Bug fix core workflow #71
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Collaborator
akgohain
commented
Nov 24, 2025
akgohain
commented
- more robust file handling
- more robust file upload preview containers
- fix inference backend
- bump pytc dependcies
- fix UI elts of inference (e.g. no need for input LABEL for inference)
- fix tensorboard implementation
- Added torch>=2.0.0 and torchvision>=0.15.0 to pyproject.toml - Fixes ModuleNotFoundError during training execution
- Use sys.executable instead of hardcoded 'python' command - Use absolute paths for pytorch_connectomics script - Add subprocess output capture with threading - Stream training logs to terminal with [TRAINING:PID] prefix - Fixes training script execution failures
- Add [SERVER_API], [SERVER_PYTC], [MODEL.PY] prefixed logs - Add startup logging for debugging server initialization - Add detailed error handling with full stack traces - Improves debugging and monitoring of training workflow
- Add /training_status endpoint to check if training is running - Implement frontend polling every 2 seconds - Auto-detect training completion and update UI - Add proper validation for outputPath and logPath - Fixes UI hanging issue when training completes
- Parse YAML config and inject DATASET.OUTPUT_PATH from UI - Use js-yaml to modify config before sending to backend - Ensures training outputs go to user-specified path - Fixes hardcoded output path issue
- Replace default upload list with custom compact layout - Add small thumbnails (32px) next to filenames - Implement scrollable container (maxHeight: 30vh) - Improve visual hierarchy and spacing
…s logs
PyTorch Connectomics writes training logs to {OUTPUT_PATH}/log{timestamp}/
not to a separate log directory. This change:
- Frontend now sends outputPath in request payload
- TensorBoard initializes monitoring outputPath instead of logPath
- TensorBoard automatically finds event files in subdirectories
- Add comprehensive logging to show which paths are being used
This ensures TensorBoard displays actual training metrics instead of
monitoring an empty directory.
…kpoint/output path handling
sam4coding
pushed a commit
to sam4coding/pytc-client
that referenced
this pull request
Dec 11, 2025
…g-fix-core-workflow Bug fix core workflow
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.