distilbert-base-cased-reward-neurallinguisticpioneers

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

• Loss: 0.2411
• Mse: 3.7748

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: 5e-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: constant
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Mse
1.4559	1.0	122	0.6534	3.4024
0.5476	2.0	244	0.5601	3.8827
0.4224	3.0	366	0.4717	3.8263
0.3534	4.0	488	0.3511	3.7530
0.2827	5.0	610	0.2960	3.8889
0.2541	6.0	732	0.2416	3.5817
0.2289	7.0	854	0.3085	4.0660
0.1997	8.0	976	0.3212	3.4440
0.1889	9.0	1098	0.2852	3.9351
0.1752	10.0	1220	0.2360	3.8505
0.1683	11.0	1342	0.2939	4.1039
0.1601	12.0	1464	0.3242	4.0499
0.155	13.0	1586	0.2297	3.8442
0.1478	14.0	1708	0.2707	3.8680
0.1439	15.0	1830	0.2582	3.8703
0.1462	16.0	1952	0.2411	3.7748

Framework versions

• Transformers 4.27.3
• Pytorch 1.13.0+cu117
• Datasets 2.12.0
• Tokenizers 0.13.3