ggerganov/llama.cpp

Pull Requests Commits

kleidiai : add Q8_0 SVE dotprod GEMV and SVE i8mm GEMM kernels (#30808)

chaxu01 committed 1 day ago

Verified dfeea520

sycl: merge the fp16 and fp32 dequant switches (#30740)

Titaniumtown committed 1 day ago

Verified e3797b18

webgpu : add Q2_0 support (#30914)

teddygood committed 1 day ago

Verified 0b1281ae

dflash : load optional FC input scale (#30945)

hthadicherla committed 1 day ago

Verified 4912ac24

ui: Add "Show tool call in progress" setting (#30681)

allozaur committed 1 day ago

Verified 84746c5c

lora : support prism.hadamard folded weights (#30910)

vieduy committed 1 day ago

Verified 0682a576

ui : key draft sidecar badges by model instead of repo (#30919)

allozaur committed 1 day ago

Verified a6b83d42

opencl: add F32-to-I32 ARGMAX (#30897)

kurquhar committed 1 day ago

Verified c7fc5094

models : add Flint model (FlintForCausalLM) (#30928)

alexey-rsi committed 1 day ago

Verified a37cc461

jinja : fix namespace attributes shadowed by dict methods (#30941)

tdakhran committed 1 day ago

Verified 99fb9971

model: Add support for SenseNova U1 model (#28919)

Maphist0 committed 1 day ago

Verified 7775ffb7

llama: create new fused op for QSA mask (#30640)

am17an committed 1 day ago

Verified d454a204

sycl: q8 fast dequant (#30775)

92es committed 2 days ago

Verified 0756e83f

jinja: support hexadecimal string escapes (#30929)

alexey-rsi committed 2 days ago

Verified 9febe63d

ggml-opencl : add IM2COL_3D support (#30895)

kurquhar committed 2 days ago

Verified 9b56ccf0

mtmd: add state object for mtmd_gen (#30902)

ngxson committed 2 days ago

Verified 0a1ef309

server : forward llama.cpp sampling params on /v1/messages (#30856)

Dgarc359 committed 2 days ago

Verified 261700b8

vendor: update BoringSSL to 0.20261020.0 (#30889)

cabelo committed 2 days ago

Verified f8490953

llama : optimize vocabulary loading (#30679)

mykhailen committed 2 days ago

Verified 824eaecf

llama : place d2t by its set_rows use so it can stay on the GPU (#30912)

am17an committed 2 days ago

Verified eb182d85

metal : fix memory leaks and cleanup logic in GPU capture path (#30890)

nikwen committed 2 days ago

Verified ea9e3446

hexagon: neon-vectorized and threaded repack (tiling) (#30901)

max-krasnyansky committed 2 days ago

Verified 0efe8e4c

opencl: add dp4a GEMM variants to vector load activation (#30753)

lhez committed 2 days ago

Verified b531e4c7

sycl : fix fa crash for Query head dimension mismatch (neq0 != HSK / D=2048/4096) (#30286)

arthw committed 2 days ago

Verified 74b115ca

sycl: collapse duplicate need_check launch bodies in mmq (#30739)

Titaniumtown committed 2 days ago

Verified 96caa800

ggml-cpu : compute f32 gelu_quick and geglu_quick without the fp16 table (#30737)

boshjerns committed 2 days ago

Verified 479a5277

llama : scale moe down projection per expert token (#30861)

ggerganov committed 2 days ago

Verified 8baabd53

Sycl iq3 crash fix (#30604)

anantshri committed 2 days ago

Verified 493253fe

opencl: fix rope program build error with newer Clang (#30893)

bashbaug committed 2 days ago

Verified 339f3242

server : fix usage.completion_tokens undercount with n > 1 (#29486)

unbug committed 2 days ago

Verified 9a3b91c6

Newer Older