[Bugfix][Harmless] Fix float16 dtype for model_is_embedding (#7566)

2024-08-15 21:26:19 -04:00
parent 21313e09e3
commit 9c8e2d1161
1 changed files with 1 additions and 1 deletions
--- a/vllm/entrypoints/openai/api_server.py
+++ b/vllm/entrypoints/openai/api_server.py
@@ -66,7 +66,7 @@ def model_is_embedding(model_name: str, trust_remote_code: bool) -> bool:
                       tokenizer_mode="auto",
                       trust_remote_code=trust_remote_code,
                       seed=0,
-                       dtype="float16").embedding_mode
+                       dtype="auto").embedding_mode


@asynccontextmanager