- 20 Jan, 2024 1 commit
-
-
Tri Dao authored
-
- 15 Jan, 2024 2 commits
- 14 Jan, 2024 5 commits
- 13 Jan, 2024 2 commits
- 12 Jan, 2024 1 commit
-
-
Tri Dao authored
-
- 10 Jan, 2024 1 commit
-
-
Tri Dao authored
-
- 08 Jan, 2024 1 commit
-
-
Erich Schubert authored
-
- 05 Jan, 2024 3 commits
- 04 Jan, 2024 1 commit
-
-
Tri Dao authored
-
- 03 Jan, 2024 1 commit
-
-
jiaxingli authored
* fix deterministic * fix deterministic
-
- 26 Dec, 2023 1 commit
-
-
Tri Dao authored
-
- 25 Dec, 2023 4 commits
- 24 Dec, 2023 3 commits
- 23 Dec, 2023 1 commit
-
-
Tri Dao authored
-
- 22 Dec, 2023 7 commits
- 20 Dec, 2023 4 commits
-
-
Tri Dao authored
-
Tri Dao authored
-
Sanghun Cho authored
* hard-code alibi in fwd * use params.h as hun_heads * hard-code alibi in bwd * add alibi on/off option * compute alibi_start, ratio outside of kernels * fix minor merge conflict * add test_alibi.py * change apply_alibi() location before masking * add alibi in splitkv kernel * fix backward func # of returns * add out-of-bound check in apply_alibi() * update test_alibi.py * update test_alibi.py for kvcache * simplify alibi parameter interface * fix performance issue by computing alibi outside of branch * update test_flash_attn_varlen_func() for left padding * implement alibi_slopes (b, nh) loading * optimize apply_alibi() a bit * update test cases for alibi_slopes loading * reflect stylistic comments * disable "seqlenq_ngroups_swapped" when using alibi --------- Co-authored-by:monk.detective <monk.detective@kakaobrain.com>
-
Tri Dao authored
-
- 17 Dec, 2023 2 commits