Pytrial

- Add new sequence patient data generation model `TWIN`: `trial_simulation.sequence.TWIN` [1]
- fix package conflicts, align sequence simulation methods with the latest `rdt` and `transformers`.


[1] TWIN: Personalized Clinical Trial Digital Twin Generation. Trisha Das*, Zifeng Wang*, and Jimeng Sun. KDD'23 (2023).

- Fix bugs in MedGAN when batch size is too large

- Fix conflict bugs caused by rdkit

- Add the state-of-the-art clinical trial outcome prediction model `SPOT` to `tasks.trial_outcome`
- Fix bugs in `MedGAN` in `tasks.trial_simulation`
- Fix bugs in `BERT` in `model_utils.bert`
- Update docs for `SPOT` and its examples in tutorial

- Add KNNSampler to `trial_simulation.sequence`
- Add presence disclosure risk and fidelity test to `trial_simulation.sequence.evaluation`
- Fix bugs in models under `indiv_outcome.sequence` when `mode='regression'`
- Update docs to reflect the updates of trial simulation models

- Add MedGAN to `trial_simulation.tabular`.
- Rename the title and demo figure of this package.

The beta release of PyTrial package!