yujunjun/ragflow
1
0
Fork 0
mirror of https://github.com/infiniflow/ragflow.git synced 2025-12-03 10:26:52 +00:00
Code Issues Packages Projects Releases Wiki Activity

Refactor: Improve How To Handle QWenEmbed (#8765)

Browse Source 
### What problem does this PR solve?

Based on https://github.com/infiniflow/ragflow/issues/8740 
1. A better handle for 'NoneType' object is not subscriptable
2. Add some logs to get the internal message

### Type of change

- [x] Refactoring
This commit is contained in:
Stephen Hu 2025-07-10 10:30:18 +08:00 committed by GitHub
parent 2a11b2c331
commit 8d027813f5
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 6 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
rag/llm/embedding_model.py
View File

@ -209,12 +209,15 @@ class QWenEmbed(Base):
for i in range(0, len(texts), batch_size):
retry_max = 5
resp = dashscope.TextEmbedding.call(model=self.model_name, input=texts[i : i + batch_size], api_key=self.key, text_type="document")
while resp["output"] is None and retry_max > 0:
while (resp["output"] is None or resp["output"].get("embeddings") is None) and retry_max > 0:
time.sleep(10)
resp = dashscope.TextEmbedding.call(model=self.model_name, input=texts[i : i + batch_size], api_key=self.key, text_type="document")
retry_max -= 1
if retry_max == 0 and resp["output"] is None:
log_exception(ValueError("Retry_max reached, calling embedding model failed"))
if retry_max == 0 and (resp["output"] is None or resp["output"].get("embeddings") is None):
if resp.get("message"):
log_exception(ValueError(f"Retry_max reached, calling embedding model failed: {resp['message']}"))
else:
log_exception(ValueError("Retry_max reached, calling embedding model failed"))
raise
try:
embds = [[] for _ in range(len(resp["output"]["embeddings"]))]