mirror of
https://github.com/katanemo/plano.git
synced 2026-04-27 01:36:33 +02:00
Adil/fix salman docs (#75)
* added the first set of docs for our technical docs * more docuemtnation changes * added support for prompt processing and updated life of a request * updated docs to including getting help sections and updated life of a request * committing local changes for getting started guide, sample applications, and full reference spec for prompt-config * updated configuration reference, added sample app skeleton, updated favico * fixed the configuration refernce file, and made minor changes to the intent detection. commit v1 for now * Updated docs with use cases and example code, updated what is arch, and made minor changes throughout * fixed imaged and minor doc fixes * add sphinx_book_theme * updated README, and make some minor fixes to documetnation * fixed README.md * fixed image width --------- Co-authored-by: Salman Paracha <salmanparacha@MacBook-Pro-261.local> Co-authored-by: Adil Hafeez <adil@katanemo.com>
This commit is contained in:
Salman Paracha
GitHub
parent
2d31aeaa36
commit
13dff3089d
33 changed files with 931 additions and 287 deletions
|
|
@ -1,78 +1,96 @@
|
|||
from flask import Flask, request, jsonify
|
||||
from datetime import datetime
|
||||
import uuid
|
||||
from langchain.memory import ConversationBufferMemory
|
||||
from langchain.schema import AIMessage, HumanMessage
|
||||
from langchain import OpenAI
|
||||
|
||||
app = Flask(__name__)
|
||||
|
||||
# Global dictionary to keep track of user conversations
|
||||
user_conversations = {}
|
||||
# Global dictionary to keep track of user memories
|
||||
user_memories = {}
|
||||
|
||||
def get_user_conversation(user_id):
|
||||
"""
|
||||
Retrieve the user's conversation history.
|
||||
If the user does not exist, initialize their conversation data.
|
||||
Retrieve the user's conversation memory using LangChain.
|
||||
If the user does not exist, initialize their conversation memory.
|
||||
"""
|
||||
if user_id not in user_conversations:
|
||||
user_conversations[user_id] = {
|
||||
'messages': []
|
||||
}
|
||||
return user_conversations[user_id]
|
||||
if user_id not in user_memories:
|
||||
user_memories[user_id] = ConversationBufferMemory(return_messages=True)
|
||||
return user_memories[user_id]
|
||||
|
||||
def update_user_conversation(user_id, client_messages, intent_changed):
|
||||
"""
|
||||
Update the user's conversation history with new messages.
|
||||
Update the user's conversation memory with new messages using LangChain.
|
||||
Each message is augmented with a UUID, timestamp, and intent change marker.
|
||||
Only new messages are added to avoid duplication.
|
||||
"""
|
||||
user_data = get_user_conversation(user_id)
|
||||
|
||||
# Existing messages in the user's conversation
|
||||
stored_messages = user_data['messages']
|
||||
memory = get_user_conversation(user_id)
|
||||
stored_messages = memory.chat_memory.messages
|
||||
|
||||
# Determine the number of stored messages
|
||||
num_stored_messages = len(stored_messages)
|
||||
|
||||
# Check for out-of-sync messages
|
||||
if num_stored_messages > len(client_messages):
|
||||
return jsonify({'error': 'Client messages are out of sync with server'}), 400
|
||||
|
||||
# Determine new messages by slicing the client messages
|
||||
new_messages = client_messages[num_stored_messages:]
|
||||
|
||||
# Process each new message
|
||||
for index, message in enumerate(new_messages):
|
||||
message_entry = {
|
||||
role = message.get('role')
|
||||
content = message.get('content')
|
||||
metadata = {
|
||||
'uuid': str(uuid.uuid4()),
|
||||
'timestamp': datetime.utcnow().isoformat(),
|
||||
'role': message.get('role'),
|
||||
'content': message.get('content'),
|
||||
'intent_changed': False # Default value
|
||||
}
|
||||
|
||||
# Mark the intent change on the last message if detected
|
||||
if intent_changed and index == len(new_messages) - 1:
|
||||
message_entry['intent_changed'] = True
|
||||
user_data['messages'].append(message_entry)
|
||||
metadata['intent_changed'] = True
|
||||
|
||||
return user_data
|
||||
# Create a new message with metadata
|
||||
if role == 'user':
|
||||
memory.chat_memory.add_message(
|
||||
HumanMessage(content=content, additional_kwargs={'metadata': metadata})
|
||||
)
|
||||
elif role == 'assistant':
|
||||
memory.chat_memory.add_message(
|
||||
AIMessage(content=content, additional_kwargs={'metadata': metadata})
|
||||
)
|
||||
else:
|
||||
# Handle other roles if necessary
|
||||
pass
|
||||
|
||||
return memory
|
||||
|
||||
def get_messages_since_last_intent(messages):
|
||||
"""
|
||||
Retrieve messages from the last intent change onwards.
