ggml-org/llama.cpp

Pull Requests Commits

llama : allow overrides for tokenizer flags

ggerganov committed 1 year ago

Verified 57349e1d

examples : Rewrite pydantic_models_to_grammar_examples.py (#8493)

maruel committed 1 year ago

Verified 22f281aa

gguf-py : fix some metadata name extraction edge cases (#8591)

compilade committed 1 year ago

Verified 328884f4

convert_hf : fix Gemma v1 conversion (#8597)

compilade committed 1 year ago

Verified c69c6303

CUDA: MMQ code deduplication + iquant support (#8495)

JohannesGaessler committed 1 year ago

Verified 69c487f4

gguf : handle null name during init (#8587)

ggerganov committed 1 year ago

Verified 07283b1a

llama : add support for Tekken pre-tokenizer (#8579)

m18coppola committed 1 year ago

Verified 94036222

llama.swiftui: fix end of generation bug (#8268)

ho2103 committed 1 year ago

Verified 69b9945b

gguf_dump.py: fix markddown kv array print (#8588)

mofosyne committed 1 year ago

Verified c3776cac

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

slaren committed 1 year ago

Verified 87e397d0

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

mofosyne committed 1 year ago

Verified 57b1d4f9

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

ggerganov committed 1 year ago

Verified d1975455

readme : fix server badge

ggerganov committed 1 year ago

Verified be0cfb41

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

HanClinto committed 1 year ago

Verified b57eb9ca

fix: typo of chatglm4 chat tmpl (#8586)

thxCode committed 1 year ago

Verified f299aa98

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

mofosyne committed 1 year ago

Verified 3d0e4367

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

JohannesGaessler committed 1 year ago

Verified a15ef8f8

cmake : install all ggml public headers (#8480)

65a committed 1 year ago

Verified 705b7ecf

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

EZForever committed 1 year ago

Verified 0d2c7321

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

mofosyne committed 1 year ago

Verified 672a6f10

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

RunningLeon committed 1 year ago

Verified 3807c3de

lookup: fibonacci hashing, fix crashes (#8548)

JohannesGaessler committed 1 year ago

Verified e02b597b

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

amochkin committed 1 year ago

Verified b3283448

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

mofosyne committed 1 year ago

Verified 30f80ca0

[CANN] Add Ascend NPU backend (#6035)

hipudding committed 1 year ago

Verified 1bdd8ae1

batched: fix n_predict parameter (#8527)

msy-kato committed 1 year ago

Verified da3913d8

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

ggerganov committed 1 year ago

Verified d65a8361

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

JohannesGaessler committed 1 year ago

Verified 5e116e8d

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

mofosyne committed 1 year ago

Verified 1666f92d

export-lora : handle help argument (#8497)

sbonds committed 1 year ago

Verified 37b12f92

Older