82%|████████▏ | 8720/10682 [1:14:02<16:05,  2.03it/s]
 82%|████████▏ | 8721/10682 [1:14:02<16:04,  2.03it/s]
 82%|████████▏ | 8722/10682 [1:14:03<16:05,  2.03it/s]
 82%|████████▏ | 8723/10682 [1:14:03<16:04,  2.03it/s]
 82%|████████▏ | 8724/10682 [1:14:04<16:04,  2.03it/s]
 82%|████████▏ | 8725/10682 [1:14:04<16:03,  2.03it/s]{'loss': 2.8428, 'grad_norm': 0.26036229729652405, 'learning_rate': 9.882098000652034e-05, 'epoch': 11.44}
+
                                                      
 82%|████████▏ | 8725/10682 [1:14:04<16:03,  2.03it/s]
 82%|████████▏ | 8726/10682 [1:14:05<16:04,  2.03it/s]
 82%|████████▏ | 8727/10682 [1:14:05<16:03,  2.03it/s]
 82%|████████▏ | 8728/10682 [1:14:06<16:02,  2.03it/s]
 82%|████████▏ | 8729/10682 [1:14:06<16:03,  2.03it/s]
 82%|████████▏ | 8730/10682 [1:14:07<16:02,  2.03it/s]
 82%|████████▏ | 8731/10682 [1:14:07<16:01,  2.03it/s]
 82%|████████▏ | 8732/10682 [1:14:08<16:01,  2.03it/s]
 82%|████████▏ | 8733/10682 [1:14:08<16:00,  2.03it/s]
 82%|████████▏ | 8734/10682 [1:14:09<16:00,  2.03it/s]
 82%|████████▏ | 8735/10682 [1:14:09<15:59,  2.03it/s]
 82%|████████▏ | 8736/10682 [1:14:10<15:59,  2.03it/s]
 82%|████████▏ | 8737/10682 [1:14:10<15:58,  2.03it/s]
 82%|████████▏ | 8738/10682 [1:14:11<15:57,  2.03it/s]
 82%|████████▏ | 8739/10682 [1:14:11<15:57,  2.03it/s]
 82%|████████▏ | 8740/10682 [1:14:12<15:56,  2.03it/s]
 82%|████████▏ | 8741/10682 [1:14:12<15:55,  2.03it/s]
 82%|████████▏ | 8742/10682 [1:14:13<15:55,  2.03it/s]
 82%|████████▏ | 8743/10682 [1:14:13<15:55,  2.03it/s]
 82%|████████▏ | 8744/10682 [1:14:14<15:55,  2.03it/s]
 82%|████████▏ | 8745/10682 [1:14:14<15:53,  2.03it/s]
 82%|████████▏ | 8746/10682 [1:14:15<15:53,  2.03it/s]
 82%|████████▏ | 8747/10682 [1:14:15<15:53,  2.03it/s]
 82%|████████▏ | 8748/10682 [1:14:16<15:52,  2.03it/s]
 82%|████████▏ | 8749/10682 [1:14:16<15:52,  2.03it/s]
 82%|████████▏ | 8750/10682 [1:14:17<15:51,  2.03it/s]{'loss': 2.8425, 'grad_norm': 0.2622228264808655, 'learning_rate': 9.639624318158335e-05, 'epoch': 11.47}
+
                                                      
 82%|████████▏ | 8750/10682 [1:14:17<15:51,  2.03it/s]
 82%|████████�� | 8751/10682 [1:14:17<15:52,  2.03it/s]
 82%|████████▏ | 8752/10682 [1:14:18<15:51,  2.03it/s]
 82%|████████▏ | 8753/10682 [1:14:18<15:50,  2.03it/s]
 82%|████████▏ | 8754/10682 [1:14:19<15:49,  2.03it/s]
 82%|████████▏ | 8755/10682 [1:14:19<15:48,  2.03it/s]
 82%|████████▏ | 8756/10682 [1:14:20<15:48,  2.03it/s]
 82%|████████▏ | 8757/10682 [1:14:20<15:47,  2.03it/s]
 82%|████████▏ | 8758/10682 [1:14:21<15:46,  2.03it/s]
 82%|████████▏ | 8759/10682 [1:14:21<15:45,  2.03it/s]
 82%|████████▏ | 8760/10682 [1:14:22<15:45,  2.03it/s]
 82%|████████▏ | 8761/10682 [1:14:22<15:44,  2.03it/s]
 82%|████████▏ | 8762/10682 [1:14:23<15:45,  2.03it/s]
 82%|████████▏ | 8763/10682 [1:14:23<15:44,  2.03it/s]
 82%|████████▏ | 8764/10682 [1:14:24<15:44,  2.03it/s]
 82%|████████▏ | 8765/10682 [1:14:24<15:44,  2.03it/s]
 82%|████████▏ | 8766/10682 [1:14:25<15:43,  2.03it/s]
 82%|████████▏ | 8767/10682 [1:14:25<15:43,  2.03it/s]
 82%|████████▏ | 8768/10682 [1:14:25<15:42,  2.03it/s]
 82%|████████▏ | 8769/10682 [1:14:26<15:40,  2.03it/s]
 82%|████████▏ | 8770/10682 [1:14:26<15:41,  2.03it/s]
 82%|████████▏ | 8771/10682 [1:14:27<15:40,  2.03it/s]
 82%|████████▏ | 8772/10682 [1:14:27<15:40,  2.03it/s]
 82%|████████▏ | 8773/10682 [1:14:28<15:40,  2.03it/s]
 82%|████████▏ | 8774/10682 [1:14:28<15:38,  2.03it/s]
 82%|████████▏ | 8775/10682 [1:14:29<15:39,  2.03it/s]{'loss': 2.8524, 'grad_norm': 0.26476430892944336, 'learning_rate': 9.399844741560781e-05, 'epoch': 11.5}
                                                      
+
 82%|████████▏ | 8775/10682 [1:14:29<15:39,  2.03it/s]
 82%|████████▏ | 8776/10682 [1:14:29<15:40,  2.03it/s]
 82%|████████▏ | 8777/10682 [1:14:30<15:41,  2.02it/s]
 82%|████████▏ | 8778/10682 [1:14:30<15:38,  2.03it/s]
 82%|████████▏ | 8779/10682 [1:14:31<15:38,  2.03it/s]
 82%|████████▏ | 8780/10682 [1:14:31<15:38,  2.03it/s]
 82%|████████▏ | 8781/10682 [1:14:32<15:36,  2.03it/s]
 82%|████████▏ | 8782/10682 [1:14:32<15:36,  2.03it/s]
 82%|████████▏ | 8783/10682 [1:14:33<15:35,  2.03it/s]
 82%|████████▏ | 8784/10682 [1:14:33<15:33,  2.03it/s]
 82%|████████▏ | 8785/10682 [1:14:34<15:34,  2.03it/s]
 82%|████████▏ | 8786/10682 [1:14:34<15:33,  2.03it/s]
 82%|████████▏ | 8787/10682 [1:14:35<15:33,  2.03it/s]
 82%|████████▏ | 8788/10682 [1:14:35<15:34,  2.03it/s]
 82%|████████▏ | 8789/10682 [1:14:36<15:33,  2.03it/s]
 82%|████████▏ | 8790/10682 [1:14:36<15:33,  2.03it/s]
 82%|████████▏ | 8791/10682 [1:14:37<15:32,  2.03it/s]
 82%|████████▏ | 8792/10682 [1:14:37<15:31,  2.03it/s]
 82%|████████▏ | 8793/10682 [1:14:38<15:31,  2.03it/s]
 82%|████████▏ | 8794/10682 [1:14:38<15:30,  2.03it/s]
 82%|████████▏ | 8795/10682 [1:14:39<15:29,  2.03it/s]
 82%|████████▏ | 8796/10682 [1:14:39<15:29,  2.03it/s]
 82%|████████▏ | 8797/10682 [1:14:40<15:28,  2.03it/s]
 82%|████████▏ | 8798/10682 [1:14:40<15:26,  2.03it/s]
 82%|████████▏ | 8799/10682 [1:14:41<15:27,  2.03it/s]
 82%|████████▏ | 8800/10682 [1:14:41<15:27,  2.03it/s]{'loss': 2.8376, 'grad_norm': 0.25995054841041565, 'learning_rate': 9.162775276448015e-05, 'epoch': 11.53}
+
                                                      
 82%|████████▏ | 8800/10682 [1:14:41<15:27,  2.03it/s]
 82%|████████▏ | 8801/10682 [1:14:42<15:28,  2.03it/s]
 82%|████████▏ | 8802/10682 [1:14:42<15:26,  2.03it/s]
 82%|████████▏ | 8803/10682 [1:14:43<15:24,  2.03it/s]
 82%|████████▏ | 8804/10682 [1:14:43<15:23,  2.03it/s]
 82%|████████▏ | 8805/10682 [1:14:44<15:23,  2.03it/s]
 82%|████████▏ | 8806/10682 [1:14:44<15:23,  2.03it/s]
 82%|████████▏ | 8807/10682 [1:14:45<15:22,  2.03it/s]
 82%|████████▏ | 8808/10682 [1:14:45<15:22,  2.03it/s]
 82%|████████▏ | 8809/10682 [1:14:46<15:22,  2.03it/s]
 82%|████████▏ | 8810/10682 [1:14:46<15:21,  2.03it/s]
 82%|████████▏ | 8811/10682 [1:14:47<15:21,  2.03it/s]
 82%|████████▏ | 8812/10682 [1:14:47<15:21,  2.03it/s]
 83%|████████▎ | 8813/10682 [1:14:48<15:20,  2.03it/s]
 83%|████████▎ | 8814/10682 [1:14:48<15:20,  2.03it/s]
 83%|���███████▎ | 8815/10682 [1:14:49<15:18,  2.03it/s]
 83%|████████▎ | 8816/10682 [1:14:49<15:18,  2.03it/s]
 83%|████████▎ | 8817/10682 [1:14:50<15:18,  2.03it/s]
 83%|████████▎ | 8818/10682 [1:14:50<15:18,  2.03it/s]
 83%|████████▎ | 8819/10682 [1:14:51<15:18,  2.03it/s]
 83%|████████▎ | 8820/10682 [1:14:51<15:16,  2.03it/s]
 83%|████████▎ | 8821/10682 [1:14:52<15:16,  2.03it/s]
 83%|████████▎ | 8822/10682 [1:14:52<15:16,  2.03it/s]
 83%|████████▎ | 8823/10682 [1:14:53<15:15,  2.03it/s]
 83%|████████▎ | 8824/10682 [1:14:53<15:15,  2.03it/s]
 83%|████████▎ | 8825/10682 [1:14:54<15:14,  2.03it/s]
                                                      
+
 83%|████████▎ | 8825/10682 [1:14:54<15:14,  2.03it/s]
 83%|████████▎ | 8826/10682 [1:14:54<15:15,  2.03it/s]
 83%|████████▎ | 8827/10682 [1:14:55<15:14,  2.03it/s]
 83%|████████▎ | 8828/10682 [1:14:55<15:13,  2.03it/s]
 83%|████████▎ | 8829/10682 [1:14:56<16:34,  1.86it/s]
 83%|████████▎ | 8830/10682 [1:14:56<16:09,  1.91it/s]
 83%|████████▎ | 8831/10682 [1:14:57<15:51,  1.94it/s]
 83%|████████▎ | 8832/10682 [1:14:57<15:38,  1.97it/s]
 83%|████████▎ | 8833/10682 [1:14:58<15:30,  1.99it/s]
 83%|████████▎ | 8834/10682 [1:14:58<15:23,  2.00it/s]
 83%|████████▎ | 8835/10682 [1:14:59<15:19,  2.01it/s]
 83%|████████▎ | 8836/10682 [1:14:59<15:15,  2.02it/s]
 83%|████████▎ | 8837/10682 [1:15:00<15:13,  2.02it/s]
 83%|████████▎ | 8838/10682 [1:15:00<15:11,  2.02it/s]
 83%|████████▎ | 8839/10682 [1:15:01<15:09,  2.03it/s]
 83%|████████▎ | 8840/10682 [1:15:01<15:09,  2.03it/s]
 83%|████████▎ | 8841/10682 [1:15:02<15:07,  2.03it/s]
 83%|████████▎ | 8842/10682 [1:15:02<15:06,  2.03it/s]
 83%|████████▎ | 8843/10682 [1:15:03<15:06,  2.03it/s]
 83%|████████▎ | 8844/10682 [1:15:03<15:05,  2.03it/s]
 83%|████████▎ | 8845/10682 [1:15:04<15:05,  2.03it/s]
 83%|████████▎ | 8846/10682 [1:15:04<15:04,  2.03it/s]
 83%|████████▎ | 8847/10682 [1:15:05<15:04,  2.03it/s]
 83%|████████▎ | 8848/10682 [1:15:05<15:03,  2.03it/s]
 83%|████████▎ | 8849/10682 [1:15:06<15:02,  2.03it/s]
 83%|████████▎ | 8850/10682 [1:15:06<15:02,  2.03it/s]{'loss': 2.8552, 'grad_norm': 0.2628968060016632, 'learning_rate': 8.6968297974584e-05, 'epoch': 11.6}
                                                      
+
 83%|████████▎ | 8850/10682 [1:15:06<15:02,  2.03it/s]
 83%|████████▎ | 8851/10682 [1:15:07<15:03,  2.03it/s]
 83%|████████▎ | 8852/10682 [1:15:07<15:02,  2.03it/s]
 83%|████████▎ | 8853/10682 [1:15:08<15:01,  2.03it/s]
 83%|████████▎ | 8854/10682 [1:15:08<15:01,  2.03it/s]
 83%|████████▎ | 8855/10682 [1:15:09<15:01,  2.03it/s]
 83%|████████▎ | 8856/10682 [1:15:09<15:00,  2.03it/s]
 83%|████████▎ | 8857/10682 [1:15:09<15:01,  2.03it/s]
 83%|████████▎ | 8858/10682 [1:15:10<14:59,  2.03it/s]
 83%|████████▎ | 8859/10682 [1:15:10<14:58,  2.03it/s]
 83%|████████▎ | 8860/10682 [1:15:11<14:58,  2.03it/s]
 83%|████████▎ | 8861/10682 [1:15:11<14:57,  2.03it/s]
 83%|████████▎ | 8862/10682 [1:15:12<14:56,  2.03it/s]
 83%|████████▎ | 8863/10682 [1:15:12<14:56,  2.03it/s]
 83%|████████▎ | 8864/10682 [1:15:13<14:55,  2.03it/s]
 83%|████████▎ | 8865/10682 [1:15:13<14:55,  2.03it/s]
 83%|████████▎ | 8866/10682 [1:15:14<14:54,  2.03it/s]
 83%|████████▎ | 8867/10682 [1:15:14<14:53,  2.03it/s]
 83%|████████▎ | 8868/10682 [1:15:15<14:53,  2.03it/s]
 83%|████████▎ | 8869/10682 [1:15:15<14:51,  2.03it/s]
 83%|████████▎ | 8870/10682 [1:15:16<14:52,  2.03it/s]
 83%|████████▎ | 8871/10682 [1:15:16<14:51,  2.03it/s]
 83%|████████▎ | 8872/10682 [1:15:17<14:51,  2.03it/s]
 83%|████████▎ | 8873/10682 [1:15:17<14:50,  2.03it/s]
 83%|████████▎ | 8874/10682 [1:15:18<14:49,  2.03it/s]
 83%|████████▎ | 8875/10682 [1:15:18<14:50,  2.03it/s]
                                                      {'loss': 2.8535, 'grad_norm': 0.2602517902851105, 'learning_rate': 8.467984886028967e-05, 'epoch': 11.63}
+
 83%|████████▎ | 8875/10682 [1:15:18<14:50,  2.03it/s]
 83%|████████▎ | 8876/10682 [1:15:19<14:50,  2.03it/s]
 83%|████████▎ | 8877/10682 [1:15:19<14:50,  2.03it/s]
 83%|████████▎ | 8878/10682 [1:15:20<14:49,  2.03it/s]
 83%|████████▎ | 8879/10682 [1:15:20<14:47,  2.03it/s]
 83%|████████▎ | 8880/10682 [1:15:21<14:47,  2.03it/s]
 83%|████████▎ | 8881/10682 [1:15:21<14:46,  2.03it/s]
 83%|████████▎ | 8882/10682 [1:15:22<14:46,  2.03it/s]
 83%|████████▎ | 8883/10682 [1:15:22<14:45,  2.03it/s]
 83%|████████▎ | 8884/10682 [1:15:23<14:44,  2.03it/s]
 83%|████████▎ | 8885/10682 [1:15:23<14:45,  2.03it/s]
 83%|████████▎ | 8886/10682 [1:15:24<14:43,  2.03it/s]
 83%|████████▎ | 8887/10682 [1:15:24<14:43,  2.03it/s]
 83%|████████▎ | 8888/10682 [1:15:25<14:43,  2.03it/s]
 83%|████████▎ | 8889/10682 [1:15:25<14:42,  2.03it/s]
 83%|████████▎ | 8890/10682 [1:15:26<14:42,  2.03it/s]
 83%|████████▎ | 8891/10682 [1:15:26<14:41,  2.03it/s]
 83%|████████▎ | 8892/10682 [1:15:27<14:41,  2.03it/s]
 83%|████████▎ | 8893/10682 [1:15:27<14:41,  2.03it/s]
 83%|████████▎ | 8894/10682 [1:15:28<14:40,  2.03it/s]
 83%|████████▎ | 8895/10682 [1:15:28<14:40,  2.03it/s]
 83%|████████▎ | 8896/10682 [1:15:29<14:38,  2.03it/s]
 83%|████████▎ | 8897/10682 [1:15:29<14:38,  2.03it/s]
 83%|████████▎ | 8898/10682 [1:15:30<14:38,  2.03it/s]
 83%|████████▎ | 8899/10682 [1:15:30<14:37,  2.03it/s]
 83%|████████▎ | 8900/10682 [1:15:31<14:37,  2.03it/s]{'loss': 2.8613, 'grad_norm': 0.264787495136261, 'learning_rate': 8.24191228890303e-05, 'epoch': 11.66}
+
                                                      
