debugging

interpretable.py

@@ -33,6 +33,8 @@ class InterpretablePolicyExtractor:
             dataset["train_label"] = dataset["train_label"][:, None]
         if dataset["train_label"].ndim == 1 and not self._action_is_discrete:
             dataset["test_label"] = dataset["test_label"][:, None]
+        dataset["train_input"] = dataset["train_input"].float()
+        dataset["test_input"] = dataset["test_input"].float()
         for k,v in dataset.items():
             print(k, v.shape, v.dtype)
         return self.policy.train(dataset, opt="LBFGS", steps=steps, loss_fn=self.loss_fn)