Update app.py

app.py

@@ -1,61 +1,21 @@
-"""
-HF Space · WFGY simulation demo
-(complete file – paste/replace your current app.py)
-"""
-
 import io
 import numpy as np
 import gradio as gr
 import matplotlib.pyplot as plt
 from transformers import AutoTokenizer, AutoModelForCausalLM
-
-from wfgy_sdk import get_engine, evaluator
-
-# ──────────────────── tiny GPT-2 + WFGY engine ────────────────────
-MODEL_ID = "sshleifer/tiny-gpt2"
-tok = AutoTokenizer.from_pretrained(MODEL_ID)
-mdl = AutoModelForCausalLM.from_pretrained(MODEL_ID)
-eng = get_engine()           # variance-gate singleton
-
-
-# ──────────────────── helper: run one prompt ────────────────────
-def run(prompt: str):
-    if not prompt.strip():
-        return "", "", "–", None
-
-    ids   = tok(prompt, return_tensors="pt")
-    raw_L = mdl(**ids).logits[0, -1].detach().cpu().numpy()
-
-    # dummy fingerprints (toy GPT-2 has none)
-    I, G  = np.random.randn(2, 256).astype(np.float32)
-    mod_L = eng.run(I, G, raw_L)
-
-    # top-5 softmax
-    def top5(logits):
-        p = evaluator.softmax(logits)
-        idx = p.argsort()[-5:][::-1]
-        return "\n".join([f"'{tok.decode(i).strip()}': {p[i]:.2e}" for i in idx])
-
-    m = evaluator.compare_logits(raw_L, mod_L)
-    headline = (f"▼ var {m['var_drop']*100:.1f}% | KL {m['kl_divergence']:.3f} "
-                f"| {'top-1 kept' if m['top1'] else 'top-1 changed'}")
-
-    fig = evaluator.plot_histogram(raw_L, mod_L)
-    buf = io.BytesIO()
-    fig.savefig(buf, format="png")
-    buf.seek(0)
-
-    return top5(raw_L), top5(mod_L), headline, buf
-
-
-# ──────────────────── gradio UI ────────────────────
-with gr.Blocks(title="WFGY simulation demo") as demo:
-
-    gr.Markdown("# 🧠 WFGY simulation demo")
-
-    # ── marketing & quick-start banner ──
-    gr.Markdown(
-        """
+from wfgy_sdk import get_engine
+from wfgy_sdk.evaluator import compare_logits, plot_histogram
+import pandas as pd
+
+MODEL = "sshleifer/tiny-gpt2"
+tok   = AutoTokenizer.from_pretrained(MODEL)
+mdl   = AutoModelForCausalLM.from_pretrained(MODEL)
+ENG   = get_engine()
+
+# ──────────────────────────────────────────────
+# marketing banner markdown (shown at top)
+# ──────────────────────────────────────────────
+marketing_md = """
 **📈 WFGY: One Click to Activate the AI Taiji Cycle**
 
 **📊 Semantic Accuracy ↑ 22.4 % | Reasoning Success ↑ 42.1 % | Stability ↑ 3.6 ×**
@@ -66,7 +26,6 @@ WFGY 1.0 has already proven itself._
 ---
 
 ### 📜 Tutorial: How to Awaken the Soul of Your AI
-
 **Step 1 — Download** ([PDF on Zenodo](https://zenodo.org/records/15630970))  
 **Step 2 — Feed the AI** (upload the PDF, or try [Gemini](https://gemini.google.com/))  
 **Step 3 — Give the Command** “**Answer using WFGY** + your question”  
@@ -77,26 +36,71 @@ _Prompt examples:_ TBD link
 