 83%|████████▎ | 8900/10682 [1:15:31<14:37,  2.03it/s]
 83%|████████▎ | 8901/10682 [1:15:31<14:37,  2.03it/s]
 83%|████████▎ | 8902/10682 [1:15:32<14:36,  2.03it/s]
 83%|████████▎ | 8903/10682 [1:15:32<14:35,  2.03it/s]
 83%|████████▎ | 8904/10682 [1:15:33<14:34,  2.03it/s]
 83%|████████▎ | 8905/10682 [1:15:33<14:34,  2.03it/s]
 83%|████████▎ | 8906/10682 [1:15:34<14:34,  2.03it/s]
 83%|████████▎ | 8907/10682 [1:15:34<14:33,  2.03it/s]
 83%|████████▎ | 8908/10682 [1:15:35<14:32,  2.03it/s]
 83%|████████▎ | 8909/10682 [1:15:35<14:32,  2.03it/s]
 83%|████████▎ | 8910/10682 [1:15:36<14:31,  2.03it/s]
 83%|████████▎ | 8911/10682 [1:15:36<14:31,  2.03it/s]
 83%|████████▎ | 8912/10682 [1:15:37<14:30,  2.03it/s]
 83%|████████▎ | 8913/10682 [1:15:37<14:30,  2.03it/s]
 83%|████████▎ | 8914/10682 [1:15:38<14:29,  2.03it/s]
 83%|████████▎ | 8915/10682 [1:15:38<14:29,  2.03it/s]
 83%|████████▎ | 8916/10682 [1:15:39<14:28,  2.03it/s]
 83%|████████▎ | 8917/10682 [1:15:39<14:28,  2.03it/s]
 83%|████████▎ | 8918/10682 [1:15:40<14:26,  2.03it/s]
 83%|████████▎ | 8919/10682 [1:15:40<14:27,  2.03it/s]
 84%|████████▎ | 8920/10682 [1:15:41<14:27,  2.03it/s]
 84%|████████▎ | 8921/10682 [1:15:41<14:26,  2.03it/s]
 84%|████████▎ | 8922/10682 [1:15:41<14:25,  2.03it/s]
 84%|████████▎ | 8923/10682 [1:15:42<14:25,  2.03it/s]
 84%|████████▎ | 8924/10682 [1:15:42<14:24,  2.03it/s]
 84%|████████▎ | 8925/10682 [1:15:43<14:24,  2.03it/s]{'loss': 2.8574, 'grad_norm': 0.26087984442710876, 'learning_rate': 8.018627096711106e-05, 'epoch': 11.7}
                                                      
+
 84%|████████▎ | 8925/10682 [1:15:43<14:24,  2.03it/s]
 84%|████████▎ | 8926/10682 [1:15:43<14:25,  2.03it/s]
 84%|████████▎ | 8927/10682 [1:15:44<14:24,  2.03it/s]
 84%|████████▎ | 8928/10682 [1:15:44<14:25,  2.03it/s]
 84%|████████▎ | 8929/10682 [1:15:45<14:24,  2.03it/s]
 84%|████████▎ | 8930/10682 [1:15:45<14:23,  2.03it/s]
 84%|████████▎ | 8931/10682 [1:15:46<14:23,  2.03it/s]
 84%|████████▎ | 8932/10682 [1:15:47<15:39,  1.86it/s]
 84%|████████▎ | 8933/10682 [1:15:47<15:15,  1.91it/s]
 84%|████████▎ | 8934/10682 [1:15:48<14:58,  1.95it/s]
 84%|████████▎ | 8935/10682 [1:15:48<14:46,  1.97it/s]
 84%|████████▎ | 8936/10682 [1:15:49<14:38,  1.99it/s]
 84%|████████▎ | 8937/10682 [1:15:49<14:32,  2.00it/s]
 84%|████████▎ | 8938/10682 [1:15:50<14:27,  2.01it/s]
 84%|████��███▎ | 8939/10682 [1:15:50<14:24,  2.02it/s]
 84%|████████▎ | 8940/10682 [1:15:51<14:22,  2.02it/s]
 84%|████████▎ | 8941/10682 [1:15:51<14:20,  2.02it/s]
 84%|████████▎ | 8942/10682 [1:15:51<14:19,  2.03it/s]
 84%|████████▎ | 8943/10682 [1:15:52<14:17,  2.03it/s]
 84%|████████▎ | 8944/10682 [1:15:52<14:16,  2.03it/s]
 84%|████████▎ | 8945/10682 [1:15:53<14:15,  2.03it/s]
 84%|████████▎ | 8946/10682 [1:15:53<14:15,  2.03it/s]
 84%|████████▍ | 8947/10682 [1:15:54<14:14,  2.03it/s]
 84%|████████▍ | 8948/10682 [1:15:54<14:13,  2.03it/s]
 84%|████████▍ | 8949/10682 [1:15:55<14:13,  2.03it/s]
 84%|████████▍ | 8950/10682 [1:15:55<14:12,  2.03it/s]{'loss': 2.8577, 'grad_norm': 0.2632240951061249, 'learning_rate': 7.798144214020909e-05, 'epoch': 11.73}
                                                      
