Allan Jie

Version

1

What is the pytorch version for running the code?

In the code https://github.com/yala/text_nn/blob/master/rationale_net/utils/learn.py#L71-L85 ```python def get_hard_mask(z, return_ind=False): ''' -z: torch Tensor where each element probablity of element being selected -args: experiment level config returns: A torch variable that is...

Experiments for MAWPS-s

5

Is the experiment for MAWPS-s using 5-fold as well? It seems yes to me as the paper reported. I got around 85.4 accuracy on MAWPS using train/dev/test. Wondering if I'm...

Reproducing Performance on DocVQA using LayoutLMv3/LayoutLMv2

3

I tried my best to reproduce the results reported in the paper, which is about 78% test set ANLS. But all I get is just 74% on the test set...

CUDA Out-of-memory using V100

1

I'm using V100 for experiments, but still out of memory in the middle of the training process. Not sure what would be the reason at this momnet ```bash Namespace(aux_loss=True, backbone='resnet50',...

Performance on CoNLL-2003

5

Can I know what's the performance you obtain with your new implementation

The speed seems to be pretty slow.

**Describe the bug** I use the HuggingFace repo to implement the inference from Galactica. The Galactica model is implemented based on OPT model in the HuggingFace implementation. While my first...

Since I haven't gone deeply into this package yet, just wonder if this package/tool can allow us to define arbitrary graphs (that have exact inference) for structured prediction task. I...

I saw this function, but never called https://github.com/paperswithcode/galai/blob/3a724f562af1a0c8ff97a096c5fbebe579e2160f/galai/model.py#L143-L172 But quite confused that it seems we never use the bos and eos tokens, even in transformer, they are `None` if we...