HIP run out of Memory

### Custom Node Testing

- [x] I have tried disabling custom nodes and the issue persists (see [how to disable custom nodes](https://docs.comfy.org/troubleshooting/custom-node-issues#step-1%3A-test-with-all-custom-nodes-disabled) if you need help)

### Expected Behavior

Text to Video

### Actual Behavior

HIP out of memory. Tried to allocate 658.00 MiB. GPU 0 has a total capacity of 19.98 GiB of which 922.00 MiB is free. Of the allocated memory 15.15 GiB is allocated by PyTorch, and 3.47 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)
This error means you ran out of memory on your GPU.

### Steps to Reproduce

Everytime

### Debug Logs

```powershell
# ComfyUI Error Report
## Error Details
- **Node ID:** 81
- **Node Type:** KSamplerAdvanced
- **Exception Type:** torch.OutOfMemoryError
- **Exception Message:** HIP out of memory. Tried to allocate 658.00 MiB. GPU 0 has a total capacity of 19.98 GiB of which 922.00 MiB is free. Of the allocated memory 15.15 GiB is allocated by PyTorch, and 3.47 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)
This error means you ran out of memory on your GPU.

TIPS: If the workflow worked before you might have accidentally set the batch_size to a large number.
## Stack Trace

File "/home/benjamin/ComfyUI/execution.py", line 496, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, hidden_inputs=hidden_inputs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/execution.py", line 315, in get_output_data
return_values = await _async_map_node_over_list(prompt_id, unique_id, obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, hidden_inputs=hidden_inputs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/execution.py", line 289, in _async_map_node_over_list
await process_inputs(input_dict, i)

File "/home/benjamin/ComfyUI/execution.py", line 277, in process_inputs
result = f(**inputs)

File "/home/benjamin/ComfyUI/nodes.py", line 1559, in sample
return common_ksampler(model, noise_seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=denoise, disable_noise=disable_noise, start_step=start_at_step, last_step=end_at_step, force_full_denoise=force_full_denoise)

File "/home/benjamin/ComfyUI/nodes.py", line 1492, in common_ksampler
samples = comfy.sample.sample(model, noise, steps, cfg, sampler_name, scheduler, positive, negative, latent_image,
denoise=denoise, disable_noise=disable_noise, start_step=start_step, last_step=last_step,
force_full_denoise=force_full_denoise, noise_mask=noise_mask, callback=callback, disable_pbar=disable_pbar, seed=seed)

File "/home/benjamin/ComfyUI/comfy/sample.py", line 45, in sample
samples = sampler.sample(noise, positive, negative, cfg=cfg, latent_image=latent_image, start_step=start_step, last_step=last_step, force_full_denoise=force_full_denoise, denoise_mask=noise_mask, sigmas=sigmas, callback=callback, disable_pbar=disable_pbar, seed=seed)

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 1154, in sample
return sample(self.model, noise, positive, negative, cfg, self.device, sampler, sigmas, self.model_options, latent_image=latent_image, denoise_mask=denoise_mask, callback=callback, disable_pbar=disable_pbar, seed=seed)

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 1044, in sample
return cfg_guider.sample(noise, latent_image, sampler, sigmas, denoise_mask, callback, disable_pbar, seed)
~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 1029, in sample
output = executor.execute(noise, latent_image, sampler, sigmas, denoise_mask, callback, disable_pbar, seed)

File "/home/benjamin/ComfyUI/comfy/patcher_extension.py", line 112, in execute
return self.original(*args, **kwargs)
~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 997, in outer_sample
output = self.inner_sample(noise, latent_image, device, sampler, sigmas, denoise_mask, callback, disable_pbar, seed)

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 980, in inner_sample
samples = executor.execute(self, sigmas, extra_args, callback, noise, latent_image, denoise_mask, disable_pbar)

File "/home/benjamin/ComfyUI/comfy/patcher_extension.py", line 112, in execute
return self.original(*args, **kwargs)
~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 752, in sample
samples = self.sampler_function(model_k, noise, sigmas, extra_args=extra_args, callback=k_callback, disable=disable_pbar, **self.extra_options)

File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/utils/_contextlib.py", line 122, in decorate_context
return func(*args, **kwargs)

File "/home/benjamin/ComfyUI/comfy/k_diffusion/sampling.py", line 199, in sample_euler
denoised = model(x, sigma_hat * s_in, **extra_args)

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 401, in __call__
out = self.inner_model(x, sigma, model_options=model_options, seed=seed)

