Is RTX 4090 not supported yet?

[YukihimeX]

Is it just me or is anyone able to get RTX 4090 to run Dreambooth SD in Windows WSL2? If you can, please do share, would really appreciate it.

Everytime I run the training cell, it always ended with

_`Traceback (most recent call last):
File "/mnt/d/linux/Dreambooth-Stable-Diffusion/main.py", line 683, in
model = load_model_from_config(config, opt.actual_resume)
File "/mnt/d/linux/Dreambooth-Stable-Diffusion/main.py", line 42, in load_model_from_config
model.cuda()
File "/home/aubuntu/miniconda3/lib/python3.9/site-packages/pytorch_lightning/core/mixins/device_dtype_mixin.py", line 138, in cuda
return super().cuda(device=device)
File "/home/aubuntu/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 689, in cuda
return self._apply(lambda t: t.cuda(device))
File "/home/aubuntu/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 579, in _apply
module._apply(fn)
File "/home/aubuntu/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 579, in _apply
module._apply(fn)
File "/home/aubuntu/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 579, in _apply
module._apply(fn)
[Previous line repeated 1 more time]
File "/home/aubuntu/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 602, in _apply
param_applied = fn(param)
File "/home/aubuntu/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 689, in
return self._apply(lambda t: t.cuda(device))
File "/home/aubuntu/miniconda3/lib/python3.9/site-packages/torch/cuda/init.py", line 217, in _lazy_init
torch._C._cuda_init()
RuntimeError: No CUDA GPUs are available

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "/mnt/d/linux/Dreambooth-Stable-Diffusion/main.py", line 896, in
if trainer.global_rank == 0:
NameError: name 'trainer' is not defined`_

Is it because my GPU is too new and is not supported by the old CUDA version? Or is it just an WSL2 problem?

=============================
Here's my driver info

`Thu Oct 20 01:59:18 2022
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 520.61.05 Driver Version: 522.30 CUDA Version: 11.8 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 NVIDIA GeForce ... On | 00000000:0B:00.0 On | Off |
| 0% 34C P8 10W / 450W | 687MiB / 24564MiB | 1% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| No running processes found |
+-----------------------------------------------------------------------------+`

[AOCGmutong]

Similar problem here, 4090 just can't run training. It even black out and crashed.

[YukihimeX]

Similar problem here, 4090 just can't run training. It even black out and crashed.

This repo can only work on Linux (I tried running it using WSL2 on Win11 but still didn't work)
But everything worked out fine on Windows 10/11 using the Dreambooth-SD-optimized Repo.