nested instantiation

[felipemello1]

Context

What is the purpose of this PR? Is it to

• add a new feature
• fix a bug
• update tests and/or documentation
• other (please add here)

Enable nested instantiation, e.g.

from omegaconf import OmegaConf
from torchtune.config import instantiate


class Spice:
    def __init__(self, heat_level):
        self.heat_level = heat_level


class Food:
    def __init__(self, seed, ingredient):
        self.seed = seed
        self.ingredient = ingredient


config = OmegaConf.create(
    {
        "_component_": "Food",
        "seed": 0,
        "ingredient": {"_component_": "Spice", "heat_level": 5},
    }
)
food = instantiate(config, seed=42)
print(food.seed)  # 42
print(food.ingredient.heat_level)  # 5
new_spice = {"_component_": "Spice", "heat_level": 10}
food = instantiate(config, ingredient=new_spice)
print(food.ingredient.heat_level)  # 10

Test plan

Unit test is blocked by #2518

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/torchtune/2519

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❌ 2 New Failures, 4 Cancelled Jobs

As of commit ddad6f3 with merge base c370348 ():

NEW FAILURES - The following jobs have failed:

• GPU tests / gpu_test (3.11, stable) (gh)
  tests/torchtune/config/test_instantiate.py::TestInstantiate::test_nested_instantiation
• Unit Test / unit_tests (3.10) (gh)
  tests/torchtune/config/test_instantiate.py::TestInstantiate::test_nested_instantiation

CANCELLED JOBS - The following jobs were cancelled. Please retry:

• GPU tests / gpu_test (3.10, stable) (gh)
  ##[error]The operation was canceled.
• GPU tests / gpu_test (3.9, stable) (gh)
  ##[error]The operation was canceled.
• Unit Test / unit_tests (3.11) (gh)
• Unit Test / unit_tests (3.9) (gh)
  ##[error]The operation was canceled.

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[ebsmothers]

I understand the motivation here, but at this point why not just use Hydra?

[felipemello1]

I understand the motivation here, but at this point why not just use Hydra?

bigger PR, but i think we should

[RdoubleA]

bigger PR, but i think we should

I'm bout to flip