gpt2_for_whole_train_result_2_4bce

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

• Loss: 0.0468
• 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
2.5829	6.8817	50	1.0807	0.64	0.7168
0.3023	13.7634	100	0.0960	0.9645	0.9675
0.0539	20.6452	150	0.0522	0.982	0.9833
0.0256	27.5269	200	0.0424	0.988	0.9888
0.0123	34.4086	250	0.0383	0.991	0.9916
0.0048	41.2903	300	0.0307	0.994	0.9944
0.0021	48.1720	350	0.0376	0.993	0.9935
0.0022	55.0538	400	0.0368	0.994	0.9944
0.0012	61.9355	450	0.0424	0.993	0.9935
0.0028	68.8172	500	0.0351	0.996	0.9962
0.0004	75.6989	550	0.0343	0.9965	0.9967
0.0003	82.5806	600	0.0370	0.996	0.9962
0.0002	89.4624	650	0.0382	0.9955	0.9958
0.0016	96.3441	700	0.0468	0.9945	0.9948

Framework versions

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