app.py

# from .demo_modelpart import InferenceDemo
import gradio as gr
import os
from threading import Thread

# import time
import cv2

import datetime
# import copy
import torch

import spaces
import numpy as np

from llava import conversation as conversation_lib
from llava.constants import DEFAULT_IMAGE_TOKEN


from llava.constants import (
    IMAGE_TOKEN_INDEX,
    DEFAULT_IMAGE_TOKEN,
    DEFAULT_IM_START_TOKEN,
    DEFAULT_IM_END_TOKEN,
)
from llava.conversation import conv_templates, SeparatorStyle
from llava.model.builder import load_pretrained_model
from llava.utils import disable_torch_init
from llava.mm_utils import (
    tokenizer_image_token,
    get_model_name_from_path,
    KeywordsStoppingCriteria,
)

from serve_constants import html_header, bibtext, learn_more_markdown, tos_markdown

from decord import VideoReader, cpu

import requests
from PIL import Image
import io
from io import BytesIO
from transformers import TextStreamer, TextIteratorStreamer

import hashlib
import PIL
import base64
import json

import datetime
import gradio as gr
import gradio_client
import subprocess
import sys

from huggingface_hub import HfApi
from huggingface_hub import login
from huggingface_hub import revision_exists

login(token=os.environ["HF_TOKEN"],
      write_permission=True)

api = HfApi()
repo_name = os.environ["LOG_REPO"]

external_log_dir = "./logs"
LOGDIR = external_log_dir
VOTEDIR = "./votes"


def install_gradio_4_35_0():
    current_version = gr.__version__
    if current_version != "4.35.0":
        print(f"Current Gradio version: {current_version}")
        print("Installing Gradio 4.35.0...")
        subprocess.check_call([sys.executable, "-m", "pip", "install", "gradio==4.35.0", "--force-reinstall"])
        print("Gradio 4.35.0 installed successfully.")
    else:
        print("Gradio 4.35.0 is already installed.")

# Call the function to install Gradio 4.35.0 if needed
install_gradio_4_35_0()

import gradio as gr
import gradio_client
print(f"Gradio version: {gr.__version__}")
print(f"Gradio-client version: {gradio_client.__version__}")

def get_conv_log_filename():
    t = datetime.datetime.now()
    name = os.path.join(LOGDIR, f"{t.year}-{t.month:02d}-{t.day:02d}-user_conv.json")
    return name

def get_conv_vote_filename():
    t = datetime.datetime.now()
    name = os.path.join(VOTEDIR, f"{t.year}-{t.month:02d}-{t.day:02d}-user_vote.json")
    if not os.path.isfile(name):
        os.makedirs(os.path.dirname(name), exist_ok=True)
    return name

def vote_last_response(state, vote_type, model_selector):
    with open(get_conv_vote_filename(), "a") as fout:
        data = {
            "type": vote_type,
            "model": model_selector,
            "state": state,
        }
        fout.write(json.dumps(data) + "\n")
    api.upload_file(
        path_or_fileobj=get_conv_vote_filename(),
        path_in_repo=get_conv_vote_filename().replace("./votes/", ""),
        repo_id=repo_name,
        repo_type="dataset")


def upvote_last_response(state):
    vote_last_response(state, "upvote", "MAmmoTH-VL-8b")
    gr.Info("Thank you for your voting!")
    return state

def downvote_last_response(state):
    vote_last_response(state, "downvote", "MAmmoTH-VL-8b")
    gr.Info("Thank you for your voting!")
    return state

class InferenceDemo(object):
    def __init__(
        self, args, model_path, tokenizer, model, image_processor, context_len
    ) -> None:
        disable_torch_init()

        self.tokenizer, self.model, self.image_processor, self.context_len = (
            tokenizer,
            model,
            image_processor,
            context_len,
        )

