scenario-KD-PR-CDF-EN-FROM-EN-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: 25.0852
• Accuracy: 0.3990
• F1: 0.3923

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	21.4007	0.3351	0.1783
No log	3.45	200	21.2647	0.3602	0.2851
No log	5.17	300	21.7287	0.3836	0.3173
No log	6.9	400	22.0257	0.3779	0.3535
22.4551	8.62	500	21.9464	0.3761	0.3596
22.4551	10.34	600	22.5105	0.3849	0.3792
22.4551	12.07	700	23.3230	0.3743	0.3466
22.4551	13.79	800	23.3751	0.3739	0.3490
22.4551	15.52	900	22.6748	0.3990	0.3990
17.2022	17.24	1000	23.5182	0.3902	0.3782
17.2022	18.97	1100	23.4775	0.3836	0.3785
17.2022	20.69	1200	23.9951	0.3893	0.3789
17.2022	22.41	1300	24.6740	0.3770	0.3500
17.2022	24.14	1400	25.2132	0.3593	0.3287
13.0792	25.86	1500	24.6053	0.3924	0.3825
13.0792	27.59	1600	25.0044	0.3854	0.3766
13.0792	29.31	1700	25.0852	0.3990	0.3923

Framework versions

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