Update app.py
Browse files
app.py
CHANGED
@@ -3,6 +3,7 @@ import subprocess,os
|
|
3 |
from datasets import load_dataset, Audio
|
4 |
import corpora
|
5 |
import ctcalign,graph
|
|
|
6 |
|
7 |
|
8 |
import matplotlib
|
@@ -53,9 +54,9 @@ def load_lang(langname):
|
|
53 |
|
54 |
def f1(langname,lang_aligner):
|
55 |
if langname=="Icelandic":
|
56 |
-
|
57 |
elif langname =="Faroese":
|
58 |
-
|
59 |
|
60 |
|
61 |
#fig = plt.figure(figsize=(10,4))
|
@@ -63,9 +64,12 @@ def f1(langname,lang_aligner):
|
|
63 |
#plt.xlabel("Vowel length (ms)")
|
64 |
#plt.ylabel("Consonant length (ms)")
|
65 |
|
66 |
-
|
|
|
|
|
|
|
67 |
#print([th for th in ds.sample()])
|
68 |
-
sound_path = ds['audio']['path']
|
69 |
transcript = ds['normalized_text']
|
70 |
|
71 |
return graph.align_and_graph(sound_path,transcript,lang_aligner)
|
|
|
3 |
from datasets import load_dataset, Audio
|
4 |
import corpora
|
5 |
import ctcalign,graph
|
6 |
+
from numpy import random
|
7 |
|
8 |
|
9 |
import matplotlib
|
|
|
54 |
|
55 |
def f1(langname,lang_aligner):
|
56 |
if langname=="Icelandic":
|
57 |
+
ds = corpora.ds_i
|
58 |
elif langname =="Faroese":
|
59 |
+
ds = corpora.ds_f
|
60 |
|
61 |
|
62 |
#fig = plt.figure(figsize=(10,4))
|
|
|
64 |
#plt.xlabel("Vowel length (ms)")
|
65 |
#plt.ylabel("Consonant length (ms)")
|
66 |
|
67 |
+
|
68 |
+
maxdat=len(df)
|
69 |
+
|
70 |
+
ds = ds.select([random.randint(maxdat-1)])
|
71 |
#print([th for th in ds.sample()])
|
72 |
+
sound_path = ds['audio'][0]['path'] # audio 0 array is the audio data itself
|
73 |
transcript = ds['normalized_text']
|
74 |
|
75 |
return graph.align_and_graph(sound_path,transcript,lang_aligner)
|