/VLMMLR_packer_contest

Primary LanguagePython

VLMMLR Packer Contest

Click here to see my slides

Run

First run

cd VLMMLR_packer_contest
python3 run.py
python3 analyze.py

Named run

cd VLMMLR_packer_contest
python3 run.py --model my_model.h5
python3 analyze.py --model my_model.h5

Using other dataset

cd VLMMLR_packer_contest
python3 run.py --training_data data/training_data.first_block.500000.csv
python3 analyze.py

Getting more training data

I have a larger training data set here

Submitting Results

When you are done training a model, run score.py --model [model name]. Don’t do this too often. Strictly speaking you shouldn’t even have access to this ability.

Add your .h5 model file and your run.py file to results/ and report the output of score.py in results.md. Push to master when you are done (may need to pull first).

Useful links

https://keras.io