distilbert_twitterfin_padding60model

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

• Loss: 1.2102
• Accuracy: 0.8673

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6858	1.0	597	0.5056	0.8086
0.4365	2.0	1194	0.3822	0.8606
0.3156	3.0	1791	0.4330	0.8626
0.228	4.0	2388	0.5520	0.8710
0.1729	5.0	2985	0.6801	0.8434
0.0844	6.0	3582	0.6752	0.8626
0.0788	7.0	4179	0.7749	0.8631
0.0598	8.0	4776	0.8199	0.8572
0.0544	9.0	5373	0.9913	0.8543
0.0422	10.0	5970	1.0151	0.8626
0.0273	11.0	6567	0.9725	0.8597
0.0235	12.0	7164	1.0349	0.8668
0.0166	13.0	7761	1.1150	0.8656
0.0131	14.0	8358	1.1363	0.8710
0.0108	15.0	8955	1.1903	0.8543
0.0079	16.0	9552	1.1553	0.8652
0.0112	17.0	10149	1.1944	0.8652
0.0035	18.0	10746	1.2227	0.8681
0.0064	19.0	11343	1.2064	0.8677
0.0033	20.0	11940	1.2102	0.8673

Framework versions

• Transformers 4.33.2
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.13.3