Update app.py
Browse files
app.py
CHANGED
|
@@ -8,12 +8,15 @@ import polars as pl
|
|
| 8 |
|
| 9 |
logging.set_verbosity_error()
|
| 10 |
favourite_langs = {"German": "de", "Romanian": "ro", "English": "en", "-----": "-----"}
|
| 11 |
-
all_langs = languagecodes.iso_languages
|
| 12 |
df = pl.read_parquet("isolanguages.parquet")
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
# Language options as list, add favourite languages first
|
| 15 |
options = list(favourite_langs.keys())
|
| 16 |
options.extend(list(all_langs.keys()))
|
|
|
|
| 17 |
models = ["Helsinki-NLP",
|
| 18 |
"facebook/nllb-200-distilled-600M", "facebook/nllb-200-distilled-1.3B", "facebook/nllb-200-1.3B", "facebook/nllb-200-3.3B",
|
| 19 |
"facebook/mbart-large-50-many-to-many-mmt", "facebook/mbart-large-50-one-to-many-mmt", "facebook/mbart-large-50-many-to-one-mmt",
|
|
@@ -110,9 +113,8 @@ class Translators:
|
|
| 110 |
except EnvironmentError as error:
|
| 111 |
try: # Last resort: multi to multi
|
| 112 |
model_name = "Helsinki-NLP/opus-mt-tc-bible-big-mul-mul"
|
| 113 |
-
pipe = pipeline("translation", model=model_name)
|
| 114 |
-
|
| 115 |
-
iso1_dict = {iso[1]: (iso[0], iso[2], iso[3]) for iso in non_empty_iso}
|
| 116 |
iso3tl = iso1_dict.get(self.tl)[2] # 'deu', 'ron', 'eng', 'fra'
|
| 117 |
translation = pipe(f'>>{iso3tl}<< {self.input_text}')
|
| 118 |
return translation[0]['translation_text'], f'Translated from {self.sl} to {self.tl} with {model_name}.'
|
|
@@ -353,8 +355,8 @@ def translate_text(input_text: str, s_language: str, t_language: str, model_name
|
|
| 353 |
("Hallo Welt", "Translated from English to German with Helsinki-NLP.")
|
| 354 |
"""
|
| 355 |
|
| 356 |
-
sl = all_langs[s_language]
|
| 357 |
-
tl = all_langs[t_language]
|
| 358 |
message_text = f'Translated from {s_language} to {t_language} with {model_name}'
|
| 359 |
try:
|
| 360 |
if model_name.startswith("Helsinki-NLP"):
|
|
|
|
| 8 |
|
| 9 |
logging.set_verbosity_error()
|
| 10 |
favourite_langs = {"German": "de", "Romanian": "ro", "English": "en", "-----": "-----"}
|
|
|
|
| 11 |
df = pl.read_parquet("isolanguages.parquet")
|
| 12 |
+
non_empty_isos = df.slice(1).filter(pl.col("ISO639-1") != "").rows()
|
| 13 |
+
# all_langs = languagecodes.iso_languages
|
| 14 |
+
all_langs = {iso[0]: (iso[1], iso[2], iso[3]) for iso in non_empty_isos}
|
| 15 |
|
| 16 |
# Language options as list, add favourite languages first
|
| 17 |
options = list(favourite_langs.keys())
|
| 18 |
options.extend(list(all_langs.keys()))
|
| 19 |
+
|
| 20 |
models = ["Helsinki-NLP",
|
| 21 |
"facebook/nllb-200-distilled-600M", "facebook/nllb-200-distilled-1.3B", "facebook/nllb-200-1.3B", "facebook/nllb-200-3.3B",
|
| 22 |
"facebook/mbart-large-50-many-to-many-mmt", "facebook/mbart-large-50-one-to-many-mmt", "facebook/mbart-large-50-many-to-one-mmt",
|
|
|
|
| 113 |
except EnvironmentError as error:
|
| 114 |
try: # Last resort: multi to multi
|
| 115 |
model_name = "Helsinki-NLP/opus-mt-tc-bible-big-mul-mul"
|
| 116 |
+
pipe = pipeline("translation", model=model_name)
|
| 117 |
+
iso1_dict = {iso[1]: (iso[0], iso[2], iso[3]) for iso in non_empty_isos}
|
|
|
|
| 118 |
iso3tl = iso1_dict.get(self.tl)[2] # 'deu', 'ron', 'eng', 'fra'
|
| 119 |
translation = pipe(f'>>{iso3tl}<< {self.input_text}')
|
| 120 |
return translation[0]['translation_text'], f'Translated from {self.sl} to {self.tl} with {model_name}.'
|
|
|
|
| 355 |
("Hallo Welt", "Translated from English to German with Helsinki-NLP.")
|
| 356 |
"""
|
| 357 |
|
| 358 |
+
sl = all_langs[s_language][0]
|
| 359 |
+
tl = all_langs[t_language][0]
|
| 360 |
message_text = f'Translated from {s_language} to {t_language} with {model_name}'
|
| 361 |
try:
|
| 362 |
if model_name.startswith("Helsinki-NLP"):
|