lfoppiano commited on
Commit
b3584a6
1 Parent(s): d520110

fix chroma embeddings when an already seen file is uplaoded

Browse files
Files changed (1) hide show
  1. document_qa/document_qa_engine.py +7 -2
document_qa/document_qa_engine.py CHANGED
@@ -205,11 +205,16 @@ class DocumentQAEngine:
205
  if doc_id:
206
  hash = doc_id
207
  else:
208
-
209
  hash = metadata[0]['hash']
210
 
211
  if hash not in self.embeddings_dict.keys():
212
- self.embeddings_dict[hash] = Chroma.from_texts(texts, embedding=self.embedding_function, metadatas=metadata, collection_name=hash)
 
 
 
 
 
 
213
 
214
  self.embeddings_root_path = None
215
 
 
205
  if doc_id:
206
  hash = doc_id
207
  else:
 
208
  hash = metadata[0]['hash']
209
 
210
  if hash not in self.embeddings_dict.keys():
211
+ self.embeddings_dict[hash] = Chroma.from_texts(texts, embedding=self.embedding_function, metadatas=metadata,
212
+ collection_name=hash)
213
+ else:
214
+ self.embeddings_dict[hash].delete(ids=self.embeddings_dict[hash].get()['ids'])
215
+ self.embeddings_dict[hash] = Chroma.from_texts(texts, embedding=self.embedding_function, metadatas=metadata,
216
+ collection_name=hash)
217
+
218
 
219
  self.embeddings_root_path = None
220