a317f3b8086eff826b3405960696df53

This model is a fine-tuned version of google/mt5-base on the Helsinki-NLP/opus_books [en-es] dataset. It achieves the following results on the evaluation set:

  • Loss: 1.6287
  • Data Size: 1.0
  • Epoch Runtime: 487.6002
  • Bleu: 9.3762

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 17.5906 0 38.7980 0.0066
No log 1 2336 11.2078 0.0078 42.4555 0.0090
0.2148 2 4672 8.1461 0.0156 46.5754 0.0132
0.2317 3 7008 3.5696 0.0312 54.7314 0.2977
3.4847 4 9344 2.5315 0.0625 69.8693 5.4395
3.1358 5 11680 2.3672 0.125 96.1808 5.3783
2.8317 6 14016 2.2252 0.25 149.3766 5.0234
2.6068 7 16352 2.0973 0.5 271.2784 6.5774
2.3911 8.0 18688 1.9709 1.0 507.0318 7.6639
2.2176 9.0 21024 1.8925 1.0 497.3987 8.1176
2.1229 10.0 23360 1.8367 1.0 483.7592 8.5419
2.0508 11.0 25696 1.7997 1.0 487.4110 8.5801
1.9534 12.0 28032 1.7722 1.0 489.1477 8.8942
1.8972 13.0 30368 1.7458 1.0 490.0701 8.6582
1.8732 14.0 32704 1.7242 1.0 490.0359 8.7500
1.8086 15.0 35040 1.7088 1.0 488.0000 9.0016
1.8172 16.0 37376 1.6885 1.0 501.7522 8.9066
1.7176 17.0 39712 1.6740 1.0 488.9021 9.0096
1.7279 18.0 42048 1.6703 1.0 514.9460 9.1182
1.687 19.0 44384 1.6622 1.0 509.9660 9.1538
1.6463 20.0 46720 1.6530 1.0 509.7470 9.2026
1.6062 21.0 49056 1.6472 1.0 511.0166 9.0921
1.5816 22.0 51392 1.6488 1.0 515.4553 8.9717
1.546 23.0 53728 1.6392 1.0 513.8395 9.1425
1.4913 24.0 56064 1.6401 1.0 509.7644 9.1420
1.5082 25.0 58400 1.6299 1.0 509.5740 9.0798
1.4748 26.0 60736 1.6231 1.0 510.2230 9.2542
1.4534 27.0 63072 1.6280 1.0 504.9012 9.1842
1.4294 28.0 65408 1.6194 1.0 502.2083 9.2138
1.4022 29.0 67744 1.6272 1.0 483.3266 9.3731
1.395 30.0 70080 1.6300 1.0 486.9168 9.4220
1.3651 31.0 72416 1.6317 1.0 485.8095 9.3675
1.3667 32.0 74752 1.6287 1.0 487.6002 9.3762

Framework versions

  • Transformers 4.57.0
  • Pytorch 2.8.0+cu128
  • Datasets 4.2.0
  • Tokenizers 0.22.1
Downloads last month
1
Safetensors
Model size
1.0B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for contemmcm/a317f3b8086eff826b3405960696df53

Base model

google/mt5-base
Finetuned
(301)
this model