+
 84%|████████▍ | 8950/10682 [1:15:55<14:12,  2.03it/s]
 84%|████████▍ | 8951/10682 [1:15:56<14:14,  2.03it/s]
 84%|████████▍ | 8952/10682 [1:15:56<14:12,  2.03it/s]
 84%|████████▍ | 8953/10682 [1:15:57<14:12,  2.03it/s]
 84%|████████▍ | 8954/10682 [1:15:57<14:11,  2.03it/s]
 84%|████████▍ | 8955/10682 [1:15:58<14:10,  2.03it/s]
 84%|████████▍ | 8956/10682 [1:15:58<14:10,  2.03it/s]
 84%|████████▍ | 8957/10682 [1:15:59<14:10,  2.03it/s]
 84%|████████▍ | 8958/10682 [1:15:59<14:07,  2.03it/s]
 84%|████████▍ | 8959/10682 [1:16:00<14:07,  2.03it/s]
 84%|████████▍ | 8960/10682 [1:16:00<14:08,  2.03it/s]
 84%|████████▍ | 8961/10682 [1:16:01<14:08,  2.03it/s]
 84%|████████▍ | 8962/10682 [1:16:01<14:07,  2.03it/s]
 84%|████████▍ | 8963/10682 [1:16:02<14:07,  2.03it/s]
 84%|████████▍ | 8964/10682 [1:16:02<14:06,  2.03it/s]
 84%|████████▍ | 8965/10682 [1:16:03<14:05,  2.03it/s]
 84%|████████▍ | 8966/10682 [1:16:03<14:05,  2.03it/s]
 84%|████████▍ | 8967/10682 [1:16:04<14:05,  2.03it/s]
 84%|████████▍ | 8968/10682 [1:16:04<14:05,  2.03it/s]
 84%|████████▍ | 8969/10682 [1:16:05<14:04,  2.03it/s]
 84%|████████▍ | 8970/10682 [1:16:05<14:03,  2.03it/s]
 84%|████████▍ | 8971/10682 [1:16:06<14:03,  2.03it/s]
 84%|████████▍ | 8972/10682 [1:16:06<14:02,  2.03it/s]
 84%|████████▍ | 8973/10682 [1:16:07<14:02,  2.03it/s]
 84%|████████▍ | 8974/10682 [1:16:07<14:01,  2.03it/s]
 84%|████████▍ | 8975/10682 [1:16:08<14:00,  2.03it/s]
                                                      
+
 84%|████████▍ | 8975/10682 [1:16:08<14:00,  2.03it/s]
 84%|████████▍ | 8976/10682 [1:16:08<14:00,  2.03it/s]
 84%|████████▍ | 8977/10682 [1:16:09<13:59,  2.03it/s]
 84%|████████▍ | 8978/10682 [1:16:09<13:59,  2.03it/s]
 84%|████████▍ | 8979/10682 [1:16:10<13:58,  2.03it/s]
 84%|████████▍ | 8980/10682 [1:16:10<13:58,  2.03it/s]
 84%|████████▍ | 8981/10682 [1:16:11<13:57,  2.03it/s]
 84%|████████▍ | 8982/10682 [1:16:11<13:57,  2.03it/s]
 84%|████████▍ | 8983/10682 [1:16:12<13:57,  2.03it/s]
 84%|████████▍ | 8984/10682 [1:16:12<13:56,  2.03it/s]
 84%|████████▍ | 8985/10682 [1:16:13<13:56,  2.03it/s]
 84%|████████▍ | 8986/10682 [1:16:13<13:55,  2.03it/s]
 84%|████████▍ | 8987/10682 [1:16:14<13:54,  2.03it/s]
 84%|████████▍ | 8988/10682 [1:16:14<13:55,  2.03it/s]
 84%|████████▍ | 8989/10682 [1:16:15<13:53,  2.03it/s]
 84%|████████▍ | 8990/10682 [1:16:15<13:52,  2.03it/s]
 84%|████████▍ | 8991/10682 [1:16:16<13:53,  2.03it/s]
 84%|████████▍ | 8992/10682 [1:16:16<13:52,  2.03it/s]
 84%|████████▍ | 8993/10682 [1:16:17<13:52,  2.03it/s]
 84%|████████▍ | 8994/10682 [1:16:17<13:51,  2.03it/s]
 84%|████████▍ | 8995/10682 [1:16:18<13:50,  2.03it/s]
 84%|████████▍ | 8996/10682 [1:16:18<13:51,  2.03it/s]
 84%|████████▍ | 8997/10682 [1:16:19<13:49,  2.03it/s]
 84%|████████▍ | 8998/10682 [1:16:19<13:48,  2.03it/s]
 84%|████████▍ | 8999/10682 [1:16:20<13:48,  2.03it/s]
 84%|████████▍ | 9000/10682 [1:16:20<13:47,  2.03it/s]
                                                      {'loss': 2.8481, 'grad_norm': 0.2617539167404175, 'learning_rate': 7.365644059145782e-05, 'epoch': 11.8}
+
 84%|████████▍ | 9000/10682 [1:16:20<13:47,  2.03it/s]
 84%|████████▍ | 9001/10682 [1:16:21<13:48,  2.03it/s]
 84%|████████▍ | 9002/10682 [1:16:21<13:47,  2.03it/s]
 84%|████████▍ | 9003/10682 [1:16:22<13:47,  2.03it/s]
 84%|████████▍ | 9004/10682 [1:16:22<13:47,  2.03it/s]
 84%|████████▍ | 9005/10682 [1:16:23<13:46,  2.03it/s]
 84%|████████▍ | 9006/10682 [1:16:23<13:46,  2.03it/s]
 84%|████████▍ | 9007/10682 [1:16:24<13:44,  2.03it/s]
 84%|████████▍ | 9008/10682 [1:16:24<13:44,  2.03it/s]
 84%|████████▍ | 9009/10682 [1:16:24<13:43,  2.03it/s]
 84%|████████▍ | 9010/10682 [1:16:25<13:42,  2.03it/s]
 84%|████████▍ | 9011/10682 [1:16:25<13:42,  2.03it/s]
 84%|████████▍ | 9012/10682 [1:16:26<13:42,  2.03it/s]
 84%|████████▍ | 9013/10682 [1:16:26<13:41,  2.03it/s]
 84%|████████▍ | 9014/10682 [1:16:27<13:41,  2.03it/s]
 84%|████████▍ | 9015/10682 [1:16:27<13:40,  2.03it/s]
 84%|████████▍ | 9016/10682 [1:16:28<13:40,  2.03it/s]
 84%|████████▍ | 9017/10682 [1:16:28<13:39,  2.03it/s]
 84%|████████▍ | 9018/10682 [1:16:29<13:38,  2.03it/s]
 84%|████████▍ | 9019/10682 [1:16:29<13:39,  2.03it/s]
 84%|████████▍ | 9020/10682 [1:16:30<13:38,  2.03it/s]
 84%|████████▍ | 9021/10682 [1:16:30<13:37,  2.03it/s]
 84%|████████▍ | 9022/10682 [1:16:31<13:37,  2.03it/s]
 84%|████████▍ | 9023/10682 [1:16:31<13:36,  2.03it/s]
 84%|████████▍ | 9024/10682 [1:16:32<13:36,  2.03it/s]
 84%|████████▍ | 9025/10682 [1:16:32<13:36,  2.03it/s]{'loss': 2.8558, 'grad_norm': 0.26141008734703064, 'learning_rate': 7.153655656890773e-05, 'epoch': 11.83}
+
                                                      
