removed custom inferencing

code/inference.py

@@ -1,74 +0,0 @@
-import json
-
-from typing import Dict, List, Any
-from setfit import SetFitModel
-  
-def model_fn(model_dir):
-        # load model
-        model = SetFitModel.from_pretrained(model_dir)
-       
-        return model
-
-
-def predict_fn(data, setfitModel):
-    # destruct model and tokenizer
-    model = setfitModel
-    
-    # get inputs
-    inputs = data.pop("inputs", data)
-    if isinstance(inputs, str):
-        inputs = [inputs]
-    
-    # exerciselabels
-    exerciselabels = ['positive experience',
-                     'power posing',
-                     'worry vs rumination',
-                     'self-confidence',
-                     'negative emotions',
-                     'sleep',
-                     'loneliness',
-                     'imaginary friend',
-                     'perfectionism',
-                     'negative self-talk',
-                     'woop',
-                     'venting',
-                     'worry window',
-                     'act of kindness',
-                     'blowing balloons',
-                     'feeling on anger',
-                     'power of smile',
-                     'body scan',
-                     'stress enhancing thoughts',
-                     'anger ball of fire',
-                     'emotions',
-                     'lean against wall',
-                     'breathing',
-                     'crossed arms']
-    
-    # run normal prediction
-    # preds = model.predict(inputs)
-    preds = model.predict(inputs) #gives predicted label with highest prob
-    scores = model.predict_proba(inputs)
-  
-    # Show predicted exercise, requires you to have stored the 'exerciselabels' somewhere
-    label = [[el for el, p in zip(exerciselabels, ps) if p] for ps in preds]
-    
-    # Modify the label array
-    modified_label = label[0]
-
-    # Extract the positives probabilities  from each inner array
-    modified_proba = [[inner[0][1]] for item, inner in zip(scores, scores)]
-
-
-    score = [[el for el, p in zip(modified_proba, ps) if p] for ps in preds]
-
-    # Modify the score array
-    modified_score = score[0]
-
-    # Use zip to combine corresponding elements from both arrays
-    # combined_array = list(zip(modified_label, modified_score))
-
-    combined_dict = [{key: value for key, value in zip(modified_label, modified_score)}]
-
-    # for element in combined_array:
-    return preds

code/requirements.txt

@@ -1 +0,0 @@
-setfit==0.3.0