hing-roberta-CM-run-3

This model is a fine-tuned version of l3cube-pune/hing-roberta on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.6968
• Accuracy: 0.7565
• Precision: 0.7045
• Recall: 0.7064
• F1: 0.7050

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: 3e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.8232	1.0	497	0.7145	0.6620	0.6319	0.6585	0.6167
0.5799	2.0	994	0.7155	0.7203	0.6718	0.6928	0.6672
0.4152	3.0	1491	0.8823	0.7485	0.6962	0.7136	0.7022
0.2657	4.0	1988	1.4502	0.7465	0.6945	0.7037	0.6968
0.16	5.0	2485	2.0667	0.7465	0.6890	0.6827	0.6855
0.0945	6.0	2982	2.0120	0.7565	0.7091	0.7159	0.7103
0.0802	7.0	3479	2.2426	0.7686	0.7253	0.7065	0.7088
0.059	8.0	3976	2.3472	0.7425	0.6844	0.6881	0.6861
0.041	9.0	4473	2.4801	0.7666	0.7258	0.7144	0.7145
0.0307	10.0	4970	2.6317	0.7545	0.7102	0.7021	0.7019
0.0471	11.0	5467	2.4626	0.7364	0.6836	0.6780	0.6788
0.0282	12.0	5964	2.3949	0.7586	0.7067	0.7108	0.7087
0.0267	13.0	6461	2.4750	0.7465	0.6938	0.6921	0.6921
0.0274	14.0	6958	2.5942	0.7565	0.7022	0.7062	0.7039
0.0212	15.0	7455	2.6728	0.7404	0.6851	0.6893	0.6867
0.026	16.0	7952	2.6683	0.7565	0.7064	0.7122	0.7085
0.0175	17.0	8449	2.6646	0.7505	0.7030	0.7087	0.7039
0.0126	18.0	8946	2.6948	0.7565	0.7021	0.7039	0.7030
0.0065	19.0	9443	2.6984	0.7565	0.7045	0.7064	0.7050
0.0103	20.0	9940	2.6968	0.7565	0.7045	0.7064	0.7050

Framework versions

• Transformers 4.20.1
• Pytorch 1.10.1+cu111
• Datasets 2.3.2
• Tokenizers 0.12.1