| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| ecco-0.1.0-py2.py3-none-any.whl | 2021-12-29 | 70.6 kB | |
| ecco-0.1.0.tar.gz | 2021-12-29 | 65.4 kB | |
| README.md | 2021-12-29 | 654 Bytes | |
| v0.1.0 - Support for T5, local models, integrated gradients, Captum integrations, and beam search decoding source code.tar.gz | 2021-12-29 | 1.1 MB | |
| v0.1.0 - Support for T5, local models, integrated gradients, Captum integrations, and beam search decoding source code.zip | 2021-12-29 | 1.2 MB | |
| Totals: 5 Items | 2.5 MB | 0 | |
Big update to Ecco! Massive contributions from @JoaoLages and @SSamDav.
- Added support of encoder-decoder models like T5
- Using Captum for feature attribution adds support to these new methods: (IntegratedGradients, Saliency, InputXGradient, DeepLift, DeepLiftShap, GuidedBackprop, GuidedGradCam, Deconvolution, LRP). This replaces the previous implementation within Ecco for Saliency and InputXGradients.
- Added support for Beam Search generation
- Added support for importing local models. Very useful for analyzing finetuned models.
- Added better support for various tokenizers in visualizations. Some more work needed on this front, still.