Using an external memory to store model and arena_size

[Huncho07]

Hi, I'm working with the tflite-micro in order to deploy a mobilenet+ssd model into a microcontroller. I don't have enough ram on the internal device, however my eval board does have an external ram of 8 megabytes, and it should be enough. When I call the Interpreter and then the interpreter.AllocateTensors(), I have this log:
Testing ObjectDetection
Schema verison is OK!Requested: 144, Available: 1048452
it stops and it's not going ahead on the IAR IDE i have this: HardFalut exception The processor has escalated a configurable priority exception to HardFault An instruction executed with an invalid EPSR.T or EPSR.IT field (CFSR.INVSTATE). Exception occurred at PC = 0x0, LR = 0x105cc095

If I use the internal memory I have something like this:
Testing ObjectDetection
Schema verison is OK!Requested: 144, Available: 643964
Requested: 144, Available: 619680
Requested: 144, Available: 619424
Requested: 144, Available: 619296
Requested: 144, Available: 618528
.
.
.

Requested: 144, Available: 419872
Requested: 144, Available: 418848
Requested: 144, Available: 418560
Requested: 144, Available: 418560
Requested: 144, Available: 418496
Requested: 144, Available: 418496
Requested: 0, Available: 418496
Requested: 3072000, Available: 418496
i cut some lines, but i hope you get what is happening.
My question is: Do I need to change something within the library in order to make it work? Maybe something about how the buffers are stored. I'll really appreciate your help ì, it's been few days that I'm working in it. Thanks.

[github-actions]

"This issue is being marked as stale due to inactivity. Remove label or comment to prevent closure in 5 days."