work on AMD gpu (rx6800) + Ubuntu 22.04.2 LTS (Jammy Jellyfish)

[tornado73]

install
git clone https://github.com/bmaltais/kohya_ss.git
cd kohya_ss
python -m venv venv
source venv/bin/activate
pip install torch torchvision --index-url https://download.pytorch.org/whl/rocm5.4.2
pip install --use-pep517 --upgrade -r requirements.txt
accelerate config

answers

No distributed training
no
no
no
all
fp16

pip install tensorflow-rocm
sudo apt install python3-tk
export HSA_OVERRIDE_GFX_VERSION=10.3.0
source venv/bin/activate
python kohya_gui.py "$@"

settings Lora of person

Caption Extension - .txt

Optimizer - AdanW

LR Scheduler - constant

Network Rank (Dimension) 128

Network Alpha 128 or 64

Enable buckets - uncheck

Advanced Configuration

Use xformers - uncheck

Memory efficient attention -check

Max num workers for DataLoader- 1

---

start next time

open gui.sh

add

export HSA_OVERRIDE_GFX_VERSION=10.3.0
python3 kohya_gui.py "$@"

with other optimizers - or not work or loss=nan

Good luck

p.s. installed by me

Package Version

---

absl-py 1.4.0
accelerate 0.18.0
addict 2.4.0
aenum 3.1.12
aiofiles 23.1.0
aiohttp 3.8.4
aiosignal 1.3.1
altair 4.2.2
antlr4-python3-runtime 4.9.3
anyio 3.6.2
apturl 0.5.2
astunparse 1.6.3
async-timeout 4.0.2
attrs 23.1.0
basicsr 1.4.2
bcrypt 3.2.0
beautifulsoup4 4.12.2
blendmodes 2022
blinker 1.4
boltons 23.0.0
Brlapi 0.8.3
cachetools 5.3.0
certifi 2020.6.20
chardet 4.0.0
charset-normalizer 3.1.0
clean-fid 0.1.29
click 8.0.3
clip 1.0
cmake 3.25.0
colorama 0.4.4
command-not-found 0.3
contourpy 1.0.7
cryptography 3.4.8
cupshelpers 1.0
cycler 0.11.0
dbus-python 1.2.18
defer 1.0.6
deprecation 2.1.0
distro 1.7.0
distro-info 1.1build1
duplicity 0.8.21
einops 0.4.1
entrypoints 0.4
facexlib 0.3.0
fastapi 0.94.0
fasteners 0.14.1
ffmpy 0.3.0
filelock 3.9.0
filterpy 1.4.5
flatbuffers 23.3.3
font-roboto 0.0.1
fonts 0.0.3
fonttools 4.39.3
frozenlist 1.3.3
fsspec 2023.4.0
ftfy 6.1.1
future 0.18.2
gast 0.4.0
gdown 4.7.1
gfpgan 1.3.8
gitdb 4.0.10
GitPython 3.1.30
google-auth 2.17.3
google-auth-oauthlib 1.0.0
google-pasta 0.2.0
gradio 3.28.1
gradio_client 0.2.0
grpcio 1.54.0
h11 0.12.0
h5py 3.8.0
httpcore 0.15.0
httplib2 0.20.2
httpx 0.24.0
huggingface-hub 0.14.1
idna 3.3
imageio 2.28.1
importlib-metadata 4.6.4
inflection 0.5.1
jeepney 0.7.1
Jinja2 3.1.2
jsonmerge 1.8.0
jsonschema 4.17.3
keras 2.11.0
keyring 23.5.0
kiwisolver 1.4.4
kornia 0.6.7
language-selector 0.1
lark 1.1.2
launchpadlib 1.10.16
lazr.restfulclient 0.14.4
lazr.uri 1.0.6
lazy_loader 0.2
libclang 16.0.0
lightning-utilities 0.8.0
linkify-it-py 2.0.2
lit 15.0.7
llvmlite 0.40.0
lmdb 1.4.1
lockfile 0.12.2
louis 3.20.0
lpips 0.1.4
macaroonbakery 1.3.1
Mako 1.1.3
Markdown 3.4.3
markdown-it-py 2.2.0
MarkupSafe 2.1.2
matplotlib 3.7.1
mdit-py-plugins 0.3.3
mdurl 0.1.2
monotonic 1.6
more-itertools 8.10.0
