google/siglip-so400m-patch14-384

[zhaoqi571436204]

ComfyUI Error Report

Error Details

• Node ID: 224
• Node Type: IPAdapterFluxLoader
• Exception Type: OSError

• Exception Message: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/siglip-so400m-patch14-384 is not the path to a directory containing a file named config.json. Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'.

  Stack Trace

  
  File "D:\AI\ComfyUI\execution.py", line 323, in execute
  output_data, output_ui, has_subgraph = get_output_data(obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb)
  
  File "D:\AI\ComfyUI\execution.py", line 198, in get_output_data
  return_values = _map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb)
  
  File "D:\AI\ComfyUI\execution.py", line 169, in _map_node_over_list
  process_inputs(input_dict, i)
  
  File "D:\AI\ComfyUI\execution.py", line 158, in process_inputs
  results.append(getattr(obj, func)(**inputs))
  
  File "D:\AI\ComfyUI\custom_nodes\ComfyUI-IPAdapter-Flux\ipadapter_flux.py", line 123, in load_model
  model = InstantXFluxIPAdapterModel(image_encoder_path=clip_vision, ip_ckpt=ipadapter, device=provider, num_tokens=128)
  
  File "D:\AI\ComfyUI\custom_nodes\ComfyUI-IPAdapter-Flux\ipadapter_flux.py", line 46, in __init__
  self.image_encoder = SiglipVisionModel.from_pretrained(self.image_encoder_path).to(self.device, dtype=torch.bfloat16)
  
  File "D:\AI\ComfyUI\python\lib\site-packages\transformers\modeling_utils.py", line 3613, in from_pretrained
  config, model_kwargs = cls.config_class.from_pretrained(
  
  File "D:\AI\ComfyUI\python\lib\site-packages\transformers\models\siglip\configuration_siglip.py", line 213, in from_pretrained
  config_dict, kwargs = cls.get_config_dict(pretrained_model_name_or_path, **kwargs)
  
  File "D:\AI\ComfyUI\python\lib\site-packages\transformers\configuration_utils.py", line 574, in get_config_dict
  config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs)
  
  File "D:\AI\ComfyUI\python\lib\site-packages\transformers\configuration_utils.py", line 633, in _get_config_dict
  resolved_config_file = cached_file(
  
  File "D:\AI\ComfyUI\python\lib\site-packages\transformers\utils\hub.py", line 446, in cached_file
  raise EnvironmentError(

[tanggogogo123]

same Error,Have you solved this problem?

[zhaoqi571436204]

同样的错误，你解决这个问题了吗？

Nope.

[ericxl277]

I'm also having the same problem, the model file is actually in the clip vision directory, but it still says it can't be found. By modifying the ipadapter_flux.py line 111 of this file, you can replace the model path with your own. but there are still other errors(issue 10), and the workflow still doesn't work.

[AaronPanXiaoFeng]

Modify "clip_vision" key in ipadapter_flux.py line 112 works, e.g.,

[ericxl277]

I'm also having the same problem, the model file is actually in the clip vision directory, but it still says it can't be found. By modifying the ipadapter_flux.py line 111 of this file, you can replace the model path with your own. but there are still other errors(issue 10), and the workflow still doesn't work.

line 112, made a mistake

[zhaoqi571436204]

修改 ipadapter_flux.py 行 112 中的 “clip_vision” 键，例如：

How does that work?

[AaronPanXiaoFeng]

修改 ipadapter_flux.py 行 112 中的 “clip_vision” 键，例如：

How does that work?

add your path of "google/siglip-so400m-patch14-384" to "clip_vision" key in ipadapter_flux.py line 112, and switch to it in your workflow,

[philipy1219]

This module reads the Hugging Face cache directory's path, not the clip_vision folder. The clip_vision folder in the original ComfyUI directory does not contain folders for Hugging Face Transformers but rather a single file. If you have an additional Hugging Face cache directory, you can modify the Hugging Face cache directory configuration. The specific method is as follows:

---

How to Modify the Hugging Face Cache Directory via Environment Variables and Global Configuration:

1. Setting the Environment Variable

You can set the HF_HOME environment variable to specify a new cache directory.

For Linux/Mac, run the following command in the terminal:

export HF_HOME=/path/to/your/cache

For Windows, use the following command:

set HF_HOME=C:\path\to\your\cache

---

2. Modifying the Global Configuration File

Hugging Face uses a global configuration file to manage settings such as the cache directory:

1. Open or create the global configuration file at the following paths:

   • Linux/Mac: ~/.cache/huggingface/hub/config.json
   • Windows: %USERPROFILE%\.cache\huggingface\hub\config.json

2. Add or update the following entry:

{
    "cache_dir": "/path/to/your/cache"
}

这个module读取的是huggingface缓存目录的地址，而不是clip_vision文件夹，comfyui原始目录的clip_vision文件夹内放置的不是供huggingface transformers访问的文件夹而是单个文件。如果您有一个额外的huggingface缓存目录，可以修改 Hugging Face 的缓存目录配置。具体方法如下：

---

如何通过环境变量和全局配置文件修改 Hugging Face 缓存目录：

1. 设置环境变量

您可以通过设置 HF_HOME 环境变量来指定新的缓存目录。

对于 Linux/Mac，在终端中运行以下命令：

export HF_HOME=/path/to/your/cache

对于 Windows，使用以下命令：

set HF_HOME=C:\path\to\your\cache

---

2. 修改全局配置文件

Hugging Face 使用全局配置文件来管理缓存目录等设置：

1. 打开或创建全局配置文件，路径如下：

   • Linux/Mac: ~/.cache/huggingface/hub/config.json
   • Windows: %USERPROFILE%\.cache\huggingface\hub\config.json

2. 添加或更新以下内容：

{
    "cache_dir": "/path/to/your/cache"
}

[Amazon90]

修改 ipadapter_flux.py 行 112 中的 “clip_vision” 键，例如：

How does that work?

add your path of "google/siglip-so400m-patch14-384" to "clip_vision" key in ipadapter_flux.py line 112, and switch to it in your workflow,

This doesn't work for me, how should I modify it? https://github.com/Shakker-Labs/ComfyUI-IPAdapter-Flux/issues/15