test_labels=test_labels,
wandb="wandb",
trial=str(trial),
)
elif args.method == "sub-enc-lstm":
print("Change method to sub-lstm")
else:
assert False, "method {} has no trainer".format(args.method)
(
results[trial][0],
results[trial][1],
results[trial][2],
results[trial][3],
) = trainer.pre_train(tr_eps, val_eps, test_eps)
np_results = results.numpy()
tresult_csv = os.path.join(args.path, "test_results" + sID + ".csv")
np.savetxt(tresult_csv, np_results, delimiter=",")
elapsed = time.time() - start_time
print("total time = ", elapsed)
if __name__ == "__main__":
parser = get_argparser()
args = parser.parse_args()
tags = ["pretraining-only"]
config = {}
config.update(vars(args))
train_encoder(args)
# -*- coding: utf-8 -*-
import os
import sklearn.metrics
from pyspark.ml.feature import VectorAssembler
from pyspark.ml.regression import LinearRegression
from pyspark.storagelevel import StorageLevel
## CUSTOM IMPORT
import conf
from src import american_community_survey as amc
from src import utils
from src import download_spark
## START
# Initiate the parser
args = utils.get_argparser().parse_args()
utils.printNowToFile("starting:")
utils.printNowToFile("downloading spark")
download_spark.download(os.getcwd())
###############################################################
if args.host and args.port:
spark = conf.load_conf(args.host, args.port)
else:
spark = conf.load_conf_default()
spark.sparkContext.addPyFile('ridge_regression.py')
import ridge_regression as rr