366f9f8cd4edc9d36e82a315e65ff9b5

This model is a fine-tuned version of Helsinki-NLP/opus-mt-en-sv on the Helsinki-NLP/opus_books [en-fr] dataset. It achieves the following results on the evaluation set:

  • Loss: 1.2721
  • Data Size: 1.0
  • Epoch Runtime: 188.7785
  • Bleu: 10.6125

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: 5e-05
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 4
  • total_train_batch_size: 32
  • total_eval_batch_size: 32
  • optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: constant
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss Data Size Epoch Runtime Bleu
No log 0 0 7.3721 0 14.8549 0.1815
No log 1 3177 5.1308 0.0078 16.7322 0.2891
0.0825 2 6354 4.3046 0.0156 17.4756 0.6095
3.9539 3 9531 3.6182 0.0312 20.2874 1.1823
3.3002 4 12708 3.0049 0.0625 27.0439 2.2036
2.6552 5 15885 2.4568 0.125 37.6736 3.7941
2.1749 6 19062 2.0286 0.25 59.7931 5.3609
1.8168 7 22239 1.7084 0.5 102.4877 6.9440
1.5644 8.0 25416 1.4729 1.0 193.0732 8.3430
1.3822 9.0 28593 1.3671 1.0 194.2319 8.9963
1.291 10.0 31770 1.3147 1.0 194.8503 9.6040
1.1788 11.0 34947 1.2803 1.0 185.6993 9.9018
1.1122 12.0 38124 1.2550 1.0 186.5323 10.1864
1.0582 13.0 41301 1.2454 1.0 192.4918 10.2136
1.0012 14.0 44478 1.2408 1.0 189.0149 10.3864
0.9536 15.0 47655 1.2377 1.0 192.5333 10.4712
0.8965 16.0 50832 1.2365 1.0 192.1877 10.4450
0.8756 17.0 54009 1.2336 1.0 190.4363 10.4975
0.8481 18.0 57186 1.2461 1.0 187.4885 10.6915
0.8002 19.0 60363 1.2565 1.0 187.3171 10.6586
0.756 20.0 63540 1.2613 1.0 189.2071 10.6488
0.7247 21.0 66717 1.2721 1.0 188.7785 10.6125

Framework versions

  • Transformers 4.57.0
  • Pytorch 2.8.0+cu128
  • Datasets 4.2.0
  • Tokenizers 0.22.1
Downloads last month
-
Safetensors
Model size
0.2B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for contemmcm/366f9f8cd4edc9d36e82a315e65ff9b5

Finetuned
(39)
this model