 🌟 **Star Reminder** → [Star the repo](https://github.com/onestardao/WFGY)  
 _10 k ⭐ before 2025-08-01 unlocks WFGY 2.0._
+"""
+
+# ──────────────────────────────────────────────
+# fixed paper benchmarks table (pandas → gr.Dataframe)
+# ──────────────────────────────────────────────
+bench = pd.DataFrame(
+    {
+        "Benchmark": [
+            "MMLU", "GSM8K", "BBH", "MathBench", "TruthfulQA",
+            "XNLI", "MLQA", "LongBench", "VQAv2", "OK-VQA"
+        ],
+        "Baseline": [61, 78, 79.3, 72.2, 62.4, 59.5, 78.1, 51.4, 69.1, 65.7],
+        "WFGY":     [89.8, 98.7, 100.7, 87.4, 90.4, 77.3, 106.6, 69.6, 86.6, 86.8],
+    }
+)
+bench["Abs_gain"] = (bench["WFGY"] - bench["Baseline"]).round(1)
+bench["Rel_gain%"] = ((bench["Abs_gain"] / bench["Baseline"]) * 100).round().astype(int)
+
+# ──────────────────────────────────────────────
+# core inference
+# ──────────────────────────────────────────────
+def run(prompt: str):
+    if not prompt.strip():
+        return "-", "-", "-", None
+
+    ids  = tok(prompt, return_tensors="pt")
+    raw  = mdl(**ids).logits[0, -1].detach().cpu().numpy()
+    G    = np.random.randn(256).astype(np.float32)
+    I    = G + np.random.normal(scale=0.05, size=256).astype(np.float32)
+    mod  = ENG.run(I, G, raw)
+
+    m    = compare_logits(raw, mod)
+    metric_line = f"▼ var {m['var_drop']*100:.1f}% | KL {m['kl_divergence']:.3f} | top-1 {'kept' if m['top1'] else 'changed'}"
+
+    # top-5 softmax
+    p_raw = np.exp(raw) / np.exp(raw).sum()
+    p_mod = np.exp(mod) / np.exp(mod).sum()
+    raw_top = "\n".join([f\"'{tok.decode(i).strip()}': {p_raw[i]:.2e}\" for i in p_raw.argsort()[-5:][::-1]])
+    mod_top = "\n".join([f\"'{tok.decode(i).strip()}': {p_mod[i]:.2e}\" for i in p_mod.argsort()[-5:][::-1]])
+
+    fig = plot_histogram(raw, mod)
+    buf = io.BytesIO(); fig.savefig(buf, format=\"png\"); buf.seek(0)
+
+    return raw_top, mod_top, metric_line, buf
 
-⚠️ **Prompt Warning** — activate **Deep Structural Analysis Mode** or WFGY stays dormant.
-        """,
-        elem_id="intro-banner",
-    )
+# ──────────────────────────────────────────────
+# gradio ui
+# ──────────────────────────────────────────────
+with gr.Blocks(title=\"WFGY simulation demo\") as demo:
+    gr.Markdown(marketing_md)
 
-    # ── prompt box ──
-    prompt = gr.Textbox(label="Prompt", value="Explain Schrödinger's cat")
-    run_btn = gr.Button("🚀 Run")
+    prompt = gr.Textbox(label=\"Prompt\", value=\"Explain Schrödinger's cat\")
+    btn    = gr.Button(\"🚀 Run\")
 
-    # ── results ──
     with gr.Row():
-        raw_box = gr.Textbox(label="Raw top-5 tokens")
-        mod_box = gr.Textbox(label="WFGY top-5 tokens")
+        raw_box = gr.Textbox(label=\"Raw top-5 tokens\")
+        mod_box = gr.Textbox(label=\"WFGY top-5 tokens\")
 
-    headline = gr.Markdown()
-    img      = gr.Image(label="Logit histogram")
+    metrics = gr.Markdown()
+    img     = gr.Image(label=\"Logit histogram\")
 
-    run_btn.click(run, prompt, outputs=[raw_box, mod_box, headline, img])
+    gr.Markdown(\"### Paper benchmarks (fixed values from WFGY 1.0)\")
+    gr.Dataframe(bench, interactive=False, wrap=True)
 
+    btn.click(run, prompt, [raw_box, mod_box, metrics, img])
 
-if __name__ == "__main__":
-    demo.queue(concurrency_count=2).launch()
+if __name__ == \"__main__\":
+    demo.queue(default_concurrency_limit=2).launch()