Harmon 1 5B

🚀 Harmon：统一多模态理解与生成的视觉表征协调框架

Harmon 是一个用于多模态理解和生成的全新统一框架。与现有的将视觉理解和生成用不同编码器模型分离处理的先进架构不同，该框架通过共享的 MAR 编码器协调理解和生成的视觉表征。Harmon 在主流的文本到图像生成基准测试中取得了先进的生成性能，并在多模态理解任务中展现出了有竞争力的结果。在本仓库中，我们提供了运行 Harmon 进行图像理解（图像到文本）和文本到图像生成的推理代码，有 Harmon - 0.5B 和 Harmon - 1.5B 两种模型变体。

🚀 快速开始

Harmon 是一个创新的多模态理解与生成统一框架。它借助共享的 MAR 编码器，协调视觉理解与生成的表征。本仓库提供了 Harmon - 0.5B 和 Harmon - 1.5B 两种模型变体的推理代码，可用于图像理解（图像到文本）和文本到图像生成任务。

✨ 主要特性

• 统一框架：通过共享的 MAR 编码器，将视觉理解和生成的表征进行协调，避免了使用不同编码器模型分离处理的方式。
• 性能优越：在主流文本到图像生成基准测试中取得先进的生成性能，在多模态理解任务中也有竞争力。
• 模型多样：提供 Harmon - 0.5B 和 Harmon - 1.5B 两种模型变体。

📦 安装指南

文档未提供安装步骤，此部分跳过。

💻 使用示例

基础用法

🖌️ 图像到文本生成

import torch
import numpy as np
from transformers import AutoTokenizer, AutoModel
from einops import rearrange
from PIL import Image
import requests


PROMPT_TEMPLATE = dict(
    SYSTEM='<|im_start|>system\n{system}<|im_end|>\n',
    INSTRUCTION='<|im_start|>user\n{input}<|im_end|>\n<|im_start|>assistant\n',
    SUFFIX='<|im_end|>',
    SUFFIX_AS_EOS=True,
    SEP='\n',
    STOP_WORDS=['<|im_end|>', '<|endoftext|>'])


def expand2square(pil_img, background_color):
    width, height = pil_img.size
    if width == height:
        return pil_img
    elif width > height:
        result = Image.new(pil_img.mode, (width, width), background_color)
        result.paste(pil_img, (0, (width - height) // 2))
        return result
    else:
        result = Image.new(pil_img.mode, (height, height), background_color)
        result.paste(pil_img, ((height - width) // 2, 0))
        return result


@torch.no_grad()
def question_answer(question,
                    image,
                    model,
                    tokenizer,
                    max_new_tokens=512,
                    image_size=512
                    ):
    assert image_size == 512
    image = expand2square(
        image, (127, 127, 127))
    image = image.resize(size=(image_size, image_size))
    image = torch.from_numpy(np.array(image)).to(dtype=model.dtype, device=model.device)
    image = rearrange(image, 'h w c -> c h w')[None]
    image = 2 * (image / 255) - 1

    prompt = PROMPT_TEMPLATE['INSTRUCTION'].format(input="<image>\n" + question)
    assert '<image>' in prompt
    image_length = (image_size // 16) ** 2 + model.mar.buffer_size
    prompt = prompt.replace('<image>', '<image>'*image_length)
    input_ids = tokenizer.encode(
        prompt, add_special_tokens=True, return_tensors='pt').cuda()
    _, z_enc = model.extract_visual_feature(model.encode(image))
    inputs_embeds = z_enc.new_zeros(*input_ids.shape, model.llm.config.hidden_size)
    inputs_embeds[input_ids == image_token_idx] = z_enc.flatten(0, 1)
    inputs_embeds[input_ids != image_token_idx] = model.llm.get_input_embeddings()(
        input_ids[input_ids != image_token_idx]
    )
    output = model.llm.generate(inputs_embeds=inputs_embeds,
                                use_cache=True,
                                do_sample=False,
                                max_new_tokens=max_new_tokens,
                                eos_token_id=tokenizer.eos_token_id,
                                pad_token_id=tokenizer.pad_token_id
                                if tokenizer.pad_token_id is not None else
                                tokenizer.eos_token_id
                                )
    return tokenizer.decode(output[0])


harmon_tokenizer = AutoTokenizer.from_pretrained("wusize/Harmon-1_5B",
                                                 trust_remote_code=True)
harmon_model = AutoModel.from_pretrained("wusize/Harmon-1_5B",
                                         trust_remote_code=True).eval().cuda().bfloat16()

