scenario-NON-KD-SCR-COPY-CDF-EN-D2_data-en-cardiff_eng_only44

This model is a fine-tuned version of microsoft/mdeberta-v3-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 5.0909
• Accuracy: 0.3682
• F1: 0.3646

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: 44
• 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	1.7241	100	1.1864	0.3651	0.3644
No log	3.4483	200	2.2167	0.3580	0.3355
No log	5.1724	300	2.7873	0.3673	0.3618
No log	6.8966	400	3.5495	0.3739	0.3714
0.4197	8.6207	500	4.2289	0.3770	0.3708
0.4197	10.3448	600	4.6578	0.3638	0.3605
0.4197	12.0690	700	4.5844	0.3690	0.3671
0.4197	13.7931	800	4.8103	0.3616	0.3462
0.4197	15.5172	900	4.8621	0.3616	0.3545
0.017	17.2414	1000	4.9407	0.3708	0.3657
0.017	18.9655	1100	5.0334	0.3699	0.3696
0.017	20.6897	1200	4.9701	0.3686	0.3676
0.017	22.4138	1300	4.9793	0.3686	0.3654
0.017	24.1379	1400	5.0299	0.3668	0.3600
0.0076	25.8621	1500	5.1558	0.3616	0.3544
0.0076	27.5862	1600	5.0915	0.3668	0.3622
0.0076	29.3103	1700	5.0909	0.3682	0.3646

Framework versions

• Transformers 4.44.2
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.19.1