Add Torch probe entries and Pi artifact sync

crypto_spot_bot/config.py

@@ -118,6 +118,10 @@ class Settings:
     time_series_max_adjustment: float
     time_series_lstm_enabled: bool
     time_series_lstm_model_path: Path
+    time_series_probe_enabled: bool
+    time_series_probe_min_edge_percent: float
+    time_series_probe_min_probability_up: float
+    time_series_probe_size_multiplier: float
     stop_loss_percent: float
     take_profit_percent: float
     trailing_stop_percent: float
@@ -266,6 +270,10 @@ def load_settings(env_file: str | Path | None = None) -> Settings:
         time_series_max_adjustment=_float_env("TIME_SERIES_MAX_ADJUSTMENT", 0.08),
         time_series_lstm_enabled=_bool_env("TIME_SERIES_LSTM_ENABLED", True),
         time_series_lstm_model_path=Path(os.getenv("TIME_SERIES_LSTM_MODEL_PATH", "runtime/lstm_forecaster.json")),
+        time_series_probe_enabled=_bool_env("TIME_SERIES_PROBE_ENABLED", True),
+        time_series_probe_min_edge_percent=_float_env("TIME_SERIES_PROBE_MIN_EDGE_PERCENT", 0.02),
+        time_series_probe_min_probability_up=_float_env("TIME_SERIES_PROBE_MIN_PROBABILITY_UP", 0.55),
+        time_series_probe_size_multiplier=_float_env("TIME_SERIES_PROBE_SIZE_MULTIPLIER", 0.40),
         stop_loss_percent=_float_env("STOP_LOSS_PERCENT", 0.04),
         take_profit_percent=_float_env("TAKE_PROFIT_PERCENT", 0.035),
         trailing_stop_percent=_float_env("TRAILING_STOP_PERCENT", 0.015),

crypto_spot_bot/dashboard.py

@@ -261,6 +261,10 @@ def _safe_config(settings: Settings) -> dict[str, Any]:
         "time_series_max_adjustment": settings.time_series_max_adjustment,
         "time_series_lstm_enabled": settings.time_series_lstm_enabled,
         "time_series_lstm_model_path": str(settings.time_series_lstm_model_path),
+        "time_series_probe_enabled": settings.time_series_probe_enabled,
+        "time_series_probe_min_edge_percent": settings.time_series_probe_min_edge_percent,
+        "time_series_probe_min_probability_up": settings.time_series_probe_min_probability_up,
+        "time_series_probe_size_multiplier": settings.time_series_probe_size_multiplier,
         "time_series_model_artifact": _time_series_model_artifact(settings),
         "stop_loss_percent": settings.stop_loss_percent,
         "take_profit_percent": settings.take_profit_percent,
@@ -678,7 +682,8 @@ HTML = r"""
             ['Horizon', String(artifact.target_horizon ?? '')],
             ['Min edge', `${num(config?.time_series_min_edge_percent, 3)}%`],
             ['Min P(up)', `${num((config?.time_series_min_probability_up || 0) * 100, 1)}%`],
-            ['Min confidence', num(config?.time_series_min_confidence, 3)]
+            ['Min confidence', num(config?.time_series_min_confidence, 3)],
+            ['Probe entry', config?.time_series_probe_enabled ? `${num(config?.time_series_probe_min_edge_percent, 3)}% / P ${num((config?.time_series_probe_min_probability_up || 0) * 100, 1)}% / size ${num((config?.time_series_probe_size_multiplier || 0) * 100, 0)}%` : 'off']
           ]))}
         </div>
         ${positionsPanel()}
@@ -699,6 +704,9 @@ HTML = r"""
       const quality = market.quality || {};
       const signal = latestSignals[symbol] || {};
       const minEdge = state.data.config?.time_series_min_edge_percent ?? 0;
+      const probeEnabled = Boolean(state.data.config?.time_series_probe_enabled);
+      const probeEdge = state.data.config?.time_series_probe_min_edge_percent ?? 0;
+      const probeProbability = state.data.config?.time_series_probe_min_probability_up ?? 0;
       const minConfidence = state.data.config?.time_series_min_confidence ?? 0;
       const diagnostics = parseDiagnostics(signal);
       const failed = Object.entries(diagnostics.checks || {}).filter(([, ok]) => !ok).map(([key]) => key);
@@ -716,6 +724,7 @@ HTML = r"""
             ['Model', modelName(forecast.model)],
             ['Edge', `${signed(forecast.expected_return_percent, 4)}% / min ${num(minEdge, 3)}%`],
             ['P(up)', num((forecast.probability_up || 0) * 100, 2) + '%'],
+            ['Probe', probeEnabled ? `${num(probeEdge, 3)}% / P ${num(probeProbability * 100, 1)}% / ${diagnostics.edge_mode || 'n/a'}` : 'off'],
             ['Confidence', `${num(signal.confidence, 4)} / min ${num(minConfidence, 2)}`],
             ['Q10/Q50/Q90', `${signed(forecast.quantile_10_percent, 2)} / ${signed(forecast.quantile_50_percent, 2)} / ${signed(forecast.quantile_90_percent, 2)}`],
             ['Blocked', forecast.block_entry ? 'yes' : 'no']

