[WIP] [float8] add float auto_filter_for_recipe

[danielvegamyhre]

Fixes #1207

Problem

• float8 rowwise + vanilla TP in torchtitan had flat perf with respect to bfloat.
• RCA In float8 rowwise vanilla TP low throughput #1207 found attention.wk and attention.wv layers were so small that float8 rowwise conversion resulted in approx ~40% slowdown for those layers, thus the perf benefits from fp8 rowwise conversion on larger linears were nullified.
• This is because the default filter_fqns for float8 model conversion are fine for the fp8 tensorwise recipe, but bad for the float8 rowwise recipe.

Solution

This has been a footgun for various users as well (including Poolside), so I created an "auto filter" (pytorch/ao#2410) which automatically filters Linears for a given float8 recipe, by checking for the following criteria:

1. dims not divisible by 16 (hardware requirement for float8)
2. dim sizes below thresholds that will result in worse perf for that given recipe, using simple heuristics based on the linked recipe perf tables above.
3. fqn matches one of the user defined filter_fqns

It prevents users from hitting this common footgun, while also preserving the flexibility to define their model-specific fqns.

Results

• llama3 8b with FSDP=2, TP=2 and default filter fn: ~1500-1700TPS (https://www.internalfb.com/phabricator/paste/view/P1845180084)
• llama3 8b with FSDP=2, TP=2 and auto filter fn: ~2150 TPS (1.26x - 1.43x improvement): (https://www.internalfb.com/phabricator/paste/view/P1845184279)