Update README.md

README.md

@@ -23,3 +23,48 @@ I will also provide dataset which can be used in Azerbaijani medical QA and rela
 
 # How to use
 
+Here is how to use this model.
+
+__Firstly, you need to build a dictionary with medical branch names and their numbers, because target is encoded and model output will be a number.__
+
+```python
+branch_dict = {0: 'Endoskopist', 1: 'Nevropatoloq',2: 'Dermato veneroloq',3: 'Qastroenteroloq',
+ 4: 'Psixoloq', 5: 'Pediatr', 6: 'Proktoloq', 7: 'Endokrinoloq',
+ 8: 'Psixoterapevt', 9: 'Allerqoloq', 10: 'Oftalmoloq', 11: 'Kardioloq', 12: 'Uroloq',
+ 13: 'Plastik cərrah', 14: 'Cərrah-proktoloq', 15: 'Ümumi cərrah',
+ 16: 'Hepatoloq', 17: 'LOR həkimi', 18: 'Ginekoloq'}
+```
+
+__Secondly, we will use a simple Python function in order to convert model result to branch name.__
+
+```python
+def result_helper_funct(model_result):
+    
+    result = model_result[0][0]
+    if result in branch_dict.keys(): 
+        return branch_dict[result]
+```
+
+__Then, we need to install simpletransformers library__
+
+```python
+!pip install simpletransformers
+```
+__After succesfully installing, use pre-trained model.__
+
+```python
+from simpletransformers.classification import ClassificationModel
+model = ClassificationModel("bert", "nijatzeynalov/azerbaijani-medical-question-classification", use_cuda=False)
+```
+
+__ At the next step, we just write down the text we want to classify and use our helper function.__
+
+```python
+sample_text = 'salam menim qulagimda agri var'
+result = model.predict([sample_text])
+
+result_helper_funct(result)
+
+'LOR həkimi'
+
+```