ngxson/llama.cpp

Pull Requests Commits

ngxson committed 1 year ago

6ad1f8da

server : (UI) fix missing async generator on safari

ngxson committed 1 year ago

5ba1fda5

rwkv6: add wkv6 support for Vulkan backend (#10829)

zhiyuan1i committed 1 year ago

Verified 160bc039

unicode : improve naming style (#10838)

ggerganov committed 1 year ago

Verified 08ea539d

sampling : refactor + optimize penalties sampler (#10803)

ggerganov committed 1 year ago

Verified 644fd71b

llava : Allow locally downloaded models for QwenVL (#10833)

bartowski1182 committed 1 year ago

Verified 4ddd199f

llama : add Deepseek MoE v1 & GigaChat models (#10827)

Inf1delis committed 1 year ago

Verified a0974156

scripts : change build path to "build-bench" for compare-commits.sh (#10836)

ggerganov committed 1 year ago

Verified 87cf323c

server: (UI) add syntax highlighting and latex math rendering (#10808)

VJHack committed 1 year ago

Verified 5478bbcd

gguf-py : bump to v0.13.0

ggerganov committed 1 year ago

Verified b5ae1ddf

server: Fix `has_next_line` in JSON response (#10818)

MichelleTanPY committed 1 year ago

Verified 89d604f2

nix: allow to override rocm gpu targets (#10794)

kurnevsky committed 1 year ago

Verified e52aba53

llama : add Qwen2VL support + multimodal RoPE (#10361)

HimariO committed 1 year ago

Verified ba1cb19c

Removes spurious \r in output that causes logging in journalctl to treat lines as binary and therefore hidden by default (#10771)

cduk committed 1 year ago

Verified 56eea078

Introducing experimental OpenCL backend with support for Qualcomm Adreno GPUs (#10693)

lhez committed 1 year ago

Verified a76c56fa

Opt class for positional argument handling (#10508)

ericcurtin committed 1 year ago

Verified c27ac678

fix: graceful shutdown for Docker images (#10815)

co42 committed 1 year ago

Verified 11e07fd6

gguf-py : numpy 2 newbyteorder fix (#9772)

jettjaniak committed 1 year ago

Verified 4601a8bb

Fix crash caused by ggml_backend_load_all when launching on Android Activity (#10812)

sienaiwun committed 1 year ago

Verified 9f35e445

vulkan: small mul_mat_vec optimizations (#10665)

netrunnereve committed 1 year ago

Verified 64ae0655

SYCL: Reduce most of the compiler warnings (#10748)

qnixsynapse committed 1 year ago

Verified 83ed24a9

ggml : Fix compilation issues on ARM platform when building without fp16 (#10811)

kkontny committed 1 year ago

Verified d583cd03

common : improve -ctv -ctk CLI arguments (#10806)

ngxson committed 1 year ago

Verified adffa6ff

contrib : add ngxson as codeowner (#10804)

ngxson committed 1 year ago

Verified 274ec65a

CUDA: faster non-contiguous concat (#10760)

A3shTnT committed 1 year ago

Verified 8faa1d4d

remove CMAKE_WINDOWS_EXPORT_ALL_SYMBOLS (#10797)

slaren committed 1 year ago

Verified cb13ef85

Vulkan: Use improved q4_k and q5_k dequant code in dequant shaders (#10798)

0cc4m committed 1 year ago

Verified 4064c0e3

Vulkan: Add VK_EXT_subgroup_size_control support to ensure full subgroups for coopmats (#10721)

0cc4m committed 1 year ago

Verified dc5301d5

common : add missing env var for speculative (#10801)

ngxson committed 1 year ago

Verified 9fdb1243

docs: update server streaming mode documentation (#9519)

CentricStorm committed 1 year ago

Verified 5555c0c1

Older