tfm-bert-sesgo

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

• Loss: 2.0691
• Accuracy: 0.2917
• Macro F1: 0.2417

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: 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
• num_epochs: 8

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Macro F1
No log	1.0	7	2.0934	0.0833	0.0739
No log	2.0	14	2.0025	0.2083	0.2173
No log	3.0	21	1.9304	0.2917	0.2524
No log	4.0	28	1.9471	0.2917	0.2357
No log	5.0	35	2.0958	0.25	0.2375
No log	6.0	42	2.0844	0.3333	0.2812
No log	7.0	49	2.0707	0.3333	0.2857
No log	8.0	56	2.0691	0.2917	0.2417

Framework versions

• Transformers 5.8.1
• Pytorch 2.10.0+cu128
• Datasets 4.8.5
• Tokenizers 0.22.2