Motivation

This series of PRs attempt to decouple the model code from vLLM dependencies. There are mainly four components we use:

from vllm.config import CacheConfig
from vllm.distributed import get_tensor_model_parallel_world_size
from vllm.model_executor.layers.rotary_embedding import get_rope
from vllm.model_executor.layers.vocab_parallel_embedding import (
   ParallelLMHead,
   VocabParallelEmbedding,
)

This PR removed the first one CacheConfig. This is the easiest one because radix attention always set page size as 1 so we don't need to set cache config.

Modifications

Remove from vllm.config import CacheConfig in all models

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.