Update (base update)

[ghstack-poisoned]

tensordict/nn/params.py

@@ -111,10 +111,8 @@ def _maybe_make_param(tensor):
 
 
 def _maybe_make_param_or_buffer(tensor):
-    if (
-        isinstance(tensor, (Tensor, ftdim.Tensor))
-        and not isinstance(tensor, (nn.Parameter, Buffer))
-        and tensor.dtype in (torch.float, torch.double, torch.half)
+    if isinstance(tensor, (Tensor, ftdim.Tensor)) and not isinstance(
+        tensor, (nn.Parameter, Buffer)
     ):
         if not tensor.requires_grad and not is_batchedtensor(tensor):
             # convert all non-parameters to buffers

tensordict/utils.py

@@ -68,12 +68,9 @@
 except ImportError:
     _has_funcdim = False
 try:
-    from torch.compiler import assume_constant_result, is_dynamo_compiling
+    from torch.compiler import assume_constant_result
 except ImportError:  # torch 2.0
-    from torch._dynamo import (
-        assume_constant_result,
-        is_compiling as is_dynamo_compiling,
-    )
+    from torch._dynamo import assume_constant_result
 
 if TYPE_CHECKING:
     from tensordict.tensordict import TensorDictBase
@@ -2825,7 +2822,8 @@ def _is_dataclass(obj):
         if isinstance(obj, type) and not isinstance(obj, GenericAlias)
         else type(obj)
     )
-    return hasattr(cls, _FIELDS)
+    # return hasattr(cls, _FIELDS)
+    return getattr(cls, _FIELDS, None) is not None
 
 
 def _is_list_tensor_compatible(t) -> Tuple[bool, tuple | None, type | None]: