Samuel Tovey
Samuel Tovey
Just playing with tasks and inner code to make things work.
**What feature would you like to see added?** In our shared network architecture, there is currently no way to use multiple optimizers in training. This could be important when, for...
Fix #48 by adding a checkpointing callback to the code #### Summary of additions and changes * Moved model saving to the protocols * Updates some type hints and comments...
**What feature would you like to see added?** I want to update the data storage in SwarmRL to handle different cases. We have the following situations: * Standard simulation storing...
Adding `vmap` and `jit` everywhere. Statistics: * Observable `vmap` + `jit` yielded ~20 % simulation time speedup when compared with `master` for gradient sensing observable.
**What feature would you like to see added?** For very long runs or when clusters / computers shut down, it would be great to have some model checkpointing available. **Describe...
Just a list of things I want to change in the code for the sake of simplicity or performance or both: - [x] Remove hidden file saving for the RL...
**What feature would you like to see added?** It might be nice to use MLFlow for variable tracking in the RL. This way we could have callbacks built into each...
**What feature would you like to see added?** Whilst we are looking into and discussing remote trajectory databases, it would be simpler and faster to also consider remotely accessible or...
**What feature would you like to see added?** In line with our guidelines on bringing state of the art analysis tools to the software, it would be nice to look...