CUTLASS submodule not found. #2047

sbstratos79 · 2022-10-09T07:21:02Z

Describe the bug
Running webui.sh with --xformers flag gives the error:
RuntimeError: CUTLASS submodule not found. Did you forget to run `git submodule update --init --recursive` ?

To Reproduce
Steps to reproduce the behavior:

Add --xformers flag to webui-user.sh
Run ./webui.sh
See the error during xformers installation.

Expected behavior
xformers should install.

Desktop (please complete the following information):

Linux
Python 3.10.7
050a6a7

Additional context
Before the --xformers flag was added, I was using the new cross attention method in a conda env after manually installing xformers and editing the attention.py in ldm modules. It was working without any issues. I have since uninstalled conda and tried using this new flag with a fresh venv.
Btw, I have tried git submodule update --init --recursive in stable-diffusion folder. It didn't help.

The text was updated successfully, but these errors were encountered:

C43H66N12O12S2 · 2022-10-09T07:49:53Z

Does entering pip install xformers outside venv give the same error?

sbstratos79 · 2022-10-09T08:22:11Z

Does entering pip install xformers outside venv give the same error?

Yes.

C43H66N12O12S2 · 2022-10-09T08:24:35Z

Do pip install -U -I xformers outside venv. If that fails as well, open an issue upstream.

r3nor · 2022-10-09T13:55:16Z

I am facing the same error. Have you found any solution? I tried all the solutions @C43H66N12O12S2 has commented, but none of them worked for me.

mifortin · 2022-10-09T15:40:00Z

Try "pip install xformers==0.0.12"? seems 0.0.13 doesn't like being installed on my PC.

r3nor · 2022-10-09T15:42:33Z

@mifortin Yes, using v0.0.12 did work. v0.0.13 is not working. Thank you.

trufty · 2022-10-09T16:53:21Z

0.0.12 was the solution for me using docker as well.

Also docker users, you have to use the devel Nvidia Cuda container or install nvcc manually for xformers to install correctly.
FROM nvidia/cuda:11.7.x-devel-ubuntu22.04

I chose 22.04 because it comes with python 3.10 by default.

luckyycode · 2022-10-09T17:24:35Z

0.0.12 was the solution for me using docker as well.

Also docker users, you have to use the devel Nvidia Cuda container or install nvcc manually for xformers to install correctly. FROM nvidia/cuda:11.7.x-devel-ubuntu22.04

I chose 22.04 because it comes with python 3.10 by default.

which dockerized sd version are you using? I get "triton" module missing if I use this image with xformers==0.0.12. I tried the same cuda container, installed pip and I get missing modules "triton", "boto" and "psycorg2"

I also tried to install it manually, but I get exactly the same error as this #2073

#xformers
set DEBIAN_FRONTEND=noninteractive
set NVCC_FLAGS="-allow-unsupported-compiler"
set TORCH_CUDA_ARCH_LIST=8.6
apt install -y g++
git clone https://github.com/facebookresearch/xformers.git repositories/xformers
cd repositories/xformers
git submodule update --init --recursive
pip install -r requirements.txt
pip install -e .
pip install cutlass
EOF```

trufty · 2022-10-09T21:07:34Z

@luckyycode build xformers in your dockerfile like this facebookresearch/xformers#473 (comment) and comment out xformers install in launch.py

That was a pain to work around...

slix · 2022-10-09T21:29:11Z

@luckyycode build xformers in your dockerfile like this facebookresearch/xformers#473 (comment) and comment out xformers install in launch.py

That was a pain to work around...

