diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/__init__.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/__init__.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..fe887e4c11579d1ee57610d666d53b95f01adc96
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/__init__.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/_gpu_trace.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/_gpu_trace.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..12e65b52edeea87aefefb492cc6fa957fe1f69dd
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/_gpu_trace.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/_memory_viz.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/_memory_viz.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..9a7fe4aa8a084a17107d858844123fe11bbe6932
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/_memory_viz.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/_sanitizer.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/_sanitizer.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..9861ab94a69953c1624df49377801f7471d6a5ff
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/_sanitizer.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/_utils.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/_utils.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..80fffd70cc79eba29222e56785d9e9061217ad76
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/_utils.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/comm.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/comm.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..32c95dc419ceb240279a8816a8125ba469b21a6d
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/comm.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/error.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/error.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..fd3882453c12746302383d15014e9638b76871af
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/error.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/gds.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/gds.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..cd7f4d6c2b0aeaf199d8c6cfa88135d25608f9aa
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/gds.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/graphs.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/graphs.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..c711deb0f84e14926921a8d2069c01c90bfe4941
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/graphs.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/jiterator.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/jiterator.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..7b76605dae88a159689af41a48cd5c61461416b4
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/jiterator.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/nccl.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/nccl.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..72a9f5753afe11b601523c0105c205887eb0f993
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/nccl.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/nvtx.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/nvtx.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..ab1489b699eaf79a0b7b46746a2793d5a5e6ff5e
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/nvtx.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/profiler.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/profiler.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..7ba90d0e7a45fcf404ce7db7cf3ad945b2215b83
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/profiler.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/random.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/random.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..430e549cfdcac0fc9d02ace87b049ab6b7397a71
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/random.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/sparse.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/sparse.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..9262a384a0414d2f8f0710037752b9a5aaabd73c
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/sparse.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/streams.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/streams.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..d8eb28667f21bc5f8cd0be7098183b9764df20d0
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/streams.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/tunable.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/tunable.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..68eec829dc829f9eaf765d5876028fbae169b4f1
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/__pycache__/tunable.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/amp/__init__.py b/.venv/lib/python3.11/site-packages/torch/cuda/amp/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..74520496372f549356a26c2c70b637b9e3ea4d4d
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/cuda/amp/__init__.py
@@ -0,0 +1,12 @@
+from .autocast_mode import autocast, custom_bwd, custom_fwd
+from .common import amp_definitely_not_available
+from .grad_scaler import GradScaler
+
+
+__all__ = [
+    "amp_definitely_not_available",
+    "autocast",
+    "custom_bwd",
+    "custom_fwd",
+    "GradScaler",
+]
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/amp/__pycache__/__init__.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/amp/__pycache__/__init__.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..670d3ebd633a0fed9a1c9ce604873bdd886848a1
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/amp/__pycache__/__init__.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/amp/__pycache__/autocast_mode.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/amp/__pycache__/autocast_mode.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..831301703e70ef4e03dfaef33e402fa6f96e2bdf
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/amp/__pycache__/autocast_mode.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/amp/__pycache__/common.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/amp/__pycache__/common.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..93dd26f04b2fcc790e0d4d1fb1f7204b1e95c06c
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/amp/__pycache__/common.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/amp/__pycache__/grad_scaler.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/cuda/amp/__pycache__/grad_scaler.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..6958220e862281e4aba6c857a2b2d0a512079b5d
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/cuda/amp/__pycache__/grad_scaler.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/amp/autocast_mode.py b/.venv/lib/python3.11/site-packages/torch/cuda/amp/autocast_mode.py
new file mode 100644
index 0000000000000000000000000000000000000000..d52ff7cf672bbdaaf4ac31a06a21cac8aabcfe0f
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/cuda/amp/autocast_mode.py
@@ -0,0 +1,90 @@
+# mypy: allow-untyped-defs
+import functools
+from typing import Any
+from typing_extensions import deprecated
+
+import torch
+
+
+__all__ = ["autocast", "custom_fwd", "custom_bwd"]
+
+
+class autocast(torch.amp.autocast_mode.autocast):
+    r"""See :class:`torch.autocast`.
+
+    ``torch.cuda.amp.autocast(args...)`` is deprecated. Please use ``torch.amp.autocast("cuda", args...)`` instead.
+    """
+
+    @deprecated(
+        "`torch.cuda.amp.autocast(args...)` is deprecated. "
+        "Please use `torch.amp.autocast('cuda', args...)` instead.",
+        category=FutureWarning,
+    )
+    def __init__(
+        self,
+        enabled: bool = True,
+        dtype: torch.dtype = torch.float16,
+        cache_enabled: bool = True,
+    ):
+        if torch._jit_internal.is_scripting():
+            self._enabled = enabled
+            self.device = "cuda"
+            self.fast_dtype = dtype
+            return
+        super().__init__(
+            "cuda", enabled=enabled, dtype=dtype, cache_enabled=cache_enabled
+        )
+
+    def __enter__(self):
+        if torch._jit_internal.is_scripting():
+            return self
+        return super().__enter__()
+
+    # TODO: discuss a unified TorchScript-friendly API for autocast
+    def __exit__(self, exc_type: Any, exc_val: Any, exc_tb: Any):  # type: ignore[override]
+        if torch._jit_internal.is_scripting():
+            return
+        return super().__exit__(exc_type, exc_val, exc_tb)
+
+    def __call__(self, func):
+        if torch._jit_internal.is_scripting():
+            return func
+        return super().__call__(func)
+
+
+# Preserved only for BC reasons
+@deprecated(
+    "`torch.cuda.amp.autocast_mode._cast(value, dtype)` is deprecated. "
+    "Please use `torch.amp.autocast_mode._cast(value, 'cuda', dtype)` instead.",
+    category=FutureWarning,
+)
+def _cast(value, dtype):
+    return torch.amp.autocast_mode._cast(value, "cuda", dtype)
+
+
+@deprecated(
+    "`torch.cuda.amp.custom_fwd(args...)` is deprecated. "
+    "Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.",
+    category=FutureWarning,
+)
+def custom_fwd(fwd=None, *, cast_inputs=None):
+    """
+    ``torch.cuda.amp.custom_fwd(args...)`` is deprecated. Please use
+    ``torch.amp.custom_fwd(args..., device_type='cuda')`` instead.
+    """
+    return functools.partial(torch.amp.custom_fwd, device_type="cuda")(
+        fwd=fwd, cast_inputs=cast_inputs
+    )
+
+
+@deprecated(
+    "`torch.cuda.amp.custom_bwd(args...)` is deprecated. "
+    "Please use `torch.amp.custom_bwd(args..., device_type='cuda')` instead.",
+    category=FutureWarning,
+)
+def custom_bwd(bwd):
+    """
+    ``torch.cuda.amp.custom_bwd(args...)`` is deprecated. Please use
+    ``torch.amp.custom_bwd(args..., device_type='cuda')`` instead.
+    """
+    return functools.partial(torch.amp.custom_bwd, device_type="cuda")(bwd)
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/amp/common.py b/.venv/lib/python3.11/site-packages/torch/cuda/amp/common.py
new file mode 100644
index 0000000000000000000000000000000000000000..915a9b4f4a9ca6c147abefd7c8ab1891ee5a8179
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/cuda/amp/common.py
@@ -0,0 +1,11 @@
+# mypy: allow-untyped-defs
+from importlib.util import find_spec
+
+import torch
+
+
+__all__ = ["amp_definitely_not_available"]
+
+
+def amp_definitely_not_available():
+    return not (torch.cuda.is_available() or find_spec("torch_xla"))
diff --git a/.venv/lib/python3.11/site-packages/torch/cuda/amp/grad_scaler.py b/.venv/lib/python3.11/site-packages/torch/cuda/amp/grad_scaler.py
new file mode 100644
index 0000000000000000000000000000000000000000..62e2020073c8ed99f7295edd1aaea4c54d815f63
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/cuda/amp/grad_scaler.py
@@ -0,0 +1,38 @@
+from typing_extensions import deprecated
+
+import torch
+
+# We need to keep this unused import for BC reasons
+from torch.amp.grad_scaler import OptState  # noqa: F401
+
+
+__all__ = ["GradScaler"]
+
+
+class GradScaler(torch.amp.GradScaler):
+    r"""
+    See :class:`torch.amp.GradScaler`.
+    ``torch.cuda.amp.GradScaler(args...)`` is deprecated. Please use ``torch.amp.GradScaler("cuda", args...)`` instead.
+    """
+
+    @deprecated(
+        "`torch.cuda.amp.GradScaler(args...)` is deprecated. "
+        "Please use `torch.amp.GradScaler('cuda', args...)` instead.",
+        category=FutureWarning,
+    )
+    def __init__(
+        self,
+        init_scale: float = 2.0**16,
+        growth_factor: float = 2.0,
+        backoff_factor: float = 0.5,
+        growth_interval: int = 2000,
+        enabled: bool = True,
+    ) -> None:
+        super().__init__(
+            "cuda",
+            init_scale=init_scale,
+            growth_factor=growth_factor,
+            backoff_factor=backoff_factor,
+            growth_interval=growth_interval,
+            enabled=enabled,
+        )
diff --git a/.venv/lib/python3.11/site-packages/torch/nn/quantizable/__init__.py b/.venv/lib/python3.11/site-packages/torch/nn/quantizable/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..c7628c5c15992efa600ea5520aed955ba42c6146
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/nn/quantizable/__init__.py
@@ -0,0 +1 @@
+from torch.nn.quantizable.modules import *  # noqa: F403
diff --git a/.venv/lib/python3.11/site-packages/torch/nn/quantizable/modules/activation.py b/.venv/lib/python3.11/site-packages/torch/nn/quantizable/modules/activation.py
new file mode 100644
index 0000000000000000000000000000000000000000..e4f7a5ca3b540edc9f9b1fc15899b63240b7ac79
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/nn/quantizable/modules/activation.py
@@ -0,0 +1,10 @@
+# flake8: noqa: F401
+r"""Quantizable Modules.
+
+This file is in the process of migration to `torch/ao/nn/quantizable`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate file under the `torch/ao/nn/quantizable/modules`,
+while adding an import statement here.
+"""
+from torch.ao.nn.quantizable.modules.activation import MultiheadAttention
diff --git a/.venv/lib/python3.11/site-packages/torch/nn/quantizable/modules/rnn.py b/.venv/lib/python3.11/site-packages/torch/nn/quantizable/modules/rnn.py
new file mode 100644
index 0000000000000000000000000000000000000000..8e355efcdd63d595580bdac22d46882b91b7d118
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/nn/quantizable/modules/rnn.py
@@ -0,0 +1,11 @@
+# flake8: noqa: F401
+r"""Quantizable Modules.
+
+This file is in the process of migration to `torch/ao/nn/quantizable`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate file under the `torch/ao/nn/quantizable/modules`,
+while adding an import statement here.
