ggml-org/llama.cpp

Pull Requests Commits

common : add -dkvc arg for enabling kv cache dumps

ggerganov committed 2 years ago

Verified f8e9f114

llama : allow exporting a view of the KV cache (#4180)

KerfuffleV2 committed 2 years ago

Verified 5df7d06c

llama : zero KV cache used upon clear

ggerganov committed 2 years ago

Verified 671f639c

llama : keep track of used KV cells + better KV cache management

ggerganov committed 2 years ago

Verified 79cb8f00

stablelm : simplify + speedup generation (#4153)

Galunid committed 2 years ago

Verified 8e672efe

finetune - update readme to mention llama support only (#4148)

Galunid committed 2 years ago

Verified 0b871f1a

readme : update ROCm Windows instructions (#4122)

jammm committed 2 years ago

Verified dfc7cd48

main : Add ChatML functionality to main example (#4046)

Sebby37 committed 2 years ago

Verified 881800d1

ci : add flake8 to github actions (python linting) (#4129)

Galunid committed 2 years ago

Verified f23c0359

speculative : fix prompt tokenization in speculative example (#4025)

AutonomicPerfectionist committed 2 years ago

Verified 40a34fe8

Revert "finetune : add --n-gpu-layers flag info to --help (#4128)"

ggerganov committed 2 years ago

Verified dae06c06

finetune : add --n-gpu-layers flag info to --help (#4128)

csaben committed 2 years ago

Verified 05e8301e

server : relay error messages (#4131)

SoftwareRenderer committed 2 years ago

Verified 936c79b2

common : comma should be semicolon (#4137)

kchro3 committed 2 years ago

Verified 262005ad

gitignore : tokenize

ggerganov committed 2 years ago

Verified 35985acf

gguf-py : export chat templates (#4125)

slaren committed 2 years ago

Verified e9370664

tokenize example: Respect normal add BOS token behavior (#4126)

KerfuffleV2 committed 2 years ago

Verified 28a2e6e7

scripts : Remove missed baichuan convert script (#4127)

Galunid committed 2 years ago

Verified 0b5c3b04

Clean up ggml-cuda.cu warnings when compiling with clang (for ROCM) (#4124)

KerfuffleV2 committed 2 years ago

Verified 2923f17f

llama : increase max nodes (#4115)

slaren committed 2 years ago

Verified bbecf3f4

build : support ppc64le build for make and CMake (#3963)

bufferoverflow committed 2 years ago

Verified 8e936108

tokenize : fix trailing whitespace

ggerganov committed 2 years ago

Verified 5ad387e9

examples : add tokenize (#4039)

zakkor committed 2 years ago

Verified 2fa02b4b

convert : use 'model' value if it exists. This allows karpathy/tinyllamas to load (#4089)

dmahurin committed 2 years ago

Verified 2ab0707a

py : Falcon HF compatibility (#4104)

cmp-nct committed 2 years ago

Verified 11173c92

common : improve yaml log escaping (#4080)

joennlae committed 2 years ago

Verified 9e87ef60

llava : fix compilation warning that fread return value is not used (#4069)

huawei-lin committed 2 years ago

Verified c7cce124

py : remove superfluous import statements (#4076)

jpodivin committed 2 years ago

Verified f7d5e975

train : move number of gpu layers argument parsing to common/train.cpp (#4074)

jpodivin committed 2 years ago

Verified ba4cf5c0

llama : add functions to get the model's metadata (#4013)

slaren committed 2 years ago

Verified e85bb1a8

Older