app.py

import gradio as gr
import torch
from transformers import AutoModel, AutoTokenizer
from PIL import Image
from decord import VideoReader, cpu
import base64
import io
import spaces

# Load model
model_path = 'openbmb/MiniCPM-V-2_6'
model = AutoModel.from_pretrained(model_path, trust_remote_code=True, torch_dtype=torch.bfloat16)
model = model.to(device='cuda')
tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True)
model.eval()

MAX_NUM_FRAMES = 64

def encode_image(image):
    if not isinstance(image, Image.Image):
        image = Image.open(image).convert("RGB")
    max_size = 448*16 
    if max(image.size) > max_size:
        w,h = image.size
        if w > h:
            new_w = max_size
            new_h = int(h * max_size / w)
        else:
            new_h = max_size
            new_w = int(w * max_size / h)
        image = image.resize((new_w, new_h), resample=Image.BICUBIC)
    return image

def encode_video(video_path):
    vr = VideoReader(video_path, ctx=cpu(0))
    sample_fps = round(vr.get_avg_fps() / 1)
    frame_idx = [i for i in range(0, len(vr), sample_fps)]
    if len(frame_idx) > MAX_NUM_FRAMES:
        frame_idx = frame_idx[:MAX_NUM_FRAMES]
    video = vr.get_batch(frame_idx).asnumpy()
    video = [Image.fromarray(v.astype('uint8')) for v in video]
    video = [encode_image(v) for v in video]
    return video

@spaces.GPU
def analyze_video(prompt, video):
    if isinstance(video, str):
        video_path = video
    else:
        video_path = video.name
    
    encoded_video = encode_video(video_path)
    
    context = [
        {"role": "user", "content": [prompt] + encoded_video}
    ]
    
    params = {
        'sampling': True,
        'top_p': 0.8,
        'top_k': 100,
        'temperature': 0.7,
        'repetition_penalty': 1.05,
        "max_new_tokens": 2048,
        "max_inp_length": 4352,
        "use_image_id": False,
        "max_slice_nums": 1 if len(encoded_video) > 16 else 2
    }
    
    response = model.chat(image=None, msgs=context, tokenizer=tokenizer, **params)
    
    return response

with gr.Blocks() as demo:
    gr.Markdown("# Video Analyzer")
    with gr.Row():
        with gr.Column():
            prompt_input = gr.Textbox(label="Prompt")
            video_input = gr.Video(label="Upload Video")
        with gr.Column():
            output = gr.Textbox(label="Analysis Result")
    
    analyze_button = gr.Button("Analyze Video")
    analyze_button.click(fn=analyze_video, inputs=[prompt_input, video_input], outputs=output)

demo.launch()