 84%|████████▍ | 9025/10682 [1:16:32<13:36,  2.03it/s]
 84%|████████▍ | 9026/10682 [1:16:33<13:37,  2.03it/s]
 85%|████████▍ | 9027/10682 [1:16:33<13:36,  2.03it/s]
 85%|████████▍ | 9028/10682 [1:16:34<13:36,  2.02it/s]
 85%|████████▍ | 9029/10682 [1:16:34<13:35,  2.03it/s]
 85%|████████▍ | 9030/10682 [1:16:35<13:34,  2.03it/s]
 85%|████████▍ | 9031/10682 [1:16:35<13:33,  2.03it/s]
 85%|████████▍ | 9032/10682 [1:16:36<13:32,  2.03it/s]
 85%|████████▍ | 9033/10682 [1:16:36<13:32,  2.03it/s]
 85%|████████▍ | 9034/10682 [1:16:37<13:31,  2.03it/s]
 85%|████████▍ | 9035/10682 [1:16:37<13:31,  2.03it/s]
 85%|████████▍ | 9036/10682 [1:16:38<13:32,  2.03it/s]
 85%|████████▍ | 9037/10682 [1:16:38<13:30,  2.03it/s]
 85%|████████▍ | 9038/10682 [1:16:39<13:30,  2.03it/s]
 85%|████████▍ | 9039/10682 [1:16:39<13:29,  2.03it/s]
 85%|████████▍ | 9040/10682 [1:16:40<13:29,  2.03it/s]
 85%|████████▍ | 9041/10682 [1:16:40<13:28,  2.03it/s]
 85%|████████▍ | 9042/10682 [1:16:41<13:27,  2.03it/s]
 85%|████████▍ | 9043/10682 [1:16:41<13:27,  2.03it/s]
 85%|████████▍ | 9044/10682 [1:16:42<13:27,  2.03it/s]
 85%|████████▍ | 9045/10682 [1:16:42<13:26,  2.03it/s]
 85%|████████▍ | 9046/10682 [1:16:43<13:25,  2.03it/s]
 85%|████████▍ | 9047/10682 [1:16:43<13:25,  2.03it/s]
 85%|████████▍ | 9048/10682 [1:16:44<13:26,  2.03it/s]
 85%|████████▍ | 9049/10682 [1:16:44<13:24,  2.03it/s]
 85%|████████▍ | 9050/10682 [1:16:45<13:24,  2.03it/s]{'loss': 2.8601, 'grad_norm': 0.2692682147026062, 'learning_rate': 6.94452730207023e-05, 'epoch': 11.86}
                                                      
