2. Limitations

unit-scaling is a new library and (despite our best efforts!) we can’t guarantee it will be bug-free or feature-complete. We’re keen to assist anyone who wants to use the library, and help them work through any issues.

Known limitations of the library include:

1. Op coverage: we’ve currently focussed on adding common transformer operations — other ops may be missing (though we can add most requested ops without difficulty).

2. Using transforms with torch.compile: currently our transforms (for example unit_scale, simulate_fp8) can’t be used directly with torch.compile. We provide a special compilation function to get around this: unit_scaling.transforms.compile (see docs for more details), though this only works with unit_scale and not simulate_fp8.

3. Distributed training: although we suspect distributed training will still work reasonably well with the current library, we haven’t tested this.

This list is not exhaustive and we encourage you to get in touch if you have feature-requests not listed here.