Training in progress, epoch 14

logs/events.out.tfevents.1715596086.sphinx2

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7e2ee8a7878e5040a6fe71b3ea54a494f4442e0c264dd37ce7e14671efb00e5e
-size 88382
+oid sha256:96af6e2a9ada45d0e4d122751106ce3acf6f8c8c062f1ad9a8930098eff5f28f
+size 95277

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0047ab60cb5e5abaa5942e3102cc692032fabd895b81c4d00013a78d31160163
+oid sha256:0f416e698c8b1f33c7778f00de4881492b2b40f0637a9d167917e1a5a2d5fe0e
 size 281715176

train_job_output.txt

@@ -502,4 +502,35 @@ command outputs:
 
                                                      
 92%|█████████▏| 9875/10682 [1:33:52<06:46,  1.99it/s]
 92%|█████████▏| 9876/10682 [1:33:52<06:45,  1.99it/s]
 92%|█████████▏| 9877/10682 [1:33:53<06:43,  1.99it/s]
 92%|█████████▏| 9878/10682 [1:33:53<06:43,  1.99it/s]
 92%|█████████▏| 9879/10682 [1:33:54<06:43,  1.99it/s]
 92%|█████████▏| 9880/10682 [1:33:54<06:43,  1.99it/s]
 93%|█████████▎| 9881/10682 [1:33:55<06:43,  1.99it/s]
 93%|█████████▎| 9882/10682 [1:33:55<06:42,  1.99it/s]
 93%|█████████▎| 9883/10682 [1:33:56<06:41,  1.99it/s]
 93%|█████████▎| 9884/10682 [1:33:56<06:40,  1.99it/s]
 93%|█████████▎| 9885/10682 [1:33:57<06:40,  1.99it/s]
 93%|█████████▎| 9886/10682 [1:33:57<06:39,  1.99it/s]
 93%|█████████▎| 9887/10682 [1:33:58<06:39,  1.99it/s]
 93%|█████████▎| 9888/10682 [1:33:58<06:39,  1.99it/s]
 93%|█████████▎| 9889/10682 [1:33:59<06:38,  1.99it/s]
 93%|█████████▎| 9890/10682 [1:33:59<06:38,  1.99it/s]
 93%|█████████▎| 9891/10682 [1:34:00<06:37,  1.99it/s]
 93%|█████████▎| 9892/10682 [1:34:00<06:36,  1.99it/s]
 93%|█████████▎| 9893/10682 [1:34:01<06:36,  1.99it/s]
 93%|█████████▎| 9894/10682 [1:34:01<06:36,  1.99it/s]
 93%|█████████▎| 9895/10682 [1:34:02<06:35,  1.99it/s]
 93%|█████████▎| 9896/10682 [1:34:02<06:34,  1.99it/s]
 93%|█████████▎| 9897/10682 [1:34:03<06:34,  1.99it/s]
 93%|█████████▎| 9898/10682 [1:34:03<06:34,  1.99it/s]
 93%|█████████▎| 9899/10682 [1:34:04<06:33,  1.99it/s]
 93%|█████████▎| 9900/10682 [1:34:04<06:32,  1.99it/s]{'loss': 3.1411, 'grad_norm': 0.2500966787338257, 'learning_rate': 1.6239414036870183e-05, 'epoch': 12.98}
                                                      
 
 93%|█████████▎| 9900/10682 [1:34:04<06:32,  1.99it/s]
 93%|█████████▎| 9901/10682 [1:34:05<06:32,  1.99it/s]
 93%|█████████▎| 9902/10682 [1:34:05<06:31,  1.99it/s]
 93%|█████████▎| 9903/10682 [1:34:06<06:31,  1.99it/s]
 93%|█████████▎| 9904/10682 [1:34:06<06:31,  1.99it/s]
 93%|█████████▎| 9905/10682 [1:34:07<06:29,  1.99it/s]
 93%|█████████▎| 9906/10682 [1:34:07<06:28,  2.00it/s]
 93%|█████████▎| 9907/10682 [1:34:08<06:28,  1.99it/s]
 93%|█████████▎| 9908/10682 [1:34:08<06:28,  1.99it/s]
 93%|█████████▎| 9909/10682 [1:34:09<06:27,  1.99it/s]
 93%|█████████▎| 9910/10682 [1:34:09<06:27,  1.99it/s]
 93%|█████████▎| 9911/10682 [1:34:10<06:26,  2.00it/s]
 93%|█████████▎| 9912/10682 [1:34:10<06:25,  2.00it/s]
 93%|█████████▎| 9913/10682 [1:34:11<06:25,  1.99it/s]
 93%|█████████▎| 9914/10682 [1:34:11<06:25,  1.99it/s]
 93%|█████████▎| 9915/10682 [1:34:12<06:24,  2.00it/s]
 93%|█████████▎| 9916/10682 [1:34:12<06:23,  2.00it/s]
 93%|█████████▎| 9917/10682 [1:34:13<06:22,  2.00it/s]
 93%|█████████▎| 9918/10682 [1:34:13<06:22,  2.00it/s]
 93%|█████████▎| 9919/10682 [1:34:14<06:17,  2.02it/s]
 93%|█████████▎| 9920/10682 [1:35:44<5:47:38, 27.37s/it]
 93%|█████████▎| 9921/10682 [1:35:44<4:04:56, 19.31s/it]
 93%|█████████▎| 9922/10682 [1:35:45<2:53:07, 13.67s/it]
 93%|█████████▎| 9923/10682 [1:35:45<2:02:58,  9.72s/it]
 93%|█████████▎| 9924/10682 [1:35:46<1:27:51,  6.96s/it]
 93%|█████████▎| 9925/10682 [1:35:46<1:03:19,  5.02s/it]{'loss': 3.1493, 'grad_norm': 0.24704374372959137, 'learning_rate': 1.5222903086944684e-05, 'epoch': 13.01}
                                                        
 
 93%|█████████▎| 9925/10682 [1:35:46<1:03:19,  5.02s/it]
 93%|█████████▎| 9926/10682 [1:35:47<46:09,  3.66s/it]  
 93%|█████████▎| 9927/10682 [1:35:47<34:09,  2.71s/it]
 93%|█████████▎| 9928/10682 [1:35:48<25:46,  2.05s/it]
 93%|█████████▎| 9929/10682 [1:35:48<19:53,  1.59s/it]
 93%|█████████▎| 9930/10682 [1:35:49<15:47,  1.26s/it]
 93%|█████████▎| 9931/10682 [1:35:49<12:55,  1.03s/it]
 93%|█████████▎| 9932/10682 [1:35:50<10:54,  1.15it/s]
 93%|█████████▎| 9933/10682 [1:35:50<09:30,  1.31it/s]
 93%|█████████▎| 9934/10682 [1:35:51<08:30,  1.47it/s]
 93%|███████���█▎| 9935/10682 [1:35:51<07:49,  1.59it/s]
 93%|█████████▎| 9936/10682 [1:35:52<07:19,  1.70it/s]
 93%|█████████▎| 9937/10682 [1:35:52<06:58,  1.78it/s]
 93%|█████████▎| 9938/10682 [1:35:53<06:44,  1.84it/s]
 93%|█████████▎| 9939/10682 [1:35:53<06:34,  1.88it/s]
 93%|█████████▎| 9940/10682 [1:35:54<06:27,  1.92it/s]
 93%|█████████▎| 9941/10682 [1:35:54<06:21,  1.94it/s]
 93%|█████████▎| 9942/10682 [1:35:55<06:17,  1.96it/s]
 93%|█████████▎| 9943/10682 [1:35:55<06:15,  1.97it/s]
 93%|█████████▎| 9944/10682 [1:35:56<06:13,  1.98it/s]
 93%|█████████▎| 9945/10682 [1:35:56<06:11,  1.98it/s]
 93%|█████████▎| 9946/10682 [1:35:57<06:09,  1.99it/s]
 93%|█████████▎| 9947/10682 [1:35:57<06:09,  1.99it/s]
 93%|█████████▎| 9948/10682 [1:35:58<06:08,  1.99it/s]
 93%|█████████▎| 9949/10682 [1:35:58<06:07,  2.00it/s]
 93%|█████████▎| 9950/10682 [1:35:59<06:06,  2.00it/s]{'loss': 3.1011, 'grad_norm': 0.24837680160999298, 'learning_rate': 1.4238751618640577e-05, 'epoch': 13.04}
                                                      
