We (Amazon SageMaker NEO) plans to host the first TVM Meetup in the Bay Area on Nov. 8th at East Palo Alto. A meetup group has been created here: https://www.meetup.com/Apache-TVM/events/265876615/
The agenda is as the following:
1:00 PM – Welcome & Logistics
1:05 PM – Graph Partitioning in Relay
1:45 PM – Quantization in TVM
2:30 PM – Dynamic Inputs
3:15 PM – Bring Your Own Topic
4:00 PM – Close
Please signup and join us if you are interested. Looking forward to seeing you.
Hi,
Are you going to publish some videos and/or summary of the discussions during the meetup?
Thanks!
Is there likely to be a remote participation method for this - it’s a bit late for me in the evening but it would be nice to be able to dial in / listen in. I’m interested in hearing more about the Graph Partitioning and the Quantization story in TVM.
regards
Ramana
The schedule is updated as the following:
1:00 PM -1:10 PM – Welcome & Logistics
1:10 PM - 1:40 PM – AWS: Graph Partitioning in Relay
1:40 PM - 2:10 PM – AWS: Quantization in TVM
2:10 PM - 2:40 PM – AWS: Dynamic inputs
2:40 PM - 2:50 PM – Breaks
2:50 PM - 3:10 PM TVM at OctoML, @jroesch
3:10 PM - 3:30 PM TVM at Facebook @ajtulloch
3:30 PM - 3:50 PM TVM at Xilinx Eliott Delaye
3:50 PM - 4:00 PM – Closing Remarks
We will host a happy hour after the meetup.
We’d like to also record the talks and share them so that they are still accessible to the people who cannot attend the meetup.
Thanks, that’s pretty neat.
The following is the summary for this meetup.
On Nov. 8th, the AWS SageMaker Neo team hosted the first TVM meetup in the Bay Area, which mainly focused on the presentation of some ongoing features the community is developing and also the successful deployment stories of TVM from several companies. The meetup was open to the public for registration and access. There were about 60 people from dozens of organizations and institutions in attendance. During the meetup, attendees freely exchanged their opinions on various topics that mentioned above in the thread.
According to the Apache spirit, we make the presentation slides available (with the permission from the presenters):
Bring You Own Codegen To TVM (AWS: @zhiics @comaniac) – slides
Compilation of Quantized Models in TVM (AWS: @janimesh) – slides
Dynamic Model in TVM (AWS: @haichen @kevinthesun) – slides
TVM at Facebook (Facebook: @ajtulloch, @llyfacebook) – slides
FPGA CNN Accelerator and TVM (Xilinx: Elliott Delaye) – slides
Many thanks to everyone who has helped us organize this meetup. Looking forward to the next one.
@jknight Sorry for the late response. We have some records, but they are just audios. We are thinking of uploading them some where, probably together with the slides in the meetup repo? Will update once that is done.