ngxson/llama.cpp

Pull Requests Commits

add comment to clarify about mtmd_image_preprocessor_dyn_size

ngxson committed 2 days ago

d58d9798

fix deepseek-ocr on server

ngxson committed 2 days ago

eae39303

ngxson committed 2 days ago

ed673769

correct some places

ngxson committed 2 days ago

f48a34ab

mtmd: refactor image pre-processing

ngxson committed 2 days ago

906b46e9

ggml-cuda: Add NVFP4 dp4a kernel (#20644)

michaelw9999 committed 2 days ago

Verified 112c7815

imatrix : fix crash when using --show-statistics with zero counts (#19532)

ssam18 committed 2 days ago

Verified 0fac87b1

CUDA & CPU: support F32 kernel type for `CONV_TRANSPOSE_2D` (#17094)

AgainstEntropy committed 2 days ago

Verified 0a524f24

common : do not delete old files from the old cache when updating (#21000)

angt committed 3 days ago

Verified c0159f9c

mtmd: Add DeepSeekOCR Support (#17400)

sfallah committed 3 days ago

Verified a970515b

common : fix verbosity setup (#20989)

angt committed 3 days ago

Verified 056b50c3

common : fix gguf selection in common_list_cached_models (#20996)

angt committed 3 days ago

Verified f2c72b8f

ci : fix parsing of vgpr counts in hip-quality-check (#20987)

IMbackK committed 3 days ago

Verified ec54ac13

model: codefuse-ai/F2LLM-v2 support

sfallah committed 3 days ago

Verified 80322ebd

model : allow causal_attn and pooling_type on all architectures (#20973)

Bing-su committed 3 days ago

Verified 44c51e52

snapdragon: add missing features to WoS scripts to achieve parity with ADB scripts (#20884)

aparmp-quic committed 3 days ago

Verified 1922f87c

Use docker in build-android.yml (#20928)

shreyajn committed 3 days ago

Verified 345de3cd

llama-bench: print `-n-cpu-moe` when offloaded layers > 1 (#20984)

am17an committed 3 days ago

Verified 9c600bcd

ci: Allow ninja to be used during unit test (#20742)

rillomas committed 3 days ago

Verified b2704f90

ci : disable self-hosted mac jobs (#20985)

ggerganov committed 3 days ago

Verified 3fab96cd

jinja: fix macro with kwargs (#20960)

ngxson committed 3 days ago

Verified 914eb5ff

gguf-split : clarify operation of gguf-split (#19749)

alosslessdev committed 3 days ago

Verified 8fc17493

llama: fix llama-model-saver (#20503)

JohannesGaessler committed 3 days ago

Verified 36dafba5

webui: Fix editing assistant message without branching (#20944)

allozaur committed 3 days ago

Verified 69e0ecef

Add SLEEPING status to the WebUI model selector (#20949)

ServeurpersoCom committed 3 days ago

Verified 062cca58

android : fix-pointer-dangling (#20974)

yikechayedan committed 3 days ago

Verified 406f4e3f

sycl : fix wrong variable check by assert (#20903)

arthw committed 3 days ago

Verified 53dc8b59

ci : bump gguf publish python version (#20982)

CISC committed 3 days ago

Verified 403c9c9c

ci : limit requirements versions (#20980)

CISC committed 3 days ago

Verified 8fc85db9

convert : register Qwen3Model architecture (#20967)

Bing-su committed 3 days ago

Verified 3a60d06a

Older