-
 93%|█████████▎| 9950/10682 [1:35:59<06:06,  2.00it/s]
 93%|█████████▎| 9951/10682 [1:35:59<06:06,  2.00it/s]
 93%|█████████▎| 9952/10682 [1:36:00<06:05,  2.00it/s]
 93%|█████████▎| 9953/10682 [1:36:00<06:05,  2.00it/s]
 93%|█████████▎| 9954/10682 [1:36:01<06:04,  2.00it/s]
 93%|█████████▎| 9955/10682 [1:36:01<06:03,  2.00it/s]
 93%|█████████▎| 9956/10682 [1:36:02<06:03,  2.00it/s]
+
 93%|█████████▎| 9950/10682 [1:35:59<06:06,  2.00it/s]
 93%|█████████▎| 9951/10682 [1:35:59<06:06,  2.00it/s]
 93%|█████████▎| 9952/10682 [1:36:00<06:05,  2.00it/s]
 93%|█████████▎| 9953/10682 [1:36:00<06:05,  2.00it/s]
 93%|█████████▎| 9954/10682 [1:36:01<06:04,  2.00it/s]
 93%|█████████▎| 9955/10682 [1:36:01<06:03,  2.00it/s]
 93%|█████████▎| 9956/10682 [1:36:02<06:03,  2.00it/s]
 93%|█████████▎| 9957/10682 [1:36:02<06:03,  1.99it/s]
 93%|█████████▎| 9958/10682 [1:36:03<06:02,  2.00it/s]
 93%|█████████▎| 9959/10682 [1:36:03<06:01,  2.00it/s]
 93%|█████████▎| 9960/10682 [1:36:04<06:01,  2.00it/s]
 93%|█████████▎| 9961/10682 [1:36:04<06:01,  2.00it/s]
 93%|█████████▎| 9962/10682 [1:36:05<06:00,  2.00it/s]
 93%|█████████▎| 9963/10682 [1:36:05<05:59,  2.00it/s]
 93%|█████████▎| 9964/10682 [1:36:06<05:59,  2.00it/s]
 93%|█████████▎| 9965/10682 [1:36:06<05:58,  2.00it/s]
 93%|█████████▎| 9966/10682 [1:36:07<05:58,  2.00it/s]
 93%|█████████▎| 9967/10682 [1:36:07<05:57,  2.00it/s]
 93%|█████████▎| 9968/10682 [1:36:08<05:57,  2.00it/s]
 93%|█████████▎| 9969/10682 [1:36:08<05:56,  2.00it/s]
 93%|█████████▎| 9970/10682 [1:36:09<05:56,  2.00it/s]
 93%|█████████▎| 9971/10682 [1:36:09<05:56,  2.00it/s]
 93%|█████████▎| 9972/10682 [1:36:10<05:55,  2.00it/s]
 93%|█████████▎| 9973/10682 [1:36:10<05:54,  2.00it/s]
 93%|█████████▎| 9974/10682 [1:36:11<05:54,  2.00it/s]
 93%|█████████▎| 9975/10682 [1:36:11<05:53,  2.00it/s]
                                                      {'loss': 3.1115, 'grad_norm': 0.24749550223350525, 'learning_rate': 1.3287025325307511e-05, 'epoch': 13.07}
+
 93%|█████████▎| 9975/10682 [1:36:11<05:53,  2.00it/s]
 93%|█████████▎| 9976/10682 [1:36:12<05:53,  2.00it/s]
 93%|█████████▎| 9977/10682 [1:36:12<05:52,  2.00it/s]
 93%|█████████▎| 9978/10682 [1:36:13<05:52,  2.00it/s]
 93%|█████████▎| 9979/10682 [1:36:13<05:51,  2.00it/s]
 93%|█████████▎| 9980/10682 [1:36:14<05:51,  2.00it/s]
 93%|█████████▎| 9981/10682 [1:36:14<05:51,  2.00it/s]
 93%|█████████▎| 9982/10682 [1:36:15<05:50,  2.00it/s]
 93%|█████████▎| 9983/10682 [1:36:15<05:50,  2.00it/s]
 93%|█████████▎| 9984/10682 [1:36:16<05:49,  2.00it/s]
 93%|█████████▎| 9985/10682 [1:36:16<05:48,  2.00it/s]
 93%|█████████▎| 9986/10682 [1:36:17<05:48,  2.00it/s]
 93%|█████████▎| 9987/10682 [1:36:17<05:48,  2.00it/s]
 94%|█████████▎| 9988/10682 [1:36:18<05:48,  1.99it/s]
 94%|█████████▎| 9989/10682 [1:36:18<05:47,  1.99it/s]
 94%|█████████▎| 9990/10682 [1:36:19<05:46,  1.99it/s]
 94%|█████████▎| 9991/10682 [1:36:19<05:45,  2.00it/s]
 94%|█████████▎| 9992/10682 [1:36:20<05:45,  2.00it/s]
 94%|█████████▎| 9993/10682 [1:36:20<05:45,  2.00it/s]
 94%|█████████▎| 9994/10682 [1:36:21<05:45,  1.99it/s]
 94%|█████████▎| 9995/10682 [1:36:21<05:45,  1.99it/s]
 94%|█████████▎| 9996/10682 [1:36:22<05:44,  1.99it/s]
 94%|█████████▎| 9997/10682 [1:36:22<05:43,  2.00it/s]
 94%|█████████▎| 9998/10682 [1:36:23<05:42,  2.00it/s]
 94%|█████████▎| 9999/10682 [1:36:23<05:42,  1.99it/s]
 94%|█████████▎| 10000/10682 [1:36:24<05:42,  1.99it/s]
                                                       
+
 94%|█████████▎| 10000/10682 [1:36:24<05:42,  1.99it/s]
 94%|█████████▎| 10001/10682 [1:36:24<05:41,  1.99it/s]
 94%|█████████▎| 10002/10682 [1:36:25<05:40,  2.00it/s]
 94%|█████████▎| 10003/10682 [1:36:25<05:39,  2.00it/s]
 94%|█████████▎| 10004/10682 [1:36:26<05:40,  1.99it/s]
 94%|█████████▎| 10005/10682 [1:36:26<05:39,  1.99it/s]
 94%|█████████▎| 10006/10682 [1:36:27<05:39,  1.99it/s]
 94%|█████████▎| 10007/10682 [1:36:27<05:38,  2.00it/s]
 94%|█████████▎| 10008/10682 [1:36:28<05:37,  2.00it/s]
 94%|█████████▎| 10009/10682 [1:36:28<05:36,  2.00it/s]
 94%|█████████▎| 10010/10682 [1:36:29<05:36,  1.99it/s]
 94%|█████████▎| 10011/10682 [1:36:29<05:36,  2.00it/s]
 94%|█████████▎| 10012/10682 [1:36:30<05:35,  2.00it/s]
 94%|█████████▎| 10013/10682 [1:36:30<05:34,  2.00it/s]
 94%|█████████▎| 10014/10682 [1:36:31<05:34,  2.00it/s]
 94%|█████████▍| 10015/10682 [1:36:31<05:34,  1.99it/s]
 94%|█████████▍| 10016/10682 [1:36:32<05:33,  1.99it/s]
 94%|█████████▍| 10017/10682 [1:36:32<05:32,  2.00it/s]
 94%|█████████▍| 10018/10682 [1:36:33<05:32,  2.00it/s]
 94%|█████████▍| 10019/10682 [1:36:33<05:32,  2.00it/s]
 94%|█████████▍| 10020/10682 [1:36:34<05:31,  1.99it/s]
 94%|█████████▍| 10021/10682 [1:36:34<05:31,  1.99it/s]
 94%|█████████▍| 10022/10682 [1:36:35<05:31,  1.99it/s]
 94%|█████████▍| 10023/10682 [1:36:35<05:30,  1.99it/s]
 94%|█████████▍| 10024/10682 [1:36:36<05:29,  2.00it/s]
 94%|█████████▍| 10025/10682 [1:36:36<05:29,  2.00it/s]
                                                       {'loss': 3.1075, 'grad_norm': 0.24997055530548096, 'learning_rate': 1.1481100210606388e-05, 'epoch': 13.14}
+
 94%|█████████▍| 10025/10682 [1:36:36<05:29,  2.00it/s]
 94%|█████████▍| 10026/10682 [1:36:37<05:29,  1.99it/s]
 94%|█████████▍| 10027/10682 [1:36:37<05:28,  1.99it/s]
 94%|█████████▍| 10028/10682 [1:36:38<05:27,  1.99it/s]
 94%|█████████▍| 10029/10682 [1:36:38<05:26,  2.00it/s]
 94%|█████████▍| 10030/10682 [1:36:39<05:26,  2.00it/s]
 94%|█████████▍| 10031/10682 [1:36:39<05:26,  2.00it/s]
 94%|█████████▍| 10032/10682 [1:36:40<05:25,  1.99it/s]
 94%|█████████▍| 10033/10682 [1:36:40<05:25,  2.00it/s]
 94%|█████████▍| 10034/10682 [1:36:41<05:24,  1.99it/s]
 94%|█████████▍| 10035/10682 [1:36:41<05:24,  1.99it/s]
 94%|█████████▍| 10036/10682 [1:36:42<05:23,  1.99it/s]
 94%|█████████▍| 10037/10682 [1:36:42<05:23,  2.00it/s]
 94%|█████████▍| 10038/10682 [1:36:43<05:22,  2.00it/s]
 94%|█████████▍| 10039/10682 [1:36:43<05:22,  2.00it/s]
 94%|█████████▍| 10040/10682 [1:36:44<05:21,  2.00it/s]
 94%|█████████▍| 10041/10682 [1:36:44<05:20,  2.00it/s]
 94%|█████████▍| 10042/10682 [1:36:45<05:20,  2.00it/s]
 94%|█████████▍| 10043/10682 [1:36:45<05:19,  2.00it/s]
 94%|█████████▍| 10044/10682 [1:36:46<05:19,  2.00it/s]
 94%|█████████▍| 10045/10682 [1:36:46<05:18,  2.00it/s]
 94%|█████████▍| 10046/10682 [1:36:47<05:18,  2.00it/s]
 94%|█████████▍| 10047/10682 [1:36:47<05:17,  2.00it/s]
 94%|█████████▍| 10048/10682 [1:36:48<05:17,  2.00it/s]
 94%|█████████▍| 10049/10682 [1:36:48<05:16,  2.00it/s]
 94%|█████████▍| 10050/10682 [1:36:49<05:16,  2.00it/s]
                                                       
