scenario_4

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: 0.1764
• Accuracy: 0.9704
• F1: 0.9704
• Precision: 0.9710
• Recall: 0.9704
• Accuracy Label Test: 0.9879
• Accuracy Label Train: 0.9536

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall	Accuracy Label Test	Accuracy Label Train
0.5515	0.1579	100	0.5194	0.7579	0.7415	0.8352	0.7579	0.5070	0.9992
0.2205	0.3157	200	0.2537	0.9300	0.9298	0.9361	0.9300	0.9883	0.8739
0.1106	0.4736	300	0.3450	0.9129	0.9124	0.9248	0.9129	0.9960	0.8329
0.0384	0.6314	400	0.1408	0.9683	0.9683	0.9687	0.9683	0.9835	0.9536
0.0631	0.7893	500	0.1517	0.9631	0.9631	0.9645	0.9631	0.9895	0.9377
0.0276	0.9471	600	0.3649	0.9387	0.9386	0.9444	0.9387	0.9948	0.8847
0.0245	1.1050	700	0.1339	0.9702	0.9702	0.9702	0.9702	0.9727	0.9679
0.0519	1.2628	800	0.4945	0.9186	0.9182	0.9299	0.9186	0.9992	0.8410
0.02	1.4207	900	0.2637	0.9549	0.9548	0.9580	0.9549	0.9960	0.9153
0.0325	1.5785	1000	0.1165	0.9708	0.9708	0.9712	0.9708	0.9851	0.9571
0.016	1.7364	1100	0.1007	0.9692	0.9692	0.9697	0.9692	0.9530	0.9849
0.0068	1.8942	1200	0.1679	0.9690	0.9690	0.9697	0.9690	0.9871	0.9516
0.0042	2.0521	1300	0.1182	0.9734	0.9734	0.9734	0.9734	0.9723	0.9745
0.0005	2.2099	1400	0.1432	0.9730	0.9730	0.9731	0.9730	0.9799	0.9663
0.0182	2.3678	1500	0.1460	0.9718	0.9718	0.9723	0.9718	0.9871	0.9571
0.0004	2.5257	1600	0.1383	0.9732	0.9732	0.9734	0.9732	0.9843	0.9625
0.0003	2.6835	1700	0.1381	0.9744	0.9744	0.9745	0.9744	0.9831	0.9660
0.0002	2.8414	1800	0.1599	0.9724	0.9724	0.9728	0.9724	0.9863	0.9590

Framework versions

• Transformers 4.44.0
• Pytorch 2.3.1
• Datasets 2.20.0
• Tokenizers 0.19.1