v0.7.3-beta

What's Changed

New Feature Update:

• Added callbacks argument. The simulate_sorn and train_sorn class methods will return state_dict and the user requested network states and parameters as a dictionary. * @Saran-nns in #43

Thrashed

• Removed async ops due to the delay in fetching operations between processes and memory. Further, the implemented operations (at current state) are not CPU intensive. * @Saran-nns in #42

Full Changelog: v0.7.0...v0.7.3

v0.7.0-beta

Minor changes:

• Added Aync class for concurrent update of plasticity rules [STDP, IP, iSTDP, SP].
• Updated call for Plasticity methods through Async at Trainer and Simulator

Todo

• I/O bound slows down the simulation. Perform optimization with respect to the number of workers. Otherwise, create an association matrix and do broadcasting over heavyside step func here and here

Version	Iterations	Pool size(ne)	Time(s)
0.6.2	5000	200	166.213
0.7.0	5000	200	179.402

• Notebooks v0.6.2 and v0.7.0

sorn: A Python package for Self Organizing Recurrent Neural Network

v0.6.2-beta

v0.6.2 update

Stable pre-release of Self Organizing Recurrent Neural Network

v0.6.1-beta

v0.6.1 update

Stable pre-release

v0.5.8

Update README.md

Stable pre-release

Merge pull request #30 from Saran-nns/freezefix

Freeze fix

Sorn Stable beta release 0.4.6

v0.4.6

update test, colab badge

Sorn Stable beta release 0.3.20

v0.3.20

built binaries, update readme and set pypi

Sorn Stable beta release 0.3.19

v0.3.19

config break fix #13

Stable alpha release

Bool func of noise added. User now can override the network configurations by explicitly passing the args inside simulate_sorn and train_sorn class methods of Simulation and Trainer respectively.