distilbert-fever

This model is a fine-tuned version of distilbert/distilbert-base-uncased on the fever dataset. It achieves the following results on the evaluation set:

• Loss: 2.1761
• Accuracy: 0.529

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: 0.0002
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	63	1.0570	0.498
No log	2.0	126	1.1103	0.505
No log	3.0	189	1.1786	0.543
No log	4.0	252	1.2573	0.506
No log	5.0	315	1.2150	0.5
No log	6.0	378	1.4350	0.512
No log	7.0	441	1.3709	0.533
0.5664	8.0	504	1.5878	0.523
0.5664	9.0	567	1.8523	0.535
0.5664	10.0	630	1.9348	0.519
0.5664	11.0	693	1.7897	0.516
0.5664	12.0	756	1.8938	0.523
0.5664	13.0	819	1.7832	0.54
0.5664	14.0	882	1.7101	0.524
0.5664	15.0	945	1.9640	0.514
0.2103	16.0	1008	2.0698	0.513
0.2103	17.0	1071	2.2193	0.523
0.2103	18.0	1134	2.2431	0.527
0.2103	19.0	1197	2.1999	0.522
0.2103	20.0	1260	2.1761	0.529

Framework versions

• Transformers 4.57.1
• Pytorch 2.9.1
• Datasets 4.4.1
• Tokenizers 0.22.1