Update app.py
Browse files
app.py
CHANGED
@@ -45,31 +45,29 @@ def local_query(query, context):
|
|
45 |
def run_query(btn, history, query):
|
46 |
|
47 |
|
|
|
48 |
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
# loader = PDFMinerLoader(file_name)
|
53 |
-
# doc = loader.load()
|
54 |
|
55 |
-
|
56 |
-
|
57 |
|
58 |
-
|
59 |
|
60 |
-
|
61 |
-
|
62 |
|
63 |
-
|
64 |
|
65 |
-
|
66 |
-
|
67 |
|
68 |
-
|
69 |
-
|
70 |
-
|
71 |
-
|
72 |
-
|
73 |
|
74 |
|
75 |
|
|
|
45 |
def run_query(btn, history, query):
|
46 |
|
47 |
|
48 |
+
file_name = btn.name
|
49 |
|
50 |
+
loader = PDFMinerLoader(file_name)
|
51 |
+
doc = loader.load()
|
|
|
|
|
|
|
52 |
|
53 |
+
text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
|
54 |
+
texts = text_splitter.split_documents(doc)
|
55 |
|
56 |
+
texts = [i.page_content for i in texts]
|
57 |
|
58 |
+
doc_emb = st_model.encode(texts)
|
59 |
+
doc_emb = doc_emb.tolist()
|
60 |
|
61 |
+
ids = [str(uuid.uuid1()) for _ in doc_emb]
|
62 |
|
63 |
+
client = chromadb.Client()
|
64 |
+
collection = client.create_collection("test_db")
|
65 |
|
66 |
+
collection.add(
|
67 |
+
embeddings=doc_emb,
|
68 |
+
documents=texts,
|
69 |
+
ids=ids
|
70 |
+
)
|
71 |
|
72 |
|
73 |
|