Data

Tools

Indexes

Applications

Experiments

Open Source Science

Recent Releases of mozuma

mozuma - 0.9.0 - Python 3.8 & PyTorch 1.13

What's Changed

  • Add links to notebooks in documentation by @smassonnet in https://github.com/mozuma/mozuma/pull/201
  • Review the documentation by @smassonnet in https://github.com/mozuma/mozuma/pull/202
  • Add Citation Section in README.me by @rlebret in https://github.com/mozuma/mozuma/pull/203
  • Migrate to PyTorch 1.13 by @smassonnet in https://github.com/mozuma/mozuma/pull/206

Full Changelog: https://github.com/mozuma/mozuma/compare/v0.8.0...0.9.0

- Python
Published by smassonnet over 2 years ago

mozuma - 0.8.0 - DistilUse tokenizer truncation option

What's Changed

  • Add truncation on DistilBert tokenizer by @smassonnet in https://github.com/mozuma/mozuma/pull/198

Full Changelog: https://github.com/mozuma/mozuma/compare/v0.7.2...v0.8.0

- Python
Published by smassonnet over 3 years ago

mozuma - 0.7.2 - Fix PyPi release process

What's Changed

  • publish action not setting the correct build version by @mromanelli9 in https://github.com/mozuma/mozuma/pull/196

Full Changelog: https://github.com/mozuma/mozuma/compare/v0.7.1...v0.7.2

- Python
Published by smassonnet over 3 years ago

mozuma - 0.7.1 - MoZuMa on PyPi

What's Changed

  • Documentation: Add a simple notebook example for demos by @smassonnet in https://github.com/mozuma/mozuma/pull/194
  • Publish python package to PyPI by @mromanelli9 in https://github.com/mozuma/mozuma/pull/184

Full Changelog: https://github.com/mozuma/mozuma/compare/v0.7.0...v0.7.1

- Python
Published by smassonnet over 3 years ago

mozuma - 0.7.0 - MoZuMa

What's Changed

  • Add issue templates by @mromanelli9 in https://github.com/mozuma/mozuma/pull/156
  • Add tensorboard logger in TorchTrainingRunner by @mromanelli9 in https://github.com/mozuma/mozuma/pull/161
  • Rename MLModule to MoZuMa by @smassonnet in https://github.com/mozuma/mozuma/pull/187
  • Rename MLModuleKit to mozuma by @smassonnet in https://github.com/mozuma/mozuma/pull/189

Full Changelog: https://github.com/mozuma/mozuma/compare/v0.6.0...v0.7.0

- Python
Published by smassonnet over 3 years ago

mozuma - 0.6.0 - New MLModule API

This is a new version of MLModule that's introducing a new API.

This version is not compatible with the previous versions of MLModule.