+"""
+
+from torch.ao.nn.quantizable.modules.rnn import LSTM, LSTMCell
diff --git a/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/__init__.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/__init__.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..1bc403123c9e44825f6fcefa3fd1e7fb05b7a98c
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/__init__.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/conv_expanded_weights.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/conv_expanded_weights.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..6ce3d66dc71a6bb411854a5b0cf01363bc770049
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/conv_expanded_weights.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/conv_utils.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/conv_utils.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..ae47a43257f4dbe6546206e96fd62c444fe977d0
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/conv_utils.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/expanded_weights_utils.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/expanded_weights_utils.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..61a1e4b87e39665bfb30049bc0f83964f9c24322
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/expanded_weights_utils.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/group_norm_expanded_weights.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/group_norm_expanded_weights.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..0c76cb825a6cbaccc8c8259aca57cf607bf133bc
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/group_norm_expanded_weights.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/instance_norm_expanded_weights.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/instance_norm_expanded_weights.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..8a7eb3114a94665de93037978656b427c95afdc4
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/instance_norm_expanded_weights.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/layer_norm_expanded_weights.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/layer_norm_expanded_weights.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..46941df277624dc68b16b4f257fc49d1ddb808ed
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/nn/utils/_expanded_weights/__pycache__/layer_norm_expanded_weights.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/__init__.py b/.venv/lib/python3.11/site-packages/torch/quantization/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..8789fea17a17ffa8e490a8d744892c5140a70ee2
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/__init__.py
@@ -0,0 +1,86 @@
+# mypy: allow-untyped-defs
+from .fake_quantize import *  # noqa: F403
+from .fuse_modules import fuse_modules
+from .fuser_method_mappings import *  # noqa: F403
+from .observer import *  # noqa: F403
+from .qconfig import *  # noqa: F403
+from .quant_type import *  # noqa: F403
+from .quantization_mappings import *  # noqa: F403
+from .quantize import *  # noqa: F403
+from .quantize_jit import *  # noqa: F403
+from .stubs import *  # noqa: F403
+
+
+def default_eval_fn(model, calib_data):
+    r"""
+    Default evaluation function takes a torch.utils.data.Dataset or a list of
+    input Tensors and run the model on the dataset
+    """
+    for data, target in calib_data:
+        model(data)
+
+
+__all__ = [
+    "QuantWrapper",
+    "QuantStub",
+    "DeQuantStub",
+    # Top level API for eager mode quantization
+    "quantize",
+    "quantize_dynamic",
+    "quantize_qat",
+    "prepare",
+    "convert",
+    "prepare_qat",
+    # Top level API for graph mode quantization on TorchScript
+    "quantize_jit",
+    "quantize_dynamic_jit",
+    "_prepare_ondevice_dynamic_jit",
+    "_convert_ondevice_dynamic_jit",
+    "_quantize_ondevice_dynamic_jit",
+    # Top level API for graph mode quantization on GraphModule(torch.fx)
+    # 'fuse_fx', 'quantize_fx',  # TODO: add quantize_dynamic_fx
+    # 'prepare_fx', 'prepare_dynamic_fx', 'convert_fx',
+    "QuantType",  # quantization type
+    # custom module APIs
+    "get_default_static_quant_module_mappings",
+    "get_static_quant_module_class",
+    "get_default_dynamic_quant_module_mappings",
+    "get_default_qat_module_mappings",
+    "get_default_qconfig_propagation_list",
+    "get_default_compare_output_module_list",
+    "get_quantized_operator",
+    "get_fuser_method",
+    # Sub functions for `prepare` and `swap_module`
+    "propagate_qconfig_",
+    "add_quant_dequant",
+    "swap_module",
+    "default_eval_fn",
+    # Observers
+    "ObserverBase",
+    "WeightObserver",
+    "HistogramObserver",
+    "observer",
+    "default_observer",
+    "default_weight_observer",
+    "default_placeholder_observer",
+    "default_per_channel_weight_observer",
+    # FakeQuantize (for qat)
+    "default_fake_quant",
+    "default_weight_fake_quant",
+    "default_fixed_qparams_range_neg1to1_fake_quant",
+    "default_fixed_qparams_range_0to1_fake_quant",
+    "default_per_channel_weight_fake_quant",
+    "default_histogram_fake_quant",
+    # QConfig
+    "QConfig",
+    "default_qconfig",
+    "default_dynamic_qconfig",
+    "float16_dynamic_qconfig",
+    "float_qparams_weight_only_qconfig",
+    # QAT utilities
+    "default_qat_qconfig",
+    "prepare_qat",
+    "quantize_qat",
+    # module transformations
+    "fuse_modules",
+]
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/_numeric_suite.py b/.venv/lib/python3.11/site-packages/torch/quantization/_numeric_suite.py
new file mode 100644
index 0000000000000000000000000000000000000000..49ccc8e69523f7dbee2335b788a2cb3a7db618a2
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/_numeric_suite.py
@@ -0,0 +1,28 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/ns/_numeric_suite.py`, while adding an import statement
+here.
+"""
+
+from torch.ao.ns._numeric_suite import (
+    _convert_tuple_to_list,
+    _dequantize_tensor_list,
+    _find_match,
+    _get_logger_dict_helper,
+    _is_identical_module_type,
+    compare_model_outputs,
+    compare_model_stub,
+    compare_weights,
+    get_logger_dict,
+    get_matching_activations,
+    Logger,
+    NON_LEAF_MODULE_TO_ADD_OBSERVER_ALLOW_LIST,
+    OutputLogger,
+    prepare_model_outputs,
+    prepare_model_with_stubs,
+    Shadow,
+    ShadowLogger,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/_numeric_suite_fx.py b/.venv/lib/python3.11/site-packages/torch/quantization/_numeric_suite_fx.py
new file mode 100644
index 0000000000000000000000000000000000000000..55cd7085740d0ce8de79491acbfc4888ebba21f8
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/_numeric_suite_fx.py
@@ -0,0 +1,26 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/ns/_numeric_suite_fx.py`, while adding an import statement
+here.
