up

wandb/run-20220325_193848-1sz5964i/files/config.yaml

@@ -8669,7 +8669,14 @@ _wandb:
       - 1
       - 5
       - 11
+      2:
+      - 1
+      - 5
+      - 11
+      - 12
       3:
+      - 1
+      - 7
       - 13
       4: 3.9.5
       5: 0.12.10

wandb/run-20220325_193848-1sz5964i/files/output.log

@@ -14173,3 +14173,15 @@ Upload file wandb/run-20220325_193848-1sz5964i/run-1sz5964i.wandb: 100%|██
   eval_samples_per_second =        3.7
   eval_steps_per_second   =      0.464
   eval_wer                =     0.1012
+[INFO|trainer.py:2369] 2022-03-26 02:55:51,800 >>   Batch size = 8 100%|█████████████| 218M/218M [00:11<00:00, 20.7MB/s]g-point operations will not be computed-26 02:44:45,865 >> `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`...
+[INFO|trainer.py:2369] 2022-03-26 02:55:51,800 >>   Batch size = 8 100%|█████████████| 218M/218M [00:11<00:00, 20.7MB/s]g-point operations will not be computed-26 02:44:45,865 >> `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`...
+[INFO|trainer.py:2369] 2022-03-26 02:55:51,800 >>   Batch size = 8 100%|█████████████| 218M/218M [00:11<00:00, 20.7MB/s]g-point operations will not be computed-26 02:44:45,865 >> `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`...
+[INFO|trainer.py:2369] 2022-03-26 02:55:51,800 >>   Batch size = 8 100%|█████████████| 218M/218M [00:11<00:00, 20.7MB/s]g-point operations will not be computed-26 02:44:45,865 >> `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`...
+[INFO|trainer.py:2369] 2022-03-26 02:55:51,800 >>   Batch size = 8 100%|█████████████| 218M/218M [00:11<00:00, 20.7MB/s]g-point operations will not be computed-26 02:44:45,865 >> `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`...
+[INFO|trainer.py:2369] 2022-03-26 02:55:51,800 >>   Batch size = 8 100%|█████████████| 218M/218M [00:11<00:00, 20.7MB/s]g-point operations will not be computed-26 02:44:45,865 >> `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`...
+[INFO|trainer.py:2369] 2022-03-26 02:55:51,800 >>   Batch size = 8 100%|█████████████| 218M/218M [00:11<00:00, 20.7MB/s]g-point operations will not be computed-26 02:44:45,865 >> `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`...
+03/26/2022 03:08:31 - WARNING - huggingface_hub.repository - To https://huggingface.co/sanchit-gandhi/wav2vec2-2-bart-large-cnn
+Upload file wandb/run-20220325_193848-1sz5964i/run-1sz5964i.wandb: 100%|█████████████| 218M/218M [00:11<00:00, 20.7MB/s]g-point operations will not be computed-26 02:44:45,865 >> `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`...
+  File "/home/sanchit_huggingface_co/gcp/lib/python3.9/site-packages/huggingface_hub/hf_api.py", line 870, in model_infog-point operations will not be computed-26 02:44:45,865 >> `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`...
+  File "/home/sanchit_huggingface_co/gcp/lib/python3.9/site-packages/huggingface_hub/hf_api.py", line 870, in model_infog-point operations will not be computed-26 02:44:45,865 >> `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`...
+  File "/home/sanchit_huggingface_co/gcp/lib/python3.9/site-packages/huggingface_hub/hf_api.py", line 870, in model_infog-point operations will not be computed-26 02:44:45,865 >> `use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`...

