ban-lan-gen
- Karma
- 6
- Created
- April 15, 2021 (5y ago)
- Submissions
- 0
-
If I have a lot of precomputed KV cache values (basically just tensors), how do I store them efficiently on disk? Ideally, I would want to lookup their values by prefix quickly as well.
- Dataflow Execution in GPU (arxiv.org)
- HBF for GPUs (tomshardware.com)