diff --git a/images/tutorial/image/hidream/hidream-e1-1-guide.jpg b/images/tutorial/image/hidream/hidream-e1-1-guide.jpg new file mode 100644 index 00000000..a53f7193 Binary files /dev/null and b/images/tutorial/image/hidream/hidream-e1-1-guide.jpg differ diff --git a/tutorials/image/hidream/hidream-e1.mdx b/tutorials/image/hidream/hidream-e1.mdx index 58dad6e8..8b6fceb6 100644 --- a/tutorials/image/hidream/hidream-e1.mdx +++ b/tutorials/image/hidream/hidream-e1.mdx @@ -1,74 +1,53 @@ --- -title: "ComfyUI Native HiDream-e1 Workflow Example" +title: "ComfyUI Native HiDream-e1 e1.1 Workflow Example" sidebarTitle: "HiDream-e1" description: "This guide will help you understand and complete the ComfyUI native HiDream-I1 text-to-image workflow example" --- +import UpdateReminder from '/snippets/tutorials/update-reminder.mdx' + ![HiDream-E1 Demo](https://raw.githubusercontent.com/HiDream-ai/HiDream-E1/refs/heads/main/assets/demo.jpg) -HiDream-E1 is an interactive image editing large model officially open-sourced by HiDream-ai on April 28, 2025, built based on HiDream-I1. +HiDream-E1 is an interactive image editing large model officially open-sourced by HiDream-ai, built based on HiDream-I1. It allows you to edit images using natural language. The model is released under the [MIT License](https://github.com/HiDream-ai/HiDream-E1?tab=MIT-1-ov-file), supporting use in personal projects, scientific research, and commercial applications. In combination with the previously released [hidream-i1](/tutorials/image/hidream/hidream-i1), it enables **creative capabilities from image generation to editing**. -**ComfyUI now natively supports HiDream E1**. In this guide, we will help you complete the workflow example of using HiDream E1 in ComfyUI. - - -For reference, this workflow takes about 500s for the first run and 370s for the second run with 28 sampling steps on Google Colab L4 with 22.5GB VRAM. - - -### HiDream-E1 Information +| Name | Update Date | Inference Steps | Resolution | HuggingFace Repository | +|-------------------|-------------------|-----------------|------------------------|-------------------------------| +| HiDream-E1-Full | 2025-4-28 | 28 | 768x768 | 🤗 [HiDream-E1-Full](https://huggingface.co/HiDream-ai/HiDream-E1-Full) | +| HiDream-E1.1 | 2025-7-16 | 28 | Dynamic (1 Megapixel) | 🤗 [HiDream-E1.1](https://huggingface.co/HiDream-ai/HiDream-E1-1) | -**HiDream-E1 Model Download** -Currently, HiDream provides a full version. Here is the model information: +[HiDream E1 - Github](https://github.com/HiDream-ai/HiDream-E1) -| Name | Inference Steps | Resolution | HuggingFace Repository | -|-----------------|----------------|------------|------------------------------------------------------------------------| -| HiDream-E1-Full | 28 | 768x768 | [🤗 HiDream-E1-Full](https://huggingface.co/HiDream-ai/HiDream-E1-Full) | + +## HiDream E1 and E1.1 Workflow Related Models --[Github](https://github.com/HiDream-ai/HiDream-E1) +All the models involved in this guide can be found [here](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/tree/main/split_files). Except for the Diffusion model, E1 and E1.1 use the same models. +The corresponding workflow files also include the relevant model