+
 94%|█████████▍| 10050/10682 [1:36:49<05:16,  2.00it/s]
 94%|█████████▍| 10051/10682 [1:36:49<05:16,  2.00it/s]
 94%|█████████▍| 10052/10682 [1:36:50<05:15,  2.00it/s]
 94%|█████████▍| 10053/10682 [1:36:50<05:15,  2.00it/s]
 94%|█████████▍| 10054/10682 [1:36:51<05:14,  1.99it/s]
 94%|█████████▍| 10055/10682 [1:36:51<05:14,  1.99it/s]
 94%|█████████▍| 10056/10682 [1:36:52<05:13,  2.00it/s]
 94%|█████████▍| 10057/10682 [1:36:52<05:12,  2.00it/s]
 94%|█████████▍| 10058/10682 [1:36:53<05:13,  1.99it/s]
 94%|█████████▍| 10059/10682 [1:36:53<05:12,  1.99it/s]
 94%|█████████▍| 10060/10682 [1:36:54<05:11,  1.99it/s]
 94%|█████████▍| 10061/10682 [1:36:54<05:10,  2.00it/s]
 94%|█████████▍| 10062/10682 [1:36:55<05:10,  2.00it/s]
 94%|█████████▍| 10063/10682 [1:36:55<05:10,  2.00it/s]
 94%|█████████▍| 10064/10682 [1:36:56<05:10,  1.99it/s]
 94%|█████████▍| 10065/10682 [1:36:56<05:09,  1.99it/s]
 94%|█████████▍| 10066/10682 [1:36:57<05:09,  1.99it/s]
 94%|█████████▍| 10067/10682 [1:36:57<05:08,  1.99it/s]
 94%|█████████▍| 10068/10682 [1:36:58<05:07,  1.99it/s]
 94%|█████████▍| 10069/10682 [1:36:58<05:07,  1.99it/s]
 94%|█████████▍| 10070/10682 [1:36:59<05:06,  1.99it/s]
 94%|█████████▍| 10071/10682 [1:36:59<05:06,  1.99it/s]
 94%|█████████▍| 10072/10682 [1:37:00<05:05,  1.99it/s]
 94%|█████████▍| 10073/10682 [1:37:00<05:05,  2.00it/s]
 94%|█████████▍| 10074/10682 [1:37:01<05:04,  2.00it/s]
 94%|█████████▍| 10075/10682 [1:37:01<05:04,  1.99it/s]
                                                       
+
 94%|█████████▍| 10075/10682 [1:37:01<05:04,  1.99it/s]
 94%|█████████▍| 10076/10682 [1:37:02<05:04,  1.99it/s]
 94%|█████████▍| 10077/10682 [1:37:02<05:03,  1.99it/s]
 94%|█████████▍| 10078/10682 [1:37:03<05:02,  1.99it/s]
 94%|█████████▍| 10079/10682 [1:37:03<05:02,  2.00it/s]
 94%|█████████▍| 10080/10682 [1:37:04<05:01,  1.99it/s]
 94%|█████████▍| 10081/10682 [1:37:04<05:01,  1.99it/s]
 94%|█████████▍| 10082/10682 [1:37:05<05:01,  1.99it/s]
 94%|█████████▍| 10083/10682 [1:37:05<05:00,  1.99it/s]
 94%|█████████▍| 10084/10682 [1:37:06<04:59,  1.99it/s]
 94%|█████████▍| 10085/10682 [1:37:06<04:59,  2.00it/s]
 94%|█████████▍| 10086/10682 [1:37:07<04:58,  1.99it/s]
 94%|█████████▍| 10087/10682 [1:37:07<04:58,  1.99it/s]
 94%|█████████▍| 10088/10682 [1:37:08<04:57,  1.99it/s]
 94%|█████████▍| 10089/10682 [1:37:08<04:57,  2.00it/s]
 94%|█████████▍| 10090/10682 [1:37:09<04:56,  2.00it/s]
 94%|█████████▍| 10091/10682 [1:37:09<04:56,  2.00it/s]
 94%|█████████▍| 10092/10682 [1:37:10<04:55,  2.00it/s]
 94%|█████████▍| 10093/10682 [1:37:10<04:54,  2.00it/s]
 94%|█████████▍| 10094/10682 [1:37:11<04:54,  2.00it/s]
 95%|█████████▍| 10095/10682 [1:37:11<04:54,  1.99it/s]
 95%|█████████▍| 10096/10682 [1:37:12<04:53,  2.00it/s]
 95%|█████████▍| 10097/10682 [1:37:12<04:52,  2.00it/s]
 95%|█████████▍| 10098/10682 [1:37:13<04:52,  2.00it/s]
 95%|█████████▍| 10099/10682 [1:37:13<04:51,  2.00it/s]
 95%|█████████▍| 10100/10682 [1:37:14<04:51,  2.00it/s]
                                                       
+
 95%|█████████▍| 10100/10682 [1:37:14<04:51,  2.00it/s]
 95%|█████████▍| 10101/10682 [1:37:15<05:15,  1.84it/s]
 95%|█████████▍| 10102/10682 [1:37:15<05:08,  1.88it/s]
 95%|█████████▍| 10103/10682 [1:37:16<05:02,  1.92it/s]
 95%|█████████▍| 10104/10682 [1:37:16<04:57,  1.94it/s]
 95%|█████████▍| 10105/10682 [1:37:17<04:55,  1.96it/s]
 95%|█████████▍| 10106/10682 [1:37:17<04:53,  1.97it/s]
 95%|█████████▍| 10107/10682 [1:37:18<04:51,  1.97it/s]
 95%|█████████▍| 10108/10682 [1:37:18<04:49,  1.98it/s]
 95%|█████████▍| 10109/10682 [1:37:19<04:48,  1.99it/s]
 95%|█████████▍| 10110/10682 [1:37:19<04:47,  1.99it/s]
 95%|█████████▍| 10111/10682 [1:37:20<04:46,  1.99it/s]
 95%|█████████▍| 10112/10682 [1:37:20<04:45,  1.99it/s]
 95%|█████████▍| 10113/10682 [1:37:21<04:45,  1.99it/s]
 95%|█████████▍| 10114/10682 [1:37:21<04:44,  2.00it/s]
 95%|█████████▍| 10115/10682 [1:37:22<04:44,  2.00it/s]
 95%|█████████▍| 10116/10682 [1:37:22<04:43,  1.99it/s]
 95%|█████████▍| 10117/10682 [1:37:23<04:43,  1.99it/s]
 95%|█████████▍| 10118/10682 [1:37:23<04:42,  2.00it/s]
 95%|█████████▍| 10119/10682 [1:37:24<04:42,  2.00it/s]
 95%|█████████▍| 10120/10682 [1:37:24<04:41,  2.00it/s]
 95%|█████████▍| 10121/10682 [1:37:25<04:41,  1.99it/s]
 95%|█████████▍| 10122/10682 [1:37:25<04:41,  1.99it/s]
 95%|█████████▍| 10123/10682 [1:37:26<04:40,  1.99it/s]
 95%|█████████▍| 10124/10682 [1:37:26<04:39,  2.00it/s]
 95%|█████████▍| 10125/10682 [1:37:27<04:39,  2.00it/s]{'loss': 3.1083, 'grad_norm': 0.24928326904773712, 'learning_rate': 8.261001828055447e-06, 'epoch': 13.27}
                                                       
+
 95%|█████████▍| 10125/10682 [1:37:27<04:39,  2.00it/s]
 95%|█████████▍| 10126/10682 [1:37:27<04:38,  1.99it/s]
 95%|█████████▍| 10127/10682 [1:37:28<04:38,  1.99it/s]
 95%|█████████▍| 10128/10682 [1:37:28<04:37,  2.00it/s]
 95%|█████████▍| 10129/10682 [1:37:29<04:36,  2.00it/s]
 95%|█████████▍| 10130/10682 [1:37:29<04:36,  2.00it/s]
 95%|█████████▍| 10131/10682 [1:37:30<04:36,  1.99it/s]
 95%|█████████▍| 10132/10682 [1:37:30<04:36,  1.99it/s]
 95%|█████████▍| 10133/10682 [1:37:31<04:35,  1.99it/s]
 95%|█████████▍| 10134/10682 [1:37:31<04:34,  1.99it/s]
 95%|█████████▍| 10135/10682 [1:37:32<04:33,  2.00it/s]
 95%|█████████▍| 10136/10682 [1:37:32<04:33,  2.00it/s]
 95%|█████████▍| 10137/10682 [1:37:33<04:32,  2.00it/s]
 95%|█████████▍| 10138/10682 [1:37:33<04:32,  1.99it/s]
 95%|█████████▍| 10139/10682 [1:37:34<04:32,  1.99it/s]
 95%|█████████▍| 10140/10682 [1:37:34<04:32,  1.99it/s]
 95%|█████████▍| 10141/10682 [1:37:35<04:31,  2.00it/s]
 95%|█████████▍| 10142/10682 [1:37:35<04:30,  2.00it/s]
 95%|█████████▍| 10143/10682 [1:37:36<04:30,  1.99it/s]
 95%|█████████▍| 10144/10682 [1:37:36<04:29,  1.99it/s]
 95%|█████████▍| 10145/10682 [1:37:37<04:29,  1.99it/s]
 95%|█████████▍| 10146/10682 [1:37:37<04:28,  1.99it/s]
 95%|█████████▍| 10147/10682 [1:37:38<04:27,  2.00it/s]
 95%|█████████▌| 10148/10682 [1:37:38<04:27,  2.00it/s]
 95%|█████████▌| 10149/10682 [1:37:39<04:27,  1.99it/s]
 95%|█████████▌| 10150/10682 [1:37:39<04:27,  1.99it/s]{'loss': 3.1143, 'grad_norm': 0.2494010180234909, 'learning_rate': 7.537908845868024e-06, 'epoch': 13.3}
+
                                                       
