Update README.md
#1
by bztxb - opened
README.md
CHANGED
|
@@ -1,3 +1,55 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: cc-by-nc-sa-4.0
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-nc-sa-4.0
|
| 3 |
+
---
|
| 4 |
+
|
| 5 |
+
# 明实录与清实录多标签分类推理模型
|
| 6 |
+
|
| 7 |
+
本模型用于对《明实录》和《清实录》文本进行多标签分类推理。基于[Jihuai/bert-ancient-chinese](https://huggingface.co/Jihuai/bert-ancient-chinese)进行任务微调，利用公开语料进行预训练，得到适合实录类型的预训练模型[shiluBERT](https://huggingface.co/bztxb/shiluBERT)。
|
| 8 |
+
|
| 9 |
+
## 中文说明
|
| 10 |
+
|
| 11 |
+
### 模型与数据来源
|
| 12 |
+
|
| 13 |
+
- 训练数据来源：[《朝鲜王朝实录》](https://sillok.history.go.kr)；
|
| 14 |
+
- 任务类型：多标签文本分类；
|
| 15 |
+
- 训练样本数：约27万。
|
| 16 |
+
|
| 17 |
+
### 评估指标
|
| 18 |
+
|
| 19 |
+
| 指标 | 数值 |
|
| 20 |
+
|---|---|
|
| 21 |
+
| Sample F1 | 0.7246 |
|
| 22 |
+
| Sample Precision | 0.7594 |
|
| 23 |
+
| Sample Recall | 0.7321 |
|
| 24 |
+
| LRAP | 0.8074 |
|
| 25 |
+
| Hamming Loss | 0.0069 |
|
| 26 |
+
|
| 27 |
+
### 示例使用方法
|
| 28 |
+
|
| 29 |
+
- 在线体验 Space： [bztxb/shiluInfer](https://huggingface.co/spaces/bztxb/shiluInfer)
|
| 30 |
+
|
| 31 |
+
|
| 32 |
+
## English Version
|
| 33 |
+
|
| 34 |
+
This model performs multi-label classification inference on texts of VERITABLE RECORDS of the Ming/Qing DYNASTY. It is fine-tuned from [Jihuai/bert-ancient-chinese](https://huggingface.co/Jihuai/bert-ancient-chinese), and further benefits from pretraining on public corpora to obtain a Shilu-oriented pretrained model, [shiluBERT](https://huggingface.co/bztxb/shiluBERT).
|
| 35 |
+
|
| 36 |
+
### Model and Data Sources
|
| 37 |
+
|
| 38 |
+
- Training data source: [VERITABLE RECORDS of the JOSEON DYNASTY](https://sillok.history.go.kr).
|
| 39 |
+
- Task type: multi-label text classification.
|
| 40 |
+
- Number of training samples: approximately 0.27 million.
|
| 41 |
+
|
| 42 |
+
### Evaluation Metrics
|
| 43 |
+
|
| 44 |
+
| Metric | Value |
|
| 45 |
+
|---|---|
|
| 46 |
+
| Sample F1 | 0.7246 |
|
| 47 |
+
| Sample Precision | 0.7594 |
|
| 48 |
+
| Sample Recall | 0.7321 |
|
| 49 |
+
| LRAP | 0.8074 |
|
| 50 |
+
| Hamming Loss | 0.0069 |
|
| 51 |
+
|
| 52 |
+
### Example Usage
|
| 53 |
+
|
| 54 |
+
- Try the online Space: [bztxb/shiluInfer](https://huggingface.co/spaces/bztxb/shiluInfer)
|
| 55 |
+
|