Merge branch 'develop' of github.com:ECP-CANDLE/Benchmarks into develop

Author: j-woz

Pilot1/Combo/combo_default_model.txt

@@ -1,4 +1,5 @@
 [Global_Params]
+model_name='combo'
 cell_features=['expression']
 drug_features=['descriptors']
 dense=[1000, 1000, 1000]

Pilot1/NT3/nt3_baseline_keras2.py

@@ -85,7 +85,6 @@ def run(gParameters):
     model = Sequential()
 
     initial_epoch = 0
-    best_metric_last = None
 
     X_train, Y_train, X_test, Y_test = load_data(train_file, test_file, gParameters)
 
@@ -167,14 +166,12 @@ def run(gParameters):
     model.add(Dense(gParameters["classes"]))
     model.add(Activation(gParameters["out_activation"]))
 
-    J = candle.restart(gParameters, model)
+    ckpt = candle.CandleCkptKeras(gParameters, verbose=False)
+    ckpt.set_model(model)
+    J = ckpt.restart(model)
     if J is not None:
         initial_epoch = J["epoch"]
-        best_metric_last = J["best_metric_last"]
-        gParameters["ckpt_best_metric_last"] = best_metric_last
-        print("initial_epoch: %i" % initial_epoch)
-
-    ckpt = candle.CandleCheckpointCallback(gParameters, verbose=False)
+        print("restarting from ckpt: initial_epoch: %i" % initial_epoch)
 
     # Reference case
     # model.add(Conv1D(filters=128, kernel_size=20, strides=1, padding='valid', input_shape=(P, 1)))

Pilot1/NT3/train.sh

@@ -0,0 +1,4 @@
+#!/bin/bash
+
+# TRAIN SH
+# IMPROVE interface to NT3 model