Ex 01 Keras Simple

Running

python ex01_keras_train.py

Output

Using TensorFlow backend.

-- imports done, starting main --

-- read 2000 samples in 0.97sec. batch_size=12, 83 batches per epoch

epoch=0 batch=0 train_loss=0.861 train_step_time=1.54

 ...

 Discussion

Data

(mlearntut)psanagpu101: ~/github/davidslac/mlearntut $ h5ls -r /reg/d/ana01/temp/davidsch/ImgMLearnSmall/amo86815_mlearn-r069-c0000.h5

/                        Group

/acq.e1.ampl             Dataset {500}
/acq.e1.pos              Dataset {500}
/acq.enPeaksLabel        Dataset {500}
/acq.peaksLabel          Dataset {500}
...
/acq.waveforms           Dataset {500, 16, 250}
/bld.ebeam.ebeamL3Energy Dataset {500}
...
/evt.fiducials           Dataset {500}
/evt.nanoseconds         Dataset {500}
/evt.seconds             Dataset {500}
/lasing                  Dataset {500}
/run                     Dataset {500}
/run.index               Dataset {500}
/xtcavimg                Dataset {500, 363, 284}

 Code

mlearntut ex01_keras_train.py

Exercises

1. Swtich code to use tensorflow channel convention - hint, adjust all convent layers