Create Nanonet model and implement Deepnano labeling

[adbailey4]

nanotensor1.config.json

• replaced blstm_layer_sizes with network
• instead of using a list of layer sizes, we use a list of dictionaries with size, type, name and/or bias. - added boolean option for binary_cost
• if binary clasification set true, if probability distribution set false

run_nanotensor.py

• parse network option in config file and create network, also created a better cost function options

utils.py

• added load and save json

create_training_data.py

• Refactored code to be a bit more clear. Moved some methods to utils folder.

data_preparation.py - Made sure that errors were handled correctly and made some minor changes

***Files Added create_training_data_test.py

[codecov[bot]]

Codecov Report

Merging #41 into master will increase coverage by 8.52%. The diff coverage is 49.34%.

@@            Coverage Diff             @@
##           master      #41      +/-   ##
==========================================
+ Coverage   25.93%   34.46%   +8.52%     
==========================================
  Files           9        9              
  Lines        1226     1300      +74     
==========================================
+ Hits          318      448     +130     
+ Misses        908      852      -56

Impacted Files	Coverage Δ
nanotensor/data.py	0% <0%> (ø)	:arrow_up:
nanotensor/run_nanotensor.py	0% <0%> (ø)	:arrow_up:
nanotensor/network.py	0% <0%> (ø)	:arrow_up:
nanotensor/data_preparation.py	84.69% <100%> (+2.93%)	:arrow_up:
nanotensor/utils.py	46.15% <56.09%> (+8.96%)	:arrow_up:
nanotensor/create_training_data.py	68.9% <73.56%> (+68.9%)	:arrow_up:

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update e391995...765513f. Read the comment docs.