stage2Only_scaffoldGPT_new

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

• Loss: 0.3643
• Accuracy: 0.8806

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: 5e-05
• train_batch_size: 24
• eval_batch_size: 24
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• total_train_batch_size: 192
• total_eval_batch_size: 192
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.4661	1.0	6015	0.4390	0.8615
0.4302	2.0	12030	0.4107	0.8688
0.4145	3.0	18045	0.3979	0.8721
0.405	4.0	24060	0.3902	0.8740
0.3985	5.0	30075	0.3850	0.8753
0.3934	6.0	36090	0.3808	0.8764
0.3896	7.0	42105	0.3778	0.8772
0.3864	8.0	48120	0.3752	0.8778
0.3839	9.0	54135	0.3731	0.8784
0.3817	10.0	60150	0.3714	0.8788
0.3799	11.0	66165	0.3700	0.8792
0.3784	12.0	72180	0.3687	0.8795
0.3771	13.0	78195	0.3677	0.8798
0.376	14.0	84210	0.3668	0.8800
0.3751	15.0	90225	0.3661	0.8802
0.3743	16.0	96240	0.3655	0.8803
0.3737	17.0	102255	0.3650	0.8804
0.3732	18.0	108270	0.3646	0.8805
0.3729	19.0	114285	0.3644	0.8806
0.3726	20.0	120300	0.3643	0.8806

Framework versions

• Transformers 4.36.0.dev0
• Pytorch 2.1.1+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0