ggerganov/llama.cpp

Pull Requests Commits
mtmd : add **vision** support for Mistral Small 3.1 (#13231)
ngxson committed 161 days ago
Verified 8936784f
arg : remove CURLINFO_EFFECTIVE_METHOD (#13228)
ngxson committed 161 days ago
Verified 13c9a331
llama-model : fix the reported size class for nomic-embed-text-v2-moe (#13223)
cebtenzzre committed 162 days ago
Verified a70183eb
sync : ggml
ggerganov committed 162 days ago
8d33d740
ggml : fix ggml_gallocr_ptr type (ggml/1205)
slaren committed 162 days ago
4254bb49
cuda : fix unused variable compile warning (whisper/0)
ggerganov committed 162 days ago
99985401
CUDA: batched+noncont MMQ, refactor bs>1 MoE code (#13199)
JohannesGaessler committed 162 days ago
Verified e1e8e099
arg : -hf do not fail if url mismatch (#13219)
ngxson committed 162 days ago
Verified 6f67cf1f
fix typo: `n_ctx_pre_seq` -> `n_ctx_per_seq` (#13221)
ddh0 committed 162 days ago
Verified 16a457fa
convert : improve model arch handling (#13122)
ngxson committed 162 days ago
Verified 3e168bed
llava : remove duplicate include (#13207)
tattn committed 162 days ago
Verified ceda28ef
common : add -jf / --json-schema-file flag (#12011)
ochafik committed 162 days ago
Verified 3b127c73
vulkan: use uint array index to avoid glslang bug (#13193)
jeffbolznv committed 162 days ago
Verified e5007a5e
ggml : fix ppc64le build (#13176)
shalinib-ibm committed 162 days ago
Verified 41631377
convert : correct typo image_mean --> image_std (#13208)
ngxson committed 162 days ago
Verified 07c2e2f7
feat(ggml-cpu): enable z17 compile (#13182)
taronaeo committed 162 days ago
Verified 44cd8d91
arg : allow using -hf offline (#13202)
ngxson committed 162 days ago
Verified 5933e6fd
docker : do not build tests (#13204)
ngxson committed 162 days ago
Verified da84c04d
rpc : fix cache directory initialization (#13188)
hbuxiaofei committed 163 days ago
Verified a0f7016d
scripts: n_depth for compare-llama-bench [no ci] (#13201)
JohannesGaessler committed 163 days ago
Verified 19e899ce
server : Prefilling assistant message in openai compatible API (#13174)
matteoserva committed 163 days ago
Verified e2e1ddb9
sampling : when top-k <= 0 -> noop (#13173)
ggerganov committed 163 days ago
Verified d9d398f8
llama-bench: fixed size of fields to correctly map to values (#13183)
Alcpz committed 163 days ago
Verified 5a639801
CUDA: fix non-cont. inputs for batched mat mul (#13155)
JohannesGaessler committed 163 days ago
Verified cdf76586
llama : llm_type order by size (#13177)
CISC committed 163 days ago
Verified 7d3af70b
mtmd : add qwen2vl and qwen2.5vl (#13141)
ngxson committed 163 days ago
Verified 00e3e5a1
llama : set qwen3 model type sizes (#13175)
CISC committed 163 days ago
Verified e98b3692
llama-graph : fix text position for mrope (#13159)
ngxson committed 164 days ago
Verified b6ce7430
model : Nomic Embed Text V2 with Mixture-of-Experts (MoE) architecture (#12466)
manyoso committed 164 days ago
Verified 5f5e39e1
clip : fix model size display (#13153)
ngxson committed 164 days ago
Verified eaea3253
Terms Of Use Privacy Policy Refund Policy Impressum