license base_model tags model-index
apache-2.0
distilbert-base-uncased
generated_from_trainer
name results
results

results

This model is a fine-tuned version of distilbert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 356.9242

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: 1
  • eval_batch_size: 1
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 500
  • num_epochs: 20

Training results

Training Loss Epoch Step Validation Loss
1033.4562 1.0 401 1658.2954
297.1692 2.0 802 841.6623
264.7683 3.0 1203 892.8900
381.3767 4.0 1604 835.1779
194.2236 5.0 2005 901.5334
180.7703 6.0 2406 552.7529
35.9909 7.0 2807 352.0545
184.5064 8.0 3208 356.3049
151.6956 9.0 3609 452.0968
58.0247 10.0 4010 390.8174
41.0552 11.0 4411 473.5340
19.0339 12.0 4812 472.2711
31.2277 13.0 5213 396.7701
38.499 14.0 5614 271.9209
48.1305 15.0 6015 334.2394
16.8594 16.0 6416 314.0363
9.0616 17.0 6817 351.6679
6.1224 18.0 7218 398.0360
16.5143 19.0 7619 372.0302
5.5345 20.0 8020 356.9242

Framework versions

  • Transformers 4.41.2
  • Pytorch 2.1.2
  • Datasets 2.19.2
  • Tokenizers 0.19.1