indictrans2-indic-indic-dist-320M-finetuned-eng_Ltn-to-mar_Deva

This model is a fine-tuned version of ai4bharat/indictrans2-indic-indic-dist-320M on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4547
• Bleu: 10.3309
• Gen Len: 20.9925

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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
0.5955	0.1703	2000	0.5251	9.1406	20.9926
0.5318	0.3407	4000	0.4925	9.7358	20.9923
0.5113	0.5110	6000	0.4738	10.0388	20.9925
0.5145	0.6814	8000	0.4618	10.2453	20.9925
0.5084	0.8517	10000	0.4547	10.3309	20.9925

Framework versions

• Transformers 4.53.2
• Pytorch 2.9.0+cu128
• Datasets 2.21.0
• Tokenizers 0.21.4