refactor(mge): polish api

- refactor(mge): add support for optimizer.step().clear_grad() idiom

- refactor(mge): rename some methods of GradManager

- refactor(mge): remove tensor_nn and TensorDict

- refactor(mge): remove Buffer

- refactor(mge): remove requires_grad flag

- refactor(mge): add a default grad=None attribute to Tensor

- refactor(mge): deprecation for 1.0

GitOrigin-RevId: 3b723d9387

imperative/python/megengine/__init__.py

@@ -74,8 +74,7 @@ from .core._imperative_rt.utils import _set_fork_exec_path_for_timed_func
 from .device import *
 from .logger import enable_debug_log, get_logger, set_log_file, set_log_level
 from .serialization import load, save
-from .tensor import Tensor, tensor
-from .tensor_nn import Buffer, Parameter
+from .tensor import Parameter, Tensor, tensor
 from .version import __version__
 
 _set_fork_exec_path_for_timed_func(

imperative/python/megengine/autodiff/grad_manager.py

@@ -22,7 +22,7 @@ class GradManager:
         self._after_backward_callback = []
         self._gradients = dict()
 
-    def register(self, params, callbacks=None):
+    def attach(self, params, callbacks=None):
         if callbacks is None:
             callbacks = []
         if isinstance(callbacks, Callable):
@@ -62,7 +62,7 @@ class GradManager:
                 if isinstance(grad, Future):
                     grad = grad.get()
                 param = self._param_dict[p]
-                if getattr(param, "grad", None) is None:
+                if param.grad is None:
                     param.grad = grad
                 else:
                     param.grad += grad
@@ -70,9 +70,9 @@ class GradManager:
             self._stop_record()
             backwarding_grad_manager = cache
 
-    def __enter__(self):
+    def record(self):
         if self._recording:
-            return self
+            raise RuntimeError("already recording")
         grad = Grad()
         self._recording = True
         self._grad = grad
@@ -88,16 +88,22 @@ class GradManager:
 
             grad.wrt(param_wrapper, callback=callback)
         grad.__enter__()
-        return self
 
-    def __exit__(self, exc_type, exc_val, exc_tb):
+    def release(self):
+        if not self._recording:
+            raise RuntimeError("not recording")
         self._stop_record()
 
-    record = __enter__
-
     def _stop_record(self):
         if self._grad is not None:
             self._grad.__exit__(None, None, None)
         self._recording = False
         self._grad = None
         self._gradients = dict()
+
+    def __enter__(self):
+        self.record()
+        return self
+
+    def __exit__(self, exc_type, exc_val, exc_tb):
+        self._stop_record()

imperative/python/megengine/core/ops/_internal/misc_ops.py

@@ -70,7 +70,7 @@ class Dimshuffle(PodOpVisitor):
             return bytes(ctypes.c_uint32(0)) + bytes(self)
 
     def __init__(self, pattern, ndim=0):
-        assert isinstance(pattern, collections.Iterable)
+        assert isinstance(pattern, collections.abc.Iterable)
         assert len(pattern) <= TensorShape.MAX_NDIM
         pattern_array = Dimshuffle.Pattern.Pattern_Array()
         for idx, v in enumerate(pattern):

imperative/python/megengine/core/tensor/megbrain_graph.py

@@ -231,13 +231,13 @@ class OpNode:
 
 
 def _wrap(x):
-    if isinstance(x, collections.Sequence):
+    if isinstance(x, collections.abc.Sequence):
         return type(x)(map(_wrap, x))
     return x.graph._wrap(x)
 
 
 def _unwrap(x):
-    if isinstance(x, collections.Sequence):
+    if isinstance(x, collections.abc.Sequence):
         return type(x)(map(_unwrap, x))
     return x._node
 

imperative/python/megengine/core/tensor/tensor_wrapper.py

@@ -166,7 +166,7 @@ def _reduce(mode):
 
             op = builtin.Reduce(mode=mode, axis=0)
             (result,) = apply(op, data)
-        elif isinstance(axis, collections.Iterable):
+        elif isinstance(axis, collections.abc.Iterable):
             axis = list(axis)
             axis.sort(reverse=True)
 
@@ -204,7 +204,9 @@ def _todo(*_):
 
 def _expand_args(args):
     if len(args) == 1:
-        if isinstance(args[0], (collections.Sequence, TensorBase, TensorWrapperBase)):
+        if isinstance(
+            args[0], (collections.abc.Sequence, TensorBase, TensorWrapperBase)
+        ):
             args = args[0]
     return args
 

imperative/python/megengine/core/tensor/utils.py

@@ -143,7 +143,7 @@ def astensor1d(x, *reference, dtype=None, device=None):
             (x,) = Const(x, dtype=dtype, device=device)(*reference)
         return x
 
-    if not isinstance(x, collections.Sequence):
+    if not isinstance(x, collections.abc.Sequence):
         raise TypeError
 
     if any(isinstance(i, (TensorBase, TensorWrapperBase)) for i in x):

imperative/python/megengine/functional/math.py

@@ -432,7 +432,7 @@ def argmin(
         [0]
 
     """
-    if isinstance(axis, collections.Iterable):
+    if isinstance(axis, collections.abc.Iterable):
         axis = list(axis)
         axis.sort(reverse=True)
 
@@ -486,7 +486,7 @@ def argmax(
         [5]
 
     """
-    if isinstance(axis, collections.Iterable):
+    if isinstance(axis, collections.abc.Iterable):
         axis = list(axis)
         axis.sort(reverse=True)
 

imperative/python/megengine/functional/types.py

@@ -15,7 +15,7 @@ def get_ndtuple(value, *, n, allow_zero=True):
 
     :type allow_zero: bool
     :param allow_zero: whether to allow zero tuple value"""
-    if not isinstance(value, collections.Iterable):
+    if not isinstance(value, collections.abc.Iterable):
         value = int(value)
         value = tuple([value for i in range(n)])
     else:

imperative/python/megengine/jit/tracing.py

@@ -502,7 +502,7 @@ class trace:
             raise TypeError(
                 "cannot specify output_names when output is already in dict format"
             )
-        if output_names and not isinstance(output_names, collections.Sequence):
+        if output_names and not isinstance(output_names, collections.abc.Sequence):
             output_names = (output_names,)
         if output_names and len(output_names) != len(self._output_bindings):
             raise ValueError(
@@ -510,7 +510,7 @@ class trace:
                     len(self._output_bindings)
                 )
             )
-        if arg_names and not isinstance(arg_names, collections.Sequence):
+        if arg_names and not isinstance(arg_names, collections.abc.Sequence):
             arg_names = (arg_names,)
         if arg_names and len(arg_names) != len(self._arg_bindings):
             raise ValueError(
@@ -646,9 +646,9 @@ class trace:
 
     def _process_outputs(self, outputs):
         output_names = None
-        if isinstance(outputs, collections.Mapping):
+        if isinstance(outputs, collections.abc.Mapping):
             output_names, outputs = zip(*sorted(outputs.items()))
-        elif not isinstance(outputs, collections.Sequence):
+        elif not isinstance(outputs, collections.abc.Sequence):
             outputs = (outputs,)
 
         if not self._untraced:

imperative/python/megengine/module/__init__.py

@@ -18,7 +18,6 @@ from .embedding import Embedding
 from .identity import Identity
 from .linear import Linear
 from .module import Module
-from .parampack import ParamPack
 from .pooling import AvgPool2d, MaxPool2d
 from .quant_dequant import DequantStub, QuantStub
 from .sequential import Sequential

imperative/python/megengine/module/activation.py

@@ -9,7 +9,7 @@
 import numpy as np
 
 from ..functional import leaky_relu, prelu, relu, sigmoid, softmax
-from ..tensor_nn import Parameter
+from ..tensor import Parameter
 from .module import Module
 
 

imperative/python/megengine/module/batchnorm.py

@@ -12,7 +12,7 @@ import numpy as np
 
 from ..distributed.group import WORLD, Group
 from ..functional import batch_norm2d, sync_batch_norm
-from ..tensor_nn import Buffer, Parameter, Tensor
+from ..tensor import Parameter, Tensor
 from . import init
 from .module import Module
 
@@ -45,8 +45,8 @@ class _BatchNorm(Module):
         tshape = (1, self.num_features, 1, 1)
 
         if self.track_running_stats:
-            self.running_mean = Buffer(np.zeros(tshape, dtype=np.float32))
-            self.running_var = Buffer(np.ones(tshape, dtype=np.float32))
+            self.running_mean = Tensor(np.zeros(tshape, dtype=np.float32))
+            self.running_var = Tensor(np.ones(tshape, dtype=np.float32))
         else:
             self.running_mean = None
             self.running_var = None

imperative/python/megengine/module/conv.py

@@ -13,7 +13,7 @@ import numpy as np
 from ..core.ops._internal import param_defs as P
 from ..functional import conv2d, conv_transpose2d, local_conv2d, relu
 from ..functional.types import _pair, _pair_nonzero
-from ..tensor_nn import Parameter
+from ..tensor import Parameter
 from . import init
 from .module import Module
 

imperative/python/megengine/module/embedding.py

@@ -11,7 +11,7 @@ from typing import Optional
 import numpy as np
 
 from ..functional import embedding as embedding_func
-from ..tensor_nn import Parameter
+from ..tensor import Parameter
 from . import init
 from .module import Module
 
