HuggingFace Integration

Lamatic.ai Studio supports both public (open-source) and private (self-hosted/dedicated) Hugging Face models.
You can integrate popular models from the Hugging Face Hub or connect your own secure, fine-tuned endpoints for production use.

Hugging Face is a leading platform for building, hosting, and deploying state-of-the-art machine learning models. With its extensive model hub and flexible deployment options, Hugging Face allows you to use both public models and your own self-hosted endpoints. Lamatic.ai Studio enables seamless integration of these models, whether you use a managed public endpoint or a private, self-hosted endpoint for custom or fine-tuned models.

Get Started

Step 1: Create a Hugging Face Account

1. Go to the Hugging Face Hub (opens in a new tab) and sign up or log in.
2. Complete account verification if prompted.

Step 2: Prepare Your Model Endpoint

Option A: Public Model Endpoint

1. Browse the Hugging Face Hub (opens in a new tab) and find the public model you wish to use.
2. Copy the model name (e.g., bert-base-uncased).
3. The public inference API endpoint is:
   https://api-inference.huggingface.co/models/your-model-name

Option B: Self-Hosted (Dedicated) Inference Endpoint

1. Go to Hugging Face Inference Endpoints (opens in a new tab).
2. Create a new endpoint for your private or fine-tuned model.
3. Once deployed, copy the Base URL of your endpoint (e.g., https://your-namespace.hf.space/proxy/your-endpoint).

Step 3: Generate Access Token

1. Visit Hugging Face Access Tokens (opens in a new tab).
2. Create a new token with read or inference permissions.
3. Copy your access token (you will need it for Lamatic.ai Studio).

Step 4: Configure in Lamatic.ai Studio

1. Open Lamatic.ai Studio (opens in a new tab).
2. Navigate to the Models section.
3. Select HuggingFace from the provider list.
4. Enter the following depending on your setup:
   • Public model:
     • Paste the public inference API URL or the model name.
     • Paste your Hugging Face access token.
   • Self-hosted endpoint:
     • Paste your dedicated endpoint Base URL.
     • Paste your Hugging Face access token.
5. Save your configuration.

Hugging Face Model Configuration in Lamatic.ai Studio

Key Features

• Custom & Fine-Tuned Models:
  Connect your own models or use domain-specific fine-tuned endpoints.
• Public Model Access:
  Quickly integrate popular open-source models from the Hugging Face Hub.
• Self-Hosted Endpoints:
  Use private, scalable endpoints for production workloads and enhanced security.
• Seamless Integration:
  Easily manage models from both public and private endpoints in Lamatic workflows.
• Secure Authentication:
  All access is managed via Hugging Face tokens.
• Provider-Level Customization:
  Override default behaviors per model or endpoint as needed.

For advanced options, see Custom Model Integration (opens in a new tab).

Available Configuration Options

• Model URL or Name: The public model name, public API endpoint URL, or self-hosted endpoint Base URL
• Access Token: Your Hugging Face access token (required for both public and private endpoints)

Best Practices

• Credential Security: Keep your access tokens private and never share them.
• Token Rotation: Regularly rotate your Hugging Face tokens.
• Testing: Test model responses after integration.
• Production: Use self-hosted, dedicated endpoints for critical or production workloads.
• Monitoring: Monitor rate limits and usage from your Hugging Face account.

Troubleshooting

Invalid Token:

• Ensure your access token is correct and has the required permissions.
• Confirm your Hugging Face account is active.

Access Denied:

• Check your token’s permissions and endpoint/model visibility.
• Make sure you have access to the deployed endpoint if using self-hosted.

Model or Endpoint Not Found:

• Double-check the model name or endpoint URL.
• Confirm you have access to the model or endpoint.

Rate Limit Exceeded:

• Wait before retrying requests.
• Upgrade your Hugging Face plan if necessary.

Integration Errors:

• Confirm Lamatic.ai Studio configuration matches your Hugging Face setup.
• Consult Hugging Face Docs (opens in a new tab) and Lamatic Support (opens in a new tab) for troubleshooting.

Additional Resources

• Hugging Face Documentation (opens in a new tab)
• Model Hub (opens in a new tab)
• Access Tokens (opens in a new tab)
• Custom Model Integration (opens in a new tab)
• Lamatic Support (opens in a new tab)

Need help? Contact Lamatic Support (opens in a new tab).