Jules Gagnon-Marchand

This project is super cool! & it works on TMUX (though the output seems to be very small), but it doesn't seem to work on GNU Screen. Was wondering if...

Cool repo. I would suggest also trying `@tf.function` for eager, which would make it more representative. Similarly, testing `@torch.jit.script` would be interesting. Not sure if these two do anything on...

This allows to just do ``` import text2digits parser = text2digits.Text2Digits() ``` instead of having to do ``` import text2digits.text2digits parser = text2digits.text2digits.Text2Digits() ``` which is just awkward

https://github.com/ShailChoksi/text2digits/issues/37 Passes all tests. Fixes the problem when there is a word number at the end of a sequence, but there is a space or punctuation after.

### 🚀 The feature, motivation, and pitch Currently TLRX, as far as I know, requires that you pass a list of strings as the prompts. However, at larger scales, if...

Logging with the root logger, like `logging.info`, removes the possibility of controlling the log level of submodules separately. `logging.getLogger(__name__)` enables this (& is the recommended practice), by doing something like...

Just a warning that the package doesn't work with Transformers 4.25.1

1

Looks like generation_beam_constraints doesn't exist or has been moved?

100% likely that two function parameters have been merged by accident

1

https://github.com/allenai/RL4LMs/blob/main/rl4lms/envs/text_generation/policy/seq2seq_policy.py#L263 

Fix ml.ExperimentDataPipe for Distributed Training

11

The current form of `ml.ExperimentDataPipe` breaks in distributed training when the amount of samples isn't split between GPUs in a way that allows for each GPUs to have the same...

[REQUEST] Add a table level "overflow" setting that is used as a default for the columns.

1

I find that most of the time when I use rich.table.Table, I don't create column objects directly, I just pass the names to the Table constructor, & most columns have...