gpt2_for_whole_train_result_4_4bce

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

• Loss: 0.0432
• Accuracy: 0.9945
• F1: 0.9948

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: 0.0001
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 64
• total_train_batch_size: 4096
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
2.7651	6.8817	50	1.0424	0.6535	0.7106
0.3142	13.7634	100	0.0996	0.965	0.9679
0.0628	20.6452	150	0.0555	0.9815	0.9828
0.0255	27.5269	200	0.0373	0.99	0.9907
0.0096	34.4086	250	0.0361	0.992	0.9925
0.0071	41.2903	300	0.0313	0.9955	0.9958
0.0028	48.1720	350	0.0384	0.992	0.9925
0.0029	55.0538	400	0.0317	0.9945	0.9948
0.0015	61.9355	450	0.0343	0.996	0.9962
0.0005	68.8172	500	0.0389	0.996	0.9963
0.001	75.6989	550	0.0354	0.995	0.9953
0.0037	82.5806	600	0.0362	0.9955	0.9958
0.0003	89.4624	650	0.0348	0.996	0.9962
0.0003	96.3441	700	0.0432	0.9945	0.9948

Framework versions

• Transformers 4.40.0
• Pytorch 2.4.1+cu121
• Datasets 3.1.0
• Tokenizers 0.19.1