        if "llama-2" in model_name.lower():
            conv_mode = "llava_llama_2"
        elif "v1" in model_name.lower():
            conv_mode = "llava_v1"
        elif "mpt" in model_name.lower():
            conv_mode = "mpt"
        elif "qwen" in model_name.lower():
            conv_mode = "qwen_1_5"
        elif "pangea" in model_name.lower():
            conv_mode = "qwen_1_5"
        elif "mammoth-vl" in model_name.lower():
            conv_mode = "qwen_2_5"
        else:
            conv_mode = "llava_v0"

        if args.conv_mode is not None and conv_mode != args.conv_mode:
            print(
                "[WARNING] the auto inferred conversation mode is {}, while `--conv-mode` is {}, using {}".format(
                    conv_mode, args.conv_mode, args.conv_mode
                )
            )
        else:
            args.conv_mode = conv_mode
        self.conv_mode = conv_mode
        self.conversation = conv_templates[args.conv_mode].copy()
        self.num_frames = args.num_frames

class ChatSessionManager:
    def __init__(self):
        self.chatbot_instance = None

    def initialize_chatbot(self, args, model_path, tokenizer, model, image_processor, context_len):
        self.chatbot_instance = InferenceDemo(args, model_path, tokenizer, model, image_processor, context_len)
        print(f"Initialized Chatbot instance with ID: {id(self.chatbot_instance)}")

    def reset_chatbot(self):
        self.chatbot_instance = None

    def get_chatbot(self, args, model_path, tokenizer, model, image_processor, context_len):
        if self.chatbot_instance is None:
            self.initialize_chatbot(args, model_path, tokenizer, model, image_processor, context_len)
        return self.chatbot_instance
        

def is_valid_video_filename(name):
    video_extensions = ["avi", "mp4", "mov", "mkv", "flv", "wmv", "mjpeg"]

    ext = name.split(".")[-1].lower()

    if ext in video_extensions:
        return True
    else:
        return False

def is_valid_image_filename(name):
    image_extensions = ["jpg", "jpeg", "png", "bmp", "gif", "tiff", "webp", "heic", "heif", "jfif", "svg", "eps", "raw"] 

    ext = name.split(".")[-1].lower()

    if ext in image_extensions:
        return True
    else:
        return False


def sample_frames_v1(video_file, num_frames):
    video = cv2.VideoCapture(video_file)
    total_frames = int(video.get(cv2.CAP_PROP_FRAME_COUNT))
    interval = total_frames // num_frames
    frames = []
    for i in range(total_frames):
        ret, frame = video.read()
        pil_img = Image.fromarray(cv2.cvtColor(frame, cv2.COLOR_BGR2RGB))
        if not ret:
            continue
        if i % interval == 0:
            frames.append(pil_img)
    video.release()
    return frames

def sample_frames_v2(video_path, frame_count=32):
    video_frames = []
    vr = VideoReader(video_path, ctx=cpu(0))
    total_frames = len(vr)
    frame_interval = max(total_frames // frame_count, 1)

    for i in range(0, total_frames, frame_interval):
        frame = vr[i].asnumpy()
        frame_image = Image.fromarray(frame)  # Convert to PIL.Image
        video_frames.append(frame_image)
        if len(video_frames) >= frame_count:
            break

    # Ensure at least one frame is returned if total frames are less than required
    if len(video_frames) < frame_count and total_frames > 0:
        for i in range(total_frames):
            frame = vr[i].asnumpy()
            frame_image = Image.fromarray(frame)  # Convert to PIL.Image
            video_frames.append(frame_image)
            if len(video_frames) >= frame_count:
                break

    return video_frames

def sample_frames(video_path, num_frames=8):
    cap = cv2.VideoCapture(video_path)
    frames = []
    total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
    indices = np.linspace(0, total_frames - 1, num_frames, dtype=int)

    for i in indices:
        cap.set(cv2.CAP_PROP_POS_FRAMES, i)
        ret, frame = cap.read()
        if ret:
            frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
            frames.append(Image.fromarray(frame))

