scenario-NON-KD-PR-COPY-CDF-EN-D2_data-en-cardiff_eng_only_alpha

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

• Loss: 4.0650
• Accuracy: 0.5150
• F1: 0.5153

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: 32
• eval_batch_size: 32
• seed: 1123
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	1.72	100	1.0715	0.4810	0.4544
No log	3.45	200	1.3491	0.5035	0.4980
No log	5.17	300	1.6657	0.5123	0.5129
No log	6.9	400	1.9846	0.5084	0.5079
0.4799	8.62	500	2.5547	0.5115	0.5117
0.4799	10.34	600	2.5722	0.5190	0.5193
0.4799	12.07	700	2.9243	0.5123	0.5118
0.4799	13.79	800	3.4404	0.5154	0.5143
0.4799	15.52	900	3.5740	0.5225	0.5210
0.0509	17.24	1000	3.6523	0.5176	0.5160
0.0509	18.97	1100	3.7591	0.5225	0.5231
0.0509	20.69	1200	3.7790	0.5137	0.5139
0.0509	22.41	1300	3.8779	0.5256	0.5271
0.0509	24.14	1400	3.8982	0.5251	0.5259
0.0107	25.86	1500	4.0718	0.5097	0.5109
0.0107	27.59	1600	3.9794	0.5234	0.5240
0.0107	29.31	1700	4.0650	0.5150	0.5153

Framework versions

• Transformers 4.33.3
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.13.3