wandb/run-20220325_193848-1sz5964i/logs/debug-internal.log

@@ -24600,3 +24600,124 @@
 2022-03-26 03:08:00,146 INFO    Thread-8  :1763281 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/output.log
 2022-03-26 03:08:02,627 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: stop_status
 2022-03-26 03:08:02,628 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: stop_status
+2022-03-26 03:08:17,670 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: stop_status
+2022-03-26 03:08:17,671 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: stop_status
+2022-03-26 03:08:21,990 DEBUG   SenderThread:1763281 [sender.py:send():235] send: stats
+2022-03-26 03:08:22,153 INFO    Thread-8  :1763281 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/output.log
+2022-03-26 03:08:24,154 INFO    Thread-8  :1763281 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/output.log
+2022-03-26 03:08:26,155 INFO    Thread-8  :1763281 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/output.log
+2022-03-26 03:08:28,155 INFO    Thread-8  :1763281 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/output.log
+2022-03-26 03:08:30,156 INFO    Thread-8  :1763281 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/output.log
+2022-03-26 03:08:32,157 INFO    Thread-8  :1763281 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/output.log
+2022-03-26 03:08:32,714 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: stop_status
+2022-03-26 03:08:32,714 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: stop_status
+2022-03-26 03:08:34,158 INFO    Thread-8  :1763281 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/output.log
+2022-03-26 03:08:37,865 DEBUG   SenderThread:1763281 [sender.py:send():235] send: telemetry
+2022-03-26 03:08:37,865 DEBUG   SenderThread:1763281 [sender.py:send():235] send: exit
+2022-03-26 03:08:37,865 INFO    SenderThread:1763281 [sender.py:send_exit():371] handling exit code: 1
+2022-03-26 03:08:37,866 INFO    SenderThread:1763281 [sender.py:send_exit():373] handling runtime: 26988
+2022-03-26 03:08:37,866 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: poll_exit
+2022-03-26 03:08:37,910 INFO    SenderThread:1763281 [sender.py:_save_file():944] saving file wandb-summary.json with policy end
+2022-03-26 03:08:37,910 INFO    SenderThread:1763281 [sender.py:send_exit():379] send defer
+2022-03-26 03:08:37,910 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: poll_exit
+2022-03-26 03:08:37,911 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: defer
+2022-03-26 03:08:37,911 INFO    HandlerThread:1763281 [handler.py:handle_request_defer():154] handle defer: 0
+2022-03-26 03:08:37,911 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: defer
+2022-03-26 03:08:37,911 INFO    SenderThread:1763281 [sender.py:send_request_defer():388] handle sender defer: 0
+2022-03-26 03:08:37,911 INFO    SenderThread:1763281 [sender.py:transition_state():392] send defer: 1
+2022-03-26 03:08:37,912 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: defer
+2022-03-26 03:08:37,912 INFO    HandlerThread:1763281 [handler.py:handle_request_defer():154] handle defer: 1
+2022-03-26 03:08:38,000 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: defer
+2022-03-26 03:08:38,000 INFO    SenderThread:1763281 [sender.py:send_request_defer():388] handle sender defer: 1
+2022-03-26 03:08:38,000 INFO    SenderThread:1763281 [sender.py:transition_state():392] send defer: 2
+2022-03-26 03:08:38,000 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: defer
+2022-03-26 03:08:38,000 DEBUG   SenderThread:1763281 [sender.py:send():235] send: stats
+2022-03-26 03:08:38,001 INFO    HandlerThread:1763281 [handler.py:handle_request_defer():154] handle defer: 2
+2022-03-26 03:08:38,001 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: defer
+2022-03-26 03:08:38,001 INFO    SenderThread:1763281 [sender.py:send_request_defer():388] handle sender defer: 2
+2022-03-26 03:08:38,001 INFO    SenderThread:1763281 [sender.py:transition_state():392] send defer: 3
+2022-03-26 03:08:38,002 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: defer
