ngxson/llama.cpp

Pull Requests Commits

server: move res_error/res_ok to static function

ngxson committed 1 year ago

4d0efa90

vendor: split httplib to cpp/h files (#17150)

ngxson committed 1 year ago

Verified 1d45b422

ggml-cpu : add RISC-V RVV (Zvfh) optimization for FP16 to FP32 conversion (#17161)

ixgbe committed 1 year ago

Verified ca484406

ggml-cpu: templateify ggml_compute_forward_rope_f32 and _f16 (#16805)

duduta committed 1 year ago

Verified 73460f62

kleidiai: add optimized per-channel kernels for Q8_0 (#16993)

chaxu01 committed 1 year ago

Verified 8c583242

cmake : add version to all shared object files (#17091)

furrysalamander committed 1 year ago

Verified 4a5b8aff

Install rpc-server when GGML_RPC is ON. (#17149)

nbp committed 1 year ago

Verified d2d62693

convert : register UMT5Model architecture for T5 conversion (#17160)

levkropp committed 1 year ago

Verified 2fc392ce

opencl: add fastdiv and use it in set_rows, ported from cuda (#17090)

lhez committed 1 year ago

Verified ece0f5c1

models : move build_inp_out_ids outside loop (#17151)

CISC committed 1 year ago

Verified 7bef6841

cpu: skip NOPs to avoid barriers (#17133)

max-krasnyansky committed 1 year ago

Verified 395e286b

metal : cap threadgroups size of set_rows (#17146)

ggerganov committed 1 year ago

Verified 13730c18

ggml-cpu : inspect -march and -mcpu to found the CPU (#16333)

angt committed 1 year ago

Verified 967eb4b2

vulkan: check glslc executable string (#17144)

0cc4m committed 1 year ago

Verified f117be18

vulkan: fix validation issue introduced by #16868 (#17145)

0cc4m committed 1 year ago

Verified 85234a4b

memory: Hybrid context shift (#17009)

gabe-l-hart committed 1 year ago

Verified 0c74f326

metal : enable tensor API for A19 (#17087)

ggerganov committed 1 year ago

Verified c27efd2b

arm64: add i8mm route with SVE ggml_vec_dot_q4_K_q8_K and ggml_vec_dot_q6_K_… (#15277)

fj-y-saito committed 1 year ago

Verified df70bedd

batched-bench : add "separate text gen" mode (#17103)

ggerganov committed 1 year ago

Verified f914544b

mtmd: fix patch_size initialized to random value in audio models (#17128)

ngxson committed 1 year ago

Verified 4b13a684

editorconfig : ignore benches/ (#17140)

ggerganov committed 1 year ago

Verified 9898b57c

cuda/vulkan : bicubic interpolation (#17022)

Acly committed 1 year ago

Verified 1032256e

benches : add eval results (#17139)

ggerganov committed 1 year ago

Verified 15274c0c

mtmd : fix embedding size for image input (#17123)

ggerganov committed 1 year ago

Verified b8595b16

vulkan: fix memory allocations (#17122)

0cc4m committed 1 year ago

Verified 392e09a6

convert : parse safetensors directly (#15667)

compilade committed 1 year ago

Verified 802cef44

convert : handle compressed-tensors quant method (#17069)

compilade committed 1 year ago

Verified 1c07c0c6

server : handle failures to restore host cache (#17078)

ggerganov committed 1 year ago

Verified cb1adf88

benches : add folder with benchmarks (#16931)

ggerganov committed 1 year ago

Verified ef1d8269

Switch to using Ubuntu 25.10 vulkan/mesa (#16497)

ericcurtin committed 1 year ago

Verified 86fde91e

Older