bmaltais/kohya_ss issues and pull requests

#2774 - [SD3-Flux.1] "train_blocks=single" not work

Issue - State: open - Opened by joneschunghk 2 months ago - 1 comment

#2773 - Can someone explain in detail how to get the flux training working?

Issue - State: open - Opened by PheonixAi420 2 months ago - 31 comments

#2772 - Attempting to train a flux lora: /workspace/kohya_ss-sd3-flux.1/venv/bin/python: can't open file '/workspace/kohya_ss-sd3-flux.1/sd-scripts/flux_train.py': [Errno 2] No such file or directory

Issue - State: closed - Opened by jygkh 2 months ago

#2771 - Flux LoRA resumes wrong epochs

Issue - State: open - Opened by joneschunghk 2 months ago - 1 comment

#2769 - flux training not working

Issue - State: closed - Opened by elen07zz 2 months ago

#2768 - latest flux branch commit not working for me [CUDA_PATH is set but CUDA wasnt able to be loaded]

Issue - State: open - Opened by elen07zz 2 months ago - 1 comment

#2766 - [MINOR] Glitch in latent caching stdout output

Issue - State: open - Opened by mbac 2 months ago - 1 comment

#2765 - Bump crate-ci/typos from 1.23.6 to 1.24.3

Pull Request - State: closed - Opened by dependabot[bot] 2 months ago
Labels: dependencies, github_actions

#2764 - multires training with toml seems to learn nothing

Issue - State: open - Opened by DarkViewAI 2 months ago - 1 comment

#2763 - SD3.Flux branch, fresh install: RuntimeError: use_libuv was requested but PyTorch was build without libuv support

Issue - State: closed - Opened by TinyForge 2 months ago - 2 comments

#2762 - Help Question - Permission set

Issue - State: closed - Opened by AsaTyr2018 2 months ago - 1 comment

#2761 - How include all folders for meta_cap.sjon and meta_lat.json

Issue - State: closed - Opened by BenDes21 2 months ago

#2760 - meta-lat.json and meta_cap.json for full finetuning mandatory ?

Issue - State: closed - Opened by BenDes21 2 months ago - 1 comment

#2757 - Convolution Rank (Dimension) and Convolutoion Alpha for Flux disappear

Issue - State: closed - Opened by DarkViewAI 3 months ago - 24 comments

#2755 - Finetuning No metadata error

Issue - State: open - Opened by hypervoxel 3 months ago - 8 comments

#2747 - Installing it on RunPod drives me crazy...

Issue - State: open - Opened by dkampien 3 months ago - 4 comments

#2745 - Flux.dev - BF16 is set instead of FP16

Issue - State: open - Opened by Samael-1976 3 months ago - 3 comments

#2743 - Which flux model, UNet? or integrated model?

Issue - State: open - Opened by carat-keeeehun 3 months ago - 2 comments

#2738 - Flux CLIP_L Text Encoder training support.

Issue - State: open - Opened by dsienra 3 months ago - 5 comments

#2735 - FLUX.1 Finetuning

Issue - State: open - Opened by CognitiveDiffusion 3 months ago - 19 comments

#2733 - RuntimeError: "cat_cuda" not implemented for 'Float8_e4m3fn'

Issue - State: open - Opened by axelblaze88 3 months ago - 2 comments

#2730 - subprocess.CalledProcessError: Command '['E:\\kohya_ss\\venv\\Scripts\\python.exe', 'E:/kohya_ss/sd-scripts/flux_train_network.py', '--config_file', './test/output/config_lora-20240822-182849.toml']' returned non-zero exit status 3221225477.

Issue - State: open - Opened by CNEA-lw 3 months ago - 13 comments

#2720 - RuntimeError: Expected is_sm80 || is_sm90 to be true, but got false. (Could this error message be improved? If so, please report an enhancement request to PyTorch.)

Issue - State: open - Opened by oracle9i88 3 months ago - 4 comments

#2717 - flux training on a 2080ti failed

Issue - State: open - Opened by chenxluo 3 months ago - 6 comments

#2714 - please add lora merge for flux for the gui, sd scripts has it

Issue - State: open - Opened by DarkViewAI 3 months ago - 3 comments

#2706 - SD1.5 Training Error

Issue - State: closed - Opened by A1yCE005 3 months ago - 5 comments

#2701 - Flux.1 LoRA training

Issue - State: open - Opened by bmaltais 3 months ago - 472 comments

#2674 - SDXL Training Error - RuntimeError: NaN detected in latents

Issue - State: open - Opened by PENGUINADELIE 4 months ago - 8 comments

#2668 - Security fix

Pull Request - State: closed - Opened by bmaltais 4 months ago

#2637 - any way to disable scheduled huber loss

Issue - State: open - Opened by DarkViewAI 4 months ago - 5 comments

#2607 - Improve "print command" output format

Pull Request - State: closed - Opened by bmaltais 5 months ago

#2500 - Could not load torch: No module named 'torch.amp'

Issue - State: open - Opened by Cookie4Free 6 months ago - 5 comments

#2427 - Bump crate-ci/typos from 1.19.0 to 1.21.0

Pull Request - State: closed - Opened by dependabot[bot] 7 months ago
Labels: dependencies, github_actions

#2396 - Can't train a lora for SDXL from a fresh install.