+"""
+
+from torch.ao.ns._numeric_suite_fx import (
+    _add_loggers_impl,
+    _add_loggers_one_model,
+    _add_shadow_loggers_impl,
+    _extract_logger_info_one_model,
+    _extract_weights_impl,
+    _extract_weights_one_model,
+    add_loggers,
+    add_shadow_loggers,
+    extend_logger_results_with_comparison,
+    extract_logger_info,
+    extract_shadow_logger_info,
+    extract_weights,
+    NSTracer,
+    OutputLogger,
+    RNNReturnType,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/_quantized_conversions.py b/.venv/lib/python3.11/site-packages/torch/quantization/_quantized_conversions.py
new file mode 100644
index 0000000000000000000000000000000000000000..8d930c366c0dd9857e463005474a2d59c04c4ae6
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/_quantized_conversions.py
@@ -0,0 +1,133 @@
+# mypy: allow-untyped-defs
+import torch
+
+
+# Pack pairs of int4 values into int8, in row major order; first int4
+# value goes into lower order bits, and second int4 value into higher
+# order bits of resulting int8 value.
+def pack_int4_to_int8(weight):
+    assert weight.dim() == 2
+    assert weight.shape[1] % 2 == 0
+    assert weight.dtype == torch.int8
+    return ((weight[:, 1::2] & 0xF) << 4) | (weight[:, 0::2] & 0xF)
+
+
+# Unpack quandruples of bits in int8 values into int4 values, in row
+# major order; lower 4 bits go into first int4 value goes, and upper 4
+# bits go into second int4 value.
+def unpack_int8_to_int4(weight):
+    assert weight.dim() == 2
+    assert weight.dtype == torch.int8
+    return torch.stack((weight & 0xF, (weight >> 4) & 0xF), dim=2).view(
+        weight.shape[0], 2 * weight.shape[1]
+    )
+
+
+# Transpose the weight matrix, and then reorder its elements according
+# to underlying requirements of CUTLASS library, so that it could be
+# used for CUTLASS-based mixed datatypes linear operation.
+def quantized_weight_reorder_for_mixed_dtypes_linear_cutlass(
+    weight, dtypeq, transpose=False
+):
+    assert weight.dim() == 2
+    assert weight.dtype == torch.int8
+    assert dtypeq == torch.int8 or dtypeq == torch.quint4x2
+    assert weight.device.type == "cuda"
+
+    device = weight.device
+
+    # subbyte_transpose
+    if not transpose:
+        if dtypeq == torch.int8:
+            outp = weight.T
+        elif dtypeq == torch.quint4x2:
+            outp = pack_int4_to_int8(unpack_int8_to_int4(weight.view(torch.int8)).T)
+    else:
+        outp = weight
+
+    ncols, nrows = outp.shape  # type: ignore[possibly-undefined]
+    assert nrows % (32 if dtypeq == torch.quint4x2 else 64) == 0
+    assert ncols % 64 == 0
+
+    # permute_B_rows_for_mixed_gemm
+    # (permute cols actually, as transpose is applied first here)
+    if dtypeq == torch.quint4x2:
+        cols_permuted = (
+            torch.tensor(
+                [0, 4, 8, 12, 1, 5, 9, 13, 2, 6, 10, 14, 3, 7, 11, 15],
+                device=device,
+            )
+            + (torch.arange(0, nrows // 16, device=device).reshape(-1, 1) * 16).expand(
+                nrows // 16, 16
+            )
+        ).view(-1)
+    else:
+        cols_permuted = (
+            torch.tensor(
+                [0, 1, 4, 5, 8, 9, 12, 13, 2, 3, 6, 7, 10, 11, 14, 15],
+                device=device,
+            )
+            + (torch.arange(0, nrows // 16, device=device).reshape(-1, 1) * 16).expand(
+                nrows // 16, 16
+            )
+        ).view(-1)
+    outp = outp.index_copy(1, cols_permuted, outp)
+
+    # interleave_column_major_tensor
+    magic0 = 4 if dtypeq == torch.quint4x2 else 2
+    magic1 = 32 // magic0
+
+    tmp0 = (
+        (torch.arange(0, ncols // magic0, device=device) * (nrows // 4 * magic0))
+        .view(-1, 1)
+        .repeat(1, nrows // 4 * magic0)
+        .view(-1)
+    )
+    tmp1 = (
+        (torch.arange(0, nrows // 4 // magic1, device=device) * (magic0 * magic1))
+        .view(-1, 1)
+        .repeat(1, magic1)
+        .view(-1)
+        .repeat(ncols)
+    )
+    tmp2 = (
+        (torch.arange(0, magic0, device=device) * magic1)
+        .view(-1, 1)
+        .repeat(1, nrows // 4)
+        .view(-1)
+        .repeat(ncols // magic0)
+    )
+    tmp3 = torch.arange(0, magic1, device=device).repeat(nrows // 4 * ncols // magic1)
+
+    outp_offsets = tmp0 + tmp1 + tmp2 + tmp3
+
+    tmp = outp.view(-1).view(torch.int32)
+    outp = torch.zeros_like(tmp)
+    outp.scatter_(0, outp_offsets, tmp)
+    outp = outp.view(weight.dtype)
+
+    # add_bias_and_interleave_quantized_tensor_inplace
+    tmp = outp.view(-1)
+
+    outp = torch.empty_like(tmp)
+    if dtypeq == torch.int8:
+        tmp = (tmp.to(torch.int) + 128).to(tmp.dtype)
+        outp[0::4] = tmp[0::4]
+        outp[1::4] = tmp[2::4]
+        outp[2::4] = tmp[1::4]
+        outp[3::4] = tmp[3::4]
+    elif dtypeq == torch.quint4x2:
+        tmp0 = ((tmp & 0xF) + 8) & 0xF
+        tmp0 = (tmp0[1::2] << 4) | tmp0[0::2]
+        tmp1 = (((tmp >> 4) & 0xF) + 8) & 0xF
+        tmp1 = (tmp1[1::2] << 4) | tmp1[0::2]
+        outp[0::4] = tmp0[0::2]
+        outp[1::4] = tmp0[1::2]
+        outp[2::4] = tmp1[0::2]
+        outp[3::4] = tmp1[1::2]
+
+    if dtypeq == torch.quint4x2:
+        nrows *= 2
+        ncols //= 2
+
+    return outp.view(nrows, ncols).view(torch.uint8)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fake_quantize.py b/.venv/lib/python3.11/site-packages/torch/quantization/fake_quantize.py
new file mode 100644
index 0000000000000000000000000000000000000000..69a5d730bfb68e89e24beb04ad13fd3fa5881ae9
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fake_quantize.py
@@ -0,0 +1,32 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/quantization/fake_quantize.py`, while adding an import statement
+here.