 95%|█████████▌| 10150/10682 [1:37:39<04:27,  1.99it/s]
 95%|█████████▌| 10151/10682 [1:37:40<04:26,  1.99it/s]
 95%|█████████▌| 10152/10682 [1:37:40<04:25,  1.99it/s]
 95%|█████████▌| 10153/10682 [1:37:41<04:25,  1.99it/s]
 95%|█████████▌| 10154/10682 [1:37:41<04:24,  1.99it/s]
 95%|█████████▌| 10155/10682 [1:37:42<04:24,  1.99it/s]
 95%|█████████▌| 10156/10682 [1:37:42<04:23,  1.99it/s]
 95%|█████████▌| 10157/10682 [1:37:43<04:23,  1.99it/s]
 95%|█████████▌| 10158/10682 [1:37:43<04:22,  2.00it/s]
 95%|█████████▌| 10159/10682 [1:37:44<04:22,  2.00it/s]
 95%|█████████▌| 10160/10682 [1:37:44<04:21,  2.00it/s]
 95%|█████████▌| 10161/10682 [1:37:45<04:20,  2.00it/s]
 95%|█████████▌| 10162/10682 [1:37:45<04:20,  2.00it/s]
 95%|█████████▌| 10163/10682 [1:37:46<04:19,  2.00it/s]
 95%|█████████▌| 10164/10682 [1:37:46<04:19,  2.00it/s]
 95%|█████████▌| 10165/10682 [1:37:47<04:19,  1.99it/s]
 95%|█████████▌| 10166/10682 [1:37:47<04:18,  2.00it/s]
 95%|█████████▌| 10167/10682 [1:37:48<04:17,  2.00it/s]
 95%|█████████▌| 10168/10682 [1:37:48<04:17,  2.00it/s]
 95%|█████████▌| 10169/10682 [1:37:49<04:16,  2.00it/s]
 95%|█████████▌| 10170/10682 [1:37:49<04:16,  2.00it/s]
 95%|█████████▌| 10171/10682 [1:37:50<04:15,  2.00it/s]
 95%|█████████▌| 10172/10682 [1:37:50<04:15,  2.00it/s]
 95%|█████████▌| 10173/10682 [1:37:51<04:15,  1.99it/s]
 95%|█████████▌| 10174/10682 [1:37:51<04:15,  1.99it/s]
 95%|█████████▌| 10175/10682 [1:37:52<04:15,  1.99it/s]{'loss': 3.1081, 'grad_norm': 0.24675361812114716, 'learning_rate': 6.847688328344037e-06, 'epoch': 13.34}
+
                                                       
 95%|█████████▌| 10175/10682 [1:37:52<04:15,  1.99it/s]
 95%|█████████▌| 10176/10682 [1:37:52<04:14,  1.99it/s]
 95%|█████████▌| 10177/10682 [1:37:53<04:13,  1.99it/s]
 95%|█████████▌| 10178/10682 [1:37:53<04:12,  1.99it/s]
 95%|█████████▌| 10179/10682 [1:37:54<04:12,  1.99it/s]
 95%|█████████▌| 10180/10682 [1:37:54<04:12,  1.99it/s]
 95%|█████████▌| 10181/10682 [1:37:55<04:11,  1.99it/s]
 95%|█████████▌| 10182/10682 [1:37:55<04:10,  1.99it/s]
 95%|█████████▌| 10183/10682 [1:37:56<04:10,  1.99it/s]
 95%|█████████▌| 10184/10682 [1:37:56<04:09,  1.99it/s]
 95%|█████████▌| 10185/10682 [1:37:57<04:09,  1.99it/s]
 95%|█████████▌| 10186/10682 [1:37:57<04:08,  2.00it/s]
 95%|█████████▌| 10187/10682 [1:37:58<04:08,  2.00it/s]
 95%|█████████▌| 10188/10682 [1:37:58<04:07,  1.99it/s]
 95%|█████████▌| 10189/10682 [1:37:59<04:07,  1.99it/s]
 95%|█████████▌| 10190/10682 [1:37:59<04:06,  1.99it/s]
 95%|█████████▌| 10191/10682 [1:38:00<04:06,  1.99it/s]
 95%|█████████▌| 10192/10682 [1:38:00<04:05,  1.99it/s]
 95%|█████████▌| 10193/10682 [1:38:01<04:05,  2.00it/s]
 95%|█████████▌| 10194/10682 [1:38:01<04:04,  1.99it/s]
 95%|█████████▌| 10195/10682 [1:38:02<04:04,  1.99it/s]
 95%|█████████▌| 10196/10682 [1:38:02<04:03,  1.99it/s]
 95%|█████████▌| 10197/10682 [1:38:03<04:03,  1.99it/s]
 95%|█████████▌| 10198/10682 [1:38:03<04:02,  2.00it/s]
 95%|█████████▌| 10199/10682 [1:38:04<04:02,  1.99it/s]
 95%|█████████▌| 10200/10682 [1:38:04<04:01,  1.99it/s]{'loss': 3.1186, 'grad_norm': 0.24594959616661072, 'learning_rate': 6.190386348572108e-06, 'epoch': 13.37}
                                                       
+
 95%|█████████▌| 10200/10682 [1:38:04<04:01,  1.99it/s]
 95%|█████████▌| 10201/10682 [1:38:05<04:01,  1.99it/s]
 96%|█████████▌| 10202/10682 [1:38:05<04:00,  1.99it/s]
 96%|█████████▌| 10203/10682 [1:38:06<03:59,  2.00it/s]
 96%|█████████▌| 10204/10682 [1:38:06<03:59,  2.00it/s]
 96%|█████████▌| 10205/10682 [1:38:07<03:58,  2.00it/s]
 96%|█████████▌| 10206/10682 [1:38:07<03:58,  1.99it/s]
 96%|█████████▌| 10207/10682 [1:38:08<03:58,  1.99it/s]
 96%|█████████▌| 10208/10682 [1:38:08<03:57,  2.00it/s]
 96%|█████████▌| 10209/10682 [1:38:09<03:56,  2.00it/s]
 96%|█████████▌| 10210/10682 [1:38:09<03:56,  2.00it/s]
 96%|█████████▌| 10211/10682 [1:38:10<04:16,  1.83it/s]
 96%|█████████▌| 10212/10682 [1:38:10<04:10,  1.88it/s]
 96%|█████████▌| 10213/10682 [1:38:11<04:05,  1.91it/s]
 96%|█████████▌| 10214/10682 [1:38:11<04:01,  1.94it/s]
 96%|█████████▌| 10215/10682 [1:38:12<03:58,  1.96it/s]
 96%|█████████▌| 10216/10682 [1:38:12<03:57,  1.97it/s]
 96%|█████████▌| 10217/10682 [1:38:13<03:55,  1.97it/s]
 96%|█████████▌| 10218/10682 [1:38:13<03:54,  1.98it/s]
 96%|█████████▌| 10219/10682 [1:38:14<03:53,  1.98it/s]
 96%|█████████▌| 10220/10682 [1:38:14<03:52,  1.99it/s]
 96%|█████████▌| 10221/10682 [1:38:15<03:51,  1.99it/s]
 96%|█████████▌| 10222/10682 [1:38:15<03:51,  1.99it/s]
 96%|█████████▌| 10223/10682 [1:38:16<03:50,  1.99it/s]
 96%|█████████▌| 10224/10682 [1:38:16<03:50,  1.99it/s]
 96%|█████████▌| 10225/10682 [1:38:17<03:49,  1.99it/s]
                                                       {'loss': 3.1084, 'grad_norm': 0.24636626243591309, 'learning_rate': 5.56604678228706e-06, 'epoch': 13.4}
+
 96%|█████████▌| 10225/10682 [1:38:17<03:49,  1.99it/s]
 96%|█████████▌| 10226/10682 [1:38:17<03:48,  1.99it/s]
 96%|█████████▌| 10227/10682 [1:38:18<03:48,  1.99it/s]
 96%|█████████▌| 10228/10682 [1:38:18<03:47,  1.99it/s]
 96%|█████████▌| 10229/10682 [1:38:19<03:46,  2.00it/s]
 96%|█████████▌| 10230/10682 [1:38:19<03:46,  1.99it/s]
 96%|█████████▌| 10231/10682 [1:38:20<03:46,  1.99it/s]
 96%|█████████▌| 10232/10682 [1:38:20<03:46,  1.99it/s]
 96%|█████████▌| 10233/10682 [1:38:21<03:45,  1.99it/s]
 96%|█████████▌| 10234/10682 [1:38:21<03:44,  1.99it/s]
 96%|█████████▌| 10235/10682 [1:38:22<03:44,  1.99it/s]
 96%|█████████▌| 10236/10682 [1:38:22<03:43,  1.99it/s]
 96%|█████████▌| 10237/10682 [1:38:23<03:43,  1.99it/s]
 96%|█████████▌| 10238/10682 [1:38:23<03:42,  1.99it/s]
 96%|█████████▌| 10239/10682 [1:38:24<03:43,  1.98it/s]
 96%|█████████▌| 10240/10682 [1:38:24<03:42,  1.99it/s]
 96%|█████████▌| 10241/10682 [1:38:25<03:41,  1.99it/s]
 96%|█████████▌| 10242/10682 [1:38:25<03:41,  1.99it/s]
 96%|█████████▌| 10243/10682 [1:38:26<03:40,  1.99it/s]
 96%|█████████▌| 10244/10682 [1:38:26<03:39,  1.99it/s]
 96%|█████████▌| 10245/10682 [1:38:27<03:39,  1.99it/s]
 96%|█████████▌| 10246/10682 [1:38:27<03:38,  1.99it/s]
 96%|█████████▌| 10247/10682 [1:38:28<03:38,  1.99it/s]
 96%|█████████▌| 10248/10682 [1:38:28<03:37,  1.99it/s]
 96%|█████████▌| 10249/10682 [1:38:29<03:37,  1.99it/s]
 96%|█████████▌| 10250/10682 [1:38:29<03:36,  2.00it/s]
                                                       {'loss': 3.1113, 'grad_norm': 0.24913303554058075, 'learning_rate': 4.974711304941093e-06, 'epoch': 13.43}
+
 96%|█████████▌| 10250/10682 [1:38:29<03:36,  2.00it/s]
 96%|█████████▌| 10251/10682 [1:38:30<03:36,  1.99it/s]
 96%|█████████▌| 10252/10682 [1:38:30<03:35,  1.99it/s]
 96%|█████████▌| 10253/10682 [1:38:31<03:35,  1.99it/s]
 96%|█████████▌| 10254/10682 [1:38:31<03:34,  1.99it/s]
 96%|█████████▌| 10255/10682 [1:38:32<03:34,  1.99it/s]
 96%|█████████▌| 10256/10682 [1:38:32<03:33,  1.99it/s]
 96%|█████████▌| 10257/10682 [1:38:33<03:33,  1.99it/s]
 96%|█████████▌| 10258/10682 [1:38:33<03:32,  1.99it/s]
 96%|█████████▌| 10259/10682 [1:38:34<03:32,  1.99it/s]
 96%|█████████▌| 10260/10682 [1:38:34<03:31,  1.99it/s]
 96%|█████████▌| 10261/10682 [1:38:35<03:31,  1.99it/s]
 96%|█████████▌| 10262/10682 [1:38:35<03:30,  1.99it/s]
 96%|█████████▌| 10263/10682 [1:38:36<03:30,  1.99it/s]
 96%|█████████▌| 10264/10682 [1:38:36<03:29,  1.99it/s]
 96%|█████████▌| 10265/10682 [1:38:37<03:29,  1.99it/s]
 96%|█████████▌| 10266/10682 [1:38:37<03:29,  1.99it/s]
 96%|█████████▌| 10267/10682 [1:38:38<03:28,  1.99it/s]
 96%|█████████▌| 10268/10682 [1:38:38<03:28,  1.99it/s]
 96%|█████████▌| 10269/10682 [1:38:39<03:27,  1.99it/s]
 96%|█████████▌| 10270/10682 [1:38:39<03:26,  1.99it/s]
 96%|█████████▌| 10271/10682 [1:38:40<03:26,  1.99it/s]
 96%|█████████▌| 10272/10682 [1:38:40<03:25,  1.99it/s]
 96%|█████████▌| 10273/10682 [1:38:41<03:25,  1.99it/s]
 96%|█████████▌| 10274/10682 [1:38:41<03:24,  1.99it/s]
 96%|█████████▌| 10275/10682 [1:38:42<03:24,  1.99it/s]
                                                       
