Spaces:
Running
Running
re-implement the conversational memory access
Browse files- document_qa/document_qa_engine.py +34 -15
- streamlit_app.py +9 -8
document_qa/document_qa_engine.py
CHANGED
@@ -4,10 +4,12 @@ from pathlib import Path
|
|
4 |
from typing import Union, Any
|
5 |
|
6 |
from grobid_client.grobid_client import GrobidClient
|
7 |
-
from langchain.chains import create_extraction_chain
|
8 |
-
from langchain.chains.question_answering import load_qa_chain
|
|
|
9 |
from langchain.prompts import SystemMessagePromptTemplate, HumanMessagePromptTemplate, ChatPromptTemplate
|
10 |
from langchain.retrievers import MultiQueryRetriever
|
|
|
11 |
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
12 |
from langchain.vectorstores import Chroma
|
13 |
from tqdm import tqdm
|
@@ -23,15 +25,28 @@ class DocumentQAEngine:
|
|
23 |
embeddings_map_from_md5 = {}
|
24 |
embeddings_map_to_md5 = {}
|
25 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
26 |
def __init__(self,
|
27 |
llm,
|
28 |
embedding_function,
|
29 |
qa_chain_type="stuff",
|
30 |
embeddings_root_path=None,
|
31 |
grobid_url=None,
|
|
|
32 |
):
|
33 |
self.embedding_function = embedding_function
|
34 |
self.llm = llm
|
|
|
|
|
|
|
|
|
|
|
35 |
self.chain = load_qa_chain(llm, chain_type=qa_chain_type)
|
36 |
|
37 |
if embeddings_root_path is not None:
|
@@ -87,14 +102,14 @@ class DocumentQAEngine:
|
|
87 |
return self.embeddings_map_from_md5[md5]
|
88 |
|
89 |
def query_document(self, query: str, doc_id, output_parser=None, context_size=4, extraction_schema=None,
|
90 |
-
verbose=False
|
91 |
Any, str):
|
92 |
# self.load_embeddings(self.embeddings_root_path)
|
93 |
|
94 |
if verbose:
|
95 |
print(query)
|
96 |
|
97 |
-
response = self._run_query(doc_id, query, context_size=context_size
|
98 |
response = response['output_text'] if 'output_text' in response else response
|
99 |
|
100 |
if verbose:
|
@@ -144,21 +159,21 @@ class DocumentQAEngine:
|
|
144 |
|
145 |
return parsed_output
|
146 |
|
147 |
-
def _run_query(self, doc_id, query, context_size=4
|
148 |
relevant_documents = self._get_context(doc_id, query, context_size)
|
149 |
-
|
150 |
-
|
151 |
-
|
152 |
-
|
153 |
-
|
154 |
-
|
155 |
-
memory=memory)
|
156 |
-
# return self.chain({"input_documents": relevant_documents, "question": prompt_chat_template}, return_only_outputs=True)
|
157 |
|
158 |
def _get_context(self, doc_id, query, context_size=4):
|
159 |
db = self.embeddings_dict[doc_id]
|
160 |
retriever = db.as_retriever(search_kwargs={"k": context_size})
|
161 |
relevant_documents = retriever.get_relevant_documents(query)
|
|
|
|
|
162 |
return relevant_documents
|
163 |
|
164 |
def get_all_context_by_document(self, doc_id):
|
@@ -222,11 +237,15 @@ class DocumentQAEngine:
|
|
222 |
hash = metadata[0]['hash']
|
223 |
|
224 |
if hash not in self.embeddings_dict.keys():
|
225 |
-
self.embeddings_dict[hash] = Chroma.from_texts(texts,
|
|
|
|
|
226 |
collection_name=hash)
|
227 |
else:
|
228 |
self.embeddings_dict[hash].delete(ids=self.embeddings_dict[hash].get()['ids'])
|
229 |
-
self.embeddings_dict[hash] = Chroma.from_texts(texts,
|
|
|
|
|
230 |
collection_name=hash)
|
231 |
|
232 |
self.embeddings_root_path = None
|
|
|
4 |
from typing import Union, Any
|
5 |
|
6 |
from grobid_client.grobid_client import GrobidClient
|
7 |
+
from langchain.chains import create_extraction_chain, ConversationChain, ConversationalRetrievalChain
|
8 |
+
from langchain.chains.question_answering import load_qa_chain, stuff_prompt, refine_prompts, map_reduce_prompt, \
|
9 |
+
map_rerank_prompt
|
10 |
from langchain.prompts import SystemMessagePromptTemplate, HumanMessagePromptTemplate, ChatPromptTemplate
|
11 |
from langchain.retrievers import MultiQueryRetriever
|
12 |
+
from langchain.schema import Document
|
13 |
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
14 |
from langchain.vectorstores import Chroma
|
15 |