    cap.release()
    return frames


def load_image(image_file):
    if image_file.startswith("http") or image_file.startswith("https"):
        response = requests.get(image_file)
        if response.status_code == 200:
            image = Image.open(BytesIO(response.content)).convert("RGB")
        else:
            print("failed to load the image")
    else:
        print("Load image from local file")
        print(image_file)
        image = Image.open(image_file).convert("RGB")

    return image


def clear_response(history):
    for index_conv in range(1, len(history)):
        # loop until get a text response from our model.
        conv = history[-index_conv]
        if not (conv[0] is None):
            break
    question = history[-index_conv][0]
    history = history[:-index_conv]
    return history, question

chat_manager = ChatSessionManager()


def clear_history(history):
    chatbot_instance = chat_manager.get_chatbot(args, model_path, tokenizer, model, image_processor, context_len)
    chatbot_instance.conversation = conv_templates[chatbot_instance.conv_mode].copy()
    return None



def add_message(history, message):
    global chat_image_num
    print("#### len(history)",len(history))
    if not history:
        history = []
        our_chatbot = chat_manager.get_chatbot(args, model_path, tokenizer, model, image_processor, context_len)
        chat_image_num = 0
    
    # if len(message["files"]) <= 1:
    #     for x in message["files"]:
    #         history.append(((x,), None))
    #         chat_image_num += 1
    #     if chat_image_num > 1:
    #         history = []
    #         chat_manager.reset_chatbot()
    #         our_chatbot = chat_manager.get_chatbot(args, model_path, tokenizer, model, image_processor, context_len)
    #         chat_image_num = 0
    #         for x in message["files"]:
    #             history.append(((x,), None))
    #             chat_image_num += 1
                
    #     if message["text"] is not None:
    #         history.append((message["text"], None))
        
    #     print(f"### Chatbot instance ID: {id(our_chatbot)}")
    #     return history, gr.MultimodalTextbox(value=None, interactive=False)
    # else:
    for x in message["files"]:
        if "realcase_video.jpg" in x:
            x = x.replace("realcase_video.jpg", "realcase_video.mp4")
        history.append(((x,), None))
    if message["text"] is not None:
        history.append((message["text"], None))
    # print(f"### Chatbot instance ID: {id(our_chatbot)}")
    return history, gr.MultimodalTextbox(value=None, interactive=False)


@spaces.GPU
def bot(history, temperature, top_p, max_output_tokens):
    our_chatbot = chat_manager.get_chatbot(args, model_path, tokenizer, model, image_processor, context_len)
    print(f"### Chatbot instance ID: {id(our_chatbot)}")
    text = history[-1][0]
    images_this_term = []
    text_this_term = ""
    
    is_video = False
    num_new_images = 0
    # previous_image = False
    for i, message in enumerate(history[:-1]):
        if type(message[0]) is tuple:
            # if previous_image:
            #     gr.Warning("Only one image can be uploaded in a conversation. Please reduce the number of images and start a new conversation.")
            #     our_chatbot.conversation = conv_templates[our_chatbot.conv_mode].copy()
            #     return None
           
            images_this_term.append(message[0][0])
            if is_valid_video_filename(message[0][0]):
                # raise ValueError("Video is not supported")
                # num_new_images += our_chatbot.num_frames
                # num_new_images += len(sample_frames(message[0][0], our_chatbot.num_frames))
                num_new_images += 1
                is_video = True
            elif is_valid_image_filename(message[0][0]):
                print("#### Load image from local file",message[0][0])
                num_new_images += 1
            else:
                raise ValueError("Invalid file format")
            # previous_image = True
        else:
            num_new_images = 0
            # previous_image = False

    
    image_list = []
    for f in images_this_term:
        if is_valid_video_filename(f):
            image_list += sample_frames(f, our_chatbot.num_frames)
        elif is_valid_image_filename(f):
            image_list.append(load_image(f))
        else:
            raise ValueError("Invalid image file")

