Training in progress, step 580

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:796907df6dad182b13ce60cc0ccb07c112a82034aaa5441dd30b61e18d46838c
+oid sha256:2b1967693c07100a438639df88eb53656c00dbba558ea2c42ffa73fba1446eee
 size 13982248

runs/Apr12_07-40-41_e5a48bec8248/events.out.tfevents.1712907642.e5a48bec8248.443.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6cb6884880e48c3450b5e32fccc7101462e6b76a60820dd08b937b4a547eb9ff
-size 11166
+oid sha256:571ae048c293de8c9d2f4fa75684f3964294c74bb8580ce257c1b6ee278e3561
+size 11377

wandb/debug-internal.log

@@ -530,3 +530,18 @@ subprocess.TimeoutExpired: Command '['conda', 'env', 'export']' timed out after
 2024-04-12 07:52:34,132 DEBUG   SenderThread:485 [sender.py:send_request():406] send_request: summary_record
 2024-04-12 07:52:34,132 INFO    SenderThread:485 [sender.py:_save_file():1390] saving file wandb-summary.json with policy end
 2024-04-12 07:52:34,237 INFO    Thread-12 :485 [dir_watcher.py:_on_file_modified():288] file/dir modified: /kaggle/working/wandb/run-20240412_074044-widovcn3/files/wandb-summary.json
+2024-04-12 07:52:37,238 INFO    Thread-12 :485 [dir_watcher.py:_on_file_modified():288] file/dir modified: /kaggle/working/wandb/run-20240412_074044-widovcn3/files/output.log
+2024-04-12 07:52:37,867 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: status_report
+2024-04-12 07:52:42,868 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: status_report
+2024-04-12 07:52:45,040 DEBUG   SenderThread:485 [sender.py:send():379] send: stats
+2024-04-12 07:52:46,049 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: internal_messages
+2024-04-12 07:52:46,050 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: stop_status
+2024-04-12 07:52:46,050 DEBUG   SenderThread:485 [sender.py:send_request():406] send_request: stop_status
+2024-04-12 07:52:48,079 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: status_report
+2024-04-12 07:52:53,080 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: status_report
+2024-04-12 07:52:57,000 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: partial_history
+2024-04-12 07:52:57,001 DEBUG   SenderThread:485 [sender.py:send():379] send: history
+2024-04-12 07:52:57,002 DEBUG   SenderThread:485 [sender.py:send_request():406] send_request: summary_record
+2024-04-12 07:52:57,002 INFO    SenderThread:485 [sender.py:_save_file():1390] saving file wandb-summary.json with policy end
+2024-04-12 07:52:57,246 INFO    Thread-12 :485 [dir_watcher.py:_on_file_modified():288] file/dir modified: /kaggle/working/wandb/run-20240412_074044-widovcn3/files/wandb-summary.json
+2024-04-12 07:52:58,985 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: status_report

wandb/run-20240412_074044-widovcn3/files/output.log

@@ -52,4 +52,6 @@
   warnings.warn(
 /opt/conda/lib/python3.10/site-packages/torch/utils/checkpoint.py:429: UserWarning: torch.utils.checkpoint: please pass in use_reentrant=True or use_reentrant=False explicitly. The default value of use_reentrant will be updated to be False in the future. To maintain current behavior, pass use_reentrant=True. It is recommended that you use use_reentrant=False. Refer to docs for more details on the differences between the two variants.
   warnings.warn(
+/opt/conda/lib/python3.10/site-packages/torch/utils/checkpoint.py:429: UserWarning: torch.utils.checkpoint: please pass in use_reentrant=True or use_reentrant=False explicitly. The default value of use_reentrant will be updated to be False in the future. To maintain current behavior, pass use_reentrant=True. It is recommended that you use use_reentrant=False. Refer to docs for more details on the differences between the two variants.
+  warnings.warn(
 /opt/conda/lib/python3.10/site-packages/torch/utils/checkpoint.py:429: UserWarning: torch.utils.checkpoint: please pass in use_reentrant=True or use_reentrant=False explicitly. The default value of use_reentrant will be updated to be False in the future. To maintain current behavior, pass use_reentrant=True. It is recommended that you use use_reentrant=False. Refer to docs for more details on the differences between the two variants.

wandb/run-20240412_074044-widovcn3/files/wandb-summary.json

@@ -1 +1 @@
-{"train/loss": 3.8248, "train/grad_norm": 23.113840103149414, "train/learning_rate": 2.8e-05, "train/epoch": 0.04, "train/global_step": 560, "_timestamp": 1712908354.1302218, "_runtime": 709.6225037574768, "_step": 27}
+{"train/loss": 3.7538, "train/grad_norm": 21.58758544921875, "train/learning_rate": 2.9e-05, "train/epoch": 0.05, "train/global_step": 580, "_timestamp": 1712908376.9997563, "_runtime": 732.4920382499695, "_step": 28}

wandb/run-20240412_074044-widovcn3/logs/debug-internal.log

@@ -530,3 +530,18 @@ subprocess.TimeoutExpired: Command '['conda', 'env', 'export']' timed out after
 2024-04-12 07:52:34,132 DEBUG   SenderThread:485 [sender.py:send_request():406] send_request: summary_record
 2024-04-12 07:52:34,132 INFO    SenderThread:485 [sender.py:_save_file():1390] saving file wandb-summary.json with policy end
 2024-04-12 07:52:34,237 INFO    Thread-12 :485 [dir_watcher.py:_on_file_modified():288] file/dir modified: /kaggle/working/wandb/run-20240412_074044-widovcn3/files/wandb-summary.json
+2024-04-12 07:52:37,238 INFO    Thread-12 :485 [dir_watcher.py:_on_file_modified():288] file/dir modified: /kaggle/working/wandb/run-20240412_074044-widovcn3/files/output.log
+2024-04-12 07:52:37,867 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: status_report
+2024-04-12 07:52:42,868 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: status_report
+2024-04-12 07:52:45,040 DEBUG   SenderThread:485 [sender.py:send():379] send: stats
+2024-04-12 07:52:46,049 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: internal_messages
+2024-04-12 07:52:46,050 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: stop_status
+2024-04-12 07:52:46,050 DEBUG   SenderThread:485 [sender.py:send_request():406] send_request: stop_status
+2024-04-12 07:52:48,079 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: status_report
+2024-04-12 07:52:53,080 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: status_report
+2024-04-12 07:52:57,000 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: partial_history
+2024-04-12 07:52:57,001 DEBUG   SenderThread:485 [sender.py:send():379] send: history
+2024-04-12 07:52:57,002 DEBUG   SenderThread:485 [sender.py:send_request():406] send_request: summary_record
+2024-04-12 07:52:57,002 INFO    SenderThread:485 [sender.py:_save_file():1390] saving file wandb-summary.json with policy end
+2024-04-12 07:52:57,246 INFO    Thread-12 :485 [dir_watcher.py:_on_file_modified():288] file/dir modified: /kaggle/working/wandb/run-20240412_074044-widovcn3/files/wandb-summary.json
+2024-04-12 07:52:58,985 DEBUG   HandlerThread:485 [handler.py:handle_request():146] handle_request: status_report