More fine-grained documentation needed for torchao.autoquant()

[suvadityamuk]

There is an important need for better documentation of options, features, and args for torchao.autoquant() along with some important gotchas, i.e. around the error_on_unseen flag among others.

[jerryzh168]

cc @HDCharles can you take a look

maybe adding error_on_unseen to the official list of args:

ao/torchao/quantization/autoquant.py

Line 558 in 53b6b78

def autoquant(

would be helpful

[jerryzh168]

@suvadityamuk please let us know if you'd like to work on this as well, also it would be helpful if you can provide a bit more detail on what the existing doc is missing as well, apart from error_on_unseen

[suvadityamuk]

@suvadityamuk please let us know if you'd like to work on this as well, also it would be helpful if you can provide a bit more detail on what the existing doc is missing as well, apart from error_on_unseen

Yeah sure, happy to work on this. I'll try to go and check again further about what is missing as well.

cc: @msaroufim We spoke today at the PyTorch Conference about this. Let me know any next steps about this as well, thanks!