max_new_tokens

app.py

@@ -10,7 +10,8 @@ tokenizer = M2M100Tokenizer.from_pretrained("facebook/m2m100_1.2B")
 
 
 this_description = '''
-Using facebook/m2m100-12B-avg-5-ckpt pre-trained model. Language code:
+Using facebook/m2m100-12B-avg-5-ckpt pre-trained model. Some language code:
+
 Chinese(zh)
 English(en)
 Hindi(hi)
@@ -23,106 +24,106 @@ Vietnamese(vi)
 
 # From facebook/m2m100-12B-avg-5-ckpt
 lang_codes = {
-  "Afrikaans": "af",
-  "Amharic": "am",
-  "Arabic": "ar",
-  "Asturian": "ast",
-  "Azerbaijani": "az",
-  "Bashkir": "ba",
-  "Belarusian": "be",
-  "Bulgarian": "bg",
-  "Bengali": "bn",
-  "Breton": "br",
-  "Bosnian": "bs",
-  "Catalan; Valencian": "ca",
-  "Cebuano": "ceb",
-  "Czech": "cs",
-  "Welsh": "cy",
-  "Danish": "da",
-  "German": "de",
-  "Greeek": "el",
-  "English": "en",
-  "Spanish": "es",
-  "Estonian": "et",
-  "Persian": "fa",
-  "Fulah": "ff",
-  "Finnish": "fi",
-  "French": "fr",
-  "Western Frisian": "fy",
-  "Irish": "ga",
-  "Gaelic; Scottish Gaelic": "gd",
-  "Galician": "gl",
-  "Gujarati": "gu",
-  "Hausa": "ha",
-  "Hebrew": "he",
-  "Hindi": "hi",
-  "Croatian": "hr",
-  "Haitian; Haitian Creole": "ht",
-  "Hungarian": "hu",
-  "Armenian": "hy",
-  "Indonesian": "id",
-  "Igbo": "ig",
-  "Iloko": "ilo",
-  "Icelandic": "is",
-  "Italian": "it",
-  "Japanese": "ja",
-  "Javanese": "jv",
-  "Georgian": "ka",
-  "Kazakh": "kk",
-  "Central Khmer": "km",
-  "Kannada": "kn",
-  "Korean": "ko",
-  "Luxembourgish; Letzeburgesch": "lb",
-  "Ganda": "lg",
-  "Lingala": "ln",
-  "Lao": "lo",
-  "Lithuanian": "lt",
-  "Latvian": "lv",
-  "Malagasy": "mg",
-  "Macedonian": "mk",
-  "Malayalam": "ml",
-  "Mongolian": "mn",
-  "Marathi": "mr",
-  "Malay": "ms",
-  "Burmese": "my",
-  "Nepali": "ne",
-  "Dutch; Flemish": "nl",
-  "Norwegian": "no",
-  "Northern Sotho": "ns",
-  "Occitan": "oc",
-  "Oriya": "or",
-  "Panjabi; Punjabi": "pa",
-  "Polish": "pl",
-  "Pushto": "ps",
-  "Portuguese": "pt",
-  "Romanian; Moldavian; Moldovan": "ro",
-  "Russian": "ru",
-  "Sindhi": "sd",
-  "Sinhala; Sinhalese": "si",
-  "Slovak": "sk",
-  "Slovenian": "sl",
-  "Somali": "so",
-  "Albanian": "sq",
-  "Serbian": "sr",
-  "Swati": "ss",
-  "Sundanese": "su",
-  "Swedish": "sv",
-  "Swahili": "sw",
-  "Tamil": "ta",
-  "Thai": "th",
-  "Tagalog": "tl",
-  "Tswana": "tn",
-  "Turkish": "tr",
-  "Ukrainian": "uk",
-  "Urdu": "ur",
-  "Uzbek": "uz",
-  "Vietnamese": "vi",
-  "Wolof": "wo",
-  "Xhosa": "xh",
-  "Yiddish": "yi",
-  "Yoruba": "yo",
-  "Chinese": "zh",
-  "Zulu": "zu"
+    "Afrikaans": "af",
+    "Amharic": "am",
+    "Arabic": "ar",
+    "Asturian": "ast",
+    "Azerbaijani": "az",
+    "Bashkir": "ba",
+    "Belarusian": "be",
+    "Bulgarian": "bg",
+    "Bengali": "bn",
+    "Breton": "br",
+    "Bosnian": "bs",
+    "Catalan; Valencian": "ca",
+    "Cebuano": "ceb",
+    "Czech": "cs",
+    "Welsh": "cy",
+    "Danish": "da",
+    "German": "de",
+    "Greeek": "el",
+    "English": "en",
+    "Spanish": "es",
+    "Estonian": "et",
+    "Persian": "fa",
+    "Fulah": "ff",
+    "Finnish": "fi",
+    "French": "fr",
+    "Western Frisian": "fy",
+    "Irish": "ga",
+    "Gaelic; Scottish Gaelic": "gd",
+    "Galician": "gl",
+    "Gujarati": "gu",
+    "Hausa": "ha",
+    "Hebrew": "he",
+    "Hindi": "hi",
+    "Croatian": "hr",
+    "Haitian; Haitian Creole": "ht",
+    "Hungarian": "hu",
+    "Armenian": "hy",
+    "Indonesian": "id",
+    "Igbo": "ig",
+    "Iloko": "ilo",
+    "Icelandic": "is",
+    "Italian": "it",
+    "Japanese": "ja",
+    "Javanese": "jv",
+    "Georgian": "ka",
+    "Kazakh": "kk",
+    "Central Khmer": "km",
+    "Kannada": "kn",
+    "Korean": "ko",
+    "Luxembourgish; Letzeburgesch": "lb",
+    "Ganda": "lg",
+    "Lingala": "ln",
+    "Lao": "lo",
+    "Lithuanian": "lt",
+    "Latvian": "lv",
+    "Malagasy": "mg",
+    "Macedonian": "mk",
+    "Malayalam": "ml",
+    "Mongolian": "mn",
+    "Marathi": "mr",
+    "Malay": "ms",
+    "Burmese": "my",
+    "Nepali": "ne",
+    "Dutch; Flemish": "nl",
+    "Norwegian": "no",
+    "Northern Sotho": "ns",
+    "Occitan": "oc",
+    "Oriya": "or",
+    "Panjabi; Punjabi": "pa",
+    "Polish": "pl",
+    "Pushto": "ps",
+    "Portuguese": "pt",
+    "Romanian; Moldavian; Moldovan": "ro",
+    "Russian": "ru",
+    "Sindhi": "sd",
+    "Sinhala; Sinhalese": "si",
+    "Slovak": "sk",
+    "Slovenian": "sl",
+    "Somali": "so",
+    "Albanian": "sq",
+    "Serbian": "sr",
+    "Swati": "ss",
+    "Sundanese": "su",
+    "Swedish": "sv",
+    "Swahili": "sw",
+    "Tamil": "ta",
+    "Thai": "th",
+    "Tagalog": "tl",
+    "Tswana": "tn",
+    "Turkish": "tr",
+    "Ukrainian": "uk",
+    "Urdu": "ur",
+    "Uzbek": "uz",
+    "Vietnamese": "vi",
+    "Wolof": "wo",
+    "Xhosa": "xh",
+    "Yiddish": "yi",
+    "Yoruba": "yo",
+    "Chinese": "zh",
+    "Zulu": "zu"
 }
 
