分布式训练cannot assign a device for operation PyFunc #1449
-
在跑learning/examples/tf/graphsage/dist_train.py 时training过程报cannot assign device for operation PyFunc: {{node PyFunc} was explicitly assigned to /job:localhost/replica:0/task:0 but available devices are [/job:worker/replica:0:task:0/device:CPU:0]. Make sure the device specification refers to a valid device. |
Beta Was this translation helpful? Give feedback.
Replies: 5 comments
-
Please try to seek to tensorflow from similar issue reports (e.g., tensorflow/tensorflow#31318 ) and see if it is helpful for your case. |
Beta Was this translation helpful? Give feedback.
-
我参照相关issue尝试了一些办法还是没有解决。 |
Beta Was this translation helpful? Give feedback.
-
我们的CI中测试过的是tensorflow 2.5.2 |
Beta Was this translation helpful? Give feedback.
-
谢谢,我使用2.2.0解决了这个问题 |
Beta Was this translation helpful? Give feedback.
-
Close as resolved. |
Beta Was this translation helpful? Give feedback.
谢谢,我使用2.2.0解决了这个问题
在import tensorflow之前我加了os.environ.pop('http_proxy'),os.environ.pop('https_proxy')