Update app.py

app.py

@@ -3,6 +3,7 @@ import subprocess,os
 from datasets import load_dataset, Audio
 import corpora
 import ctcalign,graph
+from numpy import random
 
 
 import matplotlib
@@ -53,9 +54,9 @@ def load_lang(langname):
 
 def f1(langname,lang_aligner):
     if langname=="Icelandic":
-        df = corpora.ds_i
+        ds = corpora.ds_i
     elif langname =="Faroese":
-        df = corpora.ds_f
+        ds = corpora.ds_f
 
     
     #fig = plt.figure(figsize=(10,4))
@@ -63,9 +64,12 @@ def f1(langname,lang_aligner):
     #plt.xlabel("Vowel length (ms)")
     #plt.ylabel("Consonant length (ms)")
 
-    ds = df.sample()
+
+    maxdat=len(df)
+
+    ds = ds.select([random.randint(maxdat-1)])
     #print([th for th in ds.sample()])
-    sound_path = ds['audio']['path']
+    sound_path = ds['audio'][0]['path'] # audio 0 array is the audio data itself
     transcript = ds['normalized_text']
     
     return graph.align_and_graph(sound_path,transcript,lang_aligner)