[apache/tvm-rfcs] [RFC][Frontend] Add a PaddlePaddle Frontend (#19)
Add a PaddlePaddle frontend for TVM, also I have created [a RFC topic in forum](https://discuss.tvm.apache.org/t/rfc-add-a-paddlepaddle-frontend/10692) where we can discuss at same time. The related pull request is https://github.com/apache/tvm/pull/8645, still lack of UT, which will be finished in a week. @tqchen @FrozenGene @junrushao1994 @jcf94 You can view, comment on, or merge this pull request online at: https://github.com/apache/tvm-rfcs/pull/19 -- Commit Summary -- * add rfc for paddlepaddle frontend * add rfc for paddlepaddle frontend * Update add_paddlepaddle_frontend.md * Update add_paddlepaddle_frontend.md * Update add_paddlepaddle_frontend.md * Update add_paddlepaddle_frontend.md -- File Changes -- A rfcs/add_paddlepaddle_frontend.md (104) -- Patch Links -- https://github.com/apache/tvm-rfcs/pull/19.patch https://github.com/apache/tvm-rfcs/pull/19.diff -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/tvm-rfcs/pull/19
Re: [apache/tvm-rfcs] [RFC][Frontend] Add a PaddlePaddle Frontend (#19)
> @jiangjiajun Thanks for bringing this rfc, which will help TVM to serve more > conveniently for paddlepaddle users. This rfc is good generally speaking. > However, there are many spell grammar error so that I suggest you could do > one round of proof-reading Thanks for your comments, I'll review the RFC this week and update this pull request~ -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/tvm-rfcs/pull/19#issuecomment-895014609
Re: [apache/tvm-rfcs] [RFC][Frontend] Add a PaddlePaddle Frontend (#19)
Thanks @will-jl944 for revising and polishing this document, it really helps me a lot! @FrozenGene And an example code is provided in `Guide-level explanation`, could you review this RFC again, thanks~ -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/tvm-rfcs/pull/19#issuecomment-896798011
Re: [apache/tvm-rfcs] [RFC][Frontend] Add a PaddlePaddle Frontend (#19)
> Two followup actionable items per > https://github.com/apache/tvm-rfcs#rfc-workflow > > * Please rename the rfc to the original PR number (19) > * Please create a tracking issue per guideline Okay -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/tvm-rfcs/pull/19#issuecomment-898810651
[apache/tvm] [RFC][Tracking Issue] Add PaddlePaddle Frontend (#8751)
This issue will track the progress of PaddlePaddle frontend. Here is the RFC: [Add PaddlePaddle Frontend](https://github.com/apache/tvm-rfcs/blob/main/rfcs/add_paddlepaddle_frontend.md) There're some works will be done for this frontend, - [ ] Add a PaddlePaddle frontend supports most common operators - [ ] About 200 operators supporting, covering CV/NLP models - [ ] Control-Flow operators supporting - [ ] Quantized model supporting -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/tvm/issues/8751
[apache/tvm-rfcs] Rename RFC filename (#20)
This pull request rename rfc file `add_paddlepaddle_frontend.md` to `0019-add-paddlepaddle-frontend.md`, and add a tracking issue link. You can view, comment on, or merge this pull request online at: https://github.com/apache/tvm-rfcs/pull/20 -- Commit Summary -- * Merge pull request #3 from apache/main * Rename RFC file -- File Changes -- D rfcs/add_paddlepaddle_frontend.md (114) -- Patch Links -- https://github.com/apache/tvm-rfcs/pull/20.patch https://github.com/apache/tvm-rfcs/pull/20.diff -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/tvm-rfcs/pull/20
Re: [apache/tvm] [Release] v0.8 Release Planning (#8976)
Hi, @junrushao1994 I think we could add paddlepaddle frontend as a new feature in the release note of v0.8 -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/tvm/issues/8976#issuecomment-955121366
Re: [apache/tvm] Apache TVM v0.8 Release Note Candidate (Issue #9416)
Thanks @junrushao1994 ! There are 2 part I think we may need to fix For the **Accepted RFCs** part, > `[RFC-0019] Add paddlePaddle frontend` it should be `[RFC-0019] Add PaddlePaddle frontend`, just a case problem And for the **Frontends** part, > **PaddlePaddle initial support #8645** I have checked the v0.8 branch, there are 6 pull requests merged, we could list them all like others https://github.com/apache/tvm/pull/8645 https://github.com/apache/tvm/pull/9126 https://github.com/apache/tvm/pull/9236 https://github.com/apache/tvm/pull/9283 https://github.com/apache/tvm/pull/9295 https://github.com/apache/tvm/pull/9370 -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/tvm/issues/9416#issuecomment-957048426
