Update README.md
Browse files
README.md
CHANGED
|
@@ -17,9 +17,19 @@ This model is a light SFT version from Qwen/Qwen3-14B, aimed at stable generativ
|
|
| 17 |
|
| 18 |
## Serve with vllm
|
| 19 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
|
|
|
|
| 21 |
```
|
| 22 |
vllm serve billxbf/Nano-Raccoon-Preview-1104 \
|
|
|
|
| 23 |
--trust-remote-code \
|
| 24 |
--host 0.0.0.0 \
|
| 25 |
--port 8000
|
|
|
|
| 17 |
|
| 18 |
## Serve with vllm
|
| 19 |
|
| 20 |
+
**Single GPU**
|
| 21 |
+
```
|
| 22 |
+
vllm serve billxbf/Nano-Raccoon-Preview-1104 \
|
| 23 |
+
--trust-remote-code \
|
| 24 |
+
--host 0.0.0.0 \
|
| 25 |
+
--port 8000
|
| 26 |
+
```
|
| 27 |
+
|
| 28 |
|
| 29 |
+
**Use Tensor Parallel on 8xGPU**
|
| 30 |
```
|
| 31 |
vllm serve billxbf/Nano-Raccoon-Preview-1104 \
|
| 32 |
+
--tensor-parallel-size 8 \
|
| 33 |
--trust-remote-code \
|
| 34 |
--host 0.0.0.0 \
|
| 35 |
--port 8000
|