bert_sst5_padding80model

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

• Loss: 3.9364
• Accuracy: 0.5367

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
1.3796	1.0	534	1.3990	0.3833
1.0718	2.0	1068	1.0650	0.5344
0.8586	3.0	1602	1.1671	0.5163
0.6983	4.0	2136	1.2731	0.5362
0.5278	5.0	2670	1.5800	0.5195
0.403	6.0	3204	1.8418	0.5181
0.2944	7.0	3738	1.9392	0.5317
0.238	8.0	4272	2.1068	0.5208
0.1969	9.0	4806	2.4820	0.5362
0.1571	10.0	5340	2.9684	0.5222
0.1358	11.0	5874	3.0886	0.5249
0.1232	12.0	6408	3.1854	0.5290
0.0877	13.0	6942	3.4065	0.5258
0.0738	14.0	7476	3.5549	0.5308
0.0438	15.0	8010	3.7069	0.5235
0.0429	16.0	8544	3.7658	0.5294
0.0349	17.0	9078	3.8455	0.5299
0.0218	18.0	9612	3.8512	0.5353
0.0174	19.0	10146	3.9104	0.5357
0.0159	20.0	10680	3.9364	0.5367

Framework versions

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