refactor(configs): untangle config_path/resume resolution in validate()

Split the re-parse HACK block in TrainPipelineConfig.validate() into focused
helpers (_resolve_pretrained_from_cli, _resolve_resume_checkpoint) that handle
the policy path, reward-model path, and resume config_path as separate,
readable units. Behavior-preserving.

src/lerobot/configs/train.py

@@ -144,10 +144,17 @@ class TrainPipelineConfig(HubMixin):
             return self.reward_model  # type: ignore[return-value]
         return self.policy  # type: ignore[return-value]
 
-    def validate(self) -> None:
+    def _resolve_pretrained_from_cli(self) -> None:
-        # HACK: We parse again the cli args here to get the pretrained paths if there was some.
+        """Resolve the pretrained source passed on the CLI into a loaded config.
-        policy_path = parser.get_path_arg("policy")
+
+        The pretrained paths (`--policy.path`, `--reward_model.path`) and
+        `--config_path` are only recoverable by re-reading the CLI args: draccus
+        has already consumed them by the time `validate()` runs, so they are not
+        reflected on `self`. Exactly one source applies, in priority order:
+        reward-model path, policy path, then resume.
+        """
         reward_model_path = parser.get_path_arg("reward_model")
+        policy_path = parser.get_path_arg("policy")
 
         if reward_model_path:
             cli_overrides = parser.get_cli_overrides("reward_model")
@@ -156,31 +163,35 @@ class TrainPipelineConfig(HubMixin):
             )
             self.reward_model.pretrained_path = str(Path(reward_model_path))
         elif policy_path:
-            yaml_overrides = parser.get_yaml_overrides("policy")
+            overrides = parser.get_yaml_overrides("policy") + (parser.get_cli_overrides("policy") or [])
-            cli_overrides = parser.get_cli_overrides("policy") or []
+            self.policy = PreTrainedConfig.from_pretrained(policy_path, cli_overrides=overrides)
-            self.policy = PreTrainedConfig.from_pretrained(
-                policy_path, cli_overrides=yaml_overrides + cli_overrides
-            )
             self.policy.pretrained_path = Path(policy_path)
         elif self.resume:
-            config_path = parser.parse_arg("config_path")
+            self._resolve_resume_checkpoint()
-            if not config_path:
-                raise ValueError(
-                    f"A config_path is expected when resuming a run. Please specify path to {TRAIN_CONFIG_NAME}"
-                )
 
-            if not Path(config_path).resolve().exists():
+    def _resolve_resume_checkpoint(self) -> None:
-                raise NotADirectoryError(
+        """Point the trainable config at the checkpoint named by `--config_path`."""
-                    f"{config_path=} is expected to be a local path. "
+        config_path = parser.parse_arg("config_path")
-                    "Resuming from the hub is not supported for now."
+        if not config_path:
-                )
+            raise ValueError(
+                f"A config_path is expected when resuming a run. Please specify path to {TRAIN_CONFIG_NAME}"
+            )
 
-            policy_dir = Path(config_path).parent
+        if not Path(config_path).resolve().exists():
-            if self.policy is not None:
+            raise NotADirectoryError(
-                self.policy.pretrained_path = policy_dir
+                f"{config_path=} is expected to be a local path. "
-            if self.reward_model is not None:
+                "Resuming from the hub is not supported for now."
-                self.reward_model.pretrained_path = str(policy_dir)
+            )
-            self.checkpoint_path = policy_dir.parent
+
+        policy_dir = Path(config_path).parent
+        if self.policy is not None:
+            self.policy.pretrained_path = policy_dir
+        if self.reward_model is not None:
+            self.reward_model.pretrained_path = str(policy_dir)
+        self.checkpoint_path = policy_dir.parent
+
+    def validate(self) -> None:
+        self._resolve_pretrained_from_cli()
 
         if self.policy is None and self.reward_model is None:
             raise ValueError(