Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -33,15 +33,7 @@ from transformers import AutoTokenizer, AutoModelForCausalLM
|
|
33 |
tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3")
|
34 |
model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3")
|
35 |
|
36 |
-
|
37 |
-
loader = PyPDFLoader(dataset)
|
38 |
-
data = loader.load()
|
39 |
-
text_splitter1 = CharacterTextSplitter(chunk_size=512, chunk_overlap=0,separator="\n\n")
|
40 |
-
texts = text_splitter1.split_documents(data)
|
41 |
-
db = FAISS.from_documents(texts,
|
42 |
-
HuggingFaceEmbeddings(model_name='sentence-transformers/all-MiniLM-L12-v2'))
|
43 |
-
|
44 |
-
|
45 |
# Connect query to FAISS index using a retriever
|
46 |
retriever = db.as_retriever(
|
47 |
search_type="mmr",
|
|
|
33 |
tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3")
|
34 |
model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3")
|
35 |
|
36 |
+
load_dataset("pandas", data_files="index.pkl")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
37 |
# Connect query to FAISS index using a retriever
|
38 |
retriever = db.as_retriever(
|
39 |
search_type="mmr",
|