Problem with convrot model
When I use the convrot models for generating 5sec video, it works fine. When I use the same workflow with everything the same but 10sec instead of 5sec. it fails after some time. My RTX 5090 goes beserk and the only way to stop is to restart my pc (128Gb RAM) Any idea how to fix this error? Herewith the cmd output; [INFO] got prompt
[INFO] Model MiniMaxH3TEModel_ prepared for dynamic VRAM loading. 25882MB Staged. 0 patches attached. Force pre-loaded 310 weights: 1272 KB.
[INFO] Model MiniMaxH3VideoVAE prepared for dynamic VRAM loading. 4965MB Staged. 0 patches attached. Force pre-loaded 128 weights: 348 KB.
[INFO] Model MiniMaxH3VideoVAE prepared for dynamic VRAM loading. 4965MB Staged. 0 patches attached. Force pre-loaded 128 weights: 348 KB.
[INFO] 0 models unloaded.
[INFO] Model MiniMaxH3 prepared for dynamic VRAM loading. 32427MB Staged. 0 patches attached. Force pre-loaded 210 weights: 1142 KB.
70%|█████████████████████████████████████████████████▋ | 14/20 [12:54<05:31, 55.29s/it]
[ERROR] !!! Exception during processing !!! CUDA error: unknown error
Search for cudaErrorUnknown' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information. CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1 Compile with TORCH_USE_CUDA_DSA` to enable device-side assertions.
[ERROR] Traceback (most recent call last):
File "H:\ComfyUI_windows_portable_1\ComfyUI\execution.py", line 545, 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, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\ComfyUI\execution.py", line 344, 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, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\ComfyUI\execution.py", line 318, in async_map_node_over_list
await process_inputs(input_dict, i)
File "H:\ComfyUI_windows_portable_1\ComfyUI\execution.py", line 306, in process_inputs
result = f(**inputs)
^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\ComfyUI\comfy_api\internal_init.py", line 149, in wrapped_func
return method(locked_class, **inputs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\ComfyUI\comfy_api\latest_io.py", line 1935, in EXECUTE_NORMALIZED
to_return = cls.execute(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\ComfyUI\comfy_extras\nodes_custom_sampler.py", line 1049, in execute
samples = guider.sample(noise.generate_noise(latent), latent_image, sampler, sigmas, denoise_mask=noise_mask, callback=callback, disable_pbar=disable_pbar, seed=noise.seed)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\ComfyUI\comfy\samplers.py", line 1335, in sample
output = executor.execute(noise, latent_image, sampler, sigmas, denoise_mask, callback, disable_pbar, seed, latent_shapes=latent_shapes)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\ComfyUI\comfy\patcher_extension.py", line 113, in execute
return self.original(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\ComfyUI\comfy\samplers.py", line 1262, in outer_sample
output = self.inner_sample(noise, latent_image, device, sampler, sigmas, denoise_mask, callback, disable_pbar, seed, latent_shapes=latent_shapes)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\ComfyUI\comfy\samplers.py", line 1237, in inner_sample
samples = executor.execute(self, sigmas, extra_args, callback, noise, latent_image, denoise_mask, disable_pbar)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\ComfyUI\comfy\patcher_extension.py", line 113, in execute
return self.original(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\ComfyUI\comfy\samplers.py", line 1005, in sample
samples = self.sampler_function(model_k, noise, sigmas, extra_args=extra_args, callback=k_callback, disable=disable_pbar, **self.extra_options)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\python_embeded\Lib\site-packages\torch\utils_contextlib.py", line 124, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\ComfyUI\comfy\k_diffusion\sampling.py", line 1460, in sample_res_multistep
return res_multistep(model, x, sigmas, extra_args=extra_args, callback=callback, disable=disable, s_noise=s_noise, noise_sampler=noise_sampler, eta=0., cfg_pp=False)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\python_embeded\Lib\site-packages\torch\utils_contextlib.py", line 124, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "H:\ComfyUI_windows_portable_1\ComfyUI\comfy\k_diffusion\sampling.py", line 1422, in res_multistep
if sigma_down == 0 or old_denoised is None:
^^^^^^^^^^^^^^^
torch.AcceleratorError: CUDA error: unknown error
Search for cudaErrorUnknown' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information. CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1 Compile with TORCH_USE_CUDA_DSA` to enable device-side assertions.
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "H:\ComfyUI_windows_portable_1\ComfyUI\execution.py", line 550, in execute
comfy.model_management.reset_cast_buffers()
File "H:\ComfyUI_windows_portable_1\ComfyUI\comfy\model_management.py", line 1413, in reset_cast_buffers
offload_stream.synchronize()
File "H:\ComfyUI_windows_portable_1\python_embeded\Lib\site-packages\torch\cuda\streams.py", line 106, in synchronize
super().synchronize()
torch.AcceleratorError: CUDA error: unknown error
Search for cudaErrorUnknown' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information. CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1 Compile with TORCH_USE_CUDA_DSA` to enable device-side assertions.
[INFO] Prompt executed in 00:13:01
Exception in thread Thread-20 (prompt_worker):
Traceback (most recent call last):
File "threading.py", line 1075, in bootstrap_inner
File "threading.py", line 1012, in run
File "H:\ComfyUI_windows_portable_1\ComfyUI\main.py", line 417, in prompt_worker
comfy.model_management.soft_empty_cache()
File "H:\ComfyUI_windows_portable_1\ComfyUI\comfy\model_management.py", line 2039, in soft_empty_cache
torch.cuda.synchronize()
File "H:\ComfyUI_windows_portable_1\python_embeded\Lib\site-packages\torch\cuda_init.py", line 1108, in synchronize
return torch._C._cuda_synchronize()
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
torch.AcceleratorError: CUDA error: unknown error
Search for cudaErrorUnknown' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information. CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1 Compile with TORCH_USE_CUDA_DSA` to enable device-side assertions.
NB; this only happens with the convrot models
Had this issue myself with my 5090 when using comfyui portable and all sorted when I reinstalled ComfyUI but this time via Stability Matrix and selecting it to use CU130 with Pytorch (the Cuda version most likely saved it as was on 128 before)
- I just changed the ComfyUI's "update_python_dependencies.bat" file in the "ComfyUI_windows_portable\update" folder and changed line2 like this to update everything regarding Torch, CUDA and such.
..\python_embeded\python.exe -s -m pip install --upgrade --pre torch torchvision torchaudio torchsde --extra-index-url https://download.pytorch.org/whl/cu132 --extra-index-url https://download.pytorch.org/whl/nightly/cu132 -r ../ComfyUI/requirements.txt pygit2
Save & run the .bat file
- Additionally i also installed Flash Attention 2 for Windows using the above method "..\python_embeded\python.exe -s -m pip install "flash_attn-2.9.2whatever.whl" as it seems to reduce VRAM a bit i think but I'm still experimenting - https://huggingface.co/ussoewwin/Flash-Attention-2_for_Windows/tree/main (.whl files can be downloaded here)
Then edit "run_nvidia_gpu.bat" and change line1
.\python_embeded\python.exe -s ComfyUI\main.py --windows-standalone-build --use-flash-attention
Save & run the .bat file
- If all went well, your ComfyUI startup console should report something like :
[INFO] pytorch version: 2.14.0.dev20260807+cu132
[INFO] Using Flash Attention