Af-En_update

This model is a fine-tuned version of Helsinki-NLP/opus-mt-af-en on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.7197
• Bleu: 55.3346

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: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu
1.3745	1.0	2553	1.7537	51.9270
1.0462	2.0	5106	1.6305	53.9359
0.896	3.0	7659	1.6216	54.3049
0.7824	4.0	10212	1.6108	54.9902
0.6974	5.0	12765	1.6183	55.0265
0.643	6.0	15318	1.6207	55.4137
0.5635	7.0	17871	1.6276	55.1335
0.5141	8.0	20424	1.6498	55.2215
0.4681	9.0	22977	1.6678	55.2000
0.4304	10.0	25530	1.6797	55.2748
0.425	11.0	28083	1.7004	55.0478
0.398	12.0	30636	1.7013	55.3591
0.3759	13.0	33189	1.7082	55.3225
0.3681	14.0	35742	1.7151	55.1793
0.3571	15.0	38295	1.7197	55.2729

Framework versions

• Transformers 4.21.0
• Pytorch 1.12.0+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1