ggml-org/llama.cpp

Pull Requests Commits
Faster Q5_K and Q6_K on Metal (#2294)
ikawrakow committed 2 years ago
Verified e782c9e7
[wip] chat now has parameter and cfg
SlyEcho committed 2 years ago
Verified e4db7072
Faster Q4_K on Metal (#2290)
ikawrakow committed 2 years ago
Verified 785829df
llama : fix regression from #2000 - could not load no-mmap models
ggerganov committed 2 years ago
fff0e0ea
metal: minor q4 optimization and reduce code size (#2248)
lshzh-ww committed 2 years ago
Verified 417a85a0
[wip] chat improvements
SlyEcho committed 2 years ago
Verified 082dd812
consistent semicolons
SlyEcho committed 2 years ago
Verified 43694ca8
Merge master into server-cfg
SlyEcho committed 2 years ago
Verified 890d1b84
last n tokens done
SlyEcho committed 2 years ago
Verified dd3cf576
remove "smooth factor"
SlyEcho committed 2 years ago
Verified 42591a0a
refactor evaluation logic
SlyEcho committed 2 years ago
Verified 2cb8469e
llama : extend API to get max devices at runtime (#2253)
SanftMonster committed 2 years ago
Verified 294f4245
flake : update flake.nix (#2270)
Freed-Wu committed 2 years ago
Verified 45a1b07e
cmake : install targets (#2256)
Freed-Wu committed 2 years ago
Verified b1f42909
ci : integrate with ggml-org/ci (#2250)
ggerganov committed 2 years ago
Verified d01bccde
llama : shorten quantization descriptions
ggerganov committed 2 years ago
Verified 6cbf9dfb
Support dup & cont ops on CUDA (#2242)
li-plus committed 2 years ago
Verified 7568d1a2
llama : fix t_start_sample_us initialization warning (#2238)
grencez committed 2 years ago
Verified b7647436
ggml : fixed runtime bugs and compile errors related to GGML_PERF and GGML_DEBUG (#2219)
mqy committed 2 years ago
Verified 672dda10
py : turn verify-checksum-models.py into executable (#2245)
jpodivin committed 2 years ago
Verified 27ab66e4
llama : add custom RoPE (#2054)
jxy committed 2 years ago
Verified 6e7cca40
flake : add runHook preInstall/postInstall to installPhase so hooks function (#2224)
ddellacosta committed 2 years ago
Verified a6803cab
make : use pkg-config for OpenBLAS (#2222)
Freed-Wu committed 2 years ago
Verified 7dabc66f
cuda : allocate all temporary ggml_tensor_extra_gpu from a fixed-size buffer (#2220)
bullno1 committed 2 years ago
Verified 7cdd30bf
ggml : fix static_assert with older compilers #2024 (#2218)
evanmiller committed 2 years ago
Verified e8035f14
llama : add functions that work directly on model (#2197)
bullno1 committed 2 years ago
Verified 7513b7b0
build.zig : install config header (#2216)
alichraghi committed 2 years ago
Verified de834242
examples : fixed path typos in embd-input (#2214)
xushangning committed 2 years ago
Verified c48c525f
cuda : support broadcast add & mul (#2192)
li-plus committed 2 years ago
Verified 206e01de
CUDA: mul_mat_vec_q kernels for k-quants (#2203)
JohannesGaessler committed 2 years ago
Verified 4304bd3c
Terms Of Use Privacy Policy Refund Policy Impressum