j-krzywdziak
commited on
Commit
•
edd945d
1
Parent(s):
33b6fcd
new test
Browse files
test2.py
CHANGED
@@ -185,15 +185,12 @@ class TestASR(datasets.GeneratorBasedBuilder):
|
|
185 |
audio_data = {}
|
186 |
transcripts = []
|
187 |
key = 0
|
188 |
-
print(examples, keywords)
|
189 |
-
print(local_extracted_archive)
|
190 |
for path, f in files:
|
191 |
audio_data[path] = f.read()
|
192 |
with open(keywords, encoding="utf-8") as f:
|
193 |
next(f)
|
194 |
for row in f:
|
195 |
r = row.split("\t")
|
196 |
-
print(r)
|
197 |
path = 'examples/'+r[0]
|
198 |
ngram = r[1]
|
199 |
transcripts.append({
|
@@ -202,10 +199,8 @@ class TestASR(datasets.GeneratorBasedBuilder):
|
|
202 |
"type": "keyword"
|
203 |
})
|
204 |
with open(examples, encoding="utf-8") as f2:
|
205 |
-
next(f2)
|
206 |
for row in f2:
|
207 |
r = row.split("\t")
|
208 |
-
print(r)
|
209 |
path = 'examples/'+r[0]
|
210 |
ngram = r[1]
|
211 |
transcripts.append({
|
@@ -213,8 +208,6 @@ class TestASR(datasets.GeneratorBasedBuilder):
|
|
213 |
"ngram": ngram,
|
214 |
"type": "example"
|
215 |
})
|
216 |
-
print("AUDIO DATA: ", audio_data)
|
217 |
-
print("TRANSCRIPT: ", transcripts)
|
218 |
if audio_data and len(audio_data) == len(transcripts):
|
219 |
for transcript in transcripts:
|
220 |
audio = {"path": transcript["path"], "bytes": audio_data[transcript["path"]]}
|
|
|
185 |
audio_data = {}
|
186 |
transcripts = []
|
187 |
key = 0
|
|
|
|
|
188 |
for path, f in files:
|
189 |
audio_data[path] = f.read()
|
190 |
with open(keywords, encoding="utf-8") as f:
|
191 |
next(f)
|
192 |
for row in f:
|
193 |
r = row.split("\t")
|
|
|
194 |
path = 'examples/'+r[0]
|
195 |
ngram = r[1]
|
196 |
transcripts.append({
|
|
|
199 |
"type": "keyword"
|
200 |
})
|
201 |
with open(examples, encoding="utf-8") as f2:
|
|
|
202 |
for row in f2:
|
203 |
r = row.split("\t")
|
|
|
204 |
path = 'examples/'+r[0]
|
205 |
ngram = r[1]
|
206 |
transcripts.append({
|
|
|
208 |
"ngram": ngram,
|
209 |
"type": "example"
|
210 |
})
|
|
|
|
|
211 |
if audio_data and len(audio_data) == len(transcripts):
|
212 |
for transcript in transcripts:
|
213 |
audio = {"path": transcript["path"], "bytes": audio_data[transcript["path"]]}
|