ggml-org/llama.cpp

Pull Requests Commits

moe-cache : remove the backend logic of the CPU part from the graph

am17an committed 36 minutes ago

5815ca62

scripts : replay the ClassLRU policy in moe-cache-tune.py

am17an committed 12 hours ago

0290c8da

llama-graph : make llm_graph_params::moe_cache const

ggerganov committed 12 hours ago

fed73adc

llama-graph : extract the CPU share condition into can_cpu_share

am17an committed 12 hours ago

0bd7328e

remove sync change

am17an committed 12 hours ago

740fdfc1

Fix MSVC compilation

gaugarg-nv committed 12 hours ago

b3a17ff5

moe-cache : let the scheduler copy the results of the CPU part

am17an committed 12 hours ago

f75a5af2

add backend sync for async transfer queue

am17an committed 12 hours ago

a571975b

moe-cache : upload all the misses until the cache is full

am17an committed 12 hours ago

fab91ac8

add am17an as codeowner for MoE cache

am17an committed 12 hours ago

9a2e1677

scripts : add moe-cache-tune.py to pick the MoE cache arguments

am17an committed 12 hours ago

85ed5474

moe-cache : compute part of the misses on the CPU

am17an committed 12 hours ago

b0d27534

cli : fix crash on invalid UTF-8 in prompt or files (#31563)

SofianElmotiem committed 17 hours ago

Verified 6c01da62

tests : qualify isnan and signbit with std:: (#31791)

ggerganov committed 17 hours ago

Verified 8820b088

tests : override hardware_concurrency in CI (#31797)

ggerganov committed 17 hours ago

Verified 9a1d03c7

cmake : generate RC file directly without win-app.rc.in (#31718)

angt committed 19 hours ago

Verified 18ab62a8

opencl: support repeat for f16, i32, i16, bf16 (#31782)

lhez committed 20 hours ago

Verified 52c9bad9

webgpu: add batched tensor API (`get_tensor_n` and `set_tensor_n`) (#31473)

yomaytk committed 20 hours ago

Verified 5d3ed2ae

vulkan: detect ReBAR support, avoid BAR allocations if unavailable (#31658)

0cc4m committed 20 hours ago

Verified 80088fa6

moe-cache : LRFU/MQ-like hybrid cache, ClassLRU (#31233)

sanmai committed 20 hours ago

Verified c4fb5796

ci : bump CUDA 13 release builds to 13.5 (#31788)

sam-india-007 committed 21 hours ago

Verified 3ed7284f

vendor: update BoringSSL to 0.20261105.0 (#31757)

cabelo committed 21 hours ago

Verified d9cf1870

CUDA: fix Maxwell build (#31787)

JohannesGaessler committed 21 hours ago

Verified 240d68fc

common : restrict HF token to huggingface.co hosts (#31730)

angt committed 22 hours ago

Verified bbc2e4b6

opencl: add q6_k multi column tiled GEMV for speculative verification and refactoring (#31758)

shawngu-quic committed 1 day ago

Verified 4b77dee0

hexagon: Implement set_tensor_n and improve repacking/tiling robustness (#31744)

max-krasnyansky committed 1 day ago

Verified 67107ac3

mtmd: enable cmake unity build (#31760)

ngxson committed 1 day ago

Verified e14a7797

tests : drop u8path() (#31755)

CISC committed 1 day ago

Verified d38bea34

ci: fix build-cmake-pkg ccache (#31705)

netrunnereve committed 1 day ago

Verified 186eab1c

llama: fix kv shifting, add tests, add per-model build_rope (#31683)

ngxson committed 1 day ago

Verified 6023692b

Older