Re: [apache/tvm] Apache TVM v0.8 Release Note Candidate (Issue #9416)
> > Should we wait for PyTorch TVM PR #8777? It should be merged soon. > > @masahi we can wait for it if this PR could get in this week Does this mean we will update v0.8 branch again this week, I merged a new pull request this week early https://github.com/apache/tvm/pull/9428 , will this be in v0.8? -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/tvm/issues/9416#issuecomment-960377251
Re: [apache/tvm] [VOTE] Release Apache TVM v0.8.0.rc0 (Issue #9504)
+1 -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/tvm/issues/9504#issuecomment-968217005
[TVM Discuss] [Development] Getting started with the VTA Chisel backend
Hi Luis, Both tests in [https://github.com/vegaluisjose/vta-tsim-tests](https://github.com/vegaluisjose/vta-tsim-tests) failed after updating TVM repo, please have a check. Thanks. --- [Visit Topic](https://discuss.tvm.ai/t/getting-started-with-the-vta-chisel-backend/2987/20) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/10403c5be1b5648d251ef005f9eccd469776fe35b402cf1aa7356cf2ed82af56). Tianqi Chen, UW, Seattle, WA, 98105, United States http://tracking.discuss.tvm.ai/tracking/unsubscribe?msgid=yz7I4ZKrDaCQK-d7w6pLKA2
[TVM Discuss] [Development] Getting started with the VTA Chisel backend
Luis, thanks for pointing it out. --- [Visit Topic](https://discuss.tvm.ai/t/getting-started-with-the-vta-chisel-backend/2987/22) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/b46615cf19d3a544e764cc161d7775551dd5037cee5222fef9f82eb18b61383e). Tianqi Chen, UW, Seattle, WA, 98105, United States http://tracking.discuss.tvm.ai/tracking/unsubscribe?msgid=nRXHydyvQLdyK1Mcc6CHIw2
[TVM Discuss] [Meetup] TVM Online Meetups
We here at [OctoML](https://octoml.ai) are thinking of hosting an online TVM meetup to give people a way to connect while working from home and interact in a higher bandwidth medium than here or Github issues. Following the [TVM meeting guidelines](https://discuss.tvm.ai/t/tvm-meetup-guideline/4429) and adapting them to an online setting to the extent possible, video would be recorded and put online immediately afterwards. Content would be rather informal with one or two slots for volunteers to walk through RFCs or other topics of interest and would be open for anyone to present from the community by volunteers. The main outstanding question before doing this is: * Is there enough interest from the community to join an online/video TVM meetup? Please heart this message or reply if so and I'll take that as a gauge of interest. If there was enough interest, we could also make this recurring (monhtly with rotating hosts?). If we made it recurring, then we could also alternate times to be European and Asian timezone friendly. Also, are there people who would be willing to volunteer to speak for a topic? Something like 10-20 minutes. Please let me know either here or DM if so. Thanks! --- [Visit Topic](https://discuss.tvm.ai/t/tvm-online-meetups/6382/1) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/920b8cf4ee9d8958ec46abab74a071fd8d4accaca38df1f61bbba9f18aed00cf).
[TVM Discuss] [Meetup] TVM Online Meetups
> However it is quite expensive in terms of logistics and organization @ramana-arm, to keep this light weight, my thought was initially to: 1. Put up an agenda Google doc a few weeks before the scheduled day 2. Invite people to add topics to discuss (and/or volunteer to talk about topics from anywhere from 1-15 minutes). 3. The day of, we all meet in a Zoom call (simultaneously live streamed to Youtube). Then the video recording is already live on Youtube at the end of the meeting (and we can post another copy for anyone without access to Youtube). That way there is very little overhead, and if there's not much to discuss, we end early. And given the low effort, we can just try it out as an experiment to see how it goes for a first time. Or did you have another format in mind that made you mention the expensive logistics and organization? --- Hi @bgchun, thanks for your input! Your perspective as a first time poster here is definitely appreciated. --- [Visit Topic](https://discuss.tvm.ai/t/tvm-online-meetups/6382/6) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/e897919874dd5ba287da4c9a014ae26da1b84c05d9ef7bbf8c1c32566b8e37d7).
