Invoke endpoint

documented

Invoke an Inference Endpoint to run inference (payload format depends on the task).

POST /endpoints/{endpointName}/invokeInference Endpointsno implementation yet0 runs

Input

namespace

Endpoint namespace

huggingface

endpointName

Endpoint name

my-endpoint-name

inputs

Input payload (NLP inputs)

The user query here

temperature

Inference temperature

0.7

max_new_tokens

Max new tokens

128

Calls Hugging Face as you (your API key is your account); keychains.dev holds the credentials.