IsaacZhy/roberta-large-goemotions

pretrained_model

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

• Loss: 0.0568
• F1: 0.5868
• Roc Auc: 0.7616
• Accuracy: 0.4821

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: 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.1205	1.0	679	0.0865	0.5632	0.7347	0.4458
0.0859	2.0	1358	0.0829	0.5717	0.7378	0.4521
0.0727	3.0	2037	0.0827	0.5897	0.7523	0.4753
0.0629	4.0	2716	0.0857	0.5808	0.7535	0.4652
0.0568	5.0	3395	0.0904	0.5868	0.7616	0.4821
0.0423	6.0	4074	0.0989	0.5806	0.7682	0.4724
0.0344	7.0	4753	0.1079	0.5736	0.7657	0.4650
0.0296	8.0	5432	0.1158	0.5637	0.7649	0.4504
0.0206	9.0	6111	0.1200	0.5674	0.7689	0.4486
0.0177	10.0	6790	0.1240	0.5728	0.7737	0.4547

Framework versions

• Transformers 4.26.0
• Pytorch 1.13.1+cu116
• Datasets 2.9.0
• Tokenizers 0.13.2