A possible workaround that I'm trying: facebookresearch/xformers#473 (comment)

BuffMcBigHuge · 2022-11-30T19:30:23Z

I was able to solve this by running the following script on Ubuntu:

cd stable-diffusion-webui
python -m venv venv
source venv/bin/activate
cd ..
git clone https://github.com/facebookresearch/xformers/
cd xformers
pip install --verbose --no-deps -e .
pip install -r requirements.txt
pip install functorch==0.2.1 ninja bitsandbytes
pip install -U --pre triton
deactivate
cd ../stable-diffusion-webui
bash webui.sh

TwistedFromArma · 2022-12-08T00:36:31Z

im still having this error. Pain as nohalf option gives ram space problems running sd2.1

manyhats-mike · 2022-12-13T23:34:39Z

Make sure you have 'setuptools' installed first

pip install setuptools

Then you can follow the above steps, but you will also need to initialize the submodules to grab cutlass or it will complain

cd stable-diffusion-webui
python -m venv venv
source venv/bin/activate
cd ..
git clone https://github.com/facebookresearch/xformers/
cd xformers
git submodule update --init --recursive # THIS IS NEEDED TO GET PAST CUTLASS ERROR
pip install --verbose --no-deps -e .
pip install -r requirements.txt
pip install functorch==0.2.1 ninja bitsandbytes
pip install -U --pre triton
deactivate
cd ../stable-diffusion-webui
bash webui.sh

GregHilston · 2022-12-21T05:15:49Z

@manyhats-mike I'm running into the follow error after following your instructions on a Linux host OS, and attempting to generate an image. Any ideas why?

I run the program with $ ./webui.sh --no-half --no-half-vae --medvram --opt-split-attention --xformers

