ngxson/llama.cpp

Pull Requests Commits

ci : add -no-cnv for tests

ngxson committed 342 days ago

07fe2cbb

vocab : add dummy tokens for "no_vocab" type (#11231)

ggerganov committed 342 days ago

Verified bbf3e55e

server : Improve code snippets direction between RTL text (#11221)

ebraminio committed 342 days ago

Verified c5bf0d1b

Refactor test-chat-template.cpp (#11224)

ochafik committed 342 days ago

Verified 091592d7

ggerganov committed 342 days ago

Verified 44d1e796

scripts : sync gguf (cont)

ggerganov committed 342 days ago

Verified a4f3f5d8

scripts : sync gguf

ggerganov committed 342 days ago

Verified 48e1ae0e

scripts : sync opencl

ggerganov committed 342 days ago

Verified d00a80e8

server : (UI) Improve messages bubble shape in RTL (#11220)

ebraminio committed 342 days ago

Verified 504af20e

cli : auto activate conversation mode if chat template is available (#11214)

ngxson committed 342 days ago

Verified 84a44815

cuda : CUDA Graph Compute Function Refactor (precursor for performance improvements) (#11042)

aendk committed 343 days ago

Verified 39509fb0

contrib : add naming guidelines (cont) (#11177)

ggerganov committed 343 days ago

Verified a29f0870

server : (UI) Support for RTL text as models input or output (#11208)

ebraminio committed 343 days ago

Verified 437e05f7

contrib : add naming guidelines (cont) (#11177)

ggerganov committed 343 days ago

Verified ca001f66

common : support tag-based --hf-repo like on ollama (#11195)

ngxson committed 343 days ago

Verified 00b4c3da

contrib : add naming guidelines (#11177)

ggerganov committed 343 days ago

Verified 7426a26b

llama : remove 'd' from bad special token log (#11212)

danbev committed 343 days ago

Verified 8f70fc3d

ggml : do not define GGML_USE_CUDA when building with GGML_BACKEND_DL (#11211)

rgerganov committed 343 days ago

Verified 1244cdcf

Reset color before we exit (#11205)

ericcurtin committed 344 days ago

Verified 924518e2

llama : fix chat template gguf key (#11201)

ngxson committed 344 days ago

Verified 9a483999

llama : remove notion of CLS token (#11064)

ggerganov committed 344 days ago

Verified 08f10f69

llama : add `llama_vocab`, functions -> methods, naming (#11110)

ggerganov committed 344 days ago

Verified afa8a9ec

gguf-py: fixed local detection of gguf package (#11180)

VJHack committed 345 days ago

Verified c05e8c99

convert : sort print supported models [no ci] (#11179)

danbev committed 345 days ago

Verified 2739a71e

examples : add README.md to tts example [no ci] (#11155)

danbev committed 346 days ago

Verified ba8a1f9c

convert : add --print-supported-models option (#11172)

danbev committed 346 days ago

Verified ff3fcabc

Vulkan: Fix float16 use on devices without float16 support + fix subgroup_size_control validation error (#11161)

0cc4m committed 346 days ago

Verified c3f9d257

llama: add support for QRWKV6 model architecture (#11001)

MollySophia committed 346 days ago

Verified ee7136c6

SYCL: Refactor ggml_sycl_compute_forward (#11121)

qnixsynapse committed 346 days ago

Verified c6860cc7

doc: add cuda guide for fedora (#11135)

teihome committed 347 days ago

Verified 1204f972

Older