DFL常见报错答疑搜集贴，不定时更新

bt612511

  [[node concat_1 (defined at D:\ai\GT1060\DFL_maozhihanhua_RTX2080Ti\_internal\DeepFaceLab\models\Model_SAEHD\Model.py:563) ]]
Hint: If you want to see a list of allocated tensors when OOM happens, add report_tensor_allocations_upon_oom to RunOptions for current allocation info.

这是什么意思？请大神指点！

jamesxj88

运行DeepFaceLive出现报错

RuntimeError: Unable to load kernel32 library.
我检查了windows32，kernel32.dll并没有缺失 ，我还重新安装了pyhon，还是一样的问题，看到github上好像有个朋友遇到和我一样的问题，请问有人知道是为什么吗？

wddyyd

楼主辛苦，学习了

萌新117

大佬，我是从github上面下载的代码，用Python调用的，现在报了一个这样的错误，能帮我看看吗
Traceback (most recent call last):
  File "F:\deskbak\DeepFaceLab-master\DeepFaceLab-master\mainscripts\Merger.py", line 83, in main
    if not aligned_path.exists():
           ^^^^^^^^^^^^^^^^^^^
AttributeError: 'NoneType' object has no attribute 'exists'


这是代码本体
import subprocess

def merge_faces(source_image_path, target_image_path, output_dir):
    # 定义 DeepFaceLab 的路径和命令
    deepfacelab_path = "F:/deskbak/DeepFaceLab-master/DeepFaceLab-master"
    merge_command = [
        "python",
        f"{deepfacelab_path}/main.py",
        "merge",
        "--input-dir", source_image_path,
        "--output-mask-dir", target_image_path,
        "--model-dir", f"{deepfacelab_path}/models",
        "--model", "SAEHD",
        "--output-dir", output_dir
    ]

    # 调用 DeepFaceLab 的 merge 功能
    process = subprocess.run(merge_command)

    # 检查命令是否成功执行
    if process.returncode == 0:
        print("Merge process completed successfully.")
    else:
        print("Error occurred during merge process.")

# 定义源图像路径、目标图像路径和输出路径
source_image_path = "C:/Users/admin/Desktop/byb/source_image_folder"
target_image_path = "C:/Users/admin/Desktop/byb/target_image_folder"
output_dir = "C:/Users/admin/Desktop/byb/output_folder"

# 调用 merge_faces 函数
merge_faces(source_image_path, target_image_path, output_dir)

28337

感谢楼主分享