opus-mt-finetuned-en-to-fr

This model is a fine-tuned version of DevAibest/opus-mt-finetuned-en-to-fr on the opus_books dataset. It achieves the following results on the evaluation set:

• Loss: 0.2610
• Bleu: 69.4791
• Gen Len: 32.7639

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

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
0.4587	1.0	6355	0.2063	76.6735	32.6691
0.426	2.0	12710	0.2238	74.291	32.7606
0.3935	3.0	19065	0.2385	72.4894	32.8613
0.3723	4.0	25420	0.2480	71.2701	32.635
0.3492	5.0	31775	0.2532	70.6417	32.7086
0.3315	6.0	38130	0.2588	69.8185	32.7439
0.3057	7.0	44485	0.2598	69.7941	32.7479
0.2906	8.0	50840	0.2621	69.5144	32.7624
0.283	9.0	57195	0.2619	69.27	32.7373
0.2675	10.0	63550	0.2610	69.4791	32.7639

Framework versions

• Transformers 4.33.0
• Pytorch 2.0.0
• Datasets 2.1.0
• Tokenizers 0.13.3