scenario-KD-PR-CDF-EN-FROM-EN-D2_data-en-cardiff_eng_only_beta-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: 24.9305
• Accuracy: 0.4136
• F1: 0.4137

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: 6666
• 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	21.3876	0.3382	0.2625
No log	3.45	200	21.2399	0.3602	0.2578
No log	5.17	300	21.3743	0.3805	0.3422
No log	6.9	400	22.3531	0.3946	0.3622
21.9804	8.62	500	21.9340	0.4039	0.3992
21.9804	10.34	600	23.5425	0.3929	0.3581
21.9804	12.07	700	22.3943	0.4101	0.4063
21.9804	13.79	800	22.3814	0.3951	0.3902
21.9804	15.52	900	22.3635	0.4237	0.4194
16.5213	17.24	1000	23.4647	0.4043	0.4006
16.5213	18.97	1100	23.6028	0.4083	0.4086
16.5213	20.69	1200	24.0071	0.4158	0.4161
16.5213	22.41	1300	25.3331	0.4101	0.4000
16.5213	24.14	1400	24.8164	0.4004	0.4005
11.8769	25.86	1500	24.9076	0.4220	0.4211
11.8769	27.59	1600	25.0535	0.3907	0.3912
11.8769	29.31	1700	24.9305	0.4136	0.4137

Framework versions

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