+"""
+
+from torch.ao.quantization.fake_quantize import (
+    _is_fake_quant_script_module,
+    _is_per_channel,
+    _is_per_tensor,
+    _is_symmetric_quant,
+    default_fake_quant,
+    default_fixed_qparams_range_0to1_fake_quant,
+    default_fixed_qparams_range_neg1to1_fake_quant,
+    default_fused_act_fake_quant,
+    default_fused_per_channel_wt_fake_quant,
+    default_fused_wt_fake_quant,
+    default_histogram_fake_quant,
+    default_per_channel_weight_fake_quant,
+    default_weight_fake_quant,
+    disable_fake_quant,
+    disable_observer,
+    enable_fake_quant,
+    enable_observer,
+    FakeQuantize,
+    FakeQuantizeBase,
+    FixedQParamsFakeQuantize,
+    FusedMovingAvgObsFakeQuantize,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fuse_modules.py b/.venv/lib/python3.11/site-packages/torch/quantization/fuse_modules.py
new file mode 100644
index 0000000000000000000000000000000000000000..6b704fa8094e8b367e9eba47102863ba845415b9
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fuse_modules.py
@@ -0,0 +1,22 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/quantization/fuse_modules.py`, while adding an import statement
+here.
+"""
+
+# TODO: These functions are not used outside the `fuse_modules.py`
+#       Keeping here for now, need to remove them later.
+from torch.ao.quantization.fuse_modules import (
+    _fuse_modules,
+    _get_module,
+    _set_module,
+    fuse_known_modules,
+    fuse_modules,
+    get_fuser_method,
+)
+
+# for backward compatiblity
+from torch.ao.quantization.fuser_method_mappings import fuse_conv_bn, fuse_conv_bn_relu
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fuser_method_mappings.py b/.venv/lib/python3.11/site-packages/torch/quantization/fuser_method_mappings.py
new file mode 100644
index 0000000000000000000000000000000000000000..cfb13ac96271fa7b926cc703918984760e6ede15
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fuser_method_mappings.py
@@ -0,0 +1,15 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/quantization/fuser_method_mappings.py`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.fuser_method_mappings import (
+    _DEFAULT_OP_LIST_TO_FUSER_METHOD,
+    fuse_conv_bn,
+    fuse_conv_bn_relu,
+    fuse_linear_bn,
+    get_fuser_method,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/__init__.py b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..c01cbd457374c27e40b07daca5ae1644a701767d
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__init__.py
@@ -0,0 +1,15 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate files under `torch/ao/quantization/fx/`, while adding an import statement
+here.
+"""
+
+from torch.ao.quantization.fx.convert import convert
+from torch.ao.quantization.fx.fuse import fuse
+
+# omitting files that's unlikely to be used right now, for example
+# the newly added lower_to_fbgemm etc.
+from torch.ao.quantization.fx.prepare import prepare
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/__init__.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/__init__.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..9366b0dfa2175380dd6d1d601f049929383b6e7e
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/__init__.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/_equalize.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/_equalize.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..a8cc841203a15ce4c79e7f8cd887720103c329ad
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/_equalize.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/convert.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/convert.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..81fc3c783e0ac322049bad3adb4752563695fb76
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/convert.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/fuse.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/fuse.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..2a16541001e0bce6342292b59469f7a0adcb93b6
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/fuse.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/fusion_patterns.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/fusion_patterns.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..c00a4e74bc77b9b27035172d8d1b0a6770082fa6
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/fusion_patterns.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/graph_module.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/graph_module.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..b42f548cd3050a68f1a51ebc08f8fc0b4d84bc50
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/graph_module.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/match_utils.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/match_utils.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..70ec5f590f8b92ddc91b596a9ffaf0ddbf9f74f2
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/match_utils.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/pattern_utils.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/pattern_utils.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..0bfe22ea50abbbded11c1634c46bd97867f9246e
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/pattern_utils.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/prepare.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/prepare.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..835d6857653280a9a54d6dec038116980a4db9b1
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/prepare.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/quantization_patterns.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/quantization_patterns.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..6f3a9a56aa16d3e8a07eb8d3ded65b5b819e3951
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/quantization_patterns.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/quantization_types.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/quantization_types.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..d7140696b712b7e5b6ed25b64a6f1d185b64b6f7
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/quantization_types.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/utils.cpython-311.pyc b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/utils.cpython-311.pyc
new file mode 100644
index 0000000000000000000000000000000000000000..94a6a340fb33d55518b45bd29c30991cba408845
Binary files /dev/null and b/.venv/lib/python3.11/site-packages/torch/quantization/fx/__pycache__/utils.cpython-311.pyc differ
