Update app.py
Browse files
app.py
CHANGED
@@ -1,47 +1,62 @@
|
|
1 |
import gradio as gr
|
2 |
-
from
|
|
|
|
|
|
|
|
|
|
|
3 |
|
4 |
-
|
5 |
-
For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
|
6 |
-
"""
|
7 |
-
client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
|
8 |
|
|
|
|
|
9 |
|
10 |
-
|
11 |
-
|
12 |
-
history: list[tuple[str, str]],
|
13 |
-
system_message,
|
14 |
-
max_tokens,
|
15 |
-
temperature,
|
16 |
-
top_p,
|
17 |
-
):
|
18 |
-
messages = [{"role": "system", "content": system_message}]
|
19 |
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
if val[1]:
|
24 |
-
messages.append({"role": "assistant", "content": val[1]})
|
25 |
|
26 |
-
|
|
|
|
|
|
|
27 |
|
28 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
29 |
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
)
|
37 |
-
|
38 |
|
39 |
-
|
40 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
41 |
|
42 |
-
|
43 |
-
|
44 |
-
|
|
|
|
|
|
|
|
|
45 |
demo = gr.ChatInterface(
|
46 |
respond,
|
47 |
additional_inputs=[
|
@@ -58,6 +73,5 @@ demo = gr.ChatInterface(
|
|
58 |
],
|
59 |
)
|
60 |
|
61 |
-
|
62 |
if __name__ == "__main__":
|
63 |
-
demo.launch()
|
|
|
1 |
import gradio as gr
|
2 |
+
from langchain_community.document_loaders import PyPDFLoader, DirectoryLoader
|
3 |
+
from langchain.prompts import PromptTemplate
|
4 |
+
from langchain_community.embeddings import HuggingFaceEmbeddings
|
5 |
+
from langchain_community.vectorstores import FAISS
|
6 |
+
from langchain_community.llms import CTransformers
|
7 |
+
from langchain.chains import RetrievalQA
|
8 |
|
9 |
+
DB_FAISS_PATH = 'vectorstore/db_faiss'
|
|
|
|
|
|
|
10 |
|
11 |
+
custom_prompt_template = """Use the following pieces of information to answer the user's question.
|
12 |
+
If you don't know the answer, just say that you don't know, don't try to make up an answer.
|
13 |
|
14 |
+
Context: {context}
|
15 |
+
Question: {question}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
|
17 |
+
Only return the helpful answer below and nothing else.
|
18 |
+
Helpful answer:
|
19 |
+
"""
|
|
|
|
|
20 |
|
21 |
+
def set_custom_prompt():
|
22 |
+
prompt = PromptTemplate(template=custom_prompt_template,
|
23 |
+
input_variables=['context', 'question'])
|
24 |
+
return prompt
|
25 |
|
26 |
+
def retrieval_qa_chain(llm, prompt, db):
|
27 |
+
qa_chain = RetrievalQA.from_chain_type(llm=llm,
|
28 |
+
chain_type='stuff',
|
29 |
+
retriever=db.as_retriever(search_kwargs={'k': 2}),
|
30 |
+
return_source_documents=True,
|
31 |
+
chain_type_kwargs={'prompt': prompt}
|
32 |
+
)
|
33 |
+
return qa_chain
|
34 |
|
35 |
+
def load_llm():
|
36 |
+
llm = CTransformers(
|
37 |
+
model="TheBloke/Llama-2-7B-Chat-GGML",
|
38 |
+
model_type="llama",
|
39 |
+
max_new_tokens=512,
|
40 |
+
temperature=0.5
|
41 |
+
)
|
42 |
+
return llm
|
43 |
|
44 |
+
def qa_bot():
|
45 |
+
embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2",
|
46 |
+
model_kwargs={'device': 'cpu'})
|
47 |
+
db = FAISS.load_local(DB_FAISS_PATH, embeddings)
|
48 |
+
llm = load_llm()
|
49 |
+
qa_prompt = set_custom_prompt()
|
50 |
+
qa = retrieval_qa_chain(llm, qa_prompt, db)
|
51 |
+
return qa
|
52 |
|
53 |
+
# Define a function to respond to messages using your QA model
|
54 |
+
def respond(message, history, system_message, max_tokens, temperature, top_p):
|
55 |
+
qa_result = qa_bot()
|
56 |
+
response = qa_result({'query': message})
|
57 |
+
return response
|
58 |
+
|
59 |
+
# Create a Gradio interface using the respond function
|
60 |
demo = gr.ChatInterface(
|
61 |
respond,
|
62 |
additional_inputs=[
|
|
|
73 |
],
|
74 |
)
|
75 |
|
|
|
76 |
if __name__ == "__main__":
|
77 |
+
demo.launch()
|