diff --git a/dsLightRag/Config/__pycache__/Config.cpython-310.pyc b/dsLightRag/Config/__pycache__/Config.cpython-310.pyc index 76e8086d..8c41d477 100644 Binary files a/dsLightRag/Config/__pycache__/Config.cpython-310.pyc and b/dsLightRag/Config/__pycache__/Config.cpython-310.pyc differ diff --git a/dsLightRag/T1_Train.py b/dsLightRag/T1_Train.py index fab932d8..e1e8cb83 100644 --- a/dsLightRag/T1_Train.py +++ b/dsLightRag/T1_Train.py @@ -37,9 +37,12 @@ tasks = [ # { # 数学 # "workspace": "Math", "docx_name": "Math.docx", # }, - { # 史记 - "workspace": "ShiJi", "docx_name": "少年读史记张嘉骅.docx", - }, + # { # 史记 + # "workspace": "ShiJi", "docx_name": "少年读史记张嘉骅.docx", + # }, + { # 长春市一批次高中学校介绍 + "workspace": "ChangChun", "docx_name": "长春市一批次高中学校介绍.docx", + } ] for task in tasks: task["docx_path"] = "./static/Txt/" + task["docx_name"] # 3、文档路径 python是按引用传递的& diff --git a/dsLightRag/Test/1.png b/dsLightRag/Test/1.png deleted file mode 100644 index 9d4065f8..00000000 Binary files a/dsLightRag/Test/1.png and /dev/null differ diff --git a/dsLightRag/Test/TestOcrTable.py b/dsLightRag/Test/TestOcrTable.py deleted file mode 100644 index 5c5dac71..00000000 --- a/dsLightRag/Test/TestOcrTable.py +++ /dev/null @@ -1,57 +0,0 @@ -import base64 -import json -import urllib.request - -ENCODING = 'utf-8' - - -def get_img_base64(img_file): - with open(img_file, 'rb') as infile: - s = infile.read() - return base64.b64encode(s).decode(ENCODING) - - -def predict(url, appcode, img_base64, kv_configure): - param = {} - param['image'] = img_base64 - if kv_configure is not None: - param['configure'] = json.dumps(kv_configure) - body = json.dumps(param) - data = bytes(body, "utf-8") - - headers = {'Authorization': 'APPCODE %s' % appcode} - request = urllib.request.Request(url=url, headers=headers, data=data) - try: - response = urllib.request.urlopen(request, timeout=10) - return response.code, response.headers, response.read() - except urllib.request.HTTPError as e: - return e.code, e.headers, e.read() - - -def demo(): - AppKey = '204917800' - AppSecret = 'rmIcdp5is0hM2e4KBC3Ned9aFGBYGU2F' - AppCode = '3a9b1eeacf6a4a7c9b6f3157feb61fa7' - - url = 'https://form.market.alicloudapi.com/api/predict/ocr_table_parse' - img_file = '1.png' - # configure = {'side': 'face'} - # 如果没有configure字段，configure设为None - # configure = None - configure = {"format": "html", "dir_assure": False, "line_less": True} - - img_base64data = get_img_base64(img_file) - stat, header, content = predict(url, AppCode, img_base64data, configure) - if stat != 200: - print('Http status code: ', stat) - print('Error msg in header: ', header['x-ca-error-message'] if 'x-ca-error-message' in header else '') - print('Error msg in body: ', content) - exit() - result_str = content - - print(result_str.decode(ENCODING)) - # result = json.loads(result_str) - - -if __name__ == '__main__': - demo() diff --git a/dsLightRag/Util/__pycache__/DocxUtil.cpython-310.pyc b/dsLightRag/Util/__pycache__/DocxUtil.cpython-310.pyc index 174403bf..56ac0dae 100644 Binary files a/dsLightRag/Util/__pycache__/DocxUtil.cpython-310.pyc and b/dsLightRag/Util/__pycache__/DocxUtil.cpython-310.pyc differ diff --git a/dsLightRag/Util/__pycache__/LightRagUtil.cpython-310.pyc b/dsLightRag/Util/__pycache__/LightRagUtil.cpython-310.pyc index 015d4bde..d89e8018 100644 Binary files a/dsLightRag/Util/__pycache__/LightRagUtil.cpython-310.pyc and b/dsLightRag/Util/__pycache__/LightRagUtil.cpython-310.pyc differ diff --git a/dsLightRag/static/Txt/2024长春43所高中录取分数线.md b/dsLightRag/static/Txt/2024长春43所高中录取分数线.md new file mode 100644 index 00000000..35c696c2 --- /dev/null +++ b/dsLightRag/static/Txt/2024长春43所高中录取分数线.md @@ -0,0 +1,35 @@ +| | | | | +| --------- | -------------------- | ---- | ----- | +| 一批次666 | 东北师范大学附属中学 | 统招 | 721.6 | +| | 长春市十一高 | 统招 | 710.1 | +| | 长春市二中 | 统招 | 697 | +| | 吉林省实验中学 | 统招 | 690.1 | +| | 长春外国语 | 统招 | 689.6 | +| | 长春市第二实验 | 统招 | 682.1 | +| | 长春市实验 | 统招 | 676.8 | +| | 长春市六中 | 统招 | 666 | +| | 长春汽车厂六中 | 统招 | 666 | +| 二批次586 | 师大青华 | 私立 | 705.6 | +| | 吉大 | 私立 | 701.1 | +| | 二中精致 | 私立 | 687.3 | +| | 八中 | 统招 | 677 | +| | 五中 | 统招 | 664 | +| | 省实验繁荣 | 私立 | 662.3 | +| | 希望 | 统招 | 651.4 | +| | 师大净月慧泽 | 私立 | 651 | +| | 北师大博硕 | 私立 | 646.9 | +| | 十七中 | 统招 | 643.3 | +| | 汽三 | 统招 | 637.6 | +| | 一中 | 统招 | 633.5 | +| | 北湖 | 私立 | 630.5 | +| | 二十九中 | 统招 | 624.8 | +| | 养正 | 统招 | 623.3 | +| | 十中 | 统招 | 620.5 | +| | 文理 | 私立 | 616.4 | +| | 七中 | 统招 | 614.4 | +| | 朝中 | 统招 | 613.8 | +| | 一三七中学 | 统招 | 611.8 | +| | 解放 | 私立 | 608.2 | +| | 力旺 | 私立 | 586 | +| | 日章学园 | 私立 | 586 | +| | 清浦 | 私立 | 586 | \ No newline at end of file diff --git a/dsLightRag/static/Txt/长春市2025年中考各批次录取最低控制线.txt b/dsLightRag/static/Txt/长春市2025年中考各批次录取最低控制线.txt new file mode 100644 index 00000000..c933e71b --- /dev/null +++ b/dsLightRag/static/Txt/长春市2025年中考各批次录取最低控制线.txt @@ -0,0 +1,7 @@ +长春市2025年中考各批次录取最低控制线确定如下： +第一批次城区普通高中 689分 +第二批次城区普通高中 590分 +第三批次城区普通高中 573分 +中等职业学校的综合高中班录取不低于第三批次城区普通高中最低控制线下60分。 +来源：长春市教育考试院 + diff --git a/dsLightRag/static/Txt/长春市一批次高中学校介绍.docx b/dsLightRag/static/Txt/长春市一批次高中学校介绍.docx new file mode 100644 index 00000000..b7f7ed94 Binary files /dev/null and b/dsLightRag/static/Txt/长春市一批次高中学校介绍.docx differ