 uvos
|
5ba36f6103
HIP: Cleanup hipification header (#15285)
|
5 months ago |
 Aldehir Rojas
|
b204a5a234
gpt-oss: implement harmony parsing (#15181)
|
5 months ago |
 Christian Kastner
|
646944cfa8
docker : Enable GGML_CPU_ALL_VARIANTS for ARM (#15267)
|
5 months ago |
 Georgi Gerganov
|
1a01899b61
readme : update hot topics (#15315)
|
5 months ago |
 Jeff Bolz
|
863d341eeb
vulkan: perf_logger improvements (#15246)
|
5 months ago |
|
Georgi Gerganov
|
d32e03f449
server : add SWA checkpoints (#15293)
|
5 months ago |
|
Georgi Gerganov
|
3973163bff
sync : ggml
|
5 months ago |
 Jason Ni
|
5ade3000bd
ggml: fix ggml_conv_1d_dw bug (ggml/1323)
|
5 months ago |
|
Georgi Gerganov
|
8b2483730f
tests : remove unused includes (ggml/0)
|
5 months ago |
 kallewoof
|
810b9fc8b9
perplexity : provide a helpful hint for has_cpl case in split_equal error. (#15304)
|
5 months ago |
 Sigbjørn Skjæret
|
4ebd0c125b
cuda : fix GGML_CUDA_GRAPHS=OFF (#15300)
|
5 months ago |
 Jonathan Graehl
|
5cdb27e091
finetune: SGD optimizer, more CLI args (#13873)
|
5 months ago |
|
kallewoof
|
3ea913f1ce
perplexity: give more information about constraints on failure (#15303)
|
5 months ago |
|
uvos
|
29c8fbe4e0
HIP: bump requirement to rocm 6.1 (#15296)
|
5 months ago |
 Bas Nijholt
|
1adc9812bd
fix(nix): remove non-functional llama-cpp cachix cache from flake.nix (#15295)
|
5 months ago |
|
Sigbjørn Skjæret
|
b3e16665e1
server : enable -td and -tbd parameters (#15172)
|
5 months ago |
 Judd
|
c24f4e2688
ggml : update `ggml_rope_multi` (#12665)
|
5 months ago |
 Copilot
|
d8914fc47e
common : add --override-tensor-draft, --cpu-moe-draft and --n-cpu-moe-draft parameters (#15191)
|
5 months ago |
|
Aldehir Rojas
|
e885445bc1
server : filter out harmony thought messages (#15278)
|
5 months ago |
 Ali Tariq
|
648ebcdb73
ci : Added CI with RISC-V RVV1.0 Hardware (#14439)
|
5 months ago |
|
Sigbjørn Skjæret
|
07aa869a91
ci : add more python requirements to copilot-setup-steps (#15289)
|
5 months ago |
|
Georgi Gerganov
|
00f35d509e
ggml : repack block_iq4_nlx8 (#14904)
|
5 months ago |
 Oliver Simons
|
6028bf7435
CUDA: Optimize `reduce_rows_f32` kernel, leading up to 25x perf improvement on kernel-level and 10% perf increase for Gemma3n (#15132)
|
5 months ago |
|
Sigbjørn Skjæret
|
bc5182272c
ci : add copilot-setup-steps.yml (#15214)
|
5 months ago |
 Tak-RS
|
e71d48e326
ggml-rpc: chunk send()/recv() to avoid EINVAL for very large tensors over RPC (macOS & others) (#15188)
|
5 months ago |
|
uvos
|
b0493156fa
HIP: disable sync warp shuffel operators from clr amd_warp_sync_functions.h (#15273)
|
5 months ago |
 Romain Biessy
|
f4586ee598
sycl: Fix and disable more configurations of mul_mat (#15151)
|
5 months ago |
 rmatif
|
60a7658810
opencl: allow mixed f16/f32 `add` (#15140)
|
5 months ago |
 Aman Gupta
|
efe3a90996
CUDA cmake: add `-lineinfo` for easier debug (#15260)
|
5 months ago |
 Chenguang Li
|
bbd57b7eaf
CANN: GGML_OP_CPY optimization (#15070)
|
5 months ago |