 
@@ -131,6 +132,22 @@ def m2m_translate(Input_Text, from_lang, to_lang):
 
     encoded_from_lang = tokenizer(Input_Text, return_tensors="pt")
 
+    generated_tokens = model.generate(
+        **encoded_from_lang,
+        max_new_tokens=200,
+        forced_bos_token_id=tokenizer.get_lang_id(lang_codes[to_lang])
+    )
+
+    res = tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
+
+    return res[0]
+
+
+def m2m_translate2(Input_Text, from_lang, to_lang):
+    tokenizer.src_lang = lang_codes[from_lang]
+
+    encoded_from_lang = tokenizer(Input_Text, return_tensors="pt")
+
     generated_tokens = model.generate(
         **encoded_from_lang, forced_bos_token_id=tokenizer.get_lang_id(lang_codes[to_lang]))
 
@@ -151,34 +168,34 @@ iface = gr.Interface(
         gr.Textbox(lines=5, placeholder="Enter text", label="Text input"),
 
         gr.Radio(
-    choices=[
-      'Burmese',
-        'Chinese',
-        'English',
-        'Hindi',
-        'Japanese',
-        'Sinhala',
-        'Thai',
-        'Vietnamese'
-    ],
-    value='Vietnamese',
-    label='From language'
-),
-
-gr.Radio(
-    choices=[
-      'Burmese',
-        'Chinese',
-        'English',
-        'Hindi',
-        'Japanese',
-        'Sinhala',
-        'Thai',
-        'Vietnamese'
-    ],
-    value='English',
-    label='To language'
-),
+            choices=[
+                'Burmese',
+                'Chinese',
+                'English',
+                'Hindi',
+                'Japanese',
+                'Sinhala',
+                'Thai',
+                'Vietnamese'
+            ],
+            value='Vietnamese',
+            label='From language'
+        ),
+
+        gr.Radio(
+            choices=[
+                'Burmese',
+                'Chinese',
+                'English',
+                'Hindi',
+                'Japanese',
+                'Sinhala',
+                'Thai',
+                'Vietnamese'
+            ],
+            value='English',
+            label='To language'
+        ),
     ],
     outputs="text")