feat: refactor metrics and switch to ruff

[CuriousDolphin]

Enhance Model Training Monitoring and Integrate Ruff

CONTEXT

This PR was created to modernize the development toolchain and improve the model training monitoring capabilities. The existing linting tools were fragmented, and the model training lacked comprehensive monitoring and visualization features.

KEY CHANGES

• Development Tools

  • Replaced multiple linting tools (isort, black, flake8) with Ruff for unified linting and formatting.
  • Updated pre-commit hooks to include Ruff and removed redundant hooks.
  • Updated Makefile to streamline virtual environment creation and linting processes.

• Model Training & Monitoring

  • Introduced MetricsVisualizer class for enhanced training metrics visualization.
  • Refactored RemoteModel class to improve training monitoring.
  • Added notebook_monitor_train method for real-time training monitoring in Jupyter notebooks.
  • Removed deprecated anyma_version parameter from model training configuration.

• Code Quality & Structure

  • Refactored code for better readability and consistency.
  • Improved error handling and logging across various modules.
  • Enhanced test coverage with new test cases for MetricsVisualizer.

IMPACT

• Development Workflow: Developers need to update their local pre-commit hooks and VSCode settings.
• Code Quality: Global impact on code style and formatting due to the adoption of Ruff.
• Training Pipeline: Enhanced monitoring capabilities for model training, especially in Jupyter notebooks.
• Testing: Improved test coverage and readability.

TECHNICAL DETAILS

Development Tools Configuration

Updated .pre-commit-config.yaml to use Ruff

New MetricsVisualizer Implementation

The MetricsVisualizer class provides methods for logging and visualizing training metrics:

class MetricsVisualizer:
def __init__(metrics:Metrics):
      pass
def log_training_metrics(self, metrics: Dict[str, float]) -> None:
"""Log training metrics for visualization"""
pass
def plot_metrics(self, metric_names: List[str]) -> None:
"""Visualize specified metrics over time"""
pass

RemoteModel Enhancements

Refactored RemoteModel class to improve monitoring:

class RemoteModel:
def train_info(self) -> Optional[TrainingInfo]:
"""Returns current training status and progress"""
pass
def notebook_monitor_train(self) -> None:
"""Provides real-time training monitoring in Jupyter"""
pass

Key Dependencies Updates

• Added: opencv-python, matplotlib, colorama, ipython
• Removed: isort, black, flake8
• Updated: Various dependency versions

These changes aim to streamline the development process and enhance the model training experience with better monitoring and visualization tools.

[github-actions]

Coverage Report •

File	Stmts	Miss	Cover	Missing
__init__.py	10	0	100%
focoos.py	115	0	100%
local_model.py	61	0	100%
ports.py	230	0	100%
remote_model.py	140	36	74%	227–229, 234, 238–239, 241–242, 246–247, 275–283, 285–288, 292–295, 299–304, 306–307, 364
runtime.py	138	107	22%	118–128, 130–138, 140–144, 155–157, 168–174, 186, 188, 190, 192–200, 202–211, 219–222, 224, 226, 239–244, 246, 255–258, 260, 262–263, 276–277, 279–284, 286, 295–296, 298, 300–304, 306–308, 310–315, 317–318, 328–329
utils
logger.py	32	0	100%
metrics.py	90	0	100%
system.py	80	5	93%	218–222
vision.py	73	0	100%
TOTAL	981	148	84%

Tests	Skipped	Failures	Errors	Time
100	0 💤	0 ❌	0 🔥	4.770s ⏱️

[fcdl94]

Cosa cambia lato users? Solo la parte di metriche, giusto?

Abbiamo aggiornato la doc in modo che sia allineata?