Modelpublic

TuringGame/gpt2-classifier

sourceHugging Facemitupdated 1y agoView on Hugging Face

0likes6downloads

Model Card

gpt2-classifier

This model is a fine-tuned version of openai-community/gpt2 on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 0.3035
—Accuracy: 0.8660
—F1: 0.7584

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: 2e-05
—trainbatchsize: 32
—evalbatchsize: 32
—seed: 42
—optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: linear
—num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0	0	4.5305	0.2536	0.4046
No log	0.2020	79	0.4243	0.8055	0.6281
No log	0.4041	158	0.3872	0.8213	0.6884
No log	0.6061	237	0.3544	0.8545	0.7363
No log	0.8082	316	0.3220	0.8588	0.7278
No log	1.0102	395	0.3321	0.8516	0.7392
No log	1.2123	474	0.3121	0.8746	0.7655
0.4499	1.4143	553	0.2934	0.8790	0.7717
0.4499	1.6164	632	0.3184	0.8602	0.7544
0.4499	1.8184	711	0.3035	0.8660	0.7584

Framework versions

—Transformers 4.54.0
—Pytorch 2.6.0+cu124
—Datasets 4.0.0
—Tokenizers 0.21.2