How to load safetensors Z-Image model?

[biship]

On Dev branch - 12/4/25 commit.

I used sd.next to download the reference HF Z-image model, it generates images fine, and it was saved to: G:\Models\huggingface\models--Tongyi-MAI--Z-Image-Turbo

Now I want to use a civitai safetensors z-image model.
I have it here: g:\models\Stable-diffusion\unstableRevolution_Bf16.safetensors
I also downloaded the text encoder & the vae and they load fine.

However, the safetensors model will not load. Do I need to create a json file of some type? If so, what name and contents? Do I need to fake a HF folder structure? I don't have to do this with SDXL - they work fine (of course they dont need a TE).

15:21:47-262239 DEBUG    Model requested: fn=call_queue.py:f:txt2img
15:21:47-263237 DEBUG    Search model: name="unstableRevolution_Bf16 [4f528ced41]" matched="g:\models\Stable-diffusion\unstableRevolution_Bf16.safetensors" type=alias
15:21:47-263737 INFO     Load model: select="unstableRevolution_Bf16 [4f528ced41]"
15:21:47-265737 INFO     Load model: detect="Z-Image" class=ZImagePipeline file="g:\models\Stable-diffusion\unstableRevolution_Bf16.safetensors"
15:21:47-266237 DEBUG    Cache clear
15:21:47-267237 DEBUG    Load model: type=Z-Image repo="g:\models\Stable-diffusion\unstableRevolution_Bf16.safetensors" config={'low_cpu_mem_usage': True, 'torch_dtype': torch.bfloat16, 'load_connected_pipeline':
                         True, 'safety_checker': None, 'requires_safety_checker': False} offload=none dtype=torch.bfloat16 args={'low_cpu_mem_usage': True, 'torch_dtype': torch.bfloat16, 'load_connected_pipeline':
                         True, 'safety_checker': None, 'requires_safety_checker': False}
15:21:47-268236 DEBUG    Load model: transformer="g:\models\Stable-diffusion\unstableRevolution_Bf16.safetensors" cls=ZImageTransformer2DModel subfolder=transformer quant="None" loader=default args={'torch_dtype':
                         torch.bfloat16, 'device_map': device(type='cuda')}
15:21:52-054937 ERROR    Load model: transformer="g:\models\Stable-diffusion\unstableRevolution_Bf16.safetensors" cls=ZImageTransformer2DModel It looks like the config file at
                         'g:\models\Stable-diffusion\unstableRevolution_Bf16.safetensors' is not a valid JSON file.
15:21:52-056437 ERROR    Load:: OSError
╭────────────────────────────────────────────────────────────────────────────────────────── Traceback (most recent call last) ──────────────────────────────────────────────────────────────────────────────────────────╮
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\diffusers\configuration_utils.py:443 in load_config                                                                                                                  │
│                                                                                                                                                                                                                       │
│  442 │   │   try:                                                                                                                                                                                                     │
│❱ 443 │   │   │   config_dict = cls._dict_from_json_file(config_file, dduf_entries=dduf_entries)                                                                                                                       │
│  444                                                                                                                                                                                                                  │
│                                                                                                                                                                                                                       │
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\diffusers\configuration_utils.py:573 in _dict_from_json_file                                                                                                         │
│                                                                                                                                                                                                                       │
│  572 │   │   │   with open(json_file, "r", encoding="utf-8") as reader:                                                                                                                                               │
│❱ 573 │   │   │   │   text = reader.read()                                                                                                                                                                             │
│  574 │   │   return json.loads(text)                                                                                                                                                                                  │
│in decode:322                                                                                                                                                                                                          │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xe6 in position 1: invalid continuation byte

During handling of the above exception, another exception occurred:

