Update README.md

README.md

@@ -3,4 +3,106 @@ license: apache-2.0
 language:
 - zh
 pipeline_tag: text-generation
----
+---
+---
+<div style="text-align:center">
+<!-- <img src="https://big-cheng.com/k2/k2.png" alt="k2-logo" width="200"/> -->
+<h2>📈 CFGPT: Chinese Financial Assistant with Large Language Model (CFGPT1-sft-7b-LoRA)</h2>
+</div>
+
+## Introduction
+
+We introduce **CFGPT**, an open-source language model trained by firstly further pretraining general LLMs on collected and cleaned Chinese finance text data (CFData-pt), including financial domain-specific data (announcement, finance articles, finance exams, finance news, finance research papers) and general data (Wikipedia), and secondly fine-tuning with knowledge-intensive instruction tuning data (CFData-sft). 
+As for preliminary evaluation, we use CFBenchmark-Basic. 
+CFGPT outperforms the baselines on objective and subjective tasks compared to several baseline models with similar parameters. 
+
+In this repository, we will share the supervised finetuning LoRA model.
+
+- [Supervised Finetuned Model (Lora)](https://huggingface.co/TongjiFinLab/CFGPT1-sft-7B-LoRA): Adapter model weights trained by PEFT (LoRA).
+
+## How to Use
+
+**1. Prepare the code and the environment**
+
+Clone [CFGPT]() repository, create a Python environment, and activate it via the following command
+```bash
+git clone https://github.com/TongjiFinLab/CFGPT.git
+cd CFGPT
+conda create -n env_name python=3.10   
+source activate env_name 
+pip install -r requirements.txt
+```
+
+**2. Use CFGPT1-sft-7B-LoRA**
+
+```python
+from transformers import AutoModel, AutoTokenizer
+from peft import PeftModel
+base_model = 'TongjiFinLab/CFGPT1-pt-7B'
+lora_weights = 'TongjiFinLab/CFGPT1-sft-7B-LoRA'
+device_map = 'cuda:0'
+tokenizer = AutoTokenizer.from_pretrained(base_model, trust_remote_code=True)
+model = AutoModel.from_pretrained(
+    base_model,
+    trust_remote_code=True,
+    device_map=device_map,
+    torch_dtype=torch.bfloat16
+)
+model = PeftModel.from_pretrained(
+    model,
+    lora_weights,
+    device_map=device_map,
+)
+model = model.eval()
+inputs = tokenizer("""你是一名金融从业者，请对这篇新闻进行情感分析。请从（中性、积极、消极）中选取答案。新闻内容：挖贝快讯：特步国际发布2023年第二季度中国内地业务营运状况，披露截至2023年6月30日止3个月零售销售实现高双位数同比增长(包括线上线下渠道)，零售折扣水平约七五折。同时，2022年7月MSCI首次予以特步ESG评级，一年后评级表现即迎来提升。明晟MSCI上调特步ESG评级，由“BB”升至“BBB”。\n回答：""", return_tensors='pt').to('cuda:4')
+pred = model.generate(**inputs, max_new_tokens=64, do_sample=False, repetition_penalty=1.0)
+print(tokenizer.decode(pred.cpu()[0], skip_special_tokens=True).split('回答：')[1])
+```
+
+## 简介
+
+**CFGPT**是一个开源的语言模型，首先通过在收集和清理的中国金融文本数据（CFData-pt）上进行继续预训练，包括金融领域特定数据（公告、金融文章、金融考试、金融新闻、金融研究论文）和通用数据（维基百科），然后使用知识密集的指导调整数据（CFData-sft）进行微调。
+我们使用CFBenchmark-Basic进行初步评估。与几个具有相似参数的基线模型相比，CFGPT在识别，分类和生成任务上表现优越。
+
+在这个仓库中，我们将分享以下LoRA有监督微调的模型。
+
+- [Supervised Finetuned Model (Lora)](https://huggingface.co/TongjiFinLab/CFGPT1-sft-7B-LoRA): 基于我们继续预训练模型的由PEFT（LoRA）训练的适配器模型权重。
+
+## 如何使用
+
+**1. 准备代码和环境**
+
+克隆[CFGPT]()的仓库，创建一个Python环境，并通过以下命令激活它：
+```bash
+git clone https://github.com/TongjiFinLab/CFGPT.git
+cd CFGPT
+conda create -n env_name python=3.10   
+source activate env_name 
+pip install -r requirements.txt
+```
+
+**2. 使用 CFGPT1-sft-7B-LoRA**
+
+```python
+from transformers import AutoModel, AutoTokenizer
+from peft import PeftModel
+base_model = 'TongjiFinLab/CFGPT1-pt-7B'
+lora_weights = 'TongjiFinLab/CFGPT1-sft-7B-LoRA'
+device_map = 'cuda:0'
+tokenizer = AutoTokenizer.from_pretrained(base_model, trust_remote_code=True)
+model = AutoModel.from_pretrained(
+    base_model,
+    trust_remote_code=True,
+    device_map=device_map,
+    torch_dtype=torch.bfloat16
+)
+model = PeftModel.from_pretrained(
+    model,
+    lora_weights,
+    device_map=device_map,
+)
+model = model.eval()
+inputs = tokenizer("""你是一名金融从业者，请对这篇新闻进行情感分析。请从（中性、积极、消极）中选取答案。新闻内容：挖贝快讯：特步国际发布2023年第二季度中国内地业务营运状况，披露截至2023年6月30日止3个月零售销售实现高双位数同比增长(包括线上线下渠道)，零售折扣水平约七五折。同时，2022年7月MSCI首次予以特步ESG评级，一年后评级表现即迎来提升。明晟MSCI上调特步ESG评级，由“BB”升至“BBB”。\n回答：""", return_tensors='pt').to(device_map)
+pred = model.generate(**inputs, max_new_tokens=64, do_sample=False, repetition_penalty=1.0)
+print(tokenizer.decode(pred.cpu()[0], skip_special_tokens=True).split('回答：')[1])
+```