What's Changed

  • Issues/49-Distiluse-V2-text-encoder by @smassonnet in https://github.com/LSIR/mlmodule/pull/50
  • Code quality workflows by @smassonnet in https://github.com/LSIR/mlmodule/pull/62
  • Issues/57-Faster-image-preprocessing by @mromanelli9 in https://github.com/LSIR/mlmodule/pull/63
  • Improve modularity of models definition by @smassonnet in https://github.com/LSIR/mlmodule/pull/64
  • Migrate MagFace by @smassonnet in https://github.com/LSIR/mlmodule/pull/72
  • Migrate VinVL by @smassonnet in https://github.com/LSIR/mlmodule/pull/73
  • Callback for the end of a runner by @smassonnet in https://github.com/LSIR/mlmodule/pull/77
  • Keyframe get / set state does not return the right keys by @smassonnet in https://github.com/LSIR/mlmodule/pull/80
  • Improve dataset modularity by @smassonnet in https://github.com/LSIR/mlmodule/pull/82
  • Remove logic to filter out features with NaN by @smassonnet in https://github.com/LSIR/mlmodule/pull/83
  • Set binary file dataset output to typing.BinaryIO instead of bytes by @smassonnet in https://github.com/LSIR/mlmodule/pull/85
  • BUG VinVL prediction score contains features by @smassonnet in https://github.com/LSIR/mlmodule/pull/87
  • Fixed validate options that modifies in-place by @smassonnet in https://github.com/LSIR/mlmodule/pull/90
  • Split forward predictions into forward and to predictions by @smassonnet in https://github.com/LSIR/mlmodule/pull/92
  • Removes inconsistencies in MTCNN device by @smassonnet in https://github.com/LSIR/mlmodule/pull/94
  • Video encoder dimension error on short videos by @smassonnet in https://github.com/LSIR/mlmodule/pull/97
  • ImageDataset should have a mode argument by @smassonnet in https://github.com/LSIR/mlmodule/pull/99
  • Multi-GPU inference by @mromanelli9 in https://github.com/LSIR/mlmodule/pull/100
  • Migrate DenseNet to new MLModule API by @smassonnet in https://github.com/LSIR/mlmodule/pull/106
  • Migrate sentence BERT by @smassonnet in https://github.com/LSIR/mlmodule/pull/108
  • Make a public model weights store by @smassonnet in https://github.com/LSIR/mlmodule/pull/111
  • Add exists function to stores by @smassonnet in https://github.com/LSIR/mlmodule/pull/114
  • Script to upload the model states to MLModule by @smassonnet in https://github.com/LSIR/mlmodule/pull/115
  • Migrate CLI by @smassonnet in https://github.com/LSIR/mlmodule/pull/121
  • Create LICENSE by @smassonnet in https://github.com/LSIR/mlmodule/pull/120
  • Migrate ArcFace by @smassonnet in https://github.com/LSIR/mlmodule/pull/123
  • Trainable ResNet by @mromanelli9 in https://github.com/LSIR/mlmodule/pull/105
  • Add classification module by @smassonnet in https://github.com/LSIR/mlmodule/pull/109
  • Remove old tests by @smassonnet in https://github.com/LSIR/mlmodule/pull/133
  • BUG ResNet and TorchResNetTrainingMode by @mromanelli9 in https://github.com/LSIR/mlmodule/pull/132
  • Installation instructions by @smassonnet in https://github.com/LSIR/mlmodule/pull/135
  • Remove unused code by @smassonnet in https://github.com/LSIR/mlmodule/pull/139
  • Rename contrib to models by @smassonnet in https://github.com/LSIR/mlmodule/pull/140
  • Moved callbacks module by @smassonnet in https://github.com/LSIR/mlmodule/pull/141
  • Completed models submodule by @smassonnet in https://github.com/LSIR/mlmodule/pull/142
  • Moved runners by @smassonnet in https://github.com/LSIR/mlmodule/pull/143
  • Moved predictions by @smassonnet in https://github.com/LSIR/mlmodule/pull/144
  • Cite libraries and papers used by @smassonnet in https://github.com/LSIR/mlmodule/pull/138
  • Moved states and stores by @smassonnet in https://github.com/LSIR/mlmodule/pull/145
  • Moved torch by @smassonnet in https://github.com/LSIR/mlmodule/pull/146
  • Finished re-organising modules by @smassonnet in https://github.com/LSIR/mlmodule/pull/147
  • New MLModule API by @smassonnet in https://github.com/LSIR/mlmodule/pull/148
  • BUG: Manifest contains a path to old module contrib by @smassonnet in https://github.com/LSIR/mlmodule/pull/151
  • 151 example cifar10 by @rlebret in https://github.com/LSIR/mlmodule/pull/155
  • Add model list to documentation by @smassonnet in https://github.com/LSIR/mlmodule/pull/152
  • Migrate MLModule store to GitHUB release by @smassonnet in https://github.com/LSIR/mlmodule/pull/160
  • Issues/163-Review-old-notebooks by @rlebret in https://github.com/LSIR/mlmodule/pull/164
  • Issues/157-Example-Single-GPU-inference-with-CLIP by @rlebret in https://github.com/LSIR/mlmodule/pull/165
  • Add examples to documentation by @smassonnet in https://github.com/LSIR/mlmodule/pull/167
  • Review documentation by @smassonnet in https://github.com/LSIR/mlmodule/pull/168
  • Example: Emotion face recognition by @rlebret in https://github.com/LSIR/mlmodule/pull/169
  • Example for multilingual text model by @rlebret in https://github.com/LSIR/mlmodule/pull/171
  • CIFAR DenseNet 2GPUs training by @rlebret in https://github.com/LSIR/mlmodule/pull/175
  • Example: Train Age Detection Model based on ArcFace features by @rlebret in https://github.com/LSIR/mlmodule/pull/177
  • Black formatting of example notebooks by @smassonnet in https://github.com/LSIR/mlmodule/pull/179

New Contributors

  • @mromanelli9 made their first contribution in https://github.com/LSIR/mlmodule/pull/63

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.5.1...v0.6.0

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.arcface

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: arcface

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.vinvl-vg-x152c4

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: vinvl-vg-x152c4

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.sbert-distiluse-base-multilingual

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: sbert-distiluse-base-multilingual

