FineTuned-hindi-to-english-V8

This model is a fine-tuned version of Helsinki-NLP/opus-mt-hi-en on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.4699
• Rouge1: 77.4859
• Rouge2: 54.5463
• Rougel: 70.7586
• Rougelsum: 72.9591
• Gen Len: 80.2678

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: 14
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
3.2542	1.0	750	2.4128	66.5333	38.7744	58.3682	61.5066	77.9543
2.2929	2.0	1500	2.0835	70.0324	43.5622	62.3036	65.2299	79.0161
2.0351	3.0	2250	1.8982	72.069	46.4931	64.6929	67.3732	80.1859
1.7712	4.0	3000	1.7791	73.5869	48.6666	66.3649	68.8242	79.9375
1.6431	5.0	3750	1.6935	74.5655	50.0947	67.4126	69.8943	80.2959
1.4994	6.0	4500	1.6316	75.3132	51.2895	68.3122	70.7019	80.1347
1.4154	7.0	5250	1.5856	75.9486	52.1892	69.0828	71.4087	80.0847
1.333	8.0	6000	1.5516	76.2849	52.724	69.4646	71.7547	80.0536
1.2793	9.0	6750	1.5234	76.7711	53.4504	69.9626	72.215	80.2093
1.2219	10.0	7500	1.5043	76.9627	53.7785	70.2119	72.4254	80.1563
1.196	11.0	8250	1.4870	77.2081	54.1476	70.4764	72.681	80.4921
1.1612	12.0	9000	1.4771	77.4156	54.445	70.7305	72.9198	80.2194
1.1388	13.0	9750	1.4715	77.4179	54.4518	70.6541	72.8583	80.2815
1.1277	14.0	10500	1.4699	77.4859	54.5463	70.7586	72.9591	80.2678

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 2.10.0
• Tokenizers 0.13.2