Delete agent.py

agent.py

@@ -1,97 +0,0 @@
-# agent.py – LangChain-enhanced version with voice input transcription
-
-import json
-import openai
-from gtts import gTTS
-import tempfile
-from langchain.chains import RetrievalQA
-from langchain_community.embeddings import HuggingFaceEmbeddings
-from langchain_community.vectorstores import FAISS
-from langchain_community.chat_models import ChatOpenAI
-from langchain.schema import Document
-from langchain.prompts import PromptTemplate
-from openai import OpenAI
-
-client = OpenAI()
-
-# Load dialogue lines and metadata from the grounding file
-def load_script_documents(path):
-    docs = []
-    with open(path, "r", encoding="utf-8") as f:
-        for line in f:
-            item = json.loads(line)
-            docs.append(Document(page_content=item["dialogue"], metadata={
-                "scene": item.get("scene", ""),
-                "emotion": item.get("emotion", ""),
-                "character": item.get("character", "")
-            }))
-    return docs
-
-# Build FAISS retriever using sentence-transformer embeddings
-def create_script_retriever(jsonl_path):
-    docs = load_script_documents(jsonl_path)
-    embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
-    vectorstore = FAISS.from_documents(docs, embeddings)
-    return vectorstore.as_retriever()
-
-# Create LangChain QA agent
-def create_agent():
-    retriever = create_script_retriever("the_father_segments.jsonl")
-
-    prompt = PromptTemplate.from_template("""
-    You are a compassionate companion modeled on scenes from the film "The Father".
-    Use the memories below to respond empathetically to the user's message.
-
-    Memories:
-    {context}
-
-    Message:
-    {question}
-    """)
-
-    chain = RetrievalQA.from_chain_type(
-        llm=ChatOpenAI(temperature=0.4),
-        retriever=retriever,
-        chain_type_kwargs={"prompt": prompt}
-    )
-    return chain
-
-# Instantiate the agent
-qa_chain = create_agent()
-
-# Text input
-def chat_with_agent(message, history):
-    answer = qa_chain.run(message)
-    history.append((message, answer))
-    return "", history, synthesize_voice(answer)
-    # return "", history
-
-# Voice input handler
-def handle_voice(audio, history):
-    if audio is None:
-        return history, None
-    text = transcribe_audio_to_text(audio)
-    answer = qa_chain.run(text)
-    history.append((text, answer))
-    return history, synthesize_voice(answer)
-
-
-# Audio transcription (OpenAI Whisper)    
-def transcribe_audio_to_text(audio_file):
-    if not audio_file:
-        return "No audio received."
-    try:
-        client = OpenAI()
-        with open(audio_file, "rb") as f:
-            transcript = client.audio.transcriptions.create(model="whisper-1", file=f)
-        return transcript.text  # ✅ Access as an attribute, not a dict key
-    except Exception as e:
-        return f"[Transcription Error] {e}"
-
-# Convert bot reply to spoken audio using gTTS
-def synthesize_voice(text):
-    tts = gTTS(text)
-    temp_audio = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
-    tts.save(temp_audio.name)
-    return temp_audio.name
-