なぜGradientTapeはTensorFlowで最も過小評価された機能なのか

コンテンツ概要 Automatic Differentiation and Gradients(自動差別とグラディエント) セットアップ コンピュータグレード グレードテープ Gradients with respect to a model モデル テープが何を見ているかを制御する 中間結果 パフォーマンスに関するノート 非スカラルターゲットのグラディエント フローコントロール Automatic Differentiation and Gradients(自動差別とグラディエント) 機械学習アルゴリズムの実装に役立ちます。 ニューラルネットワークのトレーニング 自動差別化 バックプロモーション このガイドでは、特に熱心な実行において、TensorFlowを使用してグレディエントを計算する方法を調べます。 セットアップ import numpy as np import matplotlib.pyplot as plt import tensorflow as tf 2024-08-15 01:30:07.003169: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:485] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered 2024-08-15 01:30:07.023862: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:8454] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered 2024-08-15 01:30:07.029954: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1452] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered コンピュータグレード 自動的に区別するには、TensorFlow は、どの操作がどの順序で発生するかを記憶する必要があります。 続いて、その間、 , TensorFlow は、グレディエントを計算するために逆順にこの操作リストを通過します。 前へ バックパス グレードテープ TensorFlowは、 API for automatic differentiation; that is, calculating the gradient of a calculation with respect to some inputs, usually s. TensorFlow は、コンテキスト内で実行された関連する操作を「レコード」します。 TensorFlow はそのテープを使用して、そのテープを使用して「記録された」計算のグレディエントを計算します。 . tf.GradientTape tf.Variable tf.GradientTape 逆モデル差異化 以下は簡単な例です。 x = tf.Variable(3.0) with tf.GradientTape() as tape: y = x**2 WARNING: All log messages before absl::InitializeLog() is called are written to STDERR I0000 00:00:1723685409.408818 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685409.412555 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685409.416343 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685409.420087 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685409.431667 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685409.435229 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685409.438777 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685409.442350 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685409.445712 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685409.449141 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685409.452491 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685409.456034 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.685265 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.687389 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.689411 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.691490 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.693542 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.695541 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.697441 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.699432 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.701351 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.703333 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.705229 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.707222 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.744994 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.747037 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.749507 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.751538 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.753500 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.755501 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.757421 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.759404 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.761363 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.763858 