feat: Update Gemini model

.gitattributes

@@ -0,0 +1,35 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,4 @@
+.env
+*.pyc
+__pycache__/
+visual_minutes_output.png

README.md

@@ -0,0 +1,28 @@
+---
+title: Visual Minutes Generator
+emoji: 🖼️
+colorFrom: blue
+colorTo: purple
+sdk: gradio
+sdk_version: 5.44.0
+app_file: app.py
+pinned: false
+license: mit
+short_description: A Space that generates visual minutes of meetings
+---
+
+# Visual Minutes Generator
+
+This is a Hugging Face Space that generates visual minutes from a meeting transcript using a template image and Google's Gemini 3 Pro model.
+
+## How to use
+
+1.  Enter your Google API Key in the input box (or set it as a Secret).
+2.  Upload a meeting transcript (text file).
+3.  Click "Generate Visual Minutes".
+
+*The template (template.jpg) is automatically loaded from the repository.*
+
+## Requirements
+
+*   Google API Key with access to Gemini 3 Pro / Gemini 2.0 Flash Exp.

app.py

@@ -0,0 +1,176 @@
+import gradio as gr
+import os
+from google import genai
+from google.genai import types
+from PIL import Image
+from pathlib import Path
+
+def generate_visual_minutes(api_key, transcript_file):
+    """
+    Generate visual minutes from a transcript using the fixed template.
+    
+    Args:
+        api_key: Google API key
+        transcript_file: Path to uploaded transcript file
+    
+    Returns:
+        PIL Image of generated visual minutes
+    """
+    # Validate API key
+    if not api_key:
+        raise gr.Error("Please enter your Google API Key.")
+    
+    if not transcript_file:
+        raise gr.Error("Please upload a transcript file.")
+    
+    # Load the fixed template from the repository
+    template_path = Path(__file__).parent / "template.jpg"
+    if not template_path.exists():
+        raise gr.Error("Template file (template.jpg) not found in the repository.")
+    
+    try:
+        template_image = Image.open(template_path)
+    except Exception as e:
+        raise gr.Error(f"Failed to load template image: {e}")
+    
+    # Read transcript
+    try:
+        with open(transcript_file, "r", encoding="utf-8") as f:
+            transcript_content = f.read()
+    except Exception as e:
+        raise gr.Error(f"Failed to read transcript file: {e}")
+    
+    # Construct the detailed prompt
+    prompt = f"""
+You are an expert visual facilitator creating graphic recording minutes for a meeting.
+
+**Task:**
+Generate a final image that looks exactly like the provided layout image (`template.jpg`), but with all the empty boxes and sections filled in with summarized information extracted from the provided `transcript_content`.
+
+**Input Transcript:**
+\"\"\"
+{transcript_content}
+\"\"\"
+
+**Layout Instructions based on the template image:**
+
+1.  **Top Banner (Title/Date):**
+    * Extract a concise, relevant Title for the meeting based on the content.
+    * Extract the Date mentioned in the transcript.
+
+2.  **Agenda Box (Top Left, blue/orange/red section):**
+    * Summarize the key discussion points into 3-4 concise bullet points.
+
+3.  **Attendees Box (Top Right):**
+    * List the names of the attendees correctly.
+    * Group them under their respective companies (e.g., 'Company 1: AVEVA' and 'Company 2: TOTAL') based on their introductions in the transcript.
+
+4.  **Objective Box (Middle Right, next to target icon):**
+    * Summarize the primary goal of this specific meeting in 1 short sentence.
+
+5.  **Next Step Box (Middle Left, next to clipboard icon):**
+    * Summarize the agreed-upon follow-up actions or future plans mentioned at the end of the meeting.
+
+6.  **Notes Columns (1.NOTE1 & 2.Note2 - The vertical avatar sections):**
+    * This is the most important part. Create a visual dialogue flow representing the key takeaways.
+    * Use the existing empty avatar slots in the template.
+    * Identify who is speaking key points in the transcript (e.g., Arthur, Marie, Antoine, Rachid).
+    * Next to their corresponding avatar slot, generate a speech bubble. check that the speech bubble is in front of the corresponding avatar.
+    * Inside the speech bubble, write a very concise summary of their main point.
+    * **Crucial:** Add relevant, expressive emojis inside the speech bubbles to visually represent their point (e.g., brains for AI, rockets for strategy, charts for data, warnings for caution).
+    * Ensure the conversation flows logically down the columns, matching the flow of the transcript.
+
+**Final Output Requirement:**
+The output must be a single image that looks like the completed infographic, maintaining the original aesthetic of the template.
+"""
+    
+    # Initialize client and call API
+    # The user's original script used "gemini-3-pro-image-preview" which supports image generation.
+    # "gemini-2.0-flash-exp" likely does not support direct image generation output.
+    MODEL_NAME = "gemini-3-pro-image-preview"
+    
+    try:
+        client = genai.Client(api_key=api_key)
+        
+        # The template is tall and narrow, so 9:16 aspect ratio is best suited.
+        aspect_ratio = "9:16"
+        resolution = "2K"
+        
+        # We pass the prompt AND the template image as contents.
+        # The model uses the image as the structural constraint.
+        response = client.models.generate_content(
+            model=MODEL_NAME,
+            contents=[
+                prompt,
+                template_image
+            ],
+            config=types.GenerateContentConfig(
+                # We only want an image back, not text explaining the image
+                response_modalities=['IMAGE'],
+                image_config=types.ImageConfig(
+                    aspect_ratio=aspect_ratio,
+                    image_size=resolution
+                ),
+            )
+        )
+        
+        # Extract and return the generated image
+        for part in response.parts:
+            if image := part.as_image():
+                return image
+        
+        raise gr.Error("The model completed but did not return an image. Check inputs or safety filters.")
+    
+    except Exception as e:
+        raise gr.Error(f"An error occurred during generation: {e}")
+
+
+# --- Gradio UI ---
+with gr.Blocks(title="Visual Minutes Generator") as demo:
+    gr.Markdown("# 🖼️ Visual Minutes Generator")
+    gr.Markdown("""
+    Generate visual minutes from a meeting transcript using Google's Gemini model.
+    
+    **The template (template.jpg) is automatically loaded from the repository.**
+    """)
+    
+    with gr.Row():
+        with gr.Column():
+            api_key_input = gr.Textbox(
+                label="Google API Key",
+                type="password",
+                placeholder="Enter your Google API Key",
+                info="Your API key is required to use the Gemini model"
+            )
+            
+            transcript_input = gr.File(
+                label="Upload Meeting Transcript (.txt)",
+                file_types=[".txt"],
+                type="filepath"
+            )
+            
+            generate_btn = gr.Button("Generate Visual Minutes", variant="primary", size="lg")
+        
+        with gr.Column():
+            output_image = gr.Image(
+                label="Generated Visual Minutes",
+                type="pil"
+            )
+    
+    gr.Markdown("""
+    ### How to use:
+    1. Enter your Google API Key (get one from [Google AI Studio](https://aistudio.google.com/app/apikey))
+    2. Upload your meeting transcript as a .txt file
+    3. Click "Generate Visual Minutes"
+    4. Wait 30-60 seconds for the AI to generate your visual minutes
+    """)
+    
+    # Connect the button to the function
+    generate_btn.click(
+        fn=generate_visual_minutes,
+        inputs=[api_key_input, transcript_input],
+        outputs=output_image
+    )
+
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -0,0 +1,5 @@
+google-genai
+pillow
+pathlib
+python-dotenv
+gradio