  0%|                                                                                                 | 0/20 [00:00<?, ?it/s]
Error completing request
Arguments: ('An astronaut riding a horse', '', 'None', 'None', 20, 0, False, False, 1, 1, 7, -1.0, -1.0, 0, 0, 0, False, 512, 512, False, 0.7, 0, 0, 0, False, False, False, False, '', 1, '', 0, '', True, False, False) {}
Traceback (most recent call last):
  File "/home/ghilston/Git/stable-diffusion-webui/modules/call_queue.py", line 45, in f
    res = list(func(*args, **kwargs))
  File "/home/ghilston/Git/stable-diffusion-webui/modules/call_queue.py", line 28, in f
    res = func(*args, **kwargs)
  File "/home/ghilston/Git/stable-diffusion-webui/modules/txt2img.py", line 49, in txt2img
    processed = process_images(p)
  File "/home/ghilston/Git/stable-diffusion-webui/modules/processing.py", line 464, in process_images
    res = process_images_inner(p)
  File "/home/ghilston/Git/stable-diffusion-webui/modules/processing.py", line 567, in process_images_inner
    samples_ddim = p.sample(conditioning=c, unconditional_conditioning=uc, seeds=seeds, subseeds=subseeds, subseed_strength=p.subseed_strength, prompts=prompts)
  File "/home/ghilston/Git/stable-diffusion-webui/modules/processing.py", line 699, in sample
    samples = self.sampler.sample(self, x, conditioning, unconditional_conditioning, image_conditioning=self.txt2img_image_conditioning(x))
  File "/home/ghilston/Git/stable-diffusion-webui/modules/sd_samplers.py", line 507, in sample
    samples = self.launch_sampling(steps, lambda: self.func(self.model_wrap_cfg, x, extra_args={
  File "/home/ghilston/Git/stable-diffusion-webui/modules/sd_samplers.py", line 422, in launch_sampling
    return func()
  File "/home/ghilston/Git/stable-diffusion-webui/modules/sd_samplers.py", line 507, in <lambda>
    samples = self.launch_sampling(steps, lambda: self.func(self.model_wrap_cfg, x, extra_args={
  File "/home/ghilston/Git/stable-diffusion-webui/venv/lib/python3.10/site-packages/torch/autograd/grad_mode.py", line 27, in decorate_context
    return func(*args, **kwargs)
  File "/home/ghilston/Git/stable-diffusion-webui/repositories/k-diffusion/k_diffusion/sampling.py", line 145, in sample_euler_ancestral
    denoised = model(x, sigmas[i] * s_in, **extra_args)
  File "/home/ghilston/Git/stable-diffusion-webui/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1194, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/ghilston/Git/stable-diffusion-webui/modules/sd_samplers.py", line 321, in forward
    x_out[a:b] = self.inner_model(x_in[a:b], sigma_in[a:b], cond={"c_crossattn": [cond_in[a:b]], "c_concat": [image_cond_in[a:b]]})
  File "/home/ghilston/Git/stable-diffusion-webui/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1194, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/ghilston/Git/stable-diffusion-webui/repositories/k-diffusion/k_diffusion/external.py", line 167, in forward
    return self.get_v(input * c_in, self.sigma_to_t(sigma), **kwargs) * c_out + input * c_skip
  File "/home/ghilston/Git/stable-diffusion-webui/repositories/k-diffusion/k_diffusion/external.py", line 177, in get_v
    return self.inner_model.apply_model(x, t, cond)
  File "/home/ghilston/Git/stable-diffusion-webui/repositories/stable-diffusion-stability-ai/ldm/models/diffusion/ddpm.py", line 858, in apply_model
    x_recon = self.model(x_noisy, t, **cond)
  File "/home/ghilston/Git/stable-diffusion-webui/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1212, in _call_impl
    result = forward_call(*input, **kwargs)
  File "/home/ghilston/Git/stable-diffusion-webui/repositories/stable-diffusion-stability-ai/ldm/models/diffusion/ddpm.py", line 1329, in forward
    out = self.diffusion_model(x, t, context=cc)
  File "/home/ghilston/Git/stable-diffusion-webui/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1194, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/ghilston/Git/stable-diffusion-webui/repositories/stable-diffusion-stability-ai/ldm/modules/diffusionmodules/openaimodel.py", line 776, in forward
    h = module(h, emb, context)
  File "/home/ghilston/Git/stable-diffusion-webui/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1194, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/ghilston/Git/stable-diffusion-webui/repositories/stable-diffusion-stability-ai/ldm/modules/diffusionmodules/openaimodel.py", line 84, in forward
    x = layer(x, context)
  File "/home/ghilston/Git/stable-diffusion-webui/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1194, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/ghilston/Git/stable-diffusion-webui/repositories/stable-diffusion-stability-ai/ldm/modules/attention.py", line 324, in forward
    x = block(x, context=context[i])
  File "/home/ghilston/Git/stable-diffusion-webui/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1194, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/ghilston/Git/stable-diffusion-webui/modules/sd_hijack_checkpoint.py", line 4, in BasicTransformerBlock_forward
    return checkpoint(self._forward, x, context)
  File "/home/ghilston/Git/stable-diffusion-webui/venv/lib/python3.10/site-packages/torch/utils/checkpoint.py", line 249, in checkpoint
    return CheckpointFunction.apply(function, preserve, *args)
  File "/home/ghilston/Git/stable-diffusion-webui/venv/lib/python3.10/site-packages/torch/utils/checkpoint.py", line 107, in forward
    outputs = run_function(*args)
  File "/home/ghilston/Git/stable-diffusion-webui/repositories/stable-diffusion-stability-ai/ldm/modules/attention.py", line 262, in _forward
    x = self.attn1(self.norm1(x), context=context if self.disable_self_attn else None) + x
  File "/home/ghilston/Git/stable-diffusion-webui/venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1194, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/ghilston/Git/stable-diffusion-webui/modules/sd_hijack_optimizations.py", line 227, in xformers_attention_forward
    out = xformers.ops.memory_efficient_attention(q, k, v, attn_bias=None)
  File "/home/ghilston/Git/xformers/xformers/ops/fmha/__init__.py", line 193, in memory_efficient_attention
    return _memory_efficient_attention(
  File "/home/ghilston/Git/xformers/xformers/ops/fmha/__init__.py", line 289, in _memory_efficient_attention
    return _memory_efficient_attention_forward(
  File "/home/ghilston/Git/xformers/xformers/ops/fmha/__init__.py", line 311, in _memory_efficient_attention_forward
    out, *_ = op.apply(inp, needs_gradient=False)
  File "/home/ghilston/Git/xformers/xformers/ops/fmha/cutlass.py", line 103, in apply
    out, lse = cls.OPERATOR(
  File "/home/ghilston/Git/stable-diffusion-webui/venv/lib/python3.10/site-packages/torch/_ops.py", line 442, in __call__
    return self._op(*args, **kwargs or {})
NotImplementedError: Could not run 'xformers::efficient_attention_forward_cutlass' with arguments from the 'CUDA' backend. This could be because the operator doesn't exist for this backend, or was omitted during the selective/custom build process (if using custom build). If you are a Facebook employee using PyTorch on mobile, please visit https://fburl.com/ptmfixes for possible resolutions. 'xformers::efficient_attention_forward_cutlass' is only available for these backends: [BackendSelect, Python, FuncTorchDynamicLayerBackMode, Functionalize, Named, Conjugate, Negative, ZeroTensor, ADInplaceOrView, AutogradOther, AutogradCPU, AutogradCUDA, AutogradXLA, AutogradMPS, AutogradXPU, AutogradHPU, AutogradLazy, Tracer, AutocastCPU, AutocastCUDA, FuncTorchBatched, FuncTorchVmapMode, Batched, VmapMode, FuncTorchGradWrapper, PythonTLSSnapshot, FuncTorchDynamicLayerFrontMode, PythonDispatcher].

