一旦我們選擇了一個架構(gòu)并設(shè)置了我們的超參數(shù),我們就進入訓(xùn)練循環(huán),我們的目標是找到最小化損失函數(shù)的參數(shù)值。訓(xùn)練后,我們將需要這些參數(shù)來進行未來的預(yù)測。此外,我們有時會希望提取參數(shù)以在其他上下文中重用它們,將我們的模型保存到磁盤以便它可以在其他軟件中執(zhí)行,或者進行檢查以期獲得科學理解。
大多數(shù)時候,我們將能夠忽略參數(shù)聲明和操作的具體細節(jié),依靠深度學習框架來完成繁重的工作。然而,當我們遠離具有標準層的堆疊架構(gòu)時,我們有時需要陷入聲明和操作參數(shù)的困境。在本節(jié)中,我們將介紹以下內(nèi)容:
-
訪問用于調(diào)試、診斷和可視化的參數(shù)。
-
跨不同模型組件共享參數(shù)。
import tensorflow as tf
我們首先關(guān)注具有一個隱藏層的 MLP。
torch.Size([2, 1])
net = nn.Sequential()
net.add(nn.Dense(8, activation='relu'))
net.add(nn.Dense(1))
net.initialize() # Use the default initialization method
X = np.random.uniform(size=(2, 4))
net(X).shape
(2, 1)
net = nn.Sequential([nn.Dense(8), nn.relu, nn.Dense(1)])
X = jax.random.uniform(d2l.get_key(), (2, 4))
params = net.init(d2l.get_key(), X)
net.apply(params, X).shape
(2, 1)
6.2.1. 參數(shù)訪問
讓我們從如何從您已知的模型中訪問參數(shù)開始。
當通過類定義模型時Sequential
,我們可以首先通過索引模型來訪問任何層,就好像它是一個列表一樣。每個層的參數(shù)都方便地位于其屬性中。
When a model is defined via the Sequential
class, we can first access any layer by indexing into the model as though it were a list. Each layer’s parameters are conveniently located in its attribute.
Flax and JAX decouple the model and the parameters as you might have observed in the models defined previously. When a model is defined via the Sequential
class, we first need to initialize the network to generate the parameters dictionary. We can access any layer’s parameters through the keys of this dictionary.
When a model is defined via the Sequential
class, we can first access any layer by indexing into the model as though it were a list. Each layer’s parameters are conveniently located in its attribute.
我們可以如下檢查第二個全連接層的參數(shù)。
OrderedDict([('weight',
tensor([[-0.2523, 0.2104, 0.2189, -0.0395, -0.0590, 0.3360, -0.0205, -0.1507]])),
('bias', tensor([0.0694]))])
dense1_ (
Parameter dense1_weight (shape=(1, 8), dtype=float32)
Parameter dense1_bias (shape=(1,), dtype=float32)
)
FrozenDict({
kernel: Array([[-0.20739523],
[ 0.16546965],
[-0.03713543],
[-0.04860032],
[-0.2102929 ],
[ 0.163712 ],
[ 0.27240783],
[-0.4046879 ]], dtype=float32),
bias: Array([0.], dtype=float32),
})
我們可以看到這個全連接層包含兩個參數(shù),分別對應(yīng)于該層的權(quán)重和偏差。
6.2.1.1. 目標參數(shù)
請注意,每個參數(shù)都表示為參數(shù)類的一個實例。要對參數(shù)做任何有用的事情,我們首先需要訪問基礎(chǔ)數(shù)值。做這件事有很多種方法。有些更簡單,有些則更通用。以下代碼從返回參數(shù)類實例的第二個神經(jīng)網(wǎng)絡(luò)層中提取偏差,并進一步訪問該參數(shù)的值。
(torch.nn.parameter.Parameter, tensor([0.0694]))
參數(shù)是復(fù)雜的對象,包含值、梯度和附加信息。這就是為什么我們需要顯式請求該值。
除了值之外,每個參數(shù)還允許我們訪問梯度。因為我們還沒有為這個網(wǎng)絡(luò)調(diào)用反向傳播,所以它處于初始狀態(tài)。
True
(mxnet.gluon.parameter.Parameter, array([0.]))
Parameters are complex objects, containing values, gradients, and additional information. That is why we need to request the value explicitly.
In addition to the value, each parameter also allows us to access the gradient. Because we have not invoked backpropagation for this network yet, it is in its initial state.
array([[0., 0., 0., 0., 0., 0., 0., 0.]])
(jaxlib.xla_extension.Array, Array([0.], dtype=float32))
Unlike the other frameworks, JAX does not keep a track of the gradients over the neural network parameters, instead the parameters and the network are decoupled. It allows the user to express their computation as a Python function, and use the grad
transformation for the same purpose.
(tensorflow.python.ops.resource_variable_ops.ResourceVariable,
<tf.Tensor: shape=(1,), dtype=float32, numpy=array([0.], dtype=float32)>)
評論
查看更多