@@ -72,6 +72,7 @@ class Embedding(Module):
         max_norm: Optional[float] = None,
         norm_type: Optional[float] = None,
         initial_weight: Parameter = None,
+        freeze: bool = False,
     ):
         super().__init__()
         if padding_idx is not None:
@@ -83,6 +84,7 @@ class Embedding(Module):
         self.norm_type = norm_type
         self.num_embeddings = num_embeddings
         self.embedding_dim = embedding_dim
+        self.freeze = freeze
         if initial_weight is None:
             self.weight = Parameter(
                 np.random.uniform(
@@ -101,7 +103,11 @@ class Embedding(Module):
         init.normal_(self.weight)
 
     def forward(self, inputs):
-        return embedding_func(inputs, self.weight)
+        if self.freeze:
+            weight = self.weight.detach()
+        else:
+            weight = self.weight
+        return embedding_func(inputs, weight)
 
     @classmethod
     def from_pretrained(
@@ -166,6 +172,6 @@ class Embedding(Module):
             padding_idx=padding_idx,
             max_norm=max_norm,
             norm_type=norm_type,
+            freeze=freeze,
         )
-        embedding.weight.requires_grad = not freeze
         return embedding

imperative/python/megengine/module/init.py

@@ -23,7 +23,7 @@ def fill_(tensor: Tensor, val: Union[float, int]) -> None:
     :param tensor: An n-dimentional tensor to be initialized
     :param val: The value to be filled throughout the tensor
     """
-    tensor.set_value(full(shape=tensor.shape, value=val, dtype=tensor.dtype))
+    tensor._reset(full(shape=tensor.shape, value=val, dtype=tensor.dtype))
 
 
 def zeros_(tensor: Tensor) -> None:
@@ -50,7 +50,7 @@ def uniform_(tensor: Tensor, a: float = 0.0, b: float = 1.0) -> None:
     :param a: Lower bound of the sampling interval
     :param b: Upper bound of the sampling interval
     """
-    tensor.set_value(uniform(tensor.shape, low=a, high=b).astype(tensor.dtype))
+    tensor._reset(uniform(tensor.shape, low=a, high=b).astype(tensor.dtype))
 
 
 def normal_(tensor: Tensor, mean: float = 0.0, std: float = 1.0) -> None:
@@ -61,7 +61,7 @@ def normal_(tensor: Tensor, mean: float = 0.0, std: float = 1.0) -> None:
     :param mean: The mean of the normal distribution
     :param std: The standard deviation of the normal distribution
     """
-    tensor.set_value(gaussian(tensor.shape, mean=mean, std=std).astype(tensor.dtype))
+    tensor._reset(gaussian(tensor.shape, mean=mean, std=std).astype(tensor.dtype))
 
 
 def calculate_gain(

imperative/python/megengine/module/linear.py

@@ -8,7 +8,7 @@
 import numpy as np
 
 from ..functional import linear
-from ..tensor_nn import Parameter
+from ..tensor import Parameter
 from . import init
 from .module import Module
 

imperative/python/megengine/module/module.py

@@ -5,6 +5,7 @@
 # Unless required by applicable law or agreed to in writing,
 # software distributed under the License is distributed on an
 # "AS IS" BASIS, WITHOUT ARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+import warnings
 from abc import ABCMeta, abstractmethod
 from collections import OrderedDict
 from typing import Any, Callable, Iterable, Optional, Set, Tuple, Union
@@ -14,8 +15,8 @@ import numpy as np
 from ..core.tensor.dtype import is_quantize
 from ..core.tensor.utils import make_shape_tuple
 from ..logger import get_logger
-from ..tensor import Tensor
-from ..tensor_nn import Buffer, Parameter
+from ..tensor import Parameter, Tensor
+from ..utils.deprecation import deprecated
 from ..utils.hook import HookHandler
 
 logger = get_logger(__name__)
@@ -48,7 +49,7 @@ def _is_parameter(obj):
 
 
 def _is_buffer(obj):
-    return isinstance(obj, Buffer)
+    return isinstance(obj, Tensor) and not isinstance(obj, Parameter)
 
 
 def _is_module(obj):
@@ -163,49 +164,43 @@ class Module(metaclass=ABCMeta):
                         seen=seen,
                     )
 
-    def parameters(
-        self, requires_grad: Optional[bool] = None, recursive: bool = True, **kwargs
-    ) -> Iterable[Parameter]:
+    def parameters(self, recursive: bool = True, **kwargs) -> Iterable[Parameter]:
         r"""Returns an iterable for the :class:`~.Parameter` of the module.
 
-        :param requires_grad: Limitation over the :attr:`~.Parameter.requires_grad`
-            attribute of returned :class:`.Parameter`. ``None`` for no limitation.
         :param recursive: If ``True``, returns all :class:`~.Parameter` within this
             module, else only returns :class:`~.Parameter` that are direct attributes
             of this module.
         """
 
+        if "requires_grad" in kwargs:
+            del kwargs["requires_grad"]
+            warnings.warn("passing requires_grad has no effect currently")
+
         def predicate(obj) -> bool:
-            return _is_parameter(obj) and (
-                requires_grad is None or obj.requires_grad == requires_grad
-            )
+            return _is_parameter(obj)
 
         yield from self._flatten(
             with_key=False, predicate=predicate, recursive=recursive, **kwargs
         )
 
     def named_parameters(
-        self,
-        requires_grad: Optional[bool] = None,
-        prefix: Optional[str] = None,
-        recursive: bool = True,
-        **kwargs
+        self, prefix: Optional[str] = None, recursive: bool = True, **kwargs
     ) -> Iterable[Tuple[str, Parameter]]:
         """Returns an iterable for key :class:`~.Parameter` pairs of the module, where
         ``key`` is the dotted path from this module to the :class:`~.Parameter` .
 
-        :param requires_grad: Limitation over the :attr:`~.Parameter.requires_grad`
-            attribute of returned :class:`~.Parameter` . ``None`` for no limitation.
         :param prefix: The prefix prepended to the keys.
         :param recursive: If ``True``, returns all :class:`~.Parameter` within this
             module, else only returns :class:`~.Parameter` that are direct attributes
             of this module.
         """
 
+        if "requires_grad" in kwargs:
+            del kwargs["requires_grad"]
+            warnings.warn("passing requires_grad has no effect currently")
+
         def predicate(obj) -> bool:
-            return _is_parameter(obj) and (
-                requires_grad is None or obj.requires_grad == requires_grad
-            )
+            return _is_parameter(obj)
 
         yield from self._flatten(
             with_key=True,
@@ -215,11 +210,13 @@ class Module(metaclass=ABCMeta):
             **kwargs,
         )
 
-    def buffers(self, recursive: bool = True, **kwargs) -> Iterable[Buffer]:
-        """Returns an iterable for the :class:`~.Buffer` of the module.
+    def buffers(self, recursive: bool = True, **kwargs) -> Iterable[Tensor]:
+        """Returns an iterable for the buffers of the module.
 
-        :param recursive: If ``True``, returns all :class:`~.Buffer` within this
-            module, else only returns :class:`~.Buffer` that are direct attributes
+        Buffer is defined to be :class:`~.Tensor` excluding :class:`~.Parameter`.
+
+        :param recursive: If ``True``, returns all buffers within this
+            module, else only returns buffers that are direct attributes
             of this module.
         """
         yield from self._flatten(
@@ -228,13 +225,15 @@ class Module(metaclass=ABCMeta):
 
     def named_buffers(
         self, prefix: Optional[str] = None, recursive: bool = True, **kwargs
-    ) -> Iterable[Tuple[str, Buffer]]:
-        """Returns an iterable for key :class:`~.Buffer` pairs of the module, where
-        ``key`` is the dotted path from this module to the :class:`~.Buffer` .
+    ) -> Iterable[Tuple[str, Tensor]]:
+        """Returns an iterable for key buffer pairs of the module, where
+        ``key`` is the dotted path from this module to the buffer.
+
+        Buffer is defined to be :class:`~.Tensor` excluding :class:`~.Parameter`.
 
