Tags
Tags give the ability to mark specific points in history as being important
master-e95b655
e95b6554
·
ggml : add Q8_0 quantization for intermediate results (#951)
·
Apr 15, 2023
master-aa485ce
aa485cee
·
ggml : use posix_memalign on non-Windows env
·
Apr 15, 2023
master-c12b14b
c12b14b7
·
benchmark : fix result validation in benchmark-q4_0-matmult (#987)
·
Apr 15, 2023
master-106faaf
106faaf2
·
cmake : add finding the OpenBLAS header file (#992)
·
Apr 15, 2023
master-c85e03d
c85e03d1
·
Revert "main : alternative instruct mode (Vicuna support, etc.) (#863)" (#982)
·
Apr 14, 2023
master-93265e9
93265e98
·
make : fix dependencies, use auto variables (#983)
·
Apr 14, 2023
master-c56b715
c56b7152
·
Expose type name from ggml (#970)
·
Apr 14, 2023
master-f4d277a
f4d277ae
·
main : alternative instruct mode (Vicuna support, etc.) (#863)
·
Apr 14, 2023
master-c9a59b7
c9a59b70
·
ggml : add unary and binary map operations (#874)
·
Apr 14, 2023
master-1623a6e
1623a6e9
·
ggml : minor
·
Apr 14, 2023
master-0f07cac
0f07cacb
·
ggml : fix q4_1 dot product types
·
Apr 14, 2023
master-c5d70f5
c5d70f5c
·
ggml : optimize rope function to avoid call powf in the tight loop (#807)
·
Apr 14, 2023
master-be87b6e
be87b6ed
·
perplexity : add support for batch size to `--perplexity` (#407)
·
Apr 14, 2023
master-0e07e6a
0e07e6a8
·
common : remove unnecessary includes (#947)
·
Apr 13, 2023
master-a3a2a0e
a3a2a0ed
·
ggml : add GGML_DEFAULT_N_THREADS
·
Apr 13, 2023
master-d990e3f
d990e3ff
·
ggml : speed-up ggml_vec_dot_q4_1() ARM_NEON + 32-bit ARM support (#900)
·
Apr 13, 2023
master-9190e8e
9190e8ea
·
llama : merge llama_internal.h into llama.h
·
Apr 13, 2023
master-6232f2d
6232f2d7
·
ggml : optimize non-SIMD Q4_0 vector dot product (#703)
·
Apr 13, 2023
master-6c24870
6c248707
·
ggml : introduce GGML_ALIGNED_MALLOC/GGML_ALIGNED_FREE macros (#884)
·
Apr 13, 2023
master-8cda5c9
8cda5c98
·
fix whitespace (#944)
·
Apr 13, 2023
1
…
27
28
29
30
31
32
33
34
35
…
38