shihab17/bn-to-en-translation

How to use

You can use this model directly with a pipeline:

from transformers import AutoTokenizer, pipeline
tokenizer = AutoTokenizer.from_pretrained("shihab17/bn-to-en-translation")
model = AutoModelForSeq2SeqLM.from_pretrained("shihab17/bn-to-en-translation")

sentence = 'ম্যাচ শেষে পুরস্কার বিতরণের মঞ্চে তামিমের মুখে মোস্তাফিজের প্রশংসা শোনা গেল'

translator = pipeline("translation_en_to_bn", model=model, tokenizer=tokenizer)
output = translator(sentence)
print(output)

bengali-en-to-bn

This model is a fine-tuned version of Helsinki-NLP/opus-mt-bn-en on the kde4 dataset. It achieves the following results on the evaluation set:

• Loss: 1.6885
• Bleu: 50.9475
• Gen Len: 6.7043

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: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
1.8866	1.0	2047	1.6397	39.6617	8.0651
1.5769	2.0	4094	1.6160	33.0247	8.9865
1.3622	3.0	6141	1.6189	53.483	6.6037
1.2317	4.0	8188	1.6280	51.6882	6.762
1.1248	5.0	10235	1.6450	53.1619	6.5515
1.0297	6.0	12282	1.6587	52.3224	6.5905
0.9632	7.0	14329	1.6733	52.3362	6.5441
0.8831	8.0	16376	1.6802	49.3544	6.8272
0.8291	9.0	18423	1.6868	49.9486	6.792
0.8175	10.0	20470	1.6885	50.9475	6.7043

Framework versions

• Transformers 4.29.1
• Pytorch 2.0.0+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3