chat appication
Browse files- .gitignore +165 -0
- app.py +157 -0
- datastore.py +107 -0
- embeddings.py +12 -0
.gitignore
ADDED
@@ -0,0 +1,165 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Byte-compiled / optimized / DLL files
|
2 |
+
__pycache__/
|
3 |
+
*.py[cod]
|
4 |
+
*$py.class
|
5 |
+
|
6 |
+
# C extensions
|
7 |
+
*.so
|
8 |
+
|
9 |
+
# Distribution / packaging
|
10 |
+
.Python
|
11 |
+
build/
|
12 |
+
develop-eggs/
|
13 |
+
dist/
|
14 |
+
downloads/
|
15 |
+
eggs/
|
16 |
+
.eggs/
|
17 |
+
lib/
|
18 |
+
lib64/
|
19 |
+
parts/
|
20 |
+
sdist/
|
21 |
+
var/
|
22 |
+
wheels/
|
23 |
+
share/python-wheels/
|
24 |
+
*.egg-info/
|
25 |
+
.installed.cfg
|
26 |
+
*.egg
|
27 |
+
MANIFEST
|
28 |
+
|
29 |
+
# PyInstaller
|
30 |
+
# Usually these files are written by a python script from a template
|
31 |
+
# before PyInstaller builds the exe, so as to inject date/other infos into it.
|
32 |
+
*.manifest
|
33 |
+
*.spec
|
34 |
+
|
35 |
+
# Installer logs
|
36 |
+
pip-log.txt
|
37 |
+
pip-delete-this-directory.txt
|
38 |
+
|
39 |
+
# Unit test / coverage reports
|
40 |
+
htmlcov/
|
41 |
+
.tox/
|
42 |
+
.nox/
|
43 |
+
.coverage
|
44 |
+
.coverage.*
|
45 |
+
.cache
|
46 |
+
nosetests.xml
|
47 |
+
coverage.xml
|
48 |
+
*.cover
|
49 |
+
*.py,cover
|
50 |
+
.hypothesis/
|
51 |
+
.pytest_cache/
|
52 |
+
cover/
|
53 |
+
|
54 |
+
# Translations
|
55 |
+
*.mo
|
56 |
+
*.pot
|
57 |
+
|
58 |
+
# Django stuff:
|
59 |
+
*.log
|
60 |
+
local_settings.py
|
61 |
+
db.sqlite3
|
62 |
+
db.sqlite3-journal
|
63 |
+
|
64 |
+
# Flask stuff:
|
65 |
+
instance/
|
66 |
+
.webassets-cache
|
67 |
+
|
68 |
+
# Scrapy stuff:
|
69 |
+
.scrapy
|
70 |
+
|
71 |
+
# Sphinx documentation
|
72 |
+
docs/_build/
|
73 |
+
|
74 |
+
# PyBuilder
|
75 |
+
.pybuilder/
|
76 |
+
target/
|
77 |
+
|
78 |
+
# Jupyter Notebook
|
79 |
+
.ipynb_checkpoints
|
80 |
+
|
81 |
+
# IPython
|
82 |
+
profile_default/
|
83 |
+
ipython_config.py
|
84 |
+
|
85 |
+
# pyenv
|
86 |
+
# For a library or package, you might want to ignore these files since the code is
|
87 |
+
# intended to run in multiple environments; otherwise, check them in:
|
88 |
+
# .python-version
|
89 |
+
|
90 |
+
# pipenv
|
91 |
+
# According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
|
92 |
+
# However, in case of collaboration, if having platform-specific dependencies or dependencies
|
93 |
+
# having no cross-platform support, pipenv may install dependencies that don't work, or not
|
94 |
+
# install all needed dependencies.
|
95 |
+
#Pipfile.lock
|
96 |
+
|
97 |
+
# poetry
|
98 |
+
# Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
|
99 |
+
# This is especially recommended for binary packages to ensure reproducibility, and is more
|
100 |
+
# commonly ignored for libraries.
|
101 |
+
# https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
|
102 |
+
#poetry.lock
|
103 |
+
|
104 |
+
# pdm
|
105 |
+
# Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
|
106 |
+
#pdm.lock
|
107 |
+
# pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
|
108 |
+
# in version control.
|
109 |
+
# https://pdm.fming.dev/#use-with-ide
|
110 |
+
.pdm.toml
|
111 |
+
|
112 |
+
# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
|
113 |
+
__pypackages__/
|
114 |
+
|
115 |
+
# Celery stuff
|
116 |
+
celerybeat-schedule
|
117 |
