1234567891011121314151617181920 |
- # flake8: noqa: F401
- r"""
- This file is in the process of migration to `torch/ao/quantization`, and
- is kept here for compatibility while the migration process is ongoing.
- If you are adding a new entry/functionality, please, add it to the
- appropriate files under `torch/ao/quantization/fx/`, while adding an import statement
- here.
- """
- from torch.ao.quantization.fx.utils import (
- get_custom_module_class_keys,
- get_linear_prepack_op_for_dtype,
- get_qconv_prepack_op,
- get_new_attr_name_with_prefix,
- graph_module_from_producer_nodes,
- assert_and_get_unique_device,
- create_getattr_from_value,
- all_node_args_have_no_tensors,
- get_non_observable_arg_indexes_and_types,
- maybe_get_next_module
- )
|