Add lora_path to chat completion

[ccchow]

Motivation

Add lora_path to ChatCompletionRequest for OpenAI chat completion API. It was previously added to OpenAI completion API #2243

Modifications

Added lora_path to ChatCompletionRequest

Checklist

• Format your code according to the Contributor Guide.
• Add unit tests as outlined in the Contributor Guide.
• Update documentation as needed, including docstrings or example tutorials.

[merrymercy]

Even if you added this, it seems it is still not used.

[ccchow]

We are working on a project to provide multi-lora serving via OpenAI compatible API, and we have validated this fix by adding lora_path to OpenAI protocol and serving a batch with different lora adapters.

[ccchow]

Thanks for merging this change!

[qingzhong1]

Hello, how to use url = "http://localhost:8000/v1/chat/completions" to request the configured lora, data = {"model": "Qwen2.5-7B-Instruct","messages": [{ "role": "user", "content": "What is the capital of France?"}]},lora name='aa'

[ccchow]

Hello, how to use url = "http://localhost:8000/v1/chat/completions" to request the configured lora, data = {"model": "Qwen2.5-7B-Instruct","messages": [{ "role": "user", "content": "What is the capital of France?"}]},lora name='aa'

curl -X POST http://127.0.0.1:30000/v1/chat/completions -d '{"model": "meta-llama/Llama-3.2-1B", "messages": [{"role": "system", "content": "You are a happy assistant that puts a positive spin on everything."}, {"role": "user", "content": "I fell off my bike today."}], "lora_path": "lora1", "max_tokens": 64}'

[qingzhong1]

v1_completions can successfully call lora, but v1/chat/completions cannot call lora, why?Comparing v1_generate_request and v1_chat_generate_request, we found that v1_chat_generate_request does not have the lora_pathes variable

[ccchow]

v1_completions can successfully call lora, but v1/chat/completions cannot call lora, why?Comparing v1_generate_request and v1_chat_generate_request, we found that v1_chat_generate_request does not have the lora_pathes variable

You are right. I missed that when cherry picking changes. Will have another PR.

[ccchow]

v1_completions can successfully call lora, but v1/chat/completions cannot call lora, why?Comparing v1_generate_request and v1_chat_generate_request, we found that v1_chat_generate_request does not have the lora_pathes variable

#2529