 Johannes Gäßler
|
424c5d00a9
ggml/examples: add backend support for numerical optimization (ggml/949)
|
1 anno fa |
 Georgi Gerganov
|
a6809c6a2e
examples : add null threadpool args where needed (ggml/0)
|
1 anno fa |
|
Johannes Gäßler
|
5cb12f6839
CUDA: fix sum.cu compilation for CUDA < 11.7 (#9562)
|
1 anno fa |
|
Georgi Gerganov
|
d39e26741f
examples : flush log upon ctrl+c (#9559)
|
1 anno fa |
 Sigbjørn Skjæret
|
722ec1eb51
perplexity : do not escape input data by default (#9548)
|
1 anno fa |
|
Georgi Gerganov
|
6026da52d6
server : clean-up completed tasks from waiting list (#9531)
|
1 anno fa |
|
Sigbjørn Skjæret
|
eca0fab44e
imatrix : disable prompt escape by default (#9543)
|
1 anno fa |
 slaren
|
64c6af3195
ggml : fix n_threads_cur initialization with one thread (#9538)
|
1 anno fa |
|
Georgi Gerganov
|
0d2f22e45c
scripts : verify py deps at the start of compare (#9520)
|
1 anno fa |
 Daniel Bevenius
|
6443ddd985
llama : use reserve/emplace_back in sampler_sample (#9534)
|
1 anno fa |
 Vinesh Janarthanan
|
8a308354f6
server : match OAI structured output response (#9527)
|
1 anno fa |
 Eric Zhang
|
f799155ab8
server : fix OpenSSL build (remove obsolete `LOG_INFO`) (#9529)
|
1 anno fa |
 Neo Zhang Jianyu
|
faf67b3de4
[SYCL]set context default value to avoid memory issue, update guide (#9476)
|
1 anno fa |
 Michael Podvitskiy
|
7be099fa81
llama-bench: correct argument parsing error message (#9524)
|
1 anno fa |
 Bert Wagner
|
8b836ae731
arg : add env variable for parallel (#9513)
|
1 anno fa |
|
Michael Podvitskiy
|
8344ef58f8
llama : fix n_vocab init for 'no_vocab' case (#9511)
|
1 anno fa |
 Max Krasnyansky
|
0226613853
threadpool : skip polling for unused threads (#9461)
|
1 anno fa |
 Yuri Khrustalev
|
503147a9f9
unicode : add <algorithm> (#9508)
|
1 anno fa |
 Gabe Goodhart
|
0d2ec43833
llama : support IBM Granite architecture (#9412)
|
1 anno fa |
|
Michael Podvitskiy
|
37f3a3810e
llama : add llama_n_head() (#9512)
|
1 anno fa |
|
slaren
|
23e0d70bac
ggml : move common CPU backend impl to new header (#9509)
|
1 anno fa |
|
Daniel Bevenius
|
acb2c32c33
llama : rename n_embed to n_embd in rwkv6_time_mix (#9504)
|
1 anno fa |
|
Michael Podvitskiy
|
a6a3a5c531
ggml : link MATH_LIBRARY not by its full path (#9339)
|
1 anno fa |
 compilade
|
d54c21df7e
convert : identify missing model files (#9397)
|
1 anno fa |
|
Georgi Gerganov
|
19514d632e
cmake : do not hide GGML options + rename option (#9465)
|
1 anno fa |
 Eve
|
5c3d0f1824
ggml : IQ4_NL sgemm + Q4_0 AVX optimization (#9422)
|
1 anno fa |
 Shane A
|
0aadac10c7
llama : support OLMoE (#9462)
|
1 anno fa |
 CarryFun
|
95ca85168b
llama : support MiniCPM3 (#9322)
|
1 anno fa |
|
Vinesh Janarthanan
|
441b72b91f
main : option to disable context shift (#9484)
|
1 anno fa |
|
Georgi Gerganov
|
c4965a64f7
metal : handle zero-sized allocs (#9466)
|
1 anno fa |
