Misc fix for min_p_sampling, --cuda-graph-bs (#2761)

python/sglang/srt/layers/quantization/__init__.py

@@ -1,8 +1,7 @@
 # Adapted from https://raw.githubusercontent.com/vllm-project/vllm/v0.5.5/vllm/model_executor/layers/quantization/__init__.py
 
-from typing import Callable, Dict, Optional, Type
+from typing import Dict, Type
 
-import torch
 from vllm.model_executor.layers.quantization.aqlm import AQLMConfig
 from vllm.model_executor.layers.quantization.awq import AWQConfig
 from vllm.model_executor.layers.quantization.awq_marlin import AWQMarlinConfig