         :param prefix: The prefix prepended to the keys.
-        :param recursive: If ``True``, returns all :class:`~.Buffer` within this
-            module, else only returns :class:`~.Buffer` that are direct attributes
+        :param recursive: If ``True``, returns all buffers within this
+            module, else only returns buffers that are direct attributes
             of this module.
         """
         yield from self._flatten(
@@ -297,6 +296,7 @@ class Module(metaclass=ABCMeta):
         for it in self.modules():
             fn(it)
 
+    @deprecated(version="1.0")
     def zero_grad(self) -> None:
         """Set all parameters' grads to zero
         """
@@ -505,7 +505,7 @@ class Module(metaclass=ABCMeta):
             # scale/zero_points maybe invalid, use pretrained dtype instead.
             if is_quantize(to_be_load.dtype) and is_quantize(var.dtype):
                 var = var.astype(to_be_load.dtype)
-            var.set_value(to_be_load)
+            var._reset(to_be_load)
             loaded.append(k)
 
         return set(loaded), set(skipped)

imperative/python/megengine/module/parampack.py

@@ -1,156 +0,0 @@
-# -*- coding: utf-8 -*-
-# MegEngine is Licensed under the Apache License, Version 2.0 (the "License")
-#
-# Copyright (c) 2014-2020 Megvii Inc. All rights reserved.
-#
-# Unless required by applicable law or agreed to in writing,
-# software distributed under the License is distributed on an
-# "AS IS" BASIS, WITHOUT ARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-import collections
-from typing import Callable, Iterable, Optional, Tuple
-
-import numpy as np
-
-from ..tensor_nn import Parameter, Tensor
-from .module import Module
-
-
-class ParamPack(Module):
-    r"""Pack module's parameters by gathering their memory to continuous address.
-    Using (device, dtype, requires_grad) as key, for example ('gpu0', float32, True),
-    parameters with same key will be packed togather.
-    It helps a lot for multimachine training by speeding up allreduce gradients.
-
-    :param model: the module you want to pack parameters.
-    :param nr_ignore_first: how many parameters will be unpacked at first.
-    :param max_size_per_group: upper bound of packed parameters' size in MB.
-    :param max_nr_params_per_group: upper bound of the number of parameters of each group.
-
-    """
-
-    def __init__(
-        self,
-        model: Module,
-        nr_ignore_first: int = 8,
-        max_size_per_group: int = 10,
-        max_nr_params_per_group: int = 100,
-        group_func: Callable = lambda name, param: 0,
-    ):
-        super().__init__()
-        self._model = model
-        self._nr_ignore_first = nr_ignore_first
-        self._max_size_per_group = max_size_per_group
-        self._max_nr_params_per_group = max_nr_params_per_group
-        self._group_func = group_func
-        self._grouped_params = []
-        self._packed_params = []
-
-        params = model.named_parameters()
-        self._pack_params(params)
-
-    def parameters(self, requires_grad: Optional[bool] = None) -> Iterable[Parameter]:
-        for param in self._packed_params:
-            if requires_grad is None or param.requires_grad == requires_grad:
-                yield param
-
-    def named_parameters(
-        self, requires_grad: Optional[bool] = None
-    ) -> Iterable[Tuple[str, Parameter]]:
-        for idx, param in enumerate(self._packed_params):
-            if requires_grad is None or param.requires_grad == requires_grad:
-                yield "packed_param_" + str(idx), param
-
-    def _pack_params(self, params: Iterable[Tuple[str, Parameter]]):
-        groups = collections.defaultdict(list)
-        ignored = 0
-        param_id = 0
-        for name, param in params:
-            if self._nr_ignore_first > ignored:
-                ignored += 1
-                self._grouped_params.append([{"shape": param.shape, "id": param_id}])
-                param.pack_group_key = self._group_func(name, param)
-                self._packed_params.append(param)
-            else:
-                key = (
-                    param.dtype,
-                    param.device,
-                    param.requires_grad,
-                    self._group_func(name, param),
-                )
-                groups[key].append({"tensor": param, "id": param_id})
-            param_id += 1
-        for (dtype, device, requires_grad, group_key) in groups.keys():
-            dtype_sz = np.dtype(dtype).itemsize
-            align = device.mem_align
-            if align < dtype_sz:
-                align = 1
-            else:
-                assert align % dtype_sz == 0
-                align //= dtype_sz
-
-            group = groups[(dtype, device, requires_grad, group_key)]
-            while group:
-                aligned_pos = []
-                offset = 0
-                params = []
-                idx = 0
-                while idx < len(group):
-                    param = group[idx]
-                    assert param["tensor"].device == device
-                    padding = (align - (offset & (align - 1))) & (align - 1)
-                    offset += padding
-                    aligned_pos.append(offset)
-                    params.append(param)
-                    offset += int(np.prod(param["tensor"].shape))
-                    idx += 1
-
-                    if (
-                        offset * dtype_sz >= self._max_size_per_group * 1024 * 1024
-                        or idx >= self._max_nr_params_per_group
-                    ):
-                        break
-                group = group[idx:]
-                if idx == 1:
-                    # ignore param packs with only one item
-                    params[0]["tensor"].pack_group_key = group_key
-                    self._packed_params.append(params[0]["tensor"])
-                    self._grouped_params.append(
-                        [{"shape": params[0]["tensor"].shape, "id": params[0]["id"]}]
-                    )
-                    continue
-
-                packed_value = np.zeros((offset,), dtype=dtype)
-                for param, pos in zip(params, aligned_pos):
-                    val = param["tensor"].numpy()
-                    packed_value[pos : pos + val.size] = val.flatten()
-                new_param = Parameter(
-                    value=packed_value,
-                    device=device,
-                    dtype=dtype,
-                    requires_grad=requires_grad,
-                )
-                new_param.pack_group_key = group_key
-                self._packed_params.append(new_param)
-                self._grouped_params.append(
-                    [{"shape": i["tensor"].shape, "id": i["id"]} for i in params]
-                )
-
-    def forward(self, *args, **kwargs):
-        replace_param = dict()
-        for i in range(len(self._packed_params)):
-            packed_param = self._packed_params[i]
-            grouped_params = self._grouped_params[i]
-            if len(grouped_params) == 1:
-                continue
-            split = param_pack_split(
-                packed_param._symvar, [i["shape"] for i in grouped_params]
-            )
-            split = [
-                Parameter(Tensor(i, requires_grad=packed_param.requires_grad))
-                for i in split
-            ]
-            for j in range(len(split)):
-                replace_param[grouped_params[j]["id"]] = split[j]
-        self._model.replace_param(replace_param, 0)
-
-        return self._model.forward(*args, **kwargs)

imperative/python/megengine/module/quantized/conv.py

@@ -12,7 +12,7 @@ import numpy as np
 from ... import module as Float
 from ...core.tensor import dtype
 from ...functional import conv_bias_activation
-from ...tensor_nn import Parameter
+from ...tensor import Parameter
 from ..qat import conv as QAT
 from .module import QuantizedModule
 

imperative/python/megengine/module/quantized/conv_bn.py

@@ -5,7 +5,7 @@
 # Unless required by applicable law or agreed to in writing,
 # software distributed under the License is distributed on an
 # "AS IS" BASIS, WITHOUT ARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-from ...tensor_nn import Parameter
+from ...tensor import Parameter
 from ..qat import conv_bn as QAT
 from .conv import Conv2d
 

imperative/python/megengine/module/quantized/linear.py

@@ -9,7 +9,7 @@ import numpy as np
 
 from ... import functional as F
 from ...core.tensor import dtype
-from ...tensor_nn import Parameter
+from ...tensor import Parameter
 from ..qat import linear as QAT
 from .module import QuantizedModule
 

imperative/python/megengine/optimizer/adadelta.py

@@ -11,7 +11,7 @@ from typing import Iterable, Union
 import numpy as np
 
 from ..functional import sqrt
-from ..tensor_nn import Parameter
+from ..tensor import Parameter
 from .optimizer import Optimizer
 
 
@@ -63,7 +63,7 @@ class Adadelta(Optimizer):
 
         for param in param_group["params"]:
 
-            if not param.requires_grad or "grad" not in param.__dict__:
+            if param.grad is None:
                 continue
 
             states = self._state[param]

imperative/python/megengine/optimizer/adagrad.py

@@ -11,7 +11,7 @@ from typing import Iterable, Union
 import numpy as np
 
 from ..functional import sqrt
-from ..tensor_nn import Parameter
+from ..tensor import Parameter
 from .optimizer import Optimizer
 
 
@@ -62,7 +62,7 @@ class Adagrad(Optimizer):
 
         for param in param_group["params"]:
 
-            if not param.requires_grad or "grad" not in param.__dict__:
+            if param.grad is None:
                 continue
 
             states = self._state[param]

imperative/python/megengine/optimizer/adam.py

@@ -8,7 +8,7 @@
 # "AS IS" BASIS, WITHOUT ARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 from typing import Iterable, Tuple, Union
 
-from ..tensor_nn import Parameter
+from ..tensor import Parameter
 from .optimizer import Optimizer
 
 
@@ -59,7 +59,7 @@ class Adam(Optimizer):
 
         for param in param_group["params"]:
 
-            if not param.requires_grad or "grad" not in param.__dict__:
+            if param.grad is None:
                 continue
 
             grad = param.grad

imperative/python/megengine/optimizer/optimizer.py

@@ -7,7 +7,7 @@
 # software distributed under the License is distributed on an
 # "AS IS" BASIS, WITHOUT ARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 from abc import ABCMeta, abstractmethod
-from collections import Iterable
+from collections.abc import Iterable
 from contextlib import contextmanager
 from typing import Dict
 from typing import Iterable as Iter
@@ -15,8 +15,7 @@ from typing import Union
 
 import numpy as np
 
-from ..tensor import Tensor, TensorDict
-from ..tensor_nn import Buffer, Parameter
+from ..tensor import Parameter, Tensor
 
 
 class _RequiredParameter:
@@ -37,7 +36,7 @@ class Optimizer(metaclass=ABCMeta):
     def __init__(  # pylint: disable=too-many-branches
         self, params: Union[Iter[Parameter], dict], defaults: dict,
     ):
-        self._state = TensorDict()
+        self._state = dict()
         self._defaults = defaults
 
         if isinstance(params, (Parameter, dict)):
@@ -93,10 +92,6 @@ class Optimizer(metaclass=ABCMeta):
                     "optimizer can only optimize Parameters, but one of the params is "
                     + type(param)
                 )
-            if not param.requires_grad:
-                raise ValueError(
-                    "optimizer can only optimize Parameters with requires_grad=True"
-                )
 
         for name, default in self._defaults.items():
             if default is required and name not in param_group:
@@ -122,7 +117,7 @@ class Optimizer(metaclass=ABCMeta):
             initializer = np.zeros(param.shape, dtype=np.float32)
         state_dict = self._state.setdefault(param, {})
         assert state_name not in state_dict
-        state = Buffer(initializer)
+        state = Tensor(initializer)
         state_dict[state_name] = state
 
     @abstractmethod
@@ -140,7 +135,7 @@ class Optimizer(metaclass=ABCMeta):
                 params.append(param)
         return params
 
-    def step(self, clear_grad=False):
+    def step(self):
         r"""Performs a single optimization step.
 
