Update README.md

README.md

@@ -1,21 +1,15 @@
-# Telugu Question-Answering model trained on Tydiqa dataset from Google
+# code-mixed Kannada-English word-level identification trained on Kanglish dataset from ICON2022 
 
 #### How to use
-Use the below script from your python terminal as the web interface for inference has few encoding issues for Telugu
+Use the below script from your python terminal 
 ```python
-from transformers.pipelines import pipeline, AutoModelForQuestionAnswering, AutoTokenizer
-model = AutoModelForQuestionAnswering.from_pretrained(model_name)
-tokenizer = AutoTokenizer.from_pretrained("kuppuluri/telugu_bertu_tydiqa",
-                                          clean_text=False,
-                                          handle_chinese_chars=False,
-                                          strip_accents=False,
-                                          wordpieces_prefix='##')
-nlp = pipeline('question-answering', model=model, tokenizer=tokenizer)
-result = nlp({'question': question, 'context': context})
+import transformers
+from transformers import AutoModelForSequenceClassification
+model AutoModelForSequenceClassification.from_pretrained("tiya1012/distilka_applied")
 ```
 
 ## Training data
-I used Tydiqa Telugu data from Google https://github.com/google-research-datasets/tydiqa
+I used code-mixed dataset from https://sites.google.com/view/kanglishicon2022/home
+
 
-PS: If you find my model useful, I would appreciate a note from you as it would encourage me to continue improving it and also add new models.