FIFO optimization for qkeras models #773
sandeep1404
started this conversation in
General
Replies: 1 comment
-
Hi, can someone look into this issue , when I run the config file, I am getting Filenotfound error, I have no idea where to get that file or something is missing in my code, below are the log file :
|
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hello all,
My Bram's and LUTs are getting overutilized so I was trying to use the concept of FIFO depth optimization, In that process when I use FIFO depth optimization, I can able to successfully synthesize my 32bit keras model without being quantized, on zcu104 and resolved the issue of overutilization, but when I try to simulate the qkeras 8bit precision model on zcu 104 still the resources are getting overutilized and when I try to use the FIFO depth optimization and then synthesize for the qkeras 8bit fp quantized model, I am facing the following issues:
Here is my config file:
please help me in resolving the issue. Thank you in advance.
Beta Was this translation helpful? Give feedback.
All reactions