mirror of
https://github.com/huggingface/text-generation-inference.git
synced 2025-04-19 22:02:06 +00:00
14 lines
1014 B
Markdown
14 lines
1014 B
Markdown
# 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:**
|
|
* **TGI CUDA backend**: This high-performance backend is optimized for NVIDIA GPUs and serves as the default option
|
|
within TGI. Developed in-house, it boasts numerous optimizations and is used in production by various projects, including those by Hugging Face.
|
|
* **[TGI TRTLLM backend](./backends/trtllm)**: This backend leverages NVIDIA's TensorRT library to accelerate LLM inference.
|
|
It utilizes specialized optimizations and custom kernels for enhanced performance.
|
|
However, it requires a model-specific compilation step for each GPU architecture.
|