se_train_run_AUTOIMMUNE

This model is a fine-tuned version of ModernBERT-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.8454
• Model Preparation Time: 0.0023
• F1: 0.9142
• Precision: 0.8907
• Recall: 0.939
• Threshold: 0.6448
• Sim Ratio: 1.8668
• Pos Sim: 0.8175
• Neg Sim: 0.4379

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: 0.0001
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	F1	Precision	Recall	Threshold	Sim Ratio	Pos Sim	Neg Sim
1.1666	0.1600	5000	3.8219	0.0023	0.8785	0.8627	0.8949	0.6512	1.8895	0.8276	0.438
1.0768	0.3201	10000	3.7588	0.0023	0.8937	0.8788	0.9092	0.6407	2.0951	0.8217	0.3922
0.9115	0.4801	15000	4.0146	0.0023	0.9012	0.872	0.9324	0.6185	2.1225	0.8194	0.386
0.8536	0.6402	20000	3.5470	0.0023	0.906	0.869	0.9464	0.6418	1.8244	0.8317	0.4559
0.8147	0.8002	25000	3.8247	0.0023	0.9115	0.8849	0.9398	0.6342	1.9331	0.8186	0.4235
0.799	0.9603	30000	3.8038	0.0023	0.914	0.8991	0.9295	0.6584	1.8577	0.8189	0.4408

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1
• Datasets 3.2.0
• Tokenizers 0.21.0