SentimentT2_GPT2 / README.md
LogischeIP's picture
End of training
d2ad407 verified
|
raw
history blame
2.04 kB
metadata
license: mit
base_model: gpt2
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
model-index:
  - name: SentimentT2_GPT2
    results: []

SentimentT2_GPT2

This model is a fine-tuned version of gpt2 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.9470
  • Accuracy: 0.8644
  • F1: 0.8728
  • Auc Roc: 0.9185
  • Log Loss: 0.9470

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: 1e-05
  • train_batch_size: 1
  • eval_batch_size: 1
  • seed: 42
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 2
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 500
  • num_epochs: 6
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy F1 Auc Roc Log Loss
1.0534 1.0 1618 0.6479 0.8694 0.8696 0.9298 0.6479
0.6971 2.0 3236 1.0859 0.8371 0.8581 0.9236 1.0859
0.5832 3.0 4854 0.9261 0.8495 0.8672 0.9255 0.9261
0.4402 4.0 6472 0.8507 0.8719 0.8804 0.9251 0.8507
0.3475 5.0 8090 0.9284 0.8657 0.8735 0.9198 0.9283
0.2985 6.0 9708 0.9470 0.8644 0.8728 0.9185 0.9470

Framework versions

  • Transformers 4.35.2
  • Pytorch 2.1.0+cu121
  • Datasets 2.16.1
  • Tokenizers 0.15.1