News for hackers. Stuff that matters.

A comment on Standardizing OpenAI’s deep learning framework on PyTorch. Parent comment

Re:Standardizing OpenAI’s deep learning framework on PyTorch by sillysaurusx on Thursday January 30, 2020 @01:56PM (#22194178)
I'll bet you $20 that GPT-3 will be in Pytorch.

My point is that we probably won't be seeing more awesome projects from OpenAI written in Tensorflow. And that's unfortunate. Memory saving gradients were particularly helpful.

[ Parent | on HN ]
Re:Standardizing OpenAI’s deep learning framework on PyTorch by chillee on Thursday January 30, 2020 @02:01PM (#22194238)
I mean, this starts to sound like your complaint isn't that OpenAI is standardizing as a framework but that the framework is PyTorch :)
[ Parent | on HN ]
Re:Standardizing OpenAI’s deep learning framework on PyTorch by SpaceManNabs on Thursday January 30, 2020 @03:17PM (#22195063)
That is usually the case with framework wars.
[ Parent | on HN ]
Re:Standardizing OpenAI’s deep learning framework on PyTorch by octbash on Thursday January 30, 2020 @07:42PM (#22197738)
My counter-arguments (as a huge PyTorch fan) are:

1. GPT hasn't really been about model/architectural experimentation, just scale. GPT-2 and GPT were architecturally very similar. Scale, especially at the scale of GPT-*, is one avenue that TensorFlow does have an edge over PyTorch 2. Work on GPT-3 probably started quite a while ago.

[ Parent | on HN ]