app tinytutor

app.py

@@ -0,0 +1,146 @@
+import os
+import asyncio
+import google.generativeai as genai
+
+from google.adk.agents import Agent
+from google.adk.models.google_llm import Gemini
+from google.adk.runners import InMemoryRunner
+from google.adk.tools import google_search
+from google.genai import types
+
+from google.cloud import texttospeech
+from pydub import AudioSegment
+import gradio as gr
+
+# --- Configure API Keys ---
+GOOGLE_API_KEY = os.getenv("GOOGLE_API_KEY")
+if not GOOGLE_API_KEY:
+    raise RuntimeError("❌ Missing GOOGLE_API_KEY environment variable.")
+genai.configure(api_key=GOOGLE_API_KEY)
+
+SERVICE_ACCOUNT_JSON = os.getenv("GCP_VI_SERVICE_ACCOUNT_JSON")
+if not SERVICE_ACCOUNT_JSON:
+    raise RuntimeError("❌ Missing GCP_VI_SERVICE_ACCOUNT_JSON environment variable.")
+
+with open("tinytutor-tss-agent.json", "w") as f:
+    f.write(SERVICE_ACCOUNT_JSON)
+
+os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = "tinytutor-tss-agent.json"
+tts_client = texttospeech.TextToSpeechClient()
+
+# --- Retry Options ---
+retry_config = types.HttpRetryOptions(
+    attempts=5,
+    exp_base=7,
+    initial_delay=1,
+    http_status_codes=[429, 500, 503, 504]
+)
+
+# --- Pedagogy Agent ---
+pedagogy_agent = Agent(
+    name="PedagogyAgent",
+    model=Gemini(model="gemini-2.5-flash-lite", retry_options=retry_config),
+    description="Explains topics in simple ELI5 style.",
+    instruction="Explain the topic like I'm 5. Use google_search if needed.",
+    tools=[google_search],
+)
+runner = InMemoryRunner(agent=pedagogy_agent)
+
+async def run_pedagogy_async(topic: str) -> str:
+    response = await runner.run_debug(topic)
+    return response[0].content.parts[0].text
+
+# --- ScriptWriter Agent ---
+SCRIPTWRITER_SYSTEM_PROMPT = """
+You are a Teacher.
+... (same as your original prompt) ...
+"""
+
+def run_scriptwriter(explanation: str) -> str:
+    model = genai.GenerativeModel(
+        model_name="gemini-2.5-flash",
+        system_instruction=SCRIPTWRITER_SYSTEM_PROMPT
+    )
+    response = model.generate_content(
+        f"Write a children's story based on this:\n{explanation}",
+        generation_config=genai.GenerationConfig(
+            temperature=0.9,
+            max_output_tokens=4096
+        )
+    )
+    try:
+        return response.text
+    except Exception:
+        try:
+            return response.candidates[0].content.parts[0].text
+        except Exception:
+            return "⚠️ ScriptWriter failed."
+
+# --- Audio Generator ---
+def chunk_text(text, max_chars=4500):
+    text = text.strip()
+    if len(text) <= max_chars:
+        return [text]
+    chunks = []
+    while len(text) > max_chars:
+        cut = text.rfind(". ", 0, max_chars)
+        if cut == -1:
+            cut = max_chars
+        chunks.append(text[:cut+1])
+        text = text[cut+1:].strip()
+    chunks.append(text)
+    return chunks
+
+def tts_segment(text):
+    synthesis_input = texttospeech.SynthesisInput(text=text)
+    voice = texttospeech.VoiceSelectionParams(
+        language_code="en-US",
+        name="en-US-Journey-F"
+    )
+    audio_cfg = texttospeech.AudioConfig(
+        audio_encoding=texttospeech.AudioEncoding.MP3,
+        speaking_rate=0.94,
+        pitch=0.0,
+        volume_gain_db=0.0
+    )
+    response = tts_client.synthesize_speech(
+        input=synthesis_input,
+        voice=voice,
+        audio_config=audio_cfg
+    )
+    return response.audio_content
+
+def audio_writer(script_text: str, out="story.mp3"):
+    chunks = chunk_text(script_text)
+    audio = AudioSegment.silent(200)
+    for i, chunk in enumerate(chunks, 1):
+        path = f"seg_{i}.mp3"
+        with open(path, "wb") as f:
+            f.write(tts_segment(chunk))
+        audio += AudioSegment.from_mp3(path)
+        audio += AudioSegment.silent(150)
+    audio.export(out, format="mp3")
+    return out
+
+# --- Full Pipeline ---
+async def full_pipeline(topic: str):
+    eli5 = await run_pedagogy_async(topic)
+    script = run_scriptwriter(eli5)
+    audio_path = audio_writer(script, "story.mp3")
+    return eli5, script, audio_path
+
+# --- Gradio App ---
+app = gr.Interface(
+    fn=full_pipeline,
+    inputs=gr.Textbox(label="Your Topic"),
+    outputs=[
+        gr.Textbox(label="ELI5 Explanation", lines=8),
+        gr.Textbox(label="Generated Story Script", lines=20),
+        gr.Audio(label="Generated Audio")
+    ],
+    title="🎧 TinyTutor — Full Pipeline",
+    css=".gradio-container { min-height: 1200px !important; }"
+)
+
+if __name__ == "__main__":
+    app.launch()