72238b54a6
Model: RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf Source: Original Platform
211 lines
11 KiB
Markdown
211 lines
11 KiB
Markdown
Quantization made by Richard Erkhov.
|
||
|
||
[Github](https://github.com/RichardErkhov)
|
||
|
||
[Discord](https://discord.gg/pvy7H8DZMG)
|
||
|
||
[Request more models](https://github.com/RichardErkhov/quant_request)
|
||
|
||
|
||
chinese-text-correction-1.5b - GGUF
|
||
- Model creator: https://huggingface.co/shibing624/
|
||
- Original model: https://huggingface.co/shibing624/chinese-text-correction-1.5b/
|
||
|
||
|
||
| Name | Quant method | Size |
|
||
| ---- | ---- | ---- |
|
||
| [chinese-text-correction-1.5b.Q2_K.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q2_K.gguf) | Q2_K | 0.63GB |
|
||
| [chinese-text-correction-1.5b.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q3_K_S.gguf) | Q3_K_S | 0.71GB |
|
||
| [chinese-text-correction-1.5b.Q3_K.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q3_K.gguf) | Q3_K | 0.77GB |
|
||
| [chinese-text-correction-1.5b.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q3_K_M.gguf) | Q3_K_M | 0.77GB |
|
||
| [chinese-text-correction-1.5b.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q3_K_L.gguf) | Q3_K_L | 0.82GB |
|
||
| [chinese-text-correction-1.5b.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.IQ4_XS.gguf) | IQ4_XS | 0.84GB |
|
||
| [chinese-text-correction-1.5b.Q4_0.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q4_0.gguf) | Q4_0 | 0.87GB |
|
||
| [chinese-text-correction-1.5b.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.IQ4_NL.gguf) | IQ4_NL | 0.88GB |
|
||
| [chinese-text-correction-1.5b.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q4_K_S.gguf) | Q4_K_S | 0.88GB |
|
||
| [chinese-text-correction-1.5b.Q4_K.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q4_K.gguf) | Q4_K | 0.92GB |
|
||
| [chinese-text-correction-1.5b.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q4_K_M.gguf) | Q4_K_M | 0.92GB |
|
||
| [chinese-text-correction-1.5b.Q4_1.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q4_1.gguf) | Q4_1 | 0.95GB |
|
||
| [chinese-text-correction-1.5b.Q5_0.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q5_0.gguf) | Q5_0 | 1.02GB |
|
||
| [chinese-text-correction-1.5b.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q5_K_S.gguf) | Q5_K_S | 1.02GB |
|
||
| [chinese-text-correction-1.5b.Q5_K.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q5_K.gguf) | Q5_K | 1.05GB |
|
||
| [chinese-text-correction-1.5b.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q5_K_M.gguf) | Q5_K_M | 1.05GB |
|
||
| [chinese-text-correction-1.5b.Q5_1.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q5_1.gguf) | Q5_1 | 1.1GB |
|
||
| [chinese-text-correction-1.5b.Q6_K.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q6_K.gguf) | Q6_K | 1.19GB |
|
||
| [chinese-text-correction-1.5b.Q8_0.gguf](https://huggingface.co/RichardErkhov/shibing624_-_chinese-text-correction-1.5b-gguf/blob/main/chinese-text-correction-1.5b.Q8_0.gguf) | Q8_0 | 1.53GB |
|
||
|
||
|
||
|
||
|
||
Original model description:
|
||
---
|
||
library_name: transformers
|
||
base_model: Qwen/Qwen2.5-1.5B-Instruct
|
||
license: apache-2.0
|
||
datasets:
|
||
- shibing624/chinese_text_correction
|
||
language:
|
||
- zh
|
||
metrics:
|
||
- f1
|
||
tags:
|
||
- text-generation-inference
|
||
widget:
|
||
- text: "文本纠错：\n少先队员因该为老人让坐。"
|
||
---
|
||
|
||
|
||
|
||
# Chinese Text Correction Model
|
||
中文文本纠错模型chinese-text-correction-1.5b：用于拼写纠错、语法纠错
|
||
|
||
`shibing624/chinese-text-correction-1.5b` evaluate test data：
|
||
|
||
The overall performance of CSC **test**:
|
||
|
||
|input_text|predict_text|
|
||
|:--- |:--- |
|
||
