[BugFix] Remove raisers in specs

ghstack-source-id: a005a62
Pull Request resolved: #2651

Author: Vincent Moens

torchrl/data/tensor_specs.py

@@ -1402,7 +1402,6 @@ def type_check(self, value: torch.Tensor, key: NestedKey | None = None) -> None:
             spec.type_check(val)
 
     def is_in(self, value) -> bool:
-        raise RuntimeError
         if self.dim == 0 and not hasattr(value, "unbind"):
             # We don't use unbind because value could be a tuple or a nested tensor
             return all(
@@ -1834,7 +1833,6 @@ def _project(self, val: torch.Tensor) -> torch.Tensor:
         return val
 
     def is_in(self, val: torch.Tensor) -> bool:
-        raise RuntimeError
         if self.mask is None:
             shape = torch.broadcast_shapes(self._safe_shape, val.shape)
             shape_match = val.shape == shape
@@ -2288,7 +2286,6 @@ def _project(self, val: torch.Tensor) -> torch.Tensor:
         return val
 
     def is_in(self, val: torch.Tensor) -> bool:
-        raise RuntimeError
         val_shape = _remove_neg_shapes(tensordict.utils._shape(val))
         shape = torch.broadcast_shapes(self._safe_shape, val_shape)
         shape = list(shape)
@@ -2489,7 +2486,6 @@ def one(self, shape=None):
         )
 
     def is_in(self, val: Any) -> bool:
-        raise RuntimeError
         shape = torch.broadcast_shapes(self._safe_shape, val.shape)
         return (
             is_non_tensor(val)
@@ -2682,7 +2678,6 @@ def rand(self, shape: torch.Size = None) -> torch.Tensor:
         return torch.empty(shape, device=self.device, dtype=self.dtype).random_()
 
     def is_in(self, val: torch.Tensor) -> bool:
-        raise RuntimeError
         shape = torch.broadcast_shapes(self._safe_shape, val.shape)
         return val.shape == shape and val.dtype == self.dtype
 
@@ -3034,7 +3029,6 @@ def index(self, index: INDEX_TYPING, tensor_to_index: torch.Tensor) -> torch.Ten
         return torch.cat(out, -1)
 
     def is_in(self, val: torch.Tensor) -> bool:
-        raise RuntimeError
         vals = self._split(val)
         if vals is None:
             return False
@@ -3435,7 +3429,6 @@ def _project(self, val: torch.Tensor) -> torch.Tensor:
         return val
 
     def is_in(self, val: torch.Tensor) -> bool:
-        raise RuntimeError
         if self.mask is None:
             shape = torch.broadcast_shapes(self._safe_shape, val.shape)
             shape_match = val.shape == shape
@@ -4066,7 +4059,6 @@ def _project(self, val: torch.Tensor) -> torch.Tensor:
         return val.squeeze(0) if val_is_scalar else val
 
     def is_in(self, val: torch.Tensor) -> bool:
-        raise RuntimeError
         if self.mask is not None:
             vals = val.unbind(-1)
             splits = self._split_self()