special_tokens_dict = {'additional_special_tokens': ["<image>", ]}
num_added_toks = harmon_tokenizer.add_special_tokens(special_tokens_dict)
assert num_added_toks == 1

image_token_idx = harmon_tokenizer.encode("<image>", add_special_tokens=False)[-1]
print(f"Image token: {harmon_tokenizer.decode(image_token_idx)}")

image_file = "http://images.cocodataset.org/val2017/000000039769.jpg"
raw_image = Image.open(requests.get(image_file, stream=True).raw).convert('RGB')

output_text = question_answer(question='Describe the image in detail.',
                              image=raw_image,
                              model=harmon_model,
                              tokenizer=harmon_tokenizer,
                              )

print(output_text)

🖼️ 文本到图像生成

import os
import torch
from transformers import AutoTokenizer, AutoModel
from einops import rearrange
from PIL import Image


PROMPT_TEMPLATE = dict(
    SYSTEM='<|im_start|>system\n{system}<|im_end|>\n',
    INSTRUCTION='<|im_start|>user\n{input}<|im_end|>\n<|im_start|>assistant\n',
    SUFFIX='<|im_end|>',
    SUFFIX_AS_EOS=True,
    SEP='\n',
    STOP_WORDS=['<|im_end|>', '<|endoftext|>'])

GENERATION_TEMPLATE = "Generate an image: {text}"


@torch.no_grad()
def generate_images(prompts,
                    negative_prompt,
                    tokenizer,
                    model,
                    output,
                    grid_size=2,   # will produce 2 x 2 images per prompt
                    num_steps=64, cfg_scale=3.0, temperature=1.0, image_size=512):
    assert image_size == 512
    m = n = image_size // 16

    prompts = [
                  PROMPT_TEMPLATE['INSTRUCTION'].format(input=prompt)
                  for prompt in prompts
              ] * (grid_size ** 2)

    if cfg_scale != 1.0:
        prompts += [PROMPT_TEMPLATE['INSTRUCTION'].format(input=negative_prompt)] * len(prompts)

    inputs = tokenizer(
        prompts, add_special_tokens=True, return_tensors='pt', padding=True).to(model.device)

    images = model.sample(**inputs, num_iter=num_steps, cfg=cfg_scale, cfg_schedule="constant",
                          temperature=temperature, progress=True, image_shape=(m, n))
    images = rearrange(images, '(m n b) c h w -> b (m h) (n w) c', m=grid_size, n=grid_size)

    images = torch.clamp(
        127.5 * images + 128.0, 0, 255).to("cpu", dtype=torch.uint8).numpy()

    os.makedirs(output, exist_ok=True)
    for idx, image in enumerate(images):
        Image.fromarray(image).save(f"{output}/{idx:08d}.jpg")


harmon_tokenizer = AutoTokenizer.from_pretrained("wusize/Harmon-1_5B",
                                                 trust_remote_code=True)
harmon_model = AutoModel.from_pretrained("wusize/Harmon-1_5B",
                                         trust_remote_code=True).cuda().bfloat16().eval()


texts = ['a dog on the left and a cat on the right.',
         'a photo of a pink stop sign.']
pos_prompts = [GENERATION_TEMPLATE.format(text=text) for text in texts]
neg_prompt = 'Generate an image.'   # for classifier-free guidance


generate_images(prompts=pos_prompts,
                negative_prompt=neg_prompt,
                tokenizer=harmon_tokenizer,
                model=harmon_model,
                output='output',)

📚 详细文档

模型变体信息

属性	详情
模型类型	Harmon - 0.5B：LLM 为 Qwen2.5 - 0.5B - Instruct，MAR 为 MAR - Base；Harmon - 1.5B：LLM 为 Qwen2.5 - 1.5B - Instruct，MAR 为 MAR - Huge
下载链接	Harmon - 0.5B：；Harmon - 1.5B：

📚 引用

如果您发现 Harmon 对您的研究或应用有用，请使用以下 BibTeX 引用我们的论文：

@misc{wu2025harmon,
      title={Harmonizing Visual Representations for Unified Multimodal Understanding and Generation}, 
      author={Size Wu and Wenwei Zhang and Lumin Xu and Sheng Jin and Zhonghua Wu and Qingyi Tao and Wentao Liu and Wei Li and Chen Change Loy},
      year={2025},
      eprint={2503.21979},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2503.21979}, 
}

📄 许可证

本项目遵循 NTU S - Lab License 1.0 许可协议。