@Jiali this could be due to the fact that the ratio of valid schedules over an entire search space is relatively low, so population sampling cannot find any valid schedule. If the evolutionary search is based on a set of invalid schedules, it is highly possible that it will be trapped in invalid schedules and will need lots of time to find the first valid one.
I have a local branch that solves this problem by making sure evolutionary search starts with a set of valid schedules. However, this change will increase the sampling time from seconds to minutes, so I need to discuss with other folks before committing. --- [Visit Topic](https://discuss.tvm.apache.org/t/auto-scheduling-for-lstm-operator/8158/8) 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/3ca4535d469a64123ec8e00332d3530534b42f32a31348182f933ebed0d497b8).
