fix(utils): 修复缓存未正确启用的问题
- 在 handle_cache 函数中添加了对 global_config 中 enable_l`lm_cache 设置的检查 - 如果配置禁用了缓存,则直接返回 None -这个修改确保了在不需要缓存的情况下,函数能够正确地跳过缓存处理
This commit is contained in:
@@ -449,7 +449,7 @@ def dequantize_embedding(
|
||||
|
||||
async def handle_cache(hashing_kv, args_hash, prompt, mode="default"):
|
||||
"""Generic cache handling function"""
|
||||
if hashing_kv is None:
|
||||
if hashing_kv is None or not hashing_kv.global_config.get("enable_l`lm_cache"):
|
||||
return None, None, None, None
|
||||
|
||||
# For naive mode, only use simple cache matching
|
||||
|
Reference in New Issue
Block a user