MEFT: Memory-Efficient Fine-Tuning through Sparse Adapter.
Jitai HaoWeiwei SunXin XinQi MengZhumin ChenPengjie RenZhaochun RenPublished in: ACL (1) (2024)
Keyphrases
- memory efficient
- fine tuning
- viable alternative
- fine tuned
- fine tune
- iterative deepening
- external memory
- sparse data
- high dimensional
- multiple sequence alignment
- sparse representation
- integral image
- compressed sensing
- sparse matrix
- compressive sensing
- random projections
- dimensionality reduction
- low latency
- canonical correlation analysis
- domain specific