         """
@@ -152,8 +147,7 @@ class Optimizer(metaclass=ABCMeta):
                     "Please use a list instead."
                 )
             self._updates(group)
-        if clear_grad:
-            self.clear_grad()
+        return self
 
     def clear_grad(self):
         r"""Clear the grad buffer.
@@ -161,8 +155,7 @@ class Optimizer(metaclass=ABCMeta):
         """
         for param_group in self.param_groups:
             for param in param_group["params"]:
-                if getattr(param, "grad", None) is not None:
-                    param.grad = None
+                param.grad = None
 
     def state_dict(self) -> Dict:
         r"""Export the optimizer state.
@@ -171,7 +164,7 @@ class Optimizer(metaclass=ABCMeta):
         """
         param_groups = []
         state = dict()
-        param2id = TensorDict()
+        param2id = dict()
 
         cur_id = 0
         for group in self.param_groups:
@@ -213,8 +206,9 @@ class Optimizer(metaclass=ABCMeta):
                 p = param_new
                 self._state[p] = state["state"][param_saved].copy()
                 for k, v in self._state[p].items():
-                    if isinstance(v, Buffer):
-                        self._state[p][k] = Buffer(v.numpy())
+                    if isinstance(v, Tensor):
+                        # TODO: maybe a more efficient way?
+                        self._state[p][k] = Tensor(v.numpy())
 
             if set(group_new.keys()) != set(group_saved.keys()):
                 raise ValueError(

imperative/python/megengine/optimizer/sgd.py

@@ -8,7 +8,7 @@
 # "AS IS" BASIS, WITHOUT ARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 from typing import Iterable, Union
 
-from ..tensor_nn import Parameter
+from ..tensor import Parameter
 from .optimizer import Optimizer
 
 
@@ -52,7 +52,7 @@ class SGD(Optimizer):
         momentum = param_group["momentum"]
 
         for param in param_group["params"]:
-            if not param.requires_grad or "grad" not in param.__dict__:
+            if param.grad is None:
                 continue
 
             grad = param.grad

imperative/python/megengine/quantization/fake_quant.py

@@ -14,8 +14,7 @@ from .. import functional as F
 from ..core.tensor.dtype import _metadata_dict, get_quantized_dtype
 from ..core.tensor.function import Function
 from ..module import Module
-from ..tensor import Tensor
-from ..tensor_nn import Parameter
+from ..tensor import Parameter, Tensor
 from .utils import QuantMode, fake_quant_tensor, get_qparam_dict
 
 

imperative/python/megengine/quantization/observer.py

@@ -13,7 +13,7 @@ import numpy as np
 from .. import functional as F
 from ..core.tensor.dtype import _metadata_dict, get_quantized_dtype
 from ..module import Module
-from ..tensor_nn import Buffer
+from ..tensor import Tensor
 from .utils import QuantMode, Round, get_qparam_dict
 
 
@@ -82,8 +82,8 @@ class MinMaxObserver(Observer):
     ):
         super().__init__(dtype, narrow_range)
         self.mode = mode
-        self.min_val = Buffer(np.finfo(np.float32).max, dtype=np.float32)
-        self.max_val = Buffer(np.finfo(np.float32).min, dtype=np.float32)
+        self.min_val = Tensor(np.finfo(np.float32).max, dtype=np.float32)
+        self.max_val = Tensor(np.finfo(np.float32).min, dtype=np.float32)
         self.scale_limit = eps
 
     def _calculate_qparams(self, inp_min_val, inp_max_val):
@@ -118,8 +118,8 @@ class MinMaxObserver(Observer):
             # stop gradient
             x = x_orig.detach()
             # find max and min
-            self.min_val.set_value(F.minimum(self.min_val, x.min()))
-            self.max_val.set_value(F.maximum(self.max_val, x.max()))
+            self.min_val._reset(F.minimum(self.min_val, x.min()))
+            self.max_val._reset(F.maximum(self.max_val, x.max()))
         return x_orig
 
 
@@ -133,22 +133,22 @@ class ExponentialMovingAverageObserver(MinMaxObserver):
         narrow_range: bool = False,
     ):
         super().__init__(mode, eps, dtype, narrow_range)
-        self.momentum = Buffer(momentum)
-        self.runtime_momentum = Buffer(0.0)
+        self.momentum = Tensor(momentum)
+        self.runtime_momentum = Tensor(0.0)
 
     def set_momentum(self, momentum):
-        self.momentum.set_value(momentum)
+        self.momentum._reset(momentum)
 
     def forward(self, x_orig):
         if self.enabled:
             # stop gradient
             x = x_orig.detach()
             # Exponential Moving Average
-            self.min_val.set_value(
+            self.min_val._reset(
                 self.min_val * self.runtime_momentum
                 + (1 - self.runtime_momentum) * x.min()
             )
-            self.max_val.set_value(
+            self.max_val._reset(
                 self.max_val * self.runtime_momentum
                 + (1 - self.runtime_momentum) * x.max()
             )
@@ -171,7 +171,7 @@ class HistogramObserver(MinMaxObserver):
         self.bins = bins
         self.upsample_rate = upsample_rate
         self.dst_nbins = _metadata_dict[dtype].qmax - _metadata_dict[dtype].qmin + 1
-        self.histogram = Buffer([-1] + [0.0] * (bins - 1))
+        self.histogram = Tensor([-1] + [0.0] * (bins - 1))
 
     def _non_linear_param_search(self):
         r"""Non-linear parameter search.
@@ -395,9 +395,9 @@ class HistogramObserver(MinMaxObserver):
                     self.bins,
                 )
 
-        self.histogram.set_value(new_histogram)
-        self.min_val.set_value(new_min)
-        self.max_val.set_value(new_max)
+        self.histogram._reset(new_histogram)
+        self.min_val._reset(new_min)
+        self.max_val._reset(new_max)
 
     def forward(self, x_orig):
         self.sideeffect_forward(x_orig)

imperative/python/megengine/tensor.py

@@ -14,10 +14,11 @@ from .core import Tensor as _Tensor
 from .core.ops.builtin import Copy
 from .core.tensor.core import apply
 from .device import get_default_device
+from .utils.deprecation import deprecated
 
 
 class Tensor(_Tensor):
-    requires_grad = False
+    grad = None
     dmap_callback = None
 
     def __init__(self, data, dtype=None, device=None):
@@ -26,15 +27,32 @@ class Tensor(_Tensor):
         self.q_dict = {"mode": None, "scale": None, "zero_point": None}
         super().__init__(data, dtype=dtype, device=device)
 
+    @deprecated(version="1.0", reason="no need to reuse an existing tensor since 1.0")
     def set_value(self, value):
         self._reset(value)
 
+    @deprecated(version="1.0", reason="use *= 0 instead")
     def reset_zero(self):
         self *= 0
 
     def to(self, cn):
         return apply(Copy(comp_node=cn), self)[0]
 
+    @property
+    def requires_grad(self):
+        raise AttributeError("requires_grad is reserved for future use")
+
+    @requires_grad.setter
+    def requires_grad(self, value):
+        raise AttributeError("requires_grad is reserved for future use")
+
+    @requires_grad.deleter
+    def requires_grad(self):
+        raise AttributeError("requires_grad is reserved for future use")
+
+    def __hash__(self):
+        return id(self)
+
     def __getstate__(self):
         r""" __getstate__ will be called for pickle serialization or deep copy
         """
@@ -73,53 +91,6 @@ class Tensor(_Tensor):
 tensor = Tensor
 
