Update app.py

app.py

@@ -4,63 +4,73 @@ import numpy as np
 from transformers import AutoTokenizer, AutoModel
 from sklearn.metrics.pairwise import cosine_similarity
 
-# Load CodeT5p embedding model
 model_name = "Salesforce/codet5p-110m-embedding"
 
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModel.from_pretrained(model_name)
+tokenizer = AutoTokenizer.from_pretrained(
+    model_name,
+    trust_remote_code=True
+)
+
+model = AutoModel.from_pretrained(
+    model_name,
+    trust_remote_code=True
+)
 
 def get_embedding(code):
     inputs = tokenizer(
         code,
         return_tensors="pt",
         truncation=True,
-        max_length=512
+        max_length=4096,   # <-- 4096 TOKENS
+        padding=True
     )
+
     with torch.no_grad():
         outputs = model(**inputs)
+
+        # mean pooling for embedding
         embedding = outputs.last_hidden_state.mean(dim=1)
+
     return embedding.numpy()
 
-student_codes = []  # stored embeddings
-raw_codes = []      # original text
+stored_embeddings = []
+stored_codes = []
 
 def analyze(code):
-    global student_codes, raw_codes
+    global stored_embeddings, stored_codes
 
     emb = get_embedding(code)
 
-    result = ""
+    text = ""
 
-    if len(student_codes) == 0:
-        result += "First submission stored. No comparison yet.\n"
+    if len(stored_embeddings) == 0:
+        text += "First submission stored. No comparisons yet.\n"
     else:
-        all_embeddings = np.vstack(student_codes)
-        sims = cosine_similarity(emb, all_embeddings)[0]
+        all_embs = np.vstack(stored_embeddings)
+        sims = cosine_similarity(emb, all_embs)[0]
 
-        max_sim = float(np.max(sims))
+        best = float(np.max(sims))
         idx = int(np.argmax(sims))
 
-        result += f"Most similar previous submission score: {max_sim:.3f}\n"
-        result += f"Most similar code index: {idx}\n\n"
+        text += f"Most similar previous submission similarity: {best:.3f}\n"
 
-        if max_sim > 0.9:
-            result += "⚠ Very high similarity — likely same approach / plagiarism\n"
-        elif max_sim > 0.7:
-            result += "🔁 Same structure / same algorithm\n"
-        elif max_sim > 0.5:
-            result += "🟡 Partially similar approach\n"
+        if best > 0.9:
+            text += "⚠ Very high similarity — likely same approach / copied\n"
+        elif best > 0.75:
+            text += "🔁 Same algorithmic structure / same method\n"
+        elif best > 0.5:
+            text += "🟡 Partial similarity\n"
         else:
-            result += "🟢 Unique solution style\n"
+            text += "🟢 Unique solution\n"
 
-    student_codes.append(emb)
-    raw_codes.append(code)
+    stored_embeddings.append(emb)
+    stored_codes.append(code)
 
-    return result
+    return text
 
 gr.Interface(
     fn=analyze,
-    inputs=gr.Textbox(lines=10, label="Student Python Code"),
-    outputs=gr.Textbox(lines=12, label="Code Similarity / Approach Analysis"),
-    title="CodeT5p — Code Similarity & Approach Clustering"
+    inputs=gr.Textbox(lines=12, label="Student Python Code"),
+    outputs=gr.Textbox(lines=12, label="Code Similarity / Approach Detection (4096 tokens)"),
+    title="CodeT5p – Code Embedding Similarity (4096-token input)"
 ).launch()