 Georgi Gerganov
|
e95b6554b4
ggml : add Q8_0 quantization for intermediate results (#951)
|
2 years ago |
|
Georgi Gerganov
|
aa485cee33
ggml : use posix_memalign on non-Windows env
|
2 years ago |
 Pavol Rusnak
|
c56b715269
Expose type name from ggml (#970)
|
2 years ago |
 Kerfuffle
|
c9a59b70a5
ggml : add unary and binary map operations (#874)
|
2 years ago |
|
Georgi Gerganov
|
1623a6e9b4
ggml : minor
|
2 years ago |
|
Georgi Gerganov
|
c14e0d2f23
ggml : always allocate buffers with size multiple of GGML_MEM_ALIGN
|
2 years ago |
|
Georgi Gerganov
|
0f07cacb05
ggml : fix q4_1 dot product types
|
2 years ago |
 Howard Su
|
c5d70f5c9e
ggml : optimize rope function to avoid call powf in the tight loop (#807)
|
2 years ago |
|
Georgi Gerganov
|
a3a2a0eda8
ggml : add GGML_DEFAULT_N_THREADS
|
2 years ago |
|
Georgi Gerganov
|
d990e3fffc
ggml : speed-up ggml_vec_dot_q4_1() ARM_NEON + 32-bit ARM support (#900)
|
2 years ago |
 Stephan Walter
|
6232f2d7fd
ggml : optimize non-SIMD Q4_0 vector dot product (#703)
|
2 years ago |
|
Pavol Rusnak
|
6c248707f5
ggml : introduce GGML_ALIGNED_MALLOC/GGML_ALIGNED_FREE macros (#884)
|
2 years ago |
 Vladimir
|
8c3ffc2f04
ggml : update cblas_sgemm columns var to be more reasonable (#838)
|
2 years ago |
|
Pavol Rusnak
|
8b679987cd
Fix whitespace, add .editorconfig, add GitHub workflow (#883)
|
2 years ago |
|
Stephan Walter
|
3e6e70d8e8
Add enum llama_ftype, sync ggml_type to model files (#709)
|
2 years ago |
 comex
|
2663d2c678
Windows fixes (#890)
|
2 years ago |
|
Georgi Gerganov
|
461ba9e66e
ggml : fix WASM build
|
2 years ago |
|
Georgi Gerganov
|
c3ac702e5e
ggml : add ggml_cont() + optimize ggml_cpy() for contiguous dst
|
2 years ago |
|
Georgi Gerganov
|
9d634ef452
ggml : remove trailing whitespaces
|
2 years ago |
 Marco Matthies
|
d9a239c410
Simplify to include lower-case windows.h always, fix compile on mingw32 (#747)
|
2 years ago |
|
Georgi Gerganov
|
684da25926
ggml : fix quantize_row_q4_1() ARM_NEON (close #876)
|
2 years ago |
|
comex
|
f963b63afa
Rewrite loading code to try to satisfy everyone:
|
2 years ago |
 unbounded
|
62cfc54f77
Add quantize-stats command for testing quantization (#728)
|
2 years ago |
|
Georgi Gerganov
|
eeaa7b0492
ggml : multi-thread ggml_rope() (~3-4 times faster on M1) (#781)
|
2 years ago |
|
Georgi Gerganov
|
986b6ce9f9
ggml, llama : avoid heavy V transpose + improvements (#775)
|
2 years ago |
 SebastianApel
|
437e77855a
10+% performance improvement of ggml_vec_dot_q4_0 on AVX2 (#654)
|
2 years ago |
 Marian Cepok
|
c0bb1d3ce2
ggml : change ne to int64_t (#626)
|
2 years ago |
|
Stephan Walter
|
3525899277
Enable -std= for cmake builds, fix warnings (#598)
|
2 years ago |
 slaren
|
1d08882afa
Optimize AVX2 ggml_vec_dot_q4_0 (#642)
|
2 years ago |
 perserk
|
02c5b27e91
Add AVX acceleration (#617)
|
2 years ago |