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.facenet-mtcnn

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: facenet-mtcnn

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.magface

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: magface

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.densenet201

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: densenet201

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.densenet169

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: densenet169

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.densenet161

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: densenet161

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.densenet121

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: densenet121

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.clip-text-vitb32

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: clip-text-vitb32

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.clip-image-vitb32

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: clip-image-vitb32

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.clip-text-rn50x4

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: clip-text-rn50x4

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.clip-image-rn50x4

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: clip-image-rn50x4

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.clip-text-rn101

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: clip-text-rn101

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.clip-image-rn101

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: clip-image-rn101

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.clip-text-rn50

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: clip-text-rn50

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.clip-image-rn50

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: clip-image-rn50

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.wide-resnet101-2

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: wide-resnet101-2

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.wide-resnet50-2

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: wide-resnet50-2

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.resnext101-32x8d

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: resnext101-32x8d

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.resnext50-32x4d

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: resnext50-32x4d

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.resnet152

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: resnet152

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.resnet101

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: resnet101

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.resnet50

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: resnet50

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.resnet34

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: resnet34

- Python
Published by smassonnet almost 4 years ago

mozuma - state.pytorch.resnet18

This release contains model state files for model with the following characteristics:

  • Backend: pytorch
  • Architecture: resnet18

- Python
Published by smassonnet almost 4 years ago

mozuma - 0.6.0a8 - New MlModule API

What's Changed

  • Video encoder dimension error on short videos by @smassonnet in https://github.com/LSIR/mlmodule/pull/97
  • ImageDataset should have a mode argument by @smassonnet in https://github.com/LSIR/mlmodule/pull/99

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.6.0a7...v0.6.0a8

- Python
Published by smassonnet almost 4 years ago

mozuma - 0.6.0a7 - New MlModule API

What's Changed

  • Removes inconsistencies in MTCNN device by @smassonnet in https://github.com/LSIR/mlmodule/pull/94

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.6.0a6...v0.6.0a7

- Python
Published by smassonnet almost 4 years ago

mozuma - 0.6.0a6 - New MlModule API

What's Changed

  • Split forward predictions into forward and to predictions by @smassonnet in https://github.com/LSIR/mlmodule/pull/92

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.6.0a5...v0.6.0a6

- Python
Published by smassonnet almost 4 years ago

mozuma - 0.6.0a5 - New MlModule API

What's Changed

  • Fixed validate options that modifies in-place by @smassonnet in https://github.com/LSIR/mlmodule/pull/90

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.6.0a4...v0.6.0a5

- Python
Published by smassonnet almost 4 years ago

mozuma - 0.6.0a4 - New MlModule API

What's Changed

  • BUG VinVL prediction score contains features by @smassonnet in https://github.com/LSIR/mlmodule/pull/87

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.6.0a3...v0.6.0a4

- Python
Published by smassonnet almost 4 years ago

mozuma - 0.6.0a3 - New MlModule API

What's Changed

  • Set binary file dataset output to typing.BinaryIO instead of bytes by @smassonnet in https://github.com/LSIR/mlmodule/pull/85

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.6.0a2...v0.6.0a3

- Python
Published by smassonnet almost 4 years ago

mozuma - 0.6.0a2 - New MlModule API

What's Changed

  • Keyframe get / set state does not return the right keys by @smassonnet in https://github.com/LSIR/mlmodule/pull/80
  • Improve dataset modularity by @smassonnet in https://github.com/LSIR/mlmodule/pull/82
  • Remove logic to filter out features with NaN by @smassonnet in https://github.com/LSIR/mlmodule/pull/83

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.6.0a1...v0.6.0a2

- Python
Published by smassonnet almost 4 years ago

mozuma - 0.6.0a1 - New MlModule API

What's Changed

  • Issues/49-Distiluse-V2-text-encoder by @smassonnet in https://github.com/LSIR/mlmodule/pull/50
  • Code quality workflows by @smassonnet in https://github.com/LSIR/mlmodule/pull/62
  • Issues/57-Faster-image-preprocessing by @mromanelli9 in https://github.com/LSIR/mlmodule/pull/63
  • Improve modularity of models definition by @smassonnet in https://github.com/LSIR/mlmodule/pull/64
  • Migrate MagFace by @smassonnet in https://github.com/LSIR/mlmodule/pull/72
  • Migrate VinVL by @smassonnet in https://github.com/LSIR/mlmodule/pull/73
  • Callback for the end of a runner by @smassonnet in https://github.com/LSIR/mlmodule/pull/77

New Contributors

  • @mromanelli9 made their first contribution in https://github.com/LSIR/mlmodule/pull/63

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.5.1...v0.6.0a1

- Python
Published by smassonnet almost 4 years ago

mozuma - 0.5.1 MTCNN small images fix

What's Changed

  • MTCNN inference error on extra small images by @smassonnet in https://github.com/LSIR/mlmodule/pull/54

