bertweet-base_3epoch10

This model is a fine-tuned version of vinai/bertweet-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.7005
• Accuracy: 0.7392
• F1: 0.4217
• Precision: 0.5789
• Recall: 0.3317
• Precision Sarcastic: 0.5789
• Recall Sarcastic: 0.3317
• F1 Sarcastic: 0.4217

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
• 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	Accuracy	F1	Precision	Recall	Precision Sarcastic	Recall Sarcastic	F1 Sarcastic
No log	1.0	174	1.3971	0.7493	0.4	0.6374	0.2915	0.6374	0.2915	0.4
No log	2.0	348	1.0371	0.7334	0.3854	0.5686	0.2915	0.5686	0.2915	0.3854
0.0617	3.0	522	1.6060	0.7147	0.4277	0.5034	0.3719	0.5034	0.3719	0.4277
0.0617	4.0	696	1.3603	0.7464	0.4172	0.6117	0.3166	0.6117	0.3166	0.4172
0.0617	5.0	870	1.5872	0.7478	0.4373	0.6071	0.3417	0.6071	0.3417	0.4373
0.032	6.0	1044	1.4206	0.7493	0.3916	0.6437	0.2814	0.6437	0.2814	0.3916
0.032	7.0	1218	1.4775	0.7507	0.4055	0.6413	0.2965	0.6413	0.2965	0.4055
0.032	8.0	1392	1.5835	0.7421	0.4389	0.5833	0.3518	0.5833	0.3518	0.4389
0.0125	9.0	1566	1.7009	0.7464	0.3846	0.6322	0.2764	0.6322	0.2764	0.3846
0.0125	10.0	1740	1.7005	0.7392	0.4217	0.5789	0.3317	0.5789	0.3317	0.4217

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1