final update

Author: hasnainnqv

.github/workflows/docker-image.yml

@@ -26,10 +26,10 @@ jobs:
       - name: Build and Push Docker Image
         id: docker_build
         run: |
-          docker build -t shizarizvi/texpose:${{ env.IMAGE_TAG }} .
-          docker tag shizarizvi/texpose:${{ env.IMAGE_TAG }} shizarizvi/texpose:latest
-          docker push shizarizvi/texpose:${{ env.IMAGE_TAG }}
-          docker push shizarizvi/texpose:latest
+          docker build -t shizarizvi/texpose-app:${{ env.IMAGE_TAG }} .
+          docker tag shizarizvi/texpose-app:${{ env.IMAGE_TAG }} shizarizvi/texpose-app:latest
+          docker push shizarizvi/texpose-app:${{ env.IMAGE_TAG }}
+          docker push shizarizvi/texpose-app:latest
         continue-on-error: true
 
       - name: Send Email Notification
@@ -55,8 +55,8 @@ jobs:
             Run: ${{ github.run_id }}
 
             Docker Tags:
-            - shizarizvi/texpose:${{ env.IMAGE_TAG }}
-            - shizarizvi/texpose:latest
+            - shizarizvi/texpose-app:${{ env.IMAGE_TAG }}
+            - shizarizvi/texpose-app:latest
 
             Check details at: https://github.com/${{ github.repository }}/actions/runs/${{ github.run_id }}
 

app/classifier_model.py

@@ -8,7 +8,7 @@
 import gdown
 import os
 
-file_path = "\app\bert_model.pth"  # ML model's path
+file_path = "bert_model.pth"  # ML model's path
 
 if not os.path.exists(file_path):
     print('Model not found locally, downloading from Drive...')
@@ -98,11 +98,13 @@ def classify_text(text, model_ai_hum, model_llm, tokenizer):
     else:
         predicted_label = "AI"
 
-        with torch.no_grad():
-            outputs_llm = model_llm(input_ids, attention_mask=attention_mask)
-            logits_llm = outputs_llm.logits
-            predicted_llm_pred = np.argmax(logits_llm.cpu().numpy(), axis=1).item()
+        # with torch.no_grad():
+        #     outputs_llm = model_llm(input_ids, attention_mask=attention_mask)
+        #     logits_llm = outputs_llm.logits
+        #     predicted_llm_pred = np.argmax(logits_llm.cpu().numpy(), axis=1).item()
 
+
+        predicted_llm_pred = 0
         if predicted_llm_pred == 0:
             predicted_label_llm = "LLAMA"
         else:

app/main.py

@@ -4,6 +4,8 @@
 from app.classifier_model import load_models, classify_text
 import os
 from fastapi.staticfiles import StaticFiles
+from fastapi.responses import JSONResponse
+from fastapi.responses import JSONResponse
 
 
 print("svc_texpose - starting")
@@ -25,16 +27,24 @@
 async def root(request: Request):
     return templates.TemplateResponse("index.html", {"request": request, "prediction": None, "input_text": ""})
 
-@app.post("/classify", response_class=HTMLResponse)
-async def classify(request: Request, input_text: str = Form(default="")):
-    if input_text.strip():  # Ensure input is not empty
+
+
+
+
+@app.post("/api/classify")
+async def classify_api(request: Request):
+    data = await request.json()
+    input_text = data.get("input_text", "")
+
+    if input_text.strip():
         result = classify_text(input_text, model_ai_hum, model_llm, tokenizer)
         prediction = result["type"]
 
-        # Ensure "llm" exists and is not empty
         if result.get("llm"):
             prediction += f" Using {result['llm']}"
 
-        return templates.TemplateResponse("index.html", {"request": request, "prediction": prediction, "input_text": input_text})
+        return JSONResponse(content={"prediction": prediction})
+    
+    return JSONResponse(content={"prediction": "No text provided."})
 
-    return templates.TemplateResponse("index.html", {"request": request, "prediction": "No text provided.", "input_text": ""})
+    

app/templates/index.html

@@ -91,10 +91,11 @@
 
     <section class="page form-section" id="formSection" style="display: none">
       <div id="formContent">
-        <form id="aiDetectorForm" method="POST" action="/classify">
+        <form id="aiDetectorForm" method="POST" action="/api/classify">
           <h1 class="form-text">Text Detector</h1>
           <textarea
-            class="{% if prediction == 'Human' %}human-text{% elif prediction %}ai-text{% endif %}"
+           style="background-color: black; color: white; opacity: 0.7; font-weight: bold;"
+           class="{% if prediction == 'Human' %}human-text{% elif prediction %}ai-text{% endif %}"
             placeholder="Enter text here..."
             rows="5"
             id="input_text"
@@ -105,11 +106,13 @@ <h1 class="form-text">Text Detector</h1>
 </textarea
           >
           <button class="submit" type="submit">Analyze</button>
-          {% if prediction and prediction != "" %}
+        
           <div id="output" style="margin-top: 30px" class="description output">
+            {% if prediction and prediction != "" %}
             Prediction : {{ prediction }}
-          </div>
+          
           {% endif %}
+        </div>
         </form>
       </div>
     </section>
@@ -192,23 +195,58 @@ <h1 class="form-text">Text Detector</h1>
         window.changeTextListenerAdded = true;
       }
 
-      document
-        .getElementById("aiDetectorForm")
-        .addEventListener("submit", function (event) {
-          //     event.preventDefault();
-          //     const input_text = document.getElementById("input_text").value.trim();
-          //     const outputElement = document.getElementById("output");
-
-          //     if (input_text.length === 0) {
-          //       outputElement.innerHTML = "Please enter some text.";
-          //       return;
-          //     }
-
-          //     // // Mock AI detection result
-          //     // const isAI = Math.random() > 0.5 ? "AI-generated" : "Human-written";
-          //     outputElement.innerHTML = `Analysis: The text is generated by <strong>{{prediction}}</strong>.`;
-          outputElement.classList.add("active");
-        });
+ 
+
+
+  document.getElementById("aiDetectorForm").addEventListener("submit", async function (event) {
+  event.preventDefault();
+
+  const inputText = document.getElementById("input_text").value.trim();
+  const outputElement = document.getElementById("output");
+
+  // Clear previous result
+  outputElement.innerHTML = "";
+  outputElement.classList.remove("active");
+
+  if (inputText.length === 0) {
+    outputElement.innerHTML = "Please enter some text.";
+    outputElement.classList.add("active");
+    return;
+  }
+
+  try {
+    const response = await fetch("/api/classify", {
+      method: "POST",
+      headers: {
+        "Content-Type": "application/json",
+      },
+      body: JSON.stringify({ input_text: inputText }),
+    });
+
+    const data = await response.json();
+
+    if (data && data.prediction) {
+      outputElement.innerHTML = `Prediction: <strong>${data.prediction}</strong>`;
+      outputElement.classList.add("active");
+
+      // Optional: Clear the textarea for next input
+    
+    } else {
+      outputElement.innerHTML = "Something went wrong. Try again.";
+      outputElement.classList.add("active");
+    }
+  } catch (error) {
+    console.error("Error:", error);
+    outputElement.innerHTML = "Error occurred while processing.";
+    outputElement.classList.add("active");
+  }
+});
+
+
+
+
+
+
     </script>
   </body>
 </html>