minoosh/finetuned_bert-base_on_shEMO_transcripts

finetuned_bert-base

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: 1.2046
• Accuracy: 0.52

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.413	1.0	75	1.3023	0.4667
1.2772	2.0	150	1.2043	0.52
1.2019	3.0	225	1.0879	0.5733
1.1463	4.0	300	1.1124	0.57
1.1566	5.0	375	1.1220	0.5367
1.1096	6.0	450	1.0675	0.5967
0.9806	7.0	525	1.0315	0.64
0.8715	8.0	600	1.0616	0.6
0.8788	9.0	675	1.1211	0.59
0.8071	10.0	750	1.1400	0.6
0.6908	11.0	825	1.1848	0.6033
0.6244	12.0	900	1.2255	0.59
0.628	13.0	975	1.2264	0.6
0.6003	14.0	1050	1.2270	0.6033
0.5283	15.0	1125	1.2399	0.5933

Framework versions

• Transformers 4.34.1
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1