bert-28

This model is a fine-tuned version of deepset/bert-base-cased-squad2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 11.6678

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
11.2204	0.18	5	12.3027
10.9471	0.36	10	12.2255
11.0769	0.54	15	12.1528
10.8237	0.71	20	12.0849
10.6924	0.89	25	12.0233
10.5051	1.07	30	11.9672
10.6612	1.25	35	11.9154
10.5235	1.43	40	11.8683
10.4814	1.61	45	11.8262
10.5109	1.79	50	11.7882
10.623	1.96	55	11.7560
10.4358	2.14	60	11.7285
10.234	2.32	65	11.7056
10.2805	2.5	70	11.6881
10.6227	2.68	75	11.6760
10.2278	2.86	80	11.6678

Framework versions

• Transformers 4.34.1
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.14.1