XMemTransfer Mistral adaptor from LLaMA-2 source (R4, 30M/30M)
This repository contains a target-side adaptor checkpoint for the XMemTransfer cross-model QA setting.
Project links
- Website: https://olaresearch.org/XMemTransfer
- Code: https://github.com/OLAResearch
- Models: https://huggingface.co/OLAResearchX
Files
adaptor_best.pt: best target-side adaptor checkpointconfig.json: adaptor configurationresults.json: summary metrics from the run
Provenance
- source memory: LLaMA-2 7B WikiText-2021 source memory
- target model family: Mistral 7B
- reader setting: dual-layer, 4-branch
- token budgets: 30M source / 30M target-side fitting
- Downloads last month
- 3
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support