╭────────────────────────────────────────────────────────────────────────────────────────── Traceback (most recent call last) ──────────────────────────────────────────────────────────────────────────────────────────╮
│J:\Code\AI\Image\automatic\pipelines\generic.py:76 in load_transformer                                                                                                                                                 │
│                                                                                                                                                                                                                       │
│   75 │   │   │   │   load_args['variant'] = variant                                                                                                                                                                   │
│❱  76 │   │   │   transformer = cls_name.from_pretrained(                                                                                                                                                              │
│   77 │   │   │   │   repo_id,                                                                                                                                                                                         │
│                                                                                                                                                                                                                       │
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\huggingface_hub\utils\_validators.py:114 in _inner_fn                                                                                                                │
│                                                                                                                                                                                                                       │
│  113 │   │                                                                                                                                                                                                            │
│❱ 114 │   │   return fn(*args, **kwargs)                                                                                                                                                                               │
│  115                                                                                                                                                                                                                  │
│                                                                                                                                                                                                                       │
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\diffusers\models\modeling_utils.py:1057 in from_pretrained                                                                                                           │
│                                                                                                                                                                                                                       │
│  1056 │   │   # load config                                                                                                                                                                                           │
│❱ 1057 │   │   config, unused_kwargs, commit_hash = cls.load_config(                                                                                                                                                   │
│  1058 │   │   │   config_path,                                                                                                                                                                                        │
│                                                                                                                                                                                                                       │
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\huggingface_hub\utils\_validators.py:114 in _inner_fn                                                                                                                │
│                                                                                                                                                                                                                       │
│  113 │   │                                                                                                                                                                                                            │
│❱ 114 │   │   return fn(*args, **kwargs)                                                                                                                                                                               │
│  115                                                                                                                                                                                                                  │
│                                                                                                                                                                                                                       │
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\diffusers\configuration_utils.py:447 in load_config                                                                                                                  │
│                                                                                                                                                                                                                       │
│  446 │   │   except (json.JSONDecodeError, UnicodeDecodeError):                                                                                                                                                       │
│❱ 447 │   │   │   raise EnvironmentError(f"It looks like the config file at '{config_file}' is not a valid JSON file.")                                                                                                │
│  448                                                                                                                                                                                                                  │
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
OSError: It looks like the config file at 'g:\models\Stable-diffusion\unstableRevolution_Bf16.safetensors' is not a valid JSON file.
15:21:52-150440 ERROR    Load model: path="g:\models\Stable-diffusion\unstableRevolution_Bf16.safetensors" It looks like the config file at 'g:\models\Stable-diffusion\unstableRevolution_Bf16.safetensors' is not a
                         valid JSON file.
Progress 90.93it/s █▊         20% 1/5 00:00 00:00 Loading pipeline components...
15:21:52-524451 ERROR    Load model: file="g:\models\Stable-diffusion\unstableRevolution_Bf16.safetensors" pipeline=Z-Image config={'low_cpu_mem_usage': True, 'torch_dtype': torch.bfloat16, 'load_connected_pipeline':
                         True, 'safety_checker': None, 'requires_safety_checker': False, 'extract_ema': False, 'use_safetensors': True, 'cache_dir': 'g:\\models\\huggingface'} Failed to load Gemma2Model. Weights for
                         this component appear to be missing in the checkpoint.
                         Please load the component before passing it in as an argument to `from_single_file`.

                         text_encoder = Gemma2Model.from_pretrained('...')
                         pipe = ZImagePipeline.from_single_file(<checkpoint path>, text_encoder=text_encoder)


15:21:52-530453 ERROR    Load model: file="g:\models\Stable-diffusion\unstableRevolution_Bf16.safetensors" is not a complete model
15:21:52-549953 ERROR    Load model: name="unstableRevolution_Bf16" not loaded

[vladmandic]

Now I want to use a civitai safetensors z-image model.

why? its the same model, but packaged differently.
not supported by sdnext as i don't see a need to support same thing just packaged differently.
if there are actual finetunes in the future, we'll add support. this, i don't see a need.

[biship]

You misunderstood.

[vladmandic]

sorry if I misunderstood, can you explain what?

[biship]

Sorry that I was not clear. It's a fine tune.

