blgpb

The losses rec, kl both becomes NAN

5

Dear Prof. Rybkin, I have read your paper "Simple and Effective VAE Training with Calibrated Decoders". It is a very nice work! I have learned a lot from it. I...

In the trainer.py, `ctx.y_true = batch.num_missing[mask]` `ctx.y_prob = pred_missing` https://github.com/alibaba/FederatedScope/blob/480b67de851df2fa02b7cc2189f1803d947998d7/federatedscope/gfl/fedsageplus/trainer.py#L42C1-L43C34 https://github.com/alibaba/FederatedScope/blob/480b67de851df2fa02b7cc2189f1803d947998d7/federatedscope/gfl/fedsageplus/trainer.py#L66C1-L67C34 It seems that 'pred_missing' (i.e., the output of the missing neighbor generator) and 'num_missing' (i.e., the groundtruth of missing...

When I run `python3 main.py --name dsprites_beta_1_pz_norm_iso --model dsprites --skip-test --epochs 20 --lr 5e-4 --batch-size 64 --latent-dim 10 --beta 1. --prior-variance iso ` an error occurred `ValueError: Expected parameter probs...