Where communities thrive

Join over 1.5M+ people
Join over 100K+ communities
Free without limits
Create your own community

Explore more communities

onnx/modelzoo

This room is deprecated - please use #onnx-modelzoo channel on https://slack.lfai.foundation/

People

Repo info

Activity

Yedong Liu

@lyd911

Hi, everyone, I just worked out the first proposal for the ONNX model zoo CI system and created a PR in the models repo: onnx/models#148, we can discuss more details here and look forward to hear from you:)

Prasanth Pulavarthi

@prasanthpul

cross posting from Lobby room: We're looking at enabling GIT LFS for the model zoo to store the ONNX models in GIT (instead of hosting in various places). Any feedback on this approach?

Yedong Liu

@lyd911

Some concerns of the Model Zoo CI project: onnx/models#168, any idea ?

Prasanth Pulavarthi

@prasanthpul

GIT LFS may help with some of the concerns you raised @lyd911

Yedong Liu

@lyd911

@prasanthpul When can this feature be online?

Prasanth Pulavarthi

@prasanthpul

soon i hope. working through some logistics

Prasanth Pulavarthi

@prasanthpul

It should be setup now!

Zhipeng Huang

@hannibalhuang

have we look at dvc ?

Prasanth Pulavarthi

@prasanthpul

@hannibalhuang looked at it briefly. downside seems to be that it needs separate storage provider (with git lfs, it is the storage) and overall seems more complicated (seems more appropriate for active model development, rather than a repository of curated final models)

Prasanth Pulavarthi

@prasanthpul

Steering Committee is happy to announce that @vinitra will help kick start the model zoo/tutorials SIG! LF AI will help setup the calendar and she will schedule mtgs. She has written a mission statement and proposed work at https://github.com/onnx/sigs/blob/master/models-tutorials/docs/MissionStatement.md

Vinitra Swamy

@vinitra

Thanks for the introduction, Prasanth! I'm excited to be leading this SIG moving forward. Keep an eye out for updates and an agenda for our first meeting in the next few days. Any feedback on the core priorities is also very welcome.

Vinitra Swamy

@vinitra

The first meeting of the Model Zoo + Tutorials SIG will be on Thursday, January 30th from 2 - 2:30 PM PST. We will discuss priorities for the SIG and proposals regarding upcoming work. Everyone is welcome.

Topic: Model Zoo + Tutorials SIG, Meeting #1
Time: Jan 30, 2020 02:00 PM Pacific Time (US and Canada)

Join Zoom Meeting
https://zoom.us/j/818755381

Meeting ID: 818 755 381

+16699006833, 818755381# US (San Jose)
+16465588656, 818755381# US (New York)

1 reply

@/all Additionally, we have now moved scheduling to the LFAI Foundation. Please join the group and subscribe to our SIG's calendar! https://lists.lfai.foundation/g/onnx-sig-modelstutorials

Svetlana Levitan

@sveta-levitan

Hi @vinitra, where can we find a summary of the call and tasks that we need to work on? Thank you!

Vinitra Swamy

@vinitra

You can find the meeting minutes + action items here, @sveta-levitan. For SIG members who missed the meeting, the zoom recording link is included in the minutes! Please sign up for tasks, or feel free to suggest new ones. https://github.com/onnx/sigs/blob/master/models-tutorials/meetings/001-20200130.md

Prasanth Pulavarthi

@prasanthpul

thanks @vinitra. collating upcoming talks workshops is a great idea. please add upcoming events to the onnx-announce calendar (https://wiki.lfai.foundation/pages/viewpage.action?pageId=18481196) instead of only in https://github.com/onnx/sigs/blob/master/models-tutorials/docs/CommunityEvents.md

Svetlana Levitan

@sveta-levitan

@prasanthpul Do you suggest we add events to both of these places or just one? I am not sure if we want to add a conference presentation into the common calendar?

Also, might be good to keep a page of past events, with materials (video, slides) from them?

Cam1256

@Cam1256

Hello. Does anybody have tutorials for making my first onnx model?

