How to swap files and making sure enough storage for HELLO AI WORLD project?

I have tried swapping file using “fallocate” command yet it says text file busy. How am i going to reduce storage? Is there any solution to this? I am on low storage memory now even though im using an SDCard with 128GB.


Do you mean you try to run this:
GitHub - dusty-nv/jetson-inference: Hello AI World guide to deploying deep-learning inference networks and deep vision primitives with TensorRT and NVIDIA Jetson.

And please share which Jetpack release you are using.

yes correct, i’m doing the hello AI world. Anyway, i’m using jetpack release 4.6.2
hoping i can swap files again as it says fallocate failed : text file busy.

thank you!

Hi @n.syafiqahme, was this file already created as swap? If so, you don’t need to use fallocate on it again. Do you have enough disk space that df -H shows?

hi yes indeed! i see but when i’m running on the classification for training, jetson shuts down and it seems like my storage has a lot of storage left. By the way, i can’t seem to proceed with the retraining cat/dog as torch module can’t be found suddenly. May i know what is the error?

One important detail is that the memory used for the GPU must be physical RAM. Virtual RAM can’t be used for many GPU uses, e.g., training. If other user space content is released from physical RAM by going to swap/virtual memory, then that would indirectly help the GPU. If you are using too much RAM though for the GPU, then swap probably won’t help.

1 Like

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.