 
-class Dict(collections.MutableMapping):
-    def __init__(self, *args, key=None, **kwargs):
-        self.data = {}
-        if key:
-            self.keyfn = key
-        for i in args:
-            self.update(i)
-        self.update(**kwargs)
-
-    @staticmethod
-    def keyfn(key):  # pylint: disable=method-hidden
-        return key
-
-    def __getitem__(self, key):
-        _, v = self.data[self.keyfn(key)]
-        return v
-
-    def __setitem__(self, key, value):
-        self.data[self.keyfn(key)] = key, value
-
-    def __delitem__(self, key):
-        del self.data[self.keyfn(key)]
-
-    def __iter__(self):
-        for _, (k, _) in self.data.items():
-            yield k
-
-    def __len__(self):
-        return len(self.data)
-
-
-class TensorDict(Dict):  # pylint: disable=too-many-ancestors
-    class keyfn:
-        def __new__(cls, x: Tensor):
-            if not isinstance(x, Tensor):
-                return x
-            return super().__new__(cls)
-
-        def __init__(self, x: Tensor):
-            self._data = x  # do not save id directly to make pickle work
-
-        def __hash__(self):
-            return id(self._data)
-
-        def __eq__(self, other):
-            # pylint: disable=undefined-variable
-            return isinstance(other, __class__) and id(self._data) == id(other._data)
-
-    def __init__(self, *args):
-        super().__init__(*args)
+class Parameter(Tensor):
+    r"""A kind of Tensor that is to be considered a module parameter.
+    """

imperative/python/megengine/tensor_nn.py

@@ -1,20 +0,0 @@
-# -*- coding: utf-8 -*-
-# MegEngine is Licensed under the Apache License, Version 2.0 (the "License")
-#
-# Copyright (c) 2014-2020 Megvii Inc. All rights reserved.
-#
-# Unless required by applicable law or agreed to in writing,
-# software distributed under the License is distributed on an
-# "AS IS" BASIS, WITHOUT ARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-from . import Tensor, tensor
-
-
-class Buffer(Tensor):
-    r"""A kind of Tensor with ``requires_grad=False``.
-    """
-
-
-class Parameter(Tensor):
-    r"""A kind of Tensor that is to be considered a module parameter.
-    """
-    requires_grad = True

imperative/python/megengine/utils/deprecation.py

@@ -0,0 +1 @@
+from deprecated.sphinx import deprecated

imperative/python/megengine/utils/types.py

@@ -15,7 +15,7 @@ def get_ndtuple(value, *, n, allow_zero=True):
 
     :type allow_zero: bool
     :param allow_zero: whether to allow zero tuple value"""
-    if not isinstance(value, collections.Iterable):
+    if not isinstance(value, collections.abc.Iterable):
         value = int(value)
         value = tuple([value for i in range(n)])
     else:

imperative/python/requires.txt

@@ -5,3 +5,4 @@ requests
 tabulate
 tqdm
 redispy
+deprecated

imperative/python/test/integration/test_advance_indexing.py

@@ -38,7 +38,7 @@ class Simple2(Module):
 def test_advance_indexing():
     net = Simple()
 
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
     optim = optimizer.SGD(net.parameters(), lr=1.0)
     optim.clear_grad()
 
@@ -48,7 +48,7 @@ def test_advance_indexing():
     data = tensor(raw_data)
     mask = tensor(raw_mask)
     answer = 1.0 - raw_data[raw_mask].sum()
-    with gm.record():
+    with gm:
         loss = net(data, mask).sum()
         gm.backward(loss)
     optim.step()
@@ -58,7 +58,7 @@ def test_advance_indexing():
 def test_advance_indexing_with_subtensor():
     net = Simple2()
 
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
     optim = optimizer.SGD(net.parameters(), lr=1.0)
     optim.clear_grad()
 
@@ -66,7 +66,7 @@ def test_advance_indexing_with_subtensor():
     raw_data = np.arange(576).reshape(dshape).astype(np.float32)
     data = tensor(raw_data)
     answer = 1.0 - raw_data[1, ..., :, 0:4:2, 0:2].sum()
-    with gm.record():
+    with gm:
         loss = net(data).sum()
         gm.backward(loss)
     optim.step()

imperative/python/test/integration/test_ai.py

@@ -28,13 +28,13 @@ class Simple(Module):
 def test_ai():
     net = Simple()
 
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
     optim = optimizer.SGD(net.parameters(), lr=1.0)
     optim.clear_grad()
 
     dshape = (10, 10)
     data = tensor(np.ones(dshape).astype(np.float32))
-    with gm.record():
+    with gm:
         loss = net(data).sum()
         gm.backward(loss)
     optim.step()

imperative/python/test/integration/test_bn.py

@@ -25,12 +25,12 @@ def test_frozen_bn():
     saved_wt = m.weight.numpy()
     saved_bias = m.bias.numpy()
 
-    gm = ad.GradManager().register(m.parameters())
+    gm = ad.GradManager().attach(m.parameters())
     optim = optimizer.SGD(m.parameters(), lr=1.0)
     optim.clear_grad()
 
     data = np.random.random((6, nchannel, 2, 2)).astype("float32")
-    with gm.record():
+    with gm:
         loss = m(data).mean()
         gm.backward(loss)
     optim.step()
@@ -46,12 +46,12 @@ def test_bn_no_track_stat():
     nchannel = 3
     m = BatchNorm2d(nchannel, track_running_stats=False)
 
-    gm = ad.GradManager().register(m.parameters())
+    gm = ad.GradManager().attach(m.parameters())
     optim = optimizer.SGD(m.parameters(), lr=1.0)
     optim.clear_grad()
 
     data = np.random.random((6, nchannel, 2, 2)).astype("float32")
-    with gm.record():
+    with gm:
         loss = m(data).sum()
         gm.backward(loss)
     optim.step()
@@ -68,12 +68,12 @@ def test_bn_no_track_stat2():
     saved_mean = m.running_mean.numpy()
     assert saved_mean is not None
 
-    gm = ad.GradManager().register(m.parameters())
+    gm = ad.GradManager().attach(m.parameters())
     optim = optimizer.SGD(m.parameters(), lr=1.0)
     optim.clear_grad()
 
     data = np.random.random((6, nchannel, 2, 2)).astype("float32")
-    with gm.record():
+    with gm:
         loss = m(data).sum()
         gm.backward(loss)
     optim.step()

imperative/python/test/integration/test_converge.py

@@ -74,13 +74,11 @@ class XORNet(Module):
 
 def test_training_converge():
     net = XORNet()
-    opt = SGD(
-        net.parameters(requires_grad=True), lr=0.01, momentum=0.9, weight_decay=5e-4
-    )
-    gm = ad.GradManager().register(net.parameters())
+    opt = SGD(net.parameters(), lr=0.01, momentum=0.9, weight_decay=5e-4)
+    gm = ad.GradManager().attach(net.parameters())
 
     def train(data, label):
-        with gm.record():
+        with gm:
             pred = net(data)
             loss = F.cross_entropy_with_softmax(pred, label)
             gm.backward(loss)

imperative/python/test/integration/test_correctness.py

@@ -91,7 +91,7 @@ class MnistNet(Module):
 
 
 def train(data, label, net, opt, gm):
-    with gm.record():
+    with gm:
         pred = net(data)
         loss = F.cross_entropy_with_softmax(pred, label)
         gm.backward(loss)
@@ -117,7 +117,7 @@ def update_model(model_path):
     net.load_state_dict(checkpoint["net_init"])
     lr = checkpoint["sgd_lr"]
     opt = SGD(net.parameters(), lr=lr)
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
 
     data = Tensor(checkpoint["data"], dtype=np.float32)
     label = Tensor(checkpoint["label"], dtype=np.int32)
@@ -152,7 +152,7 @@ def run_train(
     net.load_state_dict(checkpoint["net_init"])
     lr = checkpoint["sgd_lr"]
     opt = SGD(net.parameters(), lr=lr)
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
 
     data = Tensor(checkpoint["data"], dtype=np.float32)
     label = Tensor(checkpoint["label"], dtype=np.int32)

imperative/python/test/integration/test_detach.py

@@ -32,11 +32,11 @@ def test_detach():
 
     optim = optimizer.SGD(net.parameters(), lr=1.0)
     optim.clear_grad()
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
 
     dshape = (10, 10)
     data = tensor(np.ones(dshape).astype(np.float32))
-    with gm.record():
+    with gm:
         loss = net(data).sum()
         gm.backward(loss)
     optim.step()

imperative/python/test/integration/test_dp_correctness.py

@@ -97,7 +97,7 @@ class MnistNet(Module):
 
 def train(data, label, net, opt, gm):
     opt.clear_grad()
-    with gm.record():
+    with gm:
         pred = net(data)
         loss = F.cross_entropy_with_softmax(pred, label)
         gm.backward(loss)
@@ -125,8 +125,7 @@ def update_model(model_path):
     lr = checkpoint["sgd_lr"]
     opt = SGD(net.parameters(), lr=lr)
 
-    gm = ad.GradManager()
-    gm.register(
+    gm = ad.GradManager().attach(
         net.parameters(), callbacks=[dist.make_allreduce_cb("MEAN", dist.WORLD)]
     )
 
@@ -171,8 +170,7 @@ def run_test(
         lr = checkpoint["sgd_lr"]
         opt = SGD(net.parameters(), lr=lr)
 
-        gm = ad.GradManager()
-        gm.register(
+        gm = ad.GradManager().attach(
             net.parameters(), callbacks=[dist.make_allreduce_cb("MEAN", dist.WORLD)]
         )
 

imperative/python/test/integration/test_hello_world.py

@@ -33,10 +33,10 @@ def test_hello_world():
 
     optim = optimizer.SGD(net.parameters(), lr=1.0)
     optim.clear_grad()
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
 
     data = tensor([2.34])
-    with gm.record():
+    with gm:
         loss = net(data)
         gm.backward(loss)
     optim.step()

imperative/python/test/integration/test_optimizer.py

@@ -13,7 +13,7 @@ import megengine.functional as F
 from megengine import Parameter, optimizer
 from megengine.jit import trace
 from megengine.module import Linear, Module
-from megengine.tensor import TensorDict, tensor
+from megengine.tensor import tensor
 
 
 class MLP(Module):
@@ -44,7 +44,7 @@ def _test_optimizer(opt_str, test_case, check_class, update_lr=False):
     net = Simple()
     opt = getattr(optimizer, opt_str)(net.parameters(), **test_case)
     check_func = check_class(net, **test_case)
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
 
     step = 0
     data_shape = (2, 28)
@@ -57,12 +57,12 @@ def _test_optimizer(opt_str, test_case, check_class, update_lr=False):
         data = tensor(np.random.random(data_shape).astype(np.float32))
 
         opt.clear_grad()
-        with gm.record():
+        with gm:
             pred = net(data)
             loss = pred.sum()
             gm.backward(loss)
 
