bert_base_km_20_v1_wnli

This model is a fine-tuned version of Hartunka/bert_base_km_20_v1 on the GLUE WNLI dataset. It achieves the following results on the evaluation set:

• Loss: 0.7542
• Accuracy: 0.3944

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: 256
• eval_batch_size: 256
• seed: 10
• optimizer: Use 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: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.7377	1.0	3	0.7542	0.3944
0.7065	2.0	6	0.7640	0.2113
0.6898	3.0	9	0.7914	0.2254
0.692	4.0	12	0.8143	0.2676
0.6766	5.0	15	0.8582	0.2113
0.6757	6.0	18	0.8994	0.1690

Framework versions

• Transformers 4.50.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.21.1