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/_equalize.py b/.venv/lib/python3.11/site-packages/torch/quantization/fx/_equalize.py
new file mode 100644
index 0000000000000000000000000000000000000000..7acea4f84a2a0a82f134b6790e573f8f1cb677f2
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fx/_equalize.py
@@ -0,0 +1,38 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate files under `torch/ao/quantization/fx/`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.fx._equalize import (
+    _convert_equalization_ref,
+    _InputEqualizationObserver,
+    _WeightEqualizationObserver,
+    calculate_equalization_scale,
+    clear_weight_quant_obs_node,
+    convert_eq_obs,
+    CUSTOM_MODULE_SUPP_LIST,
+    custom_module_supports_equalization,
+    default_equalization_qconfig,
+    EqualizationQConfig,
+    fused_module_supports_equalization,
+    get_equalization_qconfig_dict,
+    get_layer_sqnr_dict,
+    get_op_node_and_weight_eq_obs,
+    input_equalization_observer,
+    is_equalization_observer,
+    maybe_get_next_equalization_scale,
+    maybe_get_next_input_eq_obs,
+    maybe_get_weight_eq_obs_node,
+    nn_module_supports_equalization,
+    node_supports_equalization,
+    remove_node,
+    reshape_scale,
+    scale_input_observer,
+    scale_weight_functional,
+    scale_weight_node,
+    update_obs_for_equalization,
+    weight_equalization_observer,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/convert.py b/.venv/lib/python3.11/site-packages/torch/quantization/fx/convert.py
new file mode 100644
index 0000000000000000000000000000000000000000..9d6ac350602bb7a97c773a3a09fec0780483379f
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fx/convert.py
@@ -0,0 +1,9 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate files under `torch/ao/quantization/fx/`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.fx.convert import convert
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/fuse.py b/.venv/lib/python3.11/site-packages/torch/quantization/fx/fuse.py
new file mode 100644
index 0000000000000000000000000000000000000000..67527080304fb31ddc54fe254533e2196f77a616
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fx/fuse.py
@@ -0,0 +1,9 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate files under `torch/ao/quantization/fx/`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.fx.fuse import fuse
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/fusion_patterns.py b/.venv/lib/python3.11/site-packages/torch/quantization/fx/fusion_patterns.py
new file mode 100644
index 0000000000000000000000000000000000000000..e29337b3f861e5b54dc9f37d39d12ad975ad1315
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fx/fusion_patterns.py
@@ -0,0 +1,9 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate files under `torch/ao/quantization/fx/`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.fx.fuse_handler import DefaultFuseHandler, FuseHandler
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/graph_module.py b/.venv/lib/python3.11/site-packages/torch/quantization/fx/graph_module.py
new file mode 100644
index 0000000000000000000000000000000000000000..a71e980a57ba141bdc5bbe9b283d69582eb8fd82
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fx/graph_module.py
@@ -0,0 +1,17 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate files under `torch/ao/quantization/fx/`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.fx.graph_module import (
+    _is_observed_module,
+    _is_observed_standalone_module,
+    FusedGraphModule,
+    GraphModule,
+    ObservedGraphModule,
+    ObservedStandaloneGraphModule,
+    QuantizedGraphModule,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/match_utils.py b/.venv/lib/python3.11/site-packages/torch/quantization/fx/match_utils.py
new file mode 100644
index 0000000000000000000000000000000000000000..8b49f7c645d8d1bc3a154d62a1295a90b155f986
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fx/match_utils.py
@@ -0,0 +1,14 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate files under `torch/ao/quantization/fx/`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.fx.match_utils import (
+    _find_matches,
+    _is_match,
+    _MatchResult,
+    MatchAllNode,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/pattern_utils.py b/.venv/lib/python3.11/site-packages/torch/quantization/fx/pattern_utils.py
new file mode 100644
index 0000000000000000000000000000000000000000..2a83e180fc4dbaa28d1d41a10037684f0afa6610
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fx/pattern_utils.py
@@ -0,0 +1,35 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate files under `torch/ao/quantization/fx/`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.fx.pattern_utils import (
+    _register_fusion_pattern,
+    _register_quant_pattern,
+    get_default_fusion_patterns,
+    get_default_output_activation_post_process_map,
+    get_default_quant_patterns,
+    QuantizeHandler,
+)
+
+
+# QuantizeHandler.__module__ = _NAMESPACE
+_register_fusion_pattern.__module__ = "torch.ao.quantization.fx.pattern_utils"
+get_default_fusion_patterns.__module__ = "torch.ao.quantization.fx.pattern_utils"
+_register_quant_pattern.__module__ = "torch.ao.quantization.fx.pattern_utils"
+get_default_quant_patterns.__module__ = "torch.ao.quantization.fx.pattern_utils"
+get_default_output_activation_post_process_map.__module__ = (
+    "torch.ao.quantization.fx.pattern_utils"
+)
+
+# __all__ = [
+#     "QuantizeHandler",
+#     "_register_fusion_pattern",
+#     "get_default_fusion_patterns",
+#     "_register_quant_pattern",
+#     "get_default_quant_patterns",
+#     "get_default_output_activation_post_process_map",
+# ]
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/prepare.py b/.venv/lib/python3.11/site-packages/torch/quantization/fx/prepare.py
new file mode 100644
index 0000000000000000000000000000000000000000..ca65dcc04dd0021f0065892ca86e209a1c218473
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fx/prepare.py
@@ -0,0 +1,9 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate files under `torch/ao/quantization/fx/`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.fx.prepare import prepare
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/quantization_patterns.py b/.venv/lib/python3.11/site-packages/torch/quantization/fx/quantization_patterns.py
new file mode 100644
index 0000000000000000000000000000000000000000..20d8cc52ee4fb16843becec5487d9d4ee46681c9
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fx/quantization_patterns.py