|
||||
Retrieve messages from the last intent change onwards using LangChain.
|
||||
"""
|
||||
messages_since_intent = []
|
||||
for message in reversed(messages):
|
||||
# Insert message at the beginning to maintain correct order
|
||||
messages_since_intent.insert(0, message)
|
||||
if message.get('intent_changed'):
|
||||
metadata = message.additional_kwargs.get('metadata', {})
|
||||
# Break if intent_changed is True
|
||||
if metadata.get('intent_changed', False) == True:
|
||||
break
|
||||
return messages_since_intent
|
||||
|
||||
def forward_to_llm(messages):
|
||||
"""
|
||||
Simulate forwarding messages to an upstream LLM.
|
||||
Replace this with the actual API call to the LLM.
|
||||
Forward messages to an upstream LLM using LangChain.
|
||||
"""
|
||||
# For demonstration purposes, we'll return a placeholder response
|
||||
return "LLM response based on provided messages."
|
||||
# Convert messages to a conversation string
|
||||
conversation = ""
|
||||
for message in messages:
|
||||
role = 'User' if isinstance(message, HumanMessage) else 'Assistant'
|
||||
content = message.content
|
||||
conversation += f"{role}: {content}\n"
|
||||
# Use LangChain's LLM to get a response. This call is proxied through Arch for end-to-end observability and traffic management
|
||||
llm = OpenAI()
|
||||
# Create a prompt that includes the conversation
|
||||
prompt = f"{conversation}Assistant:"
|
||||
response = llm(prompt)
|
||||
return response
|
||||
|
||||
@app.route('/process_rag', methods=['POST'])
|
||||
def process_rag():
|
||||
|
|
@ -98,22 +116,37 @@ def process_rag():
|
|||
return jsonify({'error': 'Invalid value for x-arch-prompt-intent-change header'}), 400
|
||||
|
||||
# Update user conversation based on intent change
|
||||
user_data = update_user_conversation(user_id, client_messages, intent_changed)
|
||||
memory = update_user_conversation(user_id, client_messages, intent_changed)
|
||||
|
||||
# Retrieve messages since last intent change for LLM
|
||||
messages_for_llm = get_messages_since_last_intent(user_data['messages'])
|
||||
messages_for_llm = get_messages_since_last_intent(memory.chat_memory.messages)
|
||||
|
||||
# Forward messages to upstream LLM
|
||||
llm_response = forward_to_llm(messages_for_llm)
|
||||
|
||||
# Prepare the messages to return
|
||||
messages_to_return = []
|
||||
for message in memory.chat_memory.messages:
|
||||
role = 'user' if isinstance(message, HumanMessage) else 'assistant'
|
||||
content = message.content
|
||||
metadata = message.additional_kwargs.get('metadata', {})
|
||||
message_entry = {
|
||||
'uuid': metadata.get('uuid'),
|
||||
'timestamp': metadata.get('timestamp'),
|
||||
'role': role,
|
||||
'content': content,
|
||||
'intent_changed': metadata.get('intent_changed', False)
|
||||
}
|
||||
messages_to_return.append(message_entry)
|
||||
|
||||
# Prepare the response
|
||||
response = {
|
||||
'user_id': user_id,
|
||||
'messages': user_data['messages'],
|
||||
'messages': messages_to_return,
|
||||
'llm_response': llm_response
|
||||
}
|
||||
|
||||
return jsonify(response), 200
|
||||
|
||||
if __name__ == '__main__':
|
||||
app.run(debug=True)
|
||||
app.run(debug=True)
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue