N_distilbert_sst5_padding90model

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: 3.9075
• Accuracy: 0.5068

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
1.4153	1.0	534	1.2718	0.4063
1.097	2.0	1068	1.1023	0.5235
0.8991	3.0	1602	1.1412	0.5258
0.726	4.0	2136	1.2640	0.5267
0.5663	5.0	2670	1.5469	0.4860
0.4446	6.0	3204	1.6275	0.5086
0.3323	7.0	3738	1.8364	0.5118
0.2447	8.0	4272	2.0486	0.5072
0.2051	9.0	4806	2.3180	0.4950
0.1546	10.0	5340	2.5455	0.5100
0.135	11.0	5874	2.7967	0.5104
0.1104	12.0	6408	3.0989	0.5009
0.0825	13.0	6942	3.3861	0.4937
0.0788	14.0	7476	3.4264	0.5027
0.0534	15.0	8010	3.5892	0.5068
0.0428	16.0	8544	3.7231	0.4968
0.0374	17.0	9078	3.8164	0.5014
0.0236	18.0	9612	3.7950	0.5086
0.0256	19.0	10146	3.8540	0.5104
0.0202	20.0	10680	3.9075	0.5068

Framework versions

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