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.766199 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 I0000 00:00:1723685410.768560 20970 cuda_executor.cc:1015] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355 いくつかの操作を記録した後、使用 いくつかのターゲット(しばしば損失)のいくつかのソース(しばしばモデルの変数)のグレディエントを計算する: GradientTape.gradient(target, sources) # dy = 2x * dx dy_dx = tape.gradient(y, x) dy_dx.numpy() 6.0 上記の例では、スカラーを使用しますが、 どのテナントでも簡単に動作します: tf.GradientTape w = tf.Variable(tf.random.normal((3, 2)), name='w') b = tf.Variable(tf.zeros(2, dtype=tf.float32), name='b') x = [[1., 2., 3.]] with tf.GradientTape(persistent=True) as tape: y = x @ w + b loss = tf.reduce_mean(y**2) グレードを取得するために 両方の変数に関して、あなたは両方をソースとして送信することができます。 The tape is flexible about how sources are passed and will accept any nested combination of lists or dictionaries and return the gradient structured the same way. テープは、ソースがどのように転送されるかについて柔軟であり、リストや辞書の組み合わせを受け入れ、同じように構造化されたグレディエントを返します(参照)。 ( ) loss gradient tf.nest [dl_dw, dl_db] = tape.gradient(loss, [w, b]) 各源に関連するグラディエントは、源の形を有する: print(w.shape) print(dl_dw.shape) (3, 2) (3, 2) 次は再びグレディント計算であり、今回は変数の辞書を通過する。 my_vars = { 'w': w, 'b': b } grad = tape.gradient(loss, my_vars) grad['b'] Gradients with respect to a model モデル 集めるのが普通 入 A あるいはその分類の一つ( で、 )の為 そして . tf.Variables tf.Module layers.Layer keras.Model チェックポイント 輸出 ほとんどの場合、あなたはモデルのトレーニング可能な変数に関してグレディエントを計算したいでしょう。 変数を合計して、the Property, you can calculate these gradients in a few lines of code: このグレディエントをいくつかのコードラインで計算できます。 tf.Module Module.trainable_variables layer = tf.keras.layers.Dense(2, activation='relu') x = tf.constant([[1., 2., 3.]]) with tf.GradientTape() as tape: # Forward pass y = layer(x) loss = tf.reduce_mean(y**2) # Calculate gradients with respect to every trainable variable grad = tape.gradient(loss, layer.trainable_variables) for var, g in zip(layer.trainable_variables, grad): print(f'{var.name}, shape: {g.shape}') kernel, shape: (3, 2) bias, shape: (2,) テープが何を見ているかを制御する デフォルトの動作は、トレーニング可能なコンテンツにアクセスした後にすべての操作を記録することです。 その理由は、 tf.Variable テープは、前進パスでどの操作を記録するかを知っておく必要があります。 テープには中間出力への参照が含まれていますので、不要な操作を記録したくありません。 最も一般的な使用例は、モデルのすべてのトレーニング可能な変数に関して損失のグレディエントを計算することです。 たとえば、次の方法では、グレディエントを計算できないため、 既定で「見る」のではなく、 トレーニングできない: tf.Tensor tf.Variable # A trainable variable x0 = tf.Variable(3.0, name='x0') # Not trainable x1 = tf.Variable(3.0, name='x1', trainable=False) # Not a Variable: A variable + tensor returns a tensor. x2 = tf.Variable(2.0, name='x2') + 1.0 # Not a variable x3 = tf.constant(3.0, name='x3') with tf.GradientTape() as tape: y = (x0**2) + (x1**2) + (x2**2) grad = tape.gradient(y, [x0, x1, x2, x3]) for g in grad: print(g) tf.Tensor(6.0, shape=(), dtype=float32) None None None You can list the variables being watched by the tape using the 方法: GradientTape.watched_variables [var.name for var in tape.watched_variables()] ['x0:0'] ユーザーに何が監視されているか、何が監視されていないかをコントロールするハックを提供します。 tf.GradientTape A に基づくグラディエントを記録する あなたは呼ぶ必要がある : tf.Tensor GradientTape.watch(x) x = tf.constant(3.0) with tf.GradientTape() as tape: tape.watch(x) y = x**2 # dy = 2x * dx dy_dx = tape.gradient(y, x) print(dy_dx.numpy()) 6.0 逆に、すべてを見るというデフォルトの行動を無効にするには、 セット この計算では、2 つの変数を使用しますが、変数のうちの 1 つだけの変数を接続します。 tf.Variables watch_accessed_variables=False x0 = tf.Variable(0.0) x1 = tf.Variable(10.0) with tf.GradientTape(watch_accessed_variables=False) as tape: tape.watch(x1) y0 = tf.math.sin(x0) y1 = tf.nn.softplus(x1) y = y0 + y1 ys = tf.reduce_sum(y) 以来 呼ばれなかった , no gradient is calculated with respect to it: グレディエントは計算されていません。 