I was able to download it directly from the Models -> CivitAI tab.
Is there information in the HF config.json you need to know to be able to run it, even when you know (or are told) the pipeline and the model, unet/TE & vae are directly specified?
ComfyUI works fine with these files.
You don't plan on supporting this type of loading?
Can we fake a json? or perhaps have your CivitAI downloader create it?
Please forgive me if it's not that simple,

https://civitai.com/models/2194671/zit-decenter-v1?modelVersionId=2471116

08:57:16-048033 DEBUG    Search model: name="ZImage\zitDecenterV1_v10" matched="g:\models\Stable-diffusion\ZImage\zitDecenterV1_v10.safetensors" type=alias
08:57:16-049031 INFO     Load model: select="ZImage\zitDecenterV1_v10"
08:57:16-050532 INFO     Load model: detect="Z-Image" class=ZImagePipeline file="g:\models\Stable-diffusion\ZImage\zitDecenterV1_v10.safetensors"
08:57:16-051532 DEBUG    Cache clear
08:57:16-056531 DEBUG    Load model: type=Z-Image repo="g:\models\Stable-diffusion\ZImage\zitDecenterV1_v10.safetensors" config={'low_cpu_mem_usage': True, 'torch_dtype': torch.bfloat16, 'load_connected_pipeline': True, 'safety_checker': None,
                         'requires_safety_checker': False} offload=none dtype=torch.bfloat16 args={'low_cpu_mem_usage': True, 'torch_dtype': torch.bfloat16, 'load_connected_pipeline': True, 'safety_checker': None, 'requires_safety_checker':
                         False}
08:57:16-058031 DEBUG    Load model: transformer="g:\models\Stable-diffusion\ZImage\zitDecenterV1_v10.safetensors" cls=ZImageTransformer2DModel subfolder=transformer quant="None" loader=default args={'torch_dtype': torch.bfloat16, 'device_map':
                         device(type='cuda')}
08:57:16-189448 INFO     API user=None code=200 http/1.1 GET /sdapi/v1/network/desc 127.0.0.1 0.002
08:58:04-670632 ERROR    Load model: transformer="g:\models\Stable-diffusion\ZImage\zitDecenterV1_v10.safetensors" cls=ZImageTransformer2DModel It looks like the config file at 'g:\models\Stable-diffusion\ZImage\zitDecenterV1_v10.safetensors'
                         is not a valid JSON file.
08:58:04-671632 ERROR    Load:: OSError
╭─────────────────────────────────────────────────────────────────────────────────────────────────────── Traceback (most recent call last) ────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\diffusers\configuration_utils.py:443 in load_config                                                                                                                                             │
│                                                                                                                                                                                                                                                  │
│  442 │   │   try:                                                                                                                                                                                                                                │
│❱ 443 │   │   │   config_dict = cls._dict_from_json_file(config_file, dduf_entries=dduf_entries)                                                                                                                                                  │
│  444                                                                                                                                                                                                                                             │
│                                                                                                                                                                                                                                                  │
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\diffusers\configuration_utils.py:573 in _dict_from_json_file                                                                                                                                    │
│                                                                                                                                                                                                                                                  │
│  572 │   │   │   with open(json_file, "r", encoding="utf-8") as reader:                                                                                                                                                                          │
│❱ 573 │   │   │   │   text = reader.read()                                                                                                                                                                                                        │
│  574 │   │   return json.loads(text)                                                                                                                                                                                                             │
│in decode:322                                                                                                                                                                                                                                     │
╰──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xd0 in position 0: invalid continuation byte

During handling of the above exception, another exception occurred:

