fix: use compositonal implementation for torch backend ivy.put_along_…

…axis

ivy/functional/backends/torch/experimental/manipulation.py

@@ -443,46 +443,6 @@ def column_stack(
     return torch.column_stack(arrays)
 
 
-@with_supported_dtypes({"2.2 and below": ("float32", "float64")}, backend_version)
-def put_along_axis(
-    arr: torch.Tensor,
-    indices: torch.Tensor,
-    values: Union[int, torch.Tensor],
-    axis: int,
-    /,
-    *,
-    mode: Literal["sum", "min", "max", "mul", "replace"] = "replace",
-    out: Optional[torch.Tensor] = None,
-) -> torch.Tensor:
-    arr, values = ivy.promote_types_of_inputs(arr, values)
-
-    mode_mappings = {
-        "sum": "sum",
-        "min": "amin",
-        "max": "amax",
-        "mul": "prod",
-        "replace": "replace",
-    }
-    mode = mode_mappings.get(mode, mode)
-    indices = indices.to(torch.int64)
-    if isinstance(values, torch.Tensor) and values.dim() == 0:
-        values = values.item()
-    if mode == "replace":
-        return torch.scatter(arr, axis, indices, values, out=out)
-    else:
-        return torch.scatter_reduce(arr, axis, indices, values, reduce=mode, out=out)
-
-
-put_along_axis.partial_mixed_handler = lambda *args, mode=None, **kwargs: mode in [
-    "replace",
-    "sum",
-    "mul",
-    "mean",
-    "max",
-    "min",
-]
-
-
 def concat_from_sequence(
     input_sequence: Union[Tuple[torch.Tensor], List[torch.Tensor]],
     /,