GradientTape.watch x0 # dys/dx1 = exp(x1) / (1 + exp(x1)) = sigmoid(x1) grad = tape.gradient(ys, {'x0': x0, 'x1': x1}) print('dy/dx0:', grad['x0']) print('dy/dx1:', grad['x1'].numpy()) dy/dx0: None dy/dx1: 0.9999546 中間結果 出力のグレディエントを、内部で計算された中間値に関して要求することもできます。 コンテキスト tf.GradientTape x = tf.constant(3.0) with tf.GradientTape() as tape: tape.watch(x) y = x * x z = y * y # Use the tape to compute the gradient of z with respect to the # intermediate value y. # dz_dy = 2 * y and y = x ** 2 = 9 print(tape.gradient(z, y).numpy()) 18.0 デフォルトでは、A が保有する資源 解放されるとすぐに、 同じ計算で複数のグレディエントを計算するには、グレディエントテープを作成します。 これは、複数の呼び出しを可能にします。 リソースがリリースされる方法は、テープオブジェクトがゴミを収集したときです。 GradientTape GradientTape.gradient persistent=True gradient x = tf.constant([1, 3.0]) with tf.GradientTape(persistent=True) as tape: tape.watch(x) y = x * x z = y * y print(tape.gradient(z, x).numpy()) # [4.0, 108.0] (4 * x**3 at x = [1.0, 3.0]) print(tape.gradient(y, x).numpy()) # [2.0, 6.0] (2 * x at x = [1.0, 3.0]) [ 4. 108.] [2. 6.] del tape # Drop the reference to the tape パフォーマンスに関するノート There is a tiny overhead associated with doing operations inside a gradient tape context. For most eager execution this will not be a noticeable cost, but you should still use tape context around the areas only where it is required. Gradient tapes use memory to store intermediate results, including inputs and outputs, for use during the backwards pass. For efficiency, some ops (like ) don't need to keep their intermediate results and they are pruned during the forward pass. However, if you use on your tape, and your peak memory usage will be higher. ReLU persistent=True nothing is discarded 非スカラルターゲットのグラディエント A gradient is fundamentally an operation on a scalar. グラディントは基本的にスカラル上の操作です。 x = tf.Variable(2.0) with tf.GradientTape(persistent=True) as tape: y0 = x**2 y1 = 1 / x print(tape.gradient(y0, x).numpy()) print(tape.gradient(y1, x).numpy()) 4.0 -0.25 したがって、複数のターゲットのグレディントを要求する場合、各ソースの結果は次のとおりです。 The gradient of the sum of the targets, or equivalent 各ターゲットのグレディエントの合計。 x = tf.Variable(2.0) with tf.GradientTape() as tape: y0 = x**2 y1 = 1 / x print(tape.gradient({'y0': y0, 'y1': y1}, x).numpy()) 3.75 同様に、ターゲット(s)がスカラーでない場合は、合計のグレディエントが計算されます: x = tf.Variable(2.) with tf.GradientTape() as tape: y = x * [3., 4.] print(tape.gradient(y, x).numpy()) 7.0 これは、損失のコレクションの合計のグレディエント、または要素の損失計算の合計のグレディエントを取ることを容易にします。 各項目に別々のグレディエントが必要な場合は、参照してください。 . ジャコビアン いくつかのケースでは、Jacobianを省略することができます. For an element-wise calculation, the gradient of the sum gives the derivative of each element with respect to its input-element, since each element is independent: 要素型の計算のために、合計のグラディエントは、それぞれの要素が独立しているので、その入力要素に関連して各要素の由来を示します。 x = tf.linspace(-10.0, 10.0, 200+1) with tf.GradientTape() as tape: tape.watch(x) y = tf.nn.sigmoid(x) dy_dx = tape.gradient(y, x) plt.plot(x, y, label='y') plt.plot(x, dy_dx, label='dy/dx') plt.legend() _ = plt.xlabel('x') フローコントロール gradient tape が操作を実行する際に記録しているため、Python コントロール フローは自然に処理されます(例えば、 そして 発言) if while Here a different variable is used on each branch of a gradient は、使用された変数にのみ接続します: if x = tf.constant(1.0) v0 = tf.Variable(2.0) v1 = tf.Variable(2.0) with tf.GradientTape(persistent=True) as tape: tape.watch(x) if x > 0.0: result = v0 else: result = v1**2 dv0, dv1 = tape.gradient(result, [v0, v1]) print(dv0) print(dv1) tf.Tensor(1.0, shape=(), dtype=float32) None ただ、コントロール声明自体は区別できないので、グレディントベースの最適化者には見えないことを覚えておいてください。 価値によっては、 上記の例では、テープまたはレコード または グレードに関して 常に . x result = v0 result = v1**2 x None dx = tape.gradient(result, x) print(dx) None 最初はTensorFlowのウェブサイトに掲載され、この記事はここに新しいタイトルの下に表示され、CC BY 4.0でライセンスを受けています。 最初はTensorFlowのウェブサイトに掲載され、この記事はここに新しいタイトルの下に表示され、CC BY 4.0でライセンスを受けています。 TensorFlowについて TensorFlowについて