Create app.py

app.py

@@ -0,0 +1,82 @@
+import os
+import time
+import gradio as gr
+import google.generativeai as genai
+
+# Configure the Gemini API key
+genai.configure(api_key=os.environ["GEMINI_API_KEY"])
+
+# Upload files to Gemini
+def upload_to_gemini(file_path, mime_type):
+    """Uploads a file to Gemini and returns its reference."""
+    file = genai.upload_file(file_path, mime_type=mime_type)
+    print(f"Uploaded file '{file.display_name}' with URI: {file.uri}")
+    return {"fileData": {"fileUri": file.uri, "mimeType": mime_type}}
+
+# Wait for file processing
+def wait_for_files_active(files):
+    """Waits until all uploaded files are active."""
+    print("Waiting for file processing...")
+    for file in files:
+        status = genai.get_file(file["fileData"]["fileUri"])
+        while status.state.name == "PROCESSING":
+            print(".", end="", flush=True)
+            time.sleep(5)
+            status = genai.get_file(file["fileData"]["fileUri"])
+        if status.state.name != "ACTIVE":
+            raise Exception(f"File {file['fileData']['fileUri']} failed to process.")
+    print("\nAll files are ready.")
+
+# Process user inputs and interact with Gemini API
+def process_inputs(image_path, audio_path, pdf_file, video_path, user_prompt):
+    """Uploads files, processes content, and queries the Gemini API."""
+    contents = []
+
+    # Upload files if provided
+    if image_path:
+        contents.append(upload_to_gemini(image_path, mime_type="image/png"))
+    if audio_path:
+        contents.append(upload_to_gemini(audio_path, mime_type="audio/ogg"))
+    if pdf_file:
+        contents.append(upload_to_gemini(pdf_file.name, mime_type="application/pdf"))
+    if video_path:
+        contents.append(upload_to_gemini(video_path, mime_type="video/quicktime"))
+
+    # Ensure files are ready
+    if contents:
+        wait_for_files_active(contents)
+
+    # Add user text prompt
+    if user_prompt:
+        contents.append({"text": user_prompt})
+
+    # Send content to Gemini API
+    model = genai.GenerativeModel(model_name="gemini-2.0-flash-exp")
+    response = model.generate_content(contents=contents)
+
+    return response.text
+
+# Gradio Interface
+with gr.Blocks() as demo:
+    gr.Markdown("## Multimodal Chat Interface with Gemini API")
+
+    with gr.Row():
+        image_input = gr.Image(label="Upload an Image")
+        audio_input = gr.Audio(label="Upload an Audio File")
+    with gr.Row():
+        pdf_input = gr.File(label="Upload a PDF")
+        video_input = gr.Video(label="Upload a Video File")
+
+    user_prompt = gr.Textbox(label="Enter Your Prompt", placeholder="Ask something about the files...")
+
+    submit_button = gr.Button("Submit")
+    output_box = gr.Textbox(label="Gemini AI Response")
+
+    submit_button.click(
+        fn=process_inputs,
+        inputs=[image_input, audio_input, pdf_input, video_input, user_prompt],
+        outputs=output_box
+    )
+
+# Launch the Gradio app
+demo.launch()