    all_image_hash = []
    all_image_path = []
    for file_path in images_this_term:
        with open(file_path, "rb") as file:
            file_data = file.read()
            file_hash = hashlib.md5(file_data).hexdigest()
            all_image_hash.append(file_hash)
            
            t = datetime.datetime.now()
            output_dir = os.path.join(
                LOGDIR,
                "serve_files",
                f"{t.year}-{t.month:02d}-{t.day:02d}"
            )
            os.makedirs(output_dir, exist_ok=True)
            
            if is_valid_image_filename(file_path):
                # Process and save images
                image = Image.open(file_path).convert("RGB")
                filename = os.path.join(output_dir, f"{file_hash}.jpg")
                all_image_path.append(filename)
                if not os.path.isfile(filename):
                    print("Image saved to", filename)
                    image.save(filename)
    
            elif is_valid_video_filename(file_path):
                # Simplified video saving
                filename = os.path.join(output_dir, f"{file_hash}.mp4")
                all_image_path.append(filename)
                if not os.path.isfile(filename):
                    print("Video saved to", filename)
                    os.makedirs(os.path.dirname(filename), exist_ok=True)
                    # Directly copy the video file
                    with open(file_path, "rb") as src, open(filename, "wb") as dst:
                        dst.write(src.read())
    
    image_tensor = []
    if is_video:
        image_tensor = our_chatbot.image_processor.preprocess(image_list, return_tensors="pt")["pixel_values"].half().to(our_chatbot.model.device)
    elif num_new_images > 0:
        image_tensor = [
            our_chatbot.image_processor.preprocess(f, return_tensors="pt")["pixel_values"][
                0
            ]
            .half()
            .to(our_chatbot.model.device)
            for f in image_list
        ]
        image_tensor = torch.stack(image_tensor)
    
    image_token = DEFAULT_IMAGE_TOKEN * num_new_images + "\n"
    
    inp = text
    inp = image_token + inp
    our_chatbot.conversation.append_message(our_chatbot.conversation.roles[0], inp)
    # image = None
    our_chatbot.conversation.append_message(our_chatbot.conversation.roles[1], None)
    prompt = our_chatbot.conversation.get_prompt()

    input_ids = tokenizer_image_token(
            prompt, our_chatbot.tokenizer, IMAGE_TOKEN_INDEX, return_tensors="pt"
        ).unsqueeze(0).to(our_chatbot.model.device)
    # print("### input_id",input_ids)
    stop_str = (
        our_chatbot.conversation.sep
        if our_chatbot.conversation.sep_style != SeparatorStyle.TWO
        else our_chatbot.conversation.sep2
    )
    keywords = [stop_str]
    stopping_criteria = KeywordsStoppingCriteria(
        keywords, our_chatbot.tokenizer, input_ids
    )
    
    streamer = TextIteratorStreamer(
        our_chatbot.tokenizer, skip_prompt=True, skip_special_tokens=True
    )
    print(our_chatbot.model.device)
    print(input_ids.device)
    # print(image_tensor.device)
    
    
    if is_video:
        input_image_tensor = [image_tensor]
    elif num_new_images > 0:
        input_image_tensor = image_tensor
    else:
        input_image_tensor = None
    
    generate_kwargs = dict(
        inputs=input_ids,
        streamer=streamer,
        images=input_image_tensor,
        do_sample=True,
        temperature=temperature,
        top_p=top_p,
        max_new_tokens=max_output_tokens,
        use_cache=False,
        stopping_criteria=[stopping_criteria],
        modalities=["video"] if is_video else ["image"]
    )
    
    t = Thread(target=our_chatbot.model.generate, kwargs=generate_kwargs)
    t.start()
    
    outputs = []
    for stream_token in streamer:
        outputs.append(stream_token)
        
        history[-1] = [text, "".join(outputs)]
        yield history
    our_chatbot.conversation.messages[-1][-1] = "".join(outputs)
    # print("### turn end history", history)
    # print("### turn end conv",our_chatbot.conversation)
        
