Closed xk-huang closed 1 year ago
tensorflow occupies all the GPU memory, and in "all" mode, the following torch code will fail with OOM. So we need to use a single sub-process.
ref: https://github.com/tensorflow/tensorflow/issues/36465#issuecomment-582749350
tensorflow occupies all the GPU memory, and in "all" mode, the following torch code will fail with OOM. So we need to use a single sub-process.
ref: https://github.com/tensorflow/tensorflow/issues/36465#issuecomment-582749350