0.2.2

📚 Documentation

• Updated code-example to work out-of-the-box
  • PR: #62

0.2.1

• no changes

0.2.0

🚀 Features

• Refactor the structure of a saved result to be easier for building benchmarks
  • PR: #54
• feat: Update result upload endpoint and payload
  • PR: #56

🐛 Fixes

• fix: Use the Parquet file size and checksum for the table content metadata when creating a dataset on the Hub
  • PR: #57
• Fix issues due to the string representation of the HubOwner
  • PR: #59

0.1.0

🚀 Features

• Redid the BenchmarkResults structure to be tabular
  • PR: #44
• Refactor dataset upload
  • PR: #50
• Extended metrics
  • PR: #48

0.0.2

🚀 Features

• Bootstrap Dataset API
  • PR: #3
• Splitter module
  • PR: #4
• First implementation of the evaluation logic
  • PR: #6
• HTML based repr for dataset and benchmark objects
  • PR: #15
• Pydantic 2.x
  • PR: #19
• Add code coverage
  • PR: #23
• First version of automated dataset curation module
  • PR: #7
• Kick-started the Polaris documentation
  • PR: #25
• Add the SIMPD (SImulated Medicinal chemistry Project Data) splitter
  • PR: #20
• Remove the splitter module
  • PR: #26
• feat: Additional model typing to match Hub API
  • PR: #28
• WIP: Integrating with the Polaris Hub
  • PR: #29
• Added a README field to the benchmark and dataset classes
  • PR: #40

🐛 Fixes

• Simplified the indices for multi-task splits and added the split to the checksum
  • PR: #12
• Fixes bugs in loading Dataset from a .zarr and in the use of pointer columns
  • PR: #13
• fix: Local Hub compatibility
  • PR: #33

🧪 Tests

• Add missing test cases for test targets access, pointer columns and multiple test sets
  • PR: #17

📚 Documentation

• Kick-started the Polaris documentation
  • PR: #25
• Improve docs and error-handling of the PolarisHubClient.
  • PR: #41

0.0.1

Release version 0.0.1