+
 96%|█████████▌| 10275/10682 [1:38:42<03:24,  1.99it/s]
 96%|█████████▌| 10276/10682 [1:38:42<03:23,  1.99it/s]
 96%|█████████▌| 10277/10682 [1:38:43<03:23,  1.99it/s]
 96%|█████████▌| 10278/10682 [1:38:43<03:22,  1.99it/s]
 96%|█████████▌| 10279/10682 [1:38:44<03:22,  1.99it/s]
 96%|█████████▌| 10280/10682 [1:38:44<03:21,  1.99it/s]
 96%|█████████▌| 10281/10682 [1:38:45<03:21,  1.99it/s]
 96%|█████████▋| 10282/10682 [1:38:45<03:20,  1.99it/s]
 96%|█████████▋| 10283/10682 [1:38:46<03:20,  1.99it/s]
 96%|█████████▋| 10284/10682 [1:38:46<03:19,  1.99it/s]
 96%|█████████▋| 10285/10682 [1:38:47<03:19,  1.99it/s]
 96%|█████████▋| 10286/10682 [1:38:47<03:18,  1.99it/s]
 96%|█████████▋| 10287/10682 [1:38:48<03:18,  1.99it/s]
 96%|█████████▋| 10288/10682 [1:38:48<03:17,  1.99it/s]
 96%|█████████▋| 10289/10682 [1:38:49<03:17,  1.99it/s]
 96%|█████████▋| 10290/10682 [1:38:50<03:17,  1.99it/s]
 96%|█████████▋| 10291/10682 [1:38:50<03:16,  1.99it/s]
 96%|█████████▋| 10292/10682 [1:38:51<03:15,  1.99it/s]
 96%|█████████▋| 10293/10682 [1:38:51<03:15,  1.99it/s]
 96%|█████████▋| 10294/10682 [1:38:52<03:14,  1.99it/s]
 96%|█████████▋| 10295/10682 [1:38:52<03:14,  1.99it/s]
 96%|█████████▋| 10296/10682 [1:38:53<03:13,  1.99it/s]
 96%|█████████▋| 10297/10682 [1:38:53<03:13,  1.99it/s]
 96%|█████████▋| 10298/10682 [1:38:54<03:12,  1.99it/s]
 96%|█████████▋| 10299/10682 [1:38:54<03:12,  1.99it/s]
 96%|█████████▋| 10300/10682 [1:38:55<03:11,  1.99it/s]{'loss': 3.111, 'grad_norm': 0.2491437792778015, 'learning_rate': 3.891208300917604e-06, 'epoch': 13.5}
+
                                                       
 96%|█████████▋| 10300/10682 [1:38:55<03:11,  1.99it/s]
 96%|█████████▋| 10301/10682 [1:38:55<03:11,  1.99it/s]
 96%|█████████▋| 10302/10682 [1:38:56<03:10,  1.99it/s]
 96%|█████████▋| 10303/10682 [1:38:56<03:10,  1.99it/s]
 96%|█████████▋| 10304/10682 [1:38:57<03:10,  1.99it/s]
 96%|█████████▋| 10305/10682 [1:38:57<03:09,  1.99it/s]
 96%|█████████▋| 10306/10682 [1:38:58<03:09,  1.99it/s]
 96%|█████████▋| 10307/10682 [1:38:58<03:08,  1.99it/s]
 96%|█████████▋| 10308/10682 [1:38:59<03:07,  1.99it/s]
 97%|█████████▋| 10309/10682 [1:38:59<03:07,  1.99it/s]
 97%|█████████▋| 10310/10682 [1:39:00<03:06,  1.99it/s]
 97%|█████████▋| 10311/10682 [1:39:00<03:06,  1.99it/s]
 97%|█████████▋| 10312/10682 [1:39:01<03:05,  1.99it/s]
 97%|█████████▋| 10313/10682 [1:39:01<03:05,  1.99it/s]
 97%|█████████▋| 10314/10682 [1:39:02<03:04,  1.99it/s]
 97%|█████████▋| 10315/10682 [1:39:02<03:04,  1.99it/s]
 97%|█████████▋| 10316/10682 [1:39:03<03:03,  1.99it/s]
 97%|█████████▋| 10317/10682 [1:39:03<03:03,  1.99it/s]
 97%|█████████▋| 10318/10682 [1:39:04<03:02,  1.99it/s]
 97%|█████████▋| 10319/10682 [1:39:04<03:02,  1.99it/s]
 97%|█████████▋| 10320/10682 [1:39:05<03:01,  1.99it/s]
 97%|█████████▋| 10321/10682 [1:39:05<03:01,  1.99it/s]
 97%|█████████▋| 10322/10682 [1:39:06<03:00,  1.99it/s]
 97%|█████████▋| 10323/10682 [1:39:06<02:59,  1.99it/s]
 97%|█████████▋| 10324/10682 [1:39:07<02:59,  1.99it/s]
 97%|█████████▋| 10325/10682 [1:39:07<02:59,  1.99it/s]{'loss': 3.1165, 'grad_norm': 0.24889500439167023, 'learning_rate': 3.3991130994299734e-06, 'epoch': 13.53}
+
                                                       
 97%|█████████▋| 10325/10682 [1:39:07<02:59,  1.99it/s]
 97%|█████████▋| 10326/10682 [1:39:08<02:58,  1.99it/s]
 97%|█████████▋| 10327/10682 [1:39:08<02:58,  1.99it/s]
 97%|█████████▋| 10328/10682 [1:39:09<02:57,  1.99it/s]
 97%|█████████▋| 10329/10682 [1:39:09<02:57,  1.99it/s]
 97%|█████████▋| 10330/10682 [1:39:10<02:56,  1.99it/s]
 97%|█████████▋| 10331/10682 [1:39:10<02:56,  1.99it/s]
 97%|█████████▋| 10332/10682 [1:39:11<02:55,  1.99it/s]
 97%|█████████▋| 10333/10682 [1:39:11<02:55,  1.99it/s]
 97%|█████████▋| 10334/10682 [1:39:12<02:54,  1.99it/s]
 97%|█████████▋| 10335/10682 [1:39:12<02:54,  1.99it/s]
 97%|█████████▋| 10336/10682 [1:39:13<02:53,  1.99it/s]
 97%|█████████▋| 10337/10682 [1:39:13<02:53,  1.99it/s]
 97%|█████████▋| 10338/10682 [1:39:14<02:52,  1.99it/s]
 97%|█████████▋| 10339/10682 [1:39:14<02:52,  1.99it/s]
 97%|█████████▋| 10340/10682 [1:39:15<02:51,  1.99it/s]
 97%|█████████▋| 10341/10682 [1:39:15<02:51,  1.99it/s]
 97%|█████████▋| 10342/10682 [1:39:16<02:50,  1.99it/s]
 97%|█████████▋| 10343/10682 [1:39:16<02:50,  1.99it/s]
 97%|█████████▋| 10344/10682 [1:39:17<02:49,  1.99it/s]
 97%|█████████▋| 10345/10682 [1:39:17<02:49,  1.99it/s]
 97%|█████████▋| 10346/10682 [1:39:18<02:48,  1.99it/s]
 97%|█████████▋| 10347/10682 [1:39:18<02:48,  1.99it/s]
 97%|█████████▋| 10348/10682 [1:39:19<02:47,  1.99it/s]
 97%|█████████▋| 10349/10682 [1:39:19<02:46,  1.99it/s]
 97%|█████████▋| 10350/10682 [1:39:20<02:46,  1.99it/s]{'loss': 3.1195, 'grad_norm': 0.2491111308336258, 'learning_rate': 2.940166632433183e-06, 'epoch': 13.56}
+
                                                       
