vLLM & TGI
Transformers.js
Transformers.js enables browser-native transformer model inference without server requirements.
About
Transformers.js allows users to perform transformer model inference directly in the browser, utilizing WebGPU for acceleration. It supports various models for tasks such as image classification and text generation, loading models directly from Hugging Face. This tool is designed for developers and researchers interested in leveraging machine learning models without the need for server infrastructure.