I want to compare the recognition rate before and after quantization.
First, i want to save the weights of the trained model as a csv file,
quantize them through an external program, and then input the csv file
as the weights of the model again
but, i don’t know how…
thx for reading my Question.
1 Like
tf.keras.Model
has explicit methods get_weights
and load_weights
with which you can store weights in any type of file and load them from any type of file.
Please take a look at this stackoverflow answer
3 Likes
I appreciate your reply.
I was able to solve this problem.
Thank you very much. ashutosh.
2 Likes