[TVM Discuss] [Meetup] TVM Online Meetups
@ramana-arm Great points and we will make sure that any discussion in an online meetup is a complement, rather than a substitute for RFCs, Github issues, and discuss threads. ## First OctoML Apache TVM Online Meetup - May 21 Okay, I'd like to go ahead and give this a try. Thursday, May 21 at 9am PST (see [timezone conversions here](https://everytimezone.com/?t=5ec5c480,3c0)) we will host a zoom meeting which will be recorded and posted onto Youtube afterwards. You'll need to complete a simple free [registration for the zoom meeting here](https://us02web.zoom.us/meeting/register/tZcudO6uqj0iHtcdcwacjFT6YPZrQPJ5t7_-) if you want to participate live (this requirement is an attempt to curb zoom "crashers"). Let me know if anyone has any issues with this. And the agenda is [live here on Google docs](https://docs.google.com/document/d/1bfE7Ydk43aMM3aiLf8GM7xEatPJwhj839ANfBheUJcM/edit?usp=sharing) feel free to make comments and suggestions on the format and content or to volunteer to present something. You can also comment here in this discuss thread. Currently we have: ### Tentative (May 5th) agenda: * 2 min - Introduction - Jason Knight * 10 min - Update on Unified IR work - Tianqi Chen * 10 min - Graph pattern matching and rewriting - Matthew Brookhart * < volunteers requested> * (Time available) - Open mic * Suggestions for future meeting topics / cadence We'll send some announcements over Twitter, but feel free to amplify in any media you have access to. --- [Visit Topic](https://discuss.tvm.ai/t/tvm-online-meetups/6382/8) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/606891d2b1c5ad178442b388041fddf2b424fee80adcd4080caaf58ab5ce2c85).
[TVM Discuss] [Meetup] TVM Online Meetups
Edit: posting this reply separately for link snapshotting purposes: @ramana-arm Great points and we will make sure that any discussion in an online meetup is a complement, rather than a substitute for RFCs, Github issues, and discuss threads. --- [Visit Topic](https://discuss.tvm.ai/t/tvm-online-meetups/6382/9) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/037e99da2b5924f809a4610b647be18bcda25fdf92bbd38892c4b47392390330).
[TVM Discuss] [Meetup] [Reminder] Online Meetup this Thursday
## [Reminder] First Online Meetup: this Thursday * Host: OctoML * Format: Zoom (then recording on Youtube) * Timezones: US and European friendly (planning to alternate to Asia friendly for the next one) * Current registered attendees: 29 (as of May 19) Just as a reminder, we're going to run our first experimental online meetup this Thursday, May 21 at 9am PST (see [timezone conversions here ](https://everytimezone.com/?t=5ec5c480,3c0)). OctoML will host a zoom meeting which will be recorded and posted onto Youtube afterwards. **You’ll need to complete a simple free** [registration for the zoom meeting here](https://us02web.zoom.us/meeting/register/tZcudO6uqj0iHtcdcwacjFT6YPZrQPJ5t7_-) **if you want to participate live** (this requirement is an attempt to curb zoom “crashers”). Let me know if anyone has any issues with this. And the agenda is [live here on Google docs ](https://docs.google.com/document/d/1bfE7Ydk43aMM3aiLf8GM7xEatPJwhj839ANfBheUJcM/edit?usp=sharing) feel free to make comments and suggestions on the format and content or to volunteer to present something. You can also comment here in this discuss thread. Currently we have: ### As of May 19th the agenda is: * 2 min - Introduction - Jason Knight * 10 min - Update on Unified IR work - Tianqi Chen * 10 min - Graph pattern matching and rewriting - Matthew Brookhart * 15 min - Ethos-N Integration using BYOC (Bring your own codegen) - Matthew Barrett * (Time available) - * Open mic * Suggestions for future meeting topics / cadence * GPU “birds of a feather” special topic session? * Office hours - discuss questions posted here in advance so they can be triaged to future meeting topics, further offline discussion, etc. Vote here by adding a comment onto the line you’d like to see discussed: * [Binding two IterVars in one loop nest to the same thread](https://discuss.tvm.ai/t/binding-two-itervars-in-one-loop-nest-to-the-same-thread/6693) Look forward to seeing you there! --- [Visit Topic](https://discuss.tvm.ai/t/reminder-online-meetup-this-thursday/6730/1) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/726957c903f0dc7a6734eac42f01150b394ef3afcb1bcb174f7480ef890bd1ef).