BackendSelect: fallthrough registered at ../aten/src/ATen/core/BackendSelectFallbackKernel.cpp:3 [backend fallback]
Python: registered at ../aten/src/ATen/core/PythonFallbackKernel.cpp:140 [backend fallback]
FuncTorchDynamicLayerBackMode: registered at ../aten/src/ATen/functorch/DynamicLayer.cpp:488 [backend fallback]
Functionalize: registered at ../aten/src/ATen/FunctionalizeFallbackKernel.cpp:291 [backend fallback]
Named: registered at ../aten/src/ATen/core/NamedRegistrations.cpp:7 [backend fallback]
Conjugate: registered at ../aten/src/ATen/ConjugateFallback.cpp:18 [backend fallback]
Negative: registered at ../aten/src/ATen/native/NegateFallback.cpp:18 [backend fallback]
ZeroTensor: registered at ../aten/src/ATen/ZeroTensorFallback.cpp:86 [backend fallback]
ADInplaceOrView: fallthrough registered at ../aten/src/ATen/core/VariableFallbackKernel.cpp:64 [backend fallback]
AutogradOther: fallthrough registered at ../aten/src/ATen/core/VariableFallbackKernel.cpp:35 [backend fallback]
AutogradCPU: fallthrough registered at ../aten/src/ATen/core/VariableFallbackKernel.cpp:39 [backend fallback]
AutogradCUDA: fallthrough registered at ../aten/src/ATen/core/VariableFallbackKernel.cpp:47 [backend fallback]
AutogradXLA: fallthrough registered at ../aten/src/ATen/core/VariableFallbackKernel.cpp:51 [backend fallback]
AutogradMPS: fallthrough registered at ../aten/src/ATen/core/VariableFallbackKernel.cpp:59 [backend fallback]
AutogradXPU: fallthrough registered at ../aten/src/ATen/core/VariableFallbackKernel.cpp:43 [backend fallback]
AutogradHPU: fallthrough registered at ../aten/src/ATen/core/VariableFallbackKernel.cpp:68 [backend fallback]
AutogradLazy: fallthrough registered at ../aten/src/ATen/core/VariableFallbackKernel.cpp:55 [backend fallback]
Tracer: registered at ../torch/csrc/autograd/TraceTypeManual.cpp:296 [backend fallback]
AutocastCPU: fallthrough registered at ../aten/src/ATen/autocast_mode.cpp:482 [backend fallback]
AutocastCUDA: fallthrough registered at ../aten/src/ATen/autocast_mode.cpp:324 [backend fallback]
FuncTorchBatched: registered at ../aten/src/ATen/functorch/LegacyBatchingRegistrations.cpp:743 [backend fallback]
FuncTorchVmapMode: fallthrough registered at ../aten/src/ATen/functorch/VmapModeRegistrations.cpp:28 [backend fallback]
Batched: registered at ../aten/src/ATen/BatchingRegistrations.cpp:1064 [backend fallback]
VmapMode: fallthrough registered at ../aten/src/ATen/VmapModeRegistrations.cpp:33 [backend fallback]
FuncTorchGradWrapper: registered at ../aten/src/ATen/functorch/TensorWrapper.cpp:189 [backend fallback]
PythonTLSSnapshot: registered at ../aten/src/ATen/core/PythonFallbackKernel.cpp:148 [backend fallback]
FuncTorchDynamicLayerFrontMode: registered at ../aten/src/ATen/functorch/DynamicLayer.cpp:484 [backend fallback]
PythonDispatcher: registered at ../aten/src/ATen/core/PythonFallbackKernel.cpp:144 [backend fallback]

