fahmiaziz's picture
Update README.md
3a22f5f
|
raw
history blame
322 Bytes
---
license: openrail
datasets:
- naver-clova-ix/cord-v2
metrics:
- accuracy
- code_eval
library_name: transformers
pipeline_tag: image-to-text
---
Fine-tuning Donut base with CORD dataset
- Accuracy 90%
Need to retrain the model, although the accuracy is high but there is something wrong with the prediction results