When I return to tensorflow 1.12, it will work properly.
Search found 2 matches
- Sat Sep 28, 2019 9:39 am
- Forum: Extract Support
- Topic: cuDNN failed to initialize
- Replies: 29
- Views: 34234
- Sat Sep 28, 2019 8:19 am
- Forum: Extract Support
- Topic: cuDNN failed to initialize
- Replies: 29
- Views: 34234
Could not Extract: create cudnn handle: CUDNN_STATUS_ALLOC_FAILED
This is an error that occurred when I tried to extract it from my notebook. The same installation method is feasible on desktop computers. Then I tried to check the Single process , but it still didn't run successfully. Please help me. 09/28/2019 16:11:17 MainProcess MainThread gpu_stats get_vram DE...