╭─────────────────────────────────────────────────────────────────────────────────────────────────────── Traceback (most recent call last) ────────────────────────────────────────────────────────────────────────────────────────────────────────╮
│J:\Code\AI\Image\automatic\pipelines\generic.py:76 in load_transformer                                                                                                                                                                            │
│                                                                                                                                                                                                                                                  │
│   75 │   │   │   │   load_args['variant'] = variant                                                                                                                                                                                              │
│❱  76 │   │   │   transformer = cls_name.from_pretrained(                                                                                                                                                                                         │
│   77 │   │   │   │   repo_id,                                                                                                                                                                                                                    │
│                                                                                                                                                                                                                                                  │
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\huggingface_hub\utils\_validators.py:114 in _inner_fn                                                                                                                                           │
│                                                                                                                                                                                                                                                  │
│  113 │   │                                                                                                                                                                                                                                       │
│❱ 114 │   │   return fn(*args, **kwargs)                                                                                                                                                                                                          │
│  115                                                                                                                                                                                                                                             │
│                                                                                                                                                                                                                                                  │
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\diffusers\models\modeling_utils.py:1062 in from_pretrained                                                                                                                                      │
│                                                                                                                                                                                                                                                  │
│  1061 │   │   # load config                                                                                                                                                                                                                      │
│❱ 1062 │   │   config, unused_kwargs, commit_hash = cls.load_config(                                                                                                                                                                              │
│  1063 │   │   │   config_path,                                                                                                                                                                                                                   │
│                                                                                                                                                                                                                                                  │
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\huggingface_hub\utils\_validators.py:114 in _inner_fn                                                                                                                                           │
│                                                                                                                                                                                                                                                  │
│  113 │   │                                                                                                                                                                                                                                       │
│❱ 114 │   │   return fn(*args, **kwargs)                                                                                                                                                                                                          │
│  115                                                                                                                                                                                                                                             │
│                                                                                                                                                                                                                                                  │
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\diffusers\configuration_utils.py:447 in load_config                                                                                                                                             │
│                                                                                                                                                                                                                                                  │
│  446 │   │   except (json.JSONDecodeError, UnicodeDecodeError):                                                                                                                                                                                  │
│❱ 447 │   │   │   raise EnvironmentError(f"It looks like the config file at '{config_file}' is not a valid JSON file.")                                                                                                                           │
│  448                                                                                                                                                                                                                                             │
╰──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
OSError: It looks like the config file at 'g:\models\Stable-diffusion\ZImage\zitDecenterV1_v10.safetensors' is not a valid JSON file.
08:58:04-827158 ERROR    Load model: path="g:\models\Stable-diffusion\ZImage\zitDecenterV1_v10.safetensors" It looks like the config file at 'g:\models\Stable-diffusion\ZImage\zitDecenterV1_v10.safetensors' is not a valid JSON file.
Progress 21970.16it/s █████████████████ 100% 11/11 00:00 00:00 Fetching 11 files
Progress 68.92it/s █▊         20% 1/5 00:00 00:00 Loading pipeline components...
08:58:04-964592 ERROR    Load model: file="g:\models\Stable-diffusion\ZImage\zitDecenterV1_v10.safetensors" pipeline=Z-Image config={'low_cpu_mem_usage': True, 'torch_dtype': torch.bfloat16, 'load_connected_pipeline': True, 'safety_checker':
                         None, 'requires_safety_checker': False, 'extract_ema': False, 'use_safetensors': True, 'cache_dir': 'g:\\models\\huggingface'} Failed to load Gemma2Model. Weights for this component appear to be missing in the
                         checkpoint.
                         Please load the component before passing it in as an argument to `from_single_file`.

                         text_encoder = Gemma2Model.from_pretrained('...')
                         pipe = ZImagePipeline.from_single_file(<checkpoint path>, text_encoder=text_encoder)


08:58:04-966091 ERROR    Load model: file="g:\models\Stable-diffusion\ZImage\zitDecenterV1_v10.safetensors" is not a complete model
08:58:04-968092 ERROR    Load model: name="ZImage\zitDecenterV1_v10" not loaded

[vladmandic]

now that you posted link to model, i can take a look at it.
yes, that looks like a valid fine tune and i'll look into adding support for z-image finetunes.
no, its not as simple as faking a json - those are packaged differently.

[biship]

great, thank you for considering this.

