 Evan Miller
|
e8035f141e
ggml : fix static_assert with older compilers #2024 (#2218)
|
пре 2 година |
 Bach Le
|
7513b7b0a1
llama : add functions that work directly on model (#2197)
|
пре 2 година |
 Ali Chraghi
|
de8342423d
build.zig : install config header (#2216)
|
пре 2 година |
 Shangning Xu
|
c48c525f87
examples : fixed path typos in embd-input (#2214)
|
пре 2 година |
 Jiahao Li
|
206e01de11
cuda : support broadcast add & mul (#2192)
|
пре 2 година |
 Johannes Gäßler
|
4304bd3cde
CUDA: mul_mat_vec_q kernels for k-quants (#2203)
|
пре 2 година |
 James Reynolds
|
229aab351c
make : fix combination of LLAMA_METAL and LLAMA_MPI (#2208)
|
пре 2 година |
 Georgi Gerganov
|
697966680b
ggml : sync (ggml_conv_2d, fix mul_mat bug, CUDA GLM rope)
|
пре 2 година |
 Kawrakow
|
27ad57a69b
Metal: faster Q4_0 and Q4_1 matrix x vector kernels (#2212)
|
пре 2 година |
 Howard Su
|
32c5411631
Revert "Support using mmap when applying LoRA (#2095)" (#2206)
|
пре 2 година |
|
Howard Su
|
ff5d58faec
Fix compile error on Windows CUDA (#2207)
|
пре 2 година |
 Bodo Graumann
|
b782422a3e
devops : add missing quotes to bash script (#2193)
|
пре 2 година |
 Shouzheng Liu
|
1cbf561466
metal : new q4_0 matrix-vector kernel (#2188)
|
пре 2 година |
|
Georgi Gerganov
|
975221e954
ggml : broadcast mul_mat + conv batch support (#2199)
|
пре 2 година |
|
Georgi Gerganov
|
4523d10d0c
ggml : add ggml_pool_1d and ggml_pool_2d
|
пре 2 година |
|
Georgi Gerganov
|
680e6f9177
cuda : add gelu support
|
пре 2 година |
|
Howard Su
|
4e7464ef88
FP16 is supported in CM=6.0 (#2177)
|
пре 2 година |
|
Johannes Gäßler
|
2b5eb72e10
Fixed __dp4a compute capability: 6.0 -> 6.1 (#2189)
|
пре 2 година |
|
Georgi Gerganov
|
f7d278faf3
ggml : revert CUDA broadcast changes from #2183 (#2191)
|
пре 2 година |
|
Georgi Gerganov
|
20d7740a9b
ggml : sync (abort callback, mul / add broadcast, fix alibi) (#2183)
|
пре 2 година |
 Spencer Sutton
|
5bf2a27718
ggml : remove src0 and src1 from ggml_tensor and rename opt to src (#2178)
|
пре 2 година |
|
Bach Le
|
c9c74b4e3f
llama : add classifier-free guidance (#2135)
|
пре 2 година |
 Jinwoo Jeong
|
3ec7e596b2
docker : add '--server' option (#2174)
|
пре 2 година |
 Chad Brewbaker
|
917831c63a
readme : fix zig build instructions (#2171)
|
пре 2 година |
|
Howard Su
|
2347463201
Support using mmap when applying LoRA (#2095)
|
пре 2 година |
 LostRuins
|
bbef28218f
Possible solution to allow K-quants on models with n_vocab!=32000 (#2148)
|
пре 2 година |
|
Evan Miller
|
5656d10599
mpi : add support for distributed inference via MPI (#2099)
|
пре 2 година |
 oobabooga
|
1d16309969
llama : remove "first token must be BOS" restriction (#2153)
|
пре 2 година |
 Nigel Bosch
|
db4047ad5c
main : escape prompt prefix/suffix (#2151)
|
пре 2 година |
 JackJollimore
|
18780e0a5e
readme : update Termux instructions (#2147)
|
пре 2 година |
