Upload 2 files

app.py

@@ -0,0 +1,112 @@
+from fastapi import FastAPI, HTTPException
+from pydantic import BaseModel
+import os
+import dotenv
+from llama_index.core.indices.vector_store.base import VectorStoreIndex
+from llama_index.vector_stores.qdrant import QdrantVectorStore
+from llama_index.embeddings.fastembed import FastEmbedEmbedding
+from llama_index.core import Settings
+import qdrant_client
+from llama_index.llms.gemini import Gemini
+from llama_index.core.memory import ChatMemoryBuffer
+from llama_index.readers.web import FireCrawlWebReader
+
+dotenv.load_dotenv()
+
+app = FastAPI()
+
+# Initialize session state equivalent
+state = {
+    'setup_complete': False,
+    'documents': None,
+    'chat_history': [],
+    'index': None,
+    'url': "",
+    'collection_name': "",
+}
+
+os.environ["GOOGLE_API_KEY"] = os.getenv("GOOGLE_API_KEY")
+
+# Setup functions
+def embed_setup():
+    Settings.embed_model = FastEmbedEmbedding(model_name="BAAI/bge-small-en-v1.5")
+    Settings.llm = Gemini(temperature=0.1, model_name="models/gemini-pro")
+
+def qdrant_setup():
+    client = qdrant_client.QdrantClient(
+        os.getenv("QDRANT_URL"),
+        api_key=os.getenv("QDRANT_API_KEY"),
+    )
+    return client
+
+def llm_setup():
+    llm = Gemini(api_key=os.getenv("GOOGLE_API_KEY"), temperature=0.1, model_name="models/gemini-pro")
+    return llm
+
+def ingest_documents(url):
+    firecrawl_reader = FireCrawlWebReader(
+        api_key=os.getenv("FIRECRAWL_API_KEY"),
+        mode="scrape",
+    )
+    documents = firecrawl_reader.load_data(url=url)
+    return documents
+
+class SetupRequest(BaseModel):
+    url: str = None
+    collection_name: str
+
+class QueryRequest(BaseModel):
+    query: str
+
+@app.post("/setup/")
+async def setup(request: SetupRequest):
+    state['url'] = request.url
+    state['collection_name'] = request.collection_name
+
+    embed_setup()
+    client = qdrant_setup()
+    llm = llm_setup()
+    vector_store = QdrantVectorStore(client=client, collection_name=state['collection_name'])
+
+    if state['url']:
+        state['documents'] = ingest_documents(state['url'])
+        state['index'] = VectorStoreIndex.from_documents(state['documents'], vector_store=vector_store)
+        state['setup_complete'] = True
+        return {"message": f"Documents ingested from {state['url']} and query engine setup completed successfully!"}
+    else:
+        state['index'] = VectorStoreIndex.from_vector_store(vector_store=vector_store)
+        state['setup_complete'] = True
+        return {"message": f"Query engine setup completed successfully using existing collection: {state['collection_name']}"}
+
+@app.post("/query/")
+async def query(request: QueryRequest):
+    if not state['setup_complete']:
+        raise HTTPException(status_code=400, detail="Please complete the setup first")
+    
+    memory = ChatMemoryBuffer.from_defaults(token_limit=4000)
+    chat_engine = state['index'].as_chat_engine(
+        chat_mode="context",
+        memory=memory,
+        system_prompt=(
+            """You are an AI assistant for developers, specializing in technical documentation. 
+            Your task is to provide accurate, concise, and helpful responses based on the given documentation context.
+            Given this context, please respond to the following user query:
+            {query_str}
+            Your response:"""
+        ),
+    )
+    
+    response = chat_engine.chat(request.query)
+    state['chat_history'].append(("User", request.query))
+    state['chat_history'].append(("Assistant", str(response.response)))
+
+    return {"response": response.response}
+
+@app.get("/chat-history/")
+async def get_chat_history():
+    return {"chat_history": state['chat_history']}
+
+@app.post("/clear-chat/")
+async def clear_chat():
+    state['chat_history'] = []
+    return {"message": "Chat history cleared!"}

requirements.txt

@@ -0,0 +1,17 @@
+llama-index==0.10.17
+llama-index-vector-stores-qdrant
+fastembed
+llama-index-embeddings-fastembed
+llama-index-llms-gemini
+llama-index-embeddings-gemini
+gradio
+llama-index-readers-web
+qdrant-client
+firecrawl-py
+streamlit-analytics2
+llama-index-readers-youtube-transcript
+pypdf2
+python-dotenv
+fastapi
+pydantic
+uvicorn