diff --git a/assets/en/sentiment_emotion_others/sentiment/SST2_Llama_7b_chat_ZeroShot.py b/assets/en/sentiment_emotion_others/sentiment/SST2_Llama_7b_chat_ZeroShot.py
index 258c83be..ba5691a6 100644
--- a/assets/en/sentiment_emotion_others/sentiment/SST2_Llama_7b_chat_ZeroShot.py
+++ b/assets/en/sentiment_emotion_others/sentiment/SST2_Llama_7b_chat_ZeroShot.py
@@ -8,7 +8,7 @@ def metadata():
         "author": "Arabic Language Technologies, QCRI, HBKU",
         "model": "Llama-2-13b-chat-hf",
         "description": "Locally hosted Llama-2-13b-chat hf model using FastChat.",
-        "scores": {"Accuracy": "0.861"},
+        "scores": {"Accuracy": "0.924"},
     }
 
 
@@ -30,29 +30,43 @@ def config():
 
 
 def prompt(input_sample):
+    prompt_string = (
+        f"You are tasked with analyzing the sentiment of the given sentence. "
+        f"Please read it carefully and determine whether the sentiment expressed is positive or negative. Provide only label.\n\n"
+        f"sentence: {input_sample.strip()}\n"
+        f"label:\n"
+    )
     return [
         {
             "role": "system",
-            "content": "You are an AI assistant that helps people find information.",
-        },
-        {
-            "role": "user",
-            "content": f'Classify the sentiment of the following sentence as "Positive" or "Negative". Output only the label and nothing else.\nSentence: {input_sample}\nLabel: ',
+            "content": "You are a data annotation expert specializing in sentiment analysis.",
         },
+        {"role": "user", "content": prompt_string},
     ]
 
 
 def post_process(response):
-    out = response["choices"][0]["message"]["content"]
-    out = out.strip().lower()
-    j = out.find("label:")
-    if j > 0:
-        out = out[j + len("label:") :]
-    out = out.strip().lower()
-
-    if out == "positive":
-        return 1
-    elif out == "negative":
-        return 0
+    mapping = {"positive": 1, "negative": 0}
+
+    pred_label = response["choices"][0]["message"]["content"].lower()
+
+    if "\n\nlabel: negative" in pred_label:
+        pred_label = "negative"
+    elif "\n\nlabel: positive" in pred_label:
+        pred_label = "positive"
+    elif "\n\nlabel:" in pred_label:
+        pred_label = pred_label.split("\n\nlabel:")[1]
+        pred_label = pred_label.strip().lower()
+    if pred_label == "positive" or pred_label == "negative":
+        return mapping[pred_label]
+    elif "\n\nnegative" in pred_label:
+        pred_label = "negative"
+    elif "\n\npositive" in pred_label:
+        pred_label = "positive"
+    else:
+        pred_label = None
+
+    if pred_label is not None:
+        return mapping[pred_label]
     else:
         return None