Skip to content
GitLab
Explore
Sign in
Tags
Tags give the ability to mark specific points in history as being important
This project is mirrored from
https://github.com/ggerganov/llama.cpp
. Pull mirroring updated
Sep 19, 2024
.
b3700
00b02bb2
·
imatrix : fix arg parser for imatrix (#9366)
·
Sep 08, 2024
b3699
a8768614
·
metal : update support condition for im2col + fix warning (#0)
·
Sep 08, 2024
b3688
fbb7fcff
·
llama : set attrs of mislabelled EOT/EOM tokens (#9348)
·
Sep 08, 2024
b3687
a5b5d9a1
·
llama.android : fix build (#9350)
·
Sep 08, 2024
b3686
f12295b8
·
llama : fix empty ring buffer push (#9358)
·
Sep 08, 2024
b3685
faf69d42
·
llama : sanitize invalid tokens (#9357)
·
Sep 08, 2024
b3684
e536426d
·
llamafile : disable sgemm for batch-size 1 (#9330)
·
Sep 07, 2024
b3683
1b9ae518
·
common : refactor arg parser (#9308)
·
Sep 07, 2024
b3682
e32d0816
·
ggml : always check bounds on get_rows operations (#9354)
·
Sep 07, 2024
b3681
df270ef7
·
llama : refactor sampling v2 (#9294)
·
Sep 07, 2024
b3680
947538ac
·
ggml : fix missing `cpu_set_t` on emscripten (#9336)
·
Sep 07, 2024
b3678
9b2c24c0
·
server : simplify state machine for slot (#9283)
·
Sep 06, 2024
b3677
134bc38e
·
llama-bench : log benchmark progress (#9287)
·
Sep 06, 2024
b3676
815b1fb2
·
batched-bench : add `--output-format jsonl` option (#9293)
·
Sep 06, 2024
b3675
409dc4f8
·
ggml : fix build break for the vulkan-debug (#9265)
·
Sep 06, 2024
b3674
4a1411b4
·
server : fix missing lock (#9334)
·
Sep 06, 2024
b3672
9bc6db28
·
ggml-quants : ternary packing for TriLMs and BitNet b1.58 (#8151)
·
Sep 05, 2024
b3671
32b2ec88
·
Update build.yml (#9184)
·
Sep 06, 2024
b3669
4db04784
·
cuda : fix defrag with quantized KV (#9319)
·
Sep 05, 2024
b3668
bdf314f3
·
llama-bench : fix NUL terminators in CPU name (#9313)
·
Sep 05, 2024
Prev
1
2
3
4
5
6
7
8
…
122
Next