About rickforescue

rickforescue · 2022-03-17

Hello Community,I am working on a machine learning project and I am using an accelerometer sensor "STEVAL-MKI190v1". The board description and documentation can be found on this link The driver files for it is available on this github link. Now I hav...

rickforescue · 2022-03-16

I have a convolutional model with convolutional layers bach normalization layers and Dense layers at the end. The model is converted to a tflite model. The inferencing works perfect on computer using tflite but When I try to deploy it on the nucleo h...

rickforescue · 2022-03-15

/* USER CODE BEGIN PV */ static ai_buffer *ai_input; static ai_buffer *ai_output; /* USER CODE END PV */ int main(void){ ... // Chunk of memory used to hold the intermediate values for neural network AI_ALIGNED(32) ai_u32 activations[AI_NETWORK_D...

rickforescue · 2022-02-25

I tried to compress a keras model. It is a versy simple Neural network model with just fully connected (dense) layers. I converted it to tflite and compressed the weight to tf.float16 which was originally tf.float32. When I upload the model, it gave ...

rickforescue · 2022-03-21

Thanks. I see. Its interesting inisight.

rickforescue · 2022-03-18

Thanks a lot @fauvarque.daniel . The solution works. :) Well, I am curious. Can you give little bit more insight about it. What do you mean by folding the bachnorm ? ThanksRick

rickforescue · 2022-03-17

Thanks for reply @Peter BENSCH . I too cam across this github repository. I am still working on it. It seems to me at this point not very straight forward, but yeah no comments on the solution untill I exhaust my ideas. Lets see what this brings up....

rickforescue · 2022-03-16

Ok Thankyou @fauvarque.daniel

rickforescue · 2022-03-16

Here is the .h5 keras file in zipped format. The model is relatively big to fit in internal flash. I quantized it using tflite Let me know if you have some other questionsThanksRick

User statistics

User Activity

STEVAL-MKI190v1 integration with python directly without using a microcontroller

TOOL ERROR: operands could not be broadcast together with shapes (32,7200) (32,)

no source available for "ai_platform_inputs_get()" X-cube.AI version 7.1.0 I attach a basic code snippet which might be helpful in recreating the problem

INTERNAL ERROR: 'FLOAT16' --> Is it that STM doesnot support tf.float16 formats ?

Re: TOOL ERROR: operands could not be broadcast together with shapes (32,7200) (32,)

Re: TOOL ERROR: operands could not be broadcast together with shapes (32,7200) (32,)

Re: STEVAL-MKI190v1 integration with python directly without using a microcontroller

Re: TOOL ERROR: operands could not be broadcast together with shapes (32,7200) (32,)

Re: TOOL ERROR: operands could not be broadcast together with shapes (32,7200) (32,)