+
 85%|████████▍ | 9050/10682 [1:16:45<13:24,  2.03it/s]
 85%|████████▍ | 9051/10682 [1:16:45<13:24,  2.03it/s]
 85%|████████▍ | 9052/10682 [1:16:46<13:24,  2.03it/s]
 85%|████████▍ | 9053/10682 [1:16:46<13:23,  2.03it/s]
 85%|████████▍ | 9054/10682 [1:16:47<13:21,  2.03it/s]
 85%|████████▍ | 9055/10682 [1:16:47<13:22,  2.03it/s]
 85%|████████▍ | 9056/10682 [1:16:48<13:21,  2.03it/s]
 85%|████████▍ | 9057/10682 [1:16:48<13:20,  2.03it/s]
 85%|████████▍ | 9058/10682 [1:16:49<13:20,  2.03it/s]
 85%|████████▍ | 9059/10682 [1:16:49<13:19,  2.03it/s]
 85%|████████▍ | 9060/10682 [1:16:50<13:18,  2.03it/s]
 85%|████████▍ | 9061/10682 [1:16:50<13:19,  2.03it/s]
 85%|████████▍ | 9062/10682 [1:16:51<13:18,  2.03it/s]
 85%|███████��▍ | 9063/10682 [1:16:51<13:17,  2.03it/s]
 85%|████████▍ | 9064/10682 [1:16:52<13:18,  2.03it/s]
 85%|████████▍ | 9065/10682 [1:16:52<13:17,  2.03it/s]
 85%|████████▍ | 9066/10682 [1:16:53<13:17,  2.03it/s]
 85%|████████▍ | 9067/10682 [1:16:53<13:16,  2.03it/s]
 85%|████████▍ | 9068/10682 [1:16:54<13:15,  2.03it/s]
 85%|████████▍ | 9069/10682 [1:16:54<13:15,  2.03it/s]
 85%|████████▍ | 9070/10682 [1:16:55<13:14,  2.03it/s]
 85%|████████▍ | 9071/10682 [1:16:55<13:13,  2.03it/s]
 85%|████████▍ | 9072/10682 [1:16:56<13:14,  2.03it/s]
 85%|████████▍ | 9073/10682 [1:16:56<13:13,  2.03it/s]
 85%|████████▍ | 9074/10682 [1:16:57<13:12,  2.03it/s]
 85%|████████▍ | 9075/10682 [1:16:57<13:11,  2.03it/s]
                                                      {'loss': 2.8627, 'grad_norm': 0.2634871006011963, 'learning_rate': 6.738272954265156e-05, 'epoch': 11.89}
+
 85%|████████▍ | 9075/10682 [1:16:57<13:11,  2.03it/s]
 85%|████████▍ | 9076/10682 [1:16:58<13:12,  2.03it/s]
 85%|████████▍ | 9077/10682 [1:16:58<13:11,  2.03it/s]
 85%|████████▍ | 9078/10682 [1:16:59<13:10,  2.03it/s]
 85%|████████▍ | 9079/10682 [1:16:59<13:09,  2.03it/s]
 85%|████████▌ | 9080/10682 [1:16:59<13:09,  2.03it/s]
 85%|████████▌ | 9081/10682 [1:17:00<13:09,  2.03it/s]
 85%|████████▌ | 9082/10682 [1:17:00<13:07,  2.03it/s]
 85%|████████▌ | 9083/10682 [1:17:01<13:06,  2.03it/s]
 85%|████████▌ | 9084/10682 [1:17:01<13:07,  2.03it/s]
 85%|████████▌ | 9085/10682 [1:17:02<13:05,  2.03it/s]
 85%|████████▌ | 9086/10682 [1:17:02<13:05,  2.03it/s]
 85%|████████▌ | 9087/10682 [1:17:03<13:05,  2.03it/s]
 85%|████████▌ | 9088/10682 [1:17:03<13:04,  2.03it/s]
 85%|████████▌ | 9089/10682 [1:17:04<13:04,  2.03it/s]
 85%|████████▌ | 9090/10682 [1:17:04<13:03,  2.03it/s]
 85%|████████▌ | 9091/10682 [1:17:05<13:03,  2.03it/s]
 85%|████████▌ | 9092/10682 [1:17:05<13:02,  2.03it/s]
 85%|████████▌ | 9093/10682 [1:17:06<13:01,  2.03it/s]
 85%|████████▌ | 9094/10682 [1:17:06<13:02,  2.03it/s]
 85%|████████▌ | 9095/10682 [1:17:07<13:00,  2.03it/s]
 85%|████████▌ | 9096/10682 [1:17:07<13:01,  2.03it/s]
 85%|████████▌ | 9097/10682 [1:17:08<13:00,  2.03it/s]
 85%|████████▌ | 9098/10682 [1:17:08<12:59,  2.03it/s]
 85%|████████▌ | 9099/10682 [1:17:09<12:59,  2.03it/s]
 85%|████████▌ | 9100/10682 [1:17:09<12:58,  2.03it/s]{'loss': 2.8535, 'grad_norm': 0.26476970314979553, 'learning_rate': 6.534906381212979e-05, 'epoch': 11.93}
+
                                                      
 85%|████████▌ | 9100/10682 [1:17:09<12:58,  2.03it/s]
 85%|████████▌ | 9101/10682 [1:17:10<13:00,  2.03it/s]
 85%|████████▌ | 9102/10682 [1:17:10<12:58,  2.03it/s]
 85%|████████▌ | 9103/10682 [1:17:11<12:58,  2.03it/s]
 85%|████████▌ | 9104/10682 [1:17:11<12:57,  2.03it/s]
 85%|████████▌ | 9105/10682 [1:17:12<12:56,  2.03it/s]
 85%|████████▌ | 9106/10682 [1:17:12<12:56,  2.03it/s]
 85%|████████▌ | 9107/10682 [1:17:13<12:55,  2.03it/s]
 85%|████████▌ | 9108/10682 [1:17:13<12:54,  2.03it/s]
 85%|████████▌ | 9109/10682 [1:17:14<12:54,  2.03it/s]
 85%|████████▌ | 9110/10682 [1:17:14<12:54,  2.03it/s]
 85%|████████▌ | 9111/10682 [1:17:15<12:53,  2.03it/s]
 85%|████████▌ | 9112/10682 [1:17:15<12:53,  2.03it/s]
 85%|████████▌ | 9113/10682 [1:17:16<12:52,  2.03it/s]
 85%|████████▌ | 9114/10682 [1:17:16<12:52,  2.03it/s]
 85%|████████▌ | 9115/10682 [1:17:17<12:52,  2.03it/s]
 85%|████████▌ | 9116/10682 [1:17:17<12:52,  2.03it/s]
 85%|████████▌ | 9117/10682 [1:17:18<12:51,  2.03it/s]
 85%|████████▌ | 9118/10682 [1:17:18<12:51,  2.03it/s]
 85%|████████▌ | 9119/10682 [1:17:19<12:50,  2.03it/s]
 85%|████████▌ | 9120/10682 [1:17:19<12:49,  2.03it/s]
 85%|████████▌ | 9121/10682 [1:17:20<12:49,  2.03it/s]
 85%|████████▌ | 9122/10682 [1:17:20<12:48,  2.03it/s]
 85%|████████▌ | 9123/10682 [1:17:21<12:48,  2.03it/s]
 85%|████████▌ | 9124/10682 [1:17:21<12:47,  2.03it/s]
 85%|████████▌ | 9125/10682 [1:17:22<12:47,  2.03it/s]{'loss': 2.8536, 'grad_norm': 0.26291805505752563, 'learning_rate': 6.334441157888504e-05, 'epoch': 11.96}
+
                                                      
