RuntimeError: Placeholder storage has not been allocated on MPS device!

[obs945]

Macbook Pro M1 chip and try to run Z-image "RuntimeError: Placeholder storage has not been allocated on MPS device!" occured

Error as below

Traceback (most recent call last):
File "/Volumes/MX500 SSD/chromaforge/modules_forge/main_thread.py", line 30, in work
self.result = self.func(*self.args, **self.kwargs)
File "/Volumes/MX500 SSD/chromaforge/modules/txt2img.py", line 131, in txt2img_function
processed = processing.process_images(p)
File "/Volumes/MX500 SSD/chromaforge/modules/processing.py", line 847, in process_images
res = process_images_inner(p)
File "/Volumes/MX500 SSD/chromaforge/modules/processing.py", line 976, in process_images_inner
p.setup_conds()
File "/Volumes/MX500 SSD/chromaforge/modules/processing.py", line 1635, in setup_conds
super().setup_conds()
File "/Volumes/MX500 SSD/chromaforge/modules/processing.py", line 505, in setup_conds
self.c = self.get_conds_with_caching(prompt_parser.get_multicond_learned_conditioning, prompts, total_steps, [self.cached_c], self.extra_network_data)
File "/Volumes/MX500 SSD/chromaforge/modules/processing.py", line 474, in get_conds_with_caching
cache[1] = function(shared.sd_model, required_prompts, steps, hires_steps, shared.opts.use_old_scheduling)
File "/Volumes/MX500 SSD/chromaforge/modules/prompt_parser.py", line 262, in get_multicond_learned_conditioning
learned_conditioning = get_learned_conditioning(model, prompt_flat_list, steps, hires_steps, use_old_scheduling)
File "/Volumes/MX500 SSD/chromaforge/modules/prompt_parser.py", line 189, in get_learned_conditioning
conds = model.get_learned_conditioning(texts)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context
return func(*args, **kwargs)
File "/Volumes/MX500 SSD/chromaforge/backend/diffusion_engine/zimage.py", line 466, in get_learned_conditioning
outputs = text_encoder(
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
return forward_call(*args, **kwargs)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/transformers/utils/generic.py", line 1072, in wrapper
outputs = func(self, *args, **kwargs)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/transformers/models/qwen3/modeling_qwen3.py", line 371, in forward
inputs_embeds = self.embed_tokens(input_ids)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
return forward_call(*args, **kwargs)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/nn/modules/sparse.py", line 192, in forward
return F.embedding(
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/nn/functional.py", line 2546, in embedding
return torch.embedding(weight, input, padding_idx, scale_grad_by_freq, sparse)
RuntimeError: Placeholder storage has not been allocated on MPS device!
Placeholder storage has not been allocated on MPS device!

Besides, text encoder seems only support safetensors (use GGUF will show AssertionError: You do not have Qwen3 text encoder state dict!)

[maybleMyers]

I have no way to test on mac, I can test a gguf if you provide a link.

[obs945]

Thanks for answer my question.
I used Qwen3-4B-Q8_0.gguf download from https://huggingface.co/Qwen/Qwen3-4B-GGUF

[maybleMyers]

seems to be working with that one now, I had to make alot of changes. It is on the gguf branch.

[obs945]

@maybleMyers Thanks for GGUF fixing.

I also did some research for the "RuntimeError: Placeholder storage has not been allocated on MPS device" issue
Possible solutions for Z-image on forge

1. no_cuda=True
2. change from device=device to device="cpu"
3. force use cpu just like flux process

Hope these can he helpful for make MAC silicon chip generate Z-image on chromaforge happened.

[maybleMyers]

Are you using export COMMANDLINE_ARGS="--use-cpu all"?

[obs945]

Yes, I using --use-cpu all in export commandline and chose cpu in swap location.

Error message as below

---

Traceback (most recent call last):
File "/Volumes/MX500 SSD/chromaforge/modules_forge/main_thread.py", line 30, in work
self.result = self.func(*self.args, **self.kwargs)
File "/Volumes/MX500 SSD/chromaforge/modules/txt2img.py", line 132, in txt2img_function
processed = processing.process_images(p)
File "/Volumes/MX500 SSD/chromaforge/modules/processing.py", line 848, in process_images
res = process_images_inner(p)
File "/Volumes/MX500 SSD/chromaforge/modules/processing.py", line 977, in process_images_inner
p.setup_conds()
File "/Volumes/MX500 SSD/chromaforge/modules/processing.py", line 1636, in setup_conds
super().setup_conds()
File "/Volumes/MX500 SSD/chromaforge/modules/processing.py", line 506, in setup_conds
self.c = self.get_conds_with_caching(prompt_parser.get_multicond_learned_conditioning, prompts, total_steps, [self.cached_c], self.extra_network_data)
File "/Volumes/MX500 SSD/chromaforge/modules/processing.py", line 475, in get_conds_with_caching
cache[1] = function(shared.sd_model, required_prompts, steps, hires_steps, shared.opts.use_old_scheduling)
File "/Volumes/MX500 SSD/chromaforge/modules/prompt_parser.py", line 262, in get_multicond_learned_conditioning
learned_conditioning = get_learned_conditioning(model, prompt_flat_list, steps, hires_steps, use_old_scheduling)
File "/Volumes/MX500 SSD/chromaforge/modules/prompt_parser.py", line 189, in get_learned_conditioning
conds = model.get_learned_conditioning(texts)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/utils/_contextlib.py", line 120, in decorate_context
return func(*args, **kwargs)
File "/Volumes/MX500 SSD/chromaforge/backend/diffusion_engine/zimage.py", line 480, in get_learned_conditioning
outputs = text_encoder(
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
return forward_call(*args, **kwargs)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/transformers/utils/generic.py", line 1072, in wrapper
outputs = func(self, *args, **kwargs)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/transformers/models/qwen3/modeling_qwen3.py", line 371, in forward
inputs_embeds = self.embed_tokens(input_ids)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1773, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1784, in _call_impl
return forward_call(*args, **kwargs)
File "/Volumes/MX500 SSD/chromaforge/backend/operations.py", line 322, in forward
return super().forward(x)
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/nn/modules/sparse.py", line 192, in forward
return F.embedding(
File "/Volumes/MX500 SSD/chromaforge/venv/lib/python3.10/site-packages/torch/nn/functional.py", line 2546, in embedding
return torch.embedding(weight, input, padding_idx, scale_grad_by_freq, sparse)
RuntimeError: Placeholder storage has not been allocated on MPS device!
Placeholder storage has not been allocated on MPS device!