 Ruben Ortlam
|
9073a73d82
vulkan: vec dot matrix multiplication fix (#16151)
|
3 ay önce |
 Xuan-Son Nguyen
|
0dd58b6877
ggml : refactor forward_dup for cpu backend (#16062)
|
4 ay önce |
 Bowen Han
|
38dbdf4c05
CUDA: Optimize PAD_REFLECT_1D (#15957)
|
4 ay önce |
 Reese Levine
|
d304f459d8
GGML WebGPU: Support for ADD, MUL, RMS_NORM, GET_ROWS operators (#16018)
|
4 ay önce |
 Georgi Gerganov
|
0320ac5264
metal : refactor + optimize v2 (#15995)
|
4 ay önce |
 Oliver Simons
|
00681dfc16
CUDA: Add `fastdiv` to `k_bin_bcast*`, giving 1-3% E2E performance (#15872)
|
4 ay önce |
 Daniel Bevenius
|
e7b6d83b52
tests : filter out no-ops from coverage report (#15900)
|
4 ay önce |
 Jeff Bolz
|
4f63cd705c
vulkan: Fix OOB accesses in soft_max_back (#15861)
|
4 ay önce |
 Aman Gupta
|
a972faebed
CUDA: Add mul_mat_id support for the mmf kernel (#15767)
|
4 ay önce |
|
Georgi Gerganov
|
f28d4f4ac9
metal : refactor + optimize (#15857)
|
4 ay önce |
|
Xuan-Son Nguyen
|
9fcb29f22f
ggml: allow casting between f32 and i32 (#15783)
|
4 ay önce |
|
Jeff Bolz
|
d413dca003
tests: large sizes for get_rows (#15687)
|
4 ay önce |
|
Jeff Bolz
|
3976dfbe00
vulkan: support im2col_3d (#15795)
|
4 ay önce |
|
Jeff Bolz
|
c97b5e5854
vulkan: Support pad_ext (#15794)
|
4 ay önce |
|
Daniel Bevenius
|
3a550b5ca4
tests : add --list-ops and --show-coverage options (#15745)
|
4 ay önce |
 leejet
|
0a1b3982cd
ggml: add ops for WAN video model (cuda && cpu) (#15669)
|
4 ay önce |
 rmatif
|
86076f92de
OpenCL: add fused group_norm/norm, mul, add (#15314)
|
4 ay önce |
 Eve
|
44b1efa41a
tests: add performance test for mul mat id (#15543)
|
4 ay önce |
|
Georgi Gerganov
|
1d8d83deaa
metal : improve `MUL_MAT_ID` (#15541)
|
4 ay önce |
|
Jeff Bolz
|
34bdbbd7c2
vulkan: Remove splitting for mul_mat_id (#15568)
|
4 ay önce |
|
Jeff Bolz
|
886b97a5d6
tests: Generate unique input values for count_equal (#15487)
|
4 ay önce |
|
Jeff Bolz
|
c9a24fb932
vulkan: Support FA with any multiple of 8 head sizes (#15537)
|
4 ay önce |
|
Jeff Bolz
|
611f419cff
vulkan: optimize rms_norm, and allow the work to spread across multiple SMs (#15281)
|
4 ay önce |
 Acly
|
0a9b43e507
vulkan : support ggml_mean (#15393)
|
4 ay önce |
|
rmatif
|
92f7f0a53c
ggml: add `conv3d` op (#15182)
|
4 ay önce |
|
Jeff Bolz
|
96452a3fa4
vulkan: Reuse conversion results in prealloc_y (#15410)
|
4 ay önce |
|
Jeff Bolz
|
de5627910d
vulkan: Optimize argsort (#15354)
|
5 ay önce |
|
Jeff Bolz
|
1fe00296f5
vulkan: fuse adds (#15252)
|
5 ay önce |
|
Jeff Bolz
|
2e2b22ba66
vulkan: Add missing bounds checking to scalar/coopmat1 mul_mat_id (#15334)
|
5 ay önce |
|
Georgi Gerganov
|
5edf1592fd
vulkan : fix out-of-bounds access in argmax kernel (#15342)
|
5 ay önce |
