bert-base-multilingual-cased-v3

This model is a fine-tuned version of bert-base-multilingual-cased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0164
• Accuracy: 0.7026
• F1: 0.6016

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.085	0.9877	40	0.7707	0.5955	0.5018
0.2531	2.0	81	0.1906	0.6979	0.6015
0.046	2.9877	121	0.0572	0.7230	0.6270
0.02	4.0	162	0.0656	0.7230	0.6282
0.015	4.9383	200	0.0540	0.7247	0.6345

Framework versions

• Transformers 4.44.2
• Pytorch 2.6.0+cu126
• Datasets 3.3.2
• Tokenizers 0.19.1