from tqdm import tqdm
|
|
|
25 |
embeddings_map_from_md5 = {}
|
26 |
embeddings_map_to_md5 = {}
|
27 |
|
28 |
+
default_prompts = {
|
29 |
+
'stuff': stuff_prompt,
|
30 |
+
'refine': refine_prompts,
|
31 |
+
"map_reduce": map_reduce_prompt,
|
32 |
+
"map_rerank": map_rerank_prompt
|
33 |
+
}
|
34 |
+
|
35 |
def __init__(self,
|
36 |
llm,
|
37 |
embedding_function,
|
38 |
qa_chain_type="stuff",
|
39 |
embeddings_root_path=None,
|
40 |
grobid_url=None,
|
41 |
+
memory=None
|
42 |
):
|
43 |
self.embedding_function = embedding_function
|
44 |
self.llm = llm
|
45 |
+
# if memory:
|
46 |
+
# prompt = self.default_prompts[qa_chain_type].PROMPT_SELECTOR.get_prompt(llm)
|
47 |
+
# self.chain = load_qa_chain(llm, chain_type=qa_chain_type, prompt=prompt, memory=memory)
|
48 |
+
# else:
|
49 |
+
self.memory = memory
|
50 |
self.chain = load_qa_chain(llm, chain_type=qa_chain_type)
|
51 |
|
52 |
if embeddings_root_path is not None:
|
|
|
102 |
return self.embeddings_map_from_md5[md5]
|
103 |
|
104 |
def query_document(self, query: str, doc_id, output_parser=None, context_size=4, extraction_schema=None,
|
105 |
+
verbose=False) -> (
|
106 |
Any, str):
|
107 |
# self.load_embeddings(self.embeddings_root_path)
|
108 |
|
109 |
if verbose:
|
110 |
print(query)
|
111 |
|
112 |
+
response = self._run_query(doc_id, query, context_size=context_size)
|
113 |
response = response['output_text'] if 'output_text' in response else response
|
114 |
|
115 |
if verbose:
|
|
|
159 |
|
160 |
return parsed_output
|
161 |
|
162 |
+
def _run_query(self, doc_id, query, context_size=4):
|
163 |
relevant_documents = self._get_context(doc_id, query, context_size)
|
164 |
+
response = self.chain.run(input_documents=relevant_documents,
|
165 |
+
question=query)
|
166 |
+
|
167 |
+
if self.memory:
|
168 |
+
self.memory.save_context({"input": query}, {"output": response})
|
169 |
+
return response
|
|
|
|
|
170 |
|
171 |
def _get_context(self, doc_id, query, context_size=4):
|
172 |
db = self.embeddings_dict[doc_id]
|
173 |
retriever = db.as_retriever(search_kwargs={"k": context_size})
|
174 |
relevant_documents = retriever.get_relevant_documents(query)
|
175 |
+
if self.memory and len(self.memory.buffer_as_messages) > 0:
|
176 |
+
relevant_documents.append(Document(page_content="Previous conversation:\n{}\n\n".format(self.memory.buffer_as_str)))
|
177 |
return relevant_documents
|
178 |
|
179 |
def get_all_context_by_document(self, doc_id):
|
|
|
237 |
hash = metadata[0]['hash']
|
238 |
|
239 |
if hash not in self.embeddings_dict.keys():
|
240 |
+
self.embeddings_dict[hash] = Chroma.from_texts(texts,
|
241 |
+
embedding=self.embedding_function,
|
242 |
+
metadatas=metadata,
|
243 |
collection_name=hash)
|
244 |
else:
|
245 |
self.embeddings_dict[hash].delete(ids=self.embeddings_dict[hash].get()['ids'])
|
246 |
+
self.embeddings_dict[hash] = Chroma.from_texts(texts,
|
247 |
+
embedding=self.embedding_function,
|
248 |
+
metadatas=metadata,
|
249 |
collection_name=hash)
|
250 |
|
251 |
self.embeddings_root_path = None
|
streamlit_app.py
CHANGED
@@ -5,6 +5,7 @@ from tempfile import NamedTemporaryFile
|
|
5 |
|
6 |
import dotenv
|
7 |
from grobid_quantities.quantities import QuantitiesAPI
|
|
|
8 |
from langchain.llms.huggingface_hub import HuggingFaceHub
|
9 |
from langchain.memory import ConversationBufferWindowMemory
|
10 |
|
@@ -80,6 +81,7 @@ def clear_memory():
|
|
80 |
|
81 |
# @st.cache_resource
|
82 |
def init_qa(model, api_key=None):
|
|
|
83 |
if model == 'chatgpt-3.5-turbo':
|
84 |
if api_key:
|
85 |
chat = ChatOpenAI(model_name="gpt-3.5-turbo",
|
@@ -108,7 +110,7 @@ def init_qa(model, api_key=None):
|
|
108 |
st.stop()
|
109 |
return
|
110 |
|
111 |
-
return DocumentQAEngine(chat, embeddings, grobid_url=os.environ['GROBID_URL'])
|
112 |
|
113 |
|
114 |
@st.cache_resource
|
@@ -315,8 +317,7 @@ if st.session_state.loaded_embeddings and question and len(question) > 0 and st.
