Teemo Ops - ML Training Resource Analyzer
Estimate, Optimize, and Manage your ML Training resources directly in VS Code.
Features
Teemo Ops analyzes your machine learning training scripts and provides:
- Resource Estimation: GPU requirements, memory usage, and compute needs
- Cost Analysis: Estimated training costs across different cloud providers
- Training Time Predictions: How long your training will take
- Hardware Recommendations: Optimal GPU/CPU configurations
- Carbon Footprint: Environmental impact estimates
- Optimization Suggestions: Tips to reduce costs and training time
Requirements
- VS Code 1.74.0 or higher
- Internet connection (to reach Teemo Ops API)
- Python ML training scripts
Usage
- Open a Python ML training script in VS Code
- Right-click in the editor
- Select "Teemo Ops: Analyze ML Training Script"
- View results in the Teemo Ops sidebar
Or use the Command Palette:
- Press
Ctrl+Shift+P (Windows/Linux) or Cmd+Shift+P (Mac)
- Type "Teemo Ops: Analyze"
- Press Enter
Supported Frameworks
- PyTorch
- TensorFlow
- Transformers (Hugging Face)
- PEFT/LoRA
- Custom training loops
What Gets Analyzed
The extension analyzes:
- Model architecture and size
- Training hyperparameters (batch size, epochs, learning rate)
- Dataset size and preprocessing
- Hardware configuration (GPUs, memory)
- Distributed training setup
Extension Settings
This extension contributes the following settings:
teemo-ops.apiUrl: Custom API endpoint (default: public Teemo Ops API)teemo-ops.timeout: Request timeout in seconds (default: 600)
Privacy
- Your scripts are sent to the Teemo Ops API for analysis
- Scripts are temporarily stored in Google Cloud Storage for processing
- No data is permanently retained after analysis
- No telemetry or tracking is performed
Known Issues
- Very large scripts (>100KB) may take longer to analyze
- Some custom training configurations may not be fully recognized
Release Notes
1.0.1
Initial release of Teemo Ops:
- ML training script analysis
- Resource estimation
- Cost predictions
- Hardware recommendations
- Carbon footprint estimates
Support
License
MIT
Enjoy optimizing your ML training! 🎯
| |