Issue - State: closed - Opened by Deejay85 7 months ago - 23 comments

#2208 - Extracting Lycoris For SDXL Issue with Text Encoder 2

Issue - State: open - Opened by DarkViewAI 7 months ago - 3 comments
Labels: Troubleshooting, LyCORIS

#2197 - How to provide Config.json AND Dataset Preparation folders and configuration via CMD?

Issue - State: closed - Opened by bjspi 8 months ago - 14 comments
Labels: enhancement

#2185 - M3 MAX 40 core GPU , 128 GB RAM

Issue - State: open - Opened by dajanaelez 8 months ago - 8 comments
Labels: MacOS

#2115 - Could not find cuda drivers on your machine, GPU will not be used

Issue - State: open - Opened by x-future 8 months ago - 14 comments
Labels: Troubleshooting

#1855 - Not installing all dependencies on Linux and another issues

Issue - State: closed - Opened by WizardlyBump17 10 months ago - 4 comments

#1484 - It's working. Training LORA of the latest version of kohya_ss on AMD GPU,Ubuntu 22.04.2 LTS ,test on RX6800 ,sd1.5&sdxl

Issue - State: closed - Opened by tornado73 about 1 year ago - 38 comments

#1248 - finally i got running and training LoRA on M1 mac

Issue - State: closed - Opened by organicmind-pixel over 1 year ago - 43 comments

#927 - Lora training not reading caption files

Issue - State: closed - Opened by Byron-Zoomry over 1 year ago - 4 comments

#664 - ValueError:no metadata F:/Aimage/trainjason/meta_lat.json

Issue - State: closed - Opened by yohancehu over 1 year ago - 13 comments

#120 - consuming LoRa finetuning with an API?

Issue - State: closed - Opened by barshag almost 2 years ago - 1 comment

#118 - v20.7.0

Pull Request - State: closed - Opened by bmaltais almost 2 years ago

#116 - This Scripts is Using too much RAM than Original Kohya script

Issue - State: closed - Opened by vsatyamesc almost 2 years ago - 2 comments

#115 - Why do I need an internet connection to train LoRA models?

Issue - State: closed - Opened by RandomInternetPreson almost 2 years ago - 1 comment

#113 - how to config the multiple GPUs?

Issue - State: closed - Opened by kemp-zz almost 2 years ago - 1 comment

#110 - accelerate : The term 'accelerate' is not recognized as the name of a cmdlet, function, script file, or operable program.

Issue - State: closed - Opened by Zetra-foilder almost 2 years ago - 2 comments

#108 - CUDA out of memory

Issue - State: closed - Opened by godless1893 almost 2 years ago - 3 comments

#107 - Add support for LoRA resizing

Pull Request - State: closed - Opened by bmaltais almost 2 years ago

#105 - v20.6.0

Pull Request - State: closed - Opened by bmaltais almost 2 years ago

#102 - ⚠D:\ai\tool\bitsandbytes\csrc\ops.cu your repo is not compatible with 1080 TI GPU.

Issue - State: closed - Opened by ZeroCool22 almost 2 years ago - 11 comments

#101 - returned non-zero exit status 1

Issue - State: closed - Opened by ZeroCool22 almost 2 years ago - 13 comments

#99 - Error no kernel image is available for execution on the device

Issue - State: closed - Opened by developer90210brr almost 2 years ago - 4 comments

#98 - Error(s) in loading state_dict for UNet2DConditionMode

Issue - State: closed - Opened by RicGdlc almost 2 years ago - 1 comment

#96 - Error while training LORA, CUDA not working

Issue - State: closed - Opened by Odawgthat almost 2 years ago - 2 comments

#94 - How do I run this while offline?

