top | item 29577435

(no title)

ritz_labringue | 4 years ago

Doesn't it cost hundreds of thousands of dollars just to train GPT-3 ? If so, that seems like a good reason to use a "managed" GPT-3.

discuss

order

dave_sullivan|4 years ago

Yes, but they didn't release the model after training and you can't take your weights with you if you finetune their model.

GPT Neo was trained at similar expense, and they released the weights. Use that.

damvigilante|4 years ago

First part is correct, the second part is not. GPT Neo is a 2.7B param model, the largest GPT is 175B (they have various flavours, up to 175B). I appreciate the sentiment and what ElutherAI is doing with GPT Neo, but there is no open source equivlenet of the full GPT-3 available for the public to use. Hopefuly it's just a matter of time.

ritz_labringue|4 years ago

Thanks ! Didn't know that. Isn't it also very expensive to run ?