Files
microsoft--rd-agent/rdagent/components/workflow/rd_loop.py
T
wehub-resource-sync e64161ec32
Release / release_and_publish (push) Waiting to run
CI / ci (3.11) (push) Has been cancelled
CI / ci (3.10) (push) Has been cancelled
CI / dependabot (push) Has been cancelled
chore: import upstream snapshot with attribution
2026-07-13 13:36:15 +08:00

242 lines
10 KiB
Python

"""
Model workflow with session control
It is from `rdagent/app/qlib_rd_loop/model.py` and try to replace `rdagent/app/qlib_rd_loop/RDAgent.py`
"""
import asyncio
import json
from multiprocessing import Queue
from pathlib import Path
from typing import Any
from rdagent.components.workflow.conf import BasePropSetting
from rdagent.core.conf import RD_AGENT_SETTINGS
from rdagent.core.developer import Developer
from rdagent.core.proposal import (
Experiment2Feedback,
ExperimentPlan,
Hypothesis,
Hypothesis2Experiment,
HypothesisFeedback,
HypothesisGen,
Trace,
)
from rdagent.core.scenario import Scenario
from rdagent.core.utils import import_class
from rdagent.log import rdagent_logger as logger
from rdagent.utils.qlib import ALPHA20, validate_qlib_features
from rdagent.utils.workflow import LoopBase, LoopMeta
class RDLoop(LoopBase, metaclass=LoopMeta):
def __init__(self, PROP_SETTING: BasePropSetting):
scen: Scenario = import_class(PROP_SETTING.scen)()
logger.log_object(scen, tag="scenario")
logger.log_object(PROP_SETTING.model_dump(), tag="RDLOOP_SETTINGS")
logger.log_object(RD_AGENT_SETTINGS.model_dump(), tag="RD_AGENT_SETTINGS")
self.hypothesis_gen: HypothesisGen = (
import_class(PROP_SETTING.hypothesis_gen)(scen)
if hasattr(PROP_SETTING, "hypothesis_gen") and PROP_SETTING.hypothesis_gen
else None
)
self.plan: ExperimentPlan = {
"features": ALPHA20,
"feature_codes": {},
} # for user interaction
self.hypothesis2experiment: Hypothesis2Experiment = (
import_class(PROP_SETTING.hypothesis2experiment)()
if hasattr(PROP_SETTING, "hypothesis2experiment") and PROP_SETTING.hypothesis2experiment
else None
)
self.coder: Developer = (
import_class(PROP_SETTING.coder)(scen) if hasattr(PROP_SETTING, "coder") and PROP_SETTING.coder else None
)
self.runner: Developer = (
import_class(PROP_SETTING.runner)(scen) if hasattr(PROP_SETTING, "runner") and PROP_SETTING.runner else None
)
self.summarizer: Experiment2Feedback = (
import_class(PROP_SETTING.summarizer)(scen)
if hasattr(PROP_SETTING, "summarizer") and PROP_SETTING.summarizer
else None
)
self.trace = Trace(scen=scen)
super().__init__()
# excluded steps
def _set_interactor(self, user_request_q: Queue, user_response_q: Queue):
self.user_request_q = user_request_q
self.user_response_q = user_response_q
def _init_base_features(self, base_features_path: str | None):
if base_features_path is not None:
try:
base_dir = Path(base_features_path)
base_factors_file = base_dir / "base_factors.json"
feature_codes: dict[str, str] = {}
for py_file in sorted(base_dir.glob("*.py")):
feature_codes[py_file.name] = py_file.read_text()
self.plan["feature_codes"] = feature_codes
if not base_factors_file.exists():
logger.info(f"No base_factors.json found under {base_dir}. Keeping default base features.")
logger.info(f"{len(feature_codes)} feature code files loaded from {base_dir}.")
else:
with base_factors_file.open("r") as f:
features = json.load(f)
if not isinstance(features, dict):
raise ValueError(
"`base_factors.json` must contain a JSON object of feature_name -> expression."
)
if validate_qlib_features(list(features.values())):
self.plan["features"] = features
logger.info(
f"Loaded base features from {base_factors_file}. {len(features)} features loaded and {len(feature_codes)} feature code files loaded."
)
else:
logger.warning(
f"Base feature validation failed for features loaded from {base_factors_file}. Using default features."
