Spaces:
Running
Running
logs off
Browse files
app.py
CHANGED
@@ -377,14 +377,14 @@ def translatePy(text, src_lang='rus_Cyrl', tgt_lang='krc_Cyrl',
|
|
377 |
text, return_tensors='pt', padding=True, truncation=True,
|
378 |
max_length=max_input_length
|
379 |
)
|
380 |
-
print(f'Inputs: {inputs}')
|
381 |
result = model_translate.generate(
|
382 |
**inputs.to(model_translate.device),
|
383 |
forced_bos_token_id=tokenizer.convert_tokens_to_ids(tgt_lang),
|
384 |
max_new_tokens=int(a + b * inputs.input_ids.shape[1]),
|
385 |
num_beams=num_beams, **kwargs
|
386 |
)
|
387 |
-
print(f'Outputs: {result}')
|
388 |
return tokenizer.batch_decode(result, skip_special_tokens=True)
|
389 |
|
390 |
|
@@ -413,13 +413,13 @@ def translateDisp(text, from_, to, dialect):
|
|
413 |
# Разбиваем текст на предложения, сохраняя знаки препинания
|
414 |
text = re.findall(r'.+?[.!?\n](?:\s|$)', text)
|
415 |
|
416 |
-
print(f'Split text: {text}')
|
417 |
|
418 |
str_ = translatePy(text, src_lang = from_, tgt_lang = to)
|
419 |
|
420 |
-
print(f'Translated text: {str_}')
|
421 |
str_ = ' '.join(str_).strip()
|
422 |
-
print(f'Jointed text: {str_}')
|
423 |
|
424 |
if to == 'krc_Cyrl':
|
425 |
str_ = fromModel(str_, dialect = dialect)
|
|
|
377 |
text, return_tensors='pt', padding=True, truncation=True,
|
378 |
max_length=max_input_length
|
379 |
)
|
380 |
+
#print(f'Inputs: {inputs}')
|
381 |
result = model_translate.generate(
|
382 |
**inputs.to(model_translate.device),
|
383 |
forced_bos_token_id=tokenizer.convert_tokens_to_ids(tgt_lang),
|
384 |
max_new_tokens=int(a + b * inputs.input_ids.shape[1]),
|
385 |
num_beams=num_beams, **kwargs
|
386 |
)
|
387 |
+
#print(f'Outputs: {result}')
|
388 |
return tokenizer.batch_decode(result, skip_special_tokens=True)
|
389 |
|
390 |
|
|
|
413 |
# Разбиваем текст на предложения, сохраняя знаки препинания
|
414 |
text = re.findall(r'.+?[.!?\n](?:\s|$)', text)
|
415 |
|
416 |
+
#print(f'Split text: {text}')
|
417 |
|
418 |
str_ = translatePy(text, src_lang = from_, tgt_lang = to)
|
419 |
|
420 |
+
#print(f'Translated text: {str_}')
|
421 |
str_ = ' '.join(str_).strip()
|
422 |
+
#print(f'Jointed text: {str_}')
|
423 |
|
424 |
if to == 'krc_Cyrl':
|
425 |
str_ = fromModel(str_, dialect = dialect)
|