ggml-org/llama.cpp

Pull Requests Commits

server : fix typo in model name (#4876)

isaac-mcfadyen committed 1 year ago

Verified 2f043328

metal : put encoder debug group behind a define (#4873)

ptsochantaris committed 1 year ago

Verified 2a7c94db

ggerganov committed 1 year ago

Verified 64802ec0

metal : fix deprecation warning (ggml/690)

ggerganov committed 1 year ago

Verified 3267c2ab

ggml : remove ggml_cpy_inplace and ggml_cont_inplace (ggml/693)

4imothy committed 1 year ago

Verified f85a973a

metal : wrap each operation in debug group (ggml/690)

jmousseau committed 1 year ago

Verified 5362e439

ggml : change GGML_MAX_NAME at compile time (ggml/682)

leejet committed 1 year ago

Verified e739de79

Fix execlp call (ggml/689)

Halalaluyafail3 committed 1 year ago

Verified c910e3c2

fix : cuda order of synchronization when setting a buffer (ggml/679)

Green-Sky committed 1 year ago

Verified f34432ca

server : update readme to document the new `/health` endpoint (#4866)

ibehnam committed 1 year ago

Verified 7a9f75c3

server : fix build + rename enums (#4870)

ggerganov committed 1 year ago

Verified 5c1980d8

server : add a `/health` endpoint (#4860)

ibehnam committed 1 year ago

Verified cd108e64

llama : add additional suffixes for model params (#4834)

mofosyne committed 1 year ago

Verified 57d016ba

llama : recognize 1B phi models (#4847)

teleprint-me committed 1 year ago

Verified 329ff615

clip : support more quantization types (#4846)

cmp-nct committed 1 year ago

Verified d34633d8

Python script to compare commits with llama-bench (#4844)

JohannesGaessler committed 1 year ago

Verified 4f56458d

convert.py : fix vanilla LLaMA model conversion (#4818)

teleprint-me committed 1 year ago

Verified 6efb8eb3

llava-cli : don't crash if --image flag is invalid (#4835)

jart committed 1 year ago

Verified 36e5a08b

metal : improve dequantize precision to match CPU (#4836)

ggerganov committed 1 year ago

Verified 4dccb38d

scripts : improve get-pg.sh (#4838)

ggerganov committed 1 year ago

Verified 9a818f7c

readme : add 3rd party collama reference to UI list (#4840)

iohub committed 1 year ago

Verified 18adb4e9

scripts : script to get Paul Graham essays in txt format (#4838)

ggerganov committed 1 year ago

Verified d9653894

server : update readme about token probs (#4777)

ibehnam committed 1 year ago

Verified 128de358

server : add api-key flag to documentation (#4832)

Zsapi committed 1 year ago

Verified 8c583303

ggml : fix vld1q_s8_x4 32-bit compat (#4828)

ggerganov committed 1 year ago

Verified 18c2e175

CUDA: faster softmax via shared memory + fp16 math (#4742)

JohannesGaessler committed 1 year ago

Verified 8f900abf

common : fix the short form of `--grp-attn-w`, not `-gat` (#4825)

howlger committed 1 year ago

Verified 1fc2f265

readme : add link to SOTA models

ggerganov committed 1 year ago

Verified a9a8c5de

SOTA 2-bit quants (#4773)

ikawrakow committed 1 year ago

Verified dd5ae064

swift : exclude ggml-metal.metal from the package (#4822)

ggerganov committed 1 year ago

Verified 668b31fc

Newer Older