Tags

Tags give the ability to mark specific points in history as being important

b3135

73bac2b1 · vulkan: select only one device for single gpu with multiple drivers (#7582) · Jun 11, 2024
b3134

ef52d1d1 · Update Vulkan RoPE implementation (#7818) · Jun 11, 2024
b3131

148995e5 · llama-bench: more compact markdown tables (#7879) · Jun 11, 2024
b3130

4bfe50f7 · tests : check the Python version (#7872) · Jun 11, 2024
b3091

2b338967 · ggml : refactor rope norm/neox (#7634) · Jun 05, 2024
b3089

c90dbe02 · Fix per token atrributes bits (#7749) · Jun 05, 2024
b3088

b90dc566 · Allow number of nodes in CUDA graph to change (#7738) · Jun 04, 2024
b3087

1442677f · common : refactor cli arg parsing (#7675) · Jun 04, 2024
b3086

554c247c · ggml : remove OpenCL (#7735) · Jun 04, 2024
b3085

0cd6bd34 · llama : remove beam search (#7736) · Jun 04, 2024
b3083

adc9ff38 · llama-bench : allow using a different printer for stderr with -oe (#7722) · Jun 04, 2024
b3082

987d743d · Improve hipBLAS support in CMake (#7696) · Jun 04, 2024
b3080

3b38d486 · Per token attributes (#7685) · Jun 04, 2024
b3079

6d161694 · ggml : prevent builds with -ffinite-math-only (#7726) · Jun 04, 2024
b3078

bde7cd3c · llama : offload to RPC in addition to other backends (#7640) · Jun 03, 2024
b3077

a5735e44 · ggml : use OpenMP as a thread pool (#7606) · Jun 03, 2024
b3076

0b832d53 · make: fix debug options not being applied to NVCC (#7714) · Jun 03, 2024
b3075

3d7ebf63 · Vulkan Mixture of Experts (MoE) support (#7628) · Jun 03, 2024
b3074

a10cda58 · cmake : add pkg-config spec file for llama.cpp (#7702) · Jun 03, 2024
b3073

6f28a333 · llama : MiniCPM support tied embeddings (#7664) · Jun 03, 2024

Prev
1
2
3
4
5
…
99
Next