metadata

license: cc-by-nc-4.0
base_model: NYTK/PULI-BERT-Large
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: kmdb_ner_model
    results: []

kmdb_ner_model

This model is a fine-tuned version of NYTK/PULI-BERT-Large on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.0405
Precision: 0.8023
Recall: 0.8261
F1: 0.8140
Accuracy: 0.9842

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: 4
eval_batch_size: 4
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.0613	0.11	1000	0.0575	0.7102	0.7359	0.7228	0.9776
0.0559	0.23	2000	0.0525	0.7108	0.7611	0.7351	0.9788
0.0569	0.34	3000	0.0512	0.7375	0.7856	0.7608	0.9798
0.0481	0.45	4000	0.0496	0.7402	0.7975	0.7678	0.9803
0.041	0.57	5000	0.0474	0.7567	0.7981	0.7769	0.9809
0.0434	0.68	6000	0.0463	0.7588	0.7853	0.7718	0.9813
0.0577	0.8	7000	0.0462	0.7592	0.7896	0.7741	0.9812
0.0422	0.91	8000	0.0428	0.7820	0.8057	0.7937	0.9828
0.0346	1.02	9000	0.0443	0.7758	0.8153	0.7951	0.9825
0.0301	1.14	10000	0.0431	0.7831	0.8124	0.7974	0.9830
0.0305	1.25	11000	0.0427	0.7955	0.8162	0.8057	0.9834
0.0347	1.36	12000	0.0420	0.7923	0.8171	0.8045	0.9834
0.0397	1.48	13000	0.0422	0.7923	0.8220	0.8069	0.9835
0.0267	1.59	14000	0.0411	0.7970	0.8209	0.8087	0.9838
0.03	1.71	15000	0.0415	0.7946	0.8230	0.8085	0.9838
0.0311	1.82	16000	0.0407	0.8023	0.8253	0.8136	0.9842
0.0283	1.93	17000	0.0405	0.8023	0.8261	0.8140	0.9842

Framework versions

Transformers 4.38.1
Pytorch 2.0.1+cu117
Datasets 2.16.1
Tokenizers 0.15.0