+
celerybeat.pid
|
118 |
+
|
119 |
+
# SageMath parsed files
|
120 |
+
*.sage.py
|
121 |
+
|
122 |
+
# Environments
|
123 |
+
.env
|
124 |
+
.venv
|
125 |
+
env/
|
126 |
+
venv/
|
127 |
+
ENV/
|
128 |
+
env.bak/
|
129 |
+
venv.bak/
|
130 |
+
|
131 |
+
# Spyder project settings
|
132 |
+
.spyderproject
|
133 |
+
.spyproject
|
134 |
+
|
135 |
+
# Rope project settings
|
136 |
+
.ropeproject
|
137 |
+
|
138 |
+
# mkdocs documentation
|
139 |
+
/site
|
140 |
+
|
141 |
+
# mypy
|
142 |
+
.mypy_cache/
|
143 |
+
.dmypy.json
|
144 |
+
dmypy.json
|
145 |
+
|
146 |
+
# Pyre type checker
|
147 |
+
.pyre/
|
148 |
+
|
149 |
+
# pytype static type analyzer
|
150 |
+
.pytype/
|
151 |
+
|
152 |
+
# Cython debug symbols
|
153 |
+
cython_debug/
|
154 |
+
|
155 |
+
# PyCharm
|
156 |
+
# JetBrains specific template is maintained in a separate JetBrains.gitignore that can
|
157 |
+
# be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
|
158 |
+
# and can be added to the global gitignore or merged into this file. For a more nuclear
|
159 |
+
# option (not recommended) you can uncomment the following to ignore the entire idea folder.
|
160 |
+
#.idea/
|
161 |
+
|
162 |
+
|
163 |
+
models
|
164 |
+
chroma
|
165 |
+
*.ipynb
|
app.py
ADDED
@@ -0,0 +1,157 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import datetime
|
2 |
+
import os
|
3 |
+
import uuid
|
4 |
+
|
5 |
+
import fitz
|
6 |
+
import streamlit as st
|
7 |
+
from langchain_text_splitters import RecursiveCharacterTextSplitter
|
8 |
+
from llama_cpp import Llama
|
9 |
+
|
10 |
+
from datastore import ChromaStore
|
11 |
+
from embeddings import Embedding
|
12 |
+
|
13 |
+
#### state
|
14 |
+
if "chat_history" not in st.session_state:
|
15 |
+
st.session_state.chat_history = []
|
16 |
+
if "document_submitted" not in st.session_state:
|
17 |
+
st.session_state.document_submitted = False
|
18 |
+
|
19 |
+
|
20 |
+
def phi3(input: str, relevant_chunks: list):
|
21 |
+
llm = Llama(
|
22 |
+
model_path=os.path.join(
|
23 |
+
os.getcwd(),
|
24 |
+
"models",
|
25 |
+
"Phi-3.1-mini-4k-instruct-Q4_K_M.gguf",
|
26 |
+
),
|
27 |
+
n_ctx=2000,
|
28 |
+
n_threads=1, # The number of CPU threads to use,
|
29 |
+
n_gpu_layers=0, # The number of layers to offload to GPU,
|
30 |
+
)
|
31 |
+
|
32 |
+
prompt = f"""CONTENT: {relevant_chunks}\n\nQUESTION: {input}\n\nFrom the given CONTENT, Please answer the QUESTION."""
|
33 |
+
|
34 |
+
output = llm(
|
35 |
+
f"<|user|>\n{prompt}<|end|>\n<|assistant|>",
|
36 |
+
max_tokens=2000,
|
37 |
+
stop=["<|end|>"],
|
38 |
+
echo=True,
|
39 |
+
)
|
40 |
+
|
41 |
+
cleaned_output = output["choices"][0]["text"].split("<|assistant|>", 1)[-1].strip()
|
42 |
+
return cleaned_output
|
43 |
+
|
44 |
+
|
45 |
+
def generate_unique_id():
|
46 |
+
unique_id = uuid.uuid4()
|
47 |
+
current_time = datetime.datetime.now().strftime("%Y%m%d%H%M%S%f")
|
48 |
+
combined_id = f"{unique_id}-{current_time}"
|
49 |
+
return combined_id
|
50 |
+
|
51 |
+
|
52 |
+
def add_to_vectorstore(content: str, chunk_size: int = 500, chunk_overlap: int = 20):
|
53 |
+
chromastore = ChromaStore(collection_name="pdf_store")
|
54 |
+
|
55 |
+
# delete if already exist
|
56 |
+
if "pdf_store" in chromastore.list_collections():
|
57 |
+
chromastore.delete("pdf_store")
|
58 |
+
st.toast("Old database cleaned!")
