DeGrimer/disaster_tweet_classification

binary_classification2

This model is a fine-tuned version of cardiffnlp/twitter-xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7178
• F1: 0.7750

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: 5

Training results

Training Loss	Epoch	Step	Validation Loss	F1
0.4132	0.2625	100	0.4500	0.7621
0.6075	0.5249	200	0.4293	0.7781
0.3672	0.7874	300	0.3937	0.7964
0.3682	1.0499	400	0.4572	0.7864
0.3747	1.3123	500	0.4268	0.7946
0.339	1.5748	600	0.4087	0.7853
0.4468	1.8373	700	0.4649	0.7972
0.2621	2.0997	800	0.4521	0.7889
0.2443	2.3622	900	0.5213	0.7927
0.289	2.6247	1000	0.5290	0.7749
0.3365	2.8871	1100	0.4803	0.7835
0.3744	3.1496	1200	0.6371	0.7861
0.2841	3.4121	1300	0.6773	0.7881
0.2847	3.6745	1400	0.6217	0.7826
0.1881	3.9370	1500	0.6898	0.7807
0.0901	4.1995	1600	0.7506	0.7726
0.1814	4.4619	1700	0.7278	0.7760
0.3393	4.7244	1800	0.7157	0.7729
0.2436	4.9869	1900	0.7178	0.7750

Framework versions

• Transformers 4.45.1
• Pytorch 2.4.0
• Datasets 3.0.1
• Tokenizers 0.20.0