roberta-base

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

• Loss: 0.7299
• Accuracy: 0.6766

Model description

roberta-base fine tuned on the kaggle Contradictory, my dear Watson dataset for that challenge.

Intended uses & limitations

Search params were heavily dependant on my (very very) small budget for training. Furthermore, there r

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.0235	1.0	682	0.8367	0.5990
0.7854	2.0	1364	0.7110	0.6741
0.6333	3.0	2046	0.7299	0.6766

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.1
• Tokenizers 0.21.1