Arnaud Stiegler

Pix2Struct: unable to overfit on a single training sample

23

### System Info - `transformers` version: 4.28.0 - Platform: Linux-5.4.0-1037-aws-x86_64-with-glibc2.27 - Python version: 3.9.16 - Huggingface_hub version: 0.13.4 - Safetensors version: 0.3.0 - PyTorch version (GPU?): 1.13.0+cu117 (True) - Tensorflow...

Hi, I'd be curious to know why you're using random padding, and how much of an impact it has on training. It doesn't seem to be explicitly discussed in the...

Issue with tokenizing '1' preceded by a char

7

Looking at the Donut tokenizer on transformers, it seems that '1' is the only digit missing a standalone token in the tokenizer. As a result, any 1 preceded by a...

Default pipeline generates many "unreadable" documents

5

Hi, I use Augraphy extensively but I've noticed that: - the default pipeline can be too destructive on my documents to the point where a human cannot read the text...