Molscore

Built in replay buffer
SMILES augmentation
Tutorial for replay buffer

- Avoid RDKit deprecation of MolStandardize.Standardizer() but maintain behaviour with original MolVS

- GUI improvements
- Select single tasks specifying the benchmark first e.g., `GuacaMol:Albuterol_similarity`
- Override config `n_jobs` by specifying command line environment variable `MOLSCORE_NJOBS` or `MOLEVAL_NJOBS`
- Bug fixes & dependency removals

Addition of curriculum learning and documention

- Correct for rDock SDF bug when using pharmacophoric constraints

Addition of MolSkill