- When first starting Tools for AI, an installation page is shown for guiding local AI development environment setup. Users can launch a one-click installer from this page to install required software packages into the default Python environment.
- We design a new start page to help users build first AI application / train first AI Model within 3 steps. The start page also contains useful AI examples and development resources.
- View local Jupyter notebooks.
- Built-in documentation search for TensorFlow / Keras / PyTorch APIs. Right-click the function name, and select “Search in Documentation Viewer” context menu.
- Docker image list is updated for job submission to remote machines. And a new all-in-one docker image including all popular AI / DL frameworks is provided.
- A more proactive feedback channel is built - will prompt user to give feedback when some conditions met.
- Azure Batch AI support is temporarily removed, will get the support of Batch AI back soon.
- Telemetry and stability improvement.