【开源实习】bart模型微调 #2026

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Merged

lvyufeng merged 1 commit into mindspore-lab:0.4 from Bireflection:bart0.4

Apr 28, 2025

+139 −0

Bireflection commented Apr 19, 2025

bart模型微调报告

任务

任务编号：#IAUOXU
任务链接：【开源实习】bart模型微调
实现内容：实现了bart模型在XSum数据集上的微调。
模型：facebook/bart-base
数据集：EdinburghNLP/xsum

结果对比

Mindnlp+D910B

Epoch	Eval Loss	R1 (ROUGE-1)	R2 (ROUGE-2)	RL (ROUGE-L)	RLsum (ROUGE-Lsum)
1	0.4504	0.5265	0.2512	0.5003	0.5004
2	0.4481	0.5272	0.2538	0.5026	0.5025
3	0.4440	0.5316	0.2580	0.5061	0.5062

Pytorch+3090

Epoch	Eval Loss	R1 (ROUGE-1)	R2 (ROUGE-2)	RL (ROUGE-L)	RLsum (ROUGE-Lsum)
1	0.4364	0.5226	0.2432	0.4965	0.4961
2	0.4297	0.5309	0.2547	0.5066	0.5065
3	0.4290	0.5318	0.2563	0.5065	0.5062


          bart finetune on xsum

7b30718

lvyufeng merged commit 94981cd into mindspore-lab:0.4

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet