bert-tiny-finetuned-wnut17-ner

This model is a fine-tuned version of google/bert_uncased_L-2_H-128_A-2 on the wnut_17 dataset. It achieves the following results on the evaluation set:

• Loss: 0.6054
• Precision: 0.0
• Recall: 0.0
• F1: 0.0
• Accuracy: 0.8961

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: 128
• eval_batch_size: 128
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	27	1.1060	0.0	0.0	0.0	0.8961
No log	2.0	54	0.9075	0.0	0.0	0.0	0.8961
No log	3.0	81	0.7978	0.0	0.0	0.0	0.8961
No log	4.0	108	0.7333	0.0	0.0	0.0	0.8961
No log	5.0	135	0.6929	0.0	0.0	0.0	0.8961
No log	6.0	162	0.6661	0.0	0.0	0.0	0.8961
No log	7.0	189	0.6477	0.0	0.0	0.0	0.8961
No log	8.0	216	0.6346	0.0	0.0	0.0	0.8961
No log	9.0	243	0.6251	0.0	0.0	0.0	0.8961
No log	10.0	270	0.6182	0.0	0.0	0.0	0.8961
No log	11.0	297	0.6132	0.0	0.0	0.0	0.8961
No log	12.0	324	0.6097	0.0	0.0	0.0	0.8961
No log	13.0	351	0.6073	0.0	0.0	0.0	0.8961
No log	14.0	378	0.6059	0.0	0.0	0.0	0.8961
No log	15.0	405	0.6054	0.0	0.0	0.0	0.8961

Framework versions

• Transformers 4.21.0
• Pytorch 1.12.0+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1