Script updates

scripts/gen-training.bash

@@ -2,4 +2,4 @@
 
 SCRIPT_DIR=$( cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )
 
-find . -name '*.ground' | nice parallel -q --eta -I foo bash -c 'F=$(dirname foo)/$(basename foo .ground); echo $F; python3 '"$SCRIPT_DIR"'/gen-training.py $F{.ground,,.csv}'
+find . -name '*.ground' | shuf | nice parallel -q --eta -I foo bash -c 'F=$(dirname foo)/$(basename foo .ground); echo $F; python3 '"$SCRIPT_DIR"'/gen-training.py $F{.ground,,.csv}'

scripts/gen-training.py

@@ -22,7 +22,7 @@ subprocess.check_output("/data/research/rose/install-latest/bin/bat-ana --no-pos
 
 def get_all_dis():
 
-    output = subprocess.check_output("/data/research/rose/install-latest/bin/bat-dis --no-bb-cfg-arrows --color=off %s 2>/dev/null" % (ananame), shell=True)
+    output = subprocess.check_output("/data/research/rose/install-latest/bin/bat-dis --no-insn-address --no-bb-cfg-arrows --color=off %s 2>/dev/null" % (ananame), shell=True)
     output = re.sub(b' +', b' ', output)
 
     func_dis = {}

scripts/import datasets.py

@@ -0,0 +1,4 @@
+import datasets
+# find /wherever -name '*.csv' | tar -T - -c | tar -C /somewhere -x
+ds = datasets.load_dataset("csv", "/somewhere")
+ds = ds.class_encode_column('Type')