Create app.py

app.py

@@ -1,63 +1,74 @@
+import os
 import gradio as gr
-from huggingface_hub import InferenceClient
-
-"""
-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
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
-
-
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
-
-    messages.append({"role": "user", "content": message})
-
-    response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
+from dotenv import load_dotenv
+from llama_index.core.callbacks import CallbackManager, LlamaDebugHandler, CBEventType
+from llama_index.core.node_parser import SentenceSplitter
+from llama_index.core.postprocessor import SimilarityPostprocessor
+from llama_index.llms.openai import OpenAI
+from llama_index.llms.groq import Groq
+from llama_index.core.base.embeddings.base import similarity
+from llama_index.llms.ollama import Ollama
+from llama_index.core import VectorStoreIndex, SimpleDirectoryReader, Settings
+from llama_index.core import StorageContext
+from llama_index.vector_stores.chroma import ChromaVectorStore
+from llama_index.embeddings.huggingface import HuggingFaceEmbedding
+from llama_index.core import load_index_from_storage
+
+
+load_dotenv()
+
+# set up LLM
+GROQ_API_KEY = os.getenv('GROQ_API_KEY')
+llm = Groq(model="llama3-70b-8192")
+Settings.llm = llm
+
+# set up callback manager
+llama_debug = LlamaDebugHandler(print_trace_on_end=True)
+callback_manager = CallbackManager([llama_debug])
+Settings.callback_manager = callback_manager
+
+# converting documents into embeddings and indexing
+embed_model = HuggingFaceEmbedding(model_name="BAAI/bge-small-en-v1.5")
+Settings.embed_model = embed_model
+
+# create splitter
+splitter = SentenceSplitter(chunk_size=1024, chunk_overlap=20)
+Settings.transformations = [splitter]
+
+if os.path.exists("./vectordb"):
+    storage_context = StorageContext.from_defaults(persist_dir="./vectordb")
+    index = load_index_from_storage(storage_context)
+else:
+    filename_fn = lambda filename: {"file_name": filename}
+    required_exts = [".pdf",".docx"]
+    reader = SimpleDirectoryReader(
+        input_dir="./data",
+        required_exts=required_exts,
+        recursive=True,
+        file_metadata=filename_fn
+    )
+    documents = reader.load_data()
+    for doc in documents:
+        doc.text = str(doc.metadata) +' '+ doc.text 
+    print("index creating with `%d` documents", len(documents))
+    index = VectorStoreIndex.from_documents(documents, embed_model=embed_model, text_splitter=splitter)
+
+    index.storage_context.persist(persist_dir="./vectordb")
+
+# set up query engine
+query_engine = index.as_query_engine(
+    similarity_top_k=5,
+    #node_postprocessors=[SimilarityPostprocessor(similarity_cutoff=0.7)],
+    verbose=True,
 )
 
+def retreive(question):
+    qns_w_source = "Answer the following question: " + question + " Followed by providing the page and file name of the source document as well, thank you!"
+    streaming_response = query_engine.query(qns_w_source)
+    #sources = streaming_response.get_formatted_sources(length=5000)
+    return str(streaming_response) # + "\n" + str(sources)
+
+demo = gr.Interface(fn=retreive, inputs="textbox", outputs="textbox")
 
 if __name__ == "__main__":
     demo.launch()