The benchmark 000300 does not exist. Please provide the right benchmark

🐛 Bug Description

This error occurs when training custom data using the code in the examples that modifies workflow_by_customer_code.py; even if you set benchmark to None, the problem is still there.

To Reproduce

Steps to reproduce the behavior:

1. prepare your data
2. run workflow_by_customer_code.py, be provided later
3. python3 examples/workflow_by_customer_code.py

Expected Behavior

This error will be gone.

Screenshot

Environment

Note: User could run cd scripts && python collect_info.py all under project directory to get system information
and paste them here directly.

• Qlib version: 0.9.8.dev31
• Python version: 3.12.13
• OS (MacOS):

Additional Notes

workflow_by_customer_code.py script

import qlib
import pandas as pd
from qlib.constant import REG_CN
from qlib.utils import init_instance_by_config, flatten_dict
from qlib.workflow import R
from qlib.workflow.record_temp import SignalRecord, PortAnaRecord, SigAnaRecord
from qlib.tests.config import CSI300_BENCH

TASK = {
    "model": {
        "class": "LGBModel",
        "module_path": "qlib.contrib.model.gbdt",
        "kwargs": {
            "loss": "mse",
            "colsample_bytree": 0.8879,
            "learning_rate": 0.0421,
            "subsample": 0.8789,
            "lambda_l1": 205.6999,
            "lambda_l2": 580.9768,
            "max_depth": 8,
            "num_leaves": 210,
            "num_threads": 10,
        },
    },
    "dataset": {
        "class": "DatasetH",
        "module_path": "qlib.data.dataset",
        "kwargs": {
            "handler": {
                "class": "Alpha158",
                "module_path": "qlib.contrib.data.handler",
                "kwargs": {
                    "start_time": "2020-01-01",
                    "end_time": "2026-05-07",
                    "fit_start_time": "2020-01-01",
                    "fit_end_time": "2024-12-31",
                    "instruments": "all",
                },
            },
            "segments": {
                "train": ("2020-01-01", "2024-12-31"),
                "valid": ("2025-01-01", "2025-06-30"),
                "test": ("2025-07-01", "2026-05-07"),
            },
        },
    },
}

if __name__ == "__main__":
    # provider_uri = "~/.qlib/qlib_data/cn_data"
    provider_uri = "~/qlib/examples/data/qlib_data"
    qlib.init(provider_uri=provider_uri, region=REG_CN)

    model = init_instance_by_config(TASK["model"])
    dataset = init_instance_by_config(TASK["dataset"])

    port_analysis_config = {
        "executor": {
            "class": "SimulatorExecutor",
            "module_path": "qlib.backtest.executor",
            "kwargs": {
                "time_per_step": "day",
                "generate_portfolio_metrics": True,
            },
        },
        "strategy": {
            "class": "TopkDropoutStrategy",
            "module_path": "qlib.contrib.strategy.signal_strategy",
            "kwargs": {
                "signal": (model, dataset),
                "topk": 5,
                "n_drop": 2,
            },
        },
        "backtest": {
            "start_time": "2025-07-01",
            "end_time": "2026-05-07",
            "account": 10000000,
            "benchmark": CSI300_BENCH,
            "exchange_kwargs": {
                "freq": "day",
                "limit_threshold": 0.095,
                "deal_price": "close",
                "open_cost": 0.0005,
                "close_cost": 0.0015,
                "min_cost": 5,
            },
        },
    }

    with R.start(experiment_name="Strategy_Research"):
        R.log_params(**flatten_dict(TASK))
        
        model.fit(dataset)
        R.save_objects(**{"params.pkl": model})

        recorder = R.get_recorder()
        sr = SignalRecord(model, dataset, recorder)
        sr.generate()

        sar = SigAnaRecord(recorder)
        sar.generate()

        par = PortAnaRecord(recorder, port_analysis_config, "day")
        par.generate()

Start with examples/workflow_by_customer_code.py and trace how the CSI300_BENCH value reaches the backtest configuration. Reproduce the failure with the provided script and custom-data setup, then verify the workflow no longer requests the missing benchmark and that the backtest completes successfully.