transformers
Fix GPT-NeoX-20B past handling, attention computation
#17811

Merged

Fix GPT-NeoX-20B past handling, attention computation #17811

sgugger merged 2 commits into huggingface:main from zphang:neox_20b_fixes_220621_v2

zphang force pushed 3 years ago

sgugger approved these changes on 2022-06-21

patrickvonplaten commented on 2022-06-21

patrickvonplaten approved these changes on 2022-06-21

zphang force pushed 3 years ago

Fix GPT-NeoX-20B past handling, swap attention computation to hopeful…

50f66010

20B tests

d2e9de90

zphang force pushed to d2e9de90 3 years ago

sgugger merged 205bc415 into main 3 years ago

Reviewers

patrickvonplaten

sgugger

Assignees

No one assigned

Labels

None yet

Milestone

No milestone