 Georgi Gerganov
|
85a7d8677b
memory : remove KV cache size padding (#16812)
|
hai 2 meses |
|
Georgi Gerganov
|
a8ca18b4b8
llama-bench : clarify benchmarked parts of the computation (#16823)
|
hai 2 meses |
 l3utterfly
|
8284efc35c
initialise buffer.device in ggml_hexagon_session (#16816)
|
hai 2 meses |
 Sam Malayek
|
1c1409e131
embedding: add raw option for --embd-output-format (#16541)
|
hai 2 meses |
 Johannes Gäßler
|
7a0e900e36
llama: consistent ctx <-> buf order for KV cache (#16746)
|
hai 2 meses |
 Aldehir Rojas
|
280d97be96
grammar : support array references in json schema (#16792)
|
hai 2 meses |
 Chenguang Li
|
3479efd112
CANN: Improve device ID handling and aclnnArange checks (#16752)
|
hai 2 meses |
 Aman Gupta
|
463bbf20bf
CUDA: add unused vars to mmvf and mmvq (#16807)
|
hai 2 meses |
 tamarPal
|
ad8d36beff
sycl: add SSM_CONV operation support (#16800)
|
hai 2 meses |
 Yuri Khrustalev
|
c053e18a66
chat: Add LFM2 tool handling (#16763)
|
hai 2 meses |
 Xuan-Son Nguyen
|
e1ab084803
mtmd : fix idefics3 preprocessing (#16806)
|
hai 2 meses |
 Diego Devesa
|
5a4ff43e7d
llama : disable pipeline parallelism if compute buffer allocation fails (#16748)
|
hai 2 meses |
 Acly
|
10640e31aa
ggml : fix interpolate with align-corners and ne=1 (#16700)
|
hai 2 meses |
|
Johannes Gäßler
|
80d28f104c
HIP: fix AMDGPU_TARGETS, update documentation (#16803)
|
hai 2 meses |
|
Xuan-Son Nguyen
|
c55d53acec
model : add LightOnOCR-1B model (#16764)
|
hai 2 meses |
|
Johannes Gäßler
|
945501f5ea
llama: fix leaked buffers for mmap + split files (#16765)
|
hai 2 meses |
|
Aman Gupta
|
75cbdd3fce
test-backend-ops: print failed tests at the end (#16785)
|
hai 2 meses |
|
tamarPal
|
2b9bd9bf4e
sycl: add ROLL operation support (#16665)
|
hai 2 meses |
 shani-f
|
59fc1ec8e8
sycl: add REPEAT_BACK operation support (#16734)
|
hai 2 meses |
|
Aman Gupta
|
75d33b9302
CUDA: support for weight clamp in top-k norm (#16702)
|
hai 2 meses |
|
Acly
|
3470a5c891
ggml-alloc : make gallocr prefer chunks that allow memory reuse (#16788)
|
hai 2 meses |
 Sigbjørn Skjæret
|
bd562fe4f7
cuda : use fast copy when src and dst are of different type and contiguous (#16789)
|
hai 2 meses |
 leejet
|
bbac6a26b2
ggml: fix cuda kernel launch configuration for k_compute_batched_ptrs to support large batch (#16744)
|
hai 2 meses |
|
Sigbjørn Skjæret
|
73a48c9790
convert : enable expert group selection for all models with it (#16691)
|
hai 2 meses |
|
Sigbjørn Skjæret
|
f696428ce8
graph : add clamping to ffn_moe_weights_sum to avoid div-by-zero (#16655)
|
hai 2 meses |
|
Sigbjørn Skjæret
|
7cce4f8158
model : set res->t_embd in SmallThinker models (#16782)
|
hai 2 meses |
 amirai21
|
8d8862829c
docs : add Jamba to Text-only models list (#16778)
|
hai 2 meses |
|
Aman Gupta
|
f77c13b91f
CUDA: General GEMV fusion (#16715)
|
hai 2 meses |
 Gilad S.
|
3cfa9c3f12
vulkan: deduplicate Microsoft Direct3D12 devices (#16689)
|
hai 2 meses |
 Galunid
|
5d195f17bc
convert : handle mmproj filename/path properly (#16760)
|
hai 2 meses |
