ngxson/llama.cpp

Pull Requests Commits
clip : fix model size display
ngxson committed 1 year ago
7135715e
mtmd : fix glm-edge redundant token count (#13139)
ngxson committed 1 year ago
Verified 4e87962e
context : do not clear output buffer on reserve (#13152)
pockers21 committed 1 year ago
Verified fb0471d1
llama : (mrope) allow using normal 1D position for text token (#13138)
ngxson committed 1 year ago
Verified d2b2031e
clip : refactor set input for cgraph + fix qwen2.5vl input (#13136)
ngxson committed 1 year ago
Verified 5fa9e63b
SYCL: Add all missing unary kernels (#13074)
qnixsynapse committed 1 year ago
Verified a4c340f9
readme : update hot topics (#13150)
ggerganov committed 1 year ago
Verified d0a417f3
common : fix noreturn compile warning (#13151)
ggerganov committed 1 year ago
Verified 43f2b071
llama-chat : fix typo GML --> GLM (#13143)
ngxson committed 1 year ago
Verified e5d6c255
musa: fix typo in cc control (#13144)
yeahdongcn committed 1 year ago
Verified f0dd6a19
CUDA: fix q_nope_absorbed prec for DS 2 Lite f16 (#13137)
JohannesGaessler committed 1 year ago
Verified 69699be4
arg : fix unused variable (#13142)
ngxson committed 1 year ago
Verified 85f36e5e
llama-bench : Add `--override-tensors` arg (#12922)
4onen committed 1 year ago
Verified c0a97b76
llama-chat : fix wrong template in GLM4-0414 (#13140)
matteoserva committed 1 year ago
Verified ced44be3
musa: fix build warning (#13129)
yeahdongcn committed 1 year ago
Verified e291450b
Fixes Qwen2.5VL segfault during inference with https://github.com/ggml-org/llama.cpp/pull/12402 as has_qwen2vl_merger migration was incomplete (#13133)
LostRuins committed 1 year ago
Verified 59e991c2
clip : Add Qwen2.5VL support (#12402)
HimariO committed 1 year ago
Verified ca2bb89e
common : add common_remote_get_content (#13123)
ngxson committed 1 year ago
Verified 2d451c80
clip : improve projector naming (#13118)
ngxson committed 1 year ago
Verified 4753791e
ggml: move fp16/bf16 conversion optimizations to CPU backend + export conversion APIs (#13107)
SongXiaoXi committed 1 year ago
Verified 77d5e9a7
grammar : handle maxItems == 0 in JSON schema (#13117)
rick-github committed 1 year ago
Verified d5fe4e81
llama : fix K-shift with quantized K and BLAS backend (#13113)
slaren committed 1 year ago
Verified 295354ea
Force FP32 compute in GLM4 FFN Down (#13101)
city96 committed 1 year ago
Verified 558a7647
clip : fix pixtral on some GPU backends (#13097)
ngxson committed 1 year ago
Verified edb18b6e
change the reorder tensor from init to execute OP (#13003)
NeoZhangJianyu committed 1 year ago
Verified 514c4560
rpc : do not wait for response when sending RPC_CMD_SET_TENSOR (#12943)
rgerganov committed 1 year ago
Verified 553a5c3a
clip : remove boi/eoi embeddings for GLM-edge model (#13081)
ngxson committed 1 year ago
Verified 13be08da
embeddings : fix batch sizes (#13076)
ggerganov committed 1 year ago
Verified 226251ed
ggml : fix trailing whitespaces (#0)
ggerganov committed 1 year ago
87616f06
sync : ggml
ggerganov committed 1 year ago
63b49114
Terms Of Use Privacy Policy Refund Policy Impressum