mljn/mdeberta-v3-base-finetuned-climate-opposition-new

mdeberta-v3-base-finetuned-climate-support-new

This model is a fine-tuned version of microsoft/mdeberta-v3-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5285
• Accuracy: 0.9018
• F1 Macro: 0.8544
• Accuracy Balanced: 0.8495
• F1 Micro: 0.9018
• Precision Macro: 0.8596
• Recall Macro: 0.8495
• Precision Micro: 0.9018
• Recall Micro: 0.9018

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• 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
0.4812	0.6046	500	0.4510	0.8817	0.7948	0.7543	0.8817	0.8832	0.7543	0.8817	0.8817
0.2994	1.2092	1000	0.4132	0.8962	0.8384	0.8190	0.8962	0.8642	0.8190	0.8962	0.8962
0.2375	1.8138	1500	0.4492	0.8990	0.8339	0.8000	0.8990	0.8910	0.8000	0.8990	0.8990
0.1788	2.4184	2000	0.5988	0.8616	0.8191	0.8614	0.8616	0.7962	0.8614	0.8616	0.8616
0.1445	3.0230	2500	0.5146	0.8896	0.8458	0.8624	0.8896	0.8324	0.8624	0.8896	0.8896
0.0928	3.6276	3000	0.5285	0.9018	0.8544	0.8495	0.9018	0.8596	0.8495	0.9018	0.9018

Framework versions

• Transformers 4.56.2
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1