ULS-MultiClinNERro-Qwen2.5-32B-procedure

This model is a fine-tuned version of Qwen/Qwen2.5-32B on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0002
• Precision: 1.0
• Recall: 1.0
• F1: 1.0
• Accuracy: 1.0

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: 0.0002
• train_batch_size: 128
• eval_batch_size: 1
• seed: 42
• 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	79	0.2073	0.5538	0.2108	0.3053	0.9380
No log	2.0	158	0.1149	0.4769	0.6030	0.5326	0.9628
No log	3.0	237	0.0483	0.7149	0.7576	0.7356	0.9857
No log	4.0	316	0.0252	0.8508	0.9016	0.8755	0.9935
No log	5.0	395	0.0097	0.9393	0.9602	0.9496	0.9976
No log	6.0	474	0.0049	0.9800	0.9731	0.9765	0.9990
0.1095	7.0	553	0.0014	0.9942	0.9965	0.9953	0.9998
0.1095	8.0	632	0.0006	0.9988	0.9977	0.9982	0.9999
0.1095	9.0	711	0.0003	1.0	1.0	1.0	1.0
0.1095	10.0	790	0.0002	1.0	1.0	1.0	1.0

Framework versions

• PEFT 0.14.0
• Transformers 4.47.0
• Pytorch 2.8.0+cu128
• Datasets 3.6.0
• Tokenizers 0.21.0