Open ZeroNyte opened 6 months ago
Looks like your graphics card doesn't get recognized. Are you sure that torch is correctly installed?
As far as I know it should be, went into venv, activated and installed it there. As it should be right?
What command did you use to install it?
oh wait I installed torch-directml, but not torch itself, shouldve done that on first startup/run right? or did I miss anything in that regard?
No, torch-directml is correct. Can you drop the skip-cuda argument and see what error it gives you now?
RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check
Try deleting the venv folder, create a new one, go into the virtual enviroment and reinstall torch with >pip install torch-directml
do I create it with running webui-user.bat or just via command line?
python
Commandline is fine. Btw, you might wanna update python as well.
so made a new venv, and get: RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check. again
also when I add it back, and try something this pops up in the terminal: AttributeError: 'NoneType' object has no attribute 'is_sdxl' and just shows in the prompt field "?/?"
try switching rng from gpu to cpu
it already was
and now it works for some reason, didnt change anything in the last hour... but I'll try again tomorrow, see if it acts up then.
ah... nvm.. had the wrong UI folder opened.. guess that's what you get for having more than 1 UI installed..
Wait a sec.... Automatic1111's SD WebUI points to a different fork for AMD on Windows, wich is lshqqytiger's, with DirectML support: https://github.com/lshqqytiger/stable-diffusion-webui-directml But this project is forked from the default WebUI, not lshqqytiger's fork. It's probably missing all the DirectML stuff. ....or better: there's indeed some DirectML stuff inside Forge, but there's also some reports saying it doesn't work properly https://github.com/lllyasviel/stable-diffusion-webui-forge/issues/58 https://github.com/lllyasviel/stable-diffusion-webui-forge/issues/570
So, your best bet until full ROCm is released on Windows, is either using lshqqytiger's fork or Linux
So been trying a couple different UIs, and A1111 directml with lobe extension, had been working for the most part, sometimes it stops eith certain prompts, but still currently running 832x1216 without much issue except that it's slow, takes about 10 mins for that res with 25 steps
Still can't get Forge to work, this system has ishqqytiger's fork working.
Forge:
Running with --directml --skip-cuda-test
, then manually activate venv
to pip install torch_directml
, get this:
Using directml with device:
Total VRAM 1024 MB, total RAM 32688 MB
Trying to enable lowvram mode because your GPU seems to have 4GB or less. If you don't want this use: --always-normal-vram
Set vram state to: LOW_VRAM
Device: privateuseone
VAE dtype: torch.float32
CUDA Stream Activated: False
Warning: caught exception 'Torch not compiled with CUDA enabled', memory monitor disabled
Using sub quadratic optimization for cross attention, if you have memory or speed issues try using: --attention-split
==============================================================================
You are running torch 2.0.0+cpu.
The program is tested to work with torch 2.1.2.
To reinstall the desired version, run with commandline flag Using directml with device:
Total VRAM 1024 MB, total RAM 32688 MB
Trying to enable lowvram mode because your GPU seems to have 4GB or less. If you don't want this use: --always-normal-vram
Set vram state to: LOW_VRAM
Device: privateuseone
VAE dtype: torch.float32
CUDA Stream Activated: False
Warning: caught exception 'Torch not compiled with CUDA enabled', memory monitor disabled
Using sub quadratic optimization for cross attention, if you have memory or speed issues try using: --attention-split
==============================================================================
You are running torch 2.0.0+cpu.
The program is tested to work with torch 2.1.2.
To reinstall the desired version, run with commandline flag --reinstall-torch.
Beware that this will cause a lot of large files to be downloaded, as well as
there are reports of issues with training tab on the latest version.
Use --skip-version-check commandline argument to disable this check..
Beware that this will cause a lot of large files to be downloaded, as well as
there are reports of issues with training tab on the latest version.
Use --skip-version-check commandline argument to disable this check.
I run again with --reinstall-torch
, get this:
ImportError: DLL load failed while importing torch_directml_native: The specified procedure could not be found.
Press any key to continue . . .
So I go back and manually activate venv
to pip install torch_directml
again, ok, browser UI appears, back to getting the same CPU indication as above:
Using directml with device:
Total VRAM 1024 MB, total RAM 32688 MB
Trying to enable lowvram mode because your GPU seems to have 4GB or less. If you don't want this use: --always-normal-vram
Set vram state to: LOW_VRAM
Device: privateuseone
VAE dtype: torch.float32
CUDA Stream Activated: False
Warning: caught exception 'Torch not compiled with CUDA enabled', memory monitor disabled
Using sub quadratic optimization for cross attention, if you have memory or speed issues try using: --attention-split
==============================================================================
You are running torch 2.0.0+cpu.
The program is tested to work with torch 2.1.2.
To reinstall the desired version, run with commandline flag --reinstall-torch.
Beware that this will cause a lot of large files to be downloaded, as well as
there are reports of issues with training tab on the latest version.
So i try to generate any random thing in text to image, without changing any settings:
Using split attention in VAE
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
Using split attention in VAE
extra {'cond_stage_model.clip_l.logit_scale', 'cond_stage_model.clip_l.text_projection'}
To load target model SD1ClipModel
Begin to load 1 model
Moving model(s) has taken 0.00 seconds
Model loaded in 3.2s (load weights from disk: 0.4s, forge load real models: 2.1s, calculate empty prompt: 0.6s).