 85%|████████▌ | 9125/10682 [1:17:22<12:47,  2.03it/s]
 85%|████████▌ | 9126/10682 [1:17:22<12:47,  2.03it/s]
 85%|████████▌ | 9127/10682 [1:17:23<12:46,  2.03it/s]
 85%|████████▌ | 9128/10682 [1:17:23<12:45,  2.03it/s]
 85%|████████▌ | 9129/10682 [1:17:24<12:45,  2.03it/s]
 85%|████████▌ | 9130/10682 [1:17:24<12:44,  2.03it/s]
 85%|████████▌ | 9131/10682 [1:17:25<12:44,  2.03it/s]
 85%|████████▌ | 9132/10682 [1:17:25<12:43,  2.03it/s]
 85%|████████▌ | 9133/10682 [1:17:26<12:43,  2.03it/s]
 86%|████████▌ | 9134/10682 [1:17:26<12:42,  2.03it/s]
 86%|████████▌ | 9135/10682 [1:17:27<12:41,  2.03it/s]
 86%|████████▌ | 9136/10682 [1:17:27<12:41,  2.03it/s]
 86%|████████▌ | 9137/10682 [1:17:28<12:40,  2.03it/s]
 86%|████████▌ | 9138/10682 [1:17:28<12:40,  2.03it/s]
 86%|████████▌ | 9139/10682 [1:17:29<12:40,  2.03it/s]
 86%|████████▌ | 9140/10682 [1:17:29<12:38,  2.03it/s]
 86%|████████▌ | 9141/10682 [1:17:30<12:38,  2.03it/s]
 86%|████████▌ | 9142/10682 [1:17:30<12:38,  2.03it/s]
 86%|████████▌ | 9143/10682 [1:17:31<12:37,  2.03it/s]
 86%|████████▌ | 9144/10682 [1:17:31<12:37,  2.03it/s]
 86%|████████▌ | 9145/10682 [1:17:32<12:36,  2.03it/s]
 86%|████████▌ | 9146/10682 [1:17:32<12:35,  2.03it/s]
 86%|████████▌ | 9147/10682 [1:17:32<12:36,  2.03it/s]
 86%|████████▌ | 9148/10682 [1:17:33<12:35,  2.03it/s]
 86%|████████▌ | 9149/10682 [1:17:33<12:35,  2.03it/s]
 86%|████████▌ | 9150/10682 [1:17:34<12:34,  2.03it/s]{'loss': 2.8583, 'grad_norm': 0.2615031898021698, 'learning_rate': 6.1368906655978e-05, 'epoch': 11.99}
                                                      
+
 86%|████████▌ | 9150/10682 [1:17:34<12:34,  2.03it/s]
 86%|████████▌ | 9151/10682 [1:17:34<12:35,  2.03it/s]
 86%|████████▌ | 9152/10682 [1:17:35<12:34,  2.03it/s]
 86%|████████▌ | 9153/10682 [1:17:35<12:33,  2.03it/s]
 86%|████████▌ | 9154/10682 [1:17:36<12:33,  2.03it/s]
 86%|████████▌ | 9155/10682 [1:17:36<12:32,  2.03it/s]
 86%|████████▌ | 9156/10682 [1:17:37<12:25,  2.05it/s]
 86%|████████▌ | 9157/10682 [1:17:52<2:01:47,  4.79s/it]
 86%|████████▌ | 9158/10682 [1:17:52<1:29:01,  3.50s/it]
 86%|████████▌ | 9159/10682 [1:17:53<1:06:01,  2.60s/it]
 86%|████████▌ | 9160/10682 [1:17:53<50:00,  1.97s/it]  
 86%|████████▌ | 9161/10682 [1:17:54<38:43,  1.53s/it]
 86%|████████▌ | 9162/10682 [1:17:54<30:50,  1.22s/it]
 86%|████████▌ | 9163/10682 [1:17:55<25:20,  1.00s/it]
 86%|████████▌ | 9164/10682 [1:17:55<21:28,  1.18it/s]
 86%|████████▌ | 9165/10682 [1:17:56<18:44,  1.35it/s]
 86%|████████▌ | 9166/10682 [1:17:56<16:50,  1.50it/s]
 86%|████████▌ | 9167/10682 [1:17:57<15:32,  1.63it/s]
 86%|████████▌ | 9168/10682 [1:17:57<14:35,  1.73it/s]
 86%|████████▌ | 9169/10682 [1:17:58<13:55,  1.81it/s]
 86%|████████▌ | 9170/10682 [1:17:58<13:28,  1.87it/s]
 86%|████████▌ | 9171/10682 [1:17:59<13:07,  1.92it/s]
 86%|████████▌ | 9172/10682 [1:17:59<12:54,  1.95it/s]
 86%|████████▌ | 9173/10682 [1:18:00<12:45,  1.97it/s]
 86%|████████▌ | 9174/10682 [1:18:00<12:38,  1.99it/s]
 86%|████████▌ | 9175/10682 [1:18:01<12:32,  2.00it/s]
                                                      
+
 86%|████████▌ | 9175/10682 [1:18:01<12:32,  2.00it/s]
 86%|████████▌ | 9176/10682 [1:18:01<12:30,  2.01it/s]
 86%|████████▌ | 9177/10682 [1:18:02<12:26,  2.01it/s]
 86%|████████▌ | 9178/10682 [1:18:02<12:24,  2.02it/s]
 86%|████████▌ | 9179/10682 [1:18:03<12:23,  2.02it/s]
 86%|████████▌ | 9180/10682 [1:18:03<12:22,  2.02it/s]
 86%|████████▌ | 9181/10682 [1:18:04<12:21,  2.02it/s]
 86%|████████▌ | 9182/10682 [1:18:04<12:20,  2.03it/s]
 86%|████████▌ | 9183/10682 [1:18:05<12:19,  2.03it/s]
 86%|████████▌ | 9184/10682 [1:18:05<12:19,  2.03it/s]
 86%|████████▌ | 9185/10682 [1:18:06<12:19,  2.03it/s]
 86%|████████▌ | 9186/10682 [1:18:06<12:17,  2.03it/s]
 86%|████████▌ | 9187/10682 [1:18:07<12:15,  2.03it/s]
 86%|████████▌ | 9188/10682 [1:18:07<12:15,  2.03it/s]