Closed AkimfromParis closed 9 months ago
Looking - can you tell us how much storage you have left in your Studio Lab environment?
Hey @MicheleMonclova,
Thank you for your help. Please found below the response...
(studiolab) studio-lab-user@default:~$ df -h Filesystem Size Used Avail Use% Mounted on overlay 32G 20K 32G 1% / tmpfs 64M 0 64M 0% /dev tmpfs 7.8G 0 7.8G 0% /sys/fs/cgroup shm 4.0G 0 4.0G 0% /dev/shm /dev/nvme0n1p1 50G 9.8G 41G 20% /opt/.sagemakerinternal /dev/nvme1n1 25G 8.7G 17G 35% /home/studio-lab-user devtmpfs 7.8G 0 7.8G 0% /dev/tty tmpfs 7.8G 0 7.8G 0% /proc/acpi tmpfs 7.8G 0 7.8G 0% /sys/firmware
Hi, sorry for inconvenience. If you need quick solution and have an AWS account, please try the following steps.
related issue: https://github.com/aws/studio-lab-examples/issues/190
Thank you Kubo-san for your reply.
But I gave up on the credit card validation for AWS... ちょっと面倒くさいですね・・・
I will toy around again with Studio Lab when the bug is fixed. ; )
@MicheleMonclova @icoxfog417
Any news about this issue? It's sad to test on AWS studio lab and not be able to download the model...
Thank you!
Hi AkimParis, this is an issue that we have in the queue to fix. The best work around is to upload the file into S3 in your AWS account. I know you didn't want to create an account because of the credit card approval. But hopefully you understand that you will be able to leverage the AWS free tier so your credit card will not be charged.
@MicheleMonclova
Thank you for your help. I subscribed to Google Colab Pro+. The simplicity of Google won it this time on the "hacky" AWS free tier. Maybe if AWS Studio Lab provides a similar product in the near future, I might switch. : )
Have a good day!
We close this issue and continue to discuss comfortable disk access integrated to S3 at #178. Dear @AkimParis , we continue to improve Studio Lab and hope you to come back Studio Lab in future!
Hello everyone!
Describe the bug After 3 hours of training a model on PyTorch, I am unable to download the model 400 MB. I got a "Failed to connect - Server" in a text file instead of the NLP model.
To Reproduce
Expected behavior Download directly the file...
Thx!