Svetlana Levitan

@sveta-levitan

You can either take one from the Model Zoo or export one using one of the converters. Start here: http://onnx.ai/get-started.html

Prasanth Pulavarthi

@prasanthpul

there are also some nice notebooks as part of the onnx-ecosystem docker image: https://github.com/onnx/onnx-docker/tree/master/onnx-ecosystem

Vinitra Swamy

@vinitra

The next SIG meeting will be on 3/6 from 2-2:30 PM PST. Hope you will join us! If you're looking for action items to accomplish before the next meeting, here's a good place to find some: https://github.com/onnx/sigs/blob/master/models-tutorials/meetings/001-20200130.md#action-items

Topic: Model Zoo + Tutorials SIG, Mtg #2
Time: Mar 6, 2020 02:00 PM Pacific Time (US and Canada)

Join Zoom Meeting
https://zoom.us/j/234284997

Meeting ID: 234 284 997

1 reply

Vinitra Swamy

@vinitra

Update -- the next meeting has been postponed to Monday 3/9 at 1-1:30 PM PST to accommodate more attendees' schedules.

Svetlana Levitan

@sveta-levitan

Great, thank you very much!

Vinitra Swamy

@vinitra

Minutes, Meeting Recording, and action items from Meeting #2: https://github.com/onnx/sigs/blob/master/models-tutorials/meetings/002-20200309.md

Prasanth Pulavarthi

@prasanthpul

anyone have thoughts on the value of the MD5 checksums for models? seems unnecessary once the files are in Git LFS

Vinitra Swamy

@vinitra

I agree, I think it can be deprecated. I don't think users are actually using them. Reminder -- if you have some free time, please port some models to Git LFS! onnx/models#271

Vinitra Swamy

@vinitra

Next SIG meeting slated for the afternoon (1 PM) of Monday, April 13th, after the ONNX workshop on April 9th. Let me know if the proposed time needs to be changed. Sign up for the online onnx workshop here: https://events.linuxfoundation.org/lf-ai-day-onnx-community-virtual-meetup/

Alex Merritt

@EmergentOrder

@vinitra I'm on the community meetup, it was mentioned earlier that the models repo has added GPT-2 & NCF (Neural Collaborative Filtering) since the last release. I found GPT-2, nice to have that. I don't see anything but the stub for NCF though, is there WIP there or was that a mistake?

Sree Vaddi

@sreev

@prasanthpul adding checksum , adding additional layer of security doesn't hurt. but is it valid , though ?

a better checksum can be, imho:
validity of the data used in making the model.
validity of model itself.

Sree Vaddi

@sreev

eg. download a picture of acquarium , use this against this model, it should identify 90% of the fish ?

or on that lines...

Vinitra Swamy

@vinitra

@EmergentOrder Hi Alex. I highlighted the Fast Neural Style Transfer and Super Resolution models in addition to GPT-2 during the ONNX workshop. I'm not aware of work on NCF currently, although we always welcome contributions! Which talk mentioned it?

Vinitra Swamy

@vinitra

@/all The next SIG meeting has been moved to this Monday, 4/13 at 2 - 2:30 PM PST. Hope to see you all there.

Topic: Model Zoo + Tutorials Meeting #3
Time: Apr 13, 2020 02:00 PM Pacific Time (US and Canada)

https://zoom.us/j/341887317
Meeting ID: 341 887 317

Alex Merritt

@EmergentOrder

@vinitra It was in the update at the beginning, I guess they must have confused NCF and Fast Neural Style Transfer. no worries, just wanted to clarify that

Vinitra Swamy

@vinitra

Reminder -- meeting today at 2! We're tracking several issues here, would love community input on these topics (opset versioning, model zoo CI, github analytics): https://github.com/onnx/sigs/issues

1 reply

Zhipeng Huang

@hannibalhuang

quick question, any specific reason the modelzoo is licensed under MIT instead of Apache or BSD ?

Vinitra Swamy

@vinitra

@hannibalhuang Some models are contributed by users / companies who prefer a specific license -- you will see a variety of licenses throughout the zoo. Do you have a preference? We can look into standardization.

Prasanth Pulavarthi

@prasanthpul

everything in ONNX is under MIT license for usage, with a CLA for contributions. with the move to LF AI, we will be moving to Apache 2.0 license with DCO for contributions (announced at recent workshop).

daquexian

@daquexian

Hi, I think the resnetv1 onnx model in model zoo https://github.com/onnx/models/tree/master/vision/classification/resnet is wrong. The spatial attr of batchnorm is incorrectly set to 0 and as a result an error is raised when running this model in onnxruntime

Vinitra Swamy

@vinitra

Hi @daquexian -- yes, that's a known issue! Thanks for reminding us. I fixed this on the MXNet exporter side: apache/incubator-mxnet#17711 a few weeks ago, but forgot to update the models. Will make a PR shortly.

Vinitra Swamy

@vinitra

onnx/models#294

Chat via Matrix