The Wayback Machine - http://web.archive.org/web/20201025092823/https://github.com/SkafteNicki

SkafteNicki Follow

SkafteNicki Follow

Nicki Skafte SkafteNicki

Ph.d student at section for Cognitive Systems (CogSys), Technical University of Denmark (DTU). Main focus: Deep Learning and Metric Learning.

Follow

34 followers · 3 following · 25

Denmark
Sign in to view email

Highlights

Arctic Code Vault Contributor

Pinned

ddtn

Repository for our upcoming code, that we used for our "Deep diffeomorphic transformer networks" paper (Accepted to CVPR 2018). Will be update during the spring of 2018.

Python 39 8
Deep_LMNN

LMNN algorithm in tensorflow

Python 2
libcpab

CPAB Transformations: finite-dimensional spaces of simple, fast, and highly-expressive diffeomorphisms derived from parametric, continuously-defined, velocity fields in Numpy, Tensorflow and Pytorch

Python 22 5
pyclust

My own personal scripts for easy working with the cluster at my work

Python 1
unsuper

Some unsupervised algorithms

Python 4 2
py_uci

Python library for loading data from the UCI Machine Learning Repository

Python 4 2

400 contributions in the last year

Contribution activity

October 2020

PyTorchLightning/pytorch-lightning 16 commits

Created a pull request in PyTorchLightning/pytorch-lightning that received 11 comments

[Metrics] Unification of regression

What does this PR do? Attempt at unification between the new class based regression metrics and their functional counterpart. In short, each funct…

+1,039 −623 • 11 comments

[Metrics] MetricCollection Oct 23
[Metrics] Detach bugfix Oct 22
New section for changelog [ci skip] Oct 21
Speedup of metric tests Oct 13
Update explained variance metric Oct 9
Remove limitation of batch scaler Oct 8
[skip ci] metric docs Oct 8
Skip tuner algorithms on fast dev Oct 6
Fix lr finder for optimizers with states Oct 6
Batch scaler docs Oct 6
[metrics] Accuracy error fix Oct 1

fix typos in docs: remove redundant whitespace in new-project.rst Oct 25
update old references to filepath in ModelCheckpoint docs Oct 25
add bug_report_model to bug_report Oct 22
add test if no loss is provided Oct 22
Update __init__.py Oct 22
[Metrics] Fix/4237 auc unstable reorder Oct 22
Set correct device ids in DDP [wip] Oct 22
Add dirpath and filename parameter in ModelCheckpoint Oct 21
Clean up optimizer code Oct 21
Fix info message when EarlyStopping 'mode' not provided [ci skip] Oct 21
add svg logo [SKIP CI] Oct 21
Skip replacing dataloader sampler if it's already a distributed sampler Oct 21
Corrected f_beta computation Oct 21
update examples Oct 21
[Metrics, fix] Fix unintended zero-only target in iou example Oct 21
update stale bot Oct 21
CI: add flake8 Oct 19
allow codecov upload to fail Oct 18
move base req. to root Oct 18
update PR template Oct 17
remove unused folder Oct 17
limit monitor callback with log_every_n_steps Oct 17
Update multi_gpu.rst Oct 17
Add load_from_checkpoint function to Loading docs Oct 16
Annotate return type of TrainerProperties.from_argparse_args(...) Oct 16
Some pull request reviews not shown.

You can’t perform that action at this time.