 Georgi Gerganov
|
5437d4aaf5
sync : ggml
|
hai 1 ano |
|
Georgi Gerganov
|
78f766768d
cmake : fix "amd64" processor string (whisper/2638)
|
hai 1 ano |
 gn64
|
8dd19a4812
vulkan : fix soft_max.comp division by zero (whisper/2633)
|
hai 1 ano |
 Daniel Bevenius
|
130d0c90bd
ggml : remove return from ggml_gallocr_allocate_node (ggml/1048)
|
hai 1 ano |
|
Daniel Bevenius
|
3919da8e33
ggml : add check for grad_accs (ggml/1046)
|
hai 1 ano |
|
Georgi Gerganov
|
0006f5a74a
ggml : update ggml_backend_cpu_device_supports_op (#10867)
|
hai 1 ano |
 krystiancha
|
05c3a444b8
server : fill usage info in embeddings and rerank responses (#10852)
|
hai 1 ano |
 Billel Mokeddem
|
382bc7f2e8
llama : add Falcon3 support (#10864)
|
hai 1 ano |
 Ruan
|
4f51968aca
readme : update typos (#10863)
|
hai 1 ano |
 Xuan Son Nguyen
|
227d7c5a7f
server : (UI) fix missing async generator on safari (#10857)
|
hai 1 ano |
 Eve
|
7b1ec53f56
vulkan: bugfixes for small subgroup size systems + llvmpipe test (#10809)
|
hai 1 ano |
 Zhiyuan Li
|
160bc039c8
rwkv6: add wkv6 support for Vulkan backend (#10829)
|
hai 1 ano |
|
Georgi Gerganov
|
08ea539df2
unicode : improve naming style (#10838)
|
hai 1 ano |
|
Georgi Gerganov
|
644fd71b44
sampling : refactor + optimize penalties sampler (#10803)
|
hai 1 ano |
 Bartowski
|
4ddd199f6f
llava : Allow locally downloaded models for QwenVL (#10833)
|
hai 1 ano |
 Valentin Mamedov
|
a0974156f3
llama : add Deepseek MoE v1 & GigaChat models (#10827)
|
hai 1 ano |
|
Georgi Gerganov
|
87cf323cef
scripts : change build path to "build-bench" for compare-commits.sh (#10836)
|
hai 1 ano |
 Vinesh Janarthanan
|
5478bbcd17
server: (UI) add syntax highlighting and latex math rendering (#10808)
|
hai 1 ano |
|
Georgi Gerganov
|
b5ae1ddff9
gguf-py : bump to v0.13.0
|
hai 1 ano |
 Michelle Tan
|
89d604f2c8
server: Fix `has_next_line` in JSON response (#10818)
|
hai 1 ano |
 Evgeny Kurnevsky
|
e52aba537a
nix: allow to override rocm gpu targets (#10794)
|
hai 1 ano |
 HimariO
|
ba1cb19cdd
llama : add Qwen2VL support + multimodal RoPE (#10361)
|
hai 1 ano |
 cduk
|
56eea0781c
Removes spurious \r in output that causes logging in journalctl to treat lines as binary and therefore hidden by default (#10771)
|
hai 1 ano |
 lhez
|
a76c56fa1a
Introducing experimental OpenCL backend with support for Qualcomm Adreno GPUs (#10693)
|
hai 1 ano |
 Eric Curtin
|
c27ac678dd
Opt class for positional argument handling (#10508)
|
hai 1 ano |
 Corentin REGAL
|
11e07fd63b
fix: graceful shutdown for Docker images (#10815)
|
hai 1 ano |
 Jett Janiak
|
4601a8bb67
gguf-py : numpy 2 newbyteorder fix (#9772)
|
hai 1 ano |
 谢乃闻
|
9f35e44592
Fix crash caused by ggml_backend_load_all when launching on Android Activity (#10812)
|
hai 1 ano |
|
Eve
|
64ae065511
vulkan: small mul_mat_vec optimizations (#10665)
|
hai 1 ano |
 Akarshan Biswas
|
83ed24a97b
SYCL: Reduce most of the compiler warnings (#10748)
|
hai 1 ano |
