german-qg-t5-e2e-quad (Work in progress)

This model is a end-to-end question generation model in German. Given a text, it generates several questions about it. This model is a fine-tuned version of valhalla/t5-base-e2e-qg on the GermanQuAD dataset from deepset.

Model description

More information needed

Training and evaluation data

Bleu_1: 0.196051
Bleu_2: 0.122380
Bleu_3: 0.079980
Bleu_4: 0.053672

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0001
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10.0

Training results

Framework versions

• Transformers 4.16.0.dev0
• Pytorch 1.10.0+cu111
• Datasets 1.16.1
• Tokenizers 0.10.3