Default Branch

11924d4c17 · test: fix some CI errors (#26415) · Updated 2026-08-01 17:16:29 -05:00

Branches

fb18934a97 · gguf-py : bump version to 0.11.0 · Updated 2024-12-11 15:13:31 -06:00    upstream-archive

5915
0
Included

4f3a7e279b · Force max subgroup size for coopmat shaders · Updated 2024-12-10 14:27:04 -06:00    upstream-archive

5923
2

b8d1b1a5e1 · server : fix infill prompt format · Updated 2024-12-08 14:12:11 -06:00    upstream-archive

5933
1

a6648b9df7 · server : chunked prefill support · Updated 2024-12-08 01:48:18 -06:00    upstream-archive

5937
1

a8046c888a · use calloc instead of malloc · Updated 2024-12-04 10:24:35 -06:00    upstream-archive

5968
3

81611bef72 · server : add tests · Updated 2024-12-04 05:11:26 -06:00    upstream-archive

5968
3

33d7b70c88 · server : do not speculate during prompt processing · Updated 2024-12-03 02:58:43 -06:00    upstream-archive

5981
1

3c8a2a83fe · shmem experiments · Updated 2024-11-26 07:17:38 -06:00    upstream-archive

6044
3

dafedd33d2 · 4x4 -> 4x · Updated 2024-11-26 06:54:02 -06:00    upstream-archive

6044
2

bf3494345e · metal : some mul_mv experiments · Updated 2024-11-26 06:48:50 -06:00    upstream-archive

6044
1

b83cae088c · speculative : add infill mode · Updated 2024-11-26 03:14:17 -06:00    upstream-archive

6049
1

4ff0831ce6 · metal : use F16 math in mul_mat kernels · Updated 2024-11-25 07:15:26 -06:00    upstream-archive

6062
1

f7b0233eca · wip · Updated 2024-11-16 02:33:55 -06:00    upstream-archive

6124
1

5e6dad9322 · speculative : experimenting with Qwen2.5 · Updated 2024-11-14 03:31:31 -06:00    upstream-archive

6146
2

33bdee667e · speculative : fix out-of-bounds access · Updated 2024-11-14 03:23:45 -06:00    upstream-archive

6146
1

8c1b186cb5 · metal : minor Q4_0 optimization · Updated 2024-11-12 07:30:51 -06:00    upstream-archive

6156
21

3d1fe1bb4d · metal : int -> short, style · Updated 2024-11-09 02:32:16 -06:00    upstream-archive

6167
2

bd1198a67a · metal : fix build and some more comments · Updated 2024-11-09 02:09:50 -06:00    upstream-archive

6167
1

a2385da59c · make : clean-up [no ci] · Updated 2024-11-08 05:46:20 -06:00    upstream-archive

6174
9

94accca4c2 · vec move mask to shmem · Updated 2024-11-07 12:58:10 -06:00    upstream-archive

6184
19