News for hackers. Stuff that matters.
Re:Show HN: GPU-Accelerated Inference Hosting by spullara on Tuesday July 13, 2021 @12:36AM (#27817669)
Does it need to reinitialize for each request or is there a warm start / cold start model like lambda? I don't really understand how you can charge per request.
Re:Show HN: GPU-Accelerated Inference Hosting by maldeh on Tuesday July 13, 2021 @01:18AM (#27817810)
The pricing appears to be static per model with a ceiling on the monthly request count, not charged per request.

Edit: Actually, I didn't spot the free tier of 1000 requests. I wonder how you avoid the problem of a lot of users leaving defunct/disused models running while still keeping them hot - presumably some kind of limit to the model count?

Re:Show HN: GPU-Accelerated Inference Hosting by theo31 on Tuesday July 13, 2021 @12:57AM (#27817746)
There is no cold start! We keep your service hot all the time.
Re:Show HN: GPU-Accelerated Inference Hosting by spullara on Tuesday July 13, 2021 @03:18AM (#27818353)
Well, I guess I know where I am going to host GPT-J-6B then. I don't think it is sustainable.
Re:Show HN: GPU-Accelerated Inference Hosting by marcooliv on Tuesday July 13, 2021 @08:37AM (#27820154)
How are you planning to put a gpt whatever when the service clearly have a model size limit?!
Re:Show HN: GPU-Accelerated Inference Hosting by spullara on Tuesday July 13, 2021 @08:51PM (#27828178)
The size limit is very close to allowing it (12GB vs 10GB). I imagine you can reduce it somewhat further and get it to fit.
Re:Show HN: GPU-Accelerated Inference Hosting by rubatuga on Tuesday July 13, 2021 @12:48AM (#27817711)
I guess GPU loading is quick? Like ~10 seconds?