model_2_stage3-seed_2025

This model is a fine-tuned version of maud-dr/model_2_stage2-seed_42 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.0962
• Precision: 0.6270
• Recall: 0.6059
• F1: 0.6163

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: 0.0003
• train_batch_size: 8
• eval_batch_size: 8
• seed: 2025
• optimizer: Use OptimizerNames.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: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1
0.1634	1.0	745	1.0563	0.6333	0.5991	0.6157
0.1316	2.0	1490	1.2545	0.5962	0.6982	0.6432
0.1042	3.0	2235	1.3465	0.6123	0.6509	0.6310
0.0982	4.0	2980	1.3607	0.6086	0.6689	0.6373
0.0584	5.0	3725	1.5593	0.6231	0.6441	0.6334
0.0523	6.0	4470	1.4722	0.6571	0.5698	0.6104
0.0403	7.0	5215	1.1651	0.6167	0.6306	0.6236
0.0547	8.0	5960	1.4298	0.6171	0.6171	0.6171
0.0266	9.0	6705	1.9920	0.6174	0.6216	0.6195
0.0275	10.0	7450	1.6296	0.6316	0.6216	0.6266
0.0262	11.0	8195	1.7069	0.5832	0.6712	0.6241
0.0285	12.0	8940	1.6690	0.6542	0.5495	0.5973
0.0032	13.0	9685	1.9120	0.6259	0.5991	0.6122
0.0012	14.0	10430	2.0423	0.6194	0.6194	0.6194
0.0001	15.0	11175	2.0962	0.6270	0.6059	0.6163

Framework versions

• Transformers 4.52.0.dev0
• Pytorch 2.7.0+cu126
• Datasets 3.6.0
• Tokenizers 0.21.1