ochafik/llama.cpp

Pull Requests Commits

grammars: comment + renaming to token_codepoints_without_partial_utf_prefix

Olivier Chafik committed 2 years ago

691c5405

grammars: faster llama_grammar_copy

Olivier Chafik committed 2 years ago

c9476e54

grammars: keep llama_grammar_copy non-quadratic optim for later

Olivier Chafik committed 2 years ago

05efa34d

grammars: fix bad merge

ochafik committed 2 years ago

24769f9a

Merge remote-tracking branch 'origin/master' into grammar-fast

ochafik committed 2 years ago

1fb300c3

llama : add option to render special/control tokens (#6807)

ggerganov committed 2 years ago

Verified 40f74e4d

grammars: faster llama_grammar_copy

ochafik committed 2 years ago

cbc75809

grammars: cache decoded tokens

ochafik committed 2 years ago

f608415d

grammars: early exit when no next_candidates to reject

ochafik committed 2 years ago

98f33bae

ggml : fix ggml_backend_cpu_supports_op() for CPY (#0)

ggerganov committed 2 years ago

Verified b9cc76d8

llama : add llama-3 chat template (#6751)

DifferentialityDevelopment committed 2 years ago

Verified 7dbdba56

gguf-py : add IQ1_M to GGML_QUANT_SIZES (#6761)

piotrek1996 committed 2 years ago

Verified c1386c93

doc : add link to falcon (#6789)

kaetemi committed 2 years ago

Verified e8d35f47

readme : add Fedora instructions (#6783)

Man2Dev committed 2 years ago

Verified 2cca09d5

llava : use logger in llava-cli (#6797)

jart committed 2 years ago

Verified 89b0bf0d

llama : support Llama 3 HF conversion (#6745)

pcuenca committed 2 years ago

Verified b97bc396

doc : server tests require llama to be built with curl enabled (#6788)

kaetemi committed 2 years ago

Verified b8109bc0

common : try to fix Android CI (#6780)

ggerganov committed 2 years ago

Verified aed82f68

ci: add ubuntu latest release and fix missing build number (mac & ubuntu) (#6748)

loonerin committed 2 years ago

Verified 0e4802b2

server: static: upstream upgrade (#6765)

phymbert committed 2 years ago

Verified 637e9a86

Implement the OLMo architecture (#6741)

nopperl committed 2 years ago

Verified 9958c81b

train : add general name (#6752)

teleprint-me committed 2 years ago

Verified 8b1b1f49

fix wrong parameter in cmd in readme-sycl.md (#6755)

arthw committed 2 years ago

Verified bca40e98

ggml : group all experts in a single ggml_mul_mat_id (#6505)

slaren committed 2 years ago

Verified 0d56246f

convert : support models with multiple chat templates (#6588)

CISC committed 2 years ago

Verified 03c0946d

Qwen2 : assume tied weights if lm_head/output weights is missing (#6738)

jklj077 committed 2 years ago

Verified e11b2e6e

llama : fix compatibility with old 2 expert models (#6735)

slaren committed 2 years ago

Verified c71bfd73

llamafile : tmp disable + build sgemm.o when needed (#6716)

ggerganov committed 2 years ago

Verified 3b8f1ec4

readme : add UI (#6724)

yaroslavyaroslav committed 2 years ago

Verified 8dd1ec8b

convert : fix autoawq gemma (#6704)

dengzheng-cloud committed 2 years ago

Verified facb8b56

Older