mpmath 1.2.1
multidict 6.0.4
netifaces 0.11.0
networkx 3.0
numba 0.57.0
numpy 1.22.4
oauthlib 3.2.0
olefile 0.46
omegaconf 2.2.3
open-clip-torch 2.7.0
opencv-python 4.7.0.72
opt-einsum 3.3.0
orjson 3.8.11
packaging 23.1
pandas 2.0.1
paramiko 2.9.3
pexpect 4.8.0
piexif 1.1.3
Pillow 9.4.0
pip 23.1.2
protobuf 3.19.6
psutil 5.9.5
ptyprocess 0.7.0
pyasn1 0.5.0
pyasn1-modules 0.3.0
pycairo 1.20.1
pycups 2.0.1
pydantic 1.10.7
pydub 0.25.1
PyGObject 3.42.1
PyJWT 2.3.0
pymacaroons 0.13.0
PyNaCl 1.5.0
pyparsing 2.4.7
pyRFC3339 1.1
pyrsistent 0.19.3
PySocks 1.7.1
python-apt 2.4.0+ubuntu1
python-dateutil 2.8.2
python-debian 0.1.43ubuntu1
python-multipart 0.0.6
pytorch-lightning 1.9.4
pytorch-triton-rocm 2.0.1
pytz 2022.1
PyWavelets 1.4.1
pyxdg 0.27
PyYAML 5.4.1
realesrgan 0.3.0
regex 2023.5.5
reportlab 3.6.8
requests 2.25.1
requests-oauthlib 1.3.1
resize-right 0.0.2
rsa 4.9
safetensors 0.3.1
scikit-image 0.19.2
scipy 1.10.1
SecretStorage 3.3.1
semantic-version 2.10.0
sentencepiece 0.1.99
setuptools 59.6.0
six 1.16.0
smmap 5.0.0
sniffio 1.3.0
soupsieve 2.4.1
starlette 0.26.1
sympy 1.11.1
systemd-python 234
tb-nightly 2.14.0a20230508
tensorboard-data-server 0.7.0
tensorboard-plugin-wit 1.8.1
tensorflow-estimator 2.11.0
tensorflow-io-gcs-filesystem 0.32.0
tensorflow-rocm 2.11.0.540
termcolor 2.3.0
tifffile 2023.4.12
timm 0.6.7
tokenizers 0.13.3
tomli 2.0.1
toolz 0.12.0
torch 2.0.1+rocm5.4.2
torchaudio 2.0.2+rocm5.4.2
torchdiffeq 0.2.3
torchmetrics 0.11.4
torchsde 0.2.5
torchvision 0.15.2+rocm5.4.2
tqdm 4.65.0
trampoline 0.1.2
transformers 4.25.1
typing_extensions 4.4.0
tzdata 2023.3
ubuntu-advantage-tools 8001
ubuntu-drivers-common 0.0.0
uc-micro-py 1.0.2
ufw 0.36.1
unattended-upgrades 0.1
urllib3 1.26.5
usb-creator 0.3.7
uvicorn 0.22.0
wadllib 1.3.6
wcwidth 0.2.6
websockets 11.0.2
Werkzeug 2.3.3
wheel 0.40.0
wrapt 1.15.0
xdg 5
xkit 0.0.0
yapf 0.33.0
yarl 1.9.2
zipp 1.0.0

[joetech]

I did get some memory issues with this still, but made a couple small adjustments and got it to run. Should it take 30 hours (same GPU as you)? I'm going to let it run because if the quality is good at least it'll be done, but if I do future ones, I'd love if I could get a quality LORA without taking 30 hours. It dropped to about 9 hours when I moved from bf16 to fp16 (now ~2.5s/it)

[hqnicolas]

@joetech @tornado73 Today I turn my Gist into an ROCm Docker REPO.
Please if a Mantainer can Analyse and ADD this AMD Docker Compose to the bmaltais/kohya_ss REPO
https://github.com/hqnicolas/bmaltaisKohya_ssROCm
With this files you can Run Kohya_ss into AMD RDNA Consumer Graphics Series with Docker rocm/pytorch:rocm6.1
if You like it don't forget to leave a star on my repo ⭐

[Insistentelk]

Unfortunately at the current moment (9-24) this no longer works as the -r requirements.txt(line 35) breaks due to not bring a python project.