Transformers documentation

GlmImage

Transformers

You are viewing main version, which requires installation from source. If you'd like regular pip install, checkout the latest stable version (v5.22.1).

Join the Hugging Face community

and get access to the augmented documentation experience

Collaborate on models, datasets and Spaces

Faster examples with accelerated inference

Switch between documentation themes

to get started

该模型于 2026-01-13 贡献至 Hugging Face Transformers。

GlmImage

概述

GLM-Image 是一款采用混合自回归 + 扩散解码器架构的图像生成模型，有效提升了视觉保真度和细粒度细节的上限。在通用图像生成质量方面，它与业界标准的基于 LDM 的方法相当，同时在知识密集型图像生成场景中展现出显著优势。

模型架构：混合自回归 + 扩散解码器设计、

自回归生成器：一个从 GLM-4-9B-0414 初始化的 9B 参数模型，其词表经过扩展以纳入视觉标记。该模型首先生成约 256 个标记的紧凑编码，随后扩展至 1K–4K 个标记，对应 1K–2K 的高分辨率图像输出。
Diffusion Decoder：一个基于单流 DiT 架构的 7B 参数解码器，用于潜在空间图像解码。它配备了 Glyph Encoder 文本模块，显著提升了图像中文字的准确渲染效果。

通过解耦强化学习进行后训练：该模型引入了基于 GRPO 算法的细粒度、模块化反馈策略，显著提升了语义理解和视觉细节质量。

自回归模块：提供专注于美学和语义对齐的低频反馈信号，从而提升指令遵循能力和艺术表现力。
解码器模块：提供针对细节保真度和文本准确性的高频反馈，从而生成高度逼真的纹理、光照和色彩还原，以及更精确的文本渲染。

GLM-Image 支持在单个模型中同时进行文本生成图像和图像生成图像

文本生成图像：根据文本描述生成高细节图像，在信息密集的场景中表现尤为出色。
图像到图像：支持多种任务，包括图像编辑、风格迁移、多主体一致性，以及针对人物和物体的身份保持生成。
GlmImageForConditionalGeneration 是 GLM-Image 模型的自回归部分，完整的图像生成流程请参见此处。

该模型由 Raushan Turganbay 和 Yuxuan Zhang 贡献。

使用示例

使用 GLM-Image 配合图像输入，为 DIT 生成视觉 token。


import torch

from transformers import AutoProcessor, GlmImageForConditionalGeneration


model = GlmImageForConditionalGeneration.from_pretrained(
    pretrained_model_name_or_path="zai-org/GLM-Image/vision_language_encoder",
    device_map=torch.accelerator.current_accelerator()
)
processor = AutoProcessor.from_pretrained(
    pretrained_model_name_or_path="zai-org/GLM-Image/processor",
    use_fast=True
)

# Case1 T2I
prompt = "现代美食杂志风格的甜点制作教程图，主题为覆盆子慕斯蛋糕。整体布局干净明亮，分为四个主要区域：顶部左侧是黑色粗体标题“覆盆子慕斯蛋糕制作指南”，右侧搭配光线柔和的成品蛋糕特写照片，蛋糕呈淡粉色，表面点缀新鲜覆盆子与薄荷叶；左下方为配料清单区域，标题“配料”使用简洁字体，下方列有“面粉 150g”“鸡蛋 3个”“细砂糖 120g”“覆盆子果泥 200g”“明胶片 10g”“淡奶油 300ml”“新鲜覆盆子”等配料，每种配料旁配有简约线图标（如面粉袋、鸡蛋、糖罐等）；右下方是四个等大的步骤方框，每个方框内含高清微距实拍图及对应操作说明，从上到下依次为：步骤1展示打蛋器打发白色泡沫（对应说明“打发蛋白至干性发泡”），步骤2展示红白相间的混合物被刮刀翻拌（对应说明“轻柔翻拌果泥与面糊”），步骤3展示粉色液体被倒入圆形模具（对应说明“倒入模具并冷藏4小时”），步骤4展示成品蛋糕表面装饰覆盆子与薄荷叶（对应说明“用覆盆子和薄荷装饰”）；底部边缘设浅棕色信息条，左侧图标分别代表“准备时间：30分钟”“烹饪时间：20分钟”“份量：8人份”。整体色调以奶油白、淡粉色为主，背景带轻微纸质纹理，图文排版紧凑有序，信息层级分明。"
target_h, target_w = 1152, 768
use_reference_images = False
reference_image_paths = None

