This model is a fine-tuned version of Shahm/bart-german on the GermanQuAD dataset. This is a model for question generation from a text corpus.
The following hyperparameters were used during training:
- learning_rate: 1e-4
- mini_batch_size: 8
- optimizer: Adam
- num_epochs: 3
- scheduler: get_linear_schedule_with_warmup