-        ori_params = TensorDict()
+        ori_params = {}
         for param in net.parameters():
             ori_params[param] = np.copy(param.numpy())
         opt.step()
@@ -75,7 +75,7 @@ def _test_optimizer(opt_str, test_case, check_class, update_lr=False):
         @trace(symbolic=symbolic)
         def train_func(data, *, opt=None, gm=None):
             opt.clear_grad()
-            with gm.record():
+            with gm:
                 pred = net(data)
                 loss = pred.sum()
                 gm.backward(loss)
@@ -84,7 +84,7 @@ def _test_optimizer(opt_str, test_case, check_class, update_lr=False):
         # reset net and opt
         net = Simple()
         opt = getattr(optimizer, opt_str)(net.parameters(), **test_case)
-        gm = ad.GradManager().register(net.parameters())
+        gm = ad.GradManager().attach(net.parameters())
         check_func = check_class(net, **test_case)
         step = 0
         for i in range(iter_num):
@@ -93,7 +93,7 @@ def _test_optimizer(opt_str, test_case, check_class, update_lr=False):
                     group["lr"] += 0.01
                 check_func.lr += 0.01
 
-            ori_params = TensorDict()
+            ori_params = {}
             for param in net.parameters():
                 ori_params[param] = np.copy(param.numpy())
 
@@ -105,7 +105,7 @@ def _test_optimizer(opt_str, test_case, check_class, update_lr=False):
 def test_sgd():
     class CheckValue:
         def __init__(self, net, **kwarg):
-            self.slots = TensorDict()
+            self.slots = {}
             for param in net.parameters():
                 self.slots[param] = np.zeros(param.shape).astype(np.float32)
             for k, v in kwarg.items():
@@ -134,8 +134,8 @@ def test_sgd():
 def test_adam():
     class CheckValue:
         def __init__(self, net, **kwarg):
-            self.m_slots = TensorDict()
-            self.v_slots = TensorDict()
+            self.m_slots = {}
+            self.v_slots = {}
             for param in net.parameters():
                 self.m_slots[param] = np.zeros(param.shape).astype(np.float32)
                 self.v_slots[param] = np.zeros(param.shape).astype(np.float32)
@@ -175,7 +175,7 @@ def test_adam():
 def test_adagrad():
     class CheckValue:
         def __init__(self, net, **kwarg):
-            self.s_slots = TensorDict()
+            self.s_slots = {}
             for param in net.parameters():
                 self.s_slots[param] = np.zeros(param.shape).astype(np.float32)
             for k, v in kwarg.items():
@@ -207,8 +207,8 @@ def test_adagrad():
 def test_adadelta():
     class CheckValue:
         def __init__(self, net, **kwarg):
-            self.s_slots = TensorDict()
-            self.a_slots = TensorDict()
+            self.s_slots = {}
+            self.a_slots = {}
             for param in net.parameters():
                 self.s_slots[param] = np.zeros(param.shape).astype(np.float32)
                 self.a_slots[param] = np.zeros(param.shape).astype(np.float32)

imperative/python/test/integration/test_save_load.py

@@ -23,11 +23,11 @@ def test_save_load():
 
     optim = optimizer.SGD(net.parameters(), lr=1.0, momentum=0.9)
     optim.clear_grad()
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
 
     data = tensor([2.34])
 
-    with gm.record():
+    with gm:
         loss = net(data)
         gm.backward(loss)
 
@@ -55,7 +55,7 @@ def test_save_load():
     optim.load_state_dict(checkpoint["opt_state"])
     print("load done")
 
-    with gm.record():
+    with gm:
         loss = net([1.23])
         gm.backward(loss)
 

imperative/python/test/integration/test_sgd_momentum.py

@@ -31,12 +31,12 @@ def test_sgd_momentum():
 
     optim = optimizer.SGD(net.parameters(), lr=1.0, momentum=0.9)
     optim.clear_grad()
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
 
     data = tensor([2.34])
 
     # do a step of train
-    with gm.record():
+    with gm:
         loss = net(data)
         gm.backward(loss)
     optim.step()
@@ -51,7 +51,7 @@ def test_sgd_momentum():
 
     # do a step of train
     optim.clear_grad()
-    with gm.record():
+    with gm:
         loss = net(data)
         gm.backward(loss)
     optim.step()
@@ -69,7 +69,7 @@ def test_sgd_momentum_trace():
         @trace(symbolic=symbolic)
         def train_func(data, *, model=None, optim=None, gm=None):
             optim.clear_grad()
-            with gm.record():
+            with gm:
                 loss = net(data)
                 gm.backward(loss)
             optim.step()
@@ -82,7 +82,7 @@ def test_sgd_momentum_trace():
 
         net = Simple()
         optim = optimizer.SGD(net.parameters(), lr=1.0, momentum=0.9)
-        gm = ad.GradManager().register(net.parameters())
+        gm = ad.GradManager().attach(net.parameters())
         data = tensor([2.34])
         train_func(data, model=net, optim=optim, gm=gm)
         np.testing.assert_almost_equal(

imperative/python/test/integration/test_trace_dump.py

@@ -61,15 +61,15 @@ class XORNet(M.Module):
 
 def test_xornet_trace_dump():
     net = XORNet()
-    opt = optim.SGD(net.parameters(requires_grad=True), lr=0.01, momentum=0.9)
-    gm = GradManager().register(net.parameters(requires_grad=True))
+    opt = optim.SGD(net.parameters(), lr=0.01, momentum=0.9)
+    gm = GradManager().attach(net.parameters())
     batch_size = 64
     train_dataset = minibatch_generator(batch_size)
     val_dataset = minibatch_generator(batch_size)
 
     @trace
     def train_fun(data, label):
-        with gm.record():
+        with gm:
             net.train()
             pred = net(data)
             loss = F.cross_entropy_with_softmax(pred, label)

imperative/python/test/unit/functional/test_functional.py

@@ -14,7 +14,7 @@ import pytest
 import megengine.core.ops.builtin as builtin
 import megengine.core.tensor.dtype as dtype
 import megengine.functional as F
-from megengine import Buffer, Parameter, is_cuda_available, tensor
+from megengine import Parameter, Tensor, is_cuda_available, tensor
 from megengine.core._trace_option import use_tensor_shape
 from megengine.core.autodiff.grad import Grad
 from megengine.core.tensor.utils import make_shape_tuple
@@ -330,7 +330,7 @@ def test_roi_pooling():
 def test_add_update():
     shape = (2, 3)
     v = np.random.random(shape).astype(np.float32)
-    b = Buffer(v)
+    b = Tensor(v)
 
     u = F.add_update(b, 1)
     assertTensorClose(u.numpy(), v + 1)
@@ -347,7 +347,7 @@ def test_add_update():
 
 def test_add_update_params():
     b = np.random.random((2, 3)).astype(np.float32)
-    y = Buffer(b)
+    y = Tensor(b)
 
     # @jit.trace
     def f(x):
@@ -355,7 +355,7 @@ def test_add_update_params():
 
     f(np.zeros((2, 3)).astype(np.float32))
 
-    z = Buffer(np.zeros((2, 3)).astype(np.float32))
+    z = Tensor(np.zeros((2, 3)).astype(np.float32))
     F.add_update(y, z, beta=0.1)
 
     res = f(np.ones((2, 3)).astype(np.float32))

imperative/python/test/unit/functional/test_tensor.py

@@ -12,7 +12,7 @@ import numpy as np
 import pytest
 
 import megengine.functional as F
-from megengine import Buffer, Parameter, is_cuda_available, tensor
+from megengine import tensor
 from megengine.core._trace_option import use_tensor_shape
 from megengine.core.tensor.utils import astensor1d
 from megengine.distributed.helper import get_device_count_by_fork

imperative/python/test/unit/module/test_batchnorm.py

@@ -14,10 +14,9 @@ import pytest
 
 import megengine as mge
 import megengine.distributed as dist
-from megengine import tensor
+from megengine import Tensor
 from megengine.core._trace_option import use_tensor_shape
 from megengine.module import BatchNorm1d, BatchNorm2d, SyncBatchNorm
-from megengine.tensor import Tensor
 from megengine.test import assertTensorClose
 
