(1) I have uploaded the dataset pima-indians-diabetes.csv to the FloydHub project folder jsmith/datasets/pima-indians-diabetes .
(2) I have previously created the FloydHub project folder jsmith/projects/keras-beta . This folder is empty.
(3) I then ran the following commands:
floyd init keras-beta
floyd run --gpu --env keras --data jsmith/datasets/pima-indians-diabetes/1:pima 'python pima_test.py'
where pima_test.py contains the line
dataset = loadtxt('floyd/input/pima', delimiter=',')
but I got the error message
"Code size too large to sync, please keep it under 100.0MiB."
This does not make sense as the dataset is not in the FloydHub project folder. What am I doing wrong?