[TVM Discuss] [Meetup] [Reminder] Online Meetup this Thursday
Thanks to @tqchen, @mbrookhart, and @matt-arm for the great talks and ~40 for attending for a great first Online TVM Meetup! Video is now up on Youtube [here](https://youtu.be/U_j1Yc-6sxM). And the agenda is [here](https://docs.google.com/document/d/1bfE7Ydk43aMM3aiLf8GM7xEatPJwhj839ANfBheUJcM/edit#). It sounds like there is interest from the community on two follow-up "directed" online sessions similar to this one but focused on a particular sub area of TVM, specifically for: * GPU efforts * Bare metal/uTVM @YuanLin has volunteered to kick off a thread to crowdsource topics to discuss for the GPU session. And once @areusch gets his latest uTVM blog post up (then we'll do something similar for the bare metal efforts). So be on the lookout for those! --- [Visit Topic](https://discuss.tvm.ai/t/reminder-online-meetup-this-thursday/6730/2) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/a8437dabfeda09d788bc7413e1326fcc175342d1e84dcf76ce3d31415155433f).
[TVM Discuss] [Meetup] uTVM (Embedded Focus) Online Meetup
There was interest at the first TVM online meetup in scheduling an embedded/bare metal focused session. Ideas of topics of discussion: * Next planned steps following the [uTVM blog post](https://tvm.apache.org/2020/06/04/tinyml-how-tvm-is-taming-tiny) work from @areusch and @weberlo * Use cases people have or would like to see examples/tutorials * Open mic, Q&A * ? Please comment here for more ideas or volunteers for discussion topics. To gauge the level of interest and for scheduling, if you'd like to attend, please take the following poll for when you could virtually attend: [poll type=multiple results=always min=1 max=2 public=true chartType=bar groups=trust_level_0] * Asia friendly - 7pm PT - Wed 6/17 * European friendly - 9am PT - Thurs 6/18 [/poll] --- [Visit Topic](https://discuss.tvm.ai/t/utvm-embedded-focus-online-meetup/6908/1) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/609300c7f1720dc6cdd9d2aa8838a5fe2b7f3e8d3160014088a3a5d9f6d8c1cf).
[TVM Discuss] [Meetup] uTVM (Embedded Focus) Online Meetup
> I wonder if via the fora here we should assemble a sort of mini list of > protoRFCs in advance of the discussion. Absolutely! Feel free to put them here or as suggested edits in [the agenda doc](https://docs.google.com/document/d/1bfE7Ydk43aMM3aiLf8GM7xEatPJwhj839ANfBheUJcM/edit#). --- [Visit Topic](https://discuss.tvm.ai/t/utvm-embedded-focus-online-meetup/6908/3) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/44c3e2e426a746410866cf1e7c2bfeb70b232c40fad166c928908b580f1d12b8).
[TVM Discuss] [Meetup] uTVM (Embedded Focus) Online Meetup
Great suggestions, @liangfu would you be able to talk next week at one of the slots above regarding your MISRA-C runtime work and plans? --- [Visit Topic](https://discuss.tvm.ai/t/utvm-embedded-focus-online-meetup/6908/6) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/75938b70e21e2aa5ad2f9e0c7010875cd9dd6e66c9da9587739854dc7935e3fb).
