跪求各位大哥，我遇到一个问题。

hahaha253

星级打分

• 1
• 2
• 3
• 4
• 5

平均分:NAN  参与人数:0  我的评分:未评

各位老哥，我今天训练了一下deepfacelab，以前3060训练很好的模型。我只是想确认一下，但是训练了几十秒。我就退出来了。我再合成的时候。我发现模型废了，全变成类似马赛克的东西。各位老哥知道是什么情况吗。
换了新卡5070ti以后，无论是切脸，还是合成，速度都比3060的时候慢了很多。我就是为了玩AI。专门买了一个显存比较大的。
但是50系以后出了很多问题。
大概变成了这个样子。

hahaha253

raceback (most recent call last):
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\python-3.6.8\lib\site-packages\tensorflow\python\client\session.py", line 1375, in _do_call
    return fn(*args)
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\python-3.6.8\lib\site-packages\tensorflow\python\client\session.py", line 1360, in _run_fn
    target_list, run_metadata)
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\python-3.6.8\lib\site-packages\tensorflow\python\client\session.py", line 1453, in _call_tf_sessionrun
    run_metadata)
tensorflow.python.framework.errors_impl.InternalError: 2 root error(s) found.
  (0) Internal: Failed to load in-memory CUBIN: CUDA_ERROR_NO_BINARY_FOR_GPU: no kernel image is available for execution on the device
         [[{{node Abs_1}}]]
         [[Sigmoid_10/_799]]
  (1) Internal: Failed to load in-memory CUBIN: CUDA_ERROR_NO_BINARY_FOR_GPU: no kernel image is available for execution on the device
         [[{{node Abs_1}}]]
0 successful operations.
0 derived errors ignored.

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\mainscripts\Merger.py", line 208, in main
    subprocess_count       = subprocess_count,
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\core\joblib\SubprocessorBase.py", line 268, in run
    if self.on_tick() and all ([cli.state == 0 for cli in self.clis]):
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\merger\InteractiveMergerSubprocessor.py", line 361, in on_tick
    io.process_messages()
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\core\interact\interact.py", line 186, in process_messages
    func()
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\core\joblib\MPClassFuncOnDemand.py", line 23, in io_callback
    self.s2c.put ( self.class_func (*func_args, **func_kwargs) )
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\facelib\XSegNet.py", line 102, in extract
    result = np.clip ( self.net_run(input_image), 0, 1.0 )
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\facelib\XSegNet.py", line 60, in net_run
    return nn.tf_sess.run ( [pred], feed_dict={self.input_t :input_np})[0]
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\python-3.6.8\lib\site-packages\tensorflow\python\client\session.py", line 968, in run
    run_metadata_ptr)
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\python-3.6.8\lib\site-packages\tensorflow\python\client\session.py", line 1191, in _run
    feed_dict_tensor, options, run_metadata)
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\python-3.6.8\lib\site-packages\tensorflow\python\client\session.py", line 1369, in _do_run
    run_metadata)
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_internal\python-3.6.8\lib\site-packages\tensorflow\python\client\session.py", line 1394, in _do_call
    raise type(e)(node_def, op, message)
tensorflow.python.framework.errors_impl.InternalError: 2 root error(s) found.
  (0) Internal: Failed to load in-memory CUBIN: CUDA_ERROR_NO_BINARY_FOR_GPU: no kernel image is available for execution on the device
         [[node Abs_1 (defined at \DeepFaceLab\core\leras\layers\FRNorm2D.py:35) ]]
         [[Sigmoid_10/_799]]
  (1) Internal: Failed to load in-memory CUBIN: CUDA_ERROR_NO_BINARY_FOR_GPU: no kernel image is available for execution on the device
         [[node Abs_1 (defined at \DeepFaceLab\core\leras\layers\FRNorm2D.py:35) ]]
0 successful operations.
0 derived errors ignored.

Errors may have originated from an input operation.
Input Source operations connected to node Abs_1:
XSeg/conv02/frn/eps/read (defined at \DeepFaceLab\core\leras\layers\FRNorm2D.py:22)

Input Source operations connected to node Abs_1:
XSeg/conv02/frn/eps/read (defined at \DeepFaceLab\core\leras\layers\FRNorm2D.py:22)

Original stack trace for 'Abs_1':
  File "\DeepFaceLab\main.py", line 336, in <module>
    arguments.func(arguments)
  File "\DeepFaceLab\main.py", line 165, in process_merge
    cpu_only               = arguments.cpu_only)
  File "\DeepFaceLab\mainscripts\Merger.py", line 208, in main
    subprocess_count       = subprocess_count,
  File "\DeepFaceLab\core\joblib\SubprocessorBase.py", line 268, in run
    if self.on_tick() and all ([cli.state == 0 for cli in self.clis]):
  File "\DeepFaceLab\merger\InteractiveMergerSubprocessor.py", line 361, in on_tick
    io.process_messages()
  File "\DeepFaceLab\core\interact\interact.py", line 186, in process_messages
    func()
  File "\DeepFaceLab\core\joblib\MPClassFuncOnDemand.py", line 22, in io_callback
    self.class_func = getattr( self.class_handle(**self.class_kwargs), self.class_func_name)
  File "\DeepFaceLab\facelib\XSegNet.py", line 57, in __init__
    _, pred = self.model(self.input_t)
  File "\DeepFaceLab\core\leras\models\ModelBase.py", line 117, in __call__
    return self.forward(*args, **kwargs)
  File "\DeepFaceLab\core\leras\models\XSeg.py", line 97, in forward
    x = x0 = self.conv02(x)
  File "\DeepFaceLab\core\leras\models\ModelBase.py", line 117, in __call__
    return self.forward(*args, **kwargs)
  File "\DeepFaceLab\core\leras\models\XSeg.py", line 16, in forward
    x = self.frn(x)
  File "\DeepFaceLab\core\leras\layers\LayerBase.py", line 14, in __call__
    return self.forward(*args, **kwargs)
  File "\DeepFaceLab\core\leras\layers\FRNorm2D.py", line 35, in forward
    x = x * ( 1.0/tf.sqrt(nu2 + tf.abs(self.eps) ) )
  File "\python-3.6.8\lib\site-packages\tensorflow\python\util\dispatch.py", line 201, in wrapper
    return target(*args, **kwargs)
  File "\python-3.6.8\lib\site-packages\tensorflow\python\ops\math_ops.py", line 401, in abs
    return gen_math_ops._abs(x, name=name)
  File "\python-3.6.8\lib\site-packages\tensorflow\python\ops\gen_math_ops.py", line 55, in _abs
    "Abs", x=x, name=name)
  File "\python-3.6.8\lib\site-packages\tensorflow\python\framework\op_def_library.py", line 750, in _apply_op_helper
    attrs=attr_protos, op_def=op_def)
  File "\python-3.6.8\lib\site-packages\tensorflow\python\framework\ops.py", line 3536, in _create_op_internal
    op_def=op_def)
  File "\python-3.6.8\lib\site-packages\tensorflow\python\framework\ops.py", line 1990, in __init__
    self._traceback = tf_stack.extract_stack()

小飞侠

50系显卡用30系列训练不兼容，用dml训练没事

注册用户

hahaha253 发表于 2025-4-10 00:17
raceback (most recent call last):
  File "H:\换脸\rocklisoft\DeepFaceLab_NVIDIA_RTX3000_series\_inte ...

路径，不要有中文

bg2kcg

注册用户 发表于 2025-4-10 00:53
路径，不要有中文

正解

小王子

5070ti的话可能还没有用DirectX 12版本的快

dada_sky

这50系显卡真的是一难难尽啊