Update app.py

app.py

@@ -1,63 +1,127 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
-
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
-
-
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
-
-    messages.append({"role": "user", "content": message})
-
-    response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
-)
+import google.generativeai as genai
+import cv2
+import os
+import shutil
+
+# Set up your API key
+genai.configure(api_key="YOUR_API_KEY")
+
+# Constants
+FRAME_EXTRACTION_DIRECTORY = "/content/frames"
+FRAME_PREFIX = "_frame"
+
+# Function to create/cleanup frame output directory
+def create_frame_output_dir(output_dir):
+    if not os.path.exists(output_dir):
+        os.makedirs(output_dir)
+    else:
+        shutil.rmtree(output_dir)
+        os.makedirs(output_dir)
+
+# Function to extract frames from video
+def extract_frame_from_video(video_file_path):
+    create_frame_output_dir(FRAME_EXTRACTION_DIRECTORY)
+    vidcap = cv2.VideoCapture(video_file_path)
+    fps = vidcap.get(cv2.CAP_PROP_FPS)
+    frame_duration = 1 / fps
+    output_file_prefix = os.path.basename(video_file_path).replace('.', '_')
+    frame_count = 0
+    count = 0
+    while vidcap.isOpened():
+        success, frame = vidcap.read()
+        if not success:
+            break
+        if int(count / fps) == frame_count:
+            min = frame_count // 60
+            sec = frame_count % 60
+            time_string = f"{min:02d}:{sec:02d}"
+            image_name = f"{output_file_prefix}{FRAME_PREFIX}{time_string}.jpg"
+            output_filename = os.path.join(FRAME_EXTRACTION_DIRECTORY, image_name)
+            cv2.imwrite(output_filename, frame)
+            frame_count += 1
+        count += 1
+    vidcap.release()
+    return frame_count
+
+# Class to represent a file
+class File:
+    def __init__(self, file_path: str, display_name: str = None):
+        self.file_path = file_path
+        if display_name:
+            self.display_name = display_name
+        self.timestamp = self.get_timestamp(file_path)
+
+    def set_file_response(self, response):
+        self.response = response
+
+    def get_timestamp(self, filename):
+        parts = filename.split(FRAME_PREFIX)
+        if len(parts) != 2:
+            return None
+        return parts[1].split('.')[0]
 
+# Function to upload files to Gemini
+def upload_files(files_to_upload):
+    uploaded_files = []
+    for file in files_to_upload:
+        response = genai.upload_file(path=file.file_path)
+        file.set_file_response(response)
+        uploaded_files.append(file)
+    return uploaded_files
+
+# Function to generate description using Gemini
+def generate_description(uploaded_files):
+    prompt = "Describe this video."
+    model = genai.GenerativeModel(model_name="models/gemini-1.5-pro-latest")
+    request = [prompt]
+    for file in uploaded_files:
+        request.append(file.timestamp)
+        request.append(file.response)
+    response = model.generate_content(request, request_options={"timeout": 600})
+    return response.text
+
+# Function to delete files from Gemini
+def delete_files(uploaded_files):
+    for file in uploaded_files:
+        genai.delete_file(file.response.name)
+
+# Gradio interface
+def process_video(video_file):
+    try:
+        # Extract frames
+        frame_count = extract_frame_from_video(video_file.name)
+        
+        # Prepare files for upload
+        files = os.listdir(FRAME_EXTRACTION_DIRECTORY)
+        files = sorted(files)
+        files_to_upload = []
+        for file in files:
+            files_to_upload.append(
+                File(file_path=os.path.join(FRAME_EXTRACTION_DIRECTORY, file))
+            )
+
+        # Upload files to Gemini
+        uploaded_files = upload_files(files_to_upload)
+        
+        # Generate description
+        description = generate_description(uploaded_files)
+
+        # Delete files from Gemini
+        delete_files(uploaded_files)
+
+        return f"Video processed successfully! Description:\n\n{description}"
+    except Exception as e:
+        return f"An error occurred: {str(e)}"
+
+# Create Gradio interface
+iface = gr.Interface(
+    fn=process_video,
+    inputs=gr.Video(type="filepath"),
+    outputs=gr.Textbox(),
+    title="Video Description with Gemini",
+    description="Upload a video to get a description using Google Gemini",
+)
 
-if __name__ == "__main__":
-    demo.launch()
+# Launch the interface
+iface.launch()