# ## Case2
# prompt = "Replace the background of the snow forest with an underground station featuring an automatic escalator."
# cond_0 = "cond.jpg"
# target_h, target_w = 1152, 768
# use_reference_images = True
# reference_image_paths = [cond_0]

## Case3
# prompt = "Make the man in the first figure and the child from the second image bow at the same time in a respectful KTV."
# cond_0 = "cond_0.jpg"
# cond_1 = "cond_1.jpg"
# target_h, target_w = 1152, 768
# use_reference_images = True
# reference_image_paths = [cond_0, cond_1]


def build_messages(prompt, use_reference_images, reference_image_paths):
    content = []
    if use_reference_images:
        for img_path in reference_image_paths:
            content.append({"type": "image", "url": img_path})
    content.append({"type": "text", "text": prompt})
    return [{"role": "user", "content": content}]


def compute_generation_params(image_grid_thw, use_reference_images):
    grid_sizes = []
    for i in range(image_grid_thw.shape[0]):
        t, h, w = image_grid_thw[i].tolist()
        grid_sizes.append(int(h * w))

    target_output_length = grid_sizes[0]

    if use_reference_images:
        max_new_tokens = grid_sizes[-1] + 1
        output_start_offset = 0
        output_length = grid_sizes[-1]
    else:
        total_tokens = sum(grid_sizes)
        max_new_tokens = total_tokens + 1
        output_start_offset = sum(grid_sizes[1:])
        output_length = target_output_length

    return max_new_tokens, output_start_offset, output_length


messages = build_messages(prompt, use_reference_images, reference_image_paths if use_reference_images else None)

inputs = processor.apply_chat_template(
    messages,
    target_h=target_h,
    target_w=target_w,
    tokenize=True,
    return_dict=True,
    return_tensors="pt",
).to(model.device)

image_grid_thw = inputs.get('image_grid_thw')
print(f"image_grid_thw: {image_grid_thw}")

max_new_tokens, output_start_offset, output_length = compute_generation_params(
    image_grid_thw, use_reference_images
)

print(f"use_reference_images: {use_reference_images}")
print(f"max_new_tokens: {max_new_tokens}")
print(f"output_start_offset: {output_start_offset}")
print(f"output_length: {output_length}")

outputs = model.generate(
    **inputs,
    max_new_tokens=max_new_tokens,
    do_sample=True
)

input_length = inputs["input_ids"].shape[-1]
output_tokens = outputs[0][input_length:][output_start_offset:output_start_offset + output_length]
print(f"Input length: {input_length}")
print(f"Total generated tokens: {outputs[0].shape[-1] - input_length}")
print(f"Extracted output tokens shape: {output_tokens.shape}")
print(f"Output tokens: {output_tokens}")

Transformers

GlmImage

概述

使用示例

GlmImageConfig

class transformers.GlmImageConfig

GlmImageVisionConfig

class transformers.GlmImageVisionConfig

GlmImageTextConfig

class transformers.GlmImageTextConfig

GlmImageVQVAEConfig

class transformers.GlmImageVQVAEConfig

GlmImageImageProcessor

class transformers.GlmImageImageProcessor

预处理

GlmImageImageProcessorPil

class transformers.GlmImageImageProcessorPil

预处理

GlmImageProcessor

class transformers.GlmImageProcessor

__call__

GlmImageVisionModel

class transformers.GlmImageVisionModel

forward

GlmImageTextModel

class transformers.GlmImageTextModel

forward

GlmImageVQVAE

class transformers.GlmImageVQVAE

forward

GlmImageModel

class transformers.GlmImageModel

forward

get_image_features

GlmImageForConditionalGeneration

类 transformers.GlmImageForConditionalGeneration

forward

get_image_features

call