[vladmandic]

added in dev branch, note is in the changelog:

• Z-Image support loading transformer file-tunes in safetensors format
  as with any transformers/unet finetunes, place them then models/unet
  and use UNET Model to load safetensors file as they are not complete models

[biship]

@vladmandic I can open a issue if you prefer, but I can't get it to work:

Using: https://civitai.com/models/2168935/z-image

When I hit Generate:

16:00:37-004583 TRACE    Server: alive=True requests=442 memory=54.94/95.71 status='idle' task='' timestamp=None current='' id='' job=0 jobs=0 total=128 step=0 steps=0 queued=0 uptime=750 elapsed=2.65 eta=None progress=0
16:00:37-318850 DEBUG    Model requested: fn=call_queue.py:f:txt2img
16:00:37-321364 DEBUG    Cache clear
16:00:37-322364 WARNING  Aborted: op=txt model not loaded

[vladmandic]

dont set pipeline, that is done automatically 99% of the time
you MUST set base model. everything else such as te/unet/vae works as replacement for equivalent component in base model.
for example, set z-image as base model, dont set text-encoder and vae since you're not using any fine tunes for those components and set unstableRevolution_Bf16 as unet.

[biship]

I assume when you say 'do not set', you mean use Autodetect or Default (because lists do not contain a blank value).

pipeline autodetect, base model set to HF diffusers z-image, te & vae are at default. unstableRevolution_Bf16 loaded as unet:

