Update README.md

README.md

@@ -100,7 +100,7 @@ You can refer to the content in [Tencent-Hunyuan-Large](https://github.com/Tence
 
 This section presents the efficiency test results of deploying various models using vLLM, including inference speed (tokens/s) under different batch sizes.
 
-| Inference Framework | Model      | Number of GPUs (series 1) | input_length | batch=1             | batch=4              |
+| Inference Framework | Model      | Number of GPUs (GPU productA) | input_length | batch=1             | batch=4              |
 |------|------------|-------------------------|-------------------------|---------------------|----------------------|
 | vLLM | hunyuan-7B | 1                       | 2048                  | 78.9                | 279.5                  |