 97%|█████████▋| 10350/10682 [1:39:20<02:46,  1.99it/s]
 97%|█████████▋| 10351/10682 [1:39:20<02:46,  1.99it/s]
 97%|█████████▋| 10352/10682 [1:39:21<02:45,  1.99it/s]
 97%|█████████▋| 10353/10682 [1:39:21<02:45,  1.99it/s]
 97%|█████████▋| 10354/10682 [1:39:22<02:44,  1.99it/s]
 97%|█████████▋| 10355/10682 [1:39:22<02:44,  1.99it/s]
 97%|█████████▋| 10356/10682 [1:39:23<02:43,  1.99it/s]
 97%|█████████▋| 10357/10682 [1:39:23<02:43,  1.99it/s]
 97%|█████████▋| 10358/10682 [1:39:24<02:42,  1.99it/s]
 97%|█████████▋| 10359/10682 [1:39:24<02:43,  1.98it/s]
 97%|█████████▋| 10360/10682 [1:39:25<02:42,  1.98it/s]
 97%|█████████▋| 10361/10682 [1:39:25<02:41,  1.98it/s]
 97%|█████████▋| 10362/10682 [1:39:26<02:40,  1.99it/s]
 97%|█████████▋| 10363/10682 [1:39:26<02:40,  1.99it/s]
 97%|█████████▋| 10364/10682 [1:39:27<02:39,  1.99it/s]
 97%|█████████▋| 10365/10682 [1:39:27<02:39,  1.99it/s]
 97%|█████████▋| 10366/10682 [1:39:28<02:38,  1.99it/s]
 97%|█████████▋| 10367/10682 [1:39:28<02:38,  1.99it/s]
 97%|█████████▋| 10368/10682 [1:39:29<02:37,  1.99it/s]
 97%|█████████▋| 10369/10682 [1:39:29<02:37,  1.99it/s]
 97%|█████████▋| 10370/10682 [1:39:30<02:36,  1.99it/s]
 97%|█████████▋| 10371/10682 [1:39:30<02:36,  1.99it/s]
 97%|█████████▋| 10372/10682 [1:39:31<02:35,  1.99it/s]
 97%|█████████▋| 10373/10682 [1:39:31<02:35,  1.99it/s]
 97%|█████████▋| 10374/10682 [1:39:32<02:34,  1.99it/s]
 97%|█████████▋| 10375/10682 [1:39:32<02:34,  1.99it/s]{'loss': 3.1152, 'grad_norm': 0.24760933220386505, 'learning_rate': 2.5143995351817882e-06, 'epoch': 13.6}
                                                       
+
 97%|█████████▋| 10375/10682 [1:39:32<02:34,  1.99it/s]
 97%|█████████▋| 10376/10682 [1:39:33<02:33,  1.99it/s]
 97%|█████████▋| 10377/10682 [1:39:33<02:33,  1.99it/s]
 97%|█████████▋| 10378/10682 [1:39:34<02:32,  1.99it/s]
 97%|█████████▋| 10379/10682 [1:39:34<02:32,  1.99it/s]
 97%|█████████▋| 10380/10682 [1:39:35<02:31,  1.99it/s]
 97%|█████████▋| 10381/10682 [1:39:35<02:31,  1.99it/s]
 97%|█████████▋| 10382/10682 [1:39:36<02:30,  1.99it/s]
 97%|█████████▋| 10383/10682 [1:39:36<02:29,  2.00it/s]
 97%|█████████▋| 10384/10682 [1:39:37<02:29,  1.99it/s]
 97%|█████████▋| 10385/10682 [1:39:37<02:28,  1.99it/s]
 97%|█████████▋| 10386/10682 [1:39:38<02:28,  1.99it/s]
 97%|█████████▋| 10387/10682 [1:39:38<02:27,  1.99it/s]
 97%|█████████▋| 10388/10682 [1:39:39<02:27,  1.99it/s]
 97%|█████████▋| 10389/10682 [1:39:39<02:26,  1.99it/s]
 97%|█████████▋| 10390/10682 [1:39:40<02:26,  1.99it/s]
 97%|█████████▋| 10391/10682 [1:39:40<02:25,  1.99it/s]
 97%|█████████▋| 10392/10682 [1:39:41<02:25,  1.99it/s]
 97%|█████████▋| 10393/10682 [1:39:41<02:25,  1.99it/s]
 97%|█████████▋| 10394/10682 [1:39:42<02:24,  2.00it/s]
 97%|█████████▋| 10395/10682 [1:39:42<02:23,  1.99it/s]
 97%|█████████▋| 10396/10682 [1:39:43<02:23,  1.99it/s]
 97%|█████████▋| 10397/10682 [1:39:43<02:23,  1.99it/s]
 97%|█████████▋| 10398/10682 [1:39:44<02:22,  1.99it/s]
 97%|█████████▋| 10399/10682 [1:39:44<02:21,  1.99it/s]
 97%|█████████▋| 10400/10682 [1:39:45<02:21,  2.00it/s]{'loss': 3.1212, 'grad_norm': 0.24558085203170776, 'learning_rate': 2.1218402281655835e-06, 'epoch': 13.63}
+
                                                       
 97%|█████████▋| 10400/10682 [1:39:45<02:21,  2.00it/s]
 97%|█████████▋| 10401/10682 [1:39:45<02:20,  1.99it/s]
 97%|█████████▋| 10402/10682 [1:39:46<02:20,  1.99it/s]
 97%|█████████▋| 10403/10682 [1:39:46<02:20,  1.99it/s]
 97%|█████████▋| 10404/10682 [1:39:47<02:19,  1.99it/s]
 97%|█████████▋| 10405/10682 [1:39:47<02:19,  1.99it/s]
 97%|█████████▋| 10406/10682 [1:39:48<02:18,  1.99it/s]
 97%|█████████▋| 10407/10682 [1:39:48<02:17,  1.99it/s]
 97%|█████████▋| 10408/10682 [1:39:49<02:17,  1.99it/s]
 97%|█████████▋| 10409/10682 [1:39:49<02:17,  1.99it/s]
 97%|█████████▋| 10410/10682 [1:39:50<02:16,  1.99it/s]
 97%|█████████▋| 10411/10682 [1:39:50<02:16,  1.99it/s]
 97%|█████████▋| 10412/10682 [1:39:51<02:15,  1.99it/s]
 97%|█████████▋| 10413/10682 [1:39:51<02:14,  1.99it/s]
 97%|█████████▋| 10414/10682 [1:39:52<02:14,  1.99it/s]
 98%|█████████▊| 10415/10682 [1:39:52<02:13,  1.99it/s]
 98%|█████████▊| 10416/10682 [1:39:53<02:13,  1.99it/s]
 98%|█████████▊| 10417/10682 [1:39:53<02:13,  1.99it/s]
 98%|█████████▊| 10418/10682 [1:39:54<02:12,  1.99it/s]
 98%|█████████▊| 10419/10682 [1:39:54<02:11,  1.99it/s]
 98%|█████████▊| 10420/10682 [1:39:55<02:11,  2.00it/s]
 98%|█████████▊| 10421/10682 [1:39:55<02:10,  1.99it/s]
 98%|█████████▊| 10422/10682 [1:39:56<02:10,  1.99it/s]
 98%|█████████▊| 10423/10682 [1:39:56<02:10,  1.99it/s]
 98%|█████████▊| 10424/10682 [1:39:57<02:09,  1.99it/s]
 98%|█████████▊| 10425/10682 [1:39:57<02:08,  1.99it/s]{'loss': 3.1095, 'grad_norm': 0.24991337954998016, 'learning_rate': 1.7625149152127318e-06, 'epoch': 13.66}
                                                       
