stduhpf
|
4ccea213bc
hellaswag: display estimated score confidence interval (#12797)
|
9 месяцев назад |
Georgi Gerganov
|
1a1ab7e7a4
cuda : fix HIP and MUSA BF16 (#0)
|
9 месяцев назад |
Georgi Gerganov
|
a4e46e28f9
sync : ggml
|
9 месяцев назад |
Georgi Gerganov
|
ff067dbcb9
ggml : simplify Arm fp16 CPU logic (ggml/1177)
|
9 месяцев назад |
Sigbjørn Skjæret
|
36ca8b3628
CUDA: don't convert BF16 weights to FP32 (ggml/1174)
|
9 месяцев назад |
cmdr2
|
995083e4ed
cpu: move all the operators into a separate c++ file (except mul_mat) (ggml/1167)
|
9 месяцев назад |
zhouwg
|
518a01480e
sycl: remove redundant memcopy in function ggml_backend_sycl_buffer_set_tensor (#12734)
|
9 месяцев назад |
Xuan-Son Nguyen
|
e391d3ee8d
ci : no curl on ggml-ci (#12796)
|
9 месяцев назад |
Xuan-Son Nguyen
|
bd3f59f812
cmake : enable curl by default (#12761)
|
9 месяцев назад |
zhouwg
|
52b3d71f12
CANN: fix typo in ggml-cann (#12733)
|
9 месяцев назад |
hipudding
|
d0d5b2232b
CANN: Refactor to reduce duplicate code (#12731)
|
9 месяцев назад |
R0CKSTAR
|
916c83bfe7
musa: fix compilation warnings in mp_22/31 (#12780)
|
9 месяцев назад |
Jeff Bolz
|
0c74b04376
vulkan: fix NaN issue in flash attention shader (#12776)
|
9 месяцев назад |
Jeff Bolz
|
80b717d493
vulkan: Use unclamped loads for flash attention mask (#12720)
|
9 месяцев назад |
0cc4m
|
6bf28f0111
Vulkan: Tune Vulkan mmq int dot shader for performance (#12767)
|
9 месяцев назад |
Sergey Fedorov
|
f1e3eb4249
common : fix includes in arg.cpp and gemma3-cli.cpp (#12766)
|
9 месяцев назад |
Xuan-Son Nguyen
|
0364178ca2
clip : refactor clip_init, add tests (#12757)
|
9 месяцев назад |
エシュナヴァリシア
|
c6ff5d2a8d
common: custom hf endpoint support (#12769)
|
9 месяцев назад |
Olivier Chafik
|
7a84777f42
sync: minja (#12739)
|
9 месяцев назад |
Georgi Gerganov
|
3e1d29348b
kv-cache : simplify + fix warning for recurrent models (#12756)
|
9 месяцев назад |
bandoti
|
1be76e4620
ci: add Linux cross-compile build (#12428)
|
9 месяцев назад |
Nauful Shaikh
|
b772394297
server : webui : Upgrade daisyui, tailwindcss. (#12735)
|
9 месяцев назад |
nick huang
|
23106f94ea
gguf-split : --merge now respects --dry-run option (#12681)
|
9 месяцев назад |
Nicolò Scipione
|
94148ba330
sycl: allow ggml-sycl configuration and compilation using Visual Studio project/solution (#12625)
|
9 месяцев назад |
Ronny Brendel
|
9ac4d611d0
cmake: fix ggml-shaders-gen compiler paths containing spaces (#12747)
|
9 месяцев назад |
Daniel Bevenius
|
348888e0dc
docs : add XCFramework section to README.md [no ci] (#12746)
|
9 месяцев назад |
Jeff Bolz
|
74d4f5b041
vulkan: Hybrid waitForFences/getFenceStatus to reduce fence latency (#12630)
|
9 месяцев назад |
Jeff Bolz
|
35e592eb30
vulkan: set cmake minimum and project name in vulkan-shaders (#12744)
|
9 месяцев назад |
lhez
|
7d7b1bafa7
opencl: update doc for OpenCL (#12702)
|
9 месяцев назад |
Gaurav Garg
|
c262beddf2
CUDA: Prefer vector flash decoding kernel for Gemma models (#12738)
|
9 месяцев назад |