-
Hello,
My team and I have encountered an unusual issue regarding the convergence dynamics of nnUNet. Initially, we trained nnUNet for 2000 epochs and observed the following learning curve:
![pro…
-
tools opening multiple terminals at once, like `tmux-resurrect` lead to conflicts like this
```
2024-09-03 11:39:47 | table `17253563876` already exists
2024-09-03 11:39:47 | table `17253563876` al…
-
Hello,
I noticed in your paper and code that you trained Far3D and StreamPETR for 6 epochs before comparison. However, the StreamPETR paper mentions training their network for 60 epochs for the cas…
-
Hi, I wonder how many epochs do you train PSALM during the finetuning phase?
In your training script the num_train_epochs is set to 10, but in your paper you stated 56k with batch_size 64 are used …
-
As the title says.
I assume that the number of epochs is different than "diffusion steps". Can you please guide me on how to change the number of epochs?
-
Dear Dr. Shehroz Khan
Thank you for providing very excellent code.
How to modify the training Epoch?
Thank you so much.
-
## Proposal
Generalize time travel query for all batch queries, which means that all batch query will be handled as time travel query.
In a single `HummockVersion`, we only provide a single vie…
-
I see that the Dataloaders have the shuffle parameter set to False and this is done because the way data loading is handled currently only allows sequential data reads. However, training the algorithm…
-
Writeup an initial spec that describe how to do nullifier epochs. The spec should give an idea of how large the changes are, and how easy it would be for us to implement. Will the complexity of the ch…
-
Epoch height 1055 is around the last height with protocol version 48 used. Apparently what matters is that we started to use `ProtocolFeature::AliasValidatorSelectionAlgorithm` after that. Looks like …