Implement lazy loading for traceable models

[kylesayrs]

Purpose

• Some models may contain imports to libraries that are not part of the base installation. We do not want users who want to use one traceable model to be forced to install libraries used by another traceable model
• Model definitions are large, so it's better to load them only when needed
• Adding new model definitions which rely on newer transformers versions affects other model definitions. For example, attempting to use transformers<4.47.0 raises an error because idefics had not been implemented at that point. While we encourage users to update to the latest transformers versions, this PR allows users to continue using the same environment without unnecessarily forcing an upgrade.

For an example of this lazy module pattern, see transformers/models/llama/__init__.py.

from typing import TYPE_CHECKING
from ...utils import _LazyModule
from ...utils.import_utils import define_import_structure

if TYPE_CHECKING:
    from .configuration_llama import *
    from .modeling_flax_llama import *
    from .modeling_llama import *
    from .tokenization_llama import *
    from .tokenization_llama_fast import *
else:
    import sys

    _file = globals()["__file__"]
    sys.modules[__name__] = _LazyModule(__name__, _file, define_import_structure(_file), module_spec=__spec__)

Changes

• Implemented _AliasableLazyModule which extends _LazyModule to allow aliases
• Dynamically replace llmcompressor.transformers.tracing with an instance of _AliasableLazyModule which lazily loads submodules as they are needed
  • Similar to lazy loading implementation by transformers

Testing

• Added passing tests in tests/llmcompressor/transformers/tracing/test_init.py

[github-actions]

👋 Hi! Thank you for contributing to llm-compressor. Please add the ready label when the PR is ready for review.

Note: This is required to complete the testing suite, please only add the label once the PR is code complete and local testing has been performed.