ggml-org/llama.cpp

Pull Requests Commits

ci : close inactive issue, increase operations per run (#6270)

phymbert committed 1 year ago

Verified ea279d56

sampling : deduplicated code for probability distribution access (#6240)

mscheong01 committed 1 year ago

Verified 586e7bc5

[SYCL] offload op (#6217)

airMeng committed 1 year ago

Verified ddf65685

Support build win release for SYCL (#6241)

NeoZhangJianyu committed 1 year ago

Verified d03224ac

use _wfopen instead of fopen on Windows (#6248)

cebtenzzre committed 1 year ago

Verified 94d1b3b4

gitignore : gguf-split

ggerganov committed 1 year ago

Verified 95562175

common: llama_load_model_from_url split support (#6192)

phymbert committed 1 year ago

Verified f482bb2e

server: docs: `--threads` and `--threads`, `--ubatch-size`, `--log-disable` (#6254)

phymbert committed 1 year ago

Verified 1997577d

llama : add grok-1 support (#6204)

arki05 committed 1 year ago

Verified 476b0251

split: add gguf-split in the make build target (#6262)

phymbert committed 1 year ago

Verified 21cad01b

server: flush stdout after logging in both text and json layout (#6253)

phymbert committed 1 year ago

Verified 1b26aebe

lookup: complement data from context with general text statistics (#5479)

JohannesGaessler committed 1 year ago

Verified 50ccaf5e

common : default --hf-file to --model (#6234)

ggerganov committed 1 year ago

Verified 56a00f0a

convert-llama2c-to-ggml : enable conversion of GQA models (#6237)

fraxy-v committed 1 year ago

Verified 92397d87

quantize: options for output and token embedding tensors qtype (#6239)

ikawrakow committed 1 year ago

Verified 1d0331c1

llama_model_loader: support multiple split/shard GGUFs (#6187)

phymbert committed 1 year ago

Verified dba1af61

ci: apply concurrency limit for github workflows (#6243)

mscheong01 committed 1 year ago

Verified ee804f62

ggml : fix CPU soft_max

ggerganov committed 1 year ago

Verified 09532120

llama : fix type of KQ_mask and KQ_pos

ggerganov committed 1 year ago

Verified 3a468e6f

Merge branch 'master' into gg/flash-attn

ggerganov committed 1 year ago

Verified 9495d398

common : add HF arg helpers (#6234)

ggerganov committed 1 year ago

Verified 80bd33bc

llama : correction of the attn.v.weight quantization for IQ3_XS (#6209)

Nexesenex committed 1 year ago

Verified e80f06d2

tests : conditional python & node json schema tests (#6207)

ochafik committed 1 year ago

Verified f77a8ffd

json-schema-to-grammar : fix order of props + non-str const/enum (#6232)

ochafik committed 1 year ago

Verified 72114edf

cuda : add LLAMA_CUDA_NO_PEER_COPY to workaround broken ROCm p2p copy (#6208)

slaren committed 1 year ago

Verified 2f0e81e0

readme : add RecurseChat to the list of UIs (#6219)

xyc committed 1 year ago

Verified 29ab270e

server : fix n_keep always showing as 0 in response (#6211)

kaetemi committed 1 year ago

Verified 6b8bb3a3

server : enable continuous batching by default (#6231)

ggerganov committed 1 year ago

Verified 68e210b3

metal : proper assert for mat-mat memory alignment (#6225)

ggerganov committed 1 year ago

Verified b3e94f26

ci : add CURL flag for the mac builds (#6214)

Vaibhavs10 committed 1 year ago

Verified b2075fd6

Newer Older