|文本纠错：\n少先队员因该为老人让坐。|少先队员应该为老人让座。|
|
||
|
||
# Models
|
||
|
||
| Name | Base Model | Download |
|
||
|-----------------|-------------------|-----------------------------------------------------------------------|
|
||
| chinese-text-correction-1.5b | Qwen/Qwen2.5-1.5B-Instruct | [🤗 Hugging Face](https://huggingface.co/shibing624/chinese-text-correction-1.5b) |
|
||
| chinese-text-correction-1.5b-lora | Qwen/Qwen2.5-1.5B-Instruct | [🤗 Hugging Face](https://huggingface.co/shibing624/chinese-text-correction-1.5b-lora) |
|
||
| chinese-text-correction-7b | Qwen/Qwen2.5-7B-Instruct | [🤗 Hugging Face](https://huggingface.co/shibing624/chinese-text-correction-7b) |
|
||
| chinese-text-correction-7b-lora | Qwen/Qwen2.5-7B-Instruct | [🤗 Hugging Face](https://huggingface.co/shibing624/chinese-text-correction-7b-lora) |
|
||
|
||
|
||
### 评估结果
|
||
- 评估指标：F1
|
||
- CSC(Chinese Spelling Correction): 拼写纠错模型，表示模型可以处理音似、形似、语法等长度对齐的错误纠正
|
||
- CTC(CHinese Text Correction): 文本纠错模型，表示模型支持拼写、语法等长度对齐的错误纠正，还可以处理多字、少字等长度不对齐的错误纠正
|
||
- GPU：Tesla V100，显存 32 GB
|
||
|
||
| Model Name | Model Link | Base Model | Avg | SIGHAN-2015 | EC-LAW | MCSC | GPU/CPU | QPS |
|
||
|:-----------------|:------------------------------------------------------------------------------------------------------------------------|:---------------------------|:-----------|:------------|:-------|:-------|:--------|:--------|
|
||
| Kenlm-CSC | [shibing624/chinese-kenlm-klm](https://huggingface.co/shibing624/chinese-kenlm-klm) | kenlm | 0.3409 | 0.3147 | 0.3763 | 0.3317 | CPU | 9 |
|
||
| Mengzi-T5-CSC | [shibing624/mengzi-t5-base-chinese-correction](https://huggingface.co/shibing624/mengzi-t5-base-chinese-correction) | mengzi-t5-base | 0.3984 | 0.7758 | 0.3156 | 0.1039 | GPU | 214 |
|
||
| ERNIE-CSC | [PaddleNLP/ernie-csc](https://github.com/PaddlePaddle/PaddleNLP/tree/develop/legacy/examples/text_correction/ernie-csc) | PaddlePaddle/ernie-1.0-base-zh | 0.4353 | 0.8383 | 0.3357 | 0.1318 | GPU | 114 |
|
||
| MacBERT-CSC | [shibing624/macbert4csc-base-chinese](https://huggingface.co/shibing624/macbert4csc-base-chinese) | hfl/chinese-macbert-base | 0.3993 | 0.8314 | 0.1610 | 0.2055 | GPU | **224** |
|
||
| ChatGLM3-6B-CSC | [shibing624/chatglm3-6b-csc-chinese-lora](https://huggingface.co/shibing624/chatglm3-6b-csc-chinese-lora) | THUDM/chatglm3-6b | 0.4538 | 0.6572 | 0.4369 | 0.2672 | GPU | 3 |
|
||
| Qwen2.5-1.5B-CTC | [shibing624/chinese-text-correction-1.5b](https://huggingface.co/shibing624/chinese-text-correction-1.5b) | Qwen/Qwen2.5-1.5B-Instruct | 0.6802 | 0.3032 | 0.7846 | 0.9529 | GPU | 6 |
|
||
| Qwen2.5-7B-CTC | [shibing624/chinese-text-correction-7b](https://huggingface.co/shibing624/chinese-text-correction-7b) | Qwen/Qwen2.5-7B-Instruct | **0.8225** | 0.4917 | 0.9798 | 0.9959 | GPU | 3 |
|
||
|
||
## Usage (pycorrector)
|
||
|
||
本项目开源在`pycorrector`项目：[pycorrector](https://github.com/shibing624/pycorrector)，可支持大模型微调后用于文本纠错，通过如下命令调用：
|
||
|
||
Install package:
|
||
```shell
|
||
pip install -U pycorrector
|
||
```
|
||
|
||
```python
|
||
from pycorrector.gpt.gpt_corrector import GptCorrector
|
||
|
||
if __name__ == '__main__':
|
||
error_sentences = [
|
||
'真麻烦你了。希望你们好好的跳无',
|
||
'少先队员因该为老人让坐',
|
||
'机七学习是人工智能领遇最能体现智能的一个分知',
|
||
'一只小鱼船浮在平净的河面上',
|
||
'我的家乡是有明的渔米之乡',
|
||
]
|
||
m = GptCorrector("shibing624/chinese-text-correction-1.5b")
|
||
|
||
batch_res = m.correct_batch(error_sentences)
|
||
for i in batch_res:
|
||
print(i)
|
||
print()
|
||
```
|
||
|
||
## Usage (HuggingFace Transformers)
|
||
Without [pycorrector](https://github.com/shibing624/pycorrector), you can use the model like this:
|
||
|
||
First, you pass your input through the transformer model, then you get the generated sentence.
