scenario-NON-KD-PR-COPY-CDF-CL-D2_data-cl-cardiff_cl_only_gamma

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: 5.4088
• Accuracy: 0.4452
• F1: 0.4432

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: 11423
• 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	1.1861	0.4599	0.4461
0.8893	2.17	500	1.2483	0.4753	0.4682
0.8893	3.26	750	1.4640	0.4877	0.4872
0.5435	4.35	1000	1.9901	0.4529	0.4440
0.5435	5.43	1250	2.1858	0.4398	0.4357
0.2767	6.52	1500	2.2484	0.4653	0.4643
0.2767	7.61	1750	2.7287	0.4653	0.4642
0.1584	8.7	2000	2.7996	0.4637	0.4616
0.1584	9.78	2250	3.2599	0.4684	0.4684
0.1119	10.87	2500	3.7690	0.4344	0.4244
0.1119	11.96	2750	3.5578	0.4591	0.4584
0.0771	13.04	3000	3.9089	0.4483	0.4490
0.0771	14.13	3250	4.1349	0.4637	0.4587
0.054	15.22	3500	4.4418	0.4506	0.4435
0.054	16.3	3750	4.4987	0.4522	0.4511
0.04	17.39	4000	4.5234	0.4514	0.4511
0.04	18.48	4250	4.7455	0.4529	0.4517
0.0241	19.57	4500	5.0606	0.4329	0.4238
0.0241	20.65	4750	5.0820	0.4414	0.4394
0.0243	21.74	5000	5.2753	0.4360	0.4304
0.0243	22.83	5250	5.1224	0.4660	0.4666
0.0155	23.91	5500	5.2712	0.4437	0.4407
0.0155	25.0	5750	5.3846	0.4421	0.4393
0.0156	26.09	6000	5.4060	0.4398	0.4352
0.0156	27.17	6250	5.3914	0.4383	0.4344
0.0105	28.26	6500	5.3427	0.4421	0.4413
0.0105	29.35	6750	5.4088	0.4452	0.4432

Framework versions

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