update

Signed-off-by: jiang1.li <[email protected]>
vllm-project · Jan 10, 2025 · 0b8bc9a · 0b8bc9a
1 parent 46621ac
commit 0b8bc9a
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/docs/source/getting_started/installation/cpu-x86.md b/docs/source/getting_started/installation/cpu-x86.md
@@ -5,7 +5,7 @@
 vLLM initially supports basic model inferencing and serving on x86 CPU platform, with data types FP32, FP16 and BF16. vLLM CPU backend supports the following vLLM features:
 
 - Tensor Parallel
-- Model Quantization (`INT8 W8A8, AWQ`)
+- Model Quantization (`INT8 W8A8, AWQ, GPTQ`)
 - Chunked-prefill
 - Prefix-caching
 - FP8-E5M2 KV-Caching (TODO)