遇到报错，求助大佬帮忙看看啥问题

godlike356

星级打分

• 1
• 2
• 3
• 4
• 5

平均分:NAN  参与人数:0  我的评分:未评

Error: Graph execution error:


Detected at node 'Pad_13' defined at (most recent call last):
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\python-3.6.8\lib\threading.py", line 885, in _bootstrap
      self._bootstrap_inner()
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\python-3.6.8\lib\threading.py", line 917, in _bootstrap_inner
      self.run()
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\python-3.6.8\lib\threading.py", line 865, in run
      self._target(*self._args, **self._kwargs)
    File "<frozen mainscripts.Trainer>", line 58, in trainerThread
    File "<frozen models.ModelBase>", line 207, in __init__
    File "<frozen models.Model_SAEHD.Model>", line 423, in on_initialize
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\DeepFaceLab\core\leras\models\ModelBase.py", line 117, in __call__
      return self.forward(*args, **kwargs)
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\DeepFaceLab\core\leras\archis\DeepFakeArchi.py", line 114, in forward
      x = self.down3(x)
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\DeepFaceLab\core\leras\models\ModelBase.py", line 117, in __call__
      return self.forward(*args, **kwargs)
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\DeepFaceLab\core\leras\archis\DeepFakeArchi.py", line 43, in forward
      x = self.conv1(x)
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\DeepFaceLab\core\leras\layers\LayerBase.py", line 14, in __call__
      return self.forward(*args, **kwargs)
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\DeepFaceLab\core\leras\layers\Conv2D.py", line 87, in forward
      x = tf.pad (x, padding, mode='CONSTANT')
Node: 'Pad_13'
Detected at node 'Pad_13' defined at (most recent call last):
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\python-3.6.8\lib\threading.py", line 885, in _bootstrap
      self._bootstrap_inner()
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\python-3.6.8\lib\threading.py", line 917, in _bootstrap_inner
      self.run()
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\python-3.6.8\lib\threading.py", line 865, in run
      self._target(*self._args, **self._kwargs)
    File "<frozen mainscripts.Trainer>", line 58, in trainerThread
    File "<frozen models.ModelBase>", line 207, in __init__
    File "<frozen models.Model_SAEHD.Model>", line 423, in on_initialize
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\DeepFaceLab\core\leras\models\ModelBase.py", line 117, in __call__
      return self.forward(*args, **kwargs)
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\DeepFaceLab\core\leras\archis\DeepFakeArchi.py", line 114, in forward
      x = self.down3(x)
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\DeepFaceLab\core\leras\models\ModelBase.py", line 117, in __call__
      return self.forward(*args, **kwargs)
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\DeepFaceLab\core\leras\archis\DeepFakeArchi.py", line 43, in forward
      x = self.conv1(x)
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\DeepFaceLab\core\leras\layers\LayerBase.py", line 14, in __call__
      return self.forward(*args, **kwargs)
    File "D:\yonghu\Desktop\DFL\DeepFaceLab  mzhh\_internal\DeepFaceLab\core\leras\layers\Conv2D.py", line 87, in forward
      x = tf.pad (x, padding, mode='CONSTANT')
Node: 'Pad_13'
2 root error(s) found.
  (0) RESOURCE_EXHAUSTED: OOM when allocating tensor with shape[4,76,76,160] and type float on /job:localhost/replica:0/task:0/device:GPU:0 by allocator GPU_0_bfc
         [[{{node Pad_13}}]]
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. This isn't available when running in Eager mode.


         [[concat_7/concat/_943]]
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. This isn't available when running in Eager mode.


  (1) RESOURCE_EXHAUSTED: OOM when allocating tensor with shape[4,76,76,160] and type float on /job:localhost/replica:0/task:0/device:GPU:0 by allocator GPU_0_bfc
         [[{{node Pad_13}}]]
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. This isn't available when running in Eager mode.


0 successful operations.
0 derived errors ignored.

lknet

https://dfldata.cc/search.php?mo ... bmit=yes&kw=oom
还有就是软件路径尽量不要用中文和空格

小飞侠

爆显存了

yxyao

现在咱有DeepSeek了，不需要翻墙用ChatGPT了，所以遇到问题可以直接贴给DeepSeek，它的回答可以说又快又好

godlike356

lknet 发表于 2025-2-16 10:55
https://dfldata.cc/search.php?mod=forum&searchid=51&orderby=lastpost&ascdesc=desc&searchsubmit=yes&k ...

谢谢谢谢大佬帮助，换个丹用上了