Update base for Update on "Enable aoti for preprocess + torch pin update"

Update torch nightly pin to 11/01 after: pytorch/pytorch#137063

Test Plan:
With pytorch/pytorch#137063:
```
pytest -c /dev/null -v -n auto examples/models/llama3_2_vision/preprocess/
```

[ghstack-poisoned]

Author: lucylq

backends/apple/coreml/README.md

@@ -65,7 +65,7 @@ To quantize a Program in a Core ML favored way, the client may utilize **CoreMLQ
 import torch
 import executorch.exir
 
-from torch._export import capture_pre_autograd_graph
+from torch.export import export_for_training
 from torch.ao.quantization.quantize_pt2e import (
     convert_pt2e,
     prepare_pt2e,
@@ -93,7 +93,7 @@ class Model(torch.nn.Module):
 source_model = Model()
 example_inputs = (torch.randn((1, 3, 256, 256)), )
 
-pre_autograd_aten_dialect = capture_pre_autograd_graph(model, example_inputs)
+pre_autograd_aten_dialect = export_for_training(model, example_inputs).module()
 
 quantization_config = LinearQuantizerConfig.from_dict(
     {

docs/source/llm/getting-started.md

@@ -144,8 +144,7 @@ import torch
 
 from executorch.exir import EdgeCompileConfig, to_edge
 from torch.nn.attention import sdpa_kernel, SDPBackend
-from torch._export import capture_pre_autograd_graph
-from torch.export import export
+from torch.export import export, export_for_training
 
 from model import GPT
 
@@ -170,7 +169,7 @@ dynamic_shape = (
 # Trace the model, converting it to a portable intermediate representation.
 # The torch.no_grad() call tells PyTorch to exclude training-specific logic.
 with torch.nn.attention.sdpa_kernel([SDPBackend.MATH]), torch.no_grad():
-    m = capture_pre_autograd_graph(model, example_inputs, dynamic_shapes=dynamic_shape)
+    m = export_for_training(model, example_inputs, dynamic_shapes=dynamic_shape).module()
     traced_model = export(m, example_inputs, dynamic_shapes=dynamic_shape)
 
 # Convert the model into a runnable ExecuTorch program.
@@ -462,7 +461,7 @@ from executorch.exir import EdgeCompileConfig, to_edge
 import torch
 from torch.export import export
 from torch.nn.attention import sdpa_kernel, SDPBackend
-from torch._export import capture_pre_autograd_graph
+from torch.export import export_for_training
 
 from model import GPT
 
@@ -489,7 +488,7 @@ dynamic_shape = (
 # Trace the model, converting it to a portable intermediate representation.
 # The torch.no_grad() call tells PyTorch to exclude training-specific logic.
 with torch.nn.attention.sdpa_kernel([SDPBackend.MATH]), torch.no_grad():
-    m = capture_pre_autograd_graph(model, example_inputs, dynamic_shapes=dynamic_shape)
+    m = export_for_training(model, example_inputs, dynamic_shapes=dynamic_shape).module()
     traced_model = export(m, example_inputs, dynamic_shapes=dynamic_shape)
 
 # Convert the model into a runnable ExecuTorch program.
@@ -635,7 +634,7 @@ xnnpack_quant_config = get_symmetric_quantization_config(
 xnnpack_quantizer = XNNPACKQuantizer()
 xnnpack_quantizer.set_global(xnnpack_quant_config)
 
-m = capture_pre_autograd_graph(model, example_inputs)
+m = export_for_training(model, example_inputs).module()
 
 # Annotate the model for quantization. This prepares the model for calibration.
 m = prepare_pt2e(m, xnnpack_quantizer)

shim/xplat/executorch/kernels/portable/op_registration_util.bzl

@@ -2,21 +2,7 @@ load("@fbsource//xplat/executorch/build:runtime_wrapper.bzl", "is_xplat", "runti
 load("@fbsource//xplat/executorch/build:selects.bzl", "selects")
 
 def get_compiler_optimization_flags():
-    # various ovr_configs are not available in oss
-    if not runtime.is_oss:
-      compiler_flags = select({
-        "DEFAULT": [],
-        "ovr_config//os:android-arm64": [
-              "-O2",
-        ],
-        "ovr_config//os:iphoneos": [
-            "-O2",
-        ],
-        "ovr_config//os:macos-arm64": [
-            "-O2",
-        ],
-      })
-      return compiler_flags
+    # App size regressons requires this to be baktraced until I have a better solution
     return []
 
 def op_target(name, deps = [], android_deps = [], _allow_third_party_deps = False, _aten_mode_deps = []):