information. You can choose to manually download and save the models, or follow the workflow prompts to download them after loading the workflow. It is recommended to use E1.1. -## ComfyUI Native HiDream-e1 Workflow Example - - -Please upgrade your ComfyUI to the latest version (latest commit) before starting to ensure your ComfyUI has the relevant support. - - -### 1. Download HiDream-e1 Workflow and Related Files - -#### 1.1 Download Workflow File - -Please download the image below and drag it into ComfyUI. The workflow already contains model download information, and after loading, it will prompt you to download the corresponding models. - -![ComfyUI Native HiDream-e1 Workflow](https://raw.githubusercontent.com/Comfy-Org/example_workflows/main/hidream_e1/hidream_e1_full.png) +This model requires a large amount of VRAM to run. Please refer to the relevant sections for specific VRAM requirements. -#### 1.2 Download Input Image +**Diffusion Model** -Please download the image below, which we will use as input +You do not need to download both models. Since E1.1 is an iterative version based on E1, our tests show that its quality and performance are significantly improved compared to E1. +- [hidream_e1_1_bf16.safetensors (Recommended)](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/resolve/main/split_files/diffusion_models/hidream_e1_1_bf16.safetensors) 34.2GB +- [hidream_e1_full_bf16.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/resolve/main/split_files/diffusion_models/hidream_e1_full_bf16.safetensors) 34.2GB -![ComfyUI Native HiDream-e1 Workflow Input Image](https://raw.githubusercontent.com/Comfy-Org/example_workflows/main/hidream_e1/input.webp) +**Text Encoder**: -### 2. Manual Model Installation for HiDream-e1 Related Models - -All models mentioned in this guide can be found [here](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/tree/main/split_files). Please download the corresponding files and save them to the appropriate folders. - -The following model files are shared models that we will use. Please click the corresponding links to download and save according to the model file storage location. We will guide you to download the corresponding **diffusion models** in the workflow. - -**text_encoders**: - -- [clip_l_hidream.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/clip_l_hidream.safetensors) -- [clip_g_hidream.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/clip_g_hidream.safetensors) -- [t5xxl_fp8_e4m3fn_scaled.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/t5xxl_fp8_e4m3fn_scaled.safetensors) This model has been used in many workflows, you may have already downloaded this file. -- [llama_3.1_8b_instruct_fp8_scaled.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/llama_3.1_8b_instruct_fp8_scaled.safetensors) +- [clip_l_hidream.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/clip_l_hidream.safetensors) 236.12MB +- [clip_g_hidream.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/clip_g_hidream.safetensors) 1.29GB +- [t5xxl_fp8_e4m3fn_scaled.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/t5xxl_fp8_e4m3fn_scaled.safetensors) 4.8GB +- [llama_3.1_8b_instruct_fp8_scaled.