 
@@ -45,10 +44,8 @@ def test_syncbn():
             return
         dist.init_process_group("localhost", port, nr_ranks, rank, rank)
         bn = SyncBatchNorm(nr_chan, momentum=momentum, eps=eps)
-        data_tensor = tensor([])
         for i in range(steps):
-            data_tensor.set_value(data[i])
-            yv = bn(data_tensor)
+            yv = bn(Tensor(data[i]))
 
         assertTensorClose(yv_expect, yv.numpy(), max_err=5e-6)
         assertTensorClose(running_mean, bn.running_mean.numpy(), max_err=5e-6)
@@ -105,7 +102,6 @@ def test_batchnorm():
     bn = BatchNorm1d(nr_chan, momentum=momentum)
     running_mean = np.zeros((1, nr_chan, 1), dtype=np.float32)
     running_var = np.ones((1, nr_chan, 1), dtype=np.float32)
-    data = tensor([])
     for i in range(3):
         xv = np.random.normal(loc=2.3, size=data_shape).astype(np.float32)
         mean = np.mean(np.mean(xv, axis=0, keepdims=True), axis=2, keepdims=True)
@@ -120,8 +116,7 @@ def test_batchnorm():
         running_mean = running_mean * momentum + mean * (1 - momentum)
         running_var = running_var * momentum + var_unbiased * (1 - momentum)
 
-        data.set_value(xv)
-        yv = bn(data)
+        yv = bn(Tensor(xv))
         yv_expect = (xv - mean) / sd
 
         assertTensorClose(yv_expect, yv.numpy(), max_err=5e-6)
@@ -137,7 +132,7 @@ def test_batchnorm():
     var_backup = bn.running_var.numpy()
     bn.training = False
     xv = np.random.normal(loc=2.3, size=data_shape).astype(np.float32)
-    data.set_value(xv)
+    data = Tensor(xv)
     yv1 = bn(data)
     yv2 = bn(data)
     assertTensorClose(yv1.numpy(), yv2.numpy(), max_err=0)
@@ -161,7 +156,6 @@ def test_syncbn1d():
     bn = SyncBatchNorm(nr_chan, momentum=momentum)
     running_mean = np.zeros((1, nr_chan, 1), dtype=np.float32)
     running_var = np.ones((1, nr_chan, 1), dtype=np.float32)
-    data = tensor([])
     for i in range(3):
         xv = np.random.normal(loc=2.3, size=data_shape).astype(np.float32)
         mean = np.mean(np.mean(xv, axis=0, keepdims=True), axis=2, keepdims=True)
@@ -176,8 +170,7 @@ def test_syncbn1d():
         running_mean = running_mean * momentum + mean * (1 - momentum)
         running_var = running_var * momentum + var_unbiased * (1 - momentum)
 
-        data.set_value(xv)
-        yv = bn(data)
+        yv = bn(Tensor(xv))
         yv_expect = (xv - mean) / sd
 
         assertTensorClose(yv_expect, yv.numpy(), max_err=5e-6)
@@ -193,7 +186,7 @@ def test_syncbn1d():
     var_backup = bn.running_var.numpy()
     bn.training = False
     xv = np.random.normal(loc=2.3, size=data_shape).astype(np.float32)
-    data.set_value(xv)
+    data = Tensor(xv)
     yv1 = bn(data)
     yv2 = bn(data)
     assertTensorClose(yv1.numpy(), yv2.numpy(), max_err=0)
@@ -210,7 +203,6 @@ def test_batchnorm2d():
     bn = BatchNorm2d(nr_chan, momentum=momentum)
     running_mean = np.zeros((1, nr_chan, 1, 1), dtype=np.float32)
     running_var = np.ones((1, nr_chan, 1, 1), dtype=np.float32)
-    data = tensor([])
     for i in range(3):
         xv = np.random.normal(loc=2.3, size=data_shape).astype(np.float32)
         xv_transposed = np.transpose(xv, [0, 2, 3, 1]).reshape(
@@ -226,8 +218,7 @@ def test_batchnorm2d():
         running_mean = running_mean * momentum + mean * (1 - momentum)
         running_var = running_var * momentum + var_unbiased * (1 - momentum)
 
-        data.set_value(xv)
-        yv = bn(data)
+        yv = bn(Tensor(xv))
         yv_expect = (xv - mean) / sd
 
         assertTensorClose(yv_expect, yv.numpy(), max_err=5e-6)
@@ -239,7 +230,7 @@ def test_batchnorm2d():
     var_backup = bn.running_var.numpy()
     bn.training = False
     xv = np.random.normal(loc=2.3, size=data_shape).astype(np.float32)
-    data.set_value(xv)
+    data = Tensor(xv)
     yv1 = bn(data)
     yv2 = bn(data)
     assertTensorClose(yv1.numpy(), yv2.numpy(), max_err=0)
@@ -263,7 +254,6 @@ def test_syncbn2d():
     bn = SyncBatchNorm(nr_chan, momentum=momentum)
     running_mean = np.zeros((1, nr_chan, 1, 1), dtype=np.float32)
     running_var = np.ones((1, nr_chan, 1, 1), dtype=np.float32)
-    data = tensor([])
     for i in range(3):
         xv = np.random.normal(loc=2.3, size=data_shape).astype(np.float32)
         xv_transposed = np.transpose(xv, [0, 2, 3, 1]).reshape(
@@ -279,8 +269,7 @@ def test_syncbn2d():
         running_mean = running_mean * momentum + mean * (1 - momentum)
         running_var = running_var * momentum + var_unbiased * (1 - momentum)
 
-        data.set_value(xv)
-        yv = bn(data)
+        yv = bn(Tensor(xv))
         yv_expect = (xv - mean) / sd
 
         assertTensorClose(yv_expect, yv.numpy(), max_err=5e-6)
@@ -292,7 +281,7 @@ def test_syncbn2d():
     var_backup = bn.running_var.numpy()
     bn.training = False
     xv = np.random.normal(loc=2.3, size=data_shape).astype(np.float32)
-    data.set_value(xv)
+    data = Tensor(xv)
     yv1 = bn(data)
     yv2 = bn(data)
     assertTensorClose(yv1.numpy(), yv2.numpy(), max_err=0)
@@ -306,7 +295,6 @@ def test_batchnorm_no_stats():
     nr_chan = 8
     data_shape = (3, nr_chan, 4)
     bn = BatchNorm1d(8, track_running_stats=False)
-    data = tensor([])
     for i in range(4):
         if i == 2:
             bn.training = False
@@ -320,8 +308,7 @@ def test_batchnorm_no_stats():
         ).reshape((1, nr_chan, 1))
         sd = np.sqrt(var + bn.eps)
 
-        data.set_value(xv)
-        yv = bn(data)
+        yv = bn(Tensor(xv))
         yv_expect = (xv - mean) / sd
 
         assertTensorClose(yv_expect, yv.numpy(), max_err=5e-6)
@@ -338,7 +325,6 @@ def test_syncbn_no_stats():
     nr_chan = 8
     data_shape = (3, nr_chan, 4)
     bn = SyncBatchNorm(8, track_running_stats=False)
-    data = tensor([])
     for i in range(4):
         if i == 2:
             bn.training = False
@@ -352,8 +338,7 @@ def test_syncbn_no_stats():
         ).reshape((1, nr_chan, 1))
         sd = np.sqrt(var + bn.eps)
 
-        data.set_value(xv)
-        yv = bn(data)
+        yv = bn(Tensor(xv))
         yv_expect = (xv - mean) / sd
 
         assertTensorClose(yv_expect, yv.numpy(), max_err=5e-6)
@@ -363,7 +348,6 @@ def test_batchnorm2d_no_stats():
     nr_chan = 8
     data_shape = (3, nr_chan, 16, 16)
     bn = BatchNorm2d(8, track_running_stats=False)
-    data = tensor([])
     for i in range(4):
         if i == 2:
             bn.training = False
@@ -376,8 +360,7 @@ def test_batchnorm2d_no_stats():
         var = np.var(xv_transposed, axis=0).reshape((1, nr_chan, 1, 1))
         sd = np.sqrt(var + bn.eps)
 
-        data.set_value(xv)
-        yv = bn(data)
+        yv = bn(Tensor(xv))
         yv_expect = (xv - mean) / sd
 
         assertTensorClose(yv_expect, yv.numpy(), max_err=5e-6)
@@ -394,7 +377,6 @@ def test_syncbn2d_no_stats():
     nr_chan = 8
     data_shape = (3, nr_chan, 16, 16)
     bn = SyncBatchNorm(8, track_running_stats=False)
-    data = tensor([])
     for i in range(4):
         if i == 2:
             bn.training = False
@@ -407,8 +389,7 @@ def test_syncbn2d_no_stats():
         var = np.var(xv_transposed, axis=0).reshape((1, nr_chan, 1, 1))
         sd = np.sqrt(var + bn.eps)
 
-        data.set_value(xv)
-        yv = bn(data)
+        yv = bn(Tensor(xv))
         yv_expect = (xv - mean) / sd
 
         assertTensorClose(yv_expect, yv.numpy(), max_err=5e-6)

imperative/python/test/unit/module/test_external.py

@@ -12,7 +12,7 @@ import numpy as np
 import pytest
 
 import megengine as mge
-from megengine import tensor
+from megengine import Tensor
 from megengine.module import Module
 
