qkeras

qkeras copied to clipboard

Error during Quantizing MobileNetV2

1

I am trying to quantize MobilnetV2 using 4-bit width, but when I run print_qstats(model), I am getting an error "A merge layer should be called on a list of inputs"...

preetam19

kshithijiyer

kshithijiyer

(Re)quantize existing QKeras model with model_quantize

3

Add to `model_quantize` the capability to accept an existing `qkeras` model and return a model with a different `quantizer_config`. This could be used, for example, to train at higher precision...

thesps

Inference of QKeras models

1

Greetings, How to use the QKeras model for prediction? Can I use method .predict for inference on test data or does QKeras defines a special function for inference of quantized...

RatkoFri

removed the integer and bits scaling in quantized bits for alpha=auto and alpha=auto_po2

1

In the quantized_bits class, when alpha=auto or alpha=auto_po2, the unquantized value "x" is manipulated in a strange way. Even if the number of integer bits is not considered when scaling...

fabrizioaymone

TypeError: Could not locate class 'QConv2D'.

5

Hi, When I run the QKerasTutorial: [https://github.com/google/qkeras/blob/master/notebook/QKerasTutorial.ipynb](url) I get the following error: > TypeError: Could not locate class 'QConv2D'. Make sure custom classes are decorated with `@keras.saving.register_keras_serializable()`. ... Below is...

OliverFaust

A function I am using calls `layer.activation.__name__`. The activation functions have been separated and called via the `QActivation` layer as shown on the figure. The error is also shown in...

Brainz22

Use `delta()` instead of `max() * 2^-bits` when `delta()` is defined

1

I'm working on floating point quantizer and I think this line doesn't work in my case: `delta = self.quantizer.max() * 2**-self.quantizer.bits` With the modification proposed one can define `delta()` in...

nghielme

mschoenb97