Instruction-level Abstraction (ILA) of HLSCNN Accelerator

This is the ILA model of HLSCNN accelerator (all right reserved - Harvard University)

This master branch contains the ILA model for the original version of HLSCNN, specifically:

• The weight bitwidth is 8
• Both spad memory size are 0x20000
• Original output activation packing and indexing