bert_sst5_padding70model

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: 4.3219
• Accuracy: 0.5195

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
1.455	1.0	534	1.3095	0.4014
1.0688	2.0	1068	1.0861	0.5231
0.8152	3.0	1602	1.1682	0.5317
0.6373	4.0	2136	1.3330	0.5330
0.4456	5.0	2670	1.6174	0.5167
0.3164	6.0	3204	1.8698	0.5190
0.2454	7.0	3738	2.1769	0.5167
0.1684	8.0	4272	2.6182	0.5281
0.1367	9.0	4806	2.9459	0.5068
0.1176	10.0	5340	3.2425	0.5104
0.1056	11.0	5874	3.3523	0.5262
0.0844	12.0	6408	3.5235	0.5258
0.0538	13.0	6942	3.7476	0.5181
0.0614	14.0	7476	3.8546	0.5131
0.0369	15.0	8010	3.9777	0.5217
0.0229	16.0	8544	4.1242	0.5176
0.0183	17.0	9078	4.1762	0.5181
0.014	18.0	9612	4.2317	0.5222
0.012	19.0	10146	4.2420	0.5235
0.0143	20.0	10680	4.3219	0.5195

Framework versions

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