Removed Sphinx config, added autogenerated API docs.

Author: Johannes Ballé

.gitignore

@@ -1,4 +1,3 @@
-/docs/_build
 .DS_Store
 *.pyc
 *.so

docs/Makefile

@@ -0,0 +1,46 @@
+<div itemscope itemtype="http://developers.google.com/ReferenceObject">
+<meta itemprop="name" content="tfc" />
+</div>
+
+# Module: tfc
+
+Data compression tools.
+
+## Classes
+
+[`class EntropyBottleneck`](./tfc/EntropyBottleneck.md): Entropy bottleneck layer.
+
+[`class GDN`](./tfc/GDN.md): Generalized divisive normalization layer.
+
+[`class IdentityInitializer`](./tfc/IdentityInitializer.md): Initialize to the identity kernel with the given shape.
+
+[`class NonnegativeParameterizer`](./tfc/NonnegativeParameterizer.md): Object encapsulating nonnegative parameterization as needed for GDN.
+
+[`class Parameterizer`](./tfc/Parameterizer.md): Parameterizer object (abstract base class).
+
+[`class RDFTParameterizer`](./tfc/RDFTParameterizer.md): Object encapsulating RDFT reparameterization.
+
+[`class SignalConv1D`](./tfc/SignalConv1D.md): 1D convolution layer.
+
+[`class SignalConv2D`](./tfc/SignalConv2D.md): 2D convolution layer.
+
+[`class SignalConv3D`](./tfc/SignalConv3D.md): 3D convolution layer.
+
+[`class StaticParameterizer`](./tfc/StaticParameterizer.md): A parameterization object that always returns a constant tensor.
+
+## Functions
+
+[`irdft_matrix(...)`](./tfc/irdft_matrix.md): Matrix for implementing kernel reparameterization with `tf.matmul`.
+
+[`lower_bound(...)`](./tfc/lower_bound.md): Same as `tf.maximum`, but with helpful gradient for `inputs < bound`.
+
+[`pmf_to_quantized_cdf(...)`](./tfc/pmf_to_quantized_cdf.md): Converts PMF to quantized CDF. This op uses floating-point operations
+
+[`range_decode(...)`](./tfc/range_decode.md): Decodes a range-coded `code` into an int32 tensor of shape `shape`.
+
+[`range_encode(...)`](./tfc/range_encode.md): Using the provided cumulative distribution functions (CDF) inside `cdf`, returns
+
+[`same_padding_for_kernel(...)`](./tfc/same_padding_for_kernel.md): Determine correct amount of padding for `same` convolution.
+
+[`upper_bound(...)`](./tfc/upper_bound.md): Same as `tf.minimum`, but with helpful gradient for `inputs > bound`.
+