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 953, in __call__
return self.outer_predict_noise(*args, **kwargs)
~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 960, in outer_predict_noise
).execute(x, timestep, model_options, seed)
~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/patcher_extension.py", line 112, in execute
return self.original(*args, **kwargs)
~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 963, in predict_noise
return sampling_function(self.inner_model, x, timestep, self.conds.get("negative", None), self.conds.get("positive", None), self.cfg, model_options=model_options, seed=seed)

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 381, in sampling_function
out = calc_cond_batch(model, conds, x, timestep, model_options)

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 206, in calc_cond_batch
return _calc_cond_batch_outer(model, conds, x_in, timestep, model_options)

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 214, in _calc_cond_batch_outer
return executor.execute(model, conds, x_in, timestep, model_options)
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/patcher_extension.py", line 112, in execute
return self.original(*args, **kwargs)
~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/samplers.py", line 326, in _calc_cond_batch
output = model.apply_model(input_x, timestep_, **c).chunk(batch_chunks)
~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/model_base.py", line 161, in apply_model
return comfy.patcher_extension.WrapperExecutor.new_class_executor(
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
...<2 lines>...
comfy.patcher_extension.get_all_wrappers(comfy.patcher_extension.WrappersMP.APPLY_MODEL, transformer_options)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
).execute(x, t, c_concat, c_crossattn, control, transformer_options, **kwargs)
~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/patcher_extension.py", line 112, in execute
return self.original(*args, **kwargs)
~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/model_base.py", line 200, in _apply_model
model_output = self.diffusion_model(xc, t, context=context, control=control, transformer_options=transformer_options, **extra_conds).float()
~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1780, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1791, in _call_impl
return forward_call(*args, **kwargs)

File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 614, in forward
return comfy.patcher_extension.WrapperExecutor.new_class_executor(
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
...<2 lines>...
comfy.patcher_extension.get_all_wrappers(comfy.patcher_extension.WrappersMP.DIFFUSION_MODEL, transformer_options)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
).execute(x, timestep, context, clip_fea, time_dim_concat, transformer_options, **kwargs)
~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/patcher_extension.py", line 112, in execute
return self.original(*args, **kwargs)
~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 634, in _forward
return self.forward_orig(x, timestep, context, clip_fea=clip_fea, freqs=freqs, transformer_options=transformer_options, **kwargs)[:, :, :t, :h, :w]
~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 579, in forward_orig
x = block(x, e=e0, freqs=freqs, context=context, context_img_len=context_img_len, transformer_options=transformer_options)

File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1780, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1791, in _call_impl
return forward_call(*args, **kwargs)

File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 235, in forward
y = self.self_attn(
torch.addcmul(repeat_e(e[0], x), self.norm1(x), 1 + repeat_e(e[1], x)),
freqs, transformer_options=transformer_options)

File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1780, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^

File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1791, in _call_impl
return forward_call(*args, **kwargs)

File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 79, in forward
k = qkv_fn_k(x)

File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 73, in qkv_fn_k
return apply_rope1(k, freqs)

File "/home/benjamin/ComfyUI/comfy/ldm/flux/math.py", line 41, in apply_rope1
x_out = freqs_cis[..., 0] * x_[..., 0]
~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~

## System Information
- **ComfyUI Version:** 0.3.66
- **Arguments:** main.py --use-pytorch-cross-attention
- **OS:** posix
- **Python Version:** 3.13.8 (main, Oct 8 2025, 08:53:24) [GCC 11.4.0]
- **Embedded Python:** false
- **PyTorch Version:** 2.10.0.dev20251020+rocm7.0
## Devices

- **Name:** cuda:0 Radeon RX 7900 XT : native
