Renumics Spotlight Versions Save

Features

• Mel scale for spectrogram lens
• Automatic retry of failed lens updates

Features

• Builtin TLS support
• Datasource for pyarrow datasets.
• optional handlesNull attribute for lenses

Fixes

• Fixed coloring on filtered similarity map

Features

• Add bounding box lens to Python layout API
• Automatically infer categories when specifying sequence of category dtype in spotlight.show

Fixes

• Fix allowed input column combinations to bounding box lens in UI
• Allow to start Spotlight with dtype but without dataset

Fixes

• Support pandas.DataFrames with (possibly non-unique) multiindex columns and/or index
• Show proper error when fail to load dataset instead of opening filebrowser immediately and possibly failing if is not allowed

Fixes

• Backup served pandas.DataFrame on spotlight.Viewer.close, so spotlight.Viewer.df is available after close.

Features

• Experimental support for arrow datasets in spotlight.show

Fixes

• Fix usage of boolean and integer pandas.DataFrame columns.

Fixes

• Fix usage of boolean, integer and float pandas.DataFrame columns with new numpy versions.

Features

• Automatically create embeddings for columns containing images (ViT), audio (Wav2Vec2), or strings (gte-base)

  spotlight.show(..., embed=['colA', 'colB'])
  

• New DType for axis aligned bounding boxes (bounding_box_dtype).
• New DType for Sequences of values. Supports bounding boxes and categories for now.
• New lens for labeled axis-aligned bounding boxes on images.
• New metrics: Levensthein distance, BLEU Score.
• New lenses for text: ROUGE & BLEU Score.

Fixes

• Fixed a problem when reading large binary data (e.g. audio files or images) from HuggingFace datasets.