sentiment_roberta_large_with_diary

This model is a fine-tuned version of klue/roberta-large on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5671
• Micro f1 score: 80.0000
• Auprc: 77.0282
• Accuracy: 0.8

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: 1e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 1.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Micro f1 score	Auprc	Accuracy
1.6198	0.13	100	1.3872	48.9362	55.5743	0.4894
0.6603	0.26	200	0.9249	65.9574	62.8759	0.6596
0.5387	0.4	300	0.7262	73.1915	71.1936	0.7319
0.4801	0.53	400	0.6623	74.0426	68.8606	0.7404
0.4597	0.66	500	0.6092	76.1702	75.7346	0.7617
0.4217	0.79	600	0.5929	78.7234	76.8709	0.7872
0.4148	0.93	700	0.5671	80.0000	77.0282	0.8

Framework versions

• Transformers 4.25.1
• Pytorch 1.13.0+cu117
• Datasets 2.7.1
• Tokenizers 0.13.2