EleutherAI/oslo Issues
TP does not work on Flan-T5
UpdatedTo apply FlashAttention
Updated 1TODO: Test Code Enhancement
UpdatedFix TP embedding layers
UpdatedLazy Parallelization
ClosedHow to pretrain T5
UpdatedTODO: documentation and readme
Updated 1PatricStar for Zero
UpdatedcoloDDP integration
Closed 1wrap transformers layer
Closedpass optimizer parameters
Closed 2Modify PP to functional
ClosedTODO: Test TP + PP
UpdatedFix some code errors
Closed