* Transformers supports Tensorflow
* Performance tuning during clustering
* Support multi-label
* Custom Embeddings, Classification, Clustering, Scoring(Learning) function
* Support TorchVision for image embeddings
* Support SentenceTransformers
* Add Least Confidence Sampling and Most Confidence Sampling
* Add Semi-supervised learning
* Add Farthest (Clustering) Sampling, Mismatch (Uncertainity) Sampling
* Add Mismatch-farthest Learning