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/llama_3.1_8b_instruct_fp8_scaled.safetensors) 8.46GB **VAE** -- [ae.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/vae/ae.safetensors) This is Flux's VAE model. If you have used Flux workflows before, you may have already downloaded this file. +- [ae.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/vae/ae.safetensors) 319.77MB +> This is the VAE model for Flux. If you have used the Flux workflow before, you may have already downloaded this file. -**diffusion models** -- [hidream_e1_full_bf16.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/resolve/main/split_files/diffusion_models/hidream_e1_full_bf16.safetensors) +Model Save Location -Model file storage location ``` 📂 ComfyUI/ ├── 📂 models/ @@ -80,9 +59,72 @@ Model file storage location │ └── 📂 vae/ │ │ └── ae.safetensors │ └── 📂 diffusion_models/ -│ └── hidream_e1_full_bf16.safetensors +│ ├── hidream_e1_1_bf16.safetensors +│ └── hidream_e1_full_bf16.safetensors ``` -### 3. Complete the HiDream-e1 Workflow Step by Step + + +## HiDream E1.1 ComfyUI Native Workflow Example + +E1.1 is an updated version released on July 16, 2025. This version supports dynamic 1-megapixel resolution, and the workflow uses the `Scale Image to Total Pixels` node to dynamically adjust the input image to 1 million pixels. + + +Here are the VRAM usage references during testing: +1. A100 40GB (VRAM usage 95%): First generation: 211s, second generation: 73s + +2. 4090D 24GB (VRAM usage 98%) +- Full version: Out of memory +- FP8_e4m3fn_fast (VRAM 98%) First generation: 120s, second generation: 91s + + +### 1. HiDream E1.1 Workflow and Related Materials + +Download the image below and drag it into ComfyUI with the corresponding workflow and models loaded: +![HiDream E1.1 Workflow](https://raw.githubusercontent.com/Comfy-Org/example_workflows/refs/heads/main/image/hidream/e1.1/hidream_e1_1.png) + +Download the image below as input: +![HiDream E1.1 Workflow Input](https://raw.githubusercontent.com/Comfy-Org/example_workflows/refs/heads/main/image/hidream/e1.1/input.webp) + +### 2. Step-by-step Guide to Running the HiDream-e1 Workflow +![hidream_e1_1_guide](/images/tutorial/image/hidream/hidream-e1-1-guide.jpg) + +Follow these steps to run the workflow: +1. Make sure the `Load Diffusion Model` node loads the `hidream_e1_1_bf16.safetensors` model. +2. Make sure the four corresponding text encoders in `QuadrupleCLIPLoader` are loaded correctly: + - clip_l_hidream.safetensors + - clip_g_hidream.safetensors + - t5xxl_fp8_e4m3fn_scaled.safetensors + - llama_3.1_8b_instruct_fp8_scaled.safetensors +3. Make sure the `Load VAE` node uses the `ae.safetensors` file. +4. In the `Load Image` node, load the provided input or your desired image. +5. In the `Empty Text Encoder(Positive)` node, enter **the modifications you want to make to the image**. +6. In the `Empty Text Encoder(Negative)` node, enter **the content you do not want to appear in the image**. +7. Click the `Run` button, or use the shortcut `Ctrl(cmd) + Enter` to execute image generation. + +### 3. Additional Notes on the Workflow + +- Since HiDream E1.1 supports dynamic input with a total of 1 million pixels, the workflow uses `Scale Image to Total Pixels` to process and convert all input images, which may cause the aspect ratio to differ from the original input image. +- When using the fp16 version of the model, in actual tests, the full version ran out of memory on both A100 40GB and 4090D 24GB, so the workflow is set by default to use `fp8_e4m3fn_fast` for inference. + + +## HiDream E1 ComfyUI Native Workflow Example + +E1 is a model released on April 28, 2025. This model only supports 768*768 resolution. + + +For reference, this workflow takes about 500s for the first run and 370s for the second run with 28 sampling steps on Google Colab L4 with 22.5GB VRAM. + + +### 1. HiDream-e1 workflow + +Please download the image below and drag it into ComfyUI. The workflow already contains model download information, and after loading, it will prompt you to download the corresponding models. + +![ComfyUI Native HiDream-e1 Workflow](https://raw.githubusercontent.com/Comfy-Org/example_workflows/main/hidream_e1/hidream_e1_full.png) + +Download this image below as input: +![ComfyUI Native HiDream-e1 Workflow Input](https://raw.githubusercontent.com/Comfy-Org/example_workflows/main/hidream_e1/input.webp) + +### 2. Complete the HiDream-e1 Workflow Step by Step ![hidream_e1_full_step_guide](/images/tutorial/advanced/hidream/hidream_e1_full_step_guide.jpg) @@ -98,7 +140,6 @@ Follow these steps to complete the workflow: 5. (Important) Enter **the prompt for how you want to modify the image** in the `Empty Text Encoder(Positive)` node 6. Click the `Run` button, or use the shortcut `Ctrl(cmd) + Enter` to generate the image - ### Additional Notes on ComfyUI HiDream-e1 Workflow - You may need to modify the prompt multiple times or generate multiple times to get better results diff --git a/zh-CN/tutorials/image/hidream/hidream-e1.mdx b/zh-CN/tutorials/image/hidream/hidream-e1.mdx index dfedb8a1..d04d9872 100644 --- a/zh-CN/tutorials/image/hidream/hidream-e1.mdx +++ b/zh-CN/tutorials/image/hidream/hidream-e1.mdx @@ -1,74 +1,53 @@ --- -title: "ComfyUI 原生版本 HiDream-e1 工作流示例" -sidebarTitle: "HiDream-I1" +title: "ComfyUI 原生版本 HiDream-e1 e1.1 工作流示例" +sidebarTitle: "HiDream-e1" description: "本篇将引导了解并完成 ComfyUI 原生版本 HiDream-I1 文生图工作流实例" --- +import UpdateReminder from '/snippets/zh/tutorials/update-reminder.mdx' + ![HiDream-E1 演示](https://raw.githubusercontent.com/HiDream-ai/HiDream-E1/refs/heads/main/assets/demo.jpg) -HiDream-E1 是智象未来(HiDream-ai) 于 2025 年 4 月 28 日正式开源的交互式图像编辑大模型，是基于 HiDream-I1 构建的图像编辑模型。 +HiDream-E1 是智象未来(HiDream-ai) 正式开源的交互式图像编辑大模型，是基于 HiDream-I1 构建的图像编辑模型。 可以使用自然语言来实现对图像的编辑，该模型基于 [MIT 许可证](https://github.com/HiDream-ai/HiDream-E1?tab=MIT-1-ov-file) 发布，支持用于个人项目、科学研究以及商用。 通过与此前发布的 [hidream-i1](/zh-CN/tutorials/image/hidream/hidream-i1)的共同组合，实现了 **从图像生成到编辑的** 创作能力。 -**目前 ComfyUI 已原生支持了 HiDream E1**, 在文本中，我们将引导你完成在 ComfyUI 中使用 HiDream e1 并完成对应工作流的示例。 - - -供参考，本文工作流在 Google Colab L4 22.5GB显存下采样步数 28步首次运行消耗 500s,第二次运行消耗 370s。 - - -### HiDream-E1 相关信息 - -*HiDream-E1 模型下载** -目前 HiDream 提供了 full 版本，下面是对应模型信息: - -| 名称 | 推理步骤 | 分辨率 | HuggingFace 仓库 | -|-----------------|---------|-----------|------------------------------------------------------------------------| -| HiDream-E1-Full | 28 | 768x768 | [🤗 HiDream-E1-Full](https://huggingface.co/HiDream-ai/HiDream-E1-Full) | - - --[Github](https://github.com/HiDream-ai/HiDream-E1) +| 名称 | 更新时间 | 推理步数 | 分辨率 | HuggingFace 仓库 | +|-------------------|-------------------|----------|--------------------|--------------------------| +| HiDream-E1-Full | 2025-4-28 | 28 | 768x768 | 🤗 [HiDream-E1-Full](https://huggingface.co/HiDream-ai/HiDream-E1-Full) | +| HiDream-E1.1 | 2025-7-16 | 28 | 动态（1百万像素） | 🤗 [HiDream-E1.1](https://huggingface.co/HiDream-ai/HiDream-E1-1) | -## ComfyUI 原生 HiDream-e1 工作流示例 +[HiDream E1 - Github](https://github.com/HiDream-ai/HiDream-E1) - -请在开始之前升级你的 ComfyUI 到最新版本（最新的 commit），以保证你的 ComfyUI 具有相关支持。 - + -### 1. HiDream-e1 工作流及相关文件下载 +## HiDream E1 及 E1.1 工作流相关模型 -#### 1.1 下载工作流文件 +本篇指南涉及的所有模型你都可以在[这里](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/tree/main/split_files)找到, E1, E1.1 除了 Diffusion model 之外都是用相同的模型 +我们在对应的工作流文件中也以包含了对应的模型信息，你可以选择手动下载模型保存，或者在加载工作流后按工作流提示进行下载，推荐使用 E1.1 -请下载下面的图片并拖入 ComfyUI 中，工作流已包含模型下载信息，加载后将会提示你进行对应的模型下载。 +这个模型的运行对显存占用要求极高，具体显存占用请参考对应部分的说明 -![ComfyUI 原生 HiDream-e1 工作流](https://raw.githubusercontent.com/Comfy-Org/example_workflows/main/hidream_e1/hidream_e1_full.png) +**Diffusion Model** -#### 1.2 下载输入图片 +你不用同时下载这两个模型，由于 E1.1 是基于 E1 的迭代版本，在实际测试中它的质量和效果较 E1 都有较大提升 +- [hidream_e1_1_bf16.safetensors(推荐)](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/resolve/main/split_files/diffusion_models/hidream_e1_1_bf16.safetensors) 34.2GB +- [hidream_e1_full_bf16.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/resolve/main/split_files/diffusion_models/hidream_e1_full_bf16.safetensors) 34.2GB -请下载下面的图片，我们将用于输入 +**Text Encoder**： -![ComfyUI 原生 HiDream-e1 工作流 输入图片](https://raw.githubusercontent.com/Comfy-Org/example_workflows/main/hidream_e1/input.webp) - -### 2.手动模型安装 HiDream-e1 相关模型 - -本篇指南涉及的所有模型你都可以在[这里](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/tree/main/split_files)找到，请下载对应文件并保存到相应的文件夹。 - -下面的模型文件是我们会共用的模型文件，请点击对应的链接进行下载，并参照模型文件保存位置进行保存，对应的 **diffusion models** 模型我们会在对应工作流中引导你进行下载。 - -**text_encoders**： - -- [clip_l_hidream.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/clip_l_hidream.safetensors) -- [clip_g_hidream.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/clip_g_hidream.safetensors) -- [t5xxl_fp8_e4m3fn_scaled.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/t5xxl_fp8_e4m3fn_scaled.safetensors) 这个模型在许多的工作流中都有使用过，你可能已经下载了这个文件。 -- [llama_3.1_8b_instruct_fp8_scaled.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/llama_3.1_8b_instruct_fp8_scaled.safetensors) +- [clip_l_hidream.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/clip_l_hidream.safetensors) 236.12MB +- [clip_g_hidream.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/clip_g_hidream.safetensors) 1.29GB +- [t5xxl_fp8_e4m3fn_scaled.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/t5xxl_fp8_e4m3fn_scaled.safetensors) 4.8GB +- [llama_3.1_8b_instruct_fp8_scaled.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/text_encoders/llama_3.1_8b_instruct_fp8_scaled.safetensors) 8.46GB **VAE** -- [ae.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/vae/ae.safetensors) 这个是 Flux 的 VAE 模型，如果你之前使用过 Flux 的工作流，你可能已经下载了这个文件。 +- [ae.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/vae/ae.safetensors) 319.77MB +> 这个是 Flux 的 VAE 模型，如果你之前使用过 Flux 的工作流，你可能已经下载了这个文件。 -**diffusion models** -- [hidream_e1_full_bf16.safetensors](https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/resolve/main/split_files/diffusion_models/hidream_e1_full_bf16.safetensors) +文件保存位置 -模型文件保存位置 ``` 📂 ComfyUI/ ├── 📂 models/ @@ -80,10 +59,75 @@ HiDream-E1 是智象未来(HiDream-ai) 于 2025 年 4 月 28 日正式开源的 │ └── 📂 vae/ │ │ └── ae.safetensors │ └── 📂 diffusion_models/ -│ └── hidream_e1_full_bf16.safetensors +│ ├── hidream_e1_1_bf16.safetensors +│ └── hidream_e1_full_bf16.safetensors ``` -### 3. 按步骤完成 HiDream-e1 工作流运行 +## HiDream E1.1 ComfyUI 原生工作流示例 + +E1.1 是于 2025年7月16日更新迭代的版本, 这个版本支持动态一百万分辨率，在工作流中使用了 `Scale Image to Total Pixels` 节点来将输入图片动态调整为 1百万像素 + + +这里是在测试使用时对应的显存占用参考： +1. A100 40GB (VRAM 使用率 95%)：第一次生成: 211s，第二次生成: 73s + +2. 4090D 24GB (VRAM 使用率 98%) +- 完整版本: Out of memory +- FP8_e4m3fn_fast (VRAM 98%) 第一次生成: 120s， 第二次生成: 91s + + +### 1. HiDream E1.1 工作流及相关素材 + +下载下面的图片并拖入 ComfyUI 已加载对应工作流及模型 +![HiDream E1.1 工作流](https://raw.githubusercontent.com/Comfy-Org/example_workflows/refs/heads/main/image/hidream/e1.1/hidream_e1_1.png) + +下载下面的图片作为输入 +![HiDream E1.1 工作流](https://raw.githubusercontent.com/Comfy-Org/example_workflows/refs/heads/main/image/hidream/e1.1/input.webp) + +### 2. 按步骤完成 HiDream-e1 工作流运行 +![hidream_e1_1_guide](/images/tutorial/image/hidream/hidream-e1-1-guide.jpg) + +按步骤完成工作流的运行 +1. 确保`Load Diffusion Model` 节点加载了 `hidream_e1_1_bf16.safetensors` 模型 +2. 确保`QuadrupleCLIPLoader` 中四个对应的 text encoder 被正确加载 + - clip_l_hidream.safetensors + - clip_g_hidream.safetensors + - t5xxl_fp8_e4m3fn_scaled.safetensors + - llama_3.1_8b_instruct_fp8_scaled.safetensors +3. 确保`Load VAE` 节点中使用的是 `ae.safetensors` 文件 +4. 在 `Load Image` 节点中加载提供的输入或你需要的图片 +5. 在`Empty Text Encoder(Positive)` 节点中输入 **想要对图片进行的修改** +6. 在`Empty Text Encoder(Negative)` 节点中输入 **不想要在画面中出现的内容** +7. 点击 `Run` 按钮，或者使用快捷键 `Ctrl(cmd) + Enter(回车)` 来执行图片生成 + +### 3. 工作流补充说明 + +- 由于 HiDream E1.1 支持的是动态总像素为一百万像素输入，所以工作流使用了 `Scale Image to Total Pixels` 来将所有输入图片进行处理转化，这可能会导致比例尺寸相对于输入图片会有所变化 +- 使用 fp16 版本的模型，在实际测试过程中，在 A100 40GB 和 4090D 24GB 时使用完整版本时会 Out of memory，所以工作流默认设置了使用 `fp8_e4m3fn_fast` 来进行推理 + +## HiDream E1 ComfyUI 原生 工作流示例 + +E1 是于 2025 年 4 月 28 日发布的，这个模型只支持 768*768 的分辨率 + +### 1. HiDream-e1 工作流及相关素材 + + +供参考，本文工作流在 Google Colab L4 22.5GB显存下采样步数 28步首次运行消耗 500s,第二次运行消耗 370s。 + + +#### 1.1 下载工作流文件 + +请下载下面的图片并拖入 ComfyUI 中，工作流已包含模型下载信息，加载后将会提示你进行对应的模型下载。 + +![ComfyUI 原生 HiDream-e1 工作流](https://raw.githubusercontent.com/Comfy-Org/example_workflows/main/hidream_e1/hidream_e1_full.png) + +#### 1.2 下载输入图片 + +请下载下面的图片，我们将用于输入 + +![ComfyUI 原生 HiDream-e1 工作流 输入图片](https://raw.githubusercontent.com/Comfy-Org/example_workflows/main/hidream_e1/input.webp) + +### 2. 按步骤完成 HiDream-e1 工作流运行 ![hidream_e1_full_step_guide](/images/tutorial/advanced/hidream/hidream_e1_full_step_guide.jpg) @@ -99,9 +143,8 @@ HiDream-E1 是智象未来(HiDream-ai) 于 2025 年 4 月 28 日正式开源的 5. （重要）在`Empty Text Encoder(Positive)` 节点中输入 **想要修改的画面的提示词** 6. 点击 `Run` 按钮，或者使用快捷键 `Ctrl(cmd) + Enter(回车)` 来执行图片生成 - ### ComfyUI HiDream-e1 工作流补充说明 - 可能需要修改多次提示词或者进行多次的生成才能得到较好的结果 - 这个模型在改变图片风格上比较难保持一致性，需要尽可能完善提示词 -- 由于模型支持的是 768*768 的分辨率，在实际测试中调整过其它尺寸，在其它尺寸下图像表现能力不佳，甚至差异较大 +- 由于模型支持的是 768*768 的分辨率，在实际测试中调整过其它尺寸，在其它尺寸下图像表现能力不佳，甚至差异较大 \ No newline at end of file