Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -35,7 +35,7 @@ app.add_middleware(
|
|
35 |
UPLOAD_DIR = "uploads"
|
36 |
os.makedirs(UPLOAD_DIR, exist_ok=True)
|
37 |
|
38 |
-
persist_directory = "
|
39 |
|
40 |
load_dotenv()
|
41 |
os.environ["HF_TOKEN"] = os.getenv("HF_TOKEN")
|
@@ -64,7 +64,8 @@ def generate_file_id(file_path):
|
|
64 |
|
65 |
def delete_existing_embedding(file_id):
|
66 |
if os.path.exists(persist_directory):
|
67 |
-
shutil.rmtree(persist_directory)
|
|
|
68 |
|
69 |
def tempUploadFile(filePath,file):
|
70 |
with open(filePath,'wb') as buffer:
|
@@ -111,7 +112,7 @@ def loadAndSplitPdfFile(filePath):
|
|
111 |
|
112 |
def prepare_retriever(filePath = "", load_from_chromadb = False):
|
113 |
if load_from_chromadb:
|
114 |
-
vector_store = Chroma(persist_directory=persist_directory, embedding_function = embeddings)
|
115 |
return vector_store.as_retriever(search_kwargs={"k": 5})
|
116 |
elif filePath:
|
117 |
doc_chunks = loadAndSplitPdfFile(filePath)
|
@@ -125,7 +126,7 @@ def prepare_retriever(filePath = "", load_from_chromadb = False):
|
|
125 |
if isinstance(value, (str, int, float, bool, Path))
|
126 |
}
|
127 |
|
128 |
-
vector_store = Chroma.from_documents(documents= doc_chunks, persist_directory=persist_directory, embedding= embeddings)
|
129 |
vector_store.persist()
|
130 |
|
131 |
def get_retriever_chain(retriever):
|
|
|
35 |
UPLOAD_DIR = "uploads"
|
36 |
os.makedirs(UPLOAD_DIR, exist_ok=True)
|
37 |
|
38 |
+
persist_directory = "/home/user/.cache/chroma_db"
|
39 |
|
40 |
load_dotenv()
|
41 |
os.environ["HF_TOKEN"] = os.getenv("HF_TOKEN")
|
|
|
64 |
|
65 |
def delete_existing_embedding(file_id):
|
66 |
if os.path.exists(persist_directory):
|
67 |
+
shutil.rmtree(persist_directory, ignore_errors = True)
|
68 |
+
os.makedirs(persist_directory, exist_ok=True)
|
69 |
|
70 |
def tempUploadFile(filePath,file):
|
71 |
with open(filePath,'wb') as buffer:
|
|
|
112 |
|
113 |
def prepare_retriever(filePath = "", load_from_chromadb = False):
|
114 |
if load_from_chromadb:
|
115 |
+
vector_store = Chroma(persist_directory=persist_directory, embedding_function = embeddings, client_settings={"allow_reset": True})
|
116 |
return vector_store.as_retriever(search_kwargs={"k": 5})
|
117 |
elif filePath:
|
118 |
doc_chunks = loadAndSplitPdfFile(filePath)
|
|
|
126 |
if isinstance(value, (str, int, float, bool, Path))
|
127 |
}
|
128 |
|
129 |
+
vector_store = Chroma.from_documents(documents= doc_chunks, persist_directory=persist_directory, embedding= embeddings, read)
|
130 |
vector_store.persist()
|
131 |
|
132 |
def get_retriever_chain(retriever):
|