DNALLM provides a comprehensive set of command-line interface tools for various DNA language model tasks.
dnallm train- Fine-tune DNA language modelsdnallm inference- Run inference with trained modelsdnallm benchmark- Compare multiple modelsdnallm mutagenesis- Perform in silico mutagenesis
dnallm mcp-server- Start MCP (Model Context Protocol) serverdnallm-mcp-server- Standalone MCP server script
dnallm model-config-generator- Interactive configuration file generator
# Generate a configuration file
dnallm model-config-generator --output finetune_config.yaml
# Train a model
dnallm train --config finetune_config.yaml
# Run inference
dnallm inference --config inference_config.yaml --model-path ./models/trained_model
# Benchmark models
dnallm benchmark --config benchmark_config.yaml
# Start MCP server
dnallm mcp-server --config mcp_server_config.yamlAll DNALLM commands use YAML configuration files that define:
- Task parameters (classification, regression, generation, etc.)
- Model settings and hyperparameters
- Data preprocessing options
- Evaluation metrics and output formats
# Show help for a specific command
dnallm train --help
# Show general help
dnallm --help- Usage Guide - Complete CLI usage instructions and examples
- Configuration Generator - Learn how to create configuration files
- MCP Server - Learn about the Model Context Protocol server
- Fine-tuning Tutorials - Learn to train models
- Benchmark Tutorials - Compare model performance
- CLI Troubleshooting - Common CLI issues and solutions