Samuel Oliveira Alves
|
276e4ea1eb
|
fix: isolate DFlash cross-device IO (#2243)
* fix: isolate DFlash cross-device IO
* remove unnecesary syncronization and simplify logic for tensors copies
|
2026-08-03 19:22:29 +03:00 |
Joel Farthing
|
1a7691fae7
|
DFlash: add Laguna XS 2.1 support (#2124)
* Add Laguna XS 2.1 DFlash support
* Fix DFlash prompt capture with microbatches
---------
Co-authored-by: Joel Farthing <262452229+joelfarthing@users.noreply.github.com>
|
2026-07-27 07:40:10 +03:00 |
|
Joel Farthing
|
64fceb70bc
|
DFlash: use persistent FA-ready K/V cache (#1997)
* Prototype physical-order DFlash KV cache
(cherry picked from commit f9093d9ee57cf66f6ce44c42524158bb1449d1c9)
* Use persistent FA-ready DFlash KV cache
(cherry picked from commit cfed6ae456b5448ac0053fbd5994037af845a69a)
* Address DFlash review cleanup
---------
Co-authored-by: Joel Farthing <262452229+joelfarthing@users.noreply.github.com>
|
2026-06-22 16:49:35 +02:00 |
|
Samuel Oliveira Alves
|
d5c04c15fd
|
clean redudance in dflash graph and small logics (#1994)
|
2026-06-19 09:04:54 +02:00 |
Kawrakow
|
71af16a6b7
|
Fix DFlash oerformance with split mode graph (#1980)
|
2026-06-17 18:40:02 +02:00 |
SamuelOliveirads
|
6cae8c7ba2
|
clean logs
|
2026-06-14 21:07:57 -03:00 |
|
SamuelOliveirads
|
0d75eee35a
|
remove duplicated code and unnecesary refactor
|
2026-06-14 16:02:02 -03:00 |
|
SamuelOliveirads
|
3b1a0f88d5
|
Add logging for DFlash statistics and clean up workspace handling
|
2026-06-13 20:14:08 -03:00 |
|
SamuelOliveirads
|
dc43cdf06b
|
move dflash for it own file
|
2026-06-02 10:22:13 -03:00 |