ggerganov/llama.cpp

Pull Requests Commits
fix missing parameters in `llama_init_from_gpt_params` (#1293)
slaren committed 2 years ago
Verified bf4b22ff
examples : add llama_init_from_gpt_params() common function (#1290)
deadprogram committed 2 years ago
Verified 67c77799
llama : fix compile warnings
ggerganov committed 2 years ago
Verified 0e6cbff1
ggml : fix 32-bit ARM
ggerganov committed 2 years ago
Verified 5d5817ca
examples : improve vertical alignment of a few variables (#1286)
deadprogram committed 2 years ago
Verified 8c9be35f
ggml : fix ppc64le build error and make cmake detect Power processors (#1284)
mgiessing committed 2 years ago
Verified cc0bb723
llama : allow 0 as a seed number. (#1275)
rbrisita committed 2 years ago
Verified 2bb992f0
main : switch input_noecho to input_echo to remove negation (#979)
deadprogram committed 2 years ago
Verified e2cd5069
ggml: add names to tensors (#1268)
slaren committed 2 years ago
Verified 2d099e51
Add git-based build information for better issue tracking (#1232)
DannyDaemonic committed 2 years ago
Verified f4cef87e
cuBLAS: refactor and optimize f16 mat mul performance (#1259)
slaren committed 2 years ago
Verified 58b367c2
llama : update stubs for systems without mmap and mlock (#1266)
xloem committed 2 years ago
Verified ea3a0ad6
ggml : fix ggml_used_mem() (#1264)
KerfuffleV2 committed 2 years ago
Verified 2bdc0964
llama : fix session load / save (#1263)
ggerganov committed 2 years ago
Verified 70269cae
cuBLAS: fall back to pageable memory if pinned alloc fails (#1233)
slaren committed 2 years ago
Verified b925f1f1
llama : let context be const when accessing const data (#1261)
grencez committed 2 years ago
Verified 90b19bd6
ggml : fix UB (int << 31)
ggerganov committed 2 years ago
Verified 7ff0dcd3
build: add armv{6,7,8} support to cmake (#1251)
prusnak committed 2 years ago
Verified 6f796992
common : better default number of threads (#934)
jon-chuang committed 2 years ago
Verified a5d30b1f
ggml : add CLBlast q5_0, q5_1, q8_0 dequant kernels (#1225)
0cc4m committed 2 years ago
Verified 76a88492
ggml : add Q5 WASM SIMD + GGML_FTYPE
ggerganov committed 2 years ago
Verified 6bc4400e
Various fixes to mat_mul benchmark (#1253)
sw committed 2 years ago
Verified f0d70f14
ggml : fix labels for GGML_OP_ALIBI
ggerganov committed 2 years ago
Verified 3e5aa8a1
ggml : fix 32-bit ARM NEON
ggerganov committed 2 years ago
Verified c3ca7a5f
ggml : use vzip instead of vuzp for consistency
ggerganov committed 2 years ago
Verified e8c05161
ggml : fix visibility and unused warnings
ggerganov committed 2 years ago
Verified 0b5a9350
ggml : fix #if for f32_f32 mul_mat (CLBlast) (#1229)
ggerganov committed 2 years ago
Verified ec728e44
ggml : adjust mul_mat_f16 work memory (#1226)
ggerganov committed 2 years ago
Verified 214b6a35
build : fix reference to old llama_util.h
ggerganov committed 2 years ago
Verified 305eb5af
examples : fix save-load-state + rename llama-util.h
ggerganov committed 2 years ago
Verified 84ca9c2e
Terms Of Use Privacy Policy Refund Policy Impressum