malayalam-bert-FakeNews-Dravidian-finalwithPP

This model is a fine-tuned version of l3cube-pune/malayalam-bert on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0597
• Accuracy: 0.9890
• Weighted f1 score: 0.9890
• Macro f1 score: 0.9890

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Weighted f1 score	Macro f1 score
0.879	1.0	255	0.6737	0.8417	0.8403	0.8403
0.5845	2.0	510	0.4242	0.9178	0.9178	0.9178
0.3641	3.0	765	0.2130	0.9656	0.9656	0.9656
0.2351	4.0	1020	0.1512	0.9681	0.9681	0.9681
0.1702	5.0	1275	0.0936	0.9816	0.9816	0.9816
0.109	6.0	1530	0.0734	0.9853	0.9853	0.9853
0.0904	7.0	1785	0.0670	0.9877	0.9877	0.9877
0.0692	8.0	2040	0.0600	0.9877	0.9877	0.9877
0.0468	9.0	2295	0.0612	0.9890	0.9890	0.9890
0.0471	10.0	2550	0.0597	0.9890	0.9890	0.9890

Framework versions

• Transformers 4.35.0
• Pytorch 2.0.0
• Datasets 2.11.0
• Tokenizers 0.14.1