+
 98%|█████████▊| 10425/10682 [1:39:57<02:08,  1.99it/s]
 98%|█████████▊| 10426/10682 [1:39:58<02:08,  1.99it/s]
 98%|█████████▊| 10427/10682 [1:39:58<02:07,  1.99it/s]
 98%|█████████▊| 10428/10682 [1:39:59<02:07,  1.99it/s]
 98%|█████████▊| 10429/10682 [1:39:59<02:07,  1.99it/s]
 98%|█████████▊| 10430/10682 [1:40:00<02:06,  1.99it/s]
 98%|█████████▊| 10431/10682 [1:40:00<02:06,  1.99it/s]
 98%|█████████▊| 10432/10682 [1:40:01<02:05,  1.99it/s]
 98%|█████████▊| 10433/10682 [1:40:01<02:04,  1.99it/s]
 98%|█████████▊| 10434/10682 [1:40:02<02:04,  1.99it/s]
 98%|█████████▊| 10435/10682 [1:40:02<02:04,  1.99it/s]
 98%|█████████▊| 10436/10682 [1:40:03<02:03,  1.99it/s]
 98%|█████████▊| 10437/10682 [1:40:03<02:03,  1.99it/s]
 98%|█████████▊| 10438/10682 [1:40:04<02:02,  1.99it/s]
 98%|█████████▊| 10439/10682 [1:40:04<02:02,  1.99it/s]
 98%|█████████▊| 10440/10682 [1:40:05<02:01,  1.99it/s]
 98%|█████████▊| 10441/10682 [1:40:05<02:01,  1.99it/s]
 98%|█████████▊| 10442/10682 [1:40:06<02:00,  1.99it/s]
 98%|█████████▊| 10443/10682 [1:40:06<02:00,  1.99it/s]
 98%|█████████▊| 10444/10682 [1:40:07<01:59,  1.99it/s]
 98%|█████████▊| 10445/10682 [1:40:07<01:58,  1.99it/s]
 98%|█████████▊| 10446/10682 [1:40:08<01:58,  1.99it/s]
 98%|█████████▊| 10447/10682 [1:40:08<01:58,  1.99it/s]
 98%|█████████▊| 10448/10682 [1:40:09<01:57,  1.99it/s]
 98%|█████████▊| 10449/10682 [1:40:09<01:57,  1.99it/s]
 98%|█████████▊| 10450/10682 [1:40:10<01:56,  1.99it/s]
                                                       
+
 98%|█████████▊| 10450/10682 [1:40:10<01:56,  1.99it/s]
 98%|█████████▊| 10451/10682 [1:40:10<01:56,  1.99it/s]
 98%|█████████▊| 10452/10682 [1:40:11<01:55,  1.99it/s]
 98%|█████████▊| 10453/10682 [1:40:11<01:54,  1.99it/s]
 98%|█████████▊| 10454/10682 [1:40:12<01:54,  1.99it/s]
 98%|█████████▊| 10455/10682 [1:40:12<01:54,  1.99it/s]
 98%|█████████▊| 10456/10682 [1:40:13<01:53,  1.99it/s]
 98%|█████████▊| 10457/10682 [1:40:13<01:52,  1.99it/s]
 98%|█████████▊| 10458/10682 [1:40:14<01:52,  1.99it/s]
 98%|█████████▊| 10459/10682 [1:40:14<01:51,  1.99it/s]
 98%|█████████▊| 10460/10682 [1:40:15<01:51,  1.99it/s]
 98%|█████████▊| 10461/10682 [1:40:15<01:50,  1.99it/s]
 98%|█████████▊| 10462/10682 [1:40:16<01:50,  2.00it/s]
 98%|█████████▊| 10463/10682 [1:40:16<01:49,  1.99it/s]
 98%|█████████▊| 10464/10682 [1:40:17<01:49,  1.99it/s]
 98%|█████████▊| 10465/10682 [1:40:17<01:48,  1.99it/s]
 98%|█████████▊| 10466/10682 [1:40:18<01:48,  1.99it/s]
 98%|█████████▊| 10467/10682 [1:40:18<01:48,  1.99it/s]
 98%|█████████▊| 10468/10682 [1:40:19<01:47,  1.99it/s]
 98%|█████████▊| 10469/10682 [1:40:19<01:46,  1.99it/s]
 98%|█████████▊| 10470/10682 [1:40:20<01:46,  1.99it/s]
 98%|█████████▊| 10471/10682 [1:40:20<01:45,  1.99it/s]
 98%|█████████▊| 10472/10682 [1:40:21<01:45,  1.99it/s]
 98%|█████████▊| 10473/10682 [1:40:21<01:44,  1.99it/s]
 98%|█████████▊| 10474/10682 [1:40:22<01:44,  1.99it/s]
 98%|█████████▊| 10475/10682 [1:40:22<01:43,  1.99it/s]
                                                       
+
 98%|█████████▊| 10475/10682 [1:40:22<01:43,  1.99it/s]
 98%|█████████▊| 10476/10682 [1:40:23<01:43,  1.99it/s]
 98%|█████████▊| 10477/10682 [1:40:23<01:42,  1.99it/s]
 98%|█████████▊| 10478/10682 [1:40:24<01:42,  1.99it/s]
 98%|█████████▊| 10479/10682 [1:40:24<01:42,  1.99it/s]
 98%|█████████▊| 10480/10682 [1:40:25<01:41,  1.99it/s]
 98%|█████████▊| 10481/10682 [1:40:25<01:40,  1.99it/s]
 98%|█████████▊| 10482/10682 [1:40:26<01:40,  1.99it/s]
 98%|█████████▊| 10483/10682 [1:40:26<01:39,  1.99it/s]
 98%|█████████▊| 10484/10682 [1:40:27<01:39,  1.99it/s]
 98%|█████████▊| 10485/10682 [1:40:27<01:38,  1.99it/s]
 98%|█████████▊| 10486/10682 [1:40:28<01:38,  1.99it/s]
 98%|█████████▊| 10487/10682 [1:40:28<01:38,  1.99it/s]
 98%|█████████▊| 10488/10682 [1:40:29<01:37,  1.99it/s]
 98%|█████████▊| 10489/10682 [1:40:29<01:36,  1.99it/s]
 98%|█████████▊| 10490/10682 [1:40:30<01:36,  1.99it/s]
 98%|█████████▊| 10491/10682 [1:40:30<01:35,  1.99it/s]
 98%|█████████▊| 10492/10682 [1:40:31<01:35,  1.99it/s]
 98%|█████████▊| 10493/10682 [1:40:31<01:34,  1.99it/s]
 98%|█████████▊| 10494/10682 [1:40:32<01:34,  1.99it/s]
 98%|█████████▊| 10495/10682 [1:40:32<01:33,  1.99it/s]
 98%|█████████▊| 10496/10682 [1:40:33<01:33,  1.99it/s]
 98%|█████████▊| 10497/10682 [1:40:33<01:32,  1.99it/s]
 98%|█████████▊| 10498/10682 [1:40:34<01:32,  2.00it/s]
 98%|█████████▊| 10499/10682 [1:40:34<01:31,  1.99it/s]
 98%|█████████▊| 10500/10682 [1:40:35<01:31,  1.99it/s]{'loss': 3.116, 'grad_norm': 0.24622003734111786, 'learning_rate': 8.841716933915555e-07, 'epoch': 13.76}
+
                                                       
 98%|█████████▊| 10500/10682 [1:40:35<01:31,  1.99it/s]
 98%|█████████▊| 10501/10682 [1:40:35<01:30,  1.99it/s]
 98%|█████████▊| 10502/10682 [1:40:36<01:30,  1.99it/s]
 98%|█████████▊| 10503/10682 [1:40:36<01:29,  1.99it/s]
 98%|█████████▊| 10504/10682 [1:40:37<01:29,  1.99it/s]
 98%|█████████▊| 10505/10682 [1:40:37<01:28,  1.99it/s]
 98%|█████████▊| 10506/10682 [1:40:38<01:28,  1.99it/s]
 98%|█████████▊| 10507/10682 [1:40:38<01:27,  1.99it/s]
 98%|█████████▊| 10508/10682 [1:40:39<01:27,  1.99it/s]
 98%|█████████▊| 10509/10682 [1:40:39<01:26,  1.99it/s]
 98%|█████████▊| 10510/10682 [1:40:40<01:26,  1.99it/s]
 98%|█████████▊| 10511/10682 [1:40:40<01:25,  1.99it/s]
 98%|█████████▊| 10512/10682 [1:40:41<01:25,  1.99it/s]
 98%|█████████▊| 10513/10682 [1:40:41<01:24,  1.99it/s]
 98%|█████████▊| 10514/10682 [1:40:42<01:24,  1.99it/s]
 98%|█████████▊| 10515/10682 [1:40:42<01:23,  1.99it/s]
 98%|█████████▊| 10516/10682 [1:40:43<01:23,  1.99it/s]
 98%|█████████▊| 10517/10682 [1:40:43<01:22,  1.99it/s]
 98%|█████████▊| 10518/10682 [1:40:44<01:22,  1.99it/s]
 98%|█████████▊| 10519/10682 [1:40:44<01:21,  1.99it/s]
 98%|█████████▊| 10520/10682 [1:40:45<01:21,  1.99it/s]
 98%|█████████▊| 10521/10682 [1:40:45<01:20,  1.99it/s]
 99%|█████████▊| 10522/10682 [1:40:46<01:20,  1.99it/s]
 99%|█████████▊| 10523/10682 [1:40:46<01:19,  1.99it/s]
 99%|█████████▊| 10524/10682 [1:40:47<01:19,  1.99it/s]
 99%|█████████▊| 10525/10682 [1:40:47<01:18,  1.99it/s]{'loss': 3.1108, 'grad_norm': 0.24499115347862244, 'learning_rate': 6.580000036264244e-07, 'epoch': 13.79}
+
                                                       