+2022-03-26 03:08:38,002 INFO    HandlerThread:1763281 [handler.py:handle_request_defer():154] handle defer: 3
+2022-03-26 03:08:38,050 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: poll_exit
+2022-03-26 03:08:38,050 DEBUG   SenderThread:1763281 [sender.py:send():235] send: summary
+2022-03-26 03:08:38,124 INFO    SenderThread:1763281 [sender.py:_save_file():944] saving file wandb-summary.json with policy end
+2022-03-26 03:08:38,124 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: defer
+2022-03-26 03:08:38,124 INFO    SenderThread:1763281 [sender.py:send_request_defer():388] handle sender defer: 3
+2022-03-26 03:08:38,124 INFO    SenderThread:1763281 [sender.py:transition_state():392] send defer: 4
+2022-03-26 03:08:38,124 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: poll_exit
+2022-03-26 03:08:38,125 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: defer
+2022-03-26 03:08:38,125 INFO    HandlerThread:1763281 [handler.py:handle_request_defer():154] handle defer: 4
+2022-03-26 03:08:38,125 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: defer
+2022-03-26 03:08:38,125 INFO    SenderThread:1763281 [sender.py:send_request_defer():388] handle sender defer: 4
+2022-03-26 03:08:38,159 INFO    Thread-8  :1763281 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/output.log
+2022-03-26 03:08:38,159 INFO    Thread-8  :1763281 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/wandb-summary.json
+2022-03-26 03:08:38,226 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: poll_exit
+2022-03-26 03:08:38,841 INFO    SenderThread:1763281 [sender.py:transition_state():392] send defer: 5
+2022-03-26 03:08:38,842 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: poll_exit
+2022-03-26 03:08:38,842 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: defer
+2022-03-26 03:08:38,842 INFO    HandlerThread:1763281 [handler.py:handle_request_defer():154] handle defer: 5
+2022-03-26 03:08:38,843 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: defer
+2022-03-26 03:08:38,843 INFO    SenderThread:1763281 [sender.py:send_request_defer():388] handle sender defer: 5
+2022-03-26 03:08:38,843 INFO    SenderThread:1763281 [dir_watcher.py:finish():283] shutting down directory watcher
+2022-03-26 03:08:38,944 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: poll_exit
+2022-03-26 03:08:39,160 INFO    Thread-8  :1763281 [dir_watcher.py:_on_file_modified():230] file/dir modified: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/config.yaml
+2022-03-26 03:08:39,160 INFO    SenderThread:1763281 [dir_watcher.py:finish():313] scan: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files
+2022-03-26 03:08:39,161 INFO    SenderThread:1763281 [dir_watcher.py:finish():327] scan save: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/wandb-metadata.json wandb-metadata.json
+2022-03-26 03:08:39,161 INFO    SenderThread:1763281 [dir_watcher.py:finish():327] scan save: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/output.log output.log
+2022-03-26 03:08:39,161 INFO    SenderThread:1763281 [dir_watcher.py:finish():327] scan save: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/wandb-summary.json wandb-summary.json
+2022-03-26 03:08:39,161 INFO    SenderThread:1763281 [dir_watcher.py:finish():327] scan save: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/requirements.txt requirements.txt
+2022-03-26 03:08:39,161 INFO    SenderThread:1763281 [dir_watcher.py:finish():327] scan save: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/config.yaml config.yaml
+2022-03-26 03:08:39,164 INFO    SenderThread:1763281 [sender.py:transition_state():392] send defer: 6
+2022-03-26 03:08:39,164 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: poll_exit
+2022-03-26 03:08:39,165 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: defer
+2022-03-26 03:08:39,165 INFO    HandlerThread:1763281 [handler.py:handle_request_defer():154] handle defer: 6
+2022-03-26 03:08:39,166 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: defer
+2022-03-26 03:08:39,168 INFO    SenderThread:1763281 [sender.py:send_request_defer():388] handle sender defer: 6
+2022-03-26 03:08:39,174 INFO    SenderThread:1763281 [file_pusher.py:finish():177] shutting down file pusher
