metadata

license: mit
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: ner_column_bert-base-NER
    results: []
language:
  - en
widget:
  - >-
    india 0S0308Z8   trudeau 3000 Ravensburger Hamnoy, Lofoten of gold
    bestseller 620463000001
  - >-
    other china lc waikiki mağazacilik hi̇zmetleri̇ ti̇c aş 630140000000 hilti
    6204699090_BD 55L Toaster Oven with Double Glass
  - >-
    611020000001 italy Apparel  other games 9W1964Z8 debenhams guangzhou hec
    fashion leather co ltd

ner_column_bert-base-NER

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

Loss: 0.1872
Precision: 0.7623
Recall: 0.7753
F1: 0.7688
Accuracy: 0.9023

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: 64
eval_batch_size: 64
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	702	0.6427	0.3025	0.2180	0.2534	0.7415
0.9329	2.0	1404	0.4771	0.4343	0.3587	0.3929	0.7955
0.546	3.0	2106	0.3983	0.5157	0.4530	0.4823	0.8242
0.546	4.0	2808	0.3748	0.5089	0.4758	0.4918	0.8305
0.4339	5.0	3510	0.2947	0.6362	0.6146	0.6252	0.8656
0.3658	6.0	4212	0.2818	0.6421	0.6231	0.6325	0.8664
0.3658	7.0	4914	0.2459	0.7108	0.6983	0.7045	0.8834
0.3221	8.0	5616	0.2665	0.6586	0.6404	0.6494	0.8701
0.2914	9.0	6318	0.2449	0.6880	0.6768	0.6823	0.8793
0.2657	10.0	7020	0.2411	0.7014	0.6862	0.6937	0.8824
0.2657	11.0	7722	0.2179	0.7261	0.7228	0.7244	0.8902
0.2453	12.0	8424	0.2301	0.6922	0.6919	0.6920	0.8858
0.2295	13.0	9126	0.2352	0.6768	0.6836	0.6802	0.8832
0.2295	14.0	9828	0.2020	0.7545	0.7499	0.7522	0.8970
0.2155	15.0	10530	0.2012	0.7449	0.7508	0.7478	0.8974
0.2064	16.0	11232	0.2036	0.7282	0.7402	0.7341	0.8960
0.2064	17.0	11934	0.1976	0.7390	0.7496	0.7443	0.8974
0.1978	18.0	12636	0.1859	0.7688	0.7828	0.7757	0.9040
0.1895	19.0	13338	0.1917	0.7574	0.7691	0.7632	0.9014
0.186	20.0	14040	0.1872	0.7623	0.7753	0.7688	0.9023

Framework versions

Transformers 4.30.2
Pytorch 1.13.1+cu116
Datasets 2.13.2
Tokenizers 0.13.3