crypto_spot_bot/strategy.py

@@ -633,6 +633,34 @@ def _torch_forecast_entry_signal(
     skill = _safe_float(forecast.get("skill"), 0.0)
     min_edge = max(0.0, settings.time_series_min_edge_percent)
     min_probability = _torch_min_probability(settings)
+    probe_min_edge = max(0.0, min(settings.time_series_probe_min_edge_percent, min_edge))
+    probe_min_probability = round(
+        _clamp(settings.time_series_probe_min_probability_up, min_probability, 0.85),
+        4,
+    )
+    full_edge_ok = expected_return >= min_edge
+    probe_edge_ok = bool(
+        settings.time_series_probe_enabled
+        and not full_edge_ok
+        and expected_return >= probe_min_edge
+        and probability_up >= probe_min_probability
+    )
+    edge_mode = "full" if full_edge_ok else ("probe" if probe_edge_ok else "blocked")
+    if probe_edge_ok and position_notional > 0:
+        probe_multiplier = _clamp(settings.time_series_probe_size_multiplier, 0.05, 1.0)
+        position_notional = round(
+            min(
+                settings.max_position_usdt,
+                max(settings.min_position_usdt, position_notional * probe_multiplier),
+            ),
+            2,
+        )
+        sizing = {
+            **sizing,
+            "notional_usdt": position_notional,
+            "probe_size_multiplier": round(probe_multiplier, 4),
+            "edge_mode": "probe",
+        }
     confidence = _torch_forecast_confidence(settings, forecast)
     spread_ok = ticker.spread_percent <= settings.max_spread_percent
     liquidity_ok = ticker.turnover_24h >= settings.min_24h_turnover_usdt
@@ -641,7 +669,7 @@ def _torch_forecast_entry_signal(
         "torch_model_ok": model_ok,
         "forecast_usable": bool(forecast.get("usable", False)),
         "forecast_not_blocked": not bool(forecast.get("block_entry", False)),
-        "expected_edge_ok": expected_return >= min_edge,
+        "expected_edge_ok": full_edge_ok or probe_edge_ok,
         "probability_ok": probability_up >= min_probability,
         "skill_ok": skill > 0.0,
         "confidence_ok": confidence >= settings.time_series_min_confidence,
@@ -659,6 +687,10 @@ def _torch_forecast_entry_signal(
         "atr_trailing_multiplier": _clamp(settings.atr_trailing_multiplier, 0.5, 10.0),
         "expected_return_percent": expected_return,
         "min_edge_percent": min_edge,
+        "probe_enabled": settings.time_series_probe_enabled,
+        "probe_min_edge_percent": probe_min_edge,
+        "probe_min_probability_up": probe_min_probability,
+        "edge_mode": edge_mode,
         "probability_up": probability_up,
         "min_probability_up": min_probability,
         "min_confidence": settings.time_series_min_confidence,
@@ -679,7 +711,8 @@ def _torch_forecast_entry_signal(
             (
                 "torch_forecast: PyTorch edge confirmed; "
                 f"model={forecast.get('model')}, p_up={probability_up:.3f}, "
-                f"expected={expected_return:.4f}%, size={position_notional:.2f} USDT"
+                f"expected={expected_return:.4f}%, edge_mode={edge_mode}, "
+                f"size={position_notional:.2f} USDT"
             ),
             diagnostics,
         )