ggerganov/llama.cpp

Pull Requests Commits

server: abstract llama_memory calls to common_memory

ngxson committed 25 days ago

f45ca59f

sycl: fix use-after-return of the SDPA scale in the oneDNN flash-attention path (#25880)

meatposes committed 25 days ago

Verified d6b61ac0

server : add extra trace log for prompt similarity (#26218)

ggerganov committed 25 days ago

Verified 9a3bf2b8

ggml-metal: FWHT kernel for metal backend (#25924)

nicklafleur committed 25 days ago

Verified f95de977

spec: add eagle3-v3 support for gpt-oss model (#25794)

ruixiang63 committed 25 days ago

Verified f8706784

contrib : add guideline about the "merge ready" label (#26178)

ggerganov committed 25 days ago

Verified c6292cfb

Disable -ffast-math on HIP (#25495)

Beinsezii committed 26 days ago

Verified 91f8c9c5

mtmd: support MiMo-V2.5 audio input (RVQ-based model) (#26190)

ngxson committed 26 days ago

Verified 1cbfd198

common : add common_print_available_devices() (#26170)

angt committed 26 days ago

Verified 0e4a0362

model: Add support for Nanbeige4.2 (#25994)

zqlcode committed 26 days ago

Verified b77d6467

fit : count nextn (MTP) blocks in n_gpu_layers so front layers stay on GPU (#26177)

John-194 committed 26 days ago

Verified 0324696b

sycl(build): parallelize ocloc invocations (#25903)

Titaniumtown committed 26 days ago

Verified 8e8681e0

ggml : adjust logic for offloading ops to weight's backend (#25832)

ggerganov committed 26 days ago

Verified dee2a846

tests : remove unnecessary sync in test-save-load-state (#26166)

ggerganov committed 26 days ago

Verified 7ef790f9

common: fix explicit -md precedence over draft sidecar resolution (#26165)

ServeurpersoCom committed 26 days ago

Verified ddfc2288

docs: add exception about weight folding (#26168)

ngxson committed 26 days ago

Verified 419b881c

ggml-cpu: Enable BF16 tiled gemm optimization on PowerPC (#26068)

shalinib-ibm committed 26 days ago

Verified b9102008

args: add `-lm mlock` where it mlocks but doesnt mmap (#26135)

taronaeo committed 26 days ago

Verified ad256ded

server + ui: fix stream routes for model names containing a slash (#26137)

ServeurpersoCom committed 26 days ago

Verified d73c1d6b

ui: Fix symbolic math tool JS sandbox prompt (#26131)

rankaiyx committed 27 days ago

Verified 88b47a75

mtmd: Add Vision Support for Minimax-M3 (#25113)

timkhronos committed 27 days ago

Verified 3d1c3a89

mtmd: fix android build (#26150)

ngxson committed 27 days ago

Verified 0d47ea74

ui: fix system message edit box not expanding to fit content (#26006)

pieroevcc committed 27 days ago

Verified d4d057b6

Keep Minimax's indexer tensors at F32 for speed and accuracy (#26144)

bartowski1182 committed 27 days ago

Verified 7657a6c2

ui: detect the conversation import format from file contents (#26121)

ServeurpersoCom committed 27 days ago

Verified 55b7d6c4

common: add `subproc.h` wrapper, disabled on android/ios (#26102)

ngxson committed 27 days ago

Verified d2a81823

mtmd: add GLM-5.2-Vision (#26126)

ehartford committed 27 days ago

Verified af285020

model: Add MiniMax-M3 (MSA: MiniMax Sparse Attention) support (#24908)

timkhronos committed 27 days ago

Verified b1d4c655

opencl: fix fused RMS norm mul view offset (#26085)

happyyzy committed 27 days ago

Verified 42fc2430

ui: fix context gauge card regressions and land at the conversation end (#26099)

ServeurpersoCom committed 28 days ago

Verified ff067f76

Older