Upload alpha_factory/config.py with huggingface_hub

alpha_factory/config.py

@@ -0,0 +1,102 @@
+"""
+Configuration — all settings in one place.
+Environment variables override defaults.
+"""
+from pydantic import BaseModel, Field
+from pathlib import Path
+import os
+
+
+class LLMConfig(BaseModel):
+    """LLM serving configuration."""
+    microfish_model: str = "Qwen/Qwen2.5-1.5B-Instruct"
+    tinyfish_model: str = "Qwen/Qwen2.5-3B-Instruct"
+    mediumfish_model: str = "Qwen/Qwen2.5-7B-Instruct"
+    bigfish_model: str = "Qwen/Qwen2.5-72B-Instruct"
+    base_url: str = Field(default="http://localhost:8000/v1", description="vLLM / Ollama endpoint")
+    api_key: str = "dummy"
+    temperature_generation: float = 0.7
+    temperature_compilation: float = 0.1
+    temperature_critique: float = 0.3
+    max_tokens: int = 4096
+
+
+class BrainConfig(BaseModel):
+    """WorldQuant BRAIN API configuration."""
+    api_url: str = "https://api.worldquantbrain.com"
+    region: str = "USA"
+    universe: str = "TOP3000"
+    delay: int = 1
+    truncation: float = 0.08
+    pasteurization: str = "ON"
+    nan_handling: str = "OFF"
+    max_concurrent: int = 4
+    submit_interval_sec: float = 15.0
+
+
+class KillSwitches(BaseModel):
+    """Hard circuit breakers — non-negotiable."""
+    daily_brain_submissions_max: int = 200
+    consecutive_lint_fail_max: int = 10
+    consecutive_kill_verdict_max: int = 30
+    daily_llm_token_budget: int = 5_000_000
+    max_credits_per_family: int = 3
+
+
+class FitnessWeights(BaseModel):
+    """Fitness function coefficients. Calibrate with hand-rankings after 20+ alphas."""
+    sharpe_os: float = 1.0
+    is_os_gap_penalty: float = 0.5
+    worst_year_penalty: float = 1.0
+    crowding_penalty: float = 0.3
+    turnover_penalty: float = 0.2
+    turnover_threshold: float = 0.40
+    drawdown_penalty: float = 0.1
+    drawdown_threshold: float = 0.05
+    novelty_bonus: float = 0.4
+
+
+class Paths(BaseModel):
+    """All filesystem paths."""
+    root: Path = Path(os.getenv("AF_ROOT", "."))
+    data: Path = Field(default=None)
+    factor_store: Path = Field(default=None)
+    prompts: Path = Field(default=None)
+    logs: Path = Field(default=None)
+
+    def model_post_init(self, __context):
+        if self.data is None:
+            self.data = self.root / "data"
+        if self.factor_store is None:
+            self.factor_store = self.root / "factor_store"
+        if self.prompts is None:
+            self.prompts = self.root / "prompts"
+        if self.logs is None:
+            self.logs = self.root / "logs"
+        # Ensure directories exist
+        for p in [self.data, self.factor_store, self.factor_store / "alphas",
+                  self.prompts, self.prompts / "templates", self.logs]:
+            p.mkdir(parents=True, exist_ok=True)
+
+
+class Config(BaseModel):
+    """Master configuration."""
+    llm: LLMConfig = LLMConfig()
+    brain: BrainConfig = BrainConfig()
+    kill: KillSwitches = KillSwitches()
+    fitness: FitnessWeights = FitnessWeights()
+    paths: Paths = Paths()
+
+    # Pipeline settings
+    batch_size: int = 10  # alphas per batch
+    max_iterations_per_family: int = 3
+    correlation_threshold: float = 0.65
+    min_sharpe_local_sim: float = 1.0
+    min_info_value_sign_sweep: float = 0.3
+
+
+def load_config() -> Config:
+    """Load config with env var overrides."""
+    return Config(
+        paths=Paths(root=Path(os.getenv("AF_ROOT", ".")))
+    )