minilm-finetuned-emotion-class-model

This model is a fine-tuned version of microsoft/MiniLM-L12-H384-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.1026
• F1 Score: 0.6649

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: 128
• eval_batch_size: 128
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Score
1.8502	1.0	270	1.4798	0.5071
1.3541	2.0	540	1.2377	0.5836
1.1809	3.0	810	1.1675	0.6202
1.0891	4.0	1080	1.1081	0.6522
1.0205	5.0	1350	1.0815	0.6603
0.9624	6.0	1620	1.0640	0.6645
0.9185	7.0	1890	1.0572	0.6689
0.8811	8.0	2160	1.0433	0.6693
0.8531	9.0	2430	1.0479	0.6746
0.8208	10.0	2700	1.0536	0.6697
0.8014	11.0	2970	1.0564	0.6713
0.7798	12.0	3240	1.0634	0.6716
0.7568	13.0	3510	1.0744	0.6698
0.7414	14.0	3780	1.0782	0.6704
0.7265	15.0	4050	1.0810	0.6694
0.7128	16.0	4320	1.0885	0.6684
0.7054	17.0	4590	1.0917	0.6631
0.6927	18.0	4860	1.0961	0.6678
0.6848	19.0	5130	1.1005	0.6644
0.6742	20.0	5400	1.1026	0.6649

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2