In that ali link look at the pictures to the left. On the battery side of the driver you will see 8 tinned pads in groups of 2. Soldering 2 pads together is bridging the connection to change modes in the mcu. The instructions for mode groups and which bridges does what are near the bottom of that linked page.
Anything with a µC would certainly have multimodes, unless reflashed.
The 7135, though, needs a certain minimum voltage to keep working, about 2.7V, so it’s not like DDing an LED and leaving it on in a closet or anything. It will cut out.
Not precise, as in sensing the input voltage and cutting out when it drops to X, but yeah, 7135s do cut out below a certain voltage.
If anyone’s interested in trying a batch, have at it. With, say, 3 chips, one might cut out at 2.73V, the next at 2.69V, the last at 2.65V. So it’s not one Master Circuit doing it, but it will crap out before killing the cell.
And when you see it getting dim, that’s Clue #1 to stop and recharge.
No need to flash the driver I linked from Ali, the driver has a settable modes with the bridges. All four bridged one mode (high).
.
The amc7135 in the datasheet list the working voltage as minium 2.7v, that doesn’t mean it wont work lower. It just means ADDtek specs all amc7135’s to work correctly at 2.7v. I have the exact driver you linked. I tested it with a DPS5015 and a XML, the light became dim at below 2.5v and kept getting dimer until 2.25v at that point the led was just barely glowing. At 2.5v just guessing, looked like maybe 5 to 10 lumens.
I don’t know where the amc7135 quits working but its definitely lower than 2.25v.
.
Removing the diode and using a jumper supposed to gain a few millivolts back also, I tested mine with the diode on.
Either way, get the 4×7135 version and set it to single mode (mode group 12) for your family to use; you can always choose a different mode group in future if you want to.