[onert] Quantization type kernel for transformer

[hseok-oh]

Below is required I/O quantization type (uint8/uint16) kernel for quantized transformer model

• [ ] MUL
  • [x] UINT8
  • [ ] INT16
• [ ] ADD
  • [x] UINT8
  • [ ] INT16
• [ ] RSQRT
  • [ ] UINT8
  • [ ] INT16
• [ ] DIV
  • [ ] UINT8
  • [ ] INT16
• [x] RESHAPE (same I/O quant param)
• [ ] TRANSPOSE (same I/O quant param)
  • [x] UINT8
  • [ ] INT16
• [ ] STRIDED_SLICE (same I/O quant param)
  • [ ] UINT8
  • [ ] INT16
• [ ] NEG
  • [ ] UINT8
  • [ ] INT16
• [ ] CONCATENATION
  • [x] UINT8
  • [ ] INT16
• [ ] BATCH_MATMUL
  • [ ] UINT8
  • [ ] INT16
• [ ] SOFTMAX
  • [x] UINT8
  • [ ] INT16
• [ ] LOGISTIC
  • [x] UINT8
  • [ ] INT16
• [ ] GATHER (indices: int32/int64)
  • [x] UINT8
  • [ ] INT16
• [ ] MEAN
  • [x] UINT8
  • [ ] INT16
• [ ] SQRT
  • [ ] UINT8
  • [ ] INT16

Quantization type change

• [ ] QUANTIZE
  • [ ] UINT8 -> INT16
  • [ ] INT16 -> UINT8

I/O and weight quantization type for transformer model

• [ ] FULLY_CONNECTED (channelwise quantization)
  • [ ] UINT4 weight, UINT8 I/O (#12741)
  • [ ] UINT8 I/O and weight
  • [ ] INT16 I/O and weight