xiulinyang/GPT2_10M_50000

10M_50000_41

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.7071
• Accuracy: 0.3920

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: 0.0006
• train_batch_size: 32
• eval_batch_size: 32
• seed: 41
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 2000
• num_epochs: 20.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	971	4.4042	0.3295
1.2033	2.0	1942	4.1879	0.3475
0.9928	3.0	2913	4.0195	0.3588
0.938	4.0	3884	3.9204	0.3674
0.895	5.0	4855	3.8516	0.3726
0.8675	6.0	5826	3.8064	0.3757
0.8468	7.0	6797	3.7914	0.3783
0.8258	8.0	7768	3.7569	0.3812
0.8125	9.0	8739	3.7413	0.3831
0.7971	10.0	9710	3.7261	0.3846
0.7838	11.0	10681	3.7196	0.3855
0.7706	12.0	11652	3.7084	0.3872
0.7576	13.0	12623	3.7034	0.3881
0.7458	14.0	13594	3.7016	0.3890
0.733	15.0	14565	3.7040	0.3895
0.721	16.0	15536	3.7076	0.3899
0.7109	17.0	16507	3.7013	0.3908
0.6999	18.0	17478	3.7044	0.3913
0.6897	19.0	18449	3.7064	0.3916
0.6796	20.0	19420	3.7071	0.3920

Framework versions

• Transformers 4.51.3
• Pytorch 2.5.1+cu121
• Datasets 3.6.0
• Tokenizers 0.21.1