Upload app.py with huggingface_hub

app.py

@@ -220,9 +220,15 @@ def compute_eval(task: str):
     predictor = load_predictor()
     y_true, y_pred = [], []
 
-    for sample in raw_samples:
+    max_samples = min(len(raw_samples), 500)  # keep eval lightweight in Spaces
+    for sample in raw_samples[:max_samples]:
         spec = _to_tensor(sample["data"])
-        res = predictor.predict(spec, return_routing=True)
+        try:
+            res = predictor.predict(spec, return_routing=True)
+        except Exception as exc:
+            # Skip problematic samples but keep going
+            print(f"[WARN] predict failed: {exc}")
+            continue
 
         if task == "comm":
             routing = res.get("routing") or []
@@ -238,7 +244,7 @@ def compute_eval(task: str):
     cm = confusion_matrix(y_true, y_pred, labels=labels)
     f1 = f1_score(y_true, y_pred, labels=labels, average="macro", zero_division=0)
     acc = (np.array(y_true) == np.array(y_pred)).mean()
-    return cm, labels, f1, acc
+    return cm, labels, f1, acc, len(y_true)
 
 
 def plot_confusion(cm: np.ndarray, labels):
@@ -257,13 +263,15 @@ def plot_confusion(cm: np.ndarray, labels):
 
 
 def run_eval(task):
-    cm, labels, f1, acc = compute_eval(task)
+    cm, labels, f1, acc, n = compute_eval(task)
     fig = plot_confusion(cm, labels)
-    summary = f"Task: {task} | Accuracy: {acc:.4f} | Macro F1: {f1:.4f}"
+    summary = f"Task: {task} | Samples: {n} | Accuracy: {acc:.4f} | Macro F1: {f1:.4f}"
     return fig, summary
 
 
+# ------------------------------------------------------------------------------
 # UI
+# ------------------------------------------------------------------------------
 with gr.Blocks(title="LWM-Spectro Demo") as demo:
     gr.Markdown("# 🔬 LWM-Spectro Interactive Demo")
     gr.Markdown("Compare embeddings vs raw for t-SNE, and view quick metrics from the latest MoE checkpoint.")
@@ -297,15 +305,21 @@ with gr.Blocks(title="LWM-Spectro Demo") as demo:
         demo.load(plot_tsne, inputs=[tech_filter, snr_filter, mod_filter, mob_filter, representation, color_by, perplexity, n_iter], outputs=[plot, status])
 
     with gr.Tab("Evaluation (MoE)"):
-        gr.Markdown("Uses the latest MoE checkpoint to score the bundled demo set. Communication uses router gating; SNR/Mobility uses the classifier head.")
+        gr.Markdown("Uses the latest MoE checkpoint to score the bundled demo set.\n\n- **comm**: predicts communication type (LTE/WiFi/5G) via router gating.\n- **snr_mobility**: predicts the SNR/Mobility class via the classifier head.")
         task_choice = gr.Radio(choices=["comm", "snr_mobility"], value="snr_mobility", label="Task")
         eval_btn = gr.Button("Run Evaluation", variant="primary")
         cm_plot = gr.Plot(label="Confusion Matrix")
         eval_summary = gr.Textbox(label="Metrics", interactive=False)
 
-        eval_btn.click(run_eval, inputs=[task_choice], outputs=[cm_plot, eval_summary])
+        def _safe_run(task):
+            try:
+                return run_eval(task)
+            except Exception as exc:
+                return None, f"Error during evaluation: {exc}"
+
+        eval_btn.click(_safe_run, inputs=[task_choice], outputs=[cm_plot, eval_summary])
         # Run once on load for convenience
-        demo.load(run_eval, inputs=[task_choice], outputs=[cm_plot, eval_summary])
+        demo.load(_safe_run, inputs=[task_choice], outputs=[cm_plot, eval_summary])
 
 if __name__ == "__main__":
     demo.launch()