opus-mt-en-id-ccmatrix-lr-5

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

• Loss: 0.6093
• Bleu: 65.4357

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: 1e-05
• 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
0.678	1.0	28125	0.6238	63.4798
0.5765	2.0	56250	0.6036	64.1162
0.5375	3.0	84375	0.5953	64.4048
0.5098	4.0	112500	0.5887	64.7167
0.4879	5.0	140625	0.5862	64.8577
0.4696	6.0	168750	0.5855	64.9321
0.4539	7.0	196875	0.5835	64.9806
0.4401	8.0	225000	0.5875	65.1012
0.4279	9.0	253125	0.5864	65.1125
0.4168	10.0	281250	0.5870	65.1402
0.4069	11.0	309375	0.5905	65.2012
0.3977	12.0	337500	0.5905	65.3486
0.3895	13.0	365625	0.5944	65.3406
0.3817	14.0	393750	0.5957	65.3218
0.3749	15.0	421875	0.5978	65.3269
0.3683	16.0	450000	0.5989	65.355
0.3624	17.0	478125	0.6009	65.4288
0.3573	18.0	506250	0.6007	65.4001
0.3525	19.0	534375	0.6035	65.4446
0.3484	20.0	562500	0.6054	65.3843
0.3448	21.0	590625	0.6060	65.392
0.3415	22.0	618750	0.6078	65.4052
0.3388	23.0	646875	0.6082	65.3898
0.3365	24.0	675000	0.6089	65.4171
0.3349	25.0	703125	0.6093	65.4357

Framework versions

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