What is this prompt chain used for?

This prompt chain configures a HuggingFace Inference Endpoint as a promptfoo provider to evaluate custom or fine-tuned models deployed on a dedicated HuggingFace Inference Endpoint.

When should I use this instead of evaluating publicly-hosted HuggingFace models?

Use this when the model you need to evaluate is deployed on a dedicated HuggingFace Inference Endpoint rather than a shared, publicly-hosted model that doesn't require a dedicated endpoint.

What credentials and configuration does this require?

It requires an `HF_API_TOKEN` and uses `promptfooconfig.yaml` for configuration, with further setup details available in promptfoo's HuggingFace provider documentation.

How do I get started with this example?

Run `npx promptfoo@latest init --example huggingface/inference-endpoint` and then navigate to the `huggingface/inference-endpoint` directory.

Prompt Chain

Deploy and test Hugging Face inference endpoints

Name: Inference Endpoint
Availability: OnlineOnly
Author: Promptfoo

Set up and evaluate Hugging Face inference endpoints with automated testing workflows and configuration templates.

Copy chain

Works with huggingface

Promptfoo

Own this? Claim it

Spark score

out of 100

Updated 19 days ago

Version 0.121.18

Add to Favorites

Why it matters

Enable developers to quickly set up, configure, and evaluate Hugging Face inference endpoints through automated testing workflows that validate model performance and API integration.

Outcomes

What it gets done

Initialize Hugging Face inference endpoint configuration files

Authenticate with Hugging Face API using environment tokens

Execute evaluation tests against deployed inference endpoints

View and analyze inference endpoint test results

Install

Add it to your toolbox

Run in your project directory:

curl -fsSL https://spark.entire.vc/get/pfoo-inference-endpoint | bash

Steps

Steps in the chain

Initialize example project

Navigate to project directory

Set HF_API_TOKEN environment variable

Edit configuration file

Run evaluation

View results

Overview

Inference Endpoint

This promptfoo example provides a starting configuration for working with Hugging Face Inference Endpoints. Use this example when you want a starting point for using promptfoo with Hugging Face Inference Endpoints.

What it does

This promptfoo example provides a starting configuration for working with Hugging Face Inference Endpoints. It includes the basic setup files needed to get started with promptfoo and Hugging Face.

When to use - and when NOT to

Use this example when you want a starting point for using promptfoo with Hugging Face Inference Endpoints. Do NOT use this if you're working with other providers - promptfoo supports many other providers that may be more appropriate for those scenarios.

Inputs and outputs

You provide your HF_API_TOKEN environment variable and edit the promptfooconfig.yaml file. After running the evaluation, you can view the results by running promptfoo view.

Integrations

This example works with Hugging Face Inference Endpoints and uses the promptfoo framework.

Who it's for

This example is designed for developers who want to use promptfoo with Hugging Face Inference Endpoints. The example provides a starting point that can be customized for specific needs.

Quick start

Initialize the example project:

npx promptfoo@latest init --example huggingface/inference-endpoint
cd huggingface/inference-endpoint

Set your HF_API_TOKEN environment variable, edit the promptfooconfig.yaml file, then run:

promptfoo eval

View results with promptfoo view.

Source README

huggingface/inference-endpoint (Huggingface Inference Endpoint)

You can run this example with:

npx promptfoo@latest init --example huggingface/inference-endpoint
cd huggingface/inference-endpoint

Usage

To get started, set your HF_API_TOKEN environment variable. See also https://promptfoo.dev/docs/providers/huggingface/ for setup details.

Next, edit promptfooconfig.yaml.

Then run:

promptfoo eval

Afterwards, you can view the results by running promptfoo view

FAQ

Common questions

Discussion