 Xuan-Son Nguyen
|
8b9cc7cdd8
llava : introduce libmtmd (#12849)
|
пре 9 месеци |
|
Xuan-Son Nguyen
|
64eda5deb9
convert : ability to lazy-load safetensors remotely without downloading to disk (#12820)
|
пре 9 месеци |
 Chenguang Li
|
fe5b78c896
CANN: Support more ops (#12841)
|
пре 9 месеци |
 Prajwal B Mehendarkar
|
11d07e1e69
Fixes #12823 (#12830)
|
пре 9 месеци |
 Rudi Servo
|
b0091ecc1e
docker : added all CPU to GPU images (#12749)
|
пре 9 месеци |
 Piotr Kubaj
|
31f7803bc4
ggml-cpu-impl.h: do not redefine bool on POWER9 (#12856)
|
пре 9 месеци |
|
Piotr Kubaj
|
2391506ace
ggml-impl.h: fix build on POWER9 (#12855)
|
пре 9 месеци |
 Bo Zheng
|
d3bd7193ba
llama : Support Qwen3 and Qwen3MoE (#12828)
|
пре 9 месеци |
 R0CKSTAR
|
d9a63b2f2e
musa: enable freediskspace for docker image build (#12839)
|
пре 9 месеци |
 Romain Biessy
|
8ed71242f4
sycl: update documentation to use -no-cnv (#12845)
|
пре 9 месеци |
 Plamen Minev
|
381603a775
ci: detach common from the library (#12827)
|
пре 9 месеци |
|
Xuan-Son Nguyen
|
65a69e6e1b
clip : do not print ftype (#12832)
|
пре 9 месеци |
 Georgi Gerganov
|
47277d6d1d
readme : add rpc backend (#12842)
|
пре 9 месеци |
|
Chenguang Li
|
6e1c4cebdb
CANN: Support Opt CONV_TRANSPOSE_1D and ELU (#12786)
|
пре 9 месеци |
 Jeff Bolz
|
0090950f67
vulkan: In coopmat2 mmq, load q4_k/q5_k scales through shared memory (#12833)
|
пре 9 месеци |
|
Jeff Bolz
|
7ecd780b1a
vulkan: Use fp16 for the flash attention P*V multiplication (#12783)
|
пре 9 месеци |
 Sigbjørn Skjæret
|
7538246e7c
cuda : add f32 to bf16 copy op (#12806)
|
пре 9 месеци |
 Matt Clayton
|
b32efad2bc
llava: improve clip_ctx destructor to not memleak load_image_size (#12834)
|
пре 9 месеци |
|
Georgi Gerganov
|
a19b5cef16
llama : fix FA when KV cache is not used (i.e. embeddings) (#12825)
|
пре 9 месеци |
|
Xuan-Son Nguyen
|
78a1ba0a4f
server : fix thread.join() on exit (#12831)
|
пре 9 месеци |
 dm4
|
2dabf759e7
llava: add more helper functions to check projector types in clip context (#12824)
|
пре 9 месеци |
|
Prajwal B Mehendarkar
|
1d343b4069
arg : Including limits file on AIX (#12822)
|
пре 9 месеци |
 characharm
|
8ca6e1c3a4
server : webui : Improve Chat Input with Auto-Sizing Textarea (#12785)
|
пре 9 месеци |
 Neo Zhang Jianyu
|
656babd6c2
Revert "sycl:remove redundant memcopy in function ggml_backend_sycl_buffer_set_tensor" (#12812)
|
пре 9 месеци |
 compilade
|
a226bc7a9a
gguf-py : support lazy tensor splitting (#12809)
|
пре 9 месеци |
|
Xuan-Son Nguyen
|
1466621e73
llama : Support llama 4 text-only (#12791)
|
пре 9 месеци |
 lhez
|
82974011f3
opencl: better identify Adreno GPU (#12760)
|
пре 9 месеци |
 stduhpf
|
4ccea213bc
hellaswag: display estimated score confidence interval (#12797)
|
пре 9 месеци |
|
Georgi Gerganov
|
1a1ab7e7a4
cuda : fix HIP and MUSA BF16 (#0)
|
пре 9 месеци |
|
Georgi Gerganov
|
a4e46e28f9
sync : ggml
|
пре 9 месеци |
