TorchStudio 0.9.9

• you can now choose a Pytorch channel (stable, test, nightly) when installing a new Python environment.
• install a full Python environment on any remote server in a single click, making all remote server compatible with TorchStudio. Just leave blank the "command" parameter when setting the remote server.
• TorchStudio files can be cleaned from remote servers using the new "Clean" button.
• torchvision models are now better documented, listing all possible weights values in the model description
• fix Menu > Settings > Open Terminal on mac

TorchStudio 0.9.8

Highlights

This release brings support for PyTorch 1.12, Torchvision 0.13, Apple Silicon, Metal acceleration, Fedora, new features and fix many issues (see full changelog below for details).

If your mac supports it, it is strongly recommended to make Metal Acceleration the default device for training and inference (Menu > Settings). If you can't see it, click Reset Python and relaunch Torchstudio.

Full Changelog

• add support for PyTorch 1.12 and future versions of PyTorch
• add support for Apple Silicon
• add support for Metal acceleration (new default device for inference and training, if not change it in Menu > Settings)
• add a RPM installer for Fedora 32 or higher and other RPM-based Linux distributions
• add support for the new optical_flow torchvision model category
• add Early Stopping feature (in the Model Tab, Hyperparameters)
• add Restore Best Epoch feature (in the Model Tab, Hyperparameters)
• add an Add Packages... button in Menu > Settings > Local Environment to add conda package to the local python environment
• add an Open Terminal... button in Menu > Settings > Local Enviroment to open a python terminal using the local environment
• python install and basic packages install now much faster and stable
• local python environment details are now displayed in Menu > Settings > Local Environment
• signals renderer module improved with more parameters (scaling)
• all parameters now have tooltips
• fix unstable ssh connections with some servers
• fix TorchStudio not launching with some Linux environments
• fix some urls not being displayed properly for some models and datasets documentation
• fix rendering issues when displaying signals with many channels
• fix shuffling setting not resetting when reloading the dataset
• fix training state sometime not properly updated when training is stopped
• TorchStudio scripts are now under MIT license

TorchStudio 0.9.7

• fix ssl certificate fail when downloading datasets on some configurations
• fix missing MSVC dependencies on Windows
• remove unnecessary torch dependencies for some analyzers

TorchStudio 0.9.6

• better check box indicator
• fix remote servers on mac
• new log system

TorchStudio 0.9.5

• improve app initialization sequence reliability
• improve app initialization status display

TorchStudio 0.9.4

• fix remote server connection issues
• fix data folder creation issue
• stronger python/pytorch environment check at startup

TorchStudio 0.9.3

• allow PyTorch 1.9 as minimum requirement
• fix PyTorch install on mac
• improve Python install on other platforms

TorchStudio 0.9.2

• fix PyTorch install on mac, in particular with M1 machines
• fix PyTorch version detection with CUDA builds
• fix Python binary selector when selecting a custom Python install
• fix error message when an error occurs during training

TorchStudio 0.9.1

• new Package model build mode added, allowing original python models to be trained on remote servers instead of compiled torchscript models.
• more efficient communication with the remote servers
• remote servers auto-update script files if needed
• fixes several issues from the closed beta