    with open(get_conv_log_filename(), "a") as fout:
        data = {
            "type": "chat",
            "model": "MAmmoTH-VL-8b",
            "state": history,
            "images": all_image_hash,
            "images_path": all_image_path
        }
        print("#### conv log",data)
        fout.write(json.dumps(data) + "\n")
    for upload_img in all_image_path:
        api.upload_file(
            path_or_fileobj=upload_img,
            path_in_repo=upload_img.replace("./logs/", ""),
            repo_id=repo_name,
            repo_type="dataset",
            # revision=revision,
            # ignore_patterns=["data*"]
        )
    # upload json
    api.upload_file(
        path_or_fileobj=get_conv_log_filename(),
        path_in_repo=get_conv_log_filename().replace("./logs/", ""),
        repo_id=repo_name,
        repo_type="dataset")
        
    

txt = gr.Textbox(
    scale=4,
    show_label=False,
    placeholder="Enter text and press enter.",
    container=False,
)

with gr.Blocks(
    css=".message-wrap.svelte-1lcyrx4>div.svelte-1lcyrx4  img {min-width: 40px}",
) as demo:

    cur_dir = os.path.dirname(os.path.abspath(__file__))
    # gr.Markdown(title_markdown)
    gr.HTML(html_header)
    
    with gr.Column():
        with gr.Accordion("Parameters", open=False) as parameter_row:
                temperature = gr.Slider(
                    minimum=0.0,
                    maximum=1.0,
                    value=0.7,
                    step=0.1,
                    interactive=True,
                    label="Temperature",
                )
                top_p = gr.Slider(
                    minimum=0.0,
                    maximum=1.0,
                    value=1,
                    step=0.1,
                    interactive=True,
                    label="Top P",
                )
                max_output_tokens = gr.Slider(
                    minimum=0,
                    maximum=8192,
                    value=4096,
                    step=256,
                    interactive=True,
                    label="Max output tokens",
                )
        with gr.Row():
            chatbot = gr.Chatbot([], elem_id="MAmmoTH-VL-8B", bubble_full_width=False, height=750)

        with gr.Row():
            upvote_btn = gr.Button(value="👍  Upvote", interactive=True)
            downvote_btn = gr.Button(value="👎  Downvote", interactive=True)
            flag_btn = gr.Button(value="⚠️  Flag", interactive=True)
            # stop_btn = gr.Button(value="⏹️  Stop Generation", interactive=True)
            regenerate_btn = gr.Button(value="🔄  Regenerate", interactive=True)
            clear_btn = gr.Button(value="🗑️  Clear history", interactive=True)
            

        chat_input = gr.MultimodalTextbox(
            interactive=True,
            file_types=["image", "video"],
            placeholder="Enter message or upload file...",
            show_label=False,
            submit_btn="🚀"
        )

