Kuangdai Leng

Minor bug in loss logging

1

Line 109 in `disvae/models/losses.py`: ```python if not is_train or self.n_train_steps % self.record_loss_every == 1: ``` This does not work when self.record_loss_every is 1 (recording all mini-batches). Fix: ```python if not...

sequence interpreted as string

1

According to YMAL standard, I think the following node foo: [1, 2, 3] should be interpreted as a sequence, while mini-ymal interprets it as a string "[1, 2, 3]". Is...

Adding at() or find()?

1

Thanks for providing such a wonderful tool. I am using yaml only for input (and I guess this is the way yaml is used frequently). Currently, random visit is supported...

### Feature request We are using langchain for non-English applications. The prefix for system is hardcoded as "System": ```python for m in messages: if isinstance(m, HumanMessage): role = human_prefix elif...

This PR creates the complete ZCS tutorial `zcs.rst`, lined from `operator.rst`. It also adds the ZCS paper to `research.rst`. The revised pages are rendered and proofread. Modules of `zcs` are...

Different kernel sizes for different dimensions

2

Can you maybe add this feature, please? I think the main change would be * From ```python class ConvolutionalBlock(nn.Module): def __init__( self, dimensions: int, in_channels: int, out_channels: int, normalization: Optional[str]...

FD-based gradient calculation seems incorrect for Burgers (with code to verify)

2

# The issue I am trying to understand how gradients are computed for Burgers, implemented by `FDM_Burgers()` in `train_utils/losses.py`, as pasted below: ```python def FDM_Burgers(u, v, D=1): batchsize = u.size(0)...

Thank you for the great effort. I have two questions: 1. Do you host a ranking for each problem? 2. Is there a workflow people can follow to contribute models...