iceclear / ldm-srtuning Goto Github PK

Your project is very enlightening, but I have a doubt, this code runs in the configuration file environment.yaml you provided, the project does not run successfully, it will show an error like this:
Traceback (most recent call last):
File "D:\17DEM\LDM-SRtuning-main\main.py", line 6, in
import pytorch_lightning as pl
File "D:\Anaconda\envs\diffusion\lib\site-packages\pytorch_lightning_init.py", line 20, in
from pytorch_lightning import metrics # noqa: E402
File "D:\Anaconda\envs\diffusion\lib\site-packages\pytorch_lightning\metrics_init_.py", line 15, in
from pytorch_lightning.metrics.classification import ( # noqa: F401
File "D:\Anaconda\envs\diffusion\lib\site-packages\pytorch_lightning\metrics\classification_init_.py", line 14, in
from pytorch_lightning.metrics.classification.accuracy import Accuracy # noqa: F401
File "D:\Anaconda\envs\diffusion\lib\site-packages\pytorch_lightning\metrics\classification\accuracy.py", line 16, in
from torchmetrics import Accuracy as Accuracy
File "D:\Anaconda\envs\diffusion\lib\site-packages\torchmetrics_init.py", line 14, in
from torchmetrics import functional # noqa: E402
File "D:\Anaconda\envs\diffusion\lib\site-packages\torchmetrics\functional_init_.py", line 14, in
from torchmetrics.functional.audio.deprecated import permutation_invariant_training as permutation_invariant_training
File "D:\Anaconda\envs\diffusion\lib\site-packages\torchmetrics\functional\audio_init.py", line 15, in
from torchmetrics.functional.audio.sdr import (
File "D:\Anaconda\envs\diffusion\lib\site-packages\torchmetrics\functional\audio\sdr.py", line 27, in
solve = torch.linalg.solve
AttributeError: module 'torch.linalg' has no attribute 'solve'

Hello, I am trying to train your model on a single A6000 from scratch with DF2K dataset. First, I have tried your original config and then I have tried batch_size=18 and LR_size=64. Nevertheless, I couldn't train the model. The loss remains at 1 and eventually explodes. Output images is just random rgb noise. Would you be able to help me with this problem?

Thank you

Hello, I got torch size mismatch error when I tried to load the pretrained model (https://huggingface.co/Iceclear/StableSR/resolve/main/ldmsr4x_finetune_119.ckpt) with the default config (configs/bsr_sr/config_sr_finetune.yaml).
Can the authors please check if the provided config parameters are consistent with the pre-trained model? I'm sorry I can't provide a screenshot of the reported error message because of the company's information management policy.
Thank you very much!

Hi, thank you for your repo, its very interesting.

I would like training a new model with my dataset, could you help me to tell me how could I achieve this goal? I guess that I have to change the root of the dataset in train and validation.

Hello, I tried your pretrained model on 256x256 images, but the output image is still 256x256.
How should I modify the parameters to achieve higher resolution output? Thank you.

Hi! I have trained a model and when I try to use it, once time even the sampling start, the next error message appear me:

File "/usr/local/lib/python3.10/dist-packages/torch/functional.py", line 378, in einsum
return _VF.einsum(equation, operands) # type: ignore[attr-defined]
torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 64.00 GiB (GPU 0; 14.75 GiB total capacity; 3.82 GiB already allocated; 10.50 GiB free; 3.94 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

Is it normal that this high GPU memory amount was allocated to my proccess? :(