Commit History

go : add beamsize/entropythold/maxcontext to context interface (#2350)
7efcda7
unverified

hsinhoyeh commited on

talk-llama : sync llama.cpp
4493ffd

ggerganov commited on

whisper : update FA call
2bfec97

ggerganov commited on

sync : ggml
7ba8c97

ggerganov commited on

sync : vulkan (skip) (llama/0)
5fe3dd6

ggerganov commited on

ggml : do not crash when quantizing q4_x_x with an imatrix (llama/9192)
d64f932

slaren commited on

metal : separate scale and mask from QKT in FA kernel (llama/9189)
90cc3cd

ggerganov commited on

ggml : add SSM Metal kernels (llama/8546)
b6e7294

ggerganov commited on

metal : gemma2 flash attention support (llama/9159)
e62fd15

slaren commited on

CPU/CUDA: Gemma 2 FlashAttention support (llama/8542)
fb8ae8b

JohannesGaessler commited on

Add a space to supress a cmake warning (llama/9133)
287612e

qnixsynapse commited on

Add oneDNN primitive support (llama/9091)
b4d8c3e

KevinLy commited on

llama : simplify Mamba with advanced batch splits (llama/8526)
f1abcb4

compilade ggerganov commited on

fallback mmvq (llama/9088)
4b1fda0

hengyu Alberto Cabrera Pérez commited on

Fix SYCL `im2col` and `convert` Overflow with Large Dims (llama/9052)
5f43886

zhentaoyu commited on

rpc : print error message when failed to connect endpoint (llama/9042)
d54b156

rgerganov commited on

rpc : prevent crashes on invalid input (llama/9040)
656ae00

rgerganov commited on

ggml : dynamic ggml_sched_max_splits based on graph_size (llama/9047)
e0dc1ad

nicoboss commited on

cmake : remove unused option GGML_CURL (llama/9011)
12634fc

ggerganov commited on

ggml : move rope type enum to ggml.h (llama/8949)
9d45f48

danbev slaren commited on

ggml: fix div-by-zero (llama/9003)
d9ee26f

DavidKorczynski commited on

Optimize Vulkan backend for better CPU performance and less GPU synchronization overhead. (llama/8943)
11bc9e6

Markus Tavenrath OccamRazor commited on

feat: ref. cross entropy, add CUDA, fix grad test (ggml/929)
e1e87a3

JohannesGaessler commited on

ggml: remove bad assert (ggml/928)
ba483f7

JohannesGaessler commited on

examples: add MNIST training + missing ops
0828065

JohannesGaessler commited on

models : add support for wget2 for fedora (#2387)
0653499
unverified

Brad Murray commited on

readme : update the path to bench.py (#2386)
57c7a6b
unverified

Peng commited on

readme : fix typo (#2383)
16e5a16
unverified

ivoputzer commited on

readme : fix broken links in implementation details section (#2382)
4863dee
unverified

stormofice commited on

whisper : fix compile warning for unused params
0e05e03
unverified

ggerganov commited on

sync : ggml vulkan (ggml/0)
c4c7e49

ggerganov commited on

ggml : fix typo in ggml-quants.c comment (ggml/922)
f158bc0

danbev commited on

feat: add new `sin` and `cos` operators (ggml/919)
f541d31

Ronsor ggerganov commited on

readme : fix broken links (#2358)
93e1056
unverified

ericcurtin commited on

examples : use colorblind friendly TTY color scheme (#2360)
09303a2
unverified

Justine Tunney commited on

sync : ggml
e6d1739
unverified

ggerganov commited on

ggml : support forward pass broadcasting in ggml_sub (ggml/914)
0af2d37
unverified

smeso commited on

metal : fix uninitialized abort_callback (llama/8968)
f971b60
unverified

slaren commited on

rpc : sanitize tensor data + warnings (llama/0)
87d58fe
unverified

ggerganov slaren commited on

cann : add Ascend NPU support (#2336)
94baae9
unverified

Mimi89757 commited on

whisper : fix compile warning (#0)
1a699ea

ggerganov commited on

sync : ggml
acf76b7

ggerganov commited on

ggml : add CANN backend (llama/0)
7c34a03

hipudding commited on

scripts : sync cann
0a74031

ggerganov commited on

ci : disable ruby workflow (#0)
4b0eff8

ggerganov commited on

ci : try to fix FreeBSD (#0)
683de5a

ggerganov commited on

build : fix aarch64 (#0)
55befbb

ggerganov commited on

talk-llama : sync llama.cpp
a40d0a7

ggerganov commited on

sync : ggml
96e8b15

ggerganov commited on