If I understood correctly, then you should simply enter the instance via the “Open V2 (Beta)” option which starts a Jupyter Lab environment where you’ll have the terminal to use. I hope this works for you.
Yup - got it to work. Thanks!
1 Like
Hello, I’m a new one in here.
I just wonder what I did is correct or not.
I already make a container guided by this page. And the machine is already running.
Is there something I’ve missed before?
I was following the instructions, but wasn’t able to get the code running (I tried both free and paid instance).
I can run the notebook, but the UI is different, there’s no top black header from the Gradient service (with Stop and Share buttons on the right) as it is in the guide, I get just pure Jupyter notebook.
After running the notebook, run the update command and git pull. I can run first few cells, but at the cell under Running Your First Notebook, shows this error
PIL.UnidentifiedImageError: cannot identify image file '/storage/data/oxford-iiit-pet/images/american_bulldog_95.jpg'
(that’s just last line of the stacktrace)
EDIT: It looks like the next cells run ok, despite this exception
Is it just me or does it take forever for a notebook on paperspace to open up? I typically have to wait 10 -15 mins when it used to be instantaneous.
I have the exact same problem. Sometimes it takes 30 mins to provision the server. Did you find a workaround? I might just switch to Colab.
See Paperspace start up extremely slow
Basically, make sure to clean your /notebooks folder before shutdown to reduce shutdown/spin-up time
I’ve switched over to JarvisLabs. Fast spin up.
1 Like
Unfortunately, non-free - their cheaper GPU is the RTX5000, at $0.49/hr. The moment you’re considering a non-free option, it’s very hard to beat the AWS spot price of their g4dn.xlarge instances (T4 GPU) - that’s about 66% the speed of an RTX5000 for training, at $0.16/hr, so basically JarvisLabs is twice the cost-per-epoch. Oh, and once you’ve configured your AMI, spin-up is instant.
*correction - JarvisLabs offer a 20% discount for FastAI on the RTX5000 (thanks guys!) - so they’re only $0.39/hr. That’s still 50% more per hour than AWS spot, but considerably cheaper than AWS on-demand.
To be clear: if you want free, do google colab or Paperspace M4000 Free-GPU. If you want to pay money, be very careful on the math by costing epochs-per-dollar. You can find an extensive list of GPUs and their relative training times for all sorts of workloads here: https://ai-benchmark.com/ranking_cpus_and_gpus_detailed.html and the cheapest global spot-price of all AWS instance types here: https://simonpbriggs.co.uk/amazonec2/
1 Like
In my experience spot instances are definitely difficult to get started with. Salamander did a great job in automating the entire setup on spot instances but later faced a lot of challenges during peak hours. There were times when you lose the instances due to the nature of Spot instances.
Definitely correct - expect to spend a day of your life learning how to set up EC2 spot instances and also EFS for persistent storage between your multiple EC2 instances (the reason you’re considering paying anything is because you’re needing clusters of GPUs, right?). But given that you’re going to be using multiple machines in parallel, that cost difference is going to more than pay for itself.
RE losing spot instances - that used to be a problem until 2018, when the spot pricing mechanism changed. Now just set your willingness to pay as higher than the spot price, and it’ll be someone else who loses their machine instead (just setting it at the on-demand price practically guarantees that you don’t get kicked off, yet you pay only the spot price as it slowly fluctuates).
I’m just starting setup, and am slightly confused. I used the default notebook (Paperspace + Fast.AI). My root on Jupyter doesn’t show coursev4 or fastbook. I have three folders (clean, images, tools) and then the course chapters starting with 01_intro. Did I miss a step?
2 Likes
I just began Chapter 3 Ethics and this is paired with the Deep Learning for Coders with fastai & Pytorch book. Maybe you setup your Notebook for the wrong course?
Having the same problem after I updated my setup and created a new paperspace notebook! There is no more persistent storage (used to be /storage folder). Anyone know what is happening?
Oh I see, instead of a separate coursev4 folder it’s now a folder within fastbook called clean. Follow-up question - how do I get access to the fastai documentation notebooks?
1 Like
I can’t get past Chapter 1 without the storage folder. Am I able to just make it myself?
No, don’t make it yourself. It is actually accessible, just not through the jupyter interface. The chapter notebooks should be able to access the folder, you just won’t see it in the interface for some reason.
1 Like
As of an hour or so ago, it still hangs me up in Chapter 1 due to that folder not being there. I skipped it as I’d been through both 1 and 2 before. I’ll post the photo sometime later today.
I am running into a ton of bad gateway errors with paperspace/gradient. Has anyone seen an uptick in this behavior? In fact the error rate is so high - I’ve been logged out every time for the past week - and each time, starting an instance - takes time - so this is really eating into my working time. If you have a workaround, I want to copy you.
OR - is there a way to take all the work I’ve done in the notebook that’s been assigned a free instance and move it to a paid instance ‘copy’?
UPDATE: So for some reason, I couldn’t change instances from free to paid. So I started a new notebook on a paid server first. Then I could see options of using free instances. I suspect this happened because my first notebook was public at inception. This is just a suspicion since that’s the only border change between files that Gradient doesn’t allow.
I think the files you keep in the /storage folder are available across all your notebook instances.
Hence, you don’t have to copy your work from the free notebook instance to the paid notebook instance.
If your notebook is taking a long time to start/stop, it probably means that you’ve kept a big file outside the /storage folder. If you move that big file to the /storage folder, then start, stop times will reduce.
Finally, files kept outside the /storage folder tend to disappear. The only way to ensure peace of mind is to put all your work (data + notebooks) inside your /storage folder.
1 Like