Is it possible to support pretrained model with dynamic adapters?
Open
Nobody has claimed this yet.
enhancement
- Dominant language
- C++
- Stars
- 4.7k
- Forks
- 536
- Avg merge
- 12h 12m
- Merged PRs (30d)
- 4
Description
Hi.
I want to implement a feature to train a generic translation model on a large dataset, where we have many translator users, each of whom has a personal memory bank they maintain. I would like to use Pretrain Transformer + Adapter Layers to support users to upload some data to finetune customized models. Is it possible for CTranslate2 to support such dynamic loading and releasing of Adapter parameters?
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
The issue names no files, tests, or entry points. Begin by tracing how pretrained Transformer models and adapter layers are represented, then establish whether per-user adapter parameters can be dynamically loaded and released.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100