bangla_idiom_translation_v1

This model is a fine-tuned version of csebuetnlp/banglat5_nmt_bn_en on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3032
• Rouge1: 85.3143
• Rouge2: 73.3973
• Rougel: 82.4163
• Rougelsum: 82.4211
• Gen Len: 21.2116

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 5000
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
0.5372	1.0	3066	0.3419	85.3363	73.5641	82.4285	82.4387	21.2325
0.4708	2.0	6132	0.3181	85.2649	73.4772	82.47	82.484	21.1232
0.437	3.0	9198	0.3076	85.3474	73.5827	82.5201	82.5292	21.1559
0.416	4.0	12264	0.3040	85.372	73.4868	82.4993	82.4963	21.1842
0.4079	5.0	15330	0.3032	85.3143	73.3973	82.4163	82.4211	21.2116

Framework versions

• Transformers 4.37.2
• Pytorch 2.1.0+cu121
• Datasets 2.17.1
• Tokenizers 0.15.2