process and clean files
Browse files- src/train_tokenizer.ipynb +45 -0
src/train_tokenizer.ipynb
CHANGED
@@ -20,6 +20,14 @@
|
|
20 |
"# load_dataset(\"balochiml/balochi-language-data\", data_dir=\"data\", cache_dir=\"../data\")"
|
21 |
]
|
22 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
23 |
{
|
24 |
"cell_type": "code",
|
25 |
"execution_count": 13,
|
@@ -55,6 +63,43 @@
|
|
55 |
"len(txt_paths)\n"
|
56 |
]
|
57 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
58 |
{
|
59 |
"cell_type": "code",
|
60 |
"execution_count": null,
|
|
|
20 |
"# load_dataset(\"balochiml/balochi-language-data\", data_dir=\"data\", cache_dir=\"../data\")"
|
21 |
]
|
22 |
},
|
23 |
+
{
|
24 |
+
"attachments": {},
|
25 |
+
"cell_type": "markdown",
|
26 |
+
"metadata": {},
|
27 |
+
"source": [
|
28 |
+
"# Generate the processed data without English characters"
|
29 |
+
]
|
30 |
+
},
|
31 |
{
|
32 |
"cell_type": "code",
|
33 |
"execution_count": 13,
|
|
|
63 |
"len(txt_paths)\n"
|
64 |
]
|
65 |
},
|
66 |
+
{
|
67 |
+
"cell_type": "code",
|
68 |
+
"execution_count": 17,
|
69 |
+
"metadata": {},
|
70 |
+
"outputs": [],
|
71 |
+
"source": [
|
72 |
+
"import re\n",
|
73 |
+
"\n",
|
74 |
+
"def clean_text(file_path):\n",
|
75 |
+
" # Open the file and read it into memory\n",
|
76 |
+
" with open(file_path, 'r', encoding='utf-8') as file:\n",
|
77 |
+
" text = file.read()\n",
|
78 |
+
"\n",
|
79 |
+
" # Remove English-language characters and numbers\n",
|
80 |
+
" text = re.sub(r'[a-zA-Z0-9]', '', text)\n",
|
81 |
+
"\n",
|
82 |
+
" # Remove any excess whitespace\n",
|
83 |
+
" text = re.sub(r'[^\\S\\n]+', ' ', text)\n",
|
84 |
+
"\n",
|
85 |
+
" return text"
|
86 |
+
]
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"cell_type": "code",
|
90 |
+
"execution_count": 18,
|
91 |
+
"metadata": {},
|
92 |
+
"outputs": [],
|
93 |
+
"source": [
|
94 |
+
"for path in txt_paths:\n",
|
95 |
+
" cleaned_text = clean_text(path)\n",
|
96 |
+
"\n",
|
97 |
+
" # write the cleaned text to a new file with an incremented filename\n",
|
98 |
+
" # write the files all into the '../data/processed_text' directory\n",
|
99 |
+
" with open(f'../data/processed_text/{path.split(\"/\")[-1]}', 'w', encoding='utf-8') as file:\n",
|
100 |
+
" file.write(cleaned_text)\n"
|
101 |
+
]
|
102 |
+
},
|
103 |
{
|
104 |
"cell_type": "code",
|
105 |
"execution_count": null,
|