bangla-hatespeech-analysis

This model is a fine-tuned version of csebuetnlp/banglabert on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4000
• Accuracy: 0.8694

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
• 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
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.3883	0.2700	1000	0.3539	0.8531
0.3848	0.5400	2000	0.3685	0.8531
0.3438	0.8099	3000	0.3294	0.8635
0.2659	1.0799	4000	0.3643	0.8605
0.2775	1.3499	5000	0.3422	0.8695
0.2956	1.6199	6000	0.3923	0.8535
0.2848	1.8898	7000	0.3431	0.8701
0.2263	2.1598	8000	0.3707	0.8729
0.1947	2.4298	9000	0.4239	0.8737
0.2111	2.6998	10000	0.3852	0.8701
0.1759	2.9698	11000	0.4000	0.8694

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0