Update README.md

README.md

@@ -71,23 +71,29 @@ Translates informal Sri Lankan code-mixed text written in Roman script into Sinh
 
 ### CTranslate2 (recommended)
 ``` python
+from huggingface_hub import snapshot_download
 import ctranslate2
 from transformers import AutoTokenizer
 
+model_path = snapshot_download("savinugunarathna/small-100-Singlish-Sinhala-CodeMix2-CT2")
+print(f"✓ Downloaded to: {model_path}")
+
 translator = ctranslate2.Translator(
-    "savinugunarathna/small-100-Singlish-Sinhala-CodeMix2-CT2",
+    model_path,
     device="cpu",
     compute_type="int8",
     intra_threads=4,
 )
+
 tokenizer = AutoTokenizer.from_pretrained(
     "savinugunarathna/small-100-Singlish-Sinhala-CodeMix2-CT2"
 )
 
 def translate(text: str) -> str:
     tokenizer.src_lang = "en"
-    ids    = tokenizer.encode(text, add_special_tokens=True)
-    tokens = tokenizer.convert_ids_to_tokens(ids)
+    tokens = tokenizer.convert_ids_to_tokens(
+        tokenizer.encode(text, add_special_tokens=True)
+    )
 
     result = translator.translate_batch(
         [tokens],
@@ -97,17 +103,14 @@ def translate(text: str) -> str:
         repetition_penalty=1.2,
     )
 
-    hyp = result[0].hypotheses[0]
-    if hyp and hyp[0] == "__si__":
-        hyp = hyp[1:]
-    if hyp and hyp[-1] == "</s>":
-        hyp = hyp[:-1]
-
-    hyp_ids = tokenizer.convert_tokens_to_ids(hyp)
+    hyp_ids = tokenizer.convert_tokens_to_ids(result[0].hypotheses[0])
     return tokenizer.decode(hyp_ids, skip_special_tokens=True).strip()
 
-print(translate("mama heta school yanne nehe"))
-# → මම හෙට පාසලට යන්නේ නෑ
+while True:
+    text = input("\nSinglish: ").strip()
+    if text.lower() == "q":
+        break
+    print(f"සිංහල: {translate(text)}")
 ```
 
 ### FastAPI REST API