AbstractPhil commited on
Commit
92b5bd8
·
verified ·
1 Parent(s): 87d05d2

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -0
README.md CHANGED
@@ -20,6 +20,7 @@ There's more than enough weights, I just need to smash math today to see if I ca
20
 
21
  It's possible that the accuracy is much higher than expected and that I'm simply not asking them the right question. That would be quite the thing wouldn't it.
22
 
 
23
 
24
 
25
  # Breakdown and assessment
 
20
 
21
  It's possible that the accuracy is much higher than expected and that I'm simply not asking them the right question. That would be quite the thing wouldn't it.
22
 
23
+ Instead of 5-10 losses like I'm currently experimenting with, I would think 1-3 would be the correct amount. The current losses are often conflicting so they must be normalized, and a proper optimizer with ema must be constructed.
24
 
25
 
26
  # Breakdown and assessment