If I remove the --xformers argument , and uninstall the xformers package, I can successfully generate images

AlexNolasco · 2022-12-23T20:51:52Z

Problem still persists Ubuntu 22.04.1 LTS

Lesani · 2022-12-28T21:20:22Z

~~problem persists in a fresh ubuntu 22.10 install as well, manually installing the 0.0.12 version worked, 0.0.13 does not want to install, going on about "CUTLASS submodule not found"~~

I stand corrected, only the installation worked, xformers itself did not, even after installing triton as it requests upon startup...

I don't know what to do, xformers worked great on windows, I switched to ubuntu because even on my 24gb vram gpu I regularly ran out when Dreambooth requestsd like 16 or so, because other stuff was reserving so much I guess... but installing A1111 ubuntu is so much more non-automatic.....

AlexNolasco · 2022-12-28T21:51:40Z

Docker may be an alternative
https://github.com/AbdBarho/stable-diffusion-webui-docker

But other things may fail, e.g. CLIP interrogator.

~~problem persists in a fresh ubuntu 22.10 install as well, manually installing the 0.0.12 version worked, 0.0.13 does not want to install, going on about "CUTLASS submodule not found"~~

I stand corrected, only the installation worked, xformers itself did not, even after installing triton as it requests upon startup...

I don't know what to do, xformers worked great on windows, I switched to ubuntu because even on my 24gb vram gpu I regularly ran out when Dreambooth requestsd like 16 or so, because other stuff was reserving so much I guess... but installing A1111 ubuntu is so much more non-automatic.....

GregHilston · 2022-12-29T01:24:33Z

I'm still unable to resolve this.

Unrelated to this thread, but @AlexNolasco , which repo have you been using to run Dreambooth?

elephantpanda · 2023-01-30T22:00:56Z

I'm running:
pip install xformers==0.0.12
and I'm getting the error:
running build_ext error: [WinError 2] The system cannot find the file specified
I am running Python3.8

nancy6o6 · 2023-03-13T08:54:09Z

I have the same error when running"pip install xformers" on python 3.9.12.
I figure it by downgrading python to 3.8.5 and run "python -m pip install --upgrade pip" and "python -m pip install --upgrade setuptools" to reload the version matching python3.8.5. Then I can install xformers 0.0.13 using pip install.

AlexNolasco · 2023-05-18T17:20:04Z

I'm still unable to resolve this.t
I simply reinstalled and it has been working since.

