scenario-KD-PR-CDF-EN-FROM-CL-D2_data-en-cardiff_eng_only_alpha-jason

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

• Loss: 15.6971
• Accuracy: 0.4105
• F1: 0.4078

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: 2222
• 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	12.5527	0.3779	0.3087
No log	3.45	200	12.9504	0.3757	0.3098
No log	5.17	300	12.9363	0.3884	0.3458
No log	6.9	400	13.3706	0.3951	0.3814
13.0874	8.62	500	13.5495	0.3849	0.3679
13.0874	10.34	600	14.1809	0.3739	0.3340
13.0874	12.07	700	13.9627	0.3814	0.3774
13.0874	13.79	800	14.6754	0.3907	0.3806
13.0874	15.52	900	14.1172	0.4136	0.4087
9.6129	17.24	1000	14.2157	0.3871	0.3752
9.6129	18.97	1100	14.6144	0.4030	0.3987
9.6129	20.69	1200	15.2951	0.3942	0.3947
9.6129	22.41	1300	15.0876	0.3933	0.3839
9.6129	24.14	1400	15.3136	0.4004	0.3987
7.2639	25.86	1500	15.5724	0.4061	0.4007
7.2639	27.59	1600	15.8778	0.3986	0.3984
7.2639	29.31	1700	15.6971	0.4105	0.4078

Framework versions

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