first working version

.gitignore

@@ -0,0 +1 @@
+.env

README.md

@@ -20,4 +20,7 @@ You are a podcast editor that specialized to create a script out of a webpage.
 [BREAKS]
 - Use the <break time="1s"/> tag to add a pause in the speech. Use it after each paragraph and after a sentence when appropriate to get to a relax tone.
 - Use <break time="3s"/> for longer pauses, to emphasize a point.
+
+[LINK]
+{link}
 ```

app.py

@@ -0,0 +1,87 @@
+# import numpy as np
+import gradio
+from openai import OpenAI
+import tempfile
+import os
+
+from dotenv import load_dotenv
+load_dotenv()
+
+import uuid
+
+
+client = OpenAI()
+
+DEFAULT_SYSTEM_PROMPT = """
+You are a podcast editor that specialized to create a script out of a webpage.
+
+[TASKS]
+- You will receive a link to a webpage about some content.
+- You will create a script out of the content.
+- The script should be 1 minute long if you read it out loud.
+- Start with an intro to peak the interest of the listener.
+- Then, summarize the content in a way that is easy to understand. Ask questions about the the content and answer them.
+- Conclude with the most intriguing part of the content.
+- Refrain from adding section headers in the script like [INTRUCTION], [CONTENT], [CONCLUSION].
+- The script should be inspiring, written in colloquialism with english words and proverbs understood around the world.
+- Write in a relax tone.
+- Use filler words like 'um', 'ah', 'well' etc. to make it sound more natural.
+
+[BREAKS]
+- Use the <break time="1s"/> tag to add a pause in the speech. Use it after each paragraph and after a sentence when appropriate to get to a relax tone.
+- Use <break time="3s"/> for longer pauses, to emphasize a point.
+"""
+
+def generate_episode(system_prompt, weblink):
+ # sr = 48000
+ # a4_freq, tones_from_a4 = 440, 12 * (octave - 4) + (note - 9)
+ # frequency = a4_freq * 2 ** (tones_from_a4 / 12)
+ # duration = int(duration)
+ # audio = np.linspace(0, duration, duration * sr)
+ # audio = (20000 * np.sin(audio * (2 * np.pi * frequency))).astype(np.int16)
+ # return sr, audio
+ response = client.chat.completions.create(
+ model="gpt-4o",
+ messages=[
+ {"role": "system", "content": system_prompt},
+ {"role": "user", "content": weblink},
+ ]
+ )
+ script = response.choices[0].message.content 
+ response = client.audio.speech.create(
+ model="tts-1",
+ voice="fable",
+ input=script,
+ )
+ 
+ with tempfile.NamedTemporaryFile(suffix=".mp3", delete=False) as temp_file:
+ temp_file.write(response.content) 
+ # state["mp3_file"] = temp_file.name
+ return temp_file.name
+
+
+def cleanup(state):
+ mp3_file_name = state.get("mp3_file") 
+ if mp3_file_name and os.path.exists(mp3_file_name):
+ os.remove(mp3_file_name)
+ print(f"Removed file: {mp3_file_name}")
+ else:
+ print(f"No file found to delete: {mp3_file_name}")
+
+demo = gradio.Interface(
+ fn=generate_episode,
+ inputs=[
+ gradio.Textbox(value=DEFAULT_SYSTEM_PROMPT, label="System Prompt"),
+ gradio.Textbox(value="https://en.wikipedia.org/wiki/Mount_Tambora", label="Weblink"),
+ # gradio.State(), # State to track the mp3 file
+
+ ],
+ # outputs=["audio", gradio.State()],
+ outputs="audio",
+
+)
+
+# demo.cleanup(cleanup)
+
+if __name__ == "__main__":
+ demo.launch()

requirements.txt

@@ -0,0 +1,4 @@
+gradio
+numpy
+python-dotenv
+openai