[Feature] Add option for auto-resetting envs in GAE

[lin-erica]

Description

Previously, when calculating the GAE, torchrl assumes that there is a valid s_(t+1) return from the environment. However, for auto-resetting environments, that is not true at the end of the episode, as it just jumps to the first state of the next episode. This PR adds a feature to the GAE class, allowing users to specify if it's an auto-resetting env, in which case it will use the s_t as an approximation for bootstrapping.

Motivation and Context

Without this change, we were getting much slower convergence rates during training for IsaacLab (an autoresetting env).

• I have raised an issue to propose this change (required for new features and bug fixes)

Types of changes

What types of changes does your code introduce? Remove all that do not apply:

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds core functionality)
• Breaking change (fix or feature that would cause existing functionality to change)
• Documentation (update in the documentation)
• Example (update in the folder of examples)

Checklist

Go over all the following points, and put an x in all the boxes that apply.
If you are unsure about any of these, don't hesitate to ask. We are here to help!

• I have read the CONTRIBUTION guide (required)
• My change requires a change to the documentation.
• I have updated the tests accordingly (required for a bug fix or a new feature).
• I have updated the documentation accordingly.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/rl/2851

• 📄 Preview Python docs built from this PR

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

❌ 11 New Failures, 1 Cancelled Job, 1 Unrelated Failure

As of commit cae5b6f with merge base 9cd95d5 ():

NEW FAILURES - The following jobs have failed:

• Continuous Benchmark (PR) / CPU Pytest benchmark (gh)
  Process completed with exit code 1.
• Continuous Benchmark (PR) / GPU Pytest benchmark (gh)
  Process completed with exit code 1.
• Continuous Benchmark / CPU Pytest benchmark (gh)
  RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.FloatTensor []] is at version 4; expected version 3 instead. Hint: enable anomaly detection to find the operation that failed to compute its gradient, with torch.autograd.set_detect_anomaly(True).
• Continuous Benchmark / GPU Pytest benchmark (gh)
  RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor []] is at version 4; expected version 3 instead. Hint: enable anomaly detection to find the operation that failed to compute its gradient, with torch.autograd.set_detect_anomaly(True).
• Generate documentation / build-docs (3.9, 12.8) / linux-job (gh)
  RuntimeError: Command docker exec -t 5c9535ae8618e6de5a8bc4f4519f0396c5662f527ebb1e13343a25f0690be085 /exec failed with exit code 2
• Habitat Tests on Linux / tests (3.9, 12.8) / linux-job (gh)
  RuntimeError: Command docker exec -t 9eb631c9230fff2cd26a69f6dd67dfea217841d96dc0f399ae675897dd86f69a /exec failed with exit code 1
• SOTA Tests on Linux / tests (3.9, 12.8) / linux-job (gh)
  RuntimeError: Command docker exec -t 6137f65af76ea8104ed7acc7bcff559964f883a6cfd68a1e23a4385ea4e41d7b /exec failed with exit code 1
• Unit-tests on Linux / tests-cpu (3.10) / linux-job (gh)
  test/test_env.py::TestNonTensorEnv::test_parallel[False-False]
• Unit-tests on Linux / tests-cpu (3.11) / linux-job (gh)
  test/test_transforms.py::TestVecNormV2::test_vecnorm_parallel_auto[2]
• Unit-tests on Linux / tests-olddeps (3.8, 11.6) / linux-job (gh)
  RuntimeError: Command docker exec -t df1e1c202ec05da3a4006281335dcbe42470dddc5f39acd09e02e4f49def47c2 /exec failed with exit code 1
• Unit-tests on Linux / tests-optdeps (3.11, 12.8) / linux-job (gh)
  test/test_cost.py::TestPPO4LLMs::test_hf[False]

CANCELLED JOB - The following job was cancelled. Please retry:

• LLM Tests on Linux / unittests (3.9, 12.8) / linux-job (gh)
  ##[error]The operation was canceled.

BROKEN TRUNK - The following job failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

• Unit-tests on Windows / unittests-cpu (3.10, windows.4xlarge, cpu) / windows-job (gh) (trunk failure)
  test/test_transforms.py::TestTimer::test_transform_env

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

[lin-erica]

@vmoens Perhaps I missed it, but I didn't see any tests that were testing the GAE class. If there is, I'm happy to add a relevant test as well.

[vmoens]

LGTM! Sorry for the delay in reviewing this!

To answer your question: Not sure what unit test we want for this feature, so I'm willing to consider merging it as is.

We do test the GAE in test_costs.py I believe... I you have an idea of what to test we can add it