Description
When using Musicgen (HF) Generator, I can generate musics of up to 40 seconds, but if I put more than 50 seconds, it gives an error and comfyui crashes.
When using vram, the vram doesn't go over 80%, the ram doesn't go over 70%, everything seems fine. When using CPU only, it doesn't go over 75%. It uses the same amount of resources if I generate a 10/20/30/40 seconds songs. But of course, my desktop isn't very strong, so it may be the cause, but when I use "one-click-installers-tts-6.0" (audio-webui) I can generate music up to 4 minutes (maybe more, didn't test) using the same "musicgen-small" model.
This is the error I get when using cpu only mode:
C:\actions-runner\_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Indexing.cu:1239: block: [4,0,0], thread: [127,0,0] Assertion srcIndex < srcSelectDimSize failed.
This is the error I get when using VRam:
C:\actions-runner\_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Indexing.cu:1239: block: [3,0,0], thread: [90,0,0] Assertion srcIndex < srcSelectDimSizefailed. C:\actions-runner\_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Indexing.cu:1239: block: [3,0,0], thread: [91,0,0] AssertionsrcIndex < srcSelectDimSizefailed. C:\actions-runner\_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Indexing.cu:1239: block: [3,0,0], thread: [92,0,0] AssertionsrcIndex < srcSelectDimSizefailed. C:\actions-runner\_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Indexing.cu:1239: block: [3,0,0], thread: [93,0,0] AssertionsrcIndex < srcSelectDimSizefailed. C:\actions-runner\_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Indexing.cu:1239: block: [3,0,0], thread: [94,0,0] AssertionsrcIndex < srcSelectDimSizefailed. C:\actions-runner\_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Indexing.cu:1239: block: [3,0,0], thread: [95,0,0] AssertionsrcIndex < srcSelectDimSize` failed.
ERROR:root:!!! Exception during processing !!!
ERROR:root:Traceback (most recent call last):
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\random.py", line 165, in fork_rng
yield
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\custom_nodes\ComfyUI-audio\musicgen_hf_nodes.py", line 124, in generate
samples = m.generate(
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\utils_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\transformers\models\musicgen\modeling_musicgen.py", line 2430, in generate
outputs = self.sample(
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\transformers\generation\utils.py", line 2642, in sample
outputs = self(
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\transformers\models\musicgen\modeling_musicgen.py", line 1916, in forward
decoder_outputs = self.decoder(
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\transformers\models\musicgen\modeling_musicgen.py", line 1029, in forward
outputs = self.model(
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\transformers\models\musicgen\modeling_musicgen.py", line 938, in forward
decoder_outputs = self.decoder(
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\transformers\models\musicgen\modeling_musicgen.py", line 848, in forward
layer_outputs = decoder_layer(
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\transformers\models\musicgen\modeling_musicgen.py", line 400, in forward
hidden_states, self_attn_weights, present_key_value = self.self_attn(
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\transformers\models\musicgen\modeling_musicgen.py", line 250, in forward
value_states = self._shape(self.v_proj(hidden_states), -1, bsz)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\nn\modules\linear.py", line 114, in forward
return F.linear(input, self.weight, self.bias)
RuntimeError: CUDA error: unknown error
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 "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\execution.py", line 152, in recursive_execute
output_data, output_ui = get_output_data(obj, input_data_all)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\execution.py", line 82, in get_output_data
return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\execution.py", line 75, in map_node_over_list
results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\custom_nodes\ComfyUI-audio\musicgen_hf_nodes.py", line 95, in generate
with (
File "contextlib.py", line 153, in exit
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\random.py", line 169, in fork_rng
device_mod.set_rng_state(device_rng_state, device)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\cuda\random.py", line 75, in set_rng_state
lazy_call(cb)
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\cuda_init.py", line 229, in _lazy_call
callable()
File "D:\AI\StabilityMatrix-win-x64\Data\Packages\ComfyUI2\venv\lib\site-packages\torch\cuda\random.py", line 73, in cb
default_generator.set_state(new_state_copy)
RuntimeError: CUDA error: unknown error
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.`
It's the same error.
This is the way I've setup the nodes, it works fine when creating smaller music.
Activity