uer
/

roberta-base-chinese-extractive-qa

Question Answering

Inference Endpoints

Model card Files Files and versions Community

uer commited on Feb 20, 2021

Commit

a3c6de6

•

1 Parent(s): d5be715

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -31,7 +31,7 @@ Training data comes from three sources: [cmrc2018](https://github.com/ymcui/cmrc
 ## Training procedure
-The model is fine-tuned by [UER-py](https://github.com/dbiir/UER-py/) on [Tencent Cloud TI-ONE](https://cloud.tencent.com/product/tione/). We fine-tune three epochs with a sequence length of 512 on the basis of the pre-trained model [chinese_roberta_L-12_H-768](https://huggingface.co/uer/chinese_roberta_L-12_H-768).
 ```
 python3 run_cmrc.py --pretrained_model_path models/cluecorpussmall_roberta_base_seq512_model.bin-250000 \

 ## Training procedure
+The model is fine-tuned by [UER-py](https://github.com/dbiir/UER-py/) on [Tencent Cloud TI-ONE](https://cloud.tencent.com/product/tione/). We fine-tune three epochs with a sequence length of 512 on the basis of the pre-trained model [chinese_roberta_L-12_H-768](https://huggingface.co/uer/chinese_roberta_L-12_H-768). At the end of each epoch, the model is saved when the best performance on development set is achieved.
 ```
 python3 run_cmrc.py --pretrained_model_path models/cluecorpussmall_roberta_base_seq512_model.bin-250000 \