Oryx 1.5 7B

🚀 Oryx-1.5-7B

Oryx-1.5-7B是一个基于Qwen2.5语言模型的多模态大模型，具有32K token的上下文窗口，在图像和视频理解任务上展现出了强大的性能。它能够无缝且高效地处理任意空间大小和时间长度的视觉输入。

基础信息

项目链接

• 代码仓库：https://github.com/Oryx-mllm/Oryx
• 项目主页：https://oryx-mllm.github.io
• 论文链接：https://arxiv.org/abs/2409.12961

🚀 快速开始

我们提供了一个简单的模型使用示例，更多详细信息请参考我们的 Github 仓库。

from oryx.model.builder import load_pretrained_model
from oryx.mm_utils import get_model_name_from_path, process_images, tokenizer_image_token
from oryx.constants import IMAGE_TOKEN_INDEX, DEFAULT_IMAGE_TOKEN, DEFAULT_IM_START_TOKEN, DEFAULT_IM_END_TOKEN, IGNORE_INDEX
from oryx.conversation import conv_templates, SeparatorStyle
from PIL import Image
import requests
import copy
import torch
import sys
import warnings
from decord import VideoReader, cpu
import numpy as np

def load_video(self, video_path, max_frames_num,fps=1,force_sample=False):
    if max_frames_num == 0:
        return np.zeros((1, 336, 336, 3))
    vr = VideoReader(video_path, ctx=cpu(0),num_threads=1)
    total_frame_num = len(vr)
    video_time = total_frame_num / vr.get_avg_fps()
    fps = round(vr.get_avg_fps()/fps)
    frame_idx = [i for i in range(0, len(vr), fps)]
    frame_time = [i/fps for i in frame_idx]
    if len(frame_idx) > max_frames_num or force_sample:
        sample_fps = max_frames_num
        uniform_sampled_frames = np.linspace(0, total_frame_num - 1, sample_fps, dtype=int)
        frame_idx = uniform_sampled_frames.tolist()
        frame_time = [i/vr.get_avg_fps() for i in frame_idx]
    frame_time = ",".join([f"{i:.2f}s" for i in frame_time])
    spare_frames = vr.get_batch(frame_idx).asnumpy()
    # import pdb;pdb.set_trace()
    return spare_frames,frame_time,video_time
pretrained = "THUdyh/Oryx-7B"
model_name = "oryx_qwen"
device = "cuda"
device_map = "auto"
tokenizer, model, image_processor, max_length = load_pretrained_model(pretrained, None, model_name, device_map=device_map)
model.eval()
video_path = ""
max_frames_num = "64"
video,frame_time,video_time = load_video(video_path, max_frames_num, 1, force_sample=True)
video = image_processor.preprocess(video, return_tensors="pt")["pixel_values"].cuda().bfloat16()
video = [video]
video_data = (video, video)
input_data = (video_data, (384, 384), "video")
conv_template = "qwen_1_5"
question = DEFAULT_IMAGE_TOKEN + "\nPlease describe this video in detail."
conv = copy.deepcopy(conv_templates[conv_template])
conv.append_message(conv.roles[0], question)
conv.append_message(conv.roles[1], None)
prompt_question = conv.get_prompt()
input_ids = tokenizer_image_token(prompt_question, tokenizer, IMAGE_TOKEN_INDEX, return_tensors="pt").unsqueeze(0).to(device)
output_ids = model.generate(
    inputs=input_ids,
    images=input_data[0][0],
    images_highres=input_data[0][1],
    modalities=video_data[2],
    do_sample=False,
    temperature=0,
    max_new_tokens=128,
    use_cache=True,
)

text_outputs = tokenizer.batch_decode(cont, skip_special_tokens=True)
print(text_outputs)

✨ 主要特性

• 多模态处理：能够无缝且高效地处理任意空间大小和时间长度的视觉输入。
• 长上下文窗口：基于Qwen2.5语言模型，具有32K token的上下文窗口。
• 多语言支持：支持英文和中文两种语言。

📚 详细文档

模型表现

通用视频基准测试

长视频理解

常见图像基准测试

3D空间理解

模型架构

• 架构：预训练的 Oryx-ViT + Qwen2.5-7B
• 数据：120万张图像/视频数据的混合
• 精度：BFloat16

硬件与软件

• 硬件：64 * NVIDIA Tesla A100
• 编排工具：HuggingFace Trainer
• 代码框架：Pytorch

📄 许可证

本项目采用 apache-2.0 许可证。

📚 引用

如果您使用了本项目的代码或模型，请引用以下论文：

@article{liu2024oryx,
title={Oryx MLLM: On-Demand Spatial-Temporal Understanding at Arbitrary Resolution},
author={Liu, Zuyan and Dong, Yuhao and Liu, Ziwei and Hu, Winston and Lu, Jiwen and Rao, Yongming},
journal={arXiv preprint arXiv:2409.12961},
year={2024}
}