[litgpt benchmark] enable force_recompute_fp8_weight_in_bwd
when torchao.float8
is used with FSDP2
#1528
Azure Pipelines / lightning-thunder (GPUs) (testing ubuntu22.04 | cuda 12.1 | python 3.10 | torch-nightly | distributed)
succeeded
Dec 17, 2024 in 20m 10s
testing ubuntu22.04 | cuda 12.1 | python 3.10 | torch-nightly | distributed succeeded
Loading