- **Type:** cuda
- **VRAM Total:** 21458059264
- **VRAM Free:** 21405630464
- **Torch VRAM Total:** 0
- **Torch VRAM Free:** 0

## Logs

2025-10-21T21:30:15.461836 - Checkpoint files will always be loaded safely.
2025-10-21T21:30:15.504493 - Total VRAM 20464 MB, total RAM 31299 MB
2025-10-21T21:30:15.504526 - pytorch version: 2.10.0.dev20251020+rocm7.0
2025-10-21T21:30:15.504626 - Set: torch.backends.cudnn.enabled = False for better AMD performance.
2025-10-21T21:30:15.504677 - AMD arch: gfx1100
2025-10-21T21:30:15.504691 - ROCm version: (7, 0)
2025-10-21T21:30:15.504760 - Set vram state to: NORMAL_VRAM
2025-10-21T21:30:15.504785 - Device: cuda:0 Radeon RX 7900 XT : native
2025-10-21T21:30:16.177609 - Using pytorch attention
2025-10-21T21:30:17.277462 - Python version: 3.13.8 (main, Oct 8 2025, 08:53:24) [GCC 11.4.0]
2025-10-21T21:30:17.277490 - ComfyUI version: 0.3.66
2025-10-21T21:30:17.279735 - ComfyUI frontend version: 1.28.7
2025-10-21T21:30:17.280267 - [Prompt Server] web root: /home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/comfyui_frontend_package/static
2025-10-21T21:30:17.743926 -
Import times for custom nodes:
2025-10-21T21:30:17.743965 - 0.0 seconds: /home/benjamin/ComfyUI/custom_nodes/websocket_image_save.py
2025-10-21T21:30:17.743983 -
2025-10-21T21:30:17.969388 - Context impl SQLiteImpl.
2025-10-21T21:30:17.969418 - Will assume non-transactional DDL.
2025-10-21T21:30:17.970011 - No target revision found.
2025-10-21T21:30:17.973702 - Starting server

2025-10-21T21:30:17.973827 - To see the GUI go to: http://127.0.0.1:8188
2025-10-21T21:30:24.510117 - got prompt
2025-10-21T21:30:24.542728 - Using split attention in VAE
2025-10-21T21:30:24.543244 - Using split attention in VAE
2025-10-21T21:30:24.668436 - VAE load device: cuda:0, offload device: cpu, dtype: torch.bfloat16
2025-10-21T21:30:24.712403 - Using scaled fp8: fp8 matrix mult: False, scale input: False
2025-10-21T21:30:25.188330 - Requested to load WanTEModel
2025-10-21T21:30:25.192627 - loaded completely 9.5367431640625e+25 6419.477203369141 True
2025-10-21T21:30:25.203918 - CLIP/text encoder model load device: cuda:0, offload device: cpu, current: cuda:0, dtype: torch.float16
2025-10-21T21:30:32.158569 - Using scaled fp8: fp8 matrix mult: False, scale input: True
2025-10-21T21:30:32.176912 - model weight dtype torch.float16, manual cast: None
2025-10-21T21:30:32.178074 - model_type FLOW
2025-10-21T21:30:37.410454 - Requested to load WAN21
2025-10-21T21:30:44.921133 - loaded partially 13502.228134581037 13502.224800109863 5
2025-10-21T21:30:44.980070 -
0%| | 0/2 [00:00...
comfy.patcher_extension.get_all_wrappers(comfy.patcher_extension.WrappersMP.APPLY_MODEL, transformer_options)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
).execute(x, t, c_concat, c_crossattn, control, transformer_options, **kwargs)
~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/comfy/patcher_extension.py", line 112, in execute
return self.original(*args, **kwargs)
~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/comfy/model_base.py", line 200, in _apply_model
model_output = self.diffusion_model(xc, t, context=context, control=control, transformer_options=transformer_options, **extra_conds).float()
~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1780, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1791, in _call_impl
return forward_call(*args, **kwargs)
File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 614, in forward
return comfy.patcher_extension.WrapperExecutor.new_class_executor(
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
...<2 lines>...
comfy.patcher_extension.get_all_wrappers(comfy.patcher_extension.WrappersMP.DIFFUSION_MODEL, transformer_options)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
).execute(x, timestep, context, clip_fea, time_dim_concat, transformer_options, **kwargs)
~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/comfy/patcher_extension.py", line 112, in execute
return self.original(*args, **kwargs)
~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 634, in _forward
return self.forward_orig(x, timestep, context, clip_fea=clip_fea, freqs=freqs, transformer_options=transformer_options, **kwargs)[:, :, :t, :h, :w]
~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 579, in forward_orig
x = block(x, e=e0, freqs=freqs, context=context, context_img_len=context_img_len, transformer_options=transformer_options)
File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1780, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1791, in _call_impl
return forward_call(*args, **kwargs)
File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 235, in forward
y = self.self_attn(
torch.addcmul(repeat_e(e[0], x), self.norm1(x), 1 + repeat_e(e[1], x)),
