ggml-org/llama.cpp

Pull Requests Commits
convert.py: Fix loading safetensors and ggml format on Windows (#991)
comex committed 2 years ago
Verified 74f5899d
Fix potential int8 overflow in non-SIMD vec_dot (#986)
sw committed 2 years ago
Verified 2f7c8e01
Refactor ggml.c for future tensor types (#1001)
sw committed 2 years ago
Verified 0ad96463
ggml : add Q8_0 quantization for intermediate results (#951)
ggerganov committed 2 years ago
Verified e95b6554
ggml : use posix_memalign on non-Windows env
ggerganov committed 2 years ago
Verified aa485cee
benchmark : fix result validation in benchmark-q4_0-matmult (#987)
dfyz committed 2 years ago
Verified c12b14b7
cmake : add finding the OpenBLAS header file (#992)
katsu560 committed 2 years ago
Verified 106faaf2
Revert "main : alternative instruct mode (Vicuna support, etc.) (#863)" (#982)
prusnak committed 2 years ago
Verified c85e03d1
py : bump sentencepiece to 0.1.98 to support Python 3.11 (#976)
prusnak committed 2 years ago
Verified 48909354
make : fix dependencies, use auto variables (#983)
sw committed 2 years ago
Verified 93265e98
Expose type name from ggml (#970)
prusnak committed 2 years ago
Verified c56b7152
main : alternative instruct mode (Vicuna support, etc.) (#863)
aroidzap committed 2 years ago
Verified f4d277ae
ggml : add unary and binary map operations (#874)
KerfuffleV2 committed 2 years ago
Verified c9a59b70
py : cleanup dependencies (#962)
prusnak committed 2 years ago
Verified a32f7acc
py : fix flake8 and isort nitpicks (#960)
prusnak committed 2 years ago
Verified 43ffdefb
ggml : minor
ggerganov committed 2 years ago
Verified 1623a6e9
ggml : always allocate buffers with size multiple of GGML_MEM_ALIGN
ggerganov committed 2 years ago
Verified c14e0d2f
py : new conversion script (#545)
comex committed 2 years ago
Verified 723dac55
ggml : fix q4_1 dot product types
ggerganov committed 2 years ago
Verified 0f07cacb
ggml : optimize rope function to avoid call powf in the tight loop (#807)
howard0su committed 2 years ago
Verified c5d70f5c
perplexity : add support for batch size to `--perplexity` (#407)
glinscott committed 2 years ago
Verified be87b6ed
common : remove unnecessary includes (#947)
CRD716 committed 2 years ago
Verified 0e07e6a8
ggml : add GGML_DEFAULT_N_THREADS
ggerganov committed 2 years ago
Verified a3a2a0ed
ggml : speed-up ggml_vec_dot_q4_1() ARM_NEON + 32-bit ARM support (#900)
ggerganov committed 2 years ago
Verified d990e3ff
llama : merge llama_internal.h into llama.h
ggerganov committed 2 years ago
Verified 9190e8ea
gitignore : benchmark
ggerganov committed 2 years ago
Verified c85980ac
ggml : optimize non-SIMD Q4_0 vector dot product (#703)
sw committed 2 years ago
Verified 6232f2d7
ggml : introduce GGML_ALIGNED_MALLOC/GGML_ALIGNED_FREE macros (#884)
prusnak committed 2 years ago
Verified 6c248707
fix whitespace (#944)
CRD716 committed 2 years ago
Verified 8cda5c98
readme : remove python 3.10 warning (#929)
CRD716 committed 2 years ago
Verified ec292721
Terms Of Use Privacy Policy Refund Policy Impressum