Running the Agent¶
The main entry point for Agentomics-ML is run.sh. This guide covers both interactive and non-interactive usage.
Interactive Mode¶
Running without arguments launches interactive mode:
Docker mode expects a .env file in the repo root (copy .env.example).
You'll be prompted to select:
- LLM Model - Choose from available models
- Dataset - Select a prepared dataset
- Iterations - Number of optimization cycles (default prompt: 5)
- Validation Metric - Optional metric to optimize (defaults:
AUROCfor classification,MAEfor regression)
Non-Interactive Mode¶
Supply parameters directly to skip prompts:
For non-interactive runs, provide at least --model, --dataset, and --iterations.
Common Options¶
| Option | Description | Example |
|---|---|---|
--model |
LLM model to use | --model openai/gpt-4 |
--dataset |
Dataset name | --dataset my_data |
--iterations |
Number of iterations | --iterations 15 |
--val-metric |
Validation metric (optional, task-based default if omitted) | --val-metric AUROC |
--timeout |
Time limit in seconds | --timeout 3600 |
--run-python-timeout |
Timeout in seconds for each run_python tool execution (see CLI options) | --run-python-timeout 43200 |
--use-provisioning-key |
Use a provisioning key for OpenRouter | --use-provisioning-key |
--spend-limit |
Spend limit for provisioning key | --spend-limit 25 |
Listing Available Options¶
# List available models
./run.sh --list-models
# List prepared datasets
./run.sh --list-datasets
# List available metrics
./run.sh --list-metrics
Deployment Flags¶
| Flag | Description |
|---|---|
--pull-images |
Pull pre-built Docker images |
--local |
Run without Docker (uses conda) |
--cpu-only |
Disable GPU acceleration |
--ollama |
Use local Ollama models |
Advanced Options¶
Foundation Models¶
Pre-download domain-specific foundation models:
Available types: dna, rna, protein, molecule
Also supported: all
Data Split and Exploration Controls¶
--split-allowed-iterations controls how many early iterations are allowed to resplit
train/validation (ignored if you provide validation.csv). --exploration-iterations
controls how long the agent spends on baseline/exploration models.
Time Limits¶
Set a deadline for the entire run:
Set timeout for each training execution (default is 6 hours):
Custom User Prompt¶
Override the default optimization goal:
See Custom Prompts for more details.
Full Help¶
What Happens During a Run¶
- Dataset Preparation - Validates and prepares data in
prepared_datasets/ - Iterative Development - Agent runs exploration, training, and evaluation cycles
- Snapshot Best Model - Tracks the best-performing iteration
- Final Evaluation - Tests on held-out test set (if provided)
- Output Results - Saves everything to
outputs/<agent_id>/
Monitoring Progress¶
During execution, you'll see:
- Current iteration number
- Agent step (exploration, training, etc.)
- Validation metrics after each iteration
- Best iteration tracking
Stopping a Run¶
Press Ctrl+C to stop. The agent will attempt to save current progress.
Next Steps¶
- CLI Options - Complete flag reference
- Understanding Outputs - What the agent produces
- Custom Prompts - Customize agent behavior