Skip to content

[WS2][PR5][Logp] Add deterministic DP-aware GRPO loss on the TP-aware logprob path - #291

Open
KJLdefeated wants to merge 10 commits into
testfrom
feat/ws2-grpo-loss-pr5
Open

[WS2][PR5][Logp] Add deterministic DP-aware GRPO loss on the TP-aware logprob path#291
KJLdefeated wants to merge 10 commits into
testfrom
feat/ws2-grpo-loss-pr5

ws2 deterministic grpo loss PR5

3627fcd
Select commit
Loading
Failed to load commit list.

Select a check to view from the sidebar