th-nuernberg/gbert-base-german-counseling-gecco

gecco-german-counseling-gbert-base

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

• Loss: 1.2480
• Accuracy: 0.7194
• F1: 0.5062

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: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 16

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
3.439	1.0	20	3.1181	0.2484	0.0474
2.9431	2.0	40	2.6841	0.3935	0.1637
2.5477	3.0	60	2.3120	0.5387	0.2802
2.1823	4.0	80	2.0526	0.5935	0.3138
1.8786	5.0	100	1.8242	0.6387	0.3541
1.6267	6.0	120	1.6720	0.6548	0.3682
1.4447	7.0	140	1.5538	0.6645	0.3718
1.2734	8.0	160	1.4655	0.6710	0.3801
1.1099	9.0	180	1.4040	0.6935	0.4202
1.0766	10.0	200	1.3541	0.6903	0.4330
0.913	11.0	220	1.3078	0.6968	0.4629
0.8557	12.0	240	1.2879	0.7161	0.5000
0.8477	13.0	260	1.2772	0.7097	0.4946
0.7412	14.0	280	1.2598	0.7161	0.5042
0.7341	15.0	300	1.2484	0.7194	0.5069
0.7029	16.0	320	1.2480	0.7194	0.5062

Framework versions

• Transformers 4.35.1
• Pytorch 1.10.1+cu111
• Datasets 2.14.7
• Tokenizers 0.14.1