 Sigbjørn Skjæret
|
759e37b0d8
tests : avoid github urls due to throttling (#13654)
|
8 months ago |
 Svetlozar Georgiev
|
4245e622e0
sycl: disable reorder for sycl mulmat (#13536)
|
8 months ago |
 0cc4m
|
c9c64dee57
Set GLM4 blk.*.attn_output.weight, kqv_out-* matmul to GGML_PREC_F32 to fix infinity values in output (#13639)
|
8 months ago |
 Georgi Gerganov
|
c00a2634be
metal : fix typo in FA kernel comments (#13651)
|
8 months ago |
|
Georgi Gerganov
|
e298d2fbd0
kv-cache : add SWA support (#13194)
|
8 months ago |
 Xinpeng Dou
|
f0adb80bf7
CANN: Update CANN model support (#13162)
|
8 months ago |
 Nicolò Scipione
|
f7c9429c85
sycl : Overcoming workaround for mmap() allocation on Windows (#13482)
|
8 months ago |
 psocolovsky
|
1dfbf2cf3a
common : add load_progress_callback (#13617)
|
8 months ago |
|
0cc4m
|
8960efd0a6
Vulkan: Add f32 accumulator support to quantized mul mat to fix GLM4 32B incoherence (#13607)
|
8 months ago |
 Alberto Cabrera Pérez
|
725f23f1f3
sycl : backend documentation review (#13544)
|
8 months ago |
 Xuan-Son Nguyen
|
92ecdcc06a
mtmd : add vision support for llama 4 (#13282)
|
8 months ago |
|
Alberto Cabrera Pérez
|
f71f40a284
ci : upgraded oneAPI version in SYCL workflows and dockerfile (#13532)
|
8 months ago |
|
Georgi Gerganov
|
d30cb5a7fa
sync : ggml
|
8 months ago |
 Johannes Gäßler
|
6c35981a64
mnist: fix segmentation fault (ggml/1227)
|
8 months ago |
 Diego Devesa
|
8b5e19aea6
ggml : fix apple OS check in ggml_print_backtrace (ggml/1229)
|
8 months ago |
 Daniel Tang
|
60aea028b5
ggml : Fix missing backtrace on Linux (ggml/1228)
|
8 months ago |
 Nick
|
9c55e5c5c2
fix: check model pointer validity before use (#13631)
|
8 months ago |
 Chenguang Li
|
33d7aed4a8
CANN: Support MOE Model MUL_MAT_ID (#13042)
|
8 months ago |
 Isaac McFadyen
|
6a2bc8bfb7
server : added --no-prefill-assistant flag (#13608)
|
8 months ago |
 Gilad S.
|
e3a7cf6c5b
cmake: use the current build config for vulkan-shaders-gen (#13595)
|
8 months ago |
|
Georgi Gerganov
|
518329b2d4
parallel : add option for non-shared and larger prompts (#13598)
|
8 months ago |
 Jeff Bolz
|
2f5a4e1e09
vulkan: move common FA code to flash_attn_base.comp (#13556)
|
8 months ago |
|
Jeff Bolz
|
4f41ee11d6
vulkan: use scalar FA rather than coopmat2 when N==1 (#13554)
|
8 months ago |
 Z
|
3e0be1cace
llguidance : official v0.7.20 release (no actual changes) [noci] (#13594)
|
8 months ago |
|
Xuan-Son Nguyen
|
6aa892ec2a
server : do not return error out of context (with ctx shift disabled) (#13577)
|
8 months ago |
|
Xuan-Son Nguyen
|
aea9f8b4e7
webui : improve accessibility for visually impaired people (#13551)
|
8 months ago |
|
Xuan-Son Nguyen
|
06c1e4abc1
readme : add list of dependencies and their license (#13591)
|
8 months ago |
|
Diego Devesa
|
415e40a357
releases : use arm version of curl for arm releases (#13592)
|
8 months ago |
|
Georgi Gerganov
|
654a67794f
metal : add FA-vec kernel for head size 64 (#13583)
|
8 months ago |
|
Diego Devesa
|
5364ae4ba5
llama : print hint when loading a model when no backends are loaded (#13589)
|
8 months ago |
