bmaltais / kohya_ss

Repository from Github https://github.combmaltais/kohya_ss

bmaltais/kohya_ss Issues

Training stopped by File "<frozen runpy>", line 198, in _run_module_as_main
Updated 8 days ago
RuntimeError: The size of tensor a (113) must match the size of tensor b (77) at non-singleton dimension 1
Updated 9 days ago
Fresh Install: Failed to install: torch-2.7.0+cu128-cp311-cp311-win_amd64.whl (torch==2.7.0+cu128)
Closed 9 days ago3
DEPRECATION messages... can you update to get rif of?
Updated 10 days ago
Huge difference in Lora training quality VS version 24.x
Updated 11 days ago
Memory leak when training on MLX!
Updated 11 days ago
[Help] "No data found" error when training model – train_data_dir issue
Updated 15 days ago4
fresh install and errors
Updated 16 days ago5
How to Train Flux using this Gui
Updated 19 days ago
Feature Request: Support multiple GPUs with IPEX
Updated 21 days ago1
Add supportion to f1 krea, qwen-image, etc.
Updated 25 days ago1
NVIDIA 50 series graphics card training VRAM usage is seriously too high
Updated a month ago2
RuntimeError: Expected all tensors , but found at least two devices, cuda:0 and cpu! Lora training not possible
Closed 2 months ago3
An error occurred during training with the RTX 5080
Closed a month ago1
Where in v.25.2.1 - Scale Weight norms, Network dropout, Rank dropout, Module dropout?
Updated a month ago
TE keys missing
Updated a month ago
Add the "qinglong" option to timestep sampling.
Updated a month ago1
please update requirements.txt with latest version of prodigy-plus-schedule-free
Updated a month ago
Resume from saved training state (path to "last-state" state folder) - FAILS
Closed a month ago1
[Bug] GUI crashes with xcb error when opening a file/folder dialog on Ubuntu
Updated a month ago
Questions about continuing training
Updated a month ago1
unable to start learning LoRA on windows
Updated a month ago6
The problem with training speed
Updated a month ago2
Kohya_SS is making JSON instead of SAFETENSORS
Closed 2 months ago3
API not responding on Kohya GUI (RunPod)
Updated 2 months ago
Cuda out of memory in Kohya_ss - what to do
Updated 2 months ago8
Nvidia RTX 5060 Ti, Linux Mint 22.1 CUDA error: no kernel image is available for execution
Updated 2 months ago3
Feature Request - lumina Lora training
Updated 2 months ago1
Kohya_ss on RunPod: Training Ends Immediately Despite Proper Dataset Setup
Closed 2 months ago4
accelerator device: cpu
Closed 2 months ago1
Weird kind of memory leakage? With each lora trained on one instance of the GUI it overfits more.
Updated 2 months ago1
can't train
Updated 2 months ago1
BLIP and BLIP2 captioning not working on XPU
Updated 2 months ago
kohya_ss expects a lower oneAPI version
Updated 2 months ago7
UserWarning: None of the inputs have requires_grad=True. Gradients will be None warnings.warn(
Updated 2 months ago
runpod_setup.md not exist
Closed 2 months ago
Training SDXL LoRA Fails v25.2.1
Updated 2 months ago5
Train a lora without metadata
Updated 2 months ago
TYpeError: cosineAnnalingLR - requires positional argumument: T_max
Closed 2 months ago1
Upgrading the PyTorch Version
Updated 2 months ago1
1 Star
Updated 2 months ago3
linux install problem? ./gui-uv.sh
Closed 2 months ago2
the problem is with the installation, I downloaded all the requirements, but it still won't install. ERROR Failed to install requirements from requirements_pytorch_windows.txt.
Closed 2 months ago16
Install via StabilityMatrix, Ubuntu 24.04, USB-Harddrive ext4
Updated 2 months ago
Training on an RTX 5090 is painfully slow, and I’m unable to configure the "(Optional) Manually configure Accelerate" option.
Closed 3 months ago7
The script crashes abnormally; it works fine after commenting out import tensorflow
Updated 2 months ago8
Prerequisite?!
Updated 2 months ago1
Localization/Language problems (Blank page)
Closed 2 months ago4
大神能不能更新训练器支持最新的FLUX KONTEXT的LORA训练
Updated 3 months ago
LoRA_paisy_09
Closed 3 months ago