ggerganov/llama.cpp

Pull Requests Commits

gguf-py : do not use title case for naming convention

compilade committed 302 days ago

1932a1b8

convert_lora : fix default filename

compilade committed 302 days ago

bf8e71b0

gguf-py : add more name metadata extraction tests

compilade committed 302 days ago

a3d154b2

gguf-py : more metadata edge cases fixes

compilade committed 303 days ago

912e6fa5

gguf-py : fix some metadata name extraction edge cases

compilade committed 303 days ago

2164c9de

ggml : fix quant dot product with odd number of blocks (#8549)

slaren committed 304 days ago

Verified 87e397d0

convert-*.py: remove add_name from ChatGLMModel class (#8590)

mofosyne committed 304 days ago

Verified 57b1d4f9

llama : bump max layers from 256 to 512 (#8530)

ggerganov committed 304 days ago

Verified d1975455

readme : fix server badge

ggerganov committed 304 days ago

Verified be0cfb41

ggml : add friendlier error message to fopen errors (#8575)

HanClinto committed 304 days ago

Verified b57eb9ca

fix: typo of chatglm4 chat tmpl (#8586)

thxCode committed 304 days ago

Verified f299aa98

convert-*.py: add general.name kv override (#8571)

mofosyne committed 304 days ago

Verified 3d0e4367

CUDA: fix partial offloading for ne0 % 256 != 0 (#8572)

JohannesGaessler committed 304 days ago

Verified a15ef8f8

cmake : install all ggml public headers (#8480)

65a committed 305 days ago

Verified 705b7ecf

server: use relative routes for static files in new UI (#8552)

EZForever committed 305 days ago

Verified 0d2c7321

convert-*.py: GGUF Naming Convention Refactor and Metadata Override Refactor (#7499)

mofosyne committed 305 days ago

Verified 672a6f10

server : respect `--special` cli arg (#8553)

RunningLeon committed 305 days ago

Verified 3807c3de

lookup: fibonacci hashing, fix crashes (#8548)

JohannesGaessler committed 305 days ago

Verified e02b597b

build : Fix docker build warnings (#8535) (#8537)

amochkin committed 305 days ago

Verified b3283448

CONTRIBUTING.md : remove mention of noci (#8541)

mofosyne committed 306 days ago

Verified 30f80ca0

[CANN] Add Ascend NPU backend (#6035)

hipudding committed 306 days ago

Verified 1bdd8ae1

batched: fix n_predict parameter (#8527)

msy-kato committed 306 days ago

Verified da3913d8

llama : disable context-shift for DeepSeek v2 (#8501)

ggerganov committed 306 days ago

Verified d65a8361

make/cmake: add missing force MMQ/cuBLAS for HIP (#8515)

JohannesGaessler committed 306 days ago

Verified 5e116e8d

gguf-hash : update clib.json to point to original xxhash repo (#8491)

mofosyne committed 307 days ago

Verified 1666f92d

export-lora : handle help argument (#8497)

sbonds committed 307 days ago

Verified 37b12f92

llama : valign + remove unused ftype (#8502)

ggerganov committed 307 days ago

Verified 0efec577

convert_hf : faster lazy safetensors (#8482)

compilade committed 307 days ago

Verified 7acfd4e8

Refactor lora adapter support (#8332)

ngxson committed 307 days ago

Verified 97bdd26e

ngxson committed 307 days ago

Verified 4db8f60f

Older