scenario-TCR_data-en-cardiff_eng_only2

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: 3.6404
• Accuracy: 0.5825
• F1: 0.5841

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: 66
• 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	1.03	60	1.0662	0.4643	0.4071
No log	2.07	120	0.9869	0.5626	0.5572
No log	3.1	180	1.1139	0.5520	0.5458
No log	4.14	240	1.1885	0.5648	0.5672
No log	5.17	300	1.4845	0.5441	0.5460
No log	6.21	360	1.5402	0.5679	0.5708
No log	7.24	420	2.0851	0.5379	0.5358
No log	8.28	480	2.2943	0.5485	0.5498
0.5216	9.31	540	2.1485	0.5661	0.5643
0.5216	10.34	600	2.4773	0.5670	0.5677
0.5216	11.38	660	2.6976	0.5635	0.5651
0.5216	12.41	720	3.1052	0.5498	0.5477
0.5216	13.45	780	2.9635	0.5661	0.5668
0.5216	14.48	840	3.1581	0.5595	0.5589
0.5216	15.52	900	3.1375	0.5551	0.5579
0.5216	16.55	960	3.2300	0.5705	0.5726
0.0612	17.59	1020	3.2335	0.5714	0.5728
0.0612	18.62	1080	3.3244	0.5776	0.5781
0.0612	19.66	1140	3.3730	0.5754	0.5749
0.0612	20.69	1200	3.3890	0.5838	0.5834
0.0612	21.72	1260	3.6022	0.5723	0.5736
0.0612	22.76	1320	3.7443	0.5578	0.5577
0.0612	23.79	1380	3.6515	0.5719	0.5722
0.0612	24.83	1440	3.5110	0.5798	0.5786
0.0092	25.86	1500	3.6586	0.5741	0.5761
0.0092	26.9	1560	3.5719	0.5838	0.5849
0.0092	27.93	1620	3.6252	0.5798	0.5817
0.0092	28.97	1680	3.6249	0.5825	0.5839
0.0092	30.0	1740	3.6404	0.5825	0.5841

Framework versions

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