Fix MAML graph breaks during Inductor inference

[msnliu]

MAML inference performs task-specific finetuning and calls torch.autograd.grad(). With PyTorch 2.14's default Dynamo settings, autograd operations are not traced, which causes graph breaks.

This change:

• enables trace_autograd_ops for MAML;
• passes parameters and BatchNorm state explicitly;
• removes deepcopy() and Tensor.item() from the compiled path.

MAML inference now compiles without the observed graph breaks. Validated on Intel XPU with numerical correctness checks and >1 Inductor speedup over eager.

[meta-cla]

Hi @msnliu!

Thank you for your pull request and welcome to our community.

Action Required

In order to merge any pull request (code, docs, etc.), we require contributors to sign our Contributor License Agreement, and we don't seem to have one on file for you.

Process

In order for us to review and merge your suggested changes, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA.

Once the CLA is signed, our tooling will perform checks and validations. Afterwards, the pull request will be tagged with CLA signed. The tagging process may take up to 1 hour after signing. Please give it that time before contacting us about it.

If you have received this in error or have any questions, please contact us at cla@meta.com. Thanks!