pogny-128-0.1

This model is a fine-tuned version of klue/roberta-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.6873
• Accuracy: 0.4376
• F1: 0.2665

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: 0.1
• train_batch_size: 128
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
42.1657	1.0	603	24.7694	0.4376	0.2665
38.0547	2.0	1206	34.6085	0.2545	0.1032
34.4507	3.0	1809	47.2521	0.0240	0.0011
33.6528	4.0	2412	23.6900	0.0702	0.0092
29.4715	5.0	3015	13.6478	0.0107	0.0002
26.073	6.0	3618	29.8545	0.4376	0.2665
23.0398	7.0	4221	17.9423	0.4376	0.2665
20.1565	8.0	4824	18.5313	0.0702	0.0092
17.9295	9.0	5427	16.2984	0.4376	0.2665
12.4633	10.0	6030	11.9847	0.2545	0.1032
9.5341	11.0	6633	10.4590	0.4376	0.2665
8.1157	12.0	7236	3.1051	0.4376	0.2665
5.1415	13.0	7839	3.4676	0.0702	0.0092
3.3891	14.0	8442	2.0901	0.4376	0.2665
1.854	15.0	9045	1.6873	0.4376	0.2665

Framework versions

• Transformers 4.34.1
• Pytorch 2.1.0a0+b5021ba
• Datasets 2.6.2
• Tokenizers 0.14.1