scenario-NON-KD-PR-COPY-CDF-CL-D2_data-cl-cardiff_cl_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: 5.4350
• Accuracy: 0.4537
• F1: 0.4493

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	1.09	250	1.2630	0.4761	0.4753
0.8829	2.17	500	1.4325	0.4529	0.4468
0.8829	3.26	750	1.7675	0.4282	0.4077
0.538	4.35	1000	1.9149	0.4298	0.4134
0.538	5.43	1250	2.1555	0.4552	0.4471
0.2691	6.52	1500	2.7676	0.4637	0.4600
0.2691	7.61	1750	2.5963	0.4491	0.4431
0.1585	8.7	2000	3.0069	0.4437	0.4364
0.1585	9.78	2250	2.9894	0.4645	0.4621
0.1083	10.87	2500	3.6542	0.4537	0.4465
0.1083	11.96	2750	3.5648	0.4684	0.4695
0.0745	13.04	3000	3.7903	0.4761	0.4771
0.0745	14.13	3250	4.0663	0.4645	0.4623
0.0545	15.22	3500	4.1840	0.4830	0.4837
0.0545	16.3	3750	4.3967	0.4653	0.4640
0.04	17.39	4000	4.3967	0.4738	0.4714
0.04	18.48	4250	4.5309	0.4637	0.4600
0.0291	19.57	4500	4.5025	0.4715	0.4711
0.0291	20.65	4750	4.9500	0.4576	0.4554
0.0217	21.74	5000	5.0171	0.4498	0.4443
0.0217	22.83	5250	4.8470	0.4622	0.4618
0.018	23.91	5500	4.9003	0.4699	0.4698
0.018	25.0	5750	5.1935	0.4522	0.4491
0.0132	26.09	6000	5.1915	0.4630	0.4631
0.0132	27.17	6250	5.3879	0.4529	0.4478
0.0077	28.26	6500	5.2970	0.4637	0.4628
0.0077	29.35	6750	5.4350	0.4537	0.4493

Framework versions

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