ochafik/llama.cpp

Pull Requests Commits

ochafik committed 1 year ago

c82ae1ac

Merge remote-tracking branch 'origin/master' into json-order

ochafik committed 1 year ago

945c1b6c

Clarify default MMQ for CUDA and LLAMA_CUDA_FORCE_MMQ flag (#8115)

isaac-mcfadyen committed 1 year ago

Verified 88540445

CUDA: fix misaligned shared memory read (#8123)

JohannesGaessler committed 1 year ago

Verified c8771ab5

llama : extend llm_build_ffn() to support _scale tensors (#8103)

Eddie-Wang1120 committed 1 year ago

Verified 494165f3

`json`: better support for "type" unions (e.g. nullable arrays w/ typed items) (#7863)

ochafik committed 1 year ago

Verified 9b2f16f8

`json`: fix additionalProperties, allow space after enum/const (#7840)

ochafik committed 1 year ago

Verified 6777c544

json: add property order integ. test

ochafik committed 1 year ago

cc5b4aae

json: remove some needless parenthesis from grammars

ochafik committed 1 year ago

48f2acfa

json: port unified properties order to JS & PY

ochafik committed 1 year ago

c6c7c19a

Merge branch 'json-additional' into json-order

ochafik committed 1 year ago

abb86272

Merge remote-tracking branch 'origin/master' into json-order

ochafik committed 1 year ago

f9a14002

update # tokens in server test: consts can now have trailing space

ochafik committed 1 year ago

23beed22

fixes #7999 (adds control vectors to all `build_XXX()` functions in `llama.cpp` [needs testing] (#8060)

jukofyork committed 1 year ago

Verified 163d50ad

Merge remote-tracking branch 'origin/master' into json-additional

ochafik committed 1 year ago

fba3a864

llama : implement Unigram tokenizer needed by T5 and FLAN-T5 model families (#5763)

fairydreaming committed 1 year ago

Verified 6fcbf682

llama : return nullptr from llama_grammar_init (#8093)

danbev committed 1 year ago

Verified e6bf0077

`json`: support integer minimum, maximum, exclusiveMinimum, exclusiveMaximum (#7797)

ochafik committed 1 year ago

Verified 84631fe1

disable docker CI on pull requests (#8110)

slaren committed 1 year ago

Verified dd047b47

Add healthchecks to llama-server containers (#8081)

codearranger committed 1 year ago

Verified 925c3095

Gguf dump start data offset via --data-offset and some extra refactor (#8054)

mofosyne committed 1 year ago

Verified c8ad3595

cvector: better prompt handling, add "mean vector" method (#8069)

ngxson committed 1 year ago

Verified 49c03c79

Add chat template support for llama-cli (#8068)

ngxson committed 1 year ago

Verified 48e6b92c

SimpleChat v3.1: Boolean chat request options in Settings UI, cache_prompt (#7950)

hanishkvc committed 1 year ago

Verified 3791ad21

Update control vector help (#8104)

HatsuneMikuUwU33 committed 1 year ago

Verified f702a90e

[SYCL] Re-enabled mul_mat_batched_sycl (#8095)

airMeng committed 1 year ago

Verified 083bacce

Merge remote-tracking branch 'origin/master' into json-additional

ochafik committed 1 year ago

c6df6cee

json: streamline order

ochafik committed 1 year ago

e10f14a0

CUDA: fix matrix multiplication algorithm choice (#8102)

JohannesGaessler committed 1 year ago

Verified 2df373ac

Merge remote-tracking branch 'origin/master' into json-additional

ochafik committed 1 year ago

35bbac1e

Older