opus-mt-en-id-ccmatrix-lr-3

This model was trained from scratch on the ccmatrix dataset. It achieves the following results on the evaluation set:

• Loss: 7.7571
• Bleu: 0.0003

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: 0.001
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 4000
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu
4.0796	1.0	28125	7.8472	0.0047
4.6004	2.0	56250	8.8066	0.0022
4.4685	3.0	84375	6.4351	0.002
4.4315	4.0	112500	6.4291	0.0091
4.338	5.0	140625	6.7652	0.2393
4.2657	6.0	168750	6.7140	0.0087
4.2072	7.0	196875	6.7277	0.0196
5.253	8.0	225000	7.0848	0.0028
4.8823	9.0	253125	7.7789	0.0012
4.7303	10.0	281250	7.3183	0.0432
4.6488	11.0	309375	7.0976	0.0961
4.581	12.0	337500	7.1433	0.0748
4.5141	13.0	365625	6.9517	0.0225
4.4603	14.0	393750	7.2024	0.0207
4.4246	15.0	421875	6.8771	0.0646
4.3863	16.0	450000	6.9140	0.012
4.337	17.0	478125	6.7126	0.0175
4.2944	18.0	506250	6.9462	0.0125
4.2578	19.0	534375	6.9446	0.0064
4.2195	20.0	562500	7.5402	0.0004
4.1806	21.0	590625	7.2321	0.0105
4.1421	22.0	618750	7.5888	0.0094
4.1026	23.0	646875	7.5273	0.0097
4.0632	24.0	675000	7.6161	0.0002
4.0269	25.0	703125	7.7571	0.0003

Framework versions

• Transformers 4.26.1
• Pytorch 2.0.0
• Datasets 2.10.1
• Tokenizers 0.11.0