xla
cff9f4e0
- Fix a bug in flash attention where kv_seq_len should divide block_k_major. (#8671)
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
264 days ago
Fix a bug in flash attention where kv_seq_len should divide block_k_major. (#8671)
References
#8671 - Fix a bug in flash attention where kv_seq_len should divide block_k_major.
Author
zhangp365
Parents
c0afda3a
Loading