如何修剪 tensorflow 层中的最高权重？ tfmot.sparsity.keras.prune_low_magnitude

Question

我想修剪 tf 层中的最高权重值。我正在考虑使用 tf.nn.top_k 但我不确定我将如何去做。

文档：https://www.tensorflow.org/model_optimization/api_docs/python/tfmot/sparsity/keras/prune_low_magnitude 代码：

pruning_params = {
    'pruning_schedule': PolynomialDecay(initial_sparsity=0.2,
        final_sparsity=0.8, begin_step=1000, end_step=2000),
    'block_size': (2, 3),
    'block_pooling_type': 'MAX'
}

model = keras.Sequential([
    layers.Dense(10, activation='relu', input_shape=(100,)),
    prune_low_magnitude(layers.Dense(2, activation='tanh'), **pruning_params)
])

Answer 1

假设 w 是你要修剪的层的权重矩阵，k 是应该修剪的权重的百分比，这应该可以解决问题：

# Convert k from percentage to integer representing the number of weights
k = tf.cast(tf.round(tf.size(w, out_type=tf.float32) * tf.constant(k)), dtype=tf.int32)
# Reshape flatten the weight matrix
w_reshaped = tf.reshape(w, [-1])
# Select the indices of the largest k weights
_, indices = tf.nn.top_k(w_reshaped, k, sorted=True, name=None)
# Set the elements matching the indices to 0
mask = tf.scatter_nd_update(tf.Variable(tf.ones_like(w_reshaped, dtype=tf.float32), name="mask", trainable=False), tf.reshape(indices, [-1, 1]), tf.zeros([k], tf.float32))
# Update the weight matrix w
w.assign(tf.reshape(w_reshaped * mask, tf.shape(w)))

这是基于此Github repo。请注意，在该项目中，我正在修剪最小的 k 权重。

如何修剪 tensorflow 层中的最高权重？ tfmot.sparsity.keras.prune_low_magnitude

How do I prune over the highest weights in tensorflow layer? tfmot.sparsity.keras.prune_low_magnitude

tensorflow

keras

tf.keras

pruning