[TVM Discuss] [Meetup] uTVM (Embedded Focus) Online Meetup
Thanks everyone for the great feedback on here and the agenda. Please [register for the event here ahead of time](https://us02web.zoom.us/meeting/register/tZ0pduyvrT4sGNFSF1LTMn9kW8S953ufskKR). A recording will be uploaded to the OctoML Youtube channel immediately afterwards as we did last time. Looking forward to seeing you there! --- [Visit Topic](https://discuss.tvm.ai/t/utvm-embedded-focus-online-meetup/6908/9) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/1bc065292a75595450e903d276569095059508636211881a5bbfd65dc5495a4a).
[TVM Discuss] [Meetup] uTVM (Embedded Focus) Online Meetup
Also @liangfu Thanks for agreeing, I gave you a 15 minute slot on the agenda (but we should be flexible for shorter or longer). Here's the timezone conversions for next week if that helps people as well:  https://everytimezone.com/s/23480098 --- [Visit Topic](https://discuss.tvm.ai/t/utvm-embedded-focus-online-meetup/6908/10) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/40e14b61de7de608418d8ad460f05807953bf563da3fc735febd0bf62db3537d).
[TVM Discuss] [Meetup] uTVM (Embedded Focus) Online Meetup
Thanks for the great meetup earlier today everyone. Video is up here: https://youtu.be/mW7dk-rXuy8 --- [Visit Topic](https://discuss.tvm.ai/t/utvm-embedded-focus-online-meetup/6908/12) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.ai/email/unsubscribe/646955932756a3f9901250be99a1cecd04bbdc1f25828b6e4f37d638c6813671).
[Apache TVM Discuss] [Development/RFC] [RFC] Rename Hybrid Script
:heart_eyes: :star_struck: Thank you for proposing this! Completely agreed that we should disambiguate. Having `tvm.script` and `te.hybrid.script` is an improvement, but I still think there is far more overlap than ideal between the two. What about `pytir`? Then it opens the door to other embedded TIR DSLs in other languages being succeessfully disambiguated as well. --- [Visit Topic](https://discuss.tvm.apache.org/t/rfc-rename-hybrid-script/7915/3) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.apache.org/email/unsubscribe/5e16cc1d5e08b1db8ea9d12b21bf7a2daf9a22c658ecc1d3fbaadce2317b800b).
[Apache TVM Discuss] [Development/RFC] [RFC] TensorIR: A schedulable IR for TVM
When I read this RFC, I was confused because I read that the compilation flow **currently** goes from > TF/PyTorch/ONNX -> Relay -> TE (based on TE schedule) -> TIR -> C++/CUDA And then I read: > TensorIR is a brand new low-level IR with full scheduling support. Here are > some [...] And then see references to `TIR` throughout the rest of the RFC, which seem unclear to me if these references are to the **new** `TensorIR`, or the **old** `TIR`. Can we clarify both here and in the code base moving forward whether we are referring to `TensorIR` or the old `TIR`? I think there are two ways of doing this going forward and we should explicitly pick one: 1. Have `TIR` refer to the old version of TIR, and always use `TensorIR` when talking about the new IR. 2. Be clear that TensorIR moving forward will be replacing the old `TIR`, so that any reference to `TIR` could be referring to the new or old version (and should be clarified if not obvious from context). Someone more knowledgeable than me should pick one of those (or correct me and/or point out other options if I've gotten anything wrong here). --- [Visit Topic](https://discuss.tvm.apache.org/t/rfc-tensorir-a-schedulable-ir-for-tvm/7872/39) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.apache.org/email/unsubscribe/793aaee1938382b9b1215945f8f86517563aa3bbf8683fc0f33e1b09fb523744).
[Apache TVM Discuss] [Meetup] Proposal : Regular µTVM Meet up
Fantastic! Thanks so much for stepping forward to lead something like this @tgall_foo. I believe @hogepodge can give you access to add the uTVM meetup calendar hit to the public TVM calendar. --- [Visit Topic](https://discuss.tvm.apache.org/t/proposal-regular-tvm-meet-up/9863/4) to respond. You are receiving this because you enabled mailing list mode. To unsubscribe from these emails, [click here](https://discuss.tvm.apache.org/email/unsubscribe/be4aeff2797400412d03f359693e674e83980a7fc6aa9eeff865ce79068c6e05).
