OpenGVLab/InternVideo2_Chat_8B_InternLM2_5

InternVideo2-Chat-8B-InternLM2.5

[📂 GitHub] [📜 Tech Report]

To further enrich the semantics embedded in InternVideo2 and improve its user-friendly in human communications, we tune InternVideo2 by incorporating it into a VideoLLM with a LLM and a video BLIP. We employ the progressive learning scheme in VideoChat by using InternVideo2 as the video encoder and train a video blip for communicating with open-sourced LLM. In training, the video encoder will be updated. Detailed training recipts are in VideoChat. This model has HD training.

The BaseLLM of this model is InternLM2.5-7B with 1M long context window.

📈 Performance

Model	MVBench	VideoMME(w/o sub)
InternVideo2-Chat-8B	60.3	41.9
InternVideo2-Chat-8B-HD	65.4	46.1
InternVideo2-Chat-8B-HD-F16	67.5	49.4
InternVideo2-Chat-8B-InternLM	61.9	49.1

🚀 How to use the model

1. make sure to have transformers >= 4.38.0, peft==0.5.0

Install the requisite Python packages from pip_requirements

2. Inference with Video input

import os
import torch

from transformers import AutoTokenizer, AutoModel

tokenizer =  AutoTokenizer.from_pretrained('OpenGVLab/InternVideo2_Chat_8B_InternLM2_5',
    trust_remote_code=True,
    use_fast=False,)
if torch.cuda.is_available():
  model = AutoModel.from_pretrained(
      'OpenGVLab/InternVideo2_Chat_8B_InternLM2_5',
      torch_dtype=torch.bfloat16,
      trust_remote_code=True).cuda()
else:
  model = AutoModel.from_pretrained(
      'OpenGVLab/InternVideo2_Chat_8B_InternLM2_5',
      torch_dtype=torch.bfloat16,
      trust_remote_code=True)


from decord import VideoReader, cpu
from PIL import Image
import numpy as np
import numpy as np
import decord
from decord import VideoReader, cpu
import torch.nn.functional as F
import torchvision.transforms as T
from torchvision.transforms import PILToTensor
from torchvision import transforms
from torchvision.transforms.functional import InterpolationMode
decord.bridge.set_bridge("torch")

def get_index(num_frames, num_segments):
    seg_size = float(num_frames - 1) / num_segments
    start = int(seg_size / 2)
    offsets = np.array([
        start + int(np.round(seg_size * idx)) for idx in range(num_segments)
    ])
    return offsets


def load_video(video_path, num_segments=8, return_msg=False, resolution=224, hd_num=4, padding=False):
    vr = VideoReader(video_path, ctx=cpu(0), num_threads=1)
    num_frames = len(vr)
    frame_indices = get_index(num_frames, num_segments)

    mean = (0.485, 0.456, 0.406)
    std = (0.229, 0.224, 0.225)

    transform = transforms.Compose([
        transforms.Lambda(lambda x: x.float().div(255.0)),
        transforms.Normalize(mean, std)
    ])

    frames = vr.get_batch(frame_indices)
    frames = frames.permute(0, 3, 1, 2)

    if padding:
        frames = HD_transform_padding(frames.float(), image_size=resolution, hd_num=hd_num)
    else:
        frames = HD_transform_no_padding(frames.float(), image_size=resolution, hd_num=hd_num)

    frames = transform(frames)
    # print(frames.shape)
    T_, C, H, W = frames.shape

    sub_img = frames.reshape(
        1, T_, 3, H//resolution, resolution, W//resolution, resolution
    ).permute(0, 3, 5, 1, 2, 4, 6).reshape(-1, T_, 3, resolution, resolution).contiguous()

    glb_img = F.interpolate(
        frames.float(), size=(resolution, resolution), mode='bicubic', align_corners=False
    ).to(sub_img.dtype).unsqueeze(0)

    frames = torch.cat([sub_img, glb_img]).unsqueeze(0)

    if return_msg:
        fps = float(vr.get_avg_fps())
        sec = ", ".join([str(round(f / fps, 1)) for f in frame_indices])
        # " " should be added in the start and end
        msg = f"The video contains {len(frame_indices)} frames sampled at {sec} seconds."
        return frames, msg
    else:
        return frames

def HD_transform_padding(frames, image_size=224, hd_num=6):
    def _padding_224(frames):
        _, _, H, W = frames.shape
        tar = int(np.ceil(H / 224) * 224)
        top_padding = (tar - H) // 2
        bottom_padding = tar - H - top_padding
        left_padding = 0
        right_padding = 0

        padded_frames = F.pad(
            frames,
            pad=[left_padding, right_padding, top_padding, bottom_padding],
            mode='constant', value=255
        )
        return padded_frames

    _, _, H, W = frames.shape
    trans = False
    if W < H:
        frames = frames.flip(-2, -1)
        trans = True
        width, height = H, W
    else:
        width, height = W, H

    ratio = width / height
    scale = 1
    while scale * np.ceil(scale / ratio) <= hd_num:
        scale += 1
    scale -= 1
    new_w = int(scale * image_size)
    new_h = int(new_w / ratio)

    resized_frames = F.interpolate(
        frames, size=(new_h, new_w),
        mode='bicubic',
        align_corners=False
    )
    padded_frames = _padding_224(resized_frames)

    if trans:
        padded_frames = padded_frames.flip(-2, -1)

    return padded_frames

def find_closest_aspect_ratio(aspect_ratio, target_ratios, width, height, image_size):
        best_ratio_diff = float('inf')
        best_ratio = (1, 1)
        area = width * height
        for ratio in target_ratios:
            target_aspect_ratio = ratio[0] / ratio[1]
            ratio_diff = abs(aspect_ratio - target_aspect_ratio)
            if ratio_diff < best_ratio_diff:
                best_ratio_diff = ratio_diff
                best_ratio = ratio
            elif ratio_diff == best_ratio_diff:
                if area > 0.5 * image_size * image_size * ratio[0] * ratio[1]:
                    best_ratio = ratio
        return best_ratio


def HD_transform_no_padding(frames, image_size=224, hd_num=6, fix_ratio=(2,1)):
    min_num = 1
    max_num = hd_num
    _, _, orig_height, orig_width = frames.shape
    aspect_ratio = orig_width / orig_height

    # calculate the existing video aspect ratio
    target_ratios = set(
        (i, j) for n in range(min_num, max_num + 1) for i in range(1, n + 1) for j in range(1, n + 1) if
        i * j <= max_num and i * j >= min_num)
    target_ratios = sorted(target_ratios, key=lambda x: x[0] * x[1])

    # find the closest aspect ratio to the target
    if fix_ratio:
        target_aspect_ratio = fix_ratio
    else:
        target_aspect_ratio = find_closest_aspect_ratio(
            aspect_ratio, target_ratios, orig_width, orig_height, image_size)

    # calculate the target width and height
    target_width = image_size * target_aspect_ratio[0]
    target_height = image_size * target_aspect_ratio[1]
    blocks = target_aspect_ratio[0] * target_aspect_ratio[1]

    # resize the frames
    resized_frame = F.interpolate(
        frames, size=(target_height, target_width),
        mode='bicubic', align_corners=False
    )
    return resized_frame

video_path = "yoga.mp4"
# sample uniformly 8 frames from the video
video_tensor = load_video(video_path, num_segments=8, return_msg=False, resolution=224, hd_num=6)
video_tensor = video_tensor.to(model.device)

chat_history = []
response, chat_history = model.chat(tokenizer, '', 'Describe the video step by step',instruction= "Carefully watch the video and pay attention to the cause and sequence of events, the detail and movement of objects, and the action and pose of persons. Based on your observations, select the best option that accurately addresses the question.\n", media_type='video', media_tensor=video_tensor, chat_history= chat_history, return_history=True,generation_config={'do_sample':False,'max_new_tokens':512,})
print(response) 

✏️ Citation

If this work is helpful for your research, please consider citing InternVideo and VideoChat.

@article{wang2024internvideo2,
  title={Internvideo2: Scaling video foundation models for multimodal video understanding},
  author={Wang, Yi and Li, Kunchang and Li, Xinhao and Yu, Jiashuo and He, Yinan and Wang, Chenting and Chen, Guo and Pei, Baoqi and Zheng, Rongkun and Xu, Jilan and Wang, Zun and others},
  journal={arXiv preprint arXiv:2403.15377},
  year={2024}
}

@article{li2023videochat,
  title={Videochat: Chat-centric video understanding},
  author={Li, KunChang and He, Yinan and Wang, Yi and Li, Yizhuo and Wang, Wenhai and Luo, Ping and Wang, Yali and Wang, Limin and Qiao, Yu},
  journal={arXiv preprint arXiv:2305.06355},
  year={2023}
}