Nodes crash monitoring

[qchapp]

This pull request updates documentation, configuration files, and scripts to improve usability, clarify execution modes, and simplify running the MMIRAGE pipeline. The most important changes are grouped below by theme.

Documentation and Usage Improvements:

• Expanded the README.md with clearer instructions for running the pipeline, checking status, and retrying failed shards, including new CLI command examples and explanations of retry behavior.
• Updated YAML configuration examples in the README.md to include state_dir and execution_params, and clarified how to control sharding and retries. [1] [2] [3] [4]

Nodes crash monitoring:

• Shards status are now tracked in a status.json custom file
• Shards failed can be automatically relaunched via a specific config parameter retry=True

Configuration Enhancements:

• Added a comprehensive example config file configs/config_comprehensive.yaml documenting all available parameters, including detailed comments for each section.
• Updated mock config files (configs/config_mock.yaml, configs/config_mock_vision.yaml) to include state_dir, set num_shards to 1 for local runs, and added execution_params for explicit execution mode and retry control. [1] [2] [3] [4]

Script and CLI Improvements:

• Deleted run.sh to delegate all execution logic to the Python CLI, removing direct SLURM commands and making the script config-driven.
• Registered a mmirage CLI entrypoint in pyproject.toml for easier command-line usage.
• Jobs can be launched using the run entrypoint
• Shards status can be checked using check entrypoint

Codebase Organization:

• Updated imports in src/mmirage/__init__.py to reflect module restructuring and ensure all relevant configuration classes are exposed at the package level.

[Copilot]

Pull request overview

Copilot reviewed 23 out of 24 changed files in this pull request and generated 4 comments.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 23 out of 24 changed files in this pull request and generated 1 comment.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.