17:01:24-063941 DEBUG    Model requested: fn=call_queue.py:f:txt2img
17:01:24-065445 DEBUG    Search model: name="Diffusers\Tongyi-MAI/Z-Image-Turbo [5f4b9cbb80]" matched="g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947" type=alias
17:01:24-065949 INFO     Load model: select="Diffusers\Tongyi-MAI/Z-Image-Turbo [5f4b9cbb80]"
17:01:24-070448 INFO     Autodetect model: detect="Z-Image" class=ZImagePipeline file="g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947"
17:01:24-071450 DEBUG    Cache clear
17:01:24-072950 DEBUG    Load model: type=Z-Image repo="Tongyi-MAI/Z-Image-Turbo" config={'low_cpu_mem_usage': True, 'torch_dtype': torch.bfloat16, 'load_connected_pipeline': True, 'safety_checker': None, 'requires_safety_checker':
                         False} offload=none dtype=torch.bfloat16 args={'low_cpu_mem_usage': True, 'torch_dtype': torch.bfloat16, 'load_connected_pipeline': True, 'safety_checker': None, 'requires_safety_checker': False}
17:01:24-074449 DEBUG    Load model: transformer="g:\models\UNET\ZImage\unstableRevolution_Bf16.safetensors" cls=ZImageTransformer2DModel quant="None" loader=default args={'torch_dtype': torch.bfloat16, 'device_map':
                         device(type='cuda')}
17:01:26-436168 ERROR    Load model: transformer="Tongyi-MAI/Z-Image-Turbo" cls=ZImageTransformer2DModel 'torch.device' object has no attribute 'items'
17:01:26-437169 ERROR    Load:: AttributeError
╭────────────────────────────────────────────────────────────────────────────────────────────────── Traceback (most recent call last) ───────────────────────────────────────────────────────────────────────────────────────────────────╮
│J:\Code\AI\Image\automatic\pipelines\generic.py:60 in load_transformer                                                                                                                                                                  │
│                                                                                                                                                                                                                                        │
│   59 │   │   │   loader = cls_name.from_single_file if hasattr(cls_name, 'from_single_file') else cls_name.from_pretrained                                                                                                             │
│❱  60 │   │   │   transformer = loader(                                                                                                                                                                                                 │
│   61 │   │   │   │   local_file,                                                                                                                                                                                                       │
│                                                                                                                                                                                                                                        │
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\huggingface_hub\utils\_validators.py:114 in _inner_fn                                                                                                                                 │
│                                                                                                                                                                                                                                        │
│  113 │   │                                                                                                                                                                                                                             │
│❱ 114 │   │   return fn(*args, **kwargs)                                                                                                                                                                                                │
│  115                                                                                                                                                                                                                                   │
│                                                                                                                                                                                                                                        │
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\diffusers\loaders\single_file_model.py:435 in from_single_file                                                                                                                        │
│                                                                                                                                                                                                                                        │
│  434 │   │   if device_map is not None:                                                                                                                                                                                                │
│❱ 435 │   │   │   expanded_device_map = _expand_device_map(device_map, model_state_dict.keys())                                                                                                                                         │
│  436 │   │   │   _caching_allocator_warmup(model, expanded_device_map, torch_dtype, hf_quantizer)                                                                                                                                      │
│                                                                                                                                                                                                                                        │
│J:\Code\AI\Image\automatic\venv\Lib\site-packages\diffusers\models\model_loading_utils.py:708 in _expand_device_map                                                                                                                     │
│                                                                                                                                                                                                                                        │
│  707 │   new_device_map = {}                                                                                                                                                                                                           │
│❱ 708 │   for module, device in device_map.items():                                                                                                                                                                                     │
│  709 │   │   new_device_map.update(                                                                                                                                                                                                    │
╰────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
AttributeError: 'torch.device' object has no attribute 'items'
17:01:26-489206 ERROR    Load model: path="g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947" 'torch.device' object has no attribute 'items'
Progress ?it/s                     0% 0/5 00:00 ? Loading pipeline components...17:01:26-494205 ERROR    An error occurred while trying to fetch g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer: Error no file named
                         diffusion_pytorch_model.safetensors found in directory g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer.
17:01:26-494205 ERROR    An error occurred while trying to fetch g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer: Error no file named
                         diffusion_pytorch_model.safetensors found in directory g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer.
17:01:26-494205 ERROR    An error occurred while trying to fetch g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer: Error no file named
                         diffusion_pytorch_model.safetensors found in directory g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer.
Progress ?it/s                     0% 0/5 00:00 ? Loading pipeline components...
Progress ?it/s                     0% 0/5 00:00 ? Loading pipeline components...17:01:26-498713 ERROR    An error occurred while trying to fetch g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer: Error no file named
                         diffusion_pytorch_model.safetensors found in directory g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer.
17:01:26-498713 ERROR    An error occurred while trying to fetch g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer: Error no file named
                         diffusion_pytorch_model.safetensors found in directory g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer.
17:01:26-498713 ERROR    An error occurred while trying to fetch g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer: Error no file named
                         diffusion_pytorch_model.safetensors found in directory g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer.
Progress ?it/s                     0% 0/5 00:00 ? Loading pipeline components...
Progress ?it/s                     0% 0/5 00:00 ? Loading pipeline components...17:01:26-502718 ERROR    An error occurred while trying to fetch g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer: Error no file named
                         diffusion_pytorch_model.safetensors found in directory g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer.
17:01:26-502718 ERROR    An error occurred while trying to fetch g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer: Error no file named
                         diffusion_pytorch_model.safetensors found in directory g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer.
17:01:26-502718 ERROR    An error occurred while trying to fetch g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer: Error no file named
                         diffusion_pytorch_model.safetensors found in directory g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer.
Progress ?it/s                     0% 0/5 00:00 ? Loading pipeline components...
Progress 13.81it/s ████▌      50% 2/4 00:00 00:00 Loading pipeline components...
17:01:26-651554 ERROR    StableDiffusionPipeline: Error no file named pytorch_model.bin, model.safetensors, tf_model.h5, model.ckpt.index or flax_model.msgpack found in directory
                         g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\text_encoder.
17:01:26-652557 ERROR    Load model: g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947 detected=Error no file named diffusion_pytorch_model.bin found in directory
                         g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer. auto=Error no file named diffusion_pytorch_model.bin found in directory
                         g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer. diffusion=Error no file named diffusion_pytorch_model.bin found in directory
                         g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\transformer. base=Error no file named pytorch_model.bin, model.safetensors, tf_model.h5,
                         model.ckpt.index or flax_model.msgpack found in directory g:\models\Diffusers\models--Tongyi-MAI--Z-Image-Turbo\snapshots\5f4b9cbb80cc95ba44fe6667dfd75710f7db2947\text_encoder.
17:01:26-655986 ERROR    Load model: name="Diffusers\Tongyi-MAI/Z-Image-Turbo" not loaded
17:01:26-656987 WARNING  Aborted: op=txt model not loaded

I might just be really dumb and need a screenshot. Sorry for the troubles.

[vladmandic]

what is your value for settings -> model loading -> Model load device map?

[biship]

It's set to GPU. Default has the same issue. I opened a Issue so i can provide more information.
#4478