biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Kernel] Expand FP8 support to Ampere GPUs using FP8 Marlin (#5975)

Browse Source
This commit is contained in:
Michael Goin
2024-07-03 13:38:00 -04:00
committed by GitHub
parent 7cd2ebb025
commit 47f0954af0
11 changed files with 1585 additions and 42 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
docs/source/quantization/fp8.rst
View File

@@ -4,7 +4,8 @@ FP8
==================
vLLM supports FP8 (8-bit floating point) weight and activation quantization using hardware acceleration on GPUs such as Nvidia H100 and AMD MI300x.
Currently, only Hopper and Ada Lovelace GPUs are supported.
Currently, only Hopper and Ada Lovelace GPUs are officially supported for W8A8.
Ampere GPUs are supported for W8A16 (weight-only FP8) utilizing Marlin kernels.
Quantization of models with FP8 allows for a 2x reduction in model memory requirements and up to a 1.6x improvement in throughput with minimal impact on accuracy.
Please visit the HF collection of `quantized FP8 checkpoints of popular LLMs ready to use with vLLM <https://huggingface.co/collections/neuralmagic/fp8-llms-for-vllm-666742ed2b78b7ac8df13127>`_.