+2022-03-26 03:08:39,267 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: poll_exit
+2022-03-26 03:08:39,268 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: poll_exit
+2022-03-26 03:08:39,370 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: poll_exit
+2022-03-26 03:08:39,370 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: poll_exit
+2022-03-26 03:08:39,463 INFO    Thread-14 :1763281 [upload_job.py:push():137] Uploaded file /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/requirements.txt
+2022-03-26 03:08:39,472 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: poll_exit
+2022-03-26 03:08:39,472 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: poll_exit
+2022-03-26 03:08:39,486 INFO    Thread-13 :1763281 [upload_job.py:push():137] Uploaded file /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/wandb-summary.json
+2022-03-26 03:08:39,542 INFO    Thread-15 :1763281 [upload_job.py:push():137] Uploaded file /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/config.yaml
+2022-03-26 03:08:39,574 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: poll_exit
+2022-03-26 03:08:39,574 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: poll_exit
+2022-03-26 03:08:39,585 INFO    Thread-12 :1763281 [upload_job.py:push():137] Uploaded file /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/files/output.log
+2022-03-26 03:08:39,676 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: poll_exit
+2022-03-26 03:08:39,676 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: poll_exit
+2022-03-26 03:08:39,778 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: poll_exit
+2022-03-26 03:08:39,778 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: poll_exit
+2022-03-26 03:08:39,785 INFO    Thread-7  :1763281 [sender.py:transition_state():392] send defer: 7
+2022-03-26 03:08:39,786 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: defer
+2022-03-26 03:08:39,786 INFO    HandlerThread:1763281 [handler.py:handle_request_defer():154] handle defer: 7
+2022-03-26 03:08:39,786 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: defer
+2022-03-26 03:08:39,786 INFO    SenderThread:1763281 [sender.py:send_request_defer():388] handle sender defer: 7
+2022-03-26 03:08:39,880 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: poll_exit
+2022-03-26 03:08:40,306 INFO    SenderThread:1763281 [sender.py:transition_state():392] send defer: 8
+2022-03-26 03:08:40,306 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: poll_exit
+2022-03-26 03:08:40,307 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: defer
+2022-03-26 03:08:40,307 INFO    HandlerThread:1763281 [handler.py:handle_request_defer():154] handle defer: 8
+2022-03-26 03:08:40,307 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: defer
+2022-03-26 03:08:40,307 INFO    SenderThread:1763281 [sender.py:send_request_defer():388] handle sender defer: 8
+2022-03-26 03:08:40,307 INFO    SenderThread:1763281 [sender.py:transition_state():392] send defer: 9
+2022-03-26 03:08:40,308 DEBUG   SenderThread:1763281 [sender.py:send():235] send: final
+2022-03-26 03:08:40,309 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: defer
+2022-03-26 03:08:40,309 INFO    HandlerThread:1763281 [handler.py:handle_request_defer():154] handle defer: 9
+2022-03-26 03:08:40,309 DEBUG   SenderThread:1763281 [sender.py:send():235] send: footer
+2022-03-26 03:08:40,309 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: defer
+2022-03-26 03:08:40,309 INFO    SenderThread:1763281 [sender.py:send_request_defer():388] handle sender defer: 9
+2022-03-26 03:08:40,408 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: poll_exit
+2022-03-26 03:08:40,409 DEBUG   SenderThread:1763281 [sender.py:send_request():249] send_request: poll_exit
+2022-03-26 03:08:40,409 INFO    SenderThread:1763281 [file_pusher.py:join():182] waiting for file pusher
+2022-03-26 03:08:40,472 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: get_summary
+2022-03-26 03:08:40,554 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: sampled_history
