bangla-clickbait-multitask

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

• Loss: 1.2984
• Binary Accuracy: 0.8662
• Binary F1 Macro: 0.8652
• Binary F1 Weighted: 0.8662
• Multi Accuracy: 0.5400
• Multi F1 Macro: 0.4968
• Multi F1 Weighted: 0.5394

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: 1e-05
• train_batch_size: 16
• 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
• lr_scheduler_warmup_steps: 0.1
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Binary Accuracy	Binary F1 Macro	Binary F1 Weighted	Multi Accuracy	Multi F1 Macro	Multi F1 Weighted
1.4507	1.0	605	1.4136	0.8233	0.8214	0.8230	0.3881	0.2760	0.3054
1.2076	2.0	1210	1.1768	0.8704	0.8701	0.8706	0.5095	0.4319	0.4811
1.0117	3.0	1815	1.0525	0.8728	0.8722	0.8730	0.5640	0.4899	0.5411
0.9297	4.0	2420	1.0100	0.8786	0.8775	0.8785	0.5739	0.4984	0.5505
0.7527	5.0	3025	1.0259	0.8671	0.8664	0.8672	0.5640	0.4943	0.5456
0.7040	6.0	3630	1.0172	0.8720	0.8713	0.8721	0.5673	0.5028	0.5580
0.6455	7.0	4235	1.0516	0.8737	0.8724	0.8735	0.5467	0.5036	0.5487
0.5461	8.0	4840	1.0951	0.8695	0.8688	0.8696	0.5533	0.5049	0.5524
0.4756	9.0	5445	1.1419	0.8646	0.8641	0.8648	0.5698	0.5117	0.5616
0.5018	10.0	6050	1.2189	0.8613	0.8609	0.8615	0.5607	0.5015	0.5540
0.4339	11.0	6655	1.2076	0.8662	0.8650	0.8661	0.5533	0.5075	0.5537
0.4350	12.0	7260	1.2696	0.8679	0.8672	0.8680	0.5186	0.4927	0.5225
0.4042	13.0	7865	1.2740	0.8654	0.8644	0.8654	0.5376	0.4912	0.5358
0.3195	14.0	8470	1.2879	0.8687	0.8677	0.8687	0.5425	0.4991	0.5421
0.3332	15.0	9075	1.2984	0.8662	0.8652	0.8662	0.5400	0.4968	0.5394

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2