Updated README.md to include serving command and inference speed.

README.md

@@ -9,8 +9,17 @@ pipeline_tag: text-generation
 library_name: transformers
 ---
 # DeepSeek V3 AWQ
-AWQ of the DeepSeek V3 chat model.
+AWQ of DeepSeek V3.
 
-This quant modified some of the model code to fix the overflow issue when using float16.
+This quant modified some of the model code to fix an overflow issue when using float16.
 
-Tested on vLLM with 8x H100, inference speed 5 tokens per second with batch size 1 and short prompt, 12 tokens per second when using `moe_wna16` kernel.
+To serve using vLLM with 8x 80GB GPUs, use the following command:
+```sh
+python -m vllm.entrypoints.openai.api_server --host 0.0.0.0 --port 12345 --max-model-len 65536 --trust-remote-code --tensor-parallel-size 8 --quantization moe_wna16 --gpu-memory-utilization 0.97 --kv-cache-dtype fp8_e5m2 --calculate-kv-scales --served-model-name deepseek-chat --model cognitivecomputations/DeepSeek-V3-AWQ
+```
+The max model length flag ensures that KV cache usage won't be higher than available memory, the `moe_wna16` kernel doubles the inference speed, but you must build vLLM from source as of 2025/2/3. \
+You can download the wheel I built for PyTorch 2.6, Python 3.12 by clicking [here](https://huggingface.co/x2ray/wheels/resolve/main/vllm-0.7.1.dev69%2Bg4f4d427a.d20220101.cu126-cp312-cp312-linux_x86_64.whl).
+
+Inference speed with batch size 1 and short prompt:
+- 8x H100: 34 TPS
+- 8x A100: 27 TPS