---
language:
- zh
datasets:
- gyr66/privacy_detection
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: RoBERTa-ext-large-chinese-finetuned-ner
  results:
  - task:
      name: Token Classification
      type: token-classification
    dataset:
        name: gyr66/privacy_detection
        type: gyr66/privacy_detection
        config: privacy_detection
        split: train
        args: privacy_detection
    metrics:
    - name: Precision
      type: precision
      value: 0.7052
    - name: Recall
      type: recall
      value: 0.7606
    - name: F1
      type: f1
      value: 0.7318
    - name: Accuracy
      type: accuracy
      value: 0.9138
---

# RoBERTa-ext-large-chinese-finetuned-ner

This model is a fine-tuned version of [chinese-roberta-wwm-ext-large](https://huggingface.co/hfl/chinese-roberta-wwm-ext-large) on the [gyr66/privacy_detection](https://huggingface.co/datasets/gyr66/privacy_detection) dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7697
- Precision: 0.7052
- Recall: 0.7606
- F1: 0.7318
- Accuracy: 0.9138

## Model description

The model is used for competition: "https://www.datafountain.cn/competitions/472"

## Training and evaluation data

The training and evaluation data is from [gyr66/privacy_detection](https://huggingface.co/datasets/gyr66/privacy_detection) dataset.

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 28
- eval_batch_size: 28
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear


### Framework versions

- Transformers 4.27.3
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.13.2