LLM inference in C/C++
C++ ggml
Commits on master
mamba2 : Flatten in/out projections to dispatch GEMM instead of GEMV (#27513)
a130532 Prabhsimran Singh committed 2 months ago
mtmd: use pillow-accurate algo, correct resize_algo for all models (#27594)
56db501 Xuan-Son Nguyen committed 2 months ago
ci : add test-llama-archs tensor split for Metal (#27598)
95b8e33 Georgi Gerganov committed 2 months ago
contrib : recommend waiting for CI before merging (#27603)
a278dce Niklas Wenzel committed 2 months ago
common : skip device_info loop if it's not going to be printed (#26692)
ba8e0ed Bartosz Taudul committed 2 months ago
[Tensor parallel] Fix meta tensor split state propagation (#27574)
d337192 Gaurav Garg committed 2 months ago
vendor : update subprocess.h (#27409)
6657ded Alessandro de Oliveira Faria (A.K.A.CABELO) committed 2 months ago
ci : add older, min and dry-run options to ccache-clear (#27504)
3a653fe Georgi Gerganov committed 3 months ago
ggml: optimize concat op by replacing per-element memcpy with row-level memcpy (#24575)
369e1cd Kartik Sirohi committed 3 months ago
common : fix draft-mtp with embeddings (#26352, #27299) (#27400)
2c6b141 Shahir BIn Zulfiker committed 3 months ago
sycl : add Q2_K reordered MMVQ and ESIMD kernels (again) (#27490)
8672290 Sigbjørn Skjæret committed 3 months ago