[V1] [Doc] Update V1 docs for Mamba models #20499
Automate choice of attention block size; update docs
f9a5f4e4
tdoublep
force pushed
to
f9a5f4e4
1 year ago
Fix typo
e618204a
More verbose logging / asserts
600ec11c
Update docs re: FlashInfer
ec6d8409
Resolve conflicts with main; Update docs based on review feedback
a5c542f6
Address review feedback
84daa123
Second attempt at auto-setting attention block size
5ea6bedf
Fix logic slightly
1df73195
Cleanup
d04dcfe9
Merge branch 'main' into auto-block-size
11680bfe
Remove unused import
2ff9a096
Revert other changs; update docs
bcd9376a
tdoublep
changed the title [V1] [Doc] Automated choice of attention block size for hybrid models in V1 [V1] [Doc] Update V1 docs for Mamba models 1 year ago
Apply suggestions from code review
61ec90b5
tdoublep
force pushed
to
61ec90b5
1 year ago
vllm-bot
merged
5358cce5
into main 1 year ago
tdoublep
deleted the auto-block-size branch 1 year ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub