QuantizedArrary Error

Zama Concrete ML
1

I am using the below command to create a quantized array for input data. The shape of my input data is - torch.Size([1, 2048]), and input is floating point numbers.

n_bits = 6, is_signed = True
q_test_data = QuantizedArray(n_bits = n_bits, values=t_ip, is_signed=is_signed)

The error I am getting is

TypeError Traceback (most recent call last)
in <cell line: 4>()
2 pt_quant = PostTrainingAffineQuantization(n_bits = n_bits, numpy_model = numpy_fc_model, is_signed = is_signed)
3 # Quantize input
----> 4 q_mnist_test_data = QuantizedArray(n_bits = n_bits, values=t_ip, is_signed=is_signed)
5 # Calibrate layers and activations
6 quant_module = pt_quant.quantize_module(t_ip)

5 frames
/usr/local/lib/python3.9/dist-packages/numpy/core/fromnumeric.py in _wrapreduction(obj, ufunc, method, axis, dtype, out, **kwargs)
82 return reduction(axis=axis, dtype=dtype, out=out, **passkwargs)
83 else:
—> 84 return reduction(axis=axis, out=out, **passkwargs)
85
86 return ufunc.reduce(obj, axis, dtype, out, **passkwargs)

TypeError: min() received an invalid combination of arguments - got (out=NoneType, axis=NoneType, ), but expected one of:

  • ()
  • (Tensor other)
  • (int dim, bool keepdim)
    didn’t match because some of the keywords were incorrect: out, axis
  • (name dim, bool keepdim)
    didn’t match because some of the keywords were incorrect: out, axis

Can I know the issue?

Thanks in advance

2

QuantizedArray does not support torch input yet so that might be the reason of you error. You will have to convert the input data to numpy.

Note that quant_module = pt_quant.quantize_module(t_ip) will do the input quantization for you. Then you can just call quant_modulde.quantize_input(t_ip) to get the quantized input.

1 Like