scenario-NON-KD-SCR-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: 5.6703
• Accuracy: 0.3619
• F1: 0.3506

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	1.3223	0.3596	0.3307
1.0598	2.17	500	1.3304	0.3727	0.3500
1.0598	3.26	750	2.0230	0.3673	0.3479
0.5145	4.35	1000	2.3446	0.3619	0.3495
0.5145	5.43	1250	3.5365	0.3495	0.3183
0.1727	6.52	1500	3.4431	0.3704	0.3671
0.1727	7.61	1750	4.3859	0.3434	0.3093
0.0869	8.7	2000	4.2371	0.3634	0.3501
0.0869	9.78	2250	4.6911	0.3534	0.3110
0.0549	10.87	2500	4.8692	0.3596	0.3409
0.0549	11.96	2750	4.5994	0.3588	0.3576
0.0302	13.04	3000	4.8291	0.3542	0.3468
0.0302	14.13	3250	5.2840	0.3472	0.3456
0.0184	15.22	3500	5.3672	0.3704	0.3519
0.0184	16.3	3750	5.6098	0.3596	0.3245
0.0211	17.39	4000	5.3263	0.3580	0.3373
0.0211	18.48	4250	5.5020	0.3619	0.3477
0.0097	19.57	4500	5.4448	0.3457	0.3384
0.0097	20.65	4750	5.5918	0.3642	0.3416
0.0077	21.74	5000	5.5070	0.3542	0.3314
0.0077	22.83	5250	5.5629	0.3657	0.3641
0.0038	23.91	5500	5.6918	0.3580	0.3533
0.0038	25.0	5750	5.6161	0.3588	0.3568
0.0033	26.09	6000	5.5866	0.3603	0.3494
0.0033	27.17	6250	5.6274	0.3657	0.3425
0.0011	28.26	6500	5.6469	0.3619	0.3494
0.0011	29.35	6750	5.6703	0.3619	0.3506

Framework versions

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