Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
|
@@ -76,6 +76,7 @@ def process_audio_to_text(audio_path):
|
|
| 76 |
"English",
|
| 77 |
api_name="/s2tt"
|
| 78 |
)
|
|
|
|
| 79 |
return result['text'] # Adjust based on the actual response
|
| 80 |
|
| 81 |
def process_text_to_audio(text, target_language="English"):
|
|
@@ -119,8 +120,6 @@ def process_input(image=None, file=None, audio=None, text=""):
|
|
| 119 |
for idx in range(len((list(ocr_prediction)[0][1]))):
|
| 120 |
final_text += " "
|
| 121 |
final_text += list((list(ocr_prediction)[0][1])[idx])[1][1]
|
| 122 |
-
# final_text += "\n" + ocr_processor.process_image(image)
|
| 123 |
-
print(final_text)
|
| 124 |
if file is not None:
|
| 125 |
if file.name.lower().endswith(('.png', '.jpg', '.jpeg')):
|
| 126 |
pil_image = Image.open(file)
|
|
@@ -129,17 +128,15 @@ def process_input(image=None, file=None, audio=None, text=""):
|
|
| 129 |
for idx in range(len((list(ocr_prediction)[0][1]))):
|
| 130 |
final_text += " "
|
| 131 |
final_text += list((list(ocr_prediction)[0][1])[idx])[1][1]
|
| 132 |
-
print(final_text)
|
| 133 |
elif file.name.lower().endswith('.pdf'):
|
| 134 |
ocr_prediction = ocr_processor.process_pdf(file.name)
|
| 135 |
# gettig text from ocr object
|
| 136 |
for idx in range(len((list(ocr_prediction)[0][1]))):
|
| 137 |
final_text += " "
|
| 138 |
final_text += list((list(ocr_prediction)[0][1])[idx])[1][1]
|
| 139 |
-
print(final_text)
|
| 140 |
-
# final_text += "\n" + ocr_processor.process_pdf(file.name)
|
| 141 |
else:
|
| 142 |
final_text += "\nUnsupported file type."
|
|
|
|
| 143 |
if audio is not None:
|
| 144 |
audio_text = process_audio_to_text(audio)
|
| 145 |
final_text += "\n" + audio_text
|
|
@@ -151,6 +148,7 @@ def process_input(image=None, file=None, audio=None, text=""):
|
|
| 151 |
temperature=0.5
|
| 152 |
)
|
| 153 |
generated_text = response.generations[0].text
|
|
|
|
| 154 |
|
| 155 |
# Process generated text with command-nightly model
|
| 156 |
response = co.generate(
|
|
|
|
| 76 |
"English",
|
| 77 |
api_name="/s2tt"
|
| 78 |
)
|
| 79 |
+
print("Audio Result: ", result)
|
| 80 |
return result['text'] # Adjust based on the actual response
|
| 81 |
|
| 82 |
def process_text_to_audio(text, target_language="English"):
|
|
|
|
| 120 |
for idx in range(len((list(ocr_prediction)[0][1]))):
|
| 121 |
final_text += " "
|
| 122 |
final_text += list((list(ocr_prediction)[0][1])[idx])[1][1]
|
|
|
|
|
|
|
| 123 |
if file is not None:
|
| 124 |
if file.name.lower().endswith(('.png', '.jpg', '.jpeg')):
|
| 125 |
pil_image = Image.open(file)
|
|
|
|
| 128 |
for idx in range(len((list(ocr_prediction)[0][1]))):
|
| 129 |
final_text += " "
|
| 130 |
final_text += list((list(ocr_prediction)[0][1])[idx])[1][1]
|
|
|
|
| 131 |
elif file.name.lower().endswith('.pdf'):
|
| 132 |
ocr_prediction = ocr_processor.process_pdf(file.name)
|
| 133 |
# gettig text from ocr object
|
| 134 |
for idx in range(len((list(ocr_prediction)[0][1]))):
|
| 135 |
final_text += " "
|
| 136 |
final_text += list((list(ocr_prediction)[0][1])[idx])[1][1]
|
|
|
|
|
|
|
| 137 |
else:
|
| 138 |
final_text += "\nUnsupported file type."
|
| 139 |
+
print(final_text)
|
| 140 |
if audio is not None:
|
| 141 |
audio_text = process_audio_to_text(audio)
|
| 142 |
final_text += "\n" + audio_text
|
|
|
|
| 148 |
temperature=0.5
|
| 149 |
)
|
| 150 |
generated_text = response.generations[0].text
|
| 151 |
+
print("Generated Text: ", generated_text)
|
| 152 |
|
| 153 |
# Process generated text with command-nightly model
|
| 154 |
response = co.generate(
|