An ML intern that autonomously researches, writes, and ships good quality ML releated code using the Hugging Face ecosystem — with deep access to docs, papers, datasets, and cloud compute.
| Template | Description | Download |
|---|---|---|
| EC2 (Simple) | Single instance, SSH access | ml-intern-stack.yaml |
| Fargate (Production) | Auto-scaling, Load Balancer | ml-intern-fargate.yaml |
| Region | Open Console |
|---|---|
| US East (N. Virginia) |  |
| US West (Oregon) |  |
| EU (Ireland) |  |
- Select "Upload a template file"
- Upload the downloaded YAML file
- Enter your HuggingFace token
- Click "Create Stack"
Prerequisites: Enable Bedrock models in your AWS account first.
See cloudformation/README.md for detailed instructions.
git clone [email protected]:huggingface/ml-intern.git
cd ml-intern
uv sync
uv tool install -e .ml-internCreate a .env file in the project root (or export these in your shell):
ANTHROPIC_API_KEY=<your-anthropic-api-key> # if using anthropic models
HF_TOKEN=<your-hugging-face-token>
GITHUB_TOKEN=<github-personal-access-token> If no HF_TOKEN is set, the CLI will prompt you to paste one on first launch. To get a GITHUB_TOKEN follow the tutorial here.
Interactive mode (start a chat session):
ml-internHeadless mode (single prompt, auto-approve):
ml-intern "fine-tune llama on my dataset"Options:
ml-intern --model anthropic/claude-opus-4-6 "your prompt"
ml-intern --max-iterations 100 "your prompt"
ml-intern --no-stream "your prompt"┌─────────────────────────────────────────────────────────────┐
│ User/CLI │
└────────────┬─────────────────────────────────────┬──────────┘
│ Operations │ Events
↓ (user_input, exec_approval, ↑
submission_queue interrupt, compact, ...) event_queue
│ │
↓ │
┌────────────────────────────────────────────────────┐ │
│ submission_loop (agent_loop.py) │ │
│ ┌──────────────────────────────────────────────┐ │ │
│ │ 1. Receive Operation from queue │ │ │
│ │ 2. Route to handler (run_agent/compact/...) │ │ │
│ └──────────────────────────────────────────────┘ │ │
│ ↓ │ │
│ ┌──────────────────────────────────────────────┐ │ │
│ │ Handlers.run_agent() │ ├──┤
│ │ │ │ │
│ │ ┌────────────────────────────────────────┐ │ │ │
│ │ │ Agentic Loop (max 300 iterations) │ │ │ │
│ │ │ │ │ │ │
│ │ │ ┌──────────────────────────────────┐ │ │ │ │
│ │ │ │ Session │ │ │ │ │
│ │ │ │ ┌────────────────────────────┐ │ │ │ │ │
│ │ │ │ │ ContextManager │ │ │ │ │ │
│ │ │ │ │ • Message history │ │ │ │ │ │
│ │ │ │ │ (litellm.Message[]) │ │ │ │ │ │
│ │ │ │ │ • Auto-compaction (170k) │ │ │ │ │ │
│ │ │ │ │ • Session upload to HF │ │ │ │ │ │
│ │ │ │ └────────────────────────────┘ │ │ │ │ │
│ │ │ │ │ │ │ │ │
│ │ │ │ ┌────────────────────────────┐ │ │ │ │ │
│ │ │ │ │ ToolRouter │ │ │ │ │ │
│ │ │ │ │ ├─ HF docs & research │ │ │ │ │ │
│ │ │ │ │ ├─ HF repos, datasets, │ │ │ │ │ │
│ │ │ │ │ │ jobs, papers │ │ │ │ │ │
│ │ │ │ │ ├─ GitHub code search │ │ │ │ │ │
│ │ │ │ │ ├─ Sandbox & local tools │ │ │ │ │ │
│ │ │ │ │ ├─ Planning │ │ │ │ │ │
│ │ │ │ │ └─ MCP server tools │ │ │ │ │ │
│ │ │ │ └────────────────────────────┘ │ │ │ │ │
│ │ │ └──────────────────────────────────┘ │ │ │ │
│ │ │ │ │ │ │
│ │ │ ┌──────────────────────────────────┐ │ │ │ │
│ │ │ │ Doom Loop Detector │ │ │ │ │
│ │ │ │ • Detects repeated tool patterns │ │ │ │ │
│ │ │ │ • Injects corrective prompts │ │ │ │ │
│ │ │ └──────────────────────────────────┘ │ │ │ │
│ │ │ │ │ │ │
│ │ │ Loop: │ │ │ │
│ │ │ 1. LLM call (litellm.acompletion) │ │ │ │
│ │ │ ↓ │ │ │ │
│ │ │ 2. Parse tool_calls[] │ │ │ │
│ │ │ ↓ │ │ │ │
│ │ │ 3. Approval check │ │ │ │
│ │ │ (jobs, sandbox, destructive ops) │ │ │ │
│ │ │ ↓ │ │ │ │
│ │ │ 4. Execute via ToolRouter │ │ │ │
│ │ │ ↓ │ │ │ │
│ │ │ 5. Add results to ContextManager │ │ │ │
│ │ │ ↓ │ │ │ │
│ │ │ 6. Repeat if tool_calls exist │ │ │ │
│ │ └────────────────────────────────────────┘ │ │ │
│ └──────────────────────────────────────────────┘ │ │
└────────────────────────────────────────────────────┴──┘
User Message
↓
[Add to ContextManager]
↓
╔═══════════════════════════════════════════╗
║ Iteration Loop (max 300) ║
║ ║
║ Get messages + tool specs ║
║ ↓ ║
║ litellm.acompletion() ║
║ ↓ ║
║ Has tool_calls? ──No──> Done ║
║ │ ║
║ Yes ║
║ ↓ ║
║ Add assistant msg (with tool_calls) ║
║ ↓ ║
║ Doom loop check ║
║ ↓ ║
║ For each tool_call: ║
║ • Needs approval? ──Yes──> Wait for ║
║ │ user confirm ║
║ No ║
║ ↓ ║
║ • ToolRouter.execute_tool() ║
║ • Add result to ContextManager ║
║ ↓ ║
║ Continue loop ─────────────────┐ ║
║ ↑ │ ║
║ └───────────────────────┘ ║
╚═══════════════════════════════════════════╝
The agent emits the following events via event_queue:
processing- Starting to process user inputready- Agent is ready for inputassistant_chunk- Streaming token chunkassistant_message- Complete LLM response textassistant_stream_end- Token stream finishedtool_call- Tool being called with argumentstool_output- Tool execution resulttool_log- Informational tool log messagetool_state_change- Tool execution state transitionapproval_required- Requesting user approval for sensitive operationsturn_complete- Agent finished processingerror- Error occurred during processinginterrupted- Agent was interruptedcompacted- Context was compactedundo_complete- Undo operation completedshutdown- Agent shutting down
Edit agent/core/tools.py:
def create_builtin_tools() -> list[ToolSpec]:
return [
ToolSpec(
name="your_tool",
description="What your tool does",
parameters={
"type": "object",
"properties": {
"param": {"type": "string", "description": "Parameter description"}
},
"required": ["param"]
},
handler=your_async_handler
),
# ... existing tools
]Edit configs/main_agent_config.json:
{
"model_name": "anthropic/claude-sonnet-4-5-20250929",
"mcpServers": {
"your-server-name": {
"transport": "http",
"url": "https://example.com/mcp",
"headers": {
"Authorization": "Bearer ${YOUR_TOKEN}"
}
}
}
}Note: Environment variables like ${YOUR_TOKEN} are auto-substituted from .env.