Multiple Dispatch In __torch_function__ PyTorch Developer podcast

Artwork

Tech PyTorch Edward Yang Team PyTorch Deep Learning Machine Learning

Innhold levert av PyTorch, Edward Yang, and Team PyTorch. Alt podcastinnhold, inkludert episoder, grafikk og podcastbeskrivelser, lastes opp og leveres direkte av PyTorch, Edward Yang, and Team PyTorch eller deres podcastplattformpartner. Hvis du tror at noen bruker det opphavsrettsbeskyttede verket ditt uten din tillatelse, kan du følge prosessen skissert her https://no.player.fm/legal.

PyTorch Developer Podcast « »
Multiple dispatch in __torch_function__

4+ y ago 14:20

Del

MP3•Episoder hjem

Innhold levert av PyTorch, Edward Yang, and Team PyTorch. Alt podcastinnhold, inkludert episoder, grafikk og podcastbeskrivelser, lastes opp og leveres direkte av PyTorch, Edward Yang, and Team PyTorch eller deres podcastplattformpartner. Hvis du tror at noen bruker det opphavsrettsbeskyttede verket ditt uten din tillatelse, kan du følge prosessen skissert her https://no.player.fm/legal.

Python is a single dispatch OO language, but there are some operations such as binary magic methods which implement a simple form of multiple dispatch. torch_function__ (through its Numpy predecessor __array_function) generalizes this mechanism so that invocations of torch.add with different subclasses work properly. This podcast describes how this mechanism works and how it can be used (in an unconventional way) to build composable subclasses ala JAX in functorch.

Further reading:

This podcast in written form https://dev-discuss.pytorch.org/t/functorch-levels-as-dynamically-allocated-classes/294
Multiple dispatch resolution rules in the RFC https://github.com/pytorch/rfcs/blob/master/RFC-0001-torch-function-for-methods.md#process-followed-during-a-functionmethod-call

… continue reading

83 episoder

#Tech #PyTorch #Edward Yang #Team PyTorch #Deep Learning #Machine Learning

Artwork

Multiple dispatch in __torch_function__

PyTorch Developer Podcast

26 subscribers

published 4+ y ago

Del

MP3•Episoder hjem

Innhold levert av PyTorch, Edward Yang, and Team PyTorch. Alt podcastinnhold, inkludert episoder, grafikk og podcastbeskrivelser, lastes opp og leveres direkte av PyTorch, Edward Yang, and Team PyTorch eller deres podcastplattformpartner. Hvis du tror at noen bruker det opphavsrettsbeskyttede verket ditt uten din tillatelse, kan du følge prosessen skissert her https://no.player.fm/legal.

Python is a single dispatch OO language, but there are some operations such as binary magic methods which implement a simple form of multiple dispatch. torch_function__ (through its Numpy predecessor __array_function) generalizes this mechanism so that invocations of torch.add with different subclasses work properly. This podcast describes how this mechanism works and how it can be used (in an unconventional way) to build composable subclasses ala JAX in functorch.

Further reading:

This podcast in written form https://dev-discuss.pytorch.org/t/functorch-levels-as-dynamically-allocated-classes/294
Multiple dispatch resolution rules in the RFC https://github.com/pytorch/rfcs/blob/master/RFC-0001-torch-function-for-methods.md#process-followed-during-a-functionmethod-call

… continue reading

83 episoder

#Tech #PyTorch #Edward Yang #Team PyTorch #Deep Learning #Machine Learning

Tất cả các tập

×

Velkommen til Player FM!

Player FM scanner netter for høykvalitets podcaster som du kan nyte nå. Det er den beste podcastappen og fungerer på Android, iPhone og internett. Registrer deg for å synkronisere abonnement på flere enheter.

Lytt til 500+ tema

Lytt til dette showet mens du utforsker