ggerganov/llama.cpp

Pull Requests Commits
vulkan: Find optimal memory type but with fallback (#5381)
luciferous committed 1 year ago
Verified 704359e2
readme : fix typo (#5490)
Rune-AI committed 1 year ago
Verified 594fca3f
llava : update README.md (#5489)
cmp-nct committed 1 year ago
Verified ccbb277f
cmake : ARM intrinsics detection for MSVC (#5401)
Xarbirus committed 1 year ago
Verified 8084d554
llava : support v1.6 (#5267)
cmp-nct committed 1 year ago
Verified aa234129
Early return for zero size calls to get_tensor. (#5482)
manyoso committed 1 year ago
Verified f5ca0548
gguf : add python reader example (#5216)
cmp-nct committed 1 year ago
Verified 6c00a066
llama : add support for Nomic Embed (#5468)
cebtenzzre committed 1 year ago
Verified ea9c8e11
llama : allow raw byte in SPM vocabs; don't crash on nl 404 (#5478)
akx committed 1 year ago
Verified c4e6dd59
llama : make load error reporting more granular (#5477)
akx committed 1 year ago
Verified 037259be
finetune : rename feed-forward tensors (w1/w2/w3) (#4839)
danbev committed 1 year ago
Verified 26397890
tests : multi-thread the tokenizer tests (#5474)
ggerganov committed 1 year ago
Verified cf45252a
llama : support batched embeddings (#5466)
iamlemec committed 1 year ago
Verified 03bf161e
make: add error message for bad CUDA version (#5444)
JohannesGaessler committed 1 year ago
Verified ad014bba
bert : add tests + fix quantization (#5475)
ggerganov committed 1 year ago
Verified 49cc1f7d
tests : disable moe test (#5473)
ggerganov committed 1 year ago
Verified 99b8b43d
ggml-quants : fix compiler warnings (shadow variable) (#5472)
ikawrakow committed 1 year ago
Verified 895407f3
llama : fix quantization when tensors are missing (#5423)
ggerganov committed 1 year ago
Verified 099afc62
swift : package no longer use ggml dependency (#5465)
ggerganov committed 1 year ago
Verified df334a11
py : fix persimmon `n_rot` conversion (#5460)
lx200916 committed 1 year ago
Verified dbd8828e
ggml-sycl: Replace 3d ops with macro (#5458)
abhilash1910 committed 1 year ago
Verified 43fe07c1
llava : remove prog parameter from ArgumentParser (#5457)
danbev committed 1 year ago
Verified 4a46d2b7
sync : ggml (#5452)
ggerganov committed 1 year ago
Verified 3b169441
CUDA: mul_mat_vec_q tiling, refactor mul mat logic (#5434)
JohannesGaessler committed 1 year ago
Verified 3bdc4cd0
Add support for BERT embedding models (#5423)
iamlemec committed 1 year ago
Verified 2891c8aa
flake.lock: Update
github-actions[bot] committed 1 year ago
97a33650
vulkan: only use M-sized matmul on Apple GPUs (#5412)
slp committed 1 year ago
Verified c88c74f9
common : use enums for sampler types (#5418)
z80maniac committed 1 year ago
Verified a803333a
server : allow to specify tokens as strings in logit_bias (#5003)
z80maniac committed 1 year ago
Verified 68478014
main : ctrl+C print timing in non-interactive mode (#3873)
ggerganov committed 1 year ago
Verified 85910c5b
Terms Of Use Privacy Policy Refund Policy Impressum