25c538fe026155262f55dc59b28f7026

This model is a fine-tuned version of google-t5/t5-large on the Helsinki-NLP/opus_books [fr-it] dataset. It achieves the following results on the evaluation set:

  • Loss: 1.6511
  • Data Size: 1.0
  • Epoch Runtime: 166.5946
  • Bleu: 8.6492

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 3.8446 0 12.2101 0.9155
No log 1 367 3.1691 0.0078 12.5952 1.5673
No log 2 734 2.7153 0.0156 15.0712 2.3566
No log 3 1101 2.5800 0.0312 18.8708 3.1036
No log 4 1468 2.4667 0.0625 23.7216 3.9287
0.1447 5 1835 2.3576 0.125 32.3061 5.2866
2.5302 6 2202 2.2269 0.25 51.9858 5.8870
2.3476 7 2569 2.0839 0.5 92.2038 6.4442
2.1358 8.0 2936 1.9428 1.0 164.3113 7.4477
1.9714 9.0 3303 1.8450 1.0 163.7945 7.6499
1.8504 10.0 3670 1.7911 1.0 163.1514 8.1984
1.7587 11.0 4037 1.7502 1.0 161.9175 8.3156
1.7041 12.0 4404 1.7103 1.0 167.2038 8.7358
1.6508 13.0 4771 1.6904 1.0 165.9298 8.7357
1.5491 14.0 5138 1.6669 1.0 159.8515 8.9149
1.5103 15.0 5505 1.6520 1.0 168.0792 8.7544
1.4479 16.0 5872 1.6488 1.0 164.1649 8.7317
1.4087 17.0 6239 1.6473 1.0 162.4921 8.7791
1.3597 18.0 6606 1.6356 1.0 166.7319 8.8385
1.3318 19.0 6973 1.6356 1.0 165.1773 8.7970
1.2954 20.0 7340 1.6262 1.0 164.0803 8.8980
1.2271 21.0 7707 1.6307 1.0 163.6734 8.8361
1.2054 22.0 8074 1.6341 1.0 165.0612 8.7217
1.1556 23.0 8441 1.6348 1.0 170.7119 8.6424
1.1252 24.0 8808 1.6511 1.0 166.5946 8.6492

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
0.8B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for contemmcm/25c538fe026155262f55dc59b28f7026

Finetuned
(171)
this model