freqs, transformer_options=transformer_options)
File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1780, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1791, in _call_impl
return forward_call(*args, **kwargs)
File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 79, in forward
k = qkv_fn_k(x)
File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 73, in qkv_fn_k
return apply_rope1(k, freqs)
File "/home/benjamin/ComfyUI/comfy/ldm/flux/math.py", line 41, in apply_rope1
x_out = freqs_cis[..., 0] * x_[..., 0]
~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~
torch.OutOfMemoryError: HIP out of memory. Tried to allocate 658.00 MiB. GPU 0 has a total capacity of 19.98 GiB of which 888.00 MiB is free. Of the allocated memory 15.34 GiB is allocated by PyTorch, and 3.36 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)

2025-10-21T21:30:47.611483 - Got an OOM, unloading all loaded models.
2025-10-21T21:30:47.769981 - Prompt executed in 23.25 seconds
2025-10-21T21:31:02.109478 - got prompt
2025-10-21T21:31:02.145679 - Requested to load WAN21
2025-10-21T21:31:05.996359 - loaded partially 13320.228134581037 13307.449348449707 12
2025-10-21T21:31:06.000114 -
0%| | 0/2 [00:00...
comfy.patcher_extension.get_all_wrappers(comfy.patcher_extension.WrappersMP.APPLY_MODEL, transformer_options)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
).execute(x, t, c_concat, c_crossattn, control, transformer_options, **kwargs)
~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/comfy/patcher_extension.py", line 112, in execute
return self.original(*args, **kwargs)
~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/comfy/model_base.py", line 200, in _apply_model
model_output = self.diffusion_model(xc, t, context=context, control=control, transformer_options=transformer_options, **extra_conds).float()
~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1780, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1791, in _call_impl
return forward_call(*args, **kwargs)
File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 614, in forward
return comfy.patcher_extension.WrapperExecutor.new_class_executor(
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
...<2 lines>...
comfy.patcher_extension.get_all_wrappers(comfy.patcher_extension.WrappersMP.DIFFUSION_MODEL, transformer_options)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
).execute(x, timestep, context, clip_fea, time_dim_concat, transformer_options, **kwargs)
~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/comfy/patcher_extension.py", line 112, in execute
return self.original(*args, **kwargs)
~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 634, in _forward
return self.forward_orig(x, timestep, context, clip_fea=clip_fea, freqs=freqs, transformer_options=transformer_options, **kwargs)[:, :, :t, :h, :w]
~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 579, in forward_orig
x = block(x, e=e0, freqs=freqs, context=context, context_img_len=context_img_len, transformer_options=transformer_options)
File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1780, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1791, in _call_impl
return forward_call(*args, **kwargs)
File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 235, in forward
y = self.self_attn(
torch.addcmul(repeat_e(e[0], x), self.norm1(x), 1 + repeat_e(e[1], x)),
freqs, transformer_options=transformer_options)
File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1780, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
File "/home/benjamin/ComfyUI/.venv/lib/python3.13/site-packages/torch/nn/modules/module.py", line 1791, in _call_impl
return forward_call(*args, **kwargs)
File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 79, in forward
k = qkv_fn_k(x)
File "/home/benjamin/ComfyUI/comfy/ldm/wan/model.py", line 73, in qkv_fn_k
return apply_rope1(k, freqs)
File "/home/benjamin/ComfyUI/comfy/ldm/flux/math.py", line 41, in apply_rope1
x_out = freqs_cis[..., 0] * x_[..., 0]
~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~
torch.OutOfMemoryError: HIP out of memory. Tried to allocate 658.00 MiB. GPU 0 has a total capacity of 19.98 GiB of which 922.00 MiB is free. Of the allocated memory 15.15 GiB is allocated by PyTorch, and 3.47 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)

2025-10-21T21:31:09.014055 - Got an OOM, unloading all loaded models.
2025-10-21T21:31:09.207522 - Prompt executed in 7.09 seconds

## Attached Workflow
Please make sure that workflow does not contain any sensitive information such as API keys or passwords.

Workflow too large. Please manually upload the workflow from local file system.

## Additional Context
(Please add any additional context or steps to reproduce the error here)
```

### Other

_No response_