+2022-03-26 03:08:40,557 DEBUG   HandlerThread:1763281 [handler.py:handle_request():131] handle_request: shutdown
+2022-03-26 03:08:40,557 INFO    HandlerThread:1763281 [handler.py:finish():739] shutting down handler
+2022-03-26 03:08:41,309 INFO    WriterThread:1763281 [datastore.py:close():281] close: /home/sanchit_huggingface_co/wav2vec2-2-bart-large-cnn/wandb/run-20220325_193848-1sz5964i/run-1sz5964i.wandb
+2022-03-26 03:08:41,471 INFO    SenderThread:1763281 [sender.py:finish():1075] shutting down sender
+2022-03-26 03:08:41,471 INFO    SenderThread:1763281 [file_pusher.py:finish():177] shutting down file pusher
+2022-03-26 03:08:41,471 INFO    SenderThread:1763281 [file_pusher.py:join():182] waiting for file pusher
+2022-03-26 03:08:41,477 INFO    MainThread:1763281 [internal.py:handle_exit():79] Internal process exited

wandb/run-20220325_193848-1sz5964i/logs/debug.log

@@ -25,3 +25,109 @@ config: {}
 2022-03-25 19:38:49,259 INFO    MainThread:1762967 [wandb_init.py:init():651] run started, returning control to user process
 2022-03-25 19:38:49,262 INFO    MainThread:1762967 [wandb_run.py:_config_callback():966] config_cb None None {'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': 'torch.float32', 'use_bfloat16': False, 'pruned_heads': {}, 'tie_word_embeddings': False, 'is_encoder_decoder': True, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'max_length': 40, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 5, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'chunk_size_feed_forward': 0, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'architectures': ['SpeechEncoderDecoderModel'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': None, 'pad_token_id': 1, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': 0, 'task_specific_params': None, 'problem_type': None, '_name_or_path': './', 'transformers_version': None, 'decoder': {'vocab_size': 50264, 'max_position_embeddings': 1024, 'd_model': 1024, 'encoder_ffn_dim': 4096, 'encoder_layers': 12, 'encoder_attention_heads': 16, 'decoder_ffn_dim': 4096, 'decoder_layers': 12, 'decoder_attention_heads': 16, 'dropout': 0.1, 'attention_dropout': 0.0, 'activation_dropout': 0.0, 'activation_function': 'gelu', 'init_std': 0.02, 'encoder_layerdrop': 0.0, 'decoder_layerdrop': 0.0, 'classifier_dropout': 0.0, 'use_cache': True, 'num_hidden_layers': 12, 'scale_embedding': False, 'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': None, 'use_bfloat16': False, 'pruned_heads': {}, 'tie_word_embeddings': True, 'is_encoder_decoder': False, 'is_decoder': True, 'cross_attention_hidden_size': None, 'add_cross_attention': True, 'tie_encoder_decoder': False, 'max_length': 142, 'min_length': 56, 'do_sample': False, 'early_stopping': True, 'num_beams': 4, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 2.0, 'no_repeat_ngram_size': 3, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'chunk_size_feed_forward': 0, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': 0, 'forced_eos_token_id': 2, 'remove_invalid_values': False, 'architectures': ['BartForConditionalGeneration'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1', 2: 'LABEL_2'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1, 'LABEL_2': 2}, 'tokenizer_class': None, 'prefix': ' ', 'bos_token_id': 0, 'pad_token_id': 1, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': 2, 'task_specific_params': {'summarization': {'early_stopping': True, 'length_penalty': 2.0, 'max_length': 142, 'min_length': 56, 'no_repeat_ngram_size': 3, 'num_beams': 4}}, 'problem_type': None, '_name_or_path': 'facebook/bart-large-cnn', 'transformers_version': '4.17.0.dev0', '_num_labels': 3, 'add_final_layer_norm': False, 'classif_dropout': 0.0, 'force_bos_token_to_be_generated': True, 'gradient_checkpointing': False, 'normalize_before': False, 'output_past': True, 'model_type': 'bart'}, 'encoder': {'return_dict': True, 'output_hidden_states': False, 'output_attentions': False, 'torchscript': False, 'torch_dtype': None, 'use_bfloat16': False, 'pruned_heads': {}, 'tie_word_embeddings': True, 'is_encoder_decoder': False, 'is_decoder': False, 'cross_attention_hidden_size': None, 'add_cross_attention': False, 'tie_encoder_decoder': False, 