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 jonso via TVM Discuss <no...@discuss.tvm.ai> on 2019/08/14 16:42:33 UTC

[TVM Discuss] [Questions] How to create a generic compute and override per target without autotvm?


I am trying to call cblas libraries when "cblas" is specified in the target libs and the target is x86. This needs to happen in two places: dense and batch_matmul.

Dense is straightforward because x86 already has an overridden compute and schedule. However, batch_matmul only has an overridden schedule. All computes call the same compute function. Further, batch_matmul doesn't support autotvm tuning. How can I make a generic compute call and override it per target?

Should I use the same methodology as the generic schedules in topi/python/topi/generic/nn.py? Or should I still use the `@autotvm.register_topi_compute` decorator?





---
[Visit Topic](https://discuss.tvm.ai/t/how-to-create-a-generic-compute-and-override-per-target-without-autotvm/3726/1) 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/4edbf6bc14fa2b3c78b418d8b57130a99a0a65930fcdd331d8436ade0955cc01).

Tianqi Chen, UW, Seattle, WA, 98105, United States
http://tracking.discuss.tvm.ai/tracking/unsubscribe?msgid=WUwOl6TTLX7ZY9rypY611A2