How to Confirm the results?

[ChenMuDX]

Hello , I have trained a VanillaVAE model using your code. The reconstructions and samples look good.
But how can I compare the results between yours(in theREADME.md) and mine? Could you provide the testing data? Thanks！

[RobertoGrabovac]

Hey, how did you manage to train VanillaVAE model? Did you use google colab and can you write exact versions of all libraries used? I would be grateful for this as I didn't manage to even start the training process due to the errors.

[ChenMuDX]

I didn't use the Colab library。
This is my requirement.txt pytorch-lightning==1.9.0
PyYAML==6.0
tensorboard>=2.2.0
torch>=1.6.1
torchsummary==1.5.1
torchvision>=0.10.1

and all libraries:

absl-py 2.1.0
aiohappyeyeballs 2.4.4
aiohttp 3.11.9
aiosignal 1.3.1
anyio 4.6.2.post1
arrow 1.3.0
asttokens 3.0.0
async-timeout 5.0.1
attrs 24.2.0
beautifulsoup4 4.12.3
blessed 1.20.0
boto3 1.35.72
botocore 1.35.72
certifi 2024.6.2
charset-normalizer 3.3.2
click 8.1.7
cmake 3.31.1
comm 0.2.2
contextlib2 21.6.0
contourpy 1.3.0
croniter 1.3.15
cycler 0.12.1
dateutils 0.6.12
debugpy 1.8.11
decorator 4.4.2
deepdiff 7.0.1
distro 1.7.0
editor 1.6.6
et-xmlfile 1.1.0
exceptiongroup 1.2.2
executing 2.1.0
fastapi 0.88.0
filelock 3.15.3
fonttools 4.54.1
frozenlist 1.5.0
fsspec 2023.12.2
future 1.0.0
grpcio 1.68.0
h11 0.14.0
idna 3.7
inquirer 3.4.0
ipykernel 6.29.5
ipython 8.30.0
itsdangerous 2.2.0
jedi 0.19.2
Jinja2 3.1.4
jmespath 1.0.1
joblib 1.4.2
jupyter_client 8.6.3
jupyter_core 5.7.2
kiwisolver 1.4.7
lightning-cloud 0.5.70
lightning-utilities 0.11.9
lit 18.1.8
Markdown 3.7
markdown-it-py 3.0.0
MarkupSafe 2.1.5
matplotlib 3.9.2
matplotlib-inline 0.1.7
mdurl 0.1.2
ml_collections 0.1.1
mpmath 1.3.0
multidict 6.1.0
nest-asyncio 1.6.0
networkx 3.3
numpy 1.26.4
nvidia-cublas-cu11 11.11.3.6
nvidia-cuda-cupti-cu11 11.8.87
nvidia-cuda-nvrtc-cu11 11.8.89
nvidia-cuda-runtime-cu11 11.8.89
nvidia-cudnn-cu11 8.7.0.84
nvidia-cufft-cu11 10.9.0.58
nvidia-curand-cu11 10.3.0.86
nvidia-cusolver-cu11 11.4.1.48
nvidia-cusparse-cu11 11.7.5.86
nvidia-nccl-cu11 2.20.5
nvidia-nvtx-cu11 11.8.86
nvidia-pyindex 1.0.9
onnx 1.17.0
opencv-python 4.10.0.84
openpyxl 3.1.5
ordered-set 4.1.0
packaging 22.0
pandas 2.2.3
parso 0.8.4
pexpect 4.9.0
pillow 10.3.0
pip 24.0
pip-autoremove 0.10.0
platformdirs 4.3.6
prompt_toolkit 3.0.48
propcache 0.2.1
protobuf 5.29.0
psutil 6.1.0
ptyprocess 0.7.0
pure_eval 0.2.3
pycairo 1.20.1
pydantic 1.10.19
pyDeprecate 0.3.1
Pygments 2.18.0
PyGObject 3.42.1
PyJWT 2.10.1
pyparsing 3.1.4
PySocks 1.7.1
python-dateutil 2.9.0.post0
python-multipart 0.0.19
pytorch-lightning 1.9.0
pytz 2024.2
PyYAML 6.0
pyzmq 26.2.0
readchar 4.2.1
requests 2.32.3
rich 13.9.4
runs 1.2.2
s3transfer 0.10.4
scikit-learn 1.5.2
scipy 1.14.1
selinux 3.3
sepolicy 3.3
setools 4.4.0
setuptools 70.1.0
six 1.16.0
sniffio 1.3.1
soupsieve 2.6
ssh-import-id 5.11
stack-data 0.6.3
starlette 0.22.0
starsessions 1.3.0
sympy 1.12.1
tensorboard 2.18.0
tensorboard-data-server 0.7.2
threadpoolctl 3.5.0
torch 2.0.0+cu118
torchaudio 2.0.1+cu118
torchmetrics 1.6.0
torchsummary 1.5.1
torchvision 0.15.1+cu118
tornado 6.4.2
tqdm 4.67.1
traitlets 5.14.3
triton 2.0.0
types-python-dateutil 2.9.0.20241003
typing_extensions 4.12.2
tzdata 2024.2
urllib3 2.2.2
uvicorn 0.32.1
wcwidth 0.2.13
websocket-client 1.8.0
websockets 11.0.3
Werkzeug 3.1.3
wheel 0.43.0
xmod 1.8.1
yarl 1.18.3

in this env,you should change the config.yaml
trainer_params:
accelerator: 'gpu'
devices: [0]
max_epochs: 100
and Comment out DDPS in run.py
#strategy=DDPStrategy(find_unused_parameters=False),

[RobertoGrabovac]

Thank you so much, it seems like its working but training process is taking too long unfortunately. Could you maybe provide me with the last checkpoint of your model (both decoder and encoder), so that I can load it? I would like to be able to encode some of my pictures and decode them, to see how does the reconstructed pictures look. This would be very helpful to me and I thank you in advance. Sorry for taking your time, but you already helped me a lot :)

[ChenMuDX]

Uploading vaeckp.zip…
I think what you want is in the zip file.

[ChenMuDX]

If the file can't be uploaded, give me your e-mail address and I'll send it to you.