Skip to content

sprint 1 follow-up finetuning experiments #223

@Locke0

Description

@Locke0

Investigate how to improve 7B CUA models' performance on perturbation eval data with finetuning

Metadata

Metadata

Labels

benchmarksEvaluation, benchmarkingdatasetRelated to datasets or data workexperimentResearch experiment or prototypemodelsModel architecture, trainingresearchresearch, experiments, papers

Type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions