Update README.md (#3)

- Update README.md (72898eec36a99f3627b924c065aec51abe0db729)
- Update README.md (3878522bf20d7e0192703ba34c43757b2c43296e)
- Update README.md (cb9a6781d102fb48dde54b69b5310f512e8376b0)
- Update README.md (5e5be73c68ca366ec85959273cc33c17e16185af)
- Update README.md (6c6ccc979e12b4f57cdc666b31523ec83540a074)


Co-authored-by: Yao Zhao <[email protected]>

README.md

@@ -31,14 +31,22 @@ To enable deployment of [Ring-Linear-2.0](https://github.com/inclusionAI/Ring-V2
 
 #### Environment Preparation
 
-Since the Pull Request (PR) has not been submitted to the vLLM community at this stage, please prepare the environment by following the steps below:
+Since the Pull Request (PR) has not been submitted to the vLLM community at this stage, please prepare the environment by following the steps below.
+
+First, create a Conda environment with Python 3.10 and CUDA 12.8 (Use the root or admin account, or ensure the current user has access to /home/admin/logs):
 ```shell
-pip install torch==2.7.0 torchvision==0.22.0 
+conda create -n vllm python=3.10
+conda activate vllm
+```
+
+Next, install our vLLM wheel package:
+```shell
+pip install https://media.githubusercontent.com/media/inclusionAI/Ring-V2/refs/heads/main/hybrid_linear/whls/vllm-0.8.5%2Bcuda12_8_gcc10_2_1-cp310-cp310-linux_x86_64.whl --force-reinstall
 ```
 
-Then you should install our vLLM wheel package:
+Finally, install compatible versions of PyTorch and Torchvision after vLLM is installed:
 ```shell
-pip install https://media.githubusercontent.com/media/inclusionAI/Ring-V2/refs/heads/main/hybrid_linear/whls/vllm-0.8.5%2Bcuda12_8_gcc10_2_1-cp310-cp310-linux_x86_64.whl --no-deps --force-reinstall
+pip install torch==2.7.0 torchvision==0.22.0 
 ```
 
 #### Offline Inference
@@ -47,35 +55,39 @@ pip install https://media.githubusercontent.com/media/inclusionAI/Ring-V2/refs/h
 from transformers import AutoTokenizer
 from vllm import LLM, SamplingParams
 
-tokenizer = AutoTokenizer.from_pretrained("inclusionAI/Ring-mini-linear-2.0-GPTQ-int4")
-
-sampling_params = SamplingParams(temperature=0.6, top_p=1.0, max_tokens=16384)
-
-
-llm = LLM(model="inclusionAI/Ring-mini-linear-2.0-GPTQ-int4", dtype='auto', enable_prefix_caching=False, max_num_seqs=128)
-
-
-prompt = "Give me a short introduction to large language models."
-messages = [
-    {"role": "user", "content": prompt}
-]
-
-text = tokenizer.apply_chat_template(
-    messages,
-    tokenize=False,
-    add_generation_prompt=True
-)
-outputs = llm.generate([text], sampling_params)
+if __name__ == '__main__':
+    tokenizer = AutoTokenizer.from_pretrained("inclusionAI/Ring-mini-linear-2.0-GPTQ-int4")
+    
+    sampling_params = SamplingParams(temperature=0.6, top_p=1.0, max_tokens=16384)
+
+    # use `max_num_seqs=1` without concurrency
+    llm = LLM(model="inclusionAI/Ring-mini-linear-2.0-GPTQ-int4", dtype='auto', enable_prefix_caching=False, max_num_seqs=128)
+    
+    
+    prompt = "Give me a short introduction to large language models."
+    messages = [
+        {"role": "user", "content": prompt}
+    ]
+    
+    text = tokenizer.apply_chat_template(
+        messages,
+        tokenize=False,
+        add_generation_prompt=True
+    )
+    outputs = llm.generate([text], sampling_params)
+    for output in outputs:
+        print(output.outputs[0].text)
 ```
 
 #### Online Inference
 ```shell
 vllm serve inclusionAI/Ring-mini-linear-2.0-GPTQ-int4 \
-              --tensor-parallel-size 2 \
+              --tensor-parallel-size 1 \
               --pipeline-parallel-size 1 \
               --gpu-memory-utilization 0.90 \
-              --max-num-seqs 512 \
+              --max-num-seqs 128 \
               --no-enable-prefix-caching
+              --api-key your-api-key
 ```