)
except Exception as e:
logger.warning(f"Failed to load base features from {base_features_path}: {e}. Using default features.")
else:
logger.info("No base features path provided. Using default features.")
def _interact_init_params(self) -> None:
if not (hasattr(self, "user_request_q") and hasattr(self, "user_response_q")):
return
logger.info("Waiting for user interaction on initial parameters...")
try:
self.user_request_q.put(
{
"user_instruction": None,
}
)
res_dict = self.user_response_q.get()
logger.info("Received user instruction response.")
self.plan.update(res_dict)
if "feature_codes" not in self.plan:
self.plan[
"user_instruction"
] += f"\n\n{str(list(self.plan['feature_codes'].keys()))} has been configured as the base factor; do not generate duplicate factors."
fea_valid_msg = ""
while True:
logger.info("Requesting base feature configuration from user.")
self.user_request_q.put(
{
"features": self.plan["features"],
"feature_validation_msg": fea_valid_msg,
}
)
self.plan["features"] = self.user_response_q.get()
logger.info("Received base feature configuration response.")
if validate_qlib_features(list(self.plan["features"].values())):
logger.info(f"Base feature validation passed. {len(self.plan['features'])} features selected.")
break
else:
logger.info("Base feature validation failed. Asking user to revise.")
fea_valid_msg = "Some features are invalid, please revise."
except (EOFError, OSError):
logger.info("User interaction failed, using default initial parameters.")
return
logger.info("Received user interaction on initial parameters.")
def _interact_hypo(self, hypo: Hypothesis) -> Hypothesis:
if not (hasattr(self, "user_request_q") and hasattr(self, "user_response_q")):
return hypo
logger.info("Waiting for user interaction on hypothesis...")
try:
self.user_request_q.put(hypo.__dict__)
res_dict = self.user_response_q.get()
modified_hypo = type(hypo)(**res_dict)
except (EOFError, OSError, TypeError):
logger.info("User interaction failed, using original hypothesis.")
return hypo
logger.info("Received user interaction on hypothesis.")
return modified_hypo
def _interact_feedback(self, feedback: HypothesisFeedback) -> HypothesisFeedback:
if not (hasattr(self, "user_request_q") and hasattr(self, "user_response_q")):
return feedback
logger.info("Waiting for user interaction on feedback...")
try:
self.user_request_q.put(feedback.__dict__)
res_dict = self.user_response_q.get()
modified_feedback = HypothesisFeedback(**res_dict)
except (EOFError, OSError, TypeError):
logger.info("User interaction failed, using original feedback.")
return feedback
logger.info("Received user interaction on feedback.")
return modified_feedback
def _propose(self):
hypothesis = self.hypothesis_gen.gen(self.trace, self.plan)
# user can change the hypothesis here
hypothesis = self._interact_hypo(hypothesis)
logger.log_object(hypothesis, tag="hypothesis generation")
return hypothesis
def _exp_gen(self, hypothesis: Hypothesis):
exp = self.hypothesis2experiment.convert(hypothesis, self.trace)
logger.log_object(exp.sub_tasks, tag="experiment generation")
return exp
# included steps
async def direct_exp_gen(self, prev_out: dict[str, Any]):
while True:
if self.get_unfinished_loop_cnt(self.loop_idx) < RD_AGENT_SETTINGS.get_max_parallel():
hypo = self._propose()
exp = self._exp_gen(hypo)
exp.base_features = self.plan["features"]
exp.base_feature_codes = self.plan["feature_codes"]
if exp.based_experiments:
exp.based_experiments[-1].base_features = self.plan["features"]
exp.based_experiments[-1].base_feature_codes = self.plan["feature_codes"]
return {"propose": hypo, "exp_gen": exp}
await asyncio.sleep(1)
def coding(self, prev_out: dict[str, Any]):
exp = self.coder.develop(prev_out["direct_exp_gen"]["exp_gen"])
logger.log_object(exp.sub_workspace_list, tag="coder result")
return exp
def running(self, prev_out: dict[str, Any]):
exp = self.runner.develop(prev_out["coding"])
logger.log_object(exp, tag="runner result")
return exp
def feedback(self, prev_out: dict[str, Any]):
# TODO: the logic branch of exception should be moved to summarizer
e = prev_out.get(self.EXCEPTION_KEY, None)
if e is not None:
feedback = HypothesisFeedback(
reason=str(e),
decision=False,
code_change_summary="",
acceptable=False,
)
else:
feedback = self.summarizer.generate_feedback(prev_out["running"], self.trace)
feedback = self._interact_feedback(feedback)
logger.log_object(feedback, tag="feedback")
return feedback
def record(self, prev_out: dict[str, Any]):
feedback = prev_out["feedback"]
exp = prev_out.get("running") or prev_out.get("coding") or prev_out.get("direct_exp_gen", {}).get("exp_gen")
self.trace.sync_dag_parent_and_hist((exp, feedback), prev_out[self.LOOP_IDX_KEY])