Skip to content

Conversation

@bradhilton
Copy link
Collaborator

No description provided.

- Set execution counts for code cells in `run.ipynb`.
- Added HTML output styling for notebook display.
- Updated model name to "003" and base model to "Qwen/Qwen2.5-32B-Instruct".
- Adjusted training configuration parameters: reduced trajectories per group, increased groups per step, modified learning rate, and changed evaluation steps and validation set size.
- Enabled evaluation during training by setting `skip_eval` to false.
- Updated GPU memory utilization setting.
- Removed unnecessary dependencies from `pyproject.toml` related to 'swebench' and added 'ruff' for development.
- Set execution counts to null for code cells in `run.ipynb`.
- Removed unnecessary HTML output styling from the notebook.
- Set execution counts for code cells to reflect the current run order.
- Added HTML output for better display of WandB tracking information and model loading status.
- Adjusted GPU memory utilization setting for improved performance during training.
@bradhilton bradhilton marked this pull request as ready for review July 9, 2025 19:01
@bradhilton bradhilton merged commit 6e455bc into main Jul 9, 2025
2 checks passed
@bradhilton bradhilton deleted the feat/tau-bench-brad-003 branch July 9, 2025 19:01
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants