Multi-backend support

TGI (Text Generation Inference) offers flexibility by supporting multiple backends for serving large language models (LLMs). With multi-backend support, you can choose the backend that best suits your needs, whether you prioritize performance, ease of use, or compatibility with specific hardware. API interaction with TGI remains consistent across backends, allowing you to switch between them seamlessly.

Supported backends:

< > Update on GitHub