ggerganov/llama.cpp

Pull Requests Commits

batched-bench : add mmq CLI arg

ggerganov committed 1 year ago

Verified 2fcdf869

batched-bench : pass custom set of PP, TG and PL

ggerganov committed 1 year ago

Verified daeb834d

batched-bench : init warm-up batch

ggerganov committed 1 year ago

Verified c062ffd1

Merge branch 'master' into batched-bench

ggerganov committed 1 year ago

Verified 76e17f8d

batched-bench : add readme + n_kv_max is now configurable

ggerganov committed 1 year ago

Verified 026bb1b1

examples : add batched.swift + improve CI for swift (#3562)

zshannon committed 1 year ago

Verified 24ba3d82

Add MPT model to supported models in README.md (#3574)

Galunid committed 1 year ago

Verified 9f6ede19

Minor improvements in GPT2 tokenizer (#3567)

goerch committed 1 year ago

Verified 233fc1c6

readme : add bloom (#3570)

xingchensong committed 1 year ago

Verified c5b49360

llm : add bloom models (#3553)

xingchensong committed 1 year ago

Verified 02d2875d

swift : improvements and fixes (#3564)

jhen0409 committed 1 year ago

Verified 0aa6595a

llm : add MPT support (#3417)

jploski committed 1 year ago

Verified f5f9121d

infill. : fix tokenization (#3508)

vvhg1 committed 1 year ago

Verified 11ea5c7d

ggml-alloc : fix assert in debug builds (#3555)

slaren committed 1 year ago

Verified 95bd60a0

refact : fix convert script + zero out KV cache to avoid nans (#3523)

ggerganov committed 1 year ago

Verified fcca0a70

metal : do not use mul_mm kernels when ne00 < 64 (#3542)

ggerganov committed 1 year ago

Verified dcc09d25

sync : ggml (ggml-backend) (#3548)

ggerganov committed 1 year ago

Verified db3abcc1

ci : add Zig CI/CD and fix build (#2996)

kassane committed 1 year ago

Verified eee42c67

batched : minor fix table

ggerganov committed 1 year ago

Verified 7438728d

batched : add bench tool

ggerganov committed 1 year ago

Verified bf06d654

api_like_OAI.py : compat with Microsoft Guidance (#2746)

ryderwishart committed 1 year ago

Verified 8e6716a1

api_like_OAI.py : simplify function (#2796)

arcrank committed 1 year ago

Verified 9c38d181

k-quants : fix comments about block sizing (#3499)

jrudolph committed 1 year ago

Verified a1202a31

ci : enable on obj-c changes + fix metal build (#3540)

ggerganov committed 1 year ago

Verified 94e502df

zig : fix build by introducing train.cpp (#3539)

robertluo committed 1 year ago

Verified 7d8b2493

metal : support MTLGPUFamily < Apple7, formatting, style (#3524)

ggerganov committed 1 year ago

Verified b0ec5218

llama : fix missing break in Persimmon arch case statements (#3535)

KerfuffleV2 committed 1 year ago

Verified 63d3b06a

Fix trying to strip newline from empty prompt and cfg prompt file content (#3534)

KerfuffleV2 committed 1 year ago

Verified a16e89ce

gguf.py : fix CI for publishing GGUF package (#3532)

monatis committed 1 year ago

Verified 4d038332

py : change version of numpy requirement to 1.24.4 (#3515)

lyjia committed 1 year ago

Verified c47066d8

Older