ngxson/llama.cpp

Pull Requests Commits

fix compilation on linux

ngxson committed 289 days ago

61d8ad6a

ngxson committed 289 days ago

ce83041e

ngxson committed 289 days ago

e185e0ac

ngxson committed 289 days ago

e618405d

ngxson committed 289 days ago

43bf237e

add mimi_model::transpose_input

ngxson committed 289 days ago

eae5f0e1

mimi : non-transposed input codes

ngxson committed 290 days ago

891273cf

put mimi_model into a shared header

ngxson committed 290 days ago

a98f1991

ngxson committed 290 days ago

efeaa571

tts : implement mimi decoder

ngxson committed 291 days ago

24a07ab6

llama : fix incorrect Qwen2Moe ffn_moe_out graph callback (#12631)

CISC committed 291 days ago

Verified 3714c3ee

metal : improve FA + improve MoE (#12612)

ggerganov committed 291 days ago

Verified b4ae5081

vulkan: fix coopmat shader generation when cross-compiling (#12272)

Icenowy committed 291 days ago

Verified b86f6007

llama: fix error on bad grammar (#12628)

JohannesGaessler committed 291 days ago

Verified dd373dd3

server : include speculative decoding stats when timings_per_token is enabled (#12603)

mostlygeek committed 291 days ago

Verified 5d016702

rpc : update README for cache usage (#12620)

rgerganov committed 291 days ago

Verified ef03229f

llamafile : ppc64le GEMV forwarding for FP32. (#12594)

amritahs-ibm committed 291 days ago

Verified 13731766

rpc : send hash when tensor data is above some fixed threshold (#12496)

rgerganov committed 291 days ago

Verified ab6ab8f8

server : Support listening on a unix socket (#12613)

Piotr committed 292 days ago

Verified 2099a9d5

media : add SVG logo [no ci] (#12616)

ggerganov committed 292 days ago

Verified 29690198

opencl: add multi and vision rope, `gelu_quick` and `im2col` (#12600)

lhez committed 292 days ago

Verified 5dec47dc

llama : add PLM GGUF Conversion & Inference Support (#12457)

Si1w committed 292 days ago

Verified f125b8dc

model : restore support for T5Encoder (#12590)

HighDoping committed 292 days ago

Verified 953c2a62

convert : Support Qwen2_5_VLForConditionalGeneration (#12595)

csabakecskemeti committed 292 days ago

Verified d5c6309d

ggerganov committed 292 days ago

029c693f

scripts : update sync + fix cmake merge

ggerganov committed 292 days ago

771d8437

ggerganov committed 292 days ago

df0665a4

cmake : sync/merge PowerPC build commands (#0)

ggerganov committed 292 days ago

0306aad1

llamafile : ppc64le MMA implementation for Q4_0. (#12489)

amritahs-ibm committed 292 days ago

Verified c7b43ab6

ggml : riscv: add 128-bit RVV support (#12530)

xctan committed 292 days ago

Verified 24feaec0

Older