ggerganov/llama.cpp

Pull Requests Commits
ci: detach common from the library (#12827)
pminev committed 305 days ago
Verified 381603a7
clip : do not print ftype (#12832)
ngxson committed 305 days ago
Verified 65a69e6e
readme : add rpc backend (#12842)
ggerganov committed 305 days ago
Verified 47277d6d
CANN: Support Opt CONV_TRANSPOSE_1D and ELU (#12786)
noemotiovon committed 305 days ago
Verified 6e1c4ceb
vulkan: In coopmat2 mmq, load q4_k/q5_k scales through shared memory (#12833)
jeffbolznv committed 305 days ago
Verified 0090950f
vulkan: Use fp16 for the flash attention P*V multiplication (#12783)
jeffbolznv committed 305 days ago
Verified 7ecd780b
gguf-py : add more clarifying comments for multi-thread writes
compilade committed 305 days ago
d8bab9ef
cuda : add f32 to bf16 copy op (#12806)
CISC committed 305 days ago
Verified 7538246e
convert : write tensors in parallel
compilade committed 305 days ago
06e1d311
llava: improve clip_ctx destructor to not memleak load_image_size (#12834)
mattjcly committed 305 days ago
Verified b32efad2
llama : fix FA when KV cache is not used (i.e. embeddings) (#12825)
ggerganov committed 305 days ago
Verified a19b5cef
server : fix thread.join() on exit (#12831)
ngxson committed 305 days ago
Verified 78a1ba0a
llava: add more helper functions to check projector types in clip context (#12824)
dm4 committed 306 days ago
Verified 2dabf759
arg : Including limits file on AIX (#12822)
mehendarkarprajwal committed 306 days ago
Verified 1d343b40
server : webui : Improve Chat Input with Auto-Sizing Textarea (#12785)
characharm committed 306 days ago
Verified 8ca6e1c3
Revert "sycl:remove redundant memcopy in function ggml_backend_sycl_buffer_set_tensor" (#12812)
NeoZhangJianyu committed 306 days ago
Verified 656babd6
gguf-py : support lazy tensor splitting (#12809)
compilade committed 306 days ago
Verified a226bc7a
llama : Support llama 4 text-only (#12791)
ngxson committed 306 days ago
Verified 1466621e
opencl: better identify Adreno GPU (#12760)
lhez committed 306 days ago
Verified 82974011
hellaswag: display estimated score confidence interval (#12797)
stduhpf committed 306 days ago
Verified 4ccea213
cuda : fix HIP and MUSA BF16 (#0)
ggerganov committed 306 days ago
1a1ab7e7
sync : ggml
ggerganov committed 306 days ago
a4e46e28
ggml : simplify Arm fp16 CPU logic (ggml/1177)
ggerganov committed 306 days ago
ff067dbc
CUDA: don't convert BF16 weights to FP32 (ggml/1174)
CISC committed 306 days ago
36ca8b36
cpu: move all the operators into a separate c++ file (except mul_mat) (ggml/1167)
cmdr2 committed 306 days ago
995083e4
sycl: remove redundant memcopy in function ggml_backend_sycl_buffer_set_tensor (#12734)
jeffzhou2000 committed 306 days ago
Verified 518a0148
ci : no curl on ggml-ci (#12796)
ngxson committed 307 days ago
Verified e391d3ee
cmake : enable curl by default (#12761)
ngxson committed 307 days ago
Verified bd3f59f8
CANN: fix typo in ggml-cann (#12733)
jeffzhou2000 committed 307 days ago
Verified 52b3d71f
CANN: Refactor to reduce duplicate code (#12731)
hipudding committed 307 days ago
Verified d0d5b223
Terms Of Use Privacy Policy Refund Policy Impressum