tiny-gpt2-br / README.md
gweltou's picture
Upload tokenizer
a8d01fb verified
|
raw
history blame
2.27 kB
metadata
license: apache-2.0
tags:
  - generated_from_trainer
base_model: distilbert/distilgpt2
model-index:
  - name: tiny-gpt2-br
    results: []

tiny-gpt2-br

This model is a fine-tuned version of distilbert/distilgpt2 on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 3.6213

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: 0.001
  • train_batch_size: 16
  • 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: 500
  • training_steps: 19000

Training results

Training Loss Epoch Step Validation Loss
5.7274 0.21 1000 4.9308
4.7191 0.42 2000 4.5347
4.4579 0.63 3000 4.3432
4.2769 0.84 4000 4.1893
4.1086 1.05 5000 4.0861
3.9327 1.26 6000 3.9992
3.8812 1.47 7000 3.9216
3.8298 1.68 8000 3.8648
3.7785 1.89 9000 3.8126
3.6099 2.1 10000 3.7931
3.471 2.31 11000 3.7539
3.4651 2.52 12000 3.7141
3.4451 2.73 13000 3.6754
3.4251 2.94 14000 3.6327
3.1855 3.15 15000 3.6779
3.0962 3.36 16000 3.6757
3.0971 3.56 17000 3.6437
3.0816 3.77 18000 3.6287
3.0582 3.98 19000 3.6213

Framework versions

  • Transformers 4.39.1
  • Pytorch 2.0.1+cu117
  • Datasets 2.18.0
  • Tokenizers 0.15.2