Update Milestone5API.py

Milestone5API.py

@@ -3,6 +3,37 @@ import re
 from pytube import YouTube
 from youtube_transcript_api import YouTubeTranscriptApi
 from transformers import MarianMTModel, MarianTokenizer
+def translate_text_file(input_text_path, output_text_path, source_lang='en', target_lang='fr', model_name="Helsinki-NLP/opus-mt-en-fr", batch_size=8):
+    model = MarianMTModel.from_pretrained(model_name)
+    tokenizer = MarianTokenizer.from_pretrained(model_name)
+
+    def translate_batch(model, tokenizer, sentences):
+        sentences = [f"{source_lang}: {sentence}" for sentence in sentences]
+        input_ids = tokenizer(sentences, return_tensors="pt", padding=True, truncation=True)["input_ids"]
+        translation_ids = model.generate(input_ids)
+        translated_texts = tokenizer.batch_decode(translation_ids, skip_special_tokens=True)
+        return translated_texts
+
+    def read_text_from_file(file_path):
+        with open(file_path, 'r', encoding='utf-8') as file:
+            text = file.readlines()
+        return text
+
+    def write_text_to_file(file_path, translated_texts):
+        with open(file_path, 'w', encoding='utf-8') as file:
+            file.writelines([f"{line}\n" for line in translated_texts])
+
+    translated_lines = []
+
+    input_lines = read_text_from_file(input_text_path)
+
+    for i in range(0, len(input_lines), batch_size):
+        batch = input_lines[i:i + batch_size]
+        translated_batch = translate_batch(model, tokenizer, batch)
+        translated_lines.extend(translated_batch)
+
+    write_text_to_file(output_text_path, translated_lines)
+    return translated_lines
 
 def download_video_transcript(video_url, source_lang='en', target_lang='fr'):
     try: