Default Branch

b6c83aad55 · [SYCL] ehance UPSCALE to support all UT cases (#20637) · Updated 2026-03-17 02:01:52 +00:00

Branches

ec125419ea · Revert "assert that event is valid in event_wait instead of skipping if it isn't" · Updated 2026-03-17 06:02:19 +00:00

65
11

840dfe4a7a · reenable transfer queue if graphics queue is not used · Updated 2026-03-16 08:17:44 +00:00

27
4

15324f905b · cont : reduce paths · Updated 2026-03-15 11:03:20 +00:00

34
3

0776a6a039 · remove event pending stage · Updated 2026-03-15 08:00:06 +00:00

65
10

5ec6569eb5 · unify scalar+vector and fix reduce function · Updated 2026-03-13 08:23:03 +00:00

66
6

95ae9982d3 · Merge branch 'master' into compilade/imatrix-neutral-prior · Updated 2026-03-12 17:20:00 +00:00

68
6

7ded1269ab · unify matmul_id shader selection · Updated 2026-03-12 13:55:12 +00:00

69
2

20fbf04cd6 · metal : fix capture_started flag · Updated 2026-03-11 21:15:16 +00:00

91
2

8c8544f9fb · metal : fix capture_compute counter logic · Updated 2026-03-11 16:35:45 +00:00

97
1

715ed28683 · use scalar sums · Updated 2026-03-07 21:11:40 +00:00

158
3

121fe62182 · test · Updated 2026-03-06 14:30:32 +00:00

174
7

4b436e4e5e · flake8 fix · Updated 2026-02-23 10:48:01 +00:00

282
20

5d45884106 · metal : fix build · Updated 2026-02-18 07:14:31 +00:00

389
23

c0c3e428dd · refactor · Updated 2026-02-16 21:02:45 +00:00

326
49

5da56dc1d8 · args : add -kvu to llama-parallel · Updated 2026-02-12 19:50:01 +00:00    sdgoij

389
17

e7fbfc9b80 · ci : tmp fixes · Updated 2026-02-11 13:48:40 +00:00    sdgoij

439
22

5372fc6461 · wip · Updated 2026-02-10 21:44:42 +00:00    sdgoij

404
18

b9b56b017e · Apply suggestion from @ggerganov (src->buffer to buf_src) v2 · Updated 2026-02-10 11:00:44 +00:00    sdgoij

404
13

5144018e7b · cont : simplify · Updated 2026-02-07 12:50:05 +00:00    sdgoij

425
4

1213a03564 · qwen3next : fix chunking · Updated 2026-02-04 08:06:38 +00:00    sdgoij

458
1