|
59 |
+
collection = chromastore.create()
|
60 |
+
# chunkify content
|
61 |
+
text_splitter = RecursiveCharacterTextSplitter(
|
62 |
+
chunk_size=chunk_size,
|
63 |
+
chunk_overlap=chunk_overlap,
|
64 |
+
length_function=len,
|
65 |
+
is_separator_regex=False,
|
66 |
+
)
|
67 |
+
chunks = text_splitter.split_text(content)
|
68 |
+
|
69 |
+
# generate embeddings and ids
|
70 |
+
embeddings, ids = [], []
|
71 |
+
for i, chunk in enumerate(chunks):
|
72 |
+
embeddings.append(Embedding.encode_text(chunk).tolist())
|
73 |
+
ids.append(generate_unique_id())
|
74 |
+
|
75 |
+
# add to vectorstore
|
76 |
+
chromastore.add(
|
77 |
+
collection=collection,
|
78 |
+
embeddings=embeddings,
|
79 |
+
documents=chunks,
|
80 |
+
ids=ids,
|
81 |
+
)
|
82 |
+
|
83 |
+
|
84 |
+
def similarity_search(query: str):
|
85 |
+
chromastore = ChromaStore(collection_name="pdf_store")
|
86 |
+
collection = chromastore.create()
|
87 |
+
query_embedding = Embedding.encode_text(query).tolist()
|
88 |
+
return chromastore.query(collection=collection, query_embedding=query_embedding)
|
89 |
+
|
90 |
+
|
91 |
+
def main():
|
92 |
+
st.set_page_config(page_icon="🤖", page_title="Phi 3 RAG", layout="wide")
|
93 |
+
st.markdown(
|
94 |
+
"""<h1 style="text-align:center;">Phi 3 RAG</h1>""", unsafe_allow_html=True
|
95 |
+
)
|
96 |
+
st.markdown(
|
97 |
+
"""<h3 style="text-align:center;">Conversational RAG application that utilizes local stack, <a href="https://huggingface.co/bartowski/Phi-3-medium-4k-instruct-GGUF">Phi-3 mini 4k instruct GGUF</a> and <a href="https://docs.trychroma.com/getting-started">ChromaDB</h3>""",
|
98 |
+
unsafe_allow_html=True,
|
99 |
+
)
|
100 |
+
layout = st.columns(2)
|
101 |
+
|
102 |
+
with layout[0]:
|
103 |
+
with st.container(border=True, height=550):
|
104 |
+
uploaded_file = st.file_uploader(
|
105 |
+
label="Upload document to search",
|
106 |
+
type="PDF",
|
107 |
+
accept_multiple_files=False,
|
108 |
+
)
|
109 |
+
submit = st.button("submit")
|
110 |
+
|
111 |
+
chunk_size = st.slider(
|
112 |
+
label="Chunk_size", min_value=100, max_value=2000, step=100
|
113 |
+
)
|
114 |
+
chunk_overlap = st.slider(
|
115 |
+
label="Chunk overlap", min_value=10, max_value=500, step=10
|
116 |
+
)
|
117 |
+
if uploaded_file is not None and submit is not False:
|
118 |
+
# load in vectorstore
|
119 |
+
doc = fitz.open(stream=uploaded_file.read(), filetype="pdf")
|
120 |
+
text = ""
|
121 |
+
for page in doc:
|
122 |
+
text += page.get_text()
|
123 |
+
doc.close()
|
124 |
+
|
125 |
+
# add to vectorstore
|
126 |
+
add_to_vectorstore(text, chunk_size, chunk_overlap)
|
127 |
+
st.session_state.document_submitted = True
|
128 |
+
st.toast("Document added successfully added to vectorstore", icon="✅")
|
129 |
+
|
130 |
+
# chats
|
131 |
+
with layout[1]:
|
132 |
+
with st.container(border=True, height=550):
|
133 |
+
if st.session_state.document_submitted:
|
134 |
+
user_input = st.chat_input("Ask me!")
|
135 |
+
if user_input is not None:
|
136 |
+
st.session_state.chat_history.append(
|
137 |
+
{"role": "user", "content": str(user_input)}
|
138 |
+
)
|
139 |
+
|
140 |
+
with st.spinner("Thinking..."):
|
141 |
+
# find on vector store
|
142 |
+
relevant_chunks = similarity_search(user_input)
|
143 |
+
response = phi3(
|
144 |
+
input=user_input, relevant_chunks=relevant_chunks
|
145 |
+
)
|
146 |
+
st.session_state.chat_history.append(
|
147 |
+
{"role": "assistant", "content": str(response)}
|
148 |
+
)
|
149 |
+
|
150 |
+
# display messages
|
151 |
+
for message in reversed(st.session_state.chat_history):
|
152 |
+
with st.chat_message(message["role"]):
|
153 |
+
st.markdown(message["content"])
|
154 |
+
|
155 |
+
|
156 |
+
if __name__ == "__main__":
|
157 |
+
main()
|
datastore.py
ADDED
@@ -0,0 +1,107 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
__import__("pysqlite3")
|
2 |
+
import sys
|
3 |
+
|
4 |
+
sys.modules["sqlite3"] = sys.modules.pop("pysqlite3")
|
5 |
+
|
6 |
+
|
7 |
+
import uuid
|
8 |
+
from collections import defaultdict
|
9 |
+
from typing import Any, List
|
10 |
+
|
11 |
+
import chromadb
|
12 |
+
import numpy as np
|
13 |
+
from chromadb import Collection
|
14 |
+
|
15 |
+
from embeddings import Embedding
|
16 |
+
|
17 |
+
|
18 |
+
class ChromaStore:
|
19 |
+
def __init__(
|
20 |
+
self,
|
21 |
+
collection_name: str,
|
22 |
+
storage_path: str = "./chroma",
|
23 |
+
database: str = "database",
|
24 |
+
metadata: dict = {"hnsw:space": "cosine"},
|
25 |
+
) -> None:
|
26 |
+
"""Initiate Chromadb
|
27 |
+
- collection_name(str): name of the collection
|
28 |
+
- metadata(dict): available options for 'hnsw:space' are 'l2', 'ip' or 'cosine'.