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.5.0...v0.5.1

- Python
Published by smassonnet about 4 years ago

mozuma - 0.5.0 - MTCNN no batch version

What's Changed

  • Code quality and pre commits by @smassonnet in https://github.com/LSIR/mlmodule/pull/47
  • Issues/51-MTCNN---Batch-image-distortion by @smassonnet in https://github.com/LSIR/mlmodule/pull/52

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.4.1...v0.5.0

- Python
Published by smassonnet about 4 years ago

mozuma - 0.4.1 - Keyframes bugfixes

What's Changed

  • CLI for video inference and bufixes by @smassonnet in https://github.com/LSIR/mlmodule/pull/44

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.4.0...v0.4.1

- Python
Published by smassonnet about 4 years ago

mozuma - 0.4.0 - Video key-frames extraction

What's Changed

  • Video Keyframes by @smassonnet in https://github.com/LSIR/mlmodule/pull/40

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.3.2...v0.4.0

- Python
Published by smassonnet over 4 years ago

mozuma - 0.3.2 - VinVL resizable input fix

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.3.1...v0.3.2

- Python
Published by smassonnet over 4 years ago

mozuma - 0.3.1 - Added back support for Python 3.7

What's Changed

  • Support for Python 3.7 by @smassonnet in https://github.com/LSIR/mlmodule/pull/39

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.3.0...v0.3.1

- Python
Published by smassonnet over 4 years ago

mozuma - 0.3.0 - VinVL object detection

What's Changed

  • Improve MLModule readability by @smassonnet in https://github.com/LSIR/mlmodule/pull/35
  • Vinvl by @rlebret in https://github.com/LSIR/mlmodule/pull/36
  • Remove MMDet dependency by @smassonnet in https://github.com/LSIR/mlmodule/pull/38

Full Changelog: https://github.com/LSIR/mlmodule/compare/v0.2.0...v0.3.0

- Python
Published by smassonnet over 4 years ago

mozuma - 0.2.0 - MagFace model

  • MLModule Kit version 2 with PyTorch 1.9.1
  • MagFace model to replace ArcFace
  • Moving to minor version increases for the future.

- Python
Published by smassonnet over 4 years ago

mozuma - 0.1.12.post1 - Device load fix

  • Fix loading of module weights on the wrong device

- Python
Published by smassonnet over 4 years ago

mozuma - 0.1.12 - Fixed region selection on RPN

  • Fixed region selection on RPN
  • Replace Region features with ResNet for faster inference

- Python
Published by smassonnet over 4 years ago

mozuma - 0.1.11 - Fix slow import in RNP

  • Dropped support for testing on PC32

  • Fixed slow import in RPN

  • Added metrics to a module for benchmarking

- Python
Published by smassonnet over 4 years ago

mozuma - 0.1.10 - Region Features bugfix

  • #22

- Python
Published by smassonnet over 4 years ago

mozuma - 0.1.9 - Updated MMDetection

  • See https://github.com/open-mmlab/mmdetection/commit/57953677687bf284f8345f62387009c36283107a

- Python
Published by smassonnet over 4 years ago

mozuma - 0.1.8 - Add image region detection

  • Added mlmodule.contrib.rpn.RegionFeatures model for image region detection.

  • Improved Docker support with MLModuleKit images.

- Python
Published by smassonnet over 4 years ago

mozuma - 0.1.7 - Docker support

  • MLModule can now be used from its Docker image lsirepfl/mlmodule
  • There is a new Docker image containing PyTorch and Torchvision (see https://hub.docker.com/repository/docker/lsirepfl/pytorch).

- Python
Published by smassonnet over 4 years ago

mozuma - 0.1.6 - Face out of memory fix and CLI improvment

  • Fixes Out of memory issue with Arcface
  • Added support to run inference on HTTP URL of images.

- Python
Published by smassonnet almost 5 years ago

mozuma - 0.1.5 - Performance improvments

  • Improved MTCNN implementation

  • Added shrink-on-load for faster image loading

- Python
Published by smassonnet almost 5 years ago

mozuma - 0.1.4 - Run model CLI

  • Faces fix
  • CLI to run a model

- Python
Published by smassonnet almost 5 years ago

mozuma - 0.1.3 - Face quality filter

  • Fix MTCNN bugs
  • Added face quality filters

- Python
Published by smassonnet almost 5 years ago

mozuma - 0.1.2 - Face detection

Added face detection (MTCNN) and face embeddings (ArcFace)

- Python
Published by smassonnet almost 5 years ago

mozuma - 0.1.1 - CLIP Integration

Added https://github.com/openai/CLIP to MLModules

- Python
Published by smassonnet almost 5 years ago

mozuma - 0.1.0 - DensetNet

- Python
Published by smassonnet almost 5 years ago