| tags: | |
| - generated_from_trainer | |
| datasets: | |
| - covid_qa_deepset | |
| model-index: | |
| - name: covid_qa_analysis_bert_base_uncased_squad2 | |
| results: [] | |
| <!-- This model card has been generated automatically according to the information the Trainer had access to. You | |
| should probably proofread and complete it, then remove this comment. --> | |
| # covid_qa_analysis_bert_base_uncased_squad2 | |
| This model is a fine-tuned version of [twmkn9/bert-base-uncased-squad2](https://huggingface.co/twmkn9/bert-base-uncased-squad2) on the covid_qa_deepset dataset. | |
| ## Model description | |
| More information needed | |
| ## Intended uses & limitations | |
| More information needed | |
| ## Training and evaluation data | |
| More information needed | |
| ## Training procedure | |
| ### Training hyperparameters | |
| The following hyperparameters were used during training: | |
| - learning_rate: 2e-05 | |
| - train_batch_size: 8 | |
| - eval_batch_size: 8 | |
| - seed: 42 | |
| - distributed_type: tpu | |
| - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 | |
| - lr_scheduler_type: linear | |
| - num_epochs: 3 | |
| ### Training results | |
| ### Framework versions | |
| - Transformers 4.16.2 | |
| - Pytorch 1.9.0+cu102 | |
| - Datasets 1.18.3 | |
| - Tokenizers 0.11.6 | |