@@ -0,0 +1,48 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate files under `torch/ao/quantization/fx/`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.fx.quantize_handler import (
+    BatchNormQuantizeHandler,
+    BinaryOpQuantizeHandler,
+    CatQuantizeHandler,
+    ConvReluQuantizeHandler,
+    CopyNodeQuantizeHandler,
+    CustomModuleQuantizeHandler,
+    DefaultNodeQuantizeHandler,
+    EmbeddingQuantizeHandler,
+    FixedQParamsOpQuantizeHandler,
+    GeneralTensorShapeOpQuantizeHandler,
+    LinearReLUQuantizeHandler,
+    QuantizeHandler,
+    RNNDynamicQuantizeHandler,
+    StandaloneModuleQuantizeHandler,
+)
+
+
+QuantizeHandler.__module__ = "torch.ao.quantization.fx.quantization_patterns"
+BinaryOpQuantizeHandler.__module__ = "torch.ao.quantization.fx.quantization_patterns"
+CatQuantizeHandler.__module__ = "torch.ao.quantization.fx.quantization_patterns"
+ConvReluQuantizeHandler.__module__ = "torch.ao.quantization.fx.quantization_patterns"
+LinearReLUQuantizeHandler.__module__ = "torch.ao.quantization.fx.quantization_patterns"
+BatchNormQuantizeHandler.__module__ = "torch.ao.quantization.fx.quantization_patterns"
+EmbeddingQuantizeHandler.__module__ = "torch.ao.quantization.fx.quantization_patterns"
+RNNDynamicQuantizeHandler.__module__ = "torch.ao.quantization.fx.quantization_patterns"
+DefaultNodeQuantizeHandler.__module__ = "torch.ao.quantization.fx.quantization_patterns"
+FixedQParamsOpQuantizeHandler.__module__ = (
+    "torch.ao.quantization.fx.quantization_patterns"
+)
+CopyNodeQuantizeHandler.__module__ = "torch.ao.quantization.fx.quantization_patterns"
+CustomModuleQuantizeHandler.__module__ = (
+    "torch.ao.quantization.fx.quantization_patterns"
+)
+GeneralTensorShapeOpQuantizeHandler.__module__ = (
+    "torch.ao.quantization.fx.quantization_patterns"
+)
+StandaloneModuleQuantizeHandler.__module__ = (
+    "torch.ao.quantization.fx.quantization_patterns"
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/quantization_types.py b/.venv/lib/python3.11/site-packages/torch/quantization/fx/quantization_types.py
new file mode 100644
index 0000000000000000000000000000000000000000..a422cdd3142e04c8d16f495cc6cd65823451810b
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fx/quantization_types.py
@@ -0,0 +1,9 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate files under `torch/ao/quantization/fx/`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.utils import Pattern, QuantizerCls
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/fx/utils.py b/.venv/lib/python3.11/site-packages/torch/quantization/fx/utils.py
new file mode 100644
index 0000000000000000000000000000000000000000..ef35559884b7c430f1d5c72b21f72979108469a5
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/fx/utils.py
@@ -0,0 +1,20 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+appropriate files under `torch/ao/quantization/fx/`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.fx.utils import (
+    all_node_args_have_no_tensors,
+    assert_and_get_unique_device,
+    create_getattr_from_value,
+    get_custom_module_class_keys,
+    get_linear_prepack_op_for_dtype,
+    get_new_attr_name_with_prefix,
+    get_non_observable_arg_indexes_and_types,
+    get_qconv_prepack_op,
+    graph_module_from_producer_nodes,
+    maybe_get_next_module,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/observer.py b/.venv/lib/python3.11/site-packages/torch/quantization/observer.py
new file mode 100644
index 0000000000000000000000000000000000000000..6e6c7c1917c83433fc19f016140b25d060284535
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/observer.py
@@ -0,0 +1,36 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/quantization/observer.py`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.observer import (
+    _is_activation_post_process,
+    _is_per_channel_script_obs_instance,
+    _ObserverBase,
+    _PartialWrapper,
+    _with_args,
+    _with_callable_args,
+    ABC,
+    default_debug_observer,
+    default_dynamic_quant_observer,
+    default_float_qparams_observer,
+    default_histogram_observer,
+    default_observer,
+    default_per_channel_weight_observer,
+    default_placeholder_observer,
+    default_weight_observer,
+    get_observer_state_dict,
+    HistogramObserver,
+    load_observer_state_dict,
+    MinMaxObserver,
+    MovingAverageMinMaxObserver,
+    MovingAveragePerChannelMinMaxObserver,
+    NoopObserver,
+    ObserverBase,
+    PerChannelMinMaxObserver,
+    PlaceholderObserver,
+    RecordingObserver,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/qconfig.py b/.venv/lib/python3.11/site-packages/torch/quantization/qconfig.py
new file mode 100644
index 0000000000000000000000000000000000000000..6bb7e14110cb9cdc4e9c2c418c6776ea6445f0d3
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/qconfig.py
@@ -0,0 +1,30 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/quantization/qconfig.py`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.qconfig import (
+    _add_module_to_qconfig_obs_ctr,
+    _assert_valid_qconfig,
+    default_activation_only_qconfig,
+    default_debug_qconfig,
+    default_dynamic_qconfig,
+    default_per_channel_qconfig,
+    default_qat_qconfig,
+    default_qat_qconfig_v2,
+    default_qconfig,
+    default_weight_only_qconfig,
+    float16_dynamic_qconfig,
+    float16_static_qconfig,
+    float_qparams_weight_only_qconfig,
+    get_default_qat_qconfig,
+    get_default_qconfig,
+    per_channel_dynamic_qconfig,
+    QConfig,
+    qconfig_equals,
+    QConfigAny,
+    QConfigDynamic,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/quant_type.py b/.venv/lib/python3.11/site-packages/torch/quantization/quant_type.py
new file mode 100644
index 0000000000000000000000000000000000000000..8555f03792661f39c85c8facf3f911786cc25d0f
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/quant_type.py
@@ -0,0 +1,10 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/quantization/quant_type.py`, while adding an import statement
+here.
+"""
+
+from torch.ao.quantization.quant_type import _get_quant_type_to_str, QuantType
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/quantization_mappings.py b/.venv/lib/python3.11/site-packages/torch/quantization/quantization_mappings.py
new file mode 100644
index 0000000000000000000000000000000000000000..8b44a980ce82fbfa5a81ad906499806cf99b876f
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/quantization_mappings.py
@@ -0,0 +1,29 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/quantization/quantization_mappings.py`, while adding an import statement
+here.
