From 645242a55269cafc06af2d08b276f9853a970de6 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E6=88=92=E9=85=92=E7=9A=84=E6=9D=8E=E7=99=BD?=
 <670939375@qq.com>
Date: Mon, 4 Aug 2025 14:06:45 +0800
Subject: [PATCH] Updated how the fine-tuned BERT model is stored.

---
 .gitignore                                    |  1 +
 .../BertChinese-Lora/README.md                | 11 +++++--
 .../BertChinese-Lora/predict.py               | 20 +++++++++--
 .../BertChinese-Lora/predict_pipeline.py      | 33 ++++++++++++++++---
 4 files changed, 55 insertions(+), 10 deletions(-)

diff --git a/.gitignore b/.gitignore
index fa4b95f..b2a1796 100644
--- a/.gitignore
+++ b/.gitignore
@@ -181,6 +181,7 @@ WeiboSentiment_Finetuned/GPT2-Lora/models/
 WeiboSentiment_Finetuned/GPT2-AdapterTuning/models/
 WeiboSentiment_Finetuned/BertChinese-Lora/models/
 WeiboSentiment_LLM/models/
+WeiboSentiment_Finetuned/BertChinese-Lora/model/
 
 # LoRA 和 Adapter 权重
 */adapter_model.safetensors
diff --git a/WeiboSentiment_Finetuned/BertChinese-Lora/README.md b/WeiboSentiment_Finetuned/BertChinese-Lora/README.md
index c164ef3..a85c403 100644
--- a/WeiboSentiment_Finetuned/BertChinese-Lora/README.md
+++ b/WeiboSentiment_Finetuned/BertChinese-Lora/README.md
@@ -64,8 +64,15 @@ print("正面情感" if prediction == 1 else "负面情感")
 - `predict_pipeline.py`: 使用pipeline方式的预测程序  
 - `README.md`: 使用说明
 
+## 模型存储
+
+- 首次运行时会自动下载模型到当前目录的 `model` 文件夹
+- 后续运行会直接从本地加载，无需重复下载
+- 模型大小约400MB，首次下载需要网络连接
+
 ## 注意事项
 
 - 首次运行时会自动下载模型，需要网络连接
-- 模型大小约400MB，下载可能需要一些时间
-- 支持GPU加速，会自动检测可用设备
\ No newline at end of file
+- 模型会保存到当前目录，方便后续使用
+- 支持GPU加速，会自动检测可用设备
+- 如需清理模型文件，删除 `model` 文件夹即可
\ No newline at end of file
diff --git a/WeiboSentiment_Finetuned/BertChinese-Lora/predict.py b/WeiboSentiment_Finetuned/BertChinese-Lora/predict.py
index e8f90ca..627d879 100644
--- a/WeiboSentiment_Finetuned/BertChinese-Lora/predict.py
+++ b/WeiboSentiment_Finetuned/BertChinese-Lora/predict.py
@@ -16,11 +16,25 @@ def main():
     
     # 使用HuggingFace预训练模型
     model_name = "wsqstar/GISchat-weibo-100k-fine-tuned-bert"
+    local_model_path = "./model"
     
     try:
-        # 加载模型和分词器
-        tokenizer = AutoTokenizer.from_pretrained(model_name)
-        model = AutoModelForSequenceClassification.from_pretrained(model_name)
+        # 检查本地是否已有模型
+        import os
+        if os.path.exists(local_model_path):
+            print("从本地加载模型...")
+            tokenizer = AutoTokenizer.from_pretrained(local_model_path)
+            model = AutoModelForSequenceClassification.from_pretrained(local_model_path)
+        else:
+            print("首次使用，正在下载模型到本地...")
+            # 下载并保存到本地
+            tokenizer = AutoTokenizer.from_pretrained(model_name)
+            model = AutoModelForSequenceClassification.from_pretrained(model_name)
+            
+            # 保存到本地
+            tokenizer.save_pretrained(local_model_path)
+            model.save_pretrained(local_model_path)
+            print(f"模型已保存到: {local_model_path}")
         
         # 设置设备
         device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
diff --git a/WeiboSentiment_Finetuned/BertChinese-Lora/predict_pipeline.py b/WeiboSentiment_Finetuned/BertChinese-Lora/predict_pipeline.py
index 4233697..b8e2482 100644
--- a/WeiboSentiment_Finetuned/BertChinese-Lora/predict_pipeline.py
+++ b/WeiboSentiment_Finetuned/BertChinese-Lora/predict_pipeline.py
@@ -15,13 +15,36 @@ def main():
     
     # 使用pipeline方式 - 更简单
     model_name = "wsqstar/GISchat-weibo-100k-fine-tuned-bert"
+    local_model_path = "./model"
     
     try:
-        classifier = pipeline(
-            "text-classification", 
-            model=model_name,
-            return_all_scores=True
-        )
+        # 检查本地是否已有模型
+        import os
+        if os.path.exists(local_model_path):
+            print("从本地加载模型...")
+            classifier = pipeline(
+                "text-classification", 
+                model=local_model_path,
+                return_all_scores=True
+            )
+        else:
+            print("首次使用，正在下载模型到本地...")
+            # 先下载模型
+            from transformers import AutoTokenizer, AutoModelForSequenceClassification
+            tokenizer = AutoTokenizer.from_pretrained(model_name)
+            model = AutoModelForSequenceClassification.from_pretrained(model_name)
+            
+            # 保存到本地
+            tokenizer.save_pretrained(local_model_path)
+            model.save_pretrained(local_model_path)
+            print(f"模型已保存到: {local_model_path}")
+            
+            # 使用本地模型创建pipeline
+            classifier = pipeline(
+                "text-classification", 
+                model=local_model_path,
+                return_all_scores=True
+            )
         print("模型加载成功!")
         
     except Exception as e: