Views
No views yet
1!pip install transformers
2!pip install sentencepiece
3
4
5from transformers import MarianMTModel, MarianTokenizer
6
7tokenizer = MarianTokenizer.from_pretrained("wolfrage89/annual_report_translation_id_en")
8model = MarianMTModel.from_pretrained("wolfrage89/annual_report_translation_id_en")
9
10#tokenizing bahasa sentence
11bahasa_sentence = "Interpretasi ini merupakan interpretasi atas PSAK 46: Pajak Penghasilan yang bertujuan untuk mengklarifikasi dan memberikan panduan dalam merefleksikan ketidakpastian perlakuan pajak penghasilan dalam laporan keuangan."
12tokenized_bahasa_sentence = tokenizer([bahasa_sentence], return_tensors='pt', max_length=104, truncation=True)
13
14#feeding tokenized sentence into model, the max_legnth have been set to 104 as the model was trained mostly on sentences with this length
15translated_tokens = model.generate(**tokenized_bahasa_sentence, max_length=104)[0]
16
17## decoding the tokens to get english sentence
18english_sentence = tokenizer.decode(translated_tokens, skip_special_tokens=True)
19
20print(english_sentence)
21# This interpretation is an interpretation of PSAK 46: Income Tax that aims to clarify and provide guidance in reflecting the uncertainty of income tax treatments in the financial statements.
22
23