gpt2-kl_01_03-hs_cn

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

• Loss: 0.5545

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: 8
• eval_batch_size: 4
• seed: 21
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss
73.5393	0.02	10	69.5854
46.1033	0.04	20	32.9856
13.5637	0.06	30	10.6461
5.8312	0.08	40	4.4538
2.9594	0.1	50	1.9261
1.4588	0.12	60	1.0456
1.0977	0.14	70	0.8724
0.8915	0.16	80	0.7484
0.7239	0.18	90	0.6722
0.7148	0.2	100	0.7310
0.636	0.22	110	0.6278
0.6422	0.24	120	0.5887
0.6416	0.26	130	0.5791
0.6548	0.28	140	0.5789
0.6148	0.3	150	0.5742
0.6129	0.32	160	0.5715
0.5267	0.34	170	0.5629
0.6428	0.36	180	0.5614
0.633	0.38	190	0.5590
0.5816	0.4	200	0.5574
0.697	0.42	210	0.5546
0.649	0.44	220	0.5521
0.5442	0.46	230	0.5541
0.5571	0.48	240	0.5528
0.5572	0.5	250	0.5545

Framework versions

• Transformers 4.29.0.dev0
• Pytorch 1.12.0a0+bd13bc6
• Datasets 2.12.0
• Tokenizers 0.13.3