 
@@ -35,12 +35,12 @@ def test_cambricon_module():
     with open(model, "rb") as f:
         data = f.read()
         m = MyModule(data)
-        inputs = []
-        inputs.append(tensor(data=[], dtype=np.float16, device="cambricon0"))
-        inputs[0].set_value(np.random.normal(size=(1, 64, 32, 32)).astype(np.float16))
+        inp = Tensor(
+            np.random.normal((1, 64, 32, 32)).astype(np.float16), device="cambricon0"
+        )
 
         def inference(inps):
             pred = m(inps)
             return pred
 
-        pred = inference(inputs)
+        pred = inference([inp])

imperative/python/test/unit/module/test_module.py

@@ -16,7 +16,7 @@ import pytest
 
 import megengine as mge
 import megengine.functional as F
-from megengine import Buffer, Parameter, Tensor, tensor
+from megengine import Parameter, Tensor, tensor
 from megengine.module import (
     BatchNorm1d,
     BatchNorm2d,
@@ -196,7 +196,7 @@ class MyModule(Module):
         self.i = self.InnerModule()
         self.bn = BatchNorm2d(4)
         self.param = Parameter(np.ones(1, dtype=np.float32))
-        self.buff = Buffer(np.ones(1, dtype=np.float32))
+        self.buff = Tensor(np.ones(1, dtype=np.float32))
 
     def forward(self, x):
         x = self.i(x)
@@ -464,8 +464,7 @@ def test_sequential_named_children():
 
 def test_state_dict():
     data_shape = (2, 28)
-    data = tensor([])
-    data.set_value(np.random.random(data_shape))
+    data = tensor(np.random.random(data_shape))
     mlp = MLP()
     pred0 = mlp(data)
 
@@ -542,8 +541,7 @@ def test_shared_param():
 
 def test_pickle_module():
     data_shape = (2, 28)
-    data = tensor([])
-    data.set_value(np.random.random(data_shape))
+    data = tensor(np.random.random(data_shape))
     mlp = MLP()
     # pickle before forward
     with BytesIO() as fout:
@@ -568,8 +566,7 @@ def test_pickle_module():
 @pytest.mark.skip(reason="under development")
 def test_dump_model():
     data_shape = (2, 28)
-    data = tensor([])
-    data.set_value(np.random.random(data_shape))
+    data = Tensor(np.random.random(data_shape))
     mlp = MLP()
     pred = mlp(data)
     f = tempfile.NamedTemporaryFile(delete=False)

imperative/python/test/unit/module/test_tensor.py

@@ -13,7 +13,7 @@ import pytest
 
 import megengine as mge
 import megengine.functional as F
-from megengine import Buffer, Parameter
+from megengine import Parameter, Tensor
 from megengine.module import Conv2d
 from megengine.test import assertTensorClose
 
@@ -33,7 +33,7 @@ def test_set_value():
 
 @pytest.mark.skip(reason="fill unsupported")
 def test_fill():
-    a = Buffer(np.zeros((2, 3), dtype=np.float32))
+    a = Tensor(np.zeros((2, 3), dtype=np.float32))
     a.fill(3)
     assertTensorClose(a.numpy(), np.full((2, 3), 3, dtype=np.float32))
     a.fill(124.568)
@@ -80,7 +80,7 @@ def test_fill():
 # def test_shape_warning():
 #     with Graph() as cg:
 #         cg.set_option("eager_evaluation", False)
-#         b = Buffer(np.ones((2, 3)).astype(np.float32))
+#         b = Tensor(np.ones((2, 3)).astype(np.float32))
 #         with pytest.warns(None) as record:
 #             print(b.shape)
 #         if len(record) != 0:

imperative/python/test/unit/test_function.py

@@ -42,11 +42,11 @@ def test_single_input():
             return x
 
     net = Simple(av)
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
     opt = optimizer.SGD(net.parameters(), lr=1.0)
 
     opt.clear_grad()
-    with gm.record():
+    with gm:
         loss = net()
         gm.backward(loss.sum())
     opt.step()
@@ -81,11 +81,11 @@ def test_multi_input():
             return x
 
     net = Simple(av, bv)
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
     opt = optimizer.SGD(net.parameters(), lr=1.0)
 
     opt.clear_grad()
-    with gm.record():
+    with gm:
         loss = net()
         gm.backward(loss.sum())
     opt.step()
@@ -121,11 +121,11 @@ def test_multi_output():
             return x + y
 
     net = Simple(av, bv)
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
     opt = optimizer.SGD(net.parameters(), lr=1.0)
 
     opt.clear_grad()
-    with gm.record():
+    with gm:
         loss = net()
         gm.backward(loss.sum())
     opt.step()
@@ -163,9 +163,9 @@ def test_skip_invalid_grad():
 
     net = Simple(av, bv)
     optim = optimizer.SGD(net.parameters(), lr=1.0)
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
     optim.clear_grad()
-    with gm.record():
+    with gm:
         loss = net().sum()
         gm.backward(loss)
     optim.step()
@@ -198,10 +198,10 @@ def test_ste():
     av = np.random.random(data_shape).astype(np.float32)
     net = Simple(av)
     optim = optimizer.SGD(net.parameters(), lr=1.0)
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
     optim.clear_grad()
 
-    with gm.record():
+    with gm:
         loss = net()
         gm.backward(loss.sum())
     optim.step()
@@ -256,9 +256,9 @@ def test_none_in_out_grad():
     b = tensor(np.array([2.0], dtype=np.float32))
     net = Simple(a, b)
     optim = optimizer.SGD(net.parameters(), lr=1.0)
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
     optim.clear_grad()
-    with gm.record():
+    with gm:
         loss, _ = net()
         gm.backward(loss)
     optim.step()
@@ -293,10 +293,10 @@ def test_zero_grad():
     a = tensor(np.array([1.0], dtype=np.float32))
     net = Simple(a)
     optim = optimizer.SGD(net.parameters(), lr=1.0)
-    gm = ad.GradManager().register(net.parameters())
+    gm = ad.GradManager().attach(net.parameters())
     optim.clear_grad()
 
-    with gm.record():
+    with gm:
         loss = net()
         gm.backward(loss.sum())
     optim.step()

imperative/python/test/unit/test_indexing_op.py

@@ -38,7 +38,7 @@ def cvt_to_shape_desc(val, inpvar, config=None):
     if isinstance(val, RawTensor):
         return as_tensor(val, device)
 
-    if not isinstance(val, collections.Iterable):
+    if not isinstance(val, collections.abc.Iterable):
         val = [val]
 
     components = []

imperative/python/test/unit/test_serialization.py

@@ -12,19 +12,18 @@ from tempfile import TemporaryFile
 import numpy as np
 
 import megengine as mge
-from megengine import Buffer, Parameter, tensor
+from megengine import Parameter, Tensor
 
 
 def test_tensor_serialization():
     def tensor_eq(a, b):
         assert a.dtype == b.dtype
         assert a.device == b.device
-        assert a.requires_grad == b.requires_grad
         np.testing.assert_equal(a.numpy(), b.numpy())
 
     with TemporaryFile() as f:
         data = np.random.randint(low=0, high=7, size=[233])
-        a = tensor(data, device="xpux", dtype=np.int32)
+        a = Tensor(data, device="xpux", dtype=np.int32)
         pickle.dump(a, f)
         f.seek(0)
         b = pickle.load(f)
@@ -39,19 +38,19 @@ def test_tensor_serialization():
         np.testing.assert_equal(a.numpy(), b.numpy())
 
     with TemporaryFile() as f:
-        a = Buffer(np.random.random(size=(2, 233)).astype(np.float32))
+        a = Tensor(np.random.random(size=(2, 233)).astype(np.float32))
         pickle.dump(a, f)
         f.seek(0)
         b = pickle.load(f)
-        assert isinstance(b, Buffer)
+        assert type(b) is Tensor
         np.testing.assert_equal(a.numpy(), b.numpy())
 
     with TemporaryFile() as f:
-        a = Buffer(np.random.random(size=(2, 233)).astype(np.float32))
+        a = Tensor(np.random.random(size=(2, 233)).astype(np.float32))
         mge.save(a, f)
         f.seek(0)
         b = mge.load(f, map_location="cpux")
-        assert isinstance(b, Buffer)
+        assert type(b) is Tensor
         assert "cpu" in str(b.device)
         np.testing.assert_equal(a.numpy(), b.numpy())
 
@@ -59,12 +58,12 @@ def test_tensor_serialization():
         if mge.is_cuda_available():
             device_org = mge.get_default_device()
             mge.set_default_device("gpu0")
-            a = Buffer(np.random.random(size=(2, 233)).astype(np.float32))
+            a = Tensor(np.random.random(size=(2, 233)).astype(np.float32))
             mge.save(a, f)
             f.seek(0)
             mge.set_default_device("cpux")
             b = mge.load(f, map_location={"gpu0": "cpu0"})
-            assert isinstance(b, Buffer)
+            assert type(b) is Tensor
             assert "cpu0" in str(b.device)
             np.testing.assert_equal(a.numpy(), b.numpy())
             mge.set_default_device(device_org)

sdk/xor-deploy/xornet.py

@@ -66,7 +66,7 @@ def main():
         mge.set_default_device("cpux")
 
     net = XORNet()
-    opt = optim.SGD(net.parameters(requires_grad=True), lr=0.01, momentum=0.9)
+    opt = optim.SGD(net.parameters(), lr=0.01, momentum=0.9)
     batch_size = 64
     train_dataset = minibatch_generator(batch_size)
     val_dataset = minibatch_generator(batch_size)