RoBERTa-Base-full-finetuned-ner-multi-label

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

• Loss: 0.0484
• F1 Micro: 0.8025
• Precision Micro: 0.8296
• Recall Micro: 0.7772

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Micro	Precision Micro	Recall Micro
No log	1.0	27	0.1227	0.6758	0.7555	0.6114
No log	2.0	54	0.0750	0.7087	0.9323	0.5716
No log	3.0	81	0.0628	0.7597	0.8531	0.6848
No log	4.0	108	0.0554	0.7868	0.8768	0.7136
No log	5.0	135	0.0522	0.7987	0.8228	0.7759
No log	6.0	162	0.0508	0.7967	0.8283	0.7674
No log	7.0	189	0.0493	0.8005	0.8263	0.7763
No log	8.0	216	0.0489	0.8032	0.8253	0.7822
No log	9.0	243	0.0490	0.8014	0.8171	0.7864
No log	10.0	270	0.0484	0.8025	0.8296	0.7772

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.0
• Tokenizers 0.19.1