 Nick
|
9c55e5c5c2
fix: check model pointer validity before use (#13631)
|
8 месяцев назад |
 Chenguang Li
|
33d7aed4a8
CANN: Support MOE Model MUL_MAT_ID (#13042)
|
8 месяцев назад |
 Isaac McFadyen
|
6a2bc8bfb7
server : added --no-prefill-assistant flag (#13608)
|
8 месяцев назад |
 Gilad S.
|
e3a7cf6c5b
cmake: use the current build config for vulkan-shaders-gen (#13595)
|
8 месяцев назад |
 Georgi Gerganov
|
518329b2d4
parallel : add option for non-shared and larger prompts (#13598)
|
8 месяцев назад |
 Jeff Bolz
|
2f5a4e1e09
vulkan: move common FA code to flash_attn_base.comp (#13556)
|
8 месяцев назад |
|
Jeff Bolz
|
4f41ee11d6
vulkan: use scalar FA rather than coopmat2 when N==1 (#13554)
|
8 месяцев назад |
 Z
|
3e0be1cace
llguidance : official v0.7.20 release (no actual changes) [noci] (#13594)
|
8 месяцев назад |
 Xuan-Son Nguyen
|
6aa892ec2a
server : do not return error out of context (with ctx shift disabled) (#13577)
|
8 месяцев назад |
|
Xuan-Son Nguyen
|
aea9f8b4e7
webui : improve accessibility for visually impaired people (#13551)
|
8 месяцев назад |
|
Xuan-Son Nguyen
|
06c1e4abc1
readme : add list of dependencies and their license (#13591)
|
8 месяцев назад |
 Diego Devesa
|
415e40a357
releases : use arm version of curl for arm releases (#13592)
|
8 месяцев назад |
|
Georgi Gerganov
|
654a67794f
metal : add FA-vec kernel for head size 64 (#13583)
|
8 месяцев назад |
|
Diego Devesa
|
5364ae4ba5
llama : print hint when loading a model when no backends are loaded (#13589)
|
8 месяцев назад |
 Sigbjørn Skjæret
|
7c07ac244d
ci : add ppc64el to build-linux-cross (#13575)
|
8 месяцев назад |
 Łukasz Ślusarczyk
|
0a338ed013
sycl : fixed compilation warnings (#13582)
|
8 месяцев назад |
 Olivier Chafik
|
bc098c3cf0
minja: sync (qwen3) (#13573)
|
8 месяцев назад |
|
Diego Devesa
|
c6a2c9e741
gguf : use ggml log system (#13571)
|
8 месяцев назад |
 Daniel Tang
|
07ad2b6db3
gguf-py : fix disconnect-before-connect in editor-gui (#13569)
|
8 месяцев назад |
|
Xuan-Son Nguyen
|
c531edfa34
convert : fix conversion for llama 4 (#13567)
|
8 месяцев назад |
 Atharva Dubey
|
02cdd2d8b0
sycl: simplify bin_bcast_kernel (#13383)
|
8 месяцев назад |
 Svetlozar Georgiev
|
64bb51cf90
sycl: reordered Q4_K MMVQ (#13109)
|
8 месяцев назад |
|
Łukasz Ślusarczyk
|
9c404ed54c
sycl: use oneDNN for matrices multiplication (#12972)
|
8 месяцев назад |
|
Diego Devesa
|
6c8b91500e
llama-bench : fix -ot with dl backends (#13563)
|
8 месяцев назад |
|
Xuan-Son Nguyen
|
3cc1f1f1d2
webui : handle PDF input (as text or image) + convert pasted long content to file (#13562)
|
8 месяцев назад |
 Piotr Wilkin (ilintar)
|
c753d7bed0
server : proper error handling for missing elements in messages array (OpenAI compatible backend) (#13540)
|
8 месяцев назад |
|
Georgi Gerganov
|
b2838049cc
bench : handle decode errors (#13548)
|
8 месяцев назад |
 Olivier Chafik
|
aa48e373f2
`server`: inject date_string in llama 3.x template + fix date for firefunction v2 (#12802)
|
8 месяцев назад |
|
Georgi Gerganov
|
e3a9421b78
kv-cache : fix out-of-bounds view during reserve graph (#13547)
|
8 месяцев назад |
 Yibo Cai
|
5ab5d5fb25
arm64: optimize q6_k_q8_k kernel with i8mm (#13519)
|
8 месяцев назад |
