Upload layer_diff_dataset/make_json copy 4.py with huggingface_hub

layer_diff_dataset/make_json copy 4.py

@@ -0,0 +1,38 @@
+import json
+import os
+# 读取im_rgba.txt中的内容作为bg prompt写入json
+
+# folder_path = '/mnt/workspace/workgroup/sihui.jsh/layer_diff_dataset/try/im'
+# folder_path_1 = '/mnt/workspace/workgroup/sihui.jsh/layer_diff_dataset/train/im'
+
+# file_list = os.listdir(folder_path)
+# file_list = [os.path.join(folder_path_1,i) for i in file_list if i.endswith('.jpg')]
+# file_list.sort()
+# print(file_list)
+# 读取原始JSON文件
+data_folder = '/mnt/workspace/workgroup/sihui.jsh/layer_diff_dataset/train'
+with open(os.path.join(data_folder,'im_rgba.json'), 'r') as file:
+    data = json.load(file)
+
+captions_dict = {}
+with open(os.path.join(data_folder, 'im_rgba.txt'), 'r') as f:
+    for line in f.readlines():
+        image_name, caption = line.strip().split('\t')
+        captions_dict[image_name] = caption
+# data_try = []
+# # 修改每个字典的"images"和"images_rgba"字段
+# for item in data:
+#     if item["images"] not in file_list:
+#         data_try.append(item)
+# print(data_try)
+for item in data:
+    image_name = item["images"].split('/')[-1]
+    # item['prompt'] = class_name + ',' + item['prompt']
+    item['bg_prompt'] = captions_dict[image_name]
+    # print(item)
+    # break
+# 写入新的JSON文件
+with open('/mnt/workspace/workgroup/sihui.jsh/layer_diff_dataset/train/im_rgba_bg.json', 'w') as file:
+    json.dump(data, file, indent=4)
+
+print("New JSON file has been created successfully.")