ggml-org/llama.cpp

Pull Requests Commits
ggml : implement backward pass for llama + small training-llama-from-scratch example (#1360)
xaedes committed 2 years ago
Verified f954edda
ggml : sync alibi fix from ggml repo
ggerganov committed 2 years ago
Verified f048af02
Adding SSE instructions to ggml_vec_dot_q4_0_q8_0 (#1413)
3ooabkhxtn committed 2 years ago
Verified ac0cd259
llama : fix various warnings
ggerganov committed 2 years ago
Verified 0cd22e19
embedding : remove unused code (#1426)
SanftMonster committed 2 years ago
Verified 6456a4eb
readme : update Q4_0 perplexities
ggerganov committed 2 years ago
Verified cdd53508
llama : free ggml context in set / copy state data (close #1425)
ggerganov committed 2 years ago
Verified 738ace39
opencl : fix kernels for the new formats (#1422)
SlyEcho committed 2 years ago
Verified 699b1ad7
llama : fix --mtest option (close #1414)
ggerganov committed 2 years ago
Verified fb62f924
CLI args use - instead of _, backwards compatible (#1416)
JohannesGaessler committed 2 years ago
Verified 773ee249
Add clang-tidy reviews to CI (#1407)
slaren committed 2 years ago
Verified 553fd4d4
readme : add C#/.NET bindings repo (#1409)
SanftMonster committed 2 years ago
Verified 089b1c93
ggml : remove bit shuffling (#1405)
ggerganov committed 2 years ago
Verified b9fd7eee
prompts : model agnostic DAN (#1304)
CRD716 committed 2 years ago
Verified b608b55a
main : add option to save full output to session (#1338)
ejones committed 2 years ago
Verified cf348a60
Locale fix for Windows (#1379)
DannyDaemonic committed 2 years ago
Verified e6a46b0e
use pause asm insn in busyloop to run the CPU (13600K) 10 °C cooler (#1314)
Safari77 committed 2 years ago
Verified 9f8dbc47
Interface improvements and `--multiline-input` (previously `--author-mode`) (#1040)
DannyDaemonic committed 2 years ago
Verified 41654efe
readme : add notice about upcoming breaking change
ggerganov committed 2 years ago
Verified 56551bc1
readme : add TOC and Pygmalion instructions (#1359)
AlpinDale committed 2 years ago
Verified fe60904e
llama : fix hparams shadow (#1367)
prusnak committed 2 years ago
Verified 003ba2fb
llama : require first token to be BOS (#1303)
ggerganov committed 2 years ago
Verified f9a63649
convert: add ability to convert safetensors files (#1276)
ubik2 committed 2 years ago
Verified 95078cc5
Documented CUDA reproducibility, added warning (#1346)
JohannesGaessler committed 2 years ago
Verified 1f48b0ab
CI: add Windows CLBlast and OpenBLAS builds (#1277)
SlyEcho committed 2 years ago
Verified e1295513
ggml : Allow usage of CLBlast alongside Accelerate.framework (#1336)
swittk committed 2 years ago
Verified 1b0fd454
Remove default arguments from sampling functions (#1343)
j-f1 committed 2 years ago
Verified 39240885
makefile: automatic Arch Linux detection (#1332)
daniandtheweb committed 2 years ago
Verified 173d0e64
ci : add cublas to windows release (#1271)
Green-Sky committed 2 years ago
Verified a3b85b28
readme: add missing info (#1324)
prusnak committed 2 years ago
Verified 921dcee0
Terms Of Use Privacy Policy Refund Policy Impressum