Transaction log file for in-memory database doesn't shrink

I am using SQL Server 2019 Enterprise Edition. We have an in-memory database used for session state data with one table stored in memory, schema only durability, simple recovery, differential backups run daily, full backup weekly. I noticed some errors in our logs recently. >Disallowing page allocations for database 'MyDatabase' due to insufficient memory in the resource pool 'default' > >XTP failed page allocation due to memory pressure: FAIL_PAGE_ALLOCATION 8 > >There is insufficient system memory in resource pool 'default' to run this query. In researching these, I also discovered that the log files for this database were several GB. It seemed odd to me that the log wasn't truncating. With experimentation, I found that if I run CHECKPOINT **twice**, I could then shrink the log file. It was my understanding that an automatic checkpoint should be happening when the transaction log file > 1.5 GB. My question is - is it normal to have transaction log files that are several GB for in-memory databases and keep growing?