ngxson/llama.cpp

Pull Requests Commits
sampling : when top-k <= 0 -> noop (#13173)
ggerganov committed 1 year ago
Verified d9d398f8
rm outdated comment
ngxson committed 1 year ago
08d0f9cf
test: mtmd_input_chunks_free
ngxson committed 1 year ago
111d5afa
rm redundant define
ngxson committed 1 year ago
92d24045
add test target
ngxson committed 1 year ago
33579615
llama-bench: fixed size of fields to correctly map to values (#13183)
Alberto Cabrera Pérez committed 1 year ago
Verified 5a639801
add mtmd::bitmaps
ngxson committed 1 year ago
f8c27b9e
CUDA: fix non-cont. inputs for batched mat mul (#13155)
JohannesGaessler committed 1 year ago
Verified cdf76586
working version
ngxson committed 1 year ago
82f4246e
llama : llm_type order by size (#13177)
CISC committed 1 year ago
Verified 7d3af70b
Merge branch 'master' into xsn/mtmd_c_api
ngxson committed 1 year ago
e0806c29
wip
ngxson committed 1 year ago
f6b6517c
mtmd : add qwen2vl and qwen2.5vl (#13141)
ngxson committed 1 year ago
Verified 00e3e5a1
llama : set qwen3 model type sizes (#13175)
CISC committed 1 year ago
Verified e98b3692
init
ngxson committed 1 year ago
4a4f35c8
llama-graph : fix text position for mrope (#13159)
ngxson committed 1 year ago
Verified b6ce7430
model : Nomic Embed Text V2 with Mixture-of-Experts (MoE) architecture (#12466)
manyoso committed 1 year ago
Verified 5f5e39e1
clip : fix model size display (#13153)
ngxson committed 1 year ago
Verified eaea3253
fix(rpc): Improve input validation and error handling (#13069)
thevilledev committed 1 year ago
Verified 43ddab6e
llama-bench: add `-d` depth arg (#13096)
thevishalagarwal committed 1 year ago
Verified 1831f538
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
Terms Of Use Privacy Policy Refund Policy Impressum