You are viewing a plain text version of this content. The canonical link for it is here.
Posted to discuss-archive@tvm.apache.org by LeoSpring via TVM Discuss <no...@discuss.tvm.ai> on 2020/05/27 02:48:28 UTC
[TVM Discuss] [Questions] Mxnet group conv not support very well by
tvm
hi all ，
I have a problem, I use tvm to speed up resnet101(mxnet) on 1080ti， I try to use group conv, but after I set group =2, the speed became slower too much. Then I try to speed up resnext101(mxnet). ResNext also became more slower . So , do tvm not support group conv (mxnet ) ?
To get more information , I try to set group = 32, after tvm speed up , it is more faster then group = 2.
if group = 1 , after tvm , process one image cost 6ms, if group = 2 , cost about 200ms, group = 3 cost 17ms. When group =2 , the speed is the slowest.
Can anyone solve the problem?
---
[Visit Topic](https://discuss.tvm.ai/t/mxnet-group-conv-not-support-very-well-by-tvm/6811/1) 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/42ce85ae03bf01868dd18df22e7492aa13d92ef9279ecd566ccc1af0a6f3da23).
[TVM Discuss] [Questions] Mxnet group conv not support very well by
tvm
Posted by LeoSpring via TVM Discuss <no...@discuss.tvm.ai>.
sorry, group=32, cost 17ms not group=3
---
[Visit Topic](https://discuss.tvm.ai/t/mxnet-group-conv-not-support-very-well-by-tvm/6811/2) 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/b7d5ad0d50079e8021ba856fddd15113f2f681410dcc8d0f33e166fecc5dc01d).
[TVM Discuss] [Questions] Mxnet group conv not support very well by
tvm
Posted by Wuwei Lin via TVM Discuss <no...@discuss.tvm.ai>.
Could you provide your workload (input/kernel shapes, stride, padding)? I’ll take a look
---
[Visit Topic](https://discuss.tvm.ai/t/mxnet-group-conv-not-support-very-well-by-tvm/6811/6) 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/b780902d842ee12f2687699ca63b1efa8efa043dd34b73d926efc5fa3b441be4).
[TVM Discuss] [Questions] Mxnet group conv not support very well by
tvm
Posted by LeoSpring via TVM Discuss <no...@discuss.tvm.ai>.
I have tuned , but it wasn't useful. I use nvidia 1080ti, cuda.
---
[Visit Topic](https://discuss.tvm.ai/t/mxnet-group-conv-not-support-very-well-by-tvm/6811/5) 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/9fe2eb035b1d7dd34df8d636452178337fae9544d32cf49f76025e763e86315d).
[TVM Discuss] [Questions] Mxnet group conv not support very well by
tvm
Posted by Wuwei Lin via TVM Discuss <no...@discuss.tvm.ai>.
have you tried using AutoTVM to tune these ops? Which platform are you using?
---
[Visit Topic](https://discuss.tvm.ai/t/mxnet-group-conv-not-support-very-well-by-tvm/6811/4) 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/bbd68ea86d0076d0069982f22361f5f0059765c5d638fff2226c036ab2f10314).
[TVM Discuss] [Questions] Mxnet group conv not support very well by
tvm
Posted by Xiaocenxiaocen via TVM Discuss <no...@discuss.tvm.ai>.
maybe group=1 fallback to a dense convolution. group=32 become a depthwise convolution. tvm has tuned optimal schedules for dense conv and depthwise conv
---
[Visit Topic](https://discuss.tvm.ai/t/mxnet-group-conv-not-support-very-well-by-tvm/6811/3) 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/7ca9590a584c2de72625739eff4bd45ecf92dbe61bae350f0419b8c483c59bf3).