Documentation Index

Fetch the complete documentation index at: https://docs.cirron.com/llms.txt

Use this file to discover all available pages before exploring further.

​

cirron compile

​

Usage

cirron compile [options]

​

Options

​

Architecture Support

​

CPU Architecture

cirron compile --arch cpu

• Optimized for CPU inference
• Compatible with all frameworks
• Lightweight deployment

​

CUDA Architecture (PyTorch)

cirron compile --arch cuda

• GPU-accelerated inference for PyTorch
• Requires CUDA-compatible hardware
• Validates CUDA availability

​

GPU Architecture (TensorFlow)

cirron compile --arch gpu

• GPU-accelerated inference for TensorFlow
• Requires GPU-compatible hardware
• Validates GPU availability

​

Framework Support

​

PyTorch

# CPU compilation
cirron compile --arch cpu

# CUDA compilation
cirron compile --arch cuda

• Loads model from src/model.py
• Optimizes for target architecture
• Saves compiled model to models/model_{arch}.pth
• Generates model metadata

​

TensorFlow

# CPU compilation
cirron compile --arch cpu

# GPU compilation
cirron compile --arch gpu

• Loads model from src/model.py
• Configures device placement
• Saves compiled model to models/model_{arch}
• Generates model metadata

​

Scikit-Learn

cirron compile --arch cpu

• Loads model from src/model.py
• Serializes model using joblib
• Saves compiled model to models/model_{arch}.joblib
• Generates model metadata

​

Validation

​

Pre-compilation Validation

cirron compile --validate

• Required files (src/model.py, requirements.txt)
• Python version compatibility
• Architecture-specific requirements
• Model creation and instantiation
• Index file format (if provided)

​

Strict Mode

cirron compile --strict --validate

• Fails fast on validation errors
• Provides detailed error messages
• Suggests recovery actions

​

Index Files

​

JSON Index

{
  "features": ["feature1", "feature2", "feature3"],
  "dataTypes": {
    "feature1": "float",
    "feature2": "int",
    "feature3": "string"
  }
}

​

YAML Index

features:
  - feature1
  - feature2
  - feature3
dataTypes:
  feature1: float
  feature2: int
  feature3: string

cirron compile --index config/index.json
cirron compile --index config/index.yaml

​

Examples

​

Basic Compilation

# Auto-detect architecture
cirron compile

# Specify CPU architecture
cirron compile --arch cpu

​

Validated Compilation

# Run with validation
cirron compile --validate

# Strict mode with validation
cirron compile --strict --validate

​

GPU Compilation

# PyTorch with CUDA
cirron compile --arch cuda

# TensorFlow with GPU
cirron compile --arch gpu

​

With Index File

# Compile with custom index
cirron compile --index data/index.json

# Full validation with index
cirron compile --validate --index data/index.yaml

​

Output

​

Generated Artifacts

models/
├── model_cpu.pth          # PyTorch CPU model
├── model_cuda.pth         # PyTorch CUDA model
└── model_gpu              # TensorFlow GPU model

artifacts/
├── model_info.json        # Model metadata
└── index_config.json      # Index configuration (if provided)

​

Model Information

{
  "framework": "pytorch",
  "architecture": "cuda",
  "parameters": 1234567,
  "compilation_time": "2024-01-15T10:30:00Z"
}

​

Error Handling

​

Common Errors

​

Project Not Found

Error: No cirron.yaml found
Run cirron init to initialize a project

​

Validation Failed

Error: Validation failed:
  • Required file missing: src/model.py
  • Python 3.9+ required, found 3.7
  • CUDA not available for PyTorch

​

Compilation Failed

Error: Compilation failed
Check compilation logs for specific errors

​

Recovery Actions

1. Missing Project Configuration
   cirron init my-project
   
2. Validation Errors
   # Fix missing files
   touch src/model.py
   
   # Update Python version
   python3 --version
   
   # Check CUDA installation
   nvidia-smi
   
3. Compilation Errors
   # Check model code
   python3 src/model.py
   
   # Verify dependencies
   pip install -r requirements.txt
   

​

Integration

​

CI/CD Pipeline

# GitHub Actions example
- name: Compile Model
  run: |
    cirron compile --validate --arch cpu
    cirron compile --validate --arch cuda

​

Docker Integration

# Multi-stage build
FROM python:3.9 as compile
COPY . /app
WORKDIR /app
RUN cirron compile --arch cpu

FROM python:3.9-slim
COPY --from=compile /app/models /app/models
COPY --from=compile /app/artifacts /app/artifacts

​

Troubleshooting

​

Performance Issues

• Use appropriate architecture for your hardware
• Enable GPU compilation for large models
• Check memory requirements

​

Validation Issues

• Ensure all required files exist
• Verify Python version compatibility
• Check framework-specific requirements

​

Compilation Failures

• Review model code for errors
• Check dependency conflicts
• Verify hardware compatibility