Upload folder using huggingface_hub

README.md

@@ -1,12 +1,10 @@
----
-title: Codesheriff Inference
-emoji: 📚
-colorFrom: red
-colorTo: gray
-sdk: gradio
-sdk_version: 6.9.0
-app_file: app.py
-pinned: false
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+---
+title: CodeSheriff Inference
+emoji: 🔍
+colorFrom: blue
+colorTo: red
+sdk: gradio
+sdk_version: "4.44.0"
+app_file: app.py
+pinned: false
+---

app.py

@@ -0,0 +1,67 @@
+"""
+CodeSheriff Inference Space
+
+A lightweight Gradio app that loads the fine-tuned CodeBERT classifier
+and exposes a /predict API endpoint.  Called remotely by the Render backend.
+"""
+
+import gradio as gr
+import torch
+from transformers import AutoTokenizer, AutoModelForSequenceClassification
+
+MODEL_ID = "jayansh21/codesheriff-bug-classifier"
+NUM_LABELS = 5
+MAX_LENGTH = 512
+LABEL_NAMES = {
+    0: "Clean",
+    1: "Null Reference Risk",
+    2: "Type Mismatch",
+    3: "Security Vulnerability",
+    4: "Logic Flaw",
+}
+
+print("Loading CodeSheriff classifier …")
+tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
+model = AutoModelForSequenceClassification.from_pretrained(
+    MODEL_ID, num_labels=NUM_LABELS
+)
+model.eval()
+print("Model loaded ✅")
+
+
+def predict(code_snippet: str) -> dict:
+    """Classify a code snippet and return label, confidence, label_id."""
+    if not code_snippet or not code_snippet.strip():
+        return {"label": "Clean", "confidence": 0.0, "label_id": 0}
+
+    encoding = tokenizer(
+        code_snippet,
+        truncation=True,
+        padding="max_length",
+        max_length=MAX_LENGTH,
+        return_tensors="pt",
+    )
+    with torch.no_grad():
+        outputs = model(**encoding)
+
+    probs = torch.softmax(outputs.logits, dim=-1).squeeze(0)
+    label_id = int(torch.argmax(probs).item())
+    confidence = float(probs[label_id].item())
+
+    return {
+        "label": LABEL_NAMES.get(label_id, f"Unknown({label_id})"),
+        "confidence": round(confidence, 4),
+        "label_id": label_id,
+    }
+
+
+demo = gr.Interface(
+    fn=predict,
+    inputs=gr.Textbox(label="Code Snippet", lines=5, placeholder="Paste code here …"),
+    outputs=gr.JSON(label="Classification"),
+    title="🔍 CodeSheriff Bug Classifier",
+    description="Fine-tuned CodeBERT model for detecting common bug patterns.",
+    api_name="predict",
+)
+
+demo.launch()

requirements.txt

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