Unbacked SymInts PyTorch Developer podcast

Artwork

Tech PyTorch Edward Yang Team PyTorch Deep Learning Machine Learning

Conteúdo fornecido por PyTorch, Edward Yang, and Team PyTorch. Todo o conteúdo do podcast, incluindo episódios, gráficos e descrições de podcast, é carregado e fornecido diretamente por PyTorch, Edward Yang, and Team PyTorch ou por seu parceiro de plataforma de podcast. Se você acredita que alguém está usando seu trabalho protegido por direitos autorais sem sua permissão, siga o processo descrito aqui https://pt.player.fm/legal.

PyTorch Developer Podcast « »
Unbacked SymInts

1+ y ago 21:31

Compartilhar

MP3•Home de episódios

Conteúdo fornecido por PyTorch, Edward Yang, and Team PyTorch. Todo o conteúdo do podcast, incluindo episódios, gráficos e descrições de podcast, é carregado e fornecido diretamente por PyTorch, Edward Yang, and Team PyTorch ou por seu parceiro de plataforma de podcast. Se você acredita que alguém está usando seu trabalho protegido por direitos autorais sem sua permissão, siga o processo descrito aqui https://pt.player.fm/legal.

This podcast goes over the basics of unbacked SymInts. You might want to listen to this one before listening to https://pytorch-dev-podcast.simplecast.com/episodes/zero-one-specialization Some questions we answer (h/t from Gregory Chanan):

- Are unbacked symints only for export? Because otherwise I could just break / wait for the actual size. But maybe I can save some retracing / graph breaks perf if I have them too? So the correct statement is "primarily" for export?

- Why am I looking into the broadcasting code at all? Naively, I would expect the export graph to be just a list of ATen ops strung together. Why do I recurse that far down? Why can't I annotate DONT_TRACE_ME_BRO?

- How does 0/1 specialization fit into this? I understand we may want to 0/1 specialize in a dynamic shape regime in "eager" mode (is there a better term?), but that doesn't seem to matter for export?

- So far we've mainly been talking about how to handle our own library code. There is a worry about pushing complicated constraints downstream, similar to torchscript. What constraints does this actually push?

… continue reading

82 episódios

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

Artwork

Unbacked SymInts

PyTorch Developer Podcast

33 subscribers

published 1+ y ago

Compartilhar

MP3•Home de episódios

Conteúdo fornecido por PyTorch, Edward Yang, and Team PyTorch. Todo o conteúdo do podcast, incluindo episódios, gráficos e descrições de podcast, é carregado e fornecido diretamente por PyTorch, Edward Yang, and Team PyTorch ou por seu parceiro de plataforma de podcast. Se você acredita que alguém está usando seu trabalho protegido por direitos autorais sem sua permissão, siga o processo descrito aqui https://pt.player.fm/legal.

This podcast goes over the basics of unbacked SymInts. You might want to listen to this one before listening to https://pytorch-dev-podcast.simplecast.com/episodes/zero-one-specialization Some questions we answer (h/t from Gregory Chanan):

- Are unbacked symints only for export? Because otherwise I could just break / wait for the actual size. But maybe I can save some retracing / graph breaks perf if I have them too? So the correct statement is "primarily" for export?

- Why am I looking into the broadcasting code at all? Naively, I would expect the export graph to be just a list of ATen ops strung together. Why do I recurse that far down? Why can't I annotate DONT_TRACE_ME_BRO?

- How does 0/1 specialization fit into this? I understand we may want to 0/1 specialize in a dynamic shape regime in "eager" mode (is there a better term?), but that doesn't seem to matter for export?

- So far we've mainly been talking about how to handle our own library code. There is a worry about pushing complicated constraints downstream, similar to torchscript. What constraints does this actually push?

… continue reading

82 episódios

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

Todos os episódios

×

Bem vindo ao Player FM!

O Player FM procura na web por podcasts de alta qualidade para você curtir agora mesmo. É o melhor app de podcast e funciona no Android, iPhone e web. Inscreva-se para sincronizar as assinaturas entre os dispositivos.

Ouça a 500+ tópicos