|
||
|
||
Install package:
|
||
```
|
||
pip install transformers
|
||
```
|
||
|
||
```python
|
||
# pip install transformers
|
||
from transformers import AutoModelForCausalLM, AutoTokenizer
|
||
checkpoint = "shibing624/chinese-text-correction-1.5b"
|
||
|
||
device = "cuda" # for GPU usage or "cpu" for CPU usage
|
||
tokenizer = AutoTokenizer.from_pretrained(checkpoint)
|
||
model = AutoModelForCausalLM.from_pretrained(checkpoint).to(device)
|
||
|
||
input_content = "文本纠错：\n少先队员因该为老人让坐。"
|
||
|
||
messages = [{"role": "user", "content": input_content}]
|
||
input_text=tokenizer.apply_chat_template(messages, tokenize=False)
|
||
|
||
print(input_text)
|
||
|
||
inputs = tokenizer.encode(input_text, return_tensors="pt").to(device)
|
||
outputs = model.generate(inputs, max_new_tokens=1024, temperature=0, do_sample=False, repetition_penalty=1.08)
|
||
|
||
print(tokenizer.decode(outputs[0]))
|
||
```
|
||
|
||
output:
|
||
```shell
|
||
少先队员应该为老人让座。
|
||
```
|
||
|
||
|
||
模型文件组成：
|
||
```
|
||
shibing624/chinese-text-correction-1.5b
|
||
|-- added_tokens.json
|
||
|-- config.json
|
||
|-- generation_config.json
|
||
|-- merges.txt
|
||
|-- model.safetensors
|
||
|-- model.safetensors.index.json
|
||
|-- README.md
|
||
|-- special_tokens_map.json
|
||
|-- tokenizer_config.json
|
||
|-- tokenizer.json
|
||
`-- vocab.json
|
||
```
|
||
|
||
#### 训练参数：
|
||
|
||
- num_epochs: 8
|
||
- batch_size: 4
|
||
- steps: 36000
|
||
- eval_loss: 0.14
|
||
- base model: Qwen/Qwen2.5-1.5B-Instruct
|
||
- train data: [shibing624/chinese_text_correction](https://huggingface.co/datasets/shibing624/chinese_text_correction)
|
||
- train time: 9 days 8 hours
|
||
- eval_loss: 
|
||
- train_loss: 
|
||
|
||
### 训练数据集
|
||
#### 中文纠错数据集
|
||
|
||
- 数据：[shibing624/chinese_text_correction](https://huggingface.co/datasets/shibing624/chinese_text_correction)
|
||
|
||
|
||
如果需要训练Qwen的纠错模型，请参考[https://github.com/shibing624/pycorrector](https://github.com/shibing624/pycorrector) 或者 [https://github.com/shibing624/MedicalGPT](https://github.com/shibing624/MedicalGPT)
|
||
|
||
## Citation
|
||
|
||
```latex
|
||
@software{pycorrector,
|
||
author = {Xu Ming},
|
||
title = {pycorrector: Implementation of language model finetune},
|
||
year = {2024},
|
||
url = {https://github.com/shibing624/pycorrector},
|
||
}
|
||
```
|
||
|
||
|
||
|