'max_length': 20, 'min_length': 0, 'do_sample': False, 'early_stopping': False, 'num_beams': 1, 'num_beam_groups': 1, 'diversity_penalty': 0.0, 'temperature': 1.0, 'top_k': 50, 'top_p': 1.0, 'repetition_penalty': 1.0, 'length_penalty': 1.0, 'no_repeat_ngram_size': 0, 'encoder_no_repeat_ngram_size': 0, 'bad_words_ids': None, 'num_return_sequences': 1, 'chunk_size_feed_forward': 0, 'output_scores': False, 'return_dict_in_generate': False, 'forced_bos_token_id': None, 'forced_eos_token_id': None, 'remove_invalid_values': False, 'architectures': ['Wav2Vec2ForPreTraining'], 'finetuning_task': None, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'tokenizer_class': None, 'prefix': None, 'bos_token_id': 1, 'pad_token_id': 0, 'eos_token_id': 2, 'sep_token_id': None, 'decoder_start_token_id': None, 'task_specific_params': None, 'problem_type': None, '_name_or_path': 'facebook/wav2vec2-large-lv60', 'transformers_version': '4.17.0.dev0', 'feat_extract_dropout': 0.0, 'gradient_checkpointing': False, 'hidden_dropout_prob': 0.1, 'num_feat_extract_layers': 7, 'hidden_size': 1024, 'feat_extract_norm': 'layer', 'feat_extract_activation': 'gelu', 'conv_dim': [512, 512, 512, 512, 512, 512, 512], 'conv_stride': [5, 2, 2, 2, 2, 2, 2], 'conv_kernel': [10, 3, 3, 3, 3, 2, 2], 'conv_bias': True, 'num_conv_pos_embeddings': 128, 'num_conv_pos_embedding_groups': 16, 'num_hidden_layers': 24, 'intermediate_size': 4096, 'hidden_act': 'gelu', 'num_attention_heads': 16, 'hidden_dropout': 0.1, 'attention_dropout': 0.1, 'activation_dropout': 0.1, 'feat_proj_dropout': 0.0, 'final_dropout': 0.0, 'layerdrop': 0.0, 'layer_norm_eps': 1e-05, 'initializer_range': 0.02, 'vocab_size': 32, 'do_stable_layer_norm': True, 'use_weighted_layer_sum': False, 'apply_spec_augment': True, 'mask_time_prob': 0.1, 'mask_time_length': 10, 'mask_time_min_masks': 2, 'mask_feature_prob': 0.0, 'mask_feature_length': 10, 'mask_feature_min_masks': 0, 'num_codevectors_per_group': 320, 'num_codevector_groups': 2, 'contrastive_logits_temperature': 0.1, 'feat_quantizer_dropout': 0.0, 'num_negatives': 100, 'codevector_dim': 768, 'proj_codevector_dim': 768, 'diversity_loss_weight': 0.1, 'ctc_loss_reduction': 'sum', 'ctc_zero_infinity': False, 'add_adapter': True, 'adapter_kernel_size': 3, 'adapter_stride': 2, 'num_adapter_layers': 3, 'output_hidden_size': 1024, 'classifier_proj_size': 256, 'tdnn_dim': [512, 512, 512, 512, 1500], 'tdnn_kernel': [5, 3, 3, 1, 1], 'tdnn_dilation': [1, 2, 3, 1, 1], 'xvector_output_dim': 512, 'model_type': 'wav2vec2'}, 'model_type': 'speech-encoder-decoder', 'processor_class': 'Wav2Vec2Processor', 'use_cache': False, 'output_dir': './', 'overwrite_output_dir': True, 'do_train': True, 'do_eval': True, 'do_predict': False, 'evaluation_strategy': 'steps', 'prediction_loss_only': False, 'per_device_train_batch_size': 8, 'per_device_eval_batch_size': 8, 'per_gpu_train_batch_size': 'None', 'per_gpu_eval_batch_size': 'None', 'gradient_accumulation_steps': 16, 'eval_accumulation_steps': 'None', 'learning_rate': 0.0003, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'max_grad_norm': 1.0, 'num_train_epochs': 5.0, 'max_steps': -1, 'lr_scheduler_type': 'linear', 'warmup_ratio': 0.0, 'warmup_steps': 500, 'log_level': -1, 'log_level_replica': -1, 'log_on_each_node': True, 'logging_dir': './runs/Mar25_19-38-20_sanchit--v100', 'logging_strategy': 'steps', 'logging_first_step': False, 'logging_steps': 1, 'logging_nan_inf_filter': True, 'save_strategy': 'steps', 'save_steps': 500, 'save_total_limit': 1, 'save_on_each_node': False, 'no_cuda': False, 'seed': 42, 'bf16': False, 'fp16': True, 'fp16_opt_level': 'O1', 'half_precision_backend': 'amp', 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': 'None', 'local_rank': -1, 'xpu_backend': 'None', 'tpu_num_cores': 'None', 'tpu_metrics_debug': False, 'debug': '[]', 'dataloader_drop_last': False, 'eval_steps': 500, 'dataloader_num_workers': 0, 'past_index': -1, 'run_name': './', 'disable_tqdm': False, 'remove_unused_columns': True, 'label_names': 'None', 'load_best_model_at_end': False, 'metric_for_best_model': 'None', 'greater_is_better': 'None', 'ignore_data_skip': False, 'sharded_ddp': '[]', 'deepspeed': 