ggerganov/llama.cpp

Pull Requests Commits
drop quantize.py (now that models are using a single file)
prusnak committed 2 years ago
9733104b
readme : update supported models
ggerganov committed 2 years ago
Verified 3df890ae
Introduce GGML migration tool for new file format
jart committed 2 years ago
ee0c40dd
Ensure --mlock works properly with mmap() support
jart committed 2 years ago
6f23ba5e
Make loading weights 10-100x faster
jart committed 2 years ago
78ca9838
Initial windows support (untested)
slaren committed 2 years ago
a0173903
Always initialize mm_addr and mm_length in llama_model
slaren committed 2 years ago
ac184d51
Unmap the file in llama_free
slaren committed 2 years ago
276e5b78
Make mmap_file static
slaren committed 2 years ago
d68c5dc4
Fix ggml_init_params in quantize
slaren committed 2 years ago
64bde3ff
Add mmap support for model files
slaren committed 2 years ago
c03ae8dc
cmake : properly invoke CTest (#629)
sw committed 2 years ago
Verified 3bcc129b
Remove unused variable (#607)
Ameobea committed 2 years ago
Verified a4755cf2
make : fix darwin f16c flags check (#615)
keen99 committed 2 years ago
Verified 1f0414fe
ggml : fix NEON signs (close #620, #622)
ggerganov committed 2 years ago
Verified 77efdf5a
Fix GGML_F32Cx8_STORE in AVX without F16C path (#619)
slaren committed 2 years ago
Verified ed3c680b
ci : re-enable AVX512 testing (Windows-MSVC) (#584)
anzz1 committed 2 years ago
Verified 9cbc404b
ggml : init time on first ggml_init() call
ggerganov committed 2 years ago
Verified b51c717d
llama : fix compile warnings when reading the vocab
ggerganov committed 2 years ago
Verified 0ba76c1e
ggml : add ARM_NEON dequantize_row_q4_1()
ggerganov committed 2 years ago
Verified cea1c859
ggml : add ARM_NEON quantize_row_q4_1()
ggerganov committed 2 years ago
Verified f202ada1
ggml : add ARM_NEON ggml_vec_dot_q4_1()
ggerganov committed 2 years ago
Verified 3b44d30d
rename convert_ggml_to_pth.py -> convert-ggml-to-pth.py (#600)
prusnak committed 2 years ago
Verified 61cbfff5
Create chat-13B.bat (#592)
Royalphax committed 2 years ago
Verified d9ad1044
readme : fix typos
ggerganov committed 2 years ago
Verified b467702b
readme : add GPT4All instructions (close #588)
ggerganov committed 2 years ago
Verified 516d88e7
py : add GPT4All conversion script
ggerganov committed 2 years ago
Verified 53635c08
llama : use the same threshold for OpenBLAS and ggml thread limiting (#577)
Piezoid committed 2 years ago
Verified 41318d70
add example of re-act pattern (#583)
tobi committed 2 years ago
Verified a6956b25
Fix GCC warning about binary literal (#595)
anzz1 committed 2 years ago
Verified 83df5639
Terms Of Use Privacy Policy Refund Policy Impressum