Issue - State: closed - Opened by RandomInternetPreson almost 2 years ago - 2 comments

#93 - error training in lora

Issue - State: closed - Opened by kotaxyz almost 2 years ago - 19 comments

#92 - Could not find a version that satisfies the requirement

Issue - State: closed - Opened by DivinoAG almost 2 years ago - 8 comments

#91 - Error during installation, missing venv folder causing error

Issue - State: closed - Opened by Odawgthat almost 2 years ago - 1 comment

#90 - Error no kernel image is available

Issue - State: closed - Opened by BobbtheBuilder almost 2 years ago - 4 comments

#89 - Error when trying to train a LoRA

Issue - State: closed - Opened by NoneIQalt almost 2 years ago - 13 comments

#88 - ERROR: xformers-some numbers is not a supported wheel on this platform

Issue - State: closed - Opened by Wutislife0022 almost 2 years ago - 7 comments

#87 - Venv issues installing?

Issue - State: closed - Opened by Mousewrites almost 2 years ago - 3 comments

#84 - Request: Can we get this on Colab?

Issue - State: closed - Opened by DarkAlchy almost 2 years ago - 7 comments

#83 - Add new features from sd-script

Pull Request - State: closed - Opened by bmaltais almost 2 years ago

#82 - "MemoryError" (not CUDA) crashes training (dreambooth)

Issue - State: closed - Opened by official-elinas almost 2 years ago - 3 comments

#81 - Unclear instructions for installing CUDNN & no indication of failure

Issue - State: closed - Opened by CodeZombie almost 2 years ago - 1 comment

#79 - AssertionError: resolution is required / resolution（解像度）指定は必須です

Issue - State: closed - Opened by StableDiffusionHub almost 2 years ago - 3 comments

#78 - Dev

Pull Request - State: closed - Opened by bmaltais almost 2 years ago

#76 - TI errors.

Issue - State: closed - Opened by DarkAlchy almost 2 years ago - 2 comments

#74 - Add Dreambooth TI embedding training

Pull Request - State: closed - Opened by bmaltais almost 2 years ago

#73 - Linux Deploy Guide?

Issue - State: closed - Opened by narugo1992 almost 2 years ago - 11 comments

#72 - permission denied when extracting LORA

Issue - State: closed - Opened by maddd13553 almost 2 years ago - 1 comment

#71 - JSON config load does not load the "save trained model as" setting properly

Issue - State: closed - Opened by alchemzy almost 2 years ago - 2 comments

#70 - Please add --keep_tokens for lora training

Issue - State: closed - Opened by alchemzy almost 2 years ago - 4 comments

#69 - "slow_conv2d_cpu" not implemented for 'Half' AND returned non-zero exit status 1.

Issue - State: closed - Opened by NaughtDZ almost 2 years ago - 1 comment

#68 - Why "python -m venv --system-site-packages venv"

Issue - State: closed - Opened by NaughtDZ almost 2 years ago - 1 comment

#67 - Accelerate config step during installation won't work

Issue - State: closed - Opened by HanJammer almost 2 years ago - 2 comments

#66 - v20.4.0

Pull Request - State: closed - Opened by bmaltais almost 2 years ago

#65 - kohya gui finetune: [WinError 2] The system cannot find the file specified

Issue - State: closed - Opened by nub2927 almost 2 years ago - 7 comments

#64 - Cannot find libcudart.so

Issue - State: closed - Opened by mtdkmgw almost 2 years ago - 3 comments

#63 - Quick fix for captioninf extension

Pull Request - State: closed - Opened by bmaltais almost 2 years ago

#62 - LoRA Recomendations / Guide

Issue - State: closed - Opened by MrKuenning almost 2 years ago - 2 comments

#61 - Out of memory

Issue - State: closed - Opened by freeload101 almost 2 years ago - 3 comments

#60 - state_dict for UNet2DConditionModel

Issue - State: closed - Opened by Pershing-Maxwell almost 2 years ago - 1 comment

#59 - v20.3.0 release

Pull Request - State: closed - Opened by bmaltais almost 2 years ago

#58 - [Proposal] Create Collab with this project

Issue - State: closed - Opened by CombainerA19 almost 2 years ago - 2 comments

#57 - I get an error message that LORA cannot run because of a missing paging file, even though the required operating environment should be in place.

Issue - State: closed - Opened by P-ara-cell almost 2 years ago - 4 comments

#55 - RuntimeWarning: Mean of empty slice. invalid value encountered in scalar divide

Issue - State: closed - Opened by healthyfat almost 2 years ago - 1 comment

#54 - Save every n epochs does not match epoch

Issue - State: closed - Opened by the0ckid82 almost 2 years ago - 1 comment

#53 - returned non-zero exit status 1.

Issue - State: closed - Opened by healthyfat almost 2 years ago - 15 comments

#52 - No data found. Please verify arguments

Issue - State: closed - Opened by healthyfat almost 2 years ago - 6 comments

GitHub / bmaltais/kohya_ss issues and pull requests