+"""
+from torch.ao.quantization.quantization_mappings import (
+    _get_special_act_post_process,
+    _has_special_act_post_process,
+    _INCLUDE_QCONFIG_PROPAGATE_LIST,
+    DEFAULT_DYNAMIC_QUANT_MODULE_MAPPINGS,
+    DEFAULT_FLOAT_TO_QUANTIZED_OPERATOR_MAPPINGS,
+    DEFAULT_MODULE_TO_ACT_POST_PROCESS,
+    DEFAULT_QAT_MODULE_MAPPINGS,
+    DEFAULT_REFERENCE_STATIC_QUANT_MODULE_MAPPINGS,
+    DEFAULT_STATIC_QUANT_MODULE_MAPPINGS,
+    get_default_compare_output_module_list,
+    get_default_dynamic_quant_module_mappings,
+    get_default_float_to_quantized_operator_mappings,
+    get_default_qat_module_mappings,
+    get_default_qconfig_propagation_list,
+    get_default_static_quant_module_mappings,
+    get_dynamic_quant_module_class,
+    get_quantized_operator,
+    get_static_quant_module_class,
+    no_observer_set,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/quantize.py b/.venv/lib/python3.11/site-packages/torch/quantization/quantize.py
new file mode 100644
index 0000000000000000000000000000000000000000..600d3a46fed0346e3ae8909872cd5bf3c733860c
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/quantize.py
@@ -0,0 +1,30 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/quantization/quantize.py`, while adding an import statement
+here.
+"""
+
+from torch.ao.quantization.quantize import (
+    _add_observer_,
+    _convert,
+    _get_observer_dict,
+    _get_unique_devices_,
+    _is_activation_post_process,
+    _observer_forward_hook,
+    _propagate_qconfig_helper,
+    _register_activation_post_process_hook,
+    _remove_activation_post_process,
+    _remove_qconfig,
+    add_quant_dequant,
+    convert,
+    prepare,
+    prepare_qat,
+    propagate_qconfig_,
+    quantize,
+    quantize_dynamic,
+    quantize_qat,
+    swap_module,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/quantize_fx.py b/.venv/lib/python3.11/site-packages/torch/quantization/quantize_fx.py
new file mode 100644
index 0000000000000000000000000000000000000000..649142c7a7eee9885d96b37f70e582f3ea9a9f8d
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/quantize_fx.py
@@ -0,0 +1,26 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/quantization/quantize_fx.py`, while adding an import statement
+here.
+"""
+
+from torch.ao.quantization.fx.graph_module import ObservedGraphModule
+from torch.ao.quantization.quantize_fx import (
+    _check_is_graph_module,
+    _convert_fx,
+    _convert_standalone_module_fx,
+    _fuse_fx,
+    _prepare_fx,
+    _prepare_standalone_module_fx,
+    _swap_ff_with_fxff,
+    convert_fx,
+    fuse_fx,
+    prepare_fx,
+    prepare_qat_fx,
+    QuantizationTracer,
+    Scope,
+    ScopeContextManager,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/quantize_jit.py b/.venv/lib/python3.11/site-packages/torch/quantization/quantize_jit.py
new file mode 100644
index 0000000000000000000000000000000000000000..aa627dc7bb51ef7ea1fde7e2e5da283c9f6c8900
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/quantize_jit.py
@@ -0,0 +1,26 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/quantization/quantize_jit.py`, while adding an import statement
+here.
+"""
+
+from torch.ao.quantization.quantize_jit import (
+    _check_forward_method,
+    _check_is_script_module,
+    _convert_jit,
+    _prepare_jit,
+    _prepare_ondevice_dynamic_jit,
+    _quantize_jit,
+    convert_dynamic_jit,
+    convert_jit,
+    fuse_conv_bn_jit,
+    prepare_dynamic_jit,
+    prepare_jit,
+    quantize_dynamic_jit,
+    quantize_jit,
+    script_qconfig,
+    script_qconfig_dict,
+)
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/stubs.py b/.venv/lib/python3.11/site-packages/torch/quantization/stubs.py
new file mode 100644
index 0000000000000000000000000000000000000000..d3fd5c63683dc572c35cabc202ee4ddb2b0053c6
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/stubs.py
@@ -0,0 +1,10 @@
+# flake8: noqa: F401
+r"""
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/quantization/stubs.py`, while adding an import statement
+here.
+"""
+
+from torch.ao.quantization.stubs import DeQuantStub, QuantStub, QuantWrapper
diff --git a/.venv/lib/python3.11/site-packages/torch/quantization/utils.py b/.venv/lib/python3.11/site-packages/torch/quantization/utils.py
new file mode 100644
index 0000000000000000000000000000000000000000..7d51d58f38d7462713f84ab62427852c1dd8e52c
--- /dev/null
+++ b/.venv/lib/python3.11/site-packages/torch/quantization/utils.py
@@ -0,0 +1,29 @@
+# flake8: noqa: F401
+r"""
+Utils shared by different modes of quantization (eager/graph)
+
+This file is in the process of migration to `torch/ao/quantization`, and
+is kept here for compatibility while the migration process is ongoing.
+If you are adding a new entry/functionality, please, add it to the
+`torch/ao/quantization/utils.py`, while adding an import statement
+here.
+"""
+
+from torch.ao.quantization.utils import (
+    activation_dtype,
+    activation_is_int8_quantized,
+    activation_is_statically_quantized,
+    calculate_qmin_qmax,
+    check_min_max_valid,
+    get_combined_dict,
+    get_qconfig_dtypes,
+    get_qparam_dict,
+    get_quant_type,
+    get_swapped_custom_module_class,
+    getattr_from_fqn,
+    is_per_channel,
+    is_per_tensor,
+    weight_dtype,
+    weight_is_quantized,
+    weight_is_statically_quantized,
+)