 Jay
|
a69f846351
cmake : fix ccache conflict (#12522)
|
9 kuukautta sitten |
 hipudding
|
d07a0d7a79
CANN : remove clang-format in ggml-cann (#12607)
|
9 kuukautta sitten |
 Sigbjørn Skjæret
|
3714c3ee1a
llama : fix incorrect Qwen2Moe ffn_moe_out graph callback (#12631)
|
9 kuukautta sitten |
 Georgi Gerganov
|
b4ae50810e
metal : improve FA + improve MoE (#12612)
|
9 kuukautta sitten |
 Icenowy Zheng
|
b86f600723
vulkan: fix coopmat shader generation when cross-compiling (#12272)
|
10 kuukautta sitten |
 Johannes Gäßler
|
dd373dd3bf
llama: fix error on bad grammar (#12628)
|
10 kuukautta sitten |
 Benson Wong
|
5d01670266
server : include speculative decoding stats when timings_per_token is enabled (#12603)
|
10 kuukautta sitten |
 Radoslav Gerganov
|
ef03229ff4
rpc : update README for cache usage (#12620)
|
10 kuukautta sitten |
 amritahs-ibm
|
13731766db
llamafile : ppc64le GEMV forwarding for FP32. (#12594)
|
10 kuukautta sitten |
|
Radoslav Gerganov
|
ab6ab8f809
rpc : send hash when tensor data is above some fixed threshold (#12496)
|
10 kuukautta sitten |
 Piotr
|
2099a9d5db
server : Support listening on a unix socket (#12613)
|
10 kuukautta sitten |
|
Georgi Gerganov
|
2969019837
media : add SVG logo [no ci] (#12616)
|
10 kuukautta sitten |
 lhez
|
5dec47dcd4
opencl: add multi and vision rope, `gelu_quick` and `im2col` (#12600)
|
10 kuukautta sitten |
 Si1w
|
f125b8dccf
llama : add PLM GGUF Conversion & Inference Support (#12457)
|
10 kuukautta sitten |
 HighDoping
|
953c2a62cf
model : restore support for T5Encoder (#12590)
|
10 kuukautta sitten |
 Csaba Kecskemeti
|
d5c6309d91
convert : Support Qwen2_5_VLForConditionalGeneration (#12595)
|
10 kuukautta sitten |
|
Georgi Gerganov
|
029c693fdc
sync : ggml
|
10 kuukautta sitten |
|
Georgi Gerganov
|
771d84371c
scripts : update sync + fix cmake merge
|
10 kuukautta sitten |
|
Georgi Gerganov
|
df0665a483
sync : ggml
|
10 kuukautta sitten |
|
Georgi Gerganov
|
0306aad1ca
cmake : sync/merge PowerPC build commands (#0)
|
10 kuukautta sitten |
|
amritahs-ibm
|
c7b43ab608
llamafile : ppc64le MMA implementation for Q4_0. (#12489)
|
10 kuukautta sitten |
 xctan
|
24feaec057
ggml : riscv: add 128-bit RVV support (#12530)
|
10 kuukautta sitten |
|
Georgi Gerganov
|
f28bc4c286
llama : make loras compatible with repacking (#12593)
|
10 kuukautta sitten |
 Akarshan Biswas
|
f17a3bb4e8
SYCL: implement memset ggml backend buffer interface (#12580)
|
10 kuukautta sitten |
 Slobodan Josic
|
bd40678df7
HIP: Add support for RDNA4 targets (#12372)
|
10 kuukautta sitten |
|
Georgi Gerganov
|
b3298fa47a
metal : refactor mat-vec code (#12569)
|
10 kuukautta sitten |
 Michał Moskal
|
2447ad8a98
upgrade to llguidance 0.7.10 (#12576)
|
10 kuukautta sitten |
 Ivy233
|
02082f1519
clip: Fix llama-llava-clip-quantize-cli quantization error under CUDA backend (#12566)
|
10 kuukautta sitten |
|
Georgi Gerganov
|
df4d20cd53
convert : fix squeeze for ssm_conv tensors (#12573)
|
10 kuukautta sitten |
|
Georgi Gerganov
|
5ed38b6852
ggml : fix MUL_MAT_ID repack with Q8_K (#12544)
|
10 kuukautta sitten |
