Paperspace cloud - Comparison between Gradient and VM

[dhoa]

I’m considering using Paperspace cloud service to learn deeplearning.

I found that recently in Paperspace, they offer a new platform called Gradient. We can run directly the FastAI notebook without doing all the configuration step. Furthermore, we can swich between different type of VM to save money. For example, doing all the test and setup notebook with a CPU VM (C2). Then when finish, switch to GPU VM to run the training. Gradient also have a free charge of Storage

I think it’s very handy. But I’m wondering, with the Gradient FastAI container, because it is free charge of storage, my new data would be saved after I stop the notebook ? Can I do more work rather just the notebook in Fastai.

What is the pros and cons between Gradient and the conventional Paperspace VM ?

Thank you for your help

[dhoa]

I have asked for support from Paperspace and found that Gradient have some advantages compare to the conventional VM. The details I have posted here in case someone need it.

Paperspace setup help

As @dkobran and @gwenf mentionned above. Paperspace just have a super handy option for FastAI learner - Gradient.

I never used VM but I think Gradient have some advantages. You don’t need to wait for approving the type of machine now. You can choose any type of machine when you open a new notebook. Sometime you need to wait a little bit depends on the current capacity. For my case, it’s not too long (5 or 10 minutes)

You can save money by switching between different kind of machine. For me, I do all the configurations, data processing, …, all stuffs can do without a GPU on C2 machine and it’s 0.009$/hour. Then when I need to train model, I switch to GPU+ (P4000).

You are free of charge for storage. If you save your new data in /storage then it’s saved in the data region.

East Coast (NY2) - GPU+, P5000, C2, C7

GCP West - K80, P100, TPU

Means that if you save you data in C2 machine. You can find it again in a GPU+.

The limits are 1TB in East and 250GB in GCP West. So it’s huge in my current state.

You have a 10$ with the promo code for Fastai learner. FASTAIG5H7 Gradient_promocode

You can also accumulate your credit when sending a referral code for new user of Paperspace. So you will get 10$ and your friend 5$.

My referral code so we can both accumulate credit : XHMDK0X

Hope that helps.

[artabalt]

Hi dhoa,
I am also considering doing this, but I have not registered yet to paperspace. My question is: can you share the memory between the C2 and the P5000 instances? Or do you have to save the notebook in C2, close down C2, open P5000, load the old notebook, etc?

Thanks for letting me know, excellent tip!
Have a nice day!
arta

[dhoa]

Sorry for the late response. I haven’t used PaperSpace for a while but what I did is close C2 and open P5000. Maybe it is possible, you can ask directly for the Paperspace support or you can just register and try it. I remember we have 10$ free for registration and 5$ for invitation. You can find a thread here about people sharing there code for invitation.

Hope you will find the answer,
Nice day,

[rvn]

Gradient is good, but it lack any customizations i like themes, if you are spending hours in front of it, might as well have some customizations to my liking
VM’s let you customize at will, but switching machine types is harder as you will have to replicate the environments and such