Update README.md for changing llm model based on validated model table

Signed-off-by: Tsai, Louie <[email protected]>

ChatQnA/docker_compose/intel/hpu/gaudi/README.md

@@ -39,6 +39,25 @@ To set up environment variables for deploying ChatQnA services, follow these ste
    source ./set_env.sh
    ```
 
+4. Change Model for LLM serving
+
+   By default, Meta-Llama-3-8B-Instruct is used for LLM serving, users could change the default model to other validated llm models.  
+   Please pick a [validated llm models](https://github.com/opea-project/GenAIComps/tree/main/comps/llms/src/text-generation#validated-llm-models) from the table.  
+   Users could just overwrite the default model defined in set_env.sh by exporting LLM_MODEL_ID to new model or changing set_env.sh and then repeat step3.  
+   For example, users can change to DeepSeek-R1-Distill-Qwen-32B by below command.
+
+   ```bash
+   export LLM_MODEL_ID=deepseek-ai/DeepSeek-R1-Distill-Qwen-32B
+   ```
+
+   Please also check [required gaudi cards for different models](https://github.com/opea-project/GenAIComps/blob/deepseek/comps/llms/src/text-generation/README.md#system-requirements-for-llm-models) for your new models.  
+   Users might need to increase number of gaudi cards for the model by exporting NUM_CARDS to new model or changing set_env.sh and then repeat step3.  
+   For example, users should increase number of gaudi cards for DeepSeek-R1-Distill-Qwen-32B by below command.
+
+   ```bash
+   export NUM_CARDS=4
+   ```
+
 ## Quick Start: 2.Run Docker Compose
 
 ```bash