ggerganov/llama.cpp

Pull Requests Commits

ggerganov committed 47 days ago

Verified 098f0e5e

CANN: Support more ops (#12841)

noemotiovon committed 48 days ago

Verified fe5b78c8

Fixes #12823 (#12830)

mehendarkarprajwal committed 48 days ago

Verified 11d07e1e

docker : added all CPU to GPU images (#12749)

rudiservo committed 48 days ago

Verified b0091ecc

ggml-cpu-impl.h: do not redefine bool on POWER9 (#12856)

pkubaj committed 48 days ago

Verified 31f7803b

ggml-impl.h: fix build on POWER9 (#12855)

pkubaj committed 48 days ago

Verified 2391506a

llama : Support Qwen3 and Qwen3MoE (#12828)

bozheng-hit committed 48 days ago

Verified d3bd7193

musa: enable freediskspace for docker image build (#12839)

yeahdongcn committed 48 days ago

Verified d9a63b2f

sycl: update documentation to use -no-cnv (#12845)

Rbiessy committed 48 days ago

Verified 8ed71242

ci: detach common from the library (#12827)

pminev committed 48 days ago

Verified 381603a7

clip : do not print ftype (#12832)

ngxson committed 48 days ago

Verified 65a69e6e

readme : add rpc backend (#12842)

ggerganov committed 48 days ago

Verified 47277d6d

CANN: Support Opt CONV_TRANSPOSE_1D and ELU (#12786)

noemotiovon committed 48 days ago

Verified 6e1c4ceb

vulkan: In coopmat2 mmq, load q4_k/q5_k scales through shared memory (#12833)

jeffbolznv committed 48 days ago

Verified 0090950f

vulkan: Use fp16 for the flash attention P*V multiplication (#12783)

jeffbolznv committed 48 days ago

Verified 7ecd780b

cuda : add f32 to bf16 copy op (#12806)

CISC committed 49 days ago

Verified 7538246e

llava: improve clip_ctx destructor to not memleak load_image_size (#12834)

mattjcly committed 49 days ago

Verified b32efad2

llama : fix FA when KV cache is not used (i.e. embeddings) (#12825)

ggerganov committed 49 days ago

Verified a19b5cef

server : fix thread.join() on exit (#12831)

ngxson committed 49 days ago

Verified 78a1ba0a

llava: add more helper functions to check projector types in clip context (#12824)

dm4 committed 49 days ago

Verified 2dabf759

arg : Including limits file on AIX (#12822)

mehendarkarprajwal committed 49 days ago

Verified 1d343b40

server : webui : Improve Chat Input with Auto-Sizing Textarea (#12785)

characharm committed 49 days ago

Verified 8ca6e1c3

Revert "sycl:remove redundant memcopy in function ggml_backend_sycl_buffer_set_tensor" (#12812)

NeoZhangJianyu committed 49 days ago

Verified 656babd6

gguf-py : support lazy tensor splitting (#12809)

compilade committed 49 days ago

Verified a226bc7a

llama : Support llama 4 text-only (#12791)

ngxson committed 50 days ago

Verified 1466621e

opencl: better identify Adreno GPU (#12760)

lhez committed 50 days ago

Verified 82974011

hellaswag: display estimated score confidence interval (#12797)

stduhpf committed 50 days ago

Verified 4ccea213

cuda : fix HIP and MUSA BF16 (#0)

ggerganov committed 50 days ago

1a1ab7e7

ggerganov committed 50 days ago

a4e46e28

ggml : simplify Arm fp16 CPU logic (ggml/1177)

ggerganov committed 50 days ago

ff067dbc

Older