Tags
Tags give the ability to mark specific points in history as being important
master-1d0824b
1d0824b2
·
llama : print help to stdout (#2338)
·
Jul 23, 2023
master-a940458
a940458e
·
llama : print max tensor size to stderr (#2336)
·
Jul 23, 2023
master-91171b8
91171b80
·
make : fix CLBLAST compile support in FreeBSD (#2331)
·
Jul 23, 2023
master-d2a4366
d2a43664
·
Speed up Q4_K (#2322)
·
Jul 23, 2023
master-b9b7d94
b9b7d94f
·
CUDA: Fixed 7b q3_K_S with mul_mat_vec_q (#2313)
·
Jul 22, 2023
master-b47b8a9
b47b8a9c
·
llama : optimize memory buffers (#2325)
·
Jul 22, 2023
master-b5fe67f
b5fe67f8
·
Perplexity: Compute scores correlated to HellaSwag (#2312)
·
Jul 22, 2023
master-d924522
d924522a
·
Custom RoPE + bettter memory management for CUDA (#2295)
·
Jul 21, 2023
master-0db14fe
0db14fef
·
ggml : fix the rope fix (
513f8619
)
·
Jul 21, 2023
master-513f861
513f8619
·
ggml : fix rope args order + assert (#2054)
·
Jul 21, 2023
master-ab0e26b
ab0e26bd
·
llama : remove cfg smooth factor as it is only a reparameterization of the guidance scale (#2280)
·
Jul 21, 2023
master-73643f5
73643f5f
·
gitignore : changes for Poetry users + chat examples (#2284)
·
Jul 21, 2023
master-a814d04
a814d04f
·
make : fix indentation
·
Jul 21, 2023
master-42c7c2e
42c7c2e2
·
make : support customized LLAMA_CUDA_NVCC and LLAMA_CUDA_CCBIN (#2275)
·
Jul 21, 2023
master-78a3d13
78a3d134
·
flake : remove intel mkl from flake.nix due to missing files (#2277)
·
Jul 21, 2023
master-ae178ab
ae178ab4
·
llama : make tensor_split ptr instead of array (#2272)
·
Jul 21, 2023
master-54e3bc7
54e3bc76
·
make : add new target for test binaries (#2244)
·
Jul 21, 2023
master-9cf022a
9cf022a1
·
make : fix embdinput library and server examples building on MSYS2 (#2235)
·
Jul 21, 2023
master-fff0e0e
fff0e0ea
·
llama : fix regression from #2000 - could not load no-mmap models
·
Jul 20, 2023
master-294f424
294f4245
·
llama : extend API to get max devices at runtime (#2253)
·
Jul 19, 2023
1
…
8
9
10
11
12
13
14
15
16
…
38