Batched entropy model model which implements Universal Quantization.

In contrast to the base class, which uses rounding for quantization, here "quantization" is performed additive uniform noise, which is implemented with Universal Quantization.

This is described in Sec. 3.2. in the paper

"Universally Quantized Neural Compression"
Eirikur Agustsson & Lucas Theis

prior A tfp.distributions.Distribution object. A density model fitting the marginal distribution of the bottleneck data with additive uniform noise, which is shared a priori between the sender and the receiver. For best results, the distribution should be flexible enough to have a unit-width uniform distribution as a special case, since this is the marginal distribution for bottleneck dimensions that are constant. The distribution parameters may not depend on data (they must be either variables or constants).
coding_rank Integer. Number of innermost dimensions considered a coding unit. Each coding unit is compressed to its own bit string, and the bits() method sums over each coding unit.
compression Boolean. If set to True, the range coding tables used by compress() and decompress() will be built on instantiation. This assumes eager mode (throws an error if in graph mode or inside a tf.function call). If set to False, these two methods will not be accessible.
laplace_tail_mass Float. If positive, will augment the prior with a laplace mixture for training stability.
expected_grads If True, will use analytical expected gradients during backpropagation w.r.t. additive uniform noise.
tail_mass Float. Approximate probability mass which is encoded using an Elias gamma code embedded into the range coder.
range_coder_precision Integer. Precision passed to the range coding op.
bottleneck_dtype tf.dtypes.DType. Data type of bottleneck tensor. Defaults to tf.keras.mixed_precision.global_policy().compute_dtype.
num_noise_levels Integer. The number of levels used to quantize the uniform noise.
stateless Boolean. If True, creates range coding tables as Tensors rather than Variables. This makes the entropy model stateless and allows it to be constructed within a tf.function body. If compression=False, then stateless=True is implied and the provided value is ignored.
decode_sanity_check Boolean. If True, an raises an error if the binary strings passed into decompress are not completely decoded.

bottleneck_dtype Data type of the bottleneck tensor.
cdf The CDFs used by range coding.
cdf_offset The CDF offsets used by range coding.
coding_rank Number of innermost dimensions considered a coding unit.
compression Whether this entropy model is prepared for compression.
expected_grads Whether to use analytical expected gradients during backpropagation.
laplace_tail_mass Whether to augment the prior with a Laplace mixture.
name Returns the name of this module as passed or determined in the ctor.

name_scope Returns a tf.name_scope instance for this class.
non_trainable_variables Sequence of non-trainable variables owned by this module and its submodules.
prior Prior distribution, used for deriving range coding tables.
prior_shape Batch shape of prior (dimensions which are not assumed i.i.d.).
prior_shape_tensor Batch shape of prior as a Tensor.
range_coder_precision Precision used in range coding op.
stateless Whether range coding tables are created as Tensors or Variables.
submodules Sequence of all sub-modules.

Submodules are modules which are properties of this module, or found as properties of modules which are properties of this module (and so on).

a = tf.Module()
b = tf.Module()
c = tf.Module()
a.b = b
b.c = c
list(a.submodules) == [b, c]
list(b.submodules) == [c]
list(c.submodules) == []

tail_mass Approximate probability mass which is range encoded with overflow.
trainable_variables Sequence of trainable variables owned by this module and its submodules.

variables Sequence of variables owned by this module and its submodules.



View source

Compresses a floating-point tensor.

Compresses the tensor to bit strings. bottleneck is first quantized as in quantize(), and then compressed using the probability tables in self.cdf derived from self.prior. The quantized tensor can later be recovered by calling decompress().

The innermost self.coding_rank dimensions are treated as one coding unit, i.e. are compressed into one string each. Any additional dimensions to the left are treated as batch dimensions.

bottleneck tf.Tensor containing the data to be compressed. Must have at least self.coding_rank dimensions, and the innermost dimensions must be broadcastable to self.prior_shape.

A tf.Tensor having the same shape as bottleneck without the self.coding_rank innermost dimensions, containing a string for each coding unit.


View source

Decompresses a tensor.

Reconstructs the quantized tensor from bit strings produced by compress(). It is necessary to provide a part of the output shape in broadcast_shape.

strings tf.Tensor containing the compressed bit strings.
broadcast_shape Iterable of ints. The part of the output tensor shape between the shape of strings on the left and self.prior_shape on the right. This must match the shape of the input to compress().

A tf.Tensor of shape strings.shape + broadcast_shape + self.prior_shape.


View source

Returns the configuration of the entropy model.

A JSON-serializable Python dict.

RuntimeError on attempting to call this method on an entropy model with compression=False or with stateless=True.


View source


View source


Decorator to automatically enter the module name scope.

class MyModule(tf.Module):
  def __call__(self, x):
    if not hasattr(self, 'w'):
      self.w = tf.Variable(tf.random.normal([x.shape[1], 3]))
    return tf.matmul(x, self.w)

Using the above module would produce tf.Variables and tf.Tensors whose names included the module name:

mod = MyModule()
mod(tf.ones([1, 2]))
<tf.Tensor: shape=(1, 3), dtype=float32, numpy=..., dtype=float32)>
<tf.Variable 'my_module/Variable:0' shape=(2, 3) dtype=float32,
numpy=..., dtype=float32)>

method The method to wrap.

The original method wrapped such that it enters the module's name scope.


View source

Perturbs a tensor with additive uniform noise and estimates bitcost.

bottleneck tf.Tensor containing a non-perturbed bottleneck. Must have at least self.coding_rank dimensions.
training Boolean. If False, computes the bitcost using discretized uniform noise. If True, estimates the differential entropy with uniform noise.

A tuple (bottleneck_perturbed, bits) where bottleneck_perturbed is bottleneck perturbed with nosie and bits is the bitcost of transmitting such a sample having the same shape as bottleneck without the self.coding_rank innermost dimensions.