        print(cur_dir)
        gr.Examples(
                examples_per_page=20,
                examples=[
                    [
                        {
                            "files": [
                                f"{cur_dir}/examples/172197131626056_P7966202.png",
                            ],
                            "text": "Why this image funny?",
                        }
                    ],
                    [    
                        {
                            "files": [
                                f"{cur_dir}/examples/realcase_doc.png",
                            ],
                            "text": "Read text in the image",
                        }
                    ],
                    [
                        {
                            "files": [
                                f"{cur_dir}/examples/realcase_weather.jpg",
                            ],
                            "text": "List the weather for Monday to Friday",
                        }
                    ],
                    [
                        {
                            "files": [
                                f"{cur_dir}/examples/realcase_knowledge.jpg",
                            ],
                            "text": "Answer the following question based on the provided image: What country do these planes belong to?",
                        }
                    ],
                    [
                        {
                            "files": [
                                f"{cur_dir}/examples/realcase_math.jpg",
                            ],
                            "text": "Find the measure of angle 3. Please provide a step by step solution.",
                        }
                    ],
                    [
                        {
                            "files": [
                                f"{cur_dir}/examples/realcase_interact.jpg",
                            ],
                            "text": "Please perfectly describe this cartoon illustration in as much detail as possible",
                        }
                    ],
                    [
                        {
                            "files": [
                                f"{cur_dir}/examples/realcase_perfer.jpg",
                            ],
                            "text": "This is an image of a room. It could either be a real image captured in the room or a rendered image from a 3D scene reconstruction technique that is trained using real images of the room. A rendered image usually contains some visible artifacts (eg. blurred regions due to under-reconstructed areas) that do not faithfully represent the actual scene. You need to decide if its a real image or a rendered image by giving each image a photorealism score between 1 and 5.",
                        }
                    ],
                    [
                        {
                            "files": [
                                f"{cur_dir}/examples/realcase_multi1.png",
                                f"{cur_dir}/examples/realcase_multi2.png",
                                f"{cur_dir}/examples/realcase_multi3.png",
                                f"{cur_dir}/examples/realcase_multi4.png",
                                f"{cur_dir}/examples/realcase_multi5.png",
                            ],
                            "text": "Based on the five species in the images, draw a food chain. Explain the role of each species in the food chain.",
                        }
                    ],
                ],
                inputs=[chat_input],
                label="Real World Image Cases",
            )
        gr.Examples(
            examples=[
                [
                    {
                        "files": [
                            f"{cur_dir}/examples/realcase_video.mp4",
                        ],
                        "text": "Please describe the video in detail.",
                    },
                ]
            ],
            inputs=[chat_input],
            label="Real World Video Case"
        )

        gr.Markdown(tos_markdown)
        gr.Markdown(learn_more_markdown)
        gr.Markdown(bibtext)

    chat_input.submit(
    add_message, [chatbot, chat_input], [chatbot, chat_input]
    ).then(bot, [chatbot, temperature, top_p, max_output_tokens], chatbot, api_name="bot_response").then(lambda: gr.MultimodalTextbox(interactive=True), None, [chat_input])
  

    # chatbot.like(print_like_dislike, None, None)
    clear_btn.click(
        fn=clear_history, inputs=[chatbot], outputs=[chatbot], api_name="clear_all"
    )

    upvote_btn.click(
        fn=upvote_last_response, inputs=chatbot, outputs=chatbot, api_name="upvote_last_response"
    )

    
    downvote_btn.click(
        fn=downvote_last_response, inputs=chatbot, outputs=chatbot, api_name="upvote_last_response"
    )


demo.queue()
    
if __name__ == "__main__":
    import argparse

    argparser = argparse.ArgumentParser()
    argparser.add_argument("--server_name", default="0.0.0.0", type=str)
    argparser.add_argument("--port", default="6123", type=str)
    argparser.add_argument(
        "--model_path", default="MMSFT/MAmmoTH-VL-8B", type=str
    )
    # argparser.add_argument("--model-path", type=str, default="facebook/opt-350m")
    argparser.add_argument("--model-base", type=str, default=None)
    argparser.add_argument("--num-gpus", type=int, default=1)
    argparser.add_argument("--conv-mode", type=str, default=None)
    argparser.add_argument("--temperature", type=float, default=0.7)
    argparser.add_argument("--max-new-tokens", type=int, default=4096)
    argparser.add_argument("--num_frames", type=int, default=32)
    argparser.add_argument("--load-8bit", action="store_true")
    argparser.add_argument("--load-4bit", action="store_true")
    argparser.add_argument("--debug", action="store_true")

    args = argparser.parse_args()

    model_path = args.model_path
    filt_invalid = "cut"
    model_name = get_model_name_from_path(args.model_path)
    tokenizer, model, image_processor, context_len = load_pretrained_model(args.model_path, args.model_base, model_name, args.load_8bit, args.load_4bit)
    model=model.to(torch.device('cuda'))
    chat_image_num = 0
    demo.launch()