|
29 |
+
"""
|
30 |
+
|
31 |
+
self.collection_name = collection_name
|
32 |
+
self.metadata = metadata
|
33 |
+
self.storage_path = storage_path
|
34 |
+
self.database = database
|
35 |
+
|
36 |
+
self.client = chromadb.PersistentClient(path=self.storage_path)
|
37 |
+
|
38 |
+
def _health_check(self) -> bool:
|
39 |
+
return isinstance(self.client.heartbeat(), int)
|
40 |
+
|
41 |
+
def create(self):
|
42 |
+
collection = self.client.get_or_create_collection(
|
43 |
+
name=self.collection_name,
|
44 |
+
)
|
45 |
+
return collection
|
46 |
+
|
47 |
+
def add(
|
48 |
+
self,
|
49 |
+
collection: Collection,
|
50 |
+
embeddings: List[float],
|
51 |
+
documents: List[str],
|
52 |
+
ids: List[str],
|
53 |
+
):
|
54 |
+
"""Add embeddings, documents to index or collection.
|
55 |
+
|
56 |
+
Args:
|
57 |
+
- collection: created collection.
|
58 |
+
- embeddings: list of embeddings
|
59 |
+
- documents: text documents
|
60 |
+
- ids: list of ids"""
|
61 |
+
try:
|
62 |
+
collection.add(
|
63 |
+
embeddings=embeddings,
|
64 |
+
ids=ids,
|
65 |
+
documents=documents,
|
66 |
+
)
|
67 |
+
except Exception as e:
|
68 |
+
raise Exception(f"Failed to add documents to Chroma store. {e}")
|
69 |
+
|
70 |
+
def query(
|
71 |
+
self,
|
72 |
+
collection: Collection,
|
73 |
+
query_embedding: List[float],
|
74 |
+
top_k: int = 3,
|
75 |
+
) -> list:
|
76 |
+
"""Retrieve relevant images from chroma database.
|
77 |
+
|
78 |
+
Args:
|
79 |
+
- collection: created collection.
|
80 |
+
- query_embedding: query image embedding.
|
81 |
+
- top_k (int): top k images to retrieve.
|
82 |
+
|
83 |
+
Returns:
|
84 |
+
- list of images along with their score.
|
85 |
+
"""
|
86 |
+
result = collection.query(query_embeddings=query_embedding, n_results=top_k)
|
87 |
+
relevant_chunks = [chunk for chunk in result["documents"][0]]
|
88 |
+
return relevant_chunks
|
89 |
+
# scores = [round(score, 3) for score in result["distances"][0]]
|
90 |
+
# return list(zip(relevant_chunks, scores))
|
91 |
+
|
92 |
+
def delete(self, collection_name: str):
|
93 |
+
try:
|
94 |
+
self.client.delete_collection(collection_name)
|
95 |
+
return True
|
96 |
+
except Exception as e:
|
97 |
+
raise Exception("Failed to delete collection", e)
|
98 |
+
|
99 |
+
def list_collections(self):
|
100 |
+
return self.client.list_collections()
|
101 |
+
|
102 |
+
@staticmethod
|
103 |
+
def collection_info(collection: Collection):
|
104 |
+
info = defaultdict(str)
|
105 |
+
info["count"] = collection.count()
|
106 |
+
info["top_10_items"] = collection.peek()
|
107 |
+
return info
|
embeddings.py
ADDED
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from typing import List
|
2 |
+
|
3 |
+
import torch
|
4 |
+
from sentence_transformers import SentenceTransformer
|
5 |
+
|
6 |
+
|
7 |
+
class Embedding:
|
8 |
+
model = SentenceTransformer("sentence-transformers/all-MiniLM-L6-v2")
|
9 |
+
|
10 |
+
@classmethod
|
11 |
+
def encode_text(cls, text: str):
|
12 |
+
return cls.model.encode(text)
|