Michael Yang 4dc5b117dd automatically set num_keep if num_keep < 0 ...

num_keep defines how many tokens to keep in the context when truncating
inputs. if left to its default value of -1, the server will calculate
num_keep to be the left of the system instructions

2023-08-07 16:19:12 -07:00

..

ggml-alloc.c

update llama.cpp

2023-08-03 11:50:24 -07:00

ggml-alloc.h

update llama.cpp

2023-08-03 11:50:24 -07:00

ggml-cuda.cu

update llama.cpp

2023-08-03 11:50:24 -07:00

ggml-cuda.h

update llama.cpp

2023-08-03 11:50:24 -07:00

ggml-metal.h

update llama.cpp

2023-08-03 11:50:24 -07:00

ggml-metal.m

update llama.cpp

2023-08-03 11:50:24 -07:00

ggml-metal.metal

update llama.cpp

2023-08-03 11:50:24 -07:00

ggml-mpi.c

update llama.cpp

2023-08-03 11:50:24 -07:00

ggml-mpi.h

update llama.cpp

2023-08-03 11:50:24 -07:00

ggml-opencl.cpp

update llama.cpp

2023-08-03 11:50:24 -07:00

ggml-opencl.h

update llama.cpp

2023-08-03 11:50:24 -07:00

ggml.c

update llama.cpp

2023-08-03 11:50:24 -07:00

ggml.h

update llama.cpp

2023-08-03 11:50:24 -07:00

k_quants.c

update llama.cpp

2023-08-03 11:50:24 -07:00

k_quants.h

update llama.cpp

2023-08-03 11:50:24 -07:00

llama_darwin.go

override ggml-metal if the file is different

2023-08-02 12:50:30 -07:00

llama-util.h

update llama.cpp

2023-08-03 11:50:24 -07:00

llama.cpp

update llama.cpp

2023-08-03 11:50:24 -07:00

llama.go

automatically set num_keep if num_keep < 0

2023-08-07 16:19:12 -07:00

llama.h

update llama.cpp

2023-08-03 11:50:24 -07:00

update-llama-cpp.sh

update llama.cpp to d91f3f0

2023-07-28 08:07:48 -04:00

utils.go

update predict code

2023-07-27 09:31:44 -07:00