Wenzhong-GPT2-110M-THUCNews

This model is a fine-tuned version of IDEA-CCNL/Wenzhong-GPT2-110M on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.2403

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: 2e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss
1.295	1.0	2500	1.1898
1.1713	2.0	5000	1.1759
1.1218	3.0	7500	1.1765
1.0846	4.0	10000	1.1780
1.0536	5.0	12500	1.1840
1.0264	6.0	15000	1.1904
1.0025	7.0	17500	1.1969
0.9812	8.0	20000	1.2026
0.9623	9.0	22500	1.2091
0.9458	10.0	25000	1.2167
0.9312	11.0	27500	1.2220
0.9181	12.0	30000	1.2292
0.9073	13.0	32500	1.2349
0.8985	14.0	35000	1.2382
0.8921	15.0	37500	1.2403

Framework versions

• Transformers 4.28.1
• Pytorch 1.13.1+cu117
• Datasets 2.11.0
• Tokenizers 0.13.2