 Johannes Gäßler
|
4227c9be42
CUDA: fix negative KV_max values in FA (#15321)
|
há 5 meses atrás |
 Georgi Gerganov
|
df36bce667
eval-callback : stop on first NaN (#15320)
|
há 5 meses atrás |
 Diego Devesa
|
f75b830647
chat : include kwargs in template example (#15309)
|
há 5 meses atrás |
 Daniel Bevenius
|
7a0de96045
llama : add 18-layer model type for Gemma 3-270m (#15319)
|
há 5 meses atrás |
 simevo
|
e4e915912c
devops : fix compile bug when the BASE_CUDA_DEV_CONTAINER is based on Ubuntu 24.04 (#15005)
|
há 5 meses atrás |
 uvos
|
5ba36f6103
HIP: Cleanup hipification header (#15285)
|
há 5 meses atrás |
 Aldehir Rojas
|
b204a5a234
gpt-oss: implement harmony parsing (#15181)
|
há 5 meses atrás |
 Christian Kastner
|
646944cfa8
docker : Enable GGML_CPU_ALL_VARIANTS for ARM (#15267)
|
há 5 meses atrás |
|
Georgi Gerganov
|
1a01899b61
readme : update hot topics (#15315)
|
há 5 meses atrás |
 Jeff Bolz
|
863d341eeb
vulkan: perf_logger improvements (#15246)
|
há 5 meses atrás |
|
Georgi Gerganov
|
d32e03f449
server : add SWA checkpoints (#15293)
|
há 5 meses atrás |
|
Georgi Gerganov
|
3973163bff
sync : ggml
|
há 5 meses atrás |
 Jason Ni
|
5ade3000bd
ggml: fix ggml_conv_1d_dw bug (ggml/1323)
|
há 5 meses atrás |
|
Georgi Gerganov
|
8b2483730f
tests : remove unused includes (ggml/0)
|
há 5 meses atrás |
 kallewoof
|
810b9fc8b9
perplexity : provide a helpful hint for has_cpl case in split_equal error. (#15304)
|
há 5 meses atrás |
 Sigbjørn Skjæret
|
4ebd0c125b
cuda : fix GGML_CUDA_GRAPHS=OFF (#15300)
|
há 5 meses atrás |
 Jonathan Graehl
|
5cdb27e091
finetune: SGD optimizer, more CLI args (#13873)
|
há 5 meses atrás |
|
kallewoof
|
3ea913f1ce
perplexity: give more information about constraints on failure (#15303)
|
há 5 meses atrás |
|
uvos
|
29c8fbe4e0
HIP: bump requirement to rocm 6.1 (#15296)
|
há 5 meses atrás |
 Bas Nijholt
|
1adc9812bd
fix(nix): remove non-functional llama-cpp cachix cache from flake.nix (#15295)
|
há 5 meses atrás |
|
Sigbjørn Skjæret
|
b3e16665e1
server : enable -td and -tbd parameters (#15172)
|
há 5 meses atrás |
 Judd
|
c24f4e2688
ggml : update `ggml_rope_multi` (#12665)
|
há 5 meses atrás |
 Copilot
|
d8914fc47e
common : add --override-tensor-draft, --cpu-moe-draft and --n-cpu-moe-draft parameters (#15191)
|
há 5 meses atrás |
|
Aldehir Rojas
|
e885445bc1
server : filter out harmony thought messages (#15278)
|
há 5 meses atrás |
 Ali Tariq
|
648ebcdb73
ci : Added CI with RISC-V RVV1.0 Hardware (#14439)
|
há 5 meses atrás |
|
Sigbjørn Skjæret
|
07aa869a91
ci : add more python requirements to copilot-setup-steps (#15289)
|
há 5 meses atrás |
|
Georgi Gerganov
|
00f35d509e
ggml : repack block_iq4_nlx8 (#14904)
|
há 5 meses atrás |
 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)
|
há 5 meses atrás |
|
Sigbjørn Skjæret
|
bc5182272c
ci : add copilot-setup-steps.yml (#15214)
|
há 5 meses atrás |
 Tak-RS
|
e71d48e326
ggml-rpc: chunk send()/recv() to avoid EINVAL for very large tensors over RPC (macOS & others) (#15188)
|
há 5 meses atrás |
