ggerganov/llama.cpp

Pull Requests Commits

llama : clean-up

ggerganov committed 1 year ago

Verified fe28a7b9

llama : suffix the internal APIs with "_impl"

ggerganov committed 1 year ago

Verified dae3cae8

llama : redirect external API to internal APIs

ggerganov committed 1 year ago

Verified 39fbaf9f

make : update llama.cpp deps [no ci]

ggerganov committed 1 year ago

Verified 66ac80f5

llama : move tokenizers into llama-vocab

ggerganov committed 1 year ago

Verified 8fef5b18

llama : deprecate llama_sample_grammar

ggerganov committed 1 year ago

Verified e7dffa6b

ggerganov committed 1 year ago

Verified 689d3779

cont : pre-fetch rules

ggerganov committed 1 year ago

Verified b4b242e6

ggerganov committed 1 year ago

Verified 5a71d1ae

llama : move grammar code into llama-grammar

ggerganov committed 1 year ago

Verified 675f305f

llama : move sampling code into llama-sampling

ggerganov committed 1 year ago

Verified 0ddc8e36

llama : fix codeshell support (#8599)

hankeke303 committed 1 year ago

Verified 081fe431

llama : add support for SmolLm pre-tokenizer (#8609)

Stillerman committed 1 year ago

Verified d94c6e0c

*.py: Stylistic adjustments for python (#8233)

jpodivin committed 1 year ago

Verified 566daa5a

llama : allow overrides for tokenizer flags (#8614)

ggerganov committed 1 year ago

Verified 6f11a83e

tests : re-enable tokenizer tests (#8611)

ggerganov committed 1 year ago

Verified e093dd23

llama : add Mistral Nemo inference support (#8604)

iamlemec committed 1 year ago

Verified 50e05353

server : update doc to clarify n_keep when there is bos token (#8619)

kaetemi committed 1 year ago

Verified 62815449

ggml: fix compile error for RISC-V (#8623)

zqb-all committed 1 year ago

Verified 04bab6b7

examples: fix android example cannot be generated continuously (#8621)

devojony committed 1 year ago

Verified b7c11d36

flake.lock: Update (#8610)

ggerganov committed 1 year ago

Verified 45f2c19c

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

Older