result
This model is a fine-tuned version of distilbert-base-uncased on an unknown dataset.
Model description
The model was trained with the NaturalQuestions dataset from https://github.com/mrqa/MRQA-Shared-Task-2019
Intended uses & limitations
This fine-tuned model is used in the context of the NLP4Web course at TU Darmstadt
Training and evaluation data
The NaturalQuestions dataset from https://github.com/mrqa/MRQA-Shared-Task-2019 was used
Training procedure
The pytorch question-answering/run_qa.py script was used to fine tune the model.
The following parameters were passed to the script
--model_name_or_path distilbert-base-uncased
--train_file $TRAIN_FILE_NAME
--validation_file $EVAL_FILE_NAME
--do_train
--do_eval
--per_device_train_batch_size 12
--learning_rate 3e-5
--num_train_epochs 2
--max_seq_length 384
--doc_stride 128
--output_dir ./tmp/result/
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 3e-05
- train_batch_size: 12
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2.0
Training results
***** train metrics ***** epoch = 2.0 train_loss = 0.9724 train_runtime = 2:57:05.43 train_samples = 148108 train_samples_per_second = 27.878 train_steps_per_second = 2.323
Framework versions
- Transformers 4.27.0.dev0
- Pytorch 1.13.1+cu116
- Datasets 2.9.0
- Tokenizers 0.13.2