Upload 2 files

app.py

@@ -0,0 +1,61 @@
+import gradio as gr
+from transformers import AutoModelForSequenceClassification, AutoTokenizer
+import torch
+
+# Load model
+model = AutoModelForSequenceClassification.from_pretrained("collapseindex/ProBERT-1.0")
+tokenizer = AutoTokenizer.from_pretrained("collapseindex/ProBERT-1.0")
+
+LABELS = ["process_clarity", "rhetorical_confidence", "scope_blur"]
+
+EXAMPLES = [
+    ["This revolutionary AI will transform your business and guarantee results."],
+    ["Step 1: Load data. Step 2: Validate schema. Step 3: Return results."],
+    ["Trust your intuition and embrace the journey. The universe has a plan."],
+    ["First, check if the input is null. If null, return error. Otherwise, process the request."],
+    ["Our cutting-edge solution leverages synergies to maximize value propositions."],
+]
+
+def classify(text):
+    inputs = tokenizer(text, return_tensors="pt", truncation=True, max_length=128)
+    with torch.no_grad():
+        outputs = model(**inputs)
+    probs = torch.softmax(outputs.logits, dim=1)[0]
+    
+    # Get top prediction and confidence
+    top_prob = float(probs.max())
+    top_idx = int(probs.argmax())
+    top_label = LABELS[top_idx]
+    
+    # Return predictions dict and confidence text
+    predictions = {LABELS[i]: float(probs[i]) for i in range(len(LABELS))}
+    confidence_text = f"**Top Prediction:** {top_label} (Confidence: {top_prob:.1%})"
+    
+    return predictions, confidence_text
+
+demo = gr.Interface(
+    fn=classify,
+    inputs=gr.Textbox(lines=3, placeholder="Enter text here...", label="Input Text"),
+    outputs=[
+        gr.Label(
+            num_top_classes=3, 
+            label="Predictions"
+        ),
+        gr.Markdown(label="Confidence Score")
+    ],
+    title="ProBERT v1.0 - Rhetorical Confidence Detection",
+    description="""
+    **Detects rhetorical overconfidence in text.**
+    
+    - 🟢 **process_clarity**: Step-by-step reasoning you can verify
+    - 🟠 **rhetorical_confidence**: Assertive claims without supporting process
+    - 🔴 **scope_blur**: Vague generalizations with ambiguous boundaries
+    
+    **Model:** [collapseindex/ProBERT-1.0](https://huggingface.co/collapseindex/ProBERT-1.0)
+    """,
+    examples=EXAMPLES,
+    theme="default",
+)
+
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -0,0 +1,3 @@
+transformers
+torch
+gradio