Update app_gradio.py

app_gradio.py

@@ -9,6 +9,10 @@ import pandas as pd
 import numpy as np  # <-- Added numpy for percentiles
 import gradio as gr
 from classify import classify_log, classify_csv
+from processor_bert import preload_models
+
+# ── Preload models in background at startup ─────────────────
+preload_models()
 
 SOURCES = [
     "ModernCRM", "ModernHR", "BillingSystem",
@@ -383,8 +387,11 @@ code, pre {
 
 # ── Functions ───────────────────────────────────────────────────────────────
 def classify_single(source: str, log_message: str):
+    from processor_bert import _model_ready
     if not log_message.strip():
         return "—", "—", "—", "—"
+    if not _model_ready:
+        return "⏳ Model loading...", "Please wait ~60s", "—", "—"
     t0 = time.perf_counter()
     result = classify_log(source, log_message)
     latency_ms = (time.perf_counter() - t0) * 1000
@@ -417,7 +424,7 @@ def classify_batch(file):
     tier_lines  = "\n".join(f"  {TIER_COLORS.get(k,'⚪')} {k}: {v} ({v/total:.0%})" for k, v in tier_counts.items())
     label_lines = "\n".join(f"  • {k}: {v}" for k, v in label_counts.items())
     
-    # Calculate Latencies (Requires 'latency_ms' column in CSV output from classify_csv)
+    # Calculate Latencies
     if "latency_ms" in df.columns and not df["latency_ms"].empty:
         latencies = df["latency_ms"].dropna()
         p50 = np.percentile(latencies, 50)
@@ -525,7 +532,7 @@ Upload a CSV with columns: **`source`**, **`log_message`** Output includes: `pre
                     batch_btn  = gr.Button("▶  CLASSIFY ALL", variant="primary")
                 with gr.Column():
                     csv_output = gr.File(label="📥 DOWNLOAD RESULTS")
-                    # Increased lines from 12 to 16 to fit the new metrics nicely
+                    # Increased lines to 16 to properly fit the latency metrics
                     stats_out  = gr.Textbox(label="📊 STATISTICS", lines=16, interactive=False)
 
             batch_btn.click(
@@ -536,29 +543,4 @@ Upload a CSV with columns: **`source`**, **`log_message`** Output includes: `pre
 
             gr.Markdown("""
 **Sample CSV format:**
-""")
-
-        # ── Tab 3: Architecture ───────────────────────────────────────────
-        with gr.Tab("🏗️ ARCHITECTURE"):
-            gr.Markdown("""
-## 3-Tier Hybrid Pipeline
-
-| Tier | Method | Coverage | Latency | Trigger |
-|------|--------|----------|---------|---------|
-| 🟢 **Regex** | Python `re` patterns | ~21% | < 1ms | Fixed patterns |
-| 🔵 **BERT** | `all-MiniLM-L6-v2` + LogReg | ~79% | 20–80ms | High-volume categories |
-| 🟡 **LLM** | HuggingFace Inference API | ~0.3% | 500–2000ms | LegacyCRM + rare patterns |
-
-## Model Performance
-- **Training data**: 2,410 synthetic enterprise logs
-- **Confidence threshold**: 0.5 (below → escalate to LLM)
-- **Source-aware routing**: `LegacyCRM` → LLM directly
-
-## Environment Variables
-| Secret | Purpose |
-|--------|---------|
-| `HF_TOKEN` | LLM inference for LegacyCRM logs |
-""")
-
-if __name__ == "__main__":
-    demo.launch(server_name="0.0.0.0", server_port=7860, theme=THEME, css=CUSTOM_CSS)
+            """)