(description)
- (list of things we need to get done)
Status: (status)
Blockers: (list of things we have to de before we can do this milestone)
Handcraft weights of T5 model to perform the identity task out of the box.
- identified a bug that blocks benchmarks on this model
Status: complete but buggy
Handcraft weights of T5 model to perform the set deduplication task out of the box.
- needs layer to replace nonduplicate characters with
<pad>
Status: mostly complete
- Use results from paper to compare to downstream performance of the synthetic initialized identity.
status: complete (from paper)
- Use results from paper to compare to downstream performance of the synthetic initialized set.
status: complete (from paper)
- Run finetuning on model with random initializations.
- Build a script for finetuning that can take an arbitrary model and return a model optimized for a given downstream task.
- Get training results and compile them to a graph.
- status: not started
- needs finetuning script for huggingface transformers and downstream tasks
blockers: none (this is ready and needs to be done soon)
- Run finetuning on model with identity synthetic initialization.
- Use the finetuning script from (baseline).
- Get training results and compile them to a graph.
status: not started
blockers:
- baseline
- bug fix for decoder
status: not started
blockers:
- baseline
- synthetic nonduplicate padding layer
status: todo, fingers crossed
status: todo, very important!
- we have been painting very localized pictures in our two demos
- for our final demo we need to clearly restate our goals and our line of thinking
- then we need to ideally show the results
- as a stretch goal it would be nice to demostrate the generalized synthetic weight generation