Unrelated to this thread, but @AlexNolasco , which repo have you been using to run Dreambooth?
Not running Dreambooth

update dev

haha889 · 2024-05-14T06:35:55Z

Can xformers only be installed on Linux systems?

sbstratos79 added the bug-report Report of a bug, yet to be confirmed label Oct 9, 2022

C43H66N12O12S2 closed this as completed Oct 9, 2022

AyoKeito mentioned this issue Jan 10, 2023

[Bug]: Tried my best to get this working on Linux. #6557

Closed

1 task

Symbiomatrix pushed a commit to Symbiomatrix/stable-diffusion-webui that referenced this issue Sep 17, 2023

Merge pull request AUTOMATIC1111#2047 from vladmandic/master

4826197

update dev

konradre mentioned this issue Dec 14, 2023

xformers install error ali-vilab/VGen#9

Open

SRagy mentioned this issue Jul 7, 2024

Unable to install Dependencies on Paperspace TheLastBen/fast-stable-diffusion#2868

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

CUTLASS submodule not found. #2047

CUTLASS submodule not found. #2047

sbstratos79 commented Oct 9, 2022 •

edited

Loading

C43H66N12O12S2 commented Oct 9, 2022

sbstratos79 commented Oct 9, 2022 •

edited

Loading

C43H66N12O12S2 commented Oct 9, 2022

r3nor commented Oct 9, 2022 •

edited

Loading

mifortin commented Oct 9, 2022

r3nor commented Oct 9, 2022

trufty commented Oct 9, 2022

luckyycode commented Oct 9, 2022 •

edited

Loading

trufty commented Oct 9, 2022 •

edited

Loading

slix commented Oct 9, 2022

BuffMcBigHuge commented Nov 30, 2022

TwistedFromArma commented Dec 8, 2022

manyhats-mike commented Dec 13, 2022 •

edited

Loading

GregHilston commented Dec 21, 2022 •

edited

Loading

AlexNolasco commented Dec 23, 2022

Lesani commented Dec 28, 2022 •

edited

Loading

AlexNolasco commented Dec 28, 2022

GregHilston commented Dec 29, 2022

elephantpanda commented Jan 30, 2023

nancy6o6 commented Mar 13, 2023

AlexNolasco commented May 18, 2023

haha889 commented May 14, 2024

CUTLASS submodule not found. #2047

CUTLASS submodule not found. #2047

Comments

sbstratos79 commented Oct 9, 2022 • edited Loading

C43H66N12O12S2 commented Oct 9, 2022

sbstratos79 commented Oct 9, 2022 • edited Loading

C43H66N12O12S2 commented Oct 9, 2022

r3nor commented Oct 9, 2022 • edited Loading

mifortin commented Oct 9, 2022

r3nor commented Oct 9, 2022

trufty commented Oct 9, 2022

luckyycode commented Oct 9, 2022 • edited Loading

trufty commented Oct 9, 2022 • edited Loading

slix commented Oct 9, 2022

BuffMcBigHuge commented Nov 30, 2022

TwistedFromArma commented Dec 8, 2022

manyhats-mike commented Dec 13, 2022 • edited Loading

GregHilston commented Dec 21, 2022 • edited Loading

AlexNolasco commented Dec 23, 2022

Lesani commented Dec 28, 2022 • edited Loading

AlexNolasco commented Dec 28, 2022

GregHilston commented Dec 29, 2022

elephantpanda commented Jan 30, 2023

nancy6o6 commented Mar 13, 2023

AlexNolasco commented May 18, 2023

haha889 commented May 14, 2024

sbstratos79 commented Oct 9, 2022 •

edited

Loading

sbstratos79 commented Oct 9, 2022 •

edited

Loading

r3nor commented Oct 9, 2022 •

edited

Loading

luckyycode commented Oct 9, 2022 •

edited

Loading

trufty commented Oct 9, 2022 •

edited

Loading

manyhats-mike commented Dec 13, 2022 •

edited

Loading

GregHilston commented Dec 21, 2022 •

edited

Loading

Lesani commented Dec 28, 2022 •

edited

Loading