ggml-org/llama.cpp

Pull Requests Commits

llama : fix BPE LF token on MSVC

cebtenzzre committed 2 years ago

d05c13b3

use wide versions of file path functions on Windows

cebtenzzre committed 2 years ago

6f4fd8f1

ggml : fix missing #defines before windows.h

cebtenzzre committed 2 years ago

14eebe23

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

ggerganov committed 2 years ago

Verified 56a00f0a

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

fraxy-v committed 2 years ago

Verified 92397d87

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

ikawrakow committed 2 years ago

Verified 1d0331c1

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

phymbert committed 2 years ago

Verified dba1af61

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

mscheong01 committed 2 years ago

Verified ee804f62

common : add HF arg helpers (#6234)

ggerganov committed 2 years ago

Verified 80bd33bc

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

Nexesenex committed 2 years ago

Verified e80f06d2

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

ochafik committed 2 years ago

Verified f77a8ffd

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

ochafik committed 2 years ago

Verified 72114edf

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

slaren committed 2 years ago

Verified 2f0e81e0

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

xyc committed 2 years ago

Verified 29ab270e

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

kaetemi committed 2 years ago

Verified 6b8bb3a3

server : enable continuous batching by default (#6231)

ggerganov committed 2 years ago

Verified 68e210b3

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

ggerganov committed 2 years ago

Verified b3e94f26

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

Vaibhavs10 committed 2 years ago

Verified b2075fd6

metal : pad n_ctx by 32 (#6177)

ggerganov committed 2 years ago

Verified 95d576b4

add blog link (#6222)

NeoZhangJianyu committed 2 years ago

Verified 59c17f02

Fix params underscore convert to dash. (#6203)

dranger003 committed 2 years ago

Verified fa046eaf

server : update readme doc from `slot_id` to `id_slot` (#6213)

kaetemi committed 2 years ago

Verified be07a032

cuda : disable host register by default (#6206)

slaren committed 2 years ago

Verified d0a71233

Corrected typo to wrong file (#6199)

semidark committed 2 years ago

Verified f372c49c

tests : disable system() calls (#6198)

ggerganov committed 2 years ago

Verified 924ce1dc

cuda : fix LLAMA_CUDA_F16 build (#6197)

slaren committed 2 years ago

Verified 03a8f8fa

ggml : same IQ4_NL quantization for CPU/CUDA/Metal (#6196)

ikawrakow committed 2 years ago

Verified cfd3be76

json-schema-to-grammar improvements (+ added to server) (#5978)

ochafik committed 2 years ago

Verified 5b7b0ac8

ci : fix indentation error (#6195)

Vaibhavs10 committed 2 years ago

Verified 1943c019

build : add mac pre-build binaries (#6182)

Vaibhavs10 committed 2 years ago

Verified 5e43ba87

Older