v2.0.0

What's Changed

• README update by @franberchez in #38
• Distribution module refactoring and documentation improvement by @victormvy in #42
• Added torch.no_grad() scope in PytorchEstimator.predict_proba by @victormvy in #46
• Removed default value for num_classes parameter of unimodal loss functions by @victormvy in #45
• Distributions by @franberchez in #48
• Tests absolute imports by @franberchez in #49
• Modification of the estimator module to return numpy arrays by @franberchez in #50
• Update the pretrained parameter to weights in the tutorials by @franberchez in #52
• Documentation by @franberchez in #57
• [API] Rename distributions module to soft_labelling by @victormvy in #59
• [API] Changed get_beta_softlabels interface to match the others by @victormvy in #60
• Minor changer in documentation and templates by @franberchez in #63
• Elimination of label_smoothing parameter by @franberchez in #64
• Dropout module by @franberchez in #61
• CLM fix to work on cpu and gpu by @franberchez in #65
• Estimator removal by @franberchez in #67
• New package version 2.0.0 by @franberchez in #68

Full Changelog: v1.0.1...v2.0.0