Create app.py

app.py

@@ -0,0 +1,56 @@
+import streamlit as st, os
+import tempfile
+from pytube import YouTube
+import whisper
+from dotenv import load_dotenv
+# langchain imports
+from langchain_community.document_loaders import TextLoader
+from langchain_community.vectorstores import FAISS
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_openai import OpenAIEmbeddings
+from langchain_openai import OpenAI
+from langchain_core.prompts import ChatPromptTemplate
+
+
+# function for getting transcribing YouTube video
+def get_yt_trans(url):
+    if not os.path.exists("transcription.txt"):
+        yt_obj = YouTube(url)
+        st.write(f"Youtube object: {yt_obj} \n\n\n")
+        video_audio = yt_obj.streams.filter(only_audio=True).first()
+        st.write(f"Audio object: {video_audio} \n\n\n")
+    # define model for audio to text conversion
+    audio_to_text_model = whisper.load_model('base')
+    st.write(f"Model object: {audio_to_text_model} \n\n\n")
+    # store transcription in local file
+    with tempfile.TemporaryDirectory() as tmpdir:
+        audio_file = video_audio.download(output_path=tmpdir)
+        st.write(f"Audio file: {audio_file} \n\n\n")
+        transcription = audio_to_text_model.transcribe(audio_file, fp16=False)["text"].strip()
+        st.write(f"Transcription: {transcription} \n\n\n")
+        with open("transcription.txt", "w") as file:
+            file.write(transcription)
+
+
+# define api keys
+load_dotenv()
+os.environ['OPEN_API_KEY'] = os.getenv('OPENAI_API_KEY')
+
+# define LLM
+llm = OpenAI(temperature=0)
+
+# Set page config
+st.set_page_config(
+    page_title="YouTube Talks",
+    page_icon='📽️'
+)
+st.header('Query YouTube Videos!')
+video_url = st.text_input("Enter video URL")
+submit_button = st.button("Submit")
+
+
+# on button press
+if submit_button:
+    get_yt_trans(video_url)
+    
+