|
|
315 |
elif mode == "LLM":
|
316 |
with st.spinner("Generating response..."):
|
317 |
_, text_response = st.session_state['rqa'][model].query_document(question, st.session_state.doc_id,
|
318 |
-
context_size=context_size
|
319 |
-
memory=st.session_state.memory)
|
320 |
|
321 |
if not text_response:
|
322 |
st.error("Something went wrong. Contact Luca Foppiano (Foppiano.Luca@nims.co.jp) to report the issue.")
|
@@ -335,11 +336,11 @@ if st.session_state.loaded_embeddings and question and len(question) > 0 and st.
|
|
335 |
st.write(text_response)
|
336 |
st.session_state.messages.append({"role": "assistant", "mode": mode, "content": text_response})
|
337 |
|
338 |
-
|
339 |
-
|
340 |
-
|
341 |
-
|
342 |
-
|
343 |
|
344 |
elif st.session_state.loaded_embeddings and st.session_state.doc_id:
|
345 |
play_old_messages()
|
|
|
5 |
|
6 |
import dotenv
|
7 |
from grobid_quantities.quantities import QuantitiesAPI
|
8 |
+
from langchain.callbacks import PromptLayerCallbackHandler
|
9 |
from langchain.llms.huggingface_hub import HuggingFaceHub
|
10 |
from langchain.memory import ConversationBufferWindowMemory
|
11 |
|
|
|
81 |
|
82 |
# @st.cache_resource
|
83 |
def init_qa(model, api_key=None):
|
84 |
+
## For debug add: callbacks=[PromptLayerCallbackHandler(pl_tags=["langchain", "chatgpt", "document-qa"])])
|
85 |
if model == 'chatgpt-3.5-turbo':
|
86 |
if api_key:
|
87 |
chat = ChatOpenAI(model_name="gpt-3.5-turbo",
|
|
|
110 |
st.stop()
|
111 |
return
|
112 |
|
113 |
+
return DocumentQAEngine(chat, embeddings, grobid_url=os.environ['GROBID_URL'], memory=st.session_state['memory'])
|
114 |
|
115 |
|
116 |
@st.cache_resource
|
|
|
317 |
elif mode == "LLM":
|
318 |
with st.spinner("Generating response..."):
|
319 |
_, text_response = st.session_state['rqa'][model].query_document(question, st.session_state.doc_id,
|
320 |
+
context_size=context_size)
|
|
|
321 |
|
322 |
if not text_response:
|
323 |
st.error("Something went wrong. Contact Luca Foppiano (Foppiano.Luca@nims.co.jp) to report the issue.")
|
|
|
336 |
st.write(text_response)
|
337 |
st.session_state.messages.append({"role": "assistant", "mode": mode, "content": text_response})
|
338 |
|
339 |
+
# if len(st.session_state.messages) > 1:
|
340 |
+
# last_answer = st.session_state.messages[len(st.session_state.messages)-1]
|
341 |
+
# if last_answer['role'] == "assistant":
|
342 |
+
# last_question = st.session_state.messages[len(st.session_state.messages)-2]
|
343 |
+
# st.session_state.memory.save_context({"input": last_question['content']}, {"output": last_answer['content']})
|
344 |
|
345 |
elif st.session_state.loaded_embeddings and st.session_state.doc_id:
|
346 |
play_old_messages()
|