mdeberta-v3-base-finetuned-environment-energy-climate-classification

This model is a fine-tuned version of microsoft/mdeberta-v3-base to classify political texts related to the environment, energy and climate change.

It achieves the following results on the evaluation set:

• Loss: 0.1030
• Accuracy: 0.9748
• F1 Macro: 0.9456
• Accuracy Balanced: 0.9519
• F1 Micro: 0.9748
• Precision Macro: 0.9395
• Recall Macro: 0.9519
• Precision Micro: 0.9748
• Recall Micro: 0.9748

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: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.06
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro	Accuracy Balanced	F1 Micro	Precision Macro	Recall Macro	Precision Micro	Recall Micro
0.1037	1.0	9662	0.0972	0.9723	0.9385	0.9325	0.9723	0.9448	0.9325	0.9723	0.9723
0.0779	2.0	19324	0.0933	0.9732	0.9427	0.9537	0.9732	0.9324	0.9537	0.9732	0.9732
0.0474	3.0	28986	0.1030	0.9748	0.9456	0.9519	0.9748	0.9395	0.9519	0.9748	0.9748
0.0397	4.0	38648	0.1263	0.9750	0.9452	0.9455	0.9750	0.9450	0.9455	0.9750	0.9750

Framework versions

• Transformers 4.31.0
• Pytorch 2.5.0+cu121
• Datasets 2.14.7
• Tokenizers 0.13.3

Usage

from transformers import pipeline

# Load pipeline for text classification
classifier = pipeline("text-classification", model="mljn/mdeberta-v3-base-finetuned-environment-energy-climate-classification")

# Example input text
text = "The government has announced new incentives for renewable energy production, focusing on wind and solar energy to reduce carbon emissions."

# Predict
result = classifier(text)

print("Predicted class:", result[0]['label'])
print("Confidence score:", result[0]['score'])