Update parameters for Arabic

Author: HugoLaurencon

ac_dc/parameters_filtering.py

@@ -94,28 +94,28 @@
     "cond_check_number_words": True,
     "tokenization": False,
     "strip_characters": special_characters_default,
-    "number_words_min_cutoff": 1,
+    "number_words_min_cutoff": 20,
     "number_words_max_cutoff": 100000,
     "cond_check_character_repetition_removal": True,
     "character_repetition_length": 10,
-    "character_repetition_max_cutoff": 0.106,
+    "character_repetition_max_cutoff": 0.20,
     "cond_check_word_repetition_removal": True,
     "word_repetition_length": 5,
-    "word_repetition_max_cutoff": 0.19,
+    "word_repetition_max_cutoff": 0.34,
     "cond_check_special_characters": True,
     "special_characters": special_characters_default,
-    "special_characters_max_cutoff": 0.45,
+    "special_characters_max_cutoff": 0.30,
     "cond_words_augmentation": False,
     "words_augmentation_group_sizes": [],
     "words_augmentation_join_char": "",
     "cond_check_stopwords": True,
-    "stopwords_min_cutoff": 0,
+    "stopwords_min_cutoff": 0.07,
     "cond_check_flagged_words": True,
-    "flagged_words_max_cutoff": 0.2,
+    "flagged_words_max_cutoff": 0.03,
     "cond_check_lang_id": True,
-    "lang_id_min_cutoff": 0.75,
+    "lang_id_min_cutoff": 0.81,
     "cond_check_perplexity": True,
-    "perplexity_max_cutoff": 1000000,
+    "perplexity_max_cutoff": 2300,
 }
 
 parameters_filtering_arz = {
@@ -868,7 +868,6 @@
 
 parameters_filtering = {
     "default": parameters_filtering_default,
-    "af": parameters_filtering_af,
     "ar": parameters_filtering_ar,
     "arz": parameters_filtering_arz,
     "as": parameters_filtering_as,