Re: [apache/tvm] [VOTE] Quarterly Releases RFC (Issue #11415)

2022-05-23 Thread Leandro Nunes
+1 (binding) -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm/issues/11415#issuecomment-1135144482 You are receiving this because you are subscribed to this thread. Message ID:

[apache/tvm] [VOTE] Quarterly Releases RFC (Issue #11415)

2022-05-23 Thread Andrew Reusch
This is a formal vote thread to determine committer and PMC support for [Quarterly Releases RFC](https://github.com/apache/tvm-rfcs/pull/67). Replying +1 here means you support the RFC and we should merge it. This vote thread is due to the fact that the RFC proposes a change in the way the commu

Re: [apache/tvm-rfcs] [RFC] Quarterly Releases (PR #67)

2022-05-23 Thread Andrew Reusch
Great. let's move this to a PMC voting thread. -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm-rfcs/pull/67#issuecomment-1135022404 You are receiving this because you are subscribed to this thread. Message ID:

Re: [apache/tvm-rfcs] Add Target Pre-processing RFC (PR #71)

2022-05-23 Thread Andrew Reusch
great! so I think we just need some updates to the text of the RFC based on discusison, then we are good to merge. -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm-rfcs/pull/71#issuecomment-1134999125 You are receiving this because you are subscribed to this th

Re: [apache/tvm-rfcs] [RFC] UMA Universal Modular Accelerator Interface (PR #60)

2022-05-23 Thread Mark Shields
``` backendA= MyUMABackendA() backendB= MyUMABackendB() backendA.register() backendB.register() mod = backendA.partition(mod) mod = backendB.partition(mod) ``` Ah, that's the example I was missing (sorry!). After registration I think calling backend.partition or letting CollagePartition 'do it fo

Re: [apache/tvm-rfcs] [RFC] DietCode: An Auto-Scheduler for Dynamic Tensor Programs (PR #72)

2022-05-23 Thread Andrew Reusch
@junrushao1994 yep, we will post meeting notes here per the TVM Community Meeting [guidelines](https://docs.google.com/document/d/179l1WzqDiLXjjLmLopvxBSr0ylFS2plyXbeTtB8CteM/edit#heading=h.yjcjro6m0xi3). -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm-rfcs/p

Re: [apache/tvm-rfcs] RFC-BYOC-DNNL-Integration (PR #73)

2022-05-23 Thread Andrew Reusch
Merged #73 into main. -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm-rfcs/pull/73#event-6662679591 You are receiving this because you are subscribed to this thread. Message ID:

Re: [apache/tvm-rfcs] [RFC] Relay Next Roadmap (PR #69)

2022-05-23 Thread Christopher Sidebottom
Merged #69 into main. -- Reply to this email directly or view it on GitHub: https://github.com/apache/tvm-rfcs/pull/69#event-6662507785 You are receiving this because you are subscribed to this thread. Message ID:

Re: [apache/tvm-rfcs] [RFC] Quarterly Releases (PR #67)

2022-05-23 Thread Ramana Radhakrishnan
> I think the main outstanding thing is the question of support here. I'd love > for a few more folks to weigh in, tagging a few folks who may have ideas: > @tqchen @jroesch @kparzysz-quic @u99127 @Mousius @leandron @comaniac @zhiics > @Hzfengsy @ZihengJiang @yzhliu @masahi @icemelon > > broadl

Re: [apache/tvm-rfcs] [RFC] UMA Universal Modular Accelerator Interface (PR #60)

2022-05-23 Thread PaulPalomeroBernardo
> ok for A1 i'm good with named phases and we can modify as necessary. i think > the A2.2 solution of directly registering target attrs makes sense to me. is > that the direction we're aligned on here? we can discuss this next week at the community meeting, or if we're in alignment on these two

Re: [apache/tvm-rfcs] [RFC] UMA Universal Modular Accelerator Interface (PR #60)

2022-05-23 Thread PaulPalomeroBernardo
Thanks @mbs-octoml for this detailed explanation. Being a Collage-supported backend is definitely something we want to achieve for UMA-integrated backends. > The registration of patterns will need to support the existing triple of > (pattern name, pattern, predicate) since the predicates are nec