Traceback (most recent call last):
File "E:\stable-diffusion-webui-forge\modules_forge\main_thread.py", line 37, in loop
task.work()
File "E:\stable-diffusion-webui-forge\modules_forge\main_thread.py", line 26, in work
self.result = self.func(*self.args, **self.kwargs)
File "E:\stable-diffusion-webui-forge\modules\txt2img.py", line 111, in txt2img_function
processed = processing.process_images(p)
File "E:\stable-diffusion-webui-forge\modules\processing.py", line 752, in process_images
res = process_images_inner(p)
File "E:\stable-diffusion-webui-forge\modules\processing.py", line 848, in process_images_inner
p.rng = rng.ImageRNG((opt_C, p.height // opt_f, p.width // opt_f), p.seeds, subseeds=p.subseeds, subseed_strength=p.subseed_strength, seed_resize_from_h=p.seed_resize_from_h, seed_resize_from_w=p.seed_resize_from_w)
File "E:\stable-diffusion-webui-forge\modules\rng.py", line 114, in __init__
self.generators = [create_generator(seed) for seed in seeds]
File "E:\stable-diffusion-webui-forge\modules\rng.py", line 114, in <listcomp>
self.generators = [create_generator(seed) for seed in seeds]
File "E:\stable-diffusion-webui-forge\modules\rng.py", line 86, in create_generator
generator = torch.Generator(device).manual_seed(int(seed))
RuntimeError: Device type privateuseone is not supported for torch.Generator() api.
Device type privateuseone is not supported for torch.Generator() api.
*** Error completing request
*** Arguments: ('task(vmmt19p6bj4fhn2)', <gradio.routes.Request object at 0x0000024889B43FD0>, 'carbon nanoparticles', '', [], 20, 'DPM++ 2M Karras', 1, 1, 7, 512, 512, False, 0.7, 2, 'Latent', 0, 0, 0, 'Use same checkpoint', 'Use same sampler', '', '', [], 0, False, '', 0.8, -1, False, -1, 0, 0, 0, ControlNetUnit(input_mode=<InputMode.SIMPLE: 'simple'>, use_preview_as_input=False, batch_image_dir='', batch_mask_dir='', batch_input_gallery=[], batch_mask_gallery=[], generated_image=None, mask_image=None, hr_option='Both', enabled=False, module='None', model='None', weight=1, image=None, resize_mode='Crop and Resize', processor_res=-1, threshold_a=-1, threshold_b=-1, guidance_start=0, guidance_end=1, pixel_perfect=False, control_mode='Balanced', save_detected_map=True), ControlNetUnit(input_mode=<InputMode.SIMPLE: 'simple'>, use_preview_as_input=False, batch_image_dir='', batch_mask_dir='', batch_input_gallery=[], batch_mask_gallery=[], generated_image=None, mask_image=None, hr_option='Both', enabled=False, module='None', model='None', weight=1, image=None, resize_mode='Crop and Resize', processor_res=-1, threshold_a=-1, threshold_b=-1, guidance_start=0, guidance_end=1, pixel_perfect=False, control_mode='Balanced', save_detected_map=True), ControlNetUnit(input_mode=<InputMode.SIMPLE: 'simple'>, use_preview_as_input=False, batch_image_dir='', batch_mask_dir='', batch_input_gallery=[], batch_mask_gallery=[], generated_image=None, mask_image=None, hr_option='Both', enabled=False, module='None', model='None', weight=1, image=None, resize_mode='Crop and Resize', processor_res=-1, threshold_a=-1, threshold_b=-1, guidance_start=0, guidance_end=1, pixel_perfect=False, control_mode='Balanced', save_detected_map=True), False, 7, 1, 'Constant', 0, 'Constant', 0, 1, 'enable', 'MEAN', 'AD', 1, False, 1.01, 1.02, 0.99, 0.95, False, 0.5, 2, False, 256, 2, 0, False, False, 3, 2, 0, 0.35, True, 'bicubic', 'bicubic', False, 0, 'anisotropic', 0, 'reinhard', 100, 0, 'subtract', 0, 0, 'gaussian', 'add', 0, 100, 127, 0, 'hard_clamp', 5, 0, 'None', 'None', False, 'MultiDiffusion', 768, 768, 64, 4, False, False, False, False, False, 'positive', 'comma', 0, False, False, 'start', '', 1, '', [], 0, '', [], 0, '', [], True, False, False, False, False, False, False, 0, False) {}
Traceback (most recent call last):
File "E:\stable-diffusion-webui-forge\modules\call_queue.py", line 57, in f
res = list(func(*args, **kwargs))
TypeError: 'NoneType' object is not iterable
---
Checklist
What happened?
Trying to run forge with my 6750 XT, to no avail, tried:
-COMMANDLINE_ARGS= --directml --skip-torch-cuda-test --always-normal-vram --skip-version-check -comment out all the @torch.inference_mode() (add # before them) on: \ldm_patched\modules\utils.py - line 407 \modules_forge\forge_loader.py - line 236, line 242 change "with torch.inference_mode():" for "with torch.no_grad():" on: \modules\processing.py - line 817
CPU and GPU seed gen, restarted my pc, nothing. only using a single SDXL checkpoint.
the only thing I get as a result is: typeerror: 'nonetype' object is not iterable
Steps to reproduce the problem
AMD gpu..
What should have happened?
should work from what I have seen on here
What browsers do you use to access the UI ?
Mozilla Firefox
Sysinfo
sysinfo-2024-03-11-18-07.json
Console logs
Additional information
No response