'None', 'label_smoothing_factor': 0.0, 'optim': 'adamw_hf', 'adafactor': False, 'group_by_length': True, 'length_column_name': 'input_length', 'report_to': "['tensorboard', 'wandb']", 'ddp_find_unused_parameters': 'None', 'ddp_bucket_cap_mb': 'None', 'dataloader_pin_memory': True, 'skip_memory_metrics': True, 'use_legacy_prediction_loop': False, 'push_to_hub': True, 'resume_from_checkpoint': 'None', 'hub_model_id': 'None', 'hub_strategy': 'every_save', 'hub_token': '<HUB_TOKEN>', 'gradient_checkpointing': True, 'fp16_backend': 'auto', 'push_to_hub_model_id': 'None', 'push_to_hub_organization': 'None', 'push_to_hub_token': '<PUSH_TO_HUB_TOKEN>', '_n_gpu': 1, 'mp_parameters': '', 'sortish_sampler': False, 'predict_with_generate': True, 'generation_max_length': 40, 'generation_num_beams': 1, 'train_batch_size': 8, 'eval_batch_size': 8}
 2022-03-25 19:38:49,265 INFO    MainThread:1762967 [wandb_watch.py:watch():43] Watching
+2022-03-26 03:08:35,667 INFO    MainThread:1762967 [wandb_run.py:_atexit_cleanup():1797] got exitcode: 1
+2022-03-26 03:08:35,670 INFO    MainThread:1762967 [wandb_run.py:_restore():1769] restore
+2022-03-26 03:08:37,911 INFO    MainThread:1762967 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
+  wandb_count: 1
+}
+pusher_stats {
+  uploaded_bytes: 2016
+  total_bytes: 2016
+}
+
+2022-03-26 03:08:38,125 INFO    MainThread:1762967 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
+  wandb_count: 1
+}
+pusher_stats {
+  uploaded_bytes: 2016
+  total_bytes: 2016
+}
+
+2022-03-26 03:08:38,842 INFO    MainThread:1762967 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
+  wandb_count: 1
+}
+pusher_stats {
+  uploaded_bytes: 2016
+  total_bytes: 2016
+}
+
+2022-03-26 03:08:39,166 INFO    MainThread:1762967 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
+  wandb_count: 3
+}
+pusher_stats {
+  uploaded_bytes: 2016
+  total_bytes: 5286519
+}
+
+2022-03-26 03:08:39,269 INFO    MainThread:1762967 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
+  wandb_count: 5
+}
+pusher_stats {
+  uploaded_bytes: 10208
+  total_bytes: 5546341
+}
+
+2022-03-26 03:08:39,371 INFO    MainThread:1762967 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
+  wandb_count: 5
+}
+pusher_stats {
+  uploaded_bytes: 5546341
+  total_bytes: 5546341
+}
+
+2022-03-26 03:08:39,473 INFO    MainThread:1762967 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
+  wandb_count: 5
+}
+pusher_stats {
+  uploaded_bytes: 5546341
+  total_bytes: 5546341
+}
+
+2022-03-26 03:08:39,575 INFO    MainThread:1762967 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
+  wandb_count: 5
+}
+pusher_stats {
+  uploaded_bytes: 5546341
+  total_bytes: 5546341
+}
+
+2022-03-26 03:08:39,677 INFO    MainThread:1762967 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
+  wandb_count: 5
+}
+pusher_stats {
+  uploaded_bytes: 5546341
+  total_bytes: 5546341
+}
+
+2022-03-26 03:08:39,779 INFO    MainThread:1762967 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
+  wandb_count: 5
+}
+pusher_stats {
+  uploaded_bytes: 5546341
+  total_bytes: 5546341
+}
+
+2022-03-26 03:08:40,307 INFO    MainThread:1762967 [wandb_run.py:_wait_for_finish():1929] got exit ret: file_counts {
+  wandb_count: 5
+}
+pusher_stats {
+  uploaded_bytes: 5546341
+  total_bytes: 5546341
+}
+
+2022-03-26 03:08:40,471 INFO    MainThread:1762967 [wandb_run.py:_wait_for_finish():1929] got exit ret: done: true
+exit_result {
+}
+file_counts {
+  wandb_count: 5
+}
+pusher_stats {
+  uploaded_bytes: 5546341
+  total_bytes: 5546341
+}
+local_info {
+}
+
+2022-03-26 03:08:41,622 INFO    MainThread:1762967 [wandb_run.py:_append_history():2144] rendering history
+2022-03-26 03:08:41,623 INFO    MainThread:1762967 [wandb_run.py:_append_summary():2102] rendering summary
+2022-03-26 03:08:41,624 INFO    MainThread:1762967 [wandb_run.py:_append_files():2194] logging synced files

wandb/run-20220325_193848-1sz5964i/run-1sz5964i.wandb

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a1516c27822fb70ec66bc1f8a81e39aec7d8372a24841988e1b148b6eb040967
-size 228560954
+oid sha256:5d9276372d0c5c37f19459a87218cc873bf5a347ff167725dcc4dea29432bc9d
+size 229877308