 99%|█████████▊| 10525/10682 [1:40:47<01:18,  1.99it/s]
 99%|█████████▊| 10526/10682 [1:40:48<01:18,  1.99it/s]
 99%|█████████▊| 10527/10682 [1:40:48<01:17,  1.99it/s]
 99%|█████████▊| 10528/10682 [1:40:49<01:17,  1.99it/s]
 99%|█████████▊| 10529/10682 [1:40:50<01:16,  1.99it/s]
 99%|█████████▊| 10530/10682 [1:40:50<01:16,  1.99it/s]
 99%|█████████▊| 10531/10682 [1:40:51<01:15,  1.99it/s]
 99%|█████████▊| 10532/10682 [1:40:51<01:15,  1.99it/s]
 99%|█████████▊| 10533/10682 [1:40:52<01:14,  1.99it/s]
 99%|█████████▊| 10534/10682 [1:40:52<01:14,  1.99it/s]
 99%|█████████▊| 10535/10682 [1:40:53<01:13,  1.99it/s]
 99%|█████████▊| 10536/10682 [1:40:53<01:13,  1.99it/s]
 99%|█████████▊| 10537/10682 [1:40:54<01:12,  1.99it/s]
 99%|█████████▊| 10538/10682 [1:40:54<01:12,  1.99it/s]
 99%|█████████▊| 10539/10682 [1:40:55<01:11,  1.99it/s]
 99%|█████████▊| 10540/10682 [1:40:55<01:11,  2.00it/s]
 99%|█████████▊| 10541/10682 [1:40:56<01:10,  2.00it/s]
 99%|█████████▊| 10542/10682 [1:40:56<01:10,  1.99it/s]
 99%|█████████▊| 10543/10682 [1:40:57<01:09,  1.99it/s]
 99%|█████████▊| 10544/10682 [1:40:57<01:09,  1.99it/s]
 99%|█████████▊| 10545/10682 [1:40:58<01:08,  1.99it/s]
 99%|█████████▊| 10546/10682 [1:40:58<01:08,  1.99it/s]
 99%|█████████▊| 10547/10682 [1:40:59<01:07,  1.99it/s]
 99%|█████████▊| 10548/10682 [1:40:59<01:07,  1.99it/s]
 99%|█████████▉| 10549/10682 [1:41:00<01:06,  1.99it/s]
 99%|█████████▉| 10550/10682 [1:41:00<01:06,  1.99it/s]
                                                       {'loss': 3.1109, 'grad_norm': 0.24714048206806183, 'learning_rate': 4.651600211027507e-07, 'epoch': 13.83}
+
 99%|█████████▉| 10550/10682 [1:41:00<01:06,  1.99it/s]
 99%|█████████▉| 10551/10682 [1:41:01<01:05,  1.99it/s]
 99%|█████████▉| 10552/10682 [1:41:01<01:05,  1.99it/s]
 99%|█████████▉| 10553/10682 [1:41:02<01:04,  1.99it/s]
 99%|█████████▉| 10554/10682 [1:41:02<01:04,  1.99it/s]
 99%|█████████▉| 10555/10682 [1:41:03<01:03,  1.99it/s]
 99%|█████████▉| 10556/10682 [1:41:03<01:03,  1.99it/s]
 99%|█████████▉| 10557/10682 [1:41:04<01:02,  1.99it/s]
 99%|█████████▉| 10558/10682 [1:41:04<01:02,  1.99it/s]
 99%|█████████▉| 10559/10682 [1:41:05<01:01,  1.99it/s]
 99%|█████████▉| 10560/10682 [1:41:05<01:01,  1.99it/s]
 99%|█████████▉| 10561/10682 [1:41:06<01:00,  1.99it/s]
 99%|█████████▉| 10562/10682 [1:41:06<01:00,  1.99it/s]
 99%|█████████▉| 10563/10682 [1:41:07<00:59,  1.99it/s]
 99%|█████████▉| 10564/10682 [1:41:07<00:59,  1.99it/s]
 99%|█████████▉| 10565/10682 [1:41:08<00:58,  1.99it/s]
 99%|█████████▉| 10566/10682 [1:41:08<00:58,  1.99it/s]
 99%|█████████▉| 10567/10682 [1:41:09<00:57,  1.99it/s]
 99%|█████████▉| 10568/10682 [1:41:09<00:57,  1.99it/s]
 99%|█████████▉| 10569/10682 [1:41:10<00:56,  1.99it/s]
 99%|█████████▉| 10570/10682 [1:41:10<00:56,  1.99it/s]
 99%|█████████▉| 10571/10682 [1:41:11<00:55,  1.99it/s]
 99%|█████████▉| 10572/10682 [1:41:11<00:55,  1.99it/s]
 99%|█████████▉| 10573/10682 [1:41:12<00:54,  1.99it/s]
 99%|█████████▉| 10574/10682 [1:41:12<00:54,  1.99it/s]
 99%|█████████▉| 10575/10682 [1:41:13<00:53,  1.99it/s]{'loss': 3.1055, 'grad_norm': 0.24866575002670288, 'learning_rate': 3.0566461813213986e-07, 'epoch': 13.86}
+
                                                       
 99%|█████████▉| 10575/10682 [1:41:13<00:53,  1.99it/s]
 99%|█████████▉| 10576/10682 [1:41:13<00:53,  1.99it/s]
 99%|█████████▉| 10577/10682 [1:41:14<00:52,  1.99it/s]
 99%|█████████▉| 10578/10682 [1:41:14<00:52,  1.99it/s]
 99%|█████████▉| 10579/10682 [1:41:15<00:51,  1.99it/s]
 99%|█████████▉| 10580/10682 [1:41:15<00:51,  1.99it/s]
 99%|█████████▉| 10581/10682 [1:41:16<00:50,  1.99it/s]
 99%|█████████▉| 10582/10682 [1:41:16<00:50,  1.99it/s]
 99%|█████████▉| 10583/10682 [1:41:17<00:49,  1.99it/s]
 99%|█████████▉| 10584/10682 [1:41:17<00:49,  1.99it/s]
 99%|█████████▉| 10585/10682 [1:41:18<00:48,  1.99it/s]
 99%|█████████▉| 10586/10682 [1:41:18<00:48,  1.99it/s]
 99%|█████████▉| 10587/10682 [1:41:19<00:47,  1.99it/s]
 99%|█████████▉| 10588/10682 [1:41:19<00:47,  1.99it/s]
 99%|█████████▉| 10589/10682 [1:41:20<00:46,  1.99it/s]
 99%|█████████▉| 10590/10682 [1:41:20<00:46,  1.99it/s]
 99%|█████████▉| 10591/10682 [1:41:21<00:45,  1.99it/s]
 99%|█████████▉| 10592/10682 [1:41:21<00:45,  1.99it/s]
 99%|█████████▉| 10593/10682 [1:41:22<00:44,  1.99it/s]
 99%|█████████▉| 10594/10682 [1:41:22<00:44,  1.99it/s]
 99%|█████████▉| 10595/10682 [1:41:23<00:43,  1.99it/s]
 99%|█████████▉| 10596/10682 [1:41:23<00:43,  1.99it/s]
 99%|█████████▉| 10597/10682 [1:41:24<00:42,  1.99it/s]
 99%|█████████▉| 10598/10682 [1:41:24<00:42,  1.97it/s]
 99%|█████████▉| 10599/10682 [1:41:25<00:41,  1.98it/s]
 99%|█████████▉| 10600/10682 [1:41:25<00:41,  1.98it/s]
                                                       
+
 99%|█████████▉| 10600/10682 [1:41:25<00:41,  1.98it/s]
 99%|█████████▉| 10601/10682 [1:41:26<00:40,  1.99it/s]
 99%|█████████▉| 10602/10682 [1:41:26<00:40,  1.99it/s]
 99%|█████████▉| 10603/10682 [1:41:27<00:39,  1.99it/s]
 99%|█████████▉| 10604/10682 [1:41:27<00:39,  1.99it/s]
 99%|█████████▉| 10605/10682 [1:41:28<00:38,  1.99it/s]
 99%|█████████▉| 10606/10682 [1:41:28<00:38,  1.99it/s]
 99%|█████████▉| 10607/10682 [1:41:29<00:37,  1.99it/s]
 99%|█████████▉| 10608/10682 [1:41:29<00:37,  2.00it/s]
 99%|█████████▉| 10609/10682 [1:41:30<00:36,  2.00it/s]
 99%|█████████▉| 10610/10682 [1:41:30<00:36,  2.00it/s]
 99%|█████████▉| 10611/10682 [1:41:31<00:35,  1.99it/s]
 99%|█████████▉| 10612/10682 [1:41:31<00:35,  2.00it/s]
 99%|█████████▉| 10613/10682 [1:41:32<00:34,  2.00it/s]
 99%|█████████▉| 10614/10682 [1:41:32<00:34,  1.99it/s]
 99%|█████████▉| 10615/10682 [1:41:33<00:33,  1.99it/s]
 99%|█████████▉| 10616/10682 [1:41:33<00:33,  1.99it/s]
 99%|█████████▉| 10617/10682 [1:41:34<00:32,  1.99it/s]
 99%|█████████▉| 10618/10682 [1:41:34<00:32,  1.99it/s]
 99%|█████████▉| 10619/10682 [1:41:35<00:31,  1.99it/s]
 99%|█████████▉| 10620/10682 [1:41:35<00:31,  1.99it/s]
 99%|█████████▉| 10621/10682 [1:41:36<00:30,  1.99it/s]
 99%|█████████▉| 10622/10682 [1:41:36<00:30,  1.99it/s]
 99%|█████████▉| 10623/10682 [1:41:37<00:29,  1.99it/s]
 99%|█████████▉| 10624/10682 [1:41:37<00:29,  2.00it/s]
 99%|█████████▉| 10625/10682 [1:41:38<00:28,  1.99it/s]{'loss': 3.1138, 'grad_norm': 0.24934372305870056, 'learning_rate': 8.674791042273533e-08, 'epoch': 13.93}
                                                       
+
 99%|█████████▉| 10625/10682 [1:41:38<00:28,  1.99it/s]
 99%|█████████▉| 10626/10682 [1:41:38<00:28,  1.99it/s]
 99%|█████████▉| 10627/10682 [1:41:39<00:27,  1.99it/s]
 99%|█████████▉| 10628/10682 [1:41:39<00:27,  2.00it/s]
                                                       
+
                                                       
+
                                                       
+
+Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.