-
It seems that the fp16 setting is not effective. I tried to use fp16 manually, and offload the autoencoder and CLIP to cpu memory during ddim denoising, and can run a 45x512x768 video on 12G GPU memor…
-
First of all, thank you so much for making this high-quality repository as well as pretrained models publicly available! This is highly useful for exploring your research.
I am currently training f…
leod updated
4 months ago
-
Hi Joao
In a denoising auto-encoder
I have:
`NeuralFit.FitLoading(NNAutoencoder, {EpochSize=}SETrainSize.Value, 0, 0, {Batch=}SEBatch.Value, {Epochs=}SEEpoch.Value, @GetTrainingData, nil, nil);…
-
Hi,
May I update the auto encoder model with the following models: https://github.com/rickwierenga/notebooks/blob/master/autoencoders.ipynb (in Swift, of course) ? The proposed models are more exte…
-
-
Paper: https://arxiv.org/abs/1312.6114
Example implementations (non-official): https://github.com/clementchadebec/benchmark_VAE
osbm updated
9 months ago
-
TODOs for the Autoencoder:
- [ ] write tests for the other activation function types
- [ ] better stopping criteria
- [x] more than one training step, continue training a `dimRedResult` object
- […
-
Hello @thushv89 , thanks for the great materials. I am trying to adapt the seq2seq NMT code as an autoencoder for sentence embedding. Do you think it's a good way to use it for such purpose?
Thank…
-
Thanks for the good work ! I assessed the quality of VQGAN on my data and it was really poor which caused poor quality as well when I used your model. So I want to not use any autoencoder anymore and …
-