ggerganov/llama.cpp

Pull Requests Commits

ci : cache-first gg_wget to avoid re-downloads

ggerganov committed 1 hour ago

d205b894

ci : fix q4_0 publish when a cached model exists

ggerganov committed 2 hours ago

499c35cb

ci : clean up scratch models after tests

ggerganov committed 3 hours ago

b659e7ea

ci : fix data race between runners sharing the model cache

ggerganov committed 4 hours ago

01240ce1

ggml : fix time overflow in ggml_time_us (#30969)

frozenblade1224 committed 5 hours ago

Verified 4231b23d

ci : add behavior check workflow (#31036)

CISC committed 5 hours ago

Verified f00c9083

CANN: missing `const` qualifiers added for consistency (#30933)

rauletorresc committed 11 hours ago

Verified f5027fb2

sycl : pass device into mul_mat_uses_onednn (#31014)

Titaniumtown committed 12 hours ago

Verified f64a4981

Extend probabilistic sampling for DFlash2 drafts (#30663)

praneshgo committed 13 hours ago

Verified 1ab09d63

vocab : fix Nemotron-3.5-Embed tokenization (#30974)

tokinasin committed 13 hours ago

Verified f9599d0c

hexagon: Op Fusion refactoring, optimizations and new fusions like `BINARY_NX` (back-to-back ADDs, etc) (#30843)

max-krasnyansky committed 14 hours ago

Verified 643b0cda

sycl: Replace host waits in fattn-mkl with explicit dependencies (#30684)

malsbat committed 16 hours ago

Verified 0a9ef006

sycl: Add graph record and replay (#28725)

malsbat committed 16 hours ago

Verified 7377bada

server: Substitute rerank placeholder from the original template only. (#30864)

Pengkun-ZHU committed 20 hours ago

Verified 184f6536

server: realtime API: send complete after committed audio drains (#30994)

ngxson committed 20 hours ago

Verified d814cbb8

opencl: add depthwise conv2d (#30900)

kurquhar committed 1 day ago

Verified 970b51ce

mtmd : add MiniCPM-OCR 1.0 support (#30979)

tc-mb committed 1 day ago

Verified a08ab75a

clef : include <algorithm> for std::find and std::copy (#30981)

kleinbem committed 1 day ago

Verified 19ab5a5e

gaugarg-nv committed 1 day ago

Verified 36cbf415

glm5 : optimize inputs (#30932)

ggerganov committed 1 day ago

Verified 5c19a5a7

model : fix qwen4exp layer inputs for DFlash drafters (#30927)

ggerganov committed 1 day ago

Verified d9ee007a

Multi-column on Intel XMX matrix engines refactored (#29864)

budiony committed 1 day ago

Verified d81d0d8f

vulkan: replace std::aligned_alloc with posix_memalign for termux compatibility (#30828)

bill88t committed 1 day ago

Verified d7725699

nemotron : support non-standard GQA, BPE tokenizer and domyn-small pre-type (#23343)

martin-cimmino committed 1 day ago

Verified 0693dbfb

server : support typed content input for RERANK endpoint (#29625)

timothywang21 committed 1 day ago

Verified a5d22276

ci : disable slow WebGPU tests (#30963)

ggerganov committed 1 day ago

Verified 5f5acfe2

ci : be able to reconfigure HTTP port for server pytest workers (#30972)

ggerganov committed 1 day ago

Verified 1f4c2eea

sycl: fix CONV_2D wrong results for non-contiguous input (#30363)

mctylr-gh committed 1 day ago

Verified 7170cad4

vocab : add clean_spaces GGUF key (#30940)

tkran2 committed 1 day ago

Verified d88040c7

vocab : use Tekken pre-tokenizer for pixtral rather than the Llama 3 one (#30960)

tkran2 committed 1 day ago

Verified e197cef4

Older