Open fegin opened 3 months ago
Stack from ghstack (oldest at bottom):
This PR adds experimental flags and functions to enable context parallelism. We currently support on ly FSDP + CP and CP only. CP + TP is being tested.
Stack from ghstack (oldest at bottom):
This PR adds experimental flags and functions to enable context parallelism. We currently support on ly FSDP + CP and CP only. CP + TP is being tested.