microsoft/qlib · error · ValueError

unknown loss `%s`

Error message

unknown loss `%s`

What it means

LocalTransformerModel.loss_fn implements only 'mse' (masked, computed on float tensors). Any other loss string raises ValueError at the first loss computation inside fit()'s training loop.

Source

Thrown at qlib/contrib/model/pytorch_localformer.py:95

        self.fitted = False
        self.model.to(self.device)

    @property
    def use_gpu(self):
        return self.device != torch.device("cpu")

    def mse(self, pred, label):
        loss = (pred.float() - label.float()) ** 2
        return torch.mean(loss)

    def loss_fn(self, pred, label):
        mask = ~torch.isnan(label)

        if self.loss == "mse":
            return self.mse(pred[mask], label[mask])

        raise ValueError("unknown loss `%s`" % self.loss)

    def metric_fn(self, pred, label):
        mask = torch.isfinite(label)

        if self.metric in ("", "loss"):
            return -self.loss_fn(pred[mask], label[mask])

        raise ValueError("unknown metric `%s`" % self.metric)

    def train_epoch(self, x_train, y_train):
        x_train_values = x_train.values
        y_train_values = np.squeeze(y_train.values)

        self.model.train()

        indices = np.arange(len(x_train_values))
        np.random.shuffle(indices)

View on GitHub (pinned to 79633dd950)

Solutions

  1. Set loss='mse'
  2. Override loss_fn in a subclass for other losses

Example fix

# before
LocalTransformerModel(loss="mae")

# after
LocalTransformerModel(loss="mse")
Defensive patterns

Strategy: validation

Validate before calling

assert loss == "mse", "LocalTransformerModel supports only loss='mse'"

Type guard

def is_supported_loss(name: str) -> bool:
    return name == "mse"

Prevention

When it happens

Trigger: LocalTransformerModel(loss != 'mse') followed by fit(); the error appears on the first batch of the first epoch.

Common situations: Shared YAML hyperparameter blocks across models; typo'd loss names.

Related errors

AI-assisted analysis of microsoft/qlib@79633dd950 (2026-08-15). Data as JSON: /api/errors/e41771615b97c58b. Report an issue: GitHub.