muril-adapted-local-FakeNews-Dravidian-muril

This model is a fine-tuned version of monsoon-nlp/muril-adapted-local on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5333
• Accuracy: 0.7448
• Weighted f1 score: 0.7448
• Macro f1 score: 0.7447

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-06
• 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: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Weighted f1 score	Macro f1 score
0.7193	1.0	204	0.7031	0.5350	0.4469	0.4477
0.6807	2.0	408	0.6694	0.5840	0.5800	0.5798
0.6606	3.0	612	0.6487	0.6160	0.6160	0.6160
0.6377	4.0	816	0.6211	0.6638	0.6602	0.6601
0.5906	5.0	1020	0.6154	0.6748	0.6552	0.6549
0.5526	6.0	1224	0.5716	0.6994	0.6956	0.6955
0.5195	7.0	1428	0.5471	0.7202	0.7200	0.7199
0.4945	8.0	1632	0.5410	0.7264	0.7256	0.7255
0.4635	9.0	1836	0.5448	0.7350	0.7332	0.7332
0.4453	10.0	2040	0.5355	0.7362	0.7358	0.7358
0.429	11.0	2244	0.5337	0.7362	0.7356	0.7356
0.4262	12.0	2448	0.5539	0.7399	0.7374	0.7373
0.4041	13.0	2652	0.5334	0.7485	0.7484	0.7484
0.3983	14.0	2856	0.5326	0.7497	0.7497	0.7497
0.3798	15.0	3060	0.5333	0.7448	0.7448	0.7447

Framework versions

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