| ## BERT-large finetuned on MNLI. | |
| The [reference finetuned model](https://github.com/google-research/bert) has an accuracy of 86.05, we get 86.7: | |
| ``` | |
| {'eval_loss': 0.3984006643295288, 'eval_accuracy': 0.8667345899133979} | |
| ``` |
| ## BERT-large finetuned on MNLI. | |
| The [reference finetuned model](https://github.com/google-research/bert) has an accuracy of 86.05, we get 86.7: | |
| ``` | |
| {'eval_loss': 0.3984006643295288, 'eval_accuracy': 0.8667345899133979} | |
| ``` |