0.6.0 #5

gabrielmfern · 2022-09-02T03:23:27Z

fix the categorical cross entropy to compute loss and derivatives as it is normally defined to compute
fix the sum buffer operation that had a problem when setting the local buffer
add the Mean Bias loss function
add the Mean Absolute loss function
improve Intricate's internal code overall to run better and sometimes faster
add the Nesterov Accelerated Gradient optimizer
add the Momentum-based optimizer
add the Adagrad optimizer

… to make it possible to actually do the calculations properly

… lightweight

…mizer

…mizers

…s as well as inplace operations and add the adagrad optimizer

…ion to another

…dent and wrong predictions

gabrielmfern added 13 commits August 27, 2022 18:36

add the momentum based optimizer and improve the optimizer methods as…

08f6d2b

… to make it possible to actually do the calculations properly

create some inplace buffer operations to make calculations a bit more…

75cd77e

… lightweight

improve error handling and add the Nesterov Accelerated Gradient Opti…

4f9866c

…mizer

add unit tests to the optimizers and fix a problem on two of the opti…

c2dc461

…mizers

fix a small bug with the buffer operations and add some new operation…

4395feb

…s as well as inplace operations and add the adagrad optimizer

improve the way the sum buffer operation waits from one buffer reduct…

fde9955

…ion to another

make a small adjustment on the buffer like trait

a4da4f9

make the XoR test not verbose to run faster

7eb2a5d

add the mean absolute and fix a small problem in the mean squared kernel

b30eefd

add the mean bias loss function

7575b30

fix a problem with the local buffer size in the sum buffer operation

d18c655

remove some dbg!'s I added for testing

32882be

fix the categorical cross entropy function to actually penalize confi…

32f2127

…dent and wrong predictions

gabrielmfern merged commit c182528 into main Sep 2, 2022

Provide feedback