Spaces:

ccm
/

chat-with-SFF

Sleeping

ccm commited on Nov 8, 2024

Commit

c759f50

verified ·

1 Parent(s): 2b9ca20

Update app.py

Files changed (1) hide show

app.py CHANGED Viewed

@@ -68,7 +68,7 @@ publication_vectorstore = langchain_community.vectorstores.FAISS.load_local(
 # Create the callable LLM
 llm = transformers.pipeline(
-    task="text-generation", model="Qwen/Qwen2.5-7B-Instruct-GPTQ-Int4", device="cuda"
 )

 # Create the callable LLM
 llm = transformers.pipeline(
+    task="text-generation", model="Qwen/Qwen2.5-7B-Instruct-AWQ", device="cuda", attn_implementation="flash_attention_2"
 )