Update the examples of earlyexit arguments which were not consistent with descriptions

Add missing image files [arghhhh.]

Add missing files :-(

- Moved the Language model and struct pruning tutorials from the Wiki to
the HTML documentation.  Love the ease of Wiki, but GitHub doesn't let
Google crawl these pages, and users can't open PRs on Wiki pages.

- Updated the pruning algorithms documentation

* Asymmetric post-training quantization (only symmetric supported so until now)
* Quantization aware training for range-based (min-max) symmetric and asymmetric quantization
* Per-channel quantization support in both training and post-training
* Added tests and examples
* Updated documentation

As @cattpku notes in issue #89, we should follow the
[PyTorch ONNX limitations](https://pytorch.org/docs/stable/onnx.html#supported-operators)
specification.  For image classifiers this change is only syntactical.

Top1: 75.52% (-0.63% from TorchVision dense ResNet50)
Total sparsity: 82.6%

In a previous commit, we chose to override the TorchVision ResNet
models with our own modified versions of ResNet because collecting
activation statistics after ReLU layers requires that each ReLU
instance is only used once in the graph.

However, overriding TorchVision models means that if someone wants
to directly access the TorchVision model dictionary, he/she won't
find the ResNet models in the dictionary keys.  This is not very
friendly and we should not break imported modules.

This commit overrides the TorchVision ResNet models when accessed
via the distiller.models.create_model API, w/o changing the
TorchVision dictionary.

Replace "concat" layers with "element-wise addition" layers,
to match the text.

This commit contains the main fix for issue #85.  It contains a couple of changes to the YAML structure pruning API, with examples.
I urge you to read the documentation in the Wiki (https://github.com/NervanaSystems/distiller/wiki/Pruning-Filters-&-Channels).

New syntax for defining Structured AGP.  I tried to make the syntax similar to fine-grained
(i.e. element-wise) pruning.  All you need to do is add: ```group_type: Filters```.
```
  low_pruner:
    class: L1RankedStructureParameterPruner_AGP
    initial_sparsity : 0.10
    final_sparsity: 0.50
    group_type: Filters
    weights: [module.layer3.0.conv2.weight,
              module.layer3.0.downsample.0.weight,
              module.layer3.1.conv2.weight,
              module.layer3.2.conv2.weight]
```

If you want to define “leader-based” pruning dependencies, add ```group_dependency: Leader```:
```
  low_pruner:
    class: L1RankedStructureParameterPruner_AGP
    initial_sparsity : 0.10
    final_sparsity: 0.50
    group_type: Filters
    group_dependency: Leader
    weights: [module.layer3.0.conv2.weight,
              module.layer3.0.downsample.0.weight,
              module.layer3.1.conv2.weight,
              module.layer3.2.conv2.weight]
```

Retired the old ```reg_regims``` API for describing one-shot structured-pruning.

The new YAML API is very similar to AGP structured-pruning, which is much better
than before.
The new API also allows us to describe data-dependencies when doing one-shot
structure pruning, just like AGP structured-pruning.

This commit also includes further code refactoring.

Old API:
```
  filter_pruner:
     class: 'L1RankedStructureParameterPruner'
     reg_regims:
       'module.layer1.0.conv1.weight': [0.6, '3D']
       'module.layer1.1.conv1.weight': [0.6, '3D']
```

New API:
```
 filter_pruner:
    class: 'L1RankedStructureParameterPruner'
    group_type: Filters
    desired_sparsity: 0.6
    weights: [
      module.layer1.0.conv1.weight,
      module.layer1.1.conv1.weight]
```

thresholding.py – separate the generation of the binary_map from the pruning_mask so that we
can cache the binary map and share it between several modules.

pruning/automated_gradual_pruner.py – major refactoring to supported “leader-based”
sub-graph pruning dependencies.  The concept is explained in issue #85


agp-pruning/resnet20_filters.schedule_agp.yaml
agp-pruning/resnet20_filters.schedule_agp_2.yaml
agp-pruning/resnet20_filters.schedule_agp_3.yaml
network_trimming/resnet56_cifar_activation_apoz.yaml
network_trimming/resnet56_cifar_activation_apoz_v2.yaml

Add support to models that contain named empty layers (i.e. have type: NoneType).
This fix catches the AttributeError that is raised when a named NoneType layer is detected, and the layer is ignored.

Thanks to @haim-barad for notifying us about this.

When ranking filters by APoZ, we need to change the sort order, to match
that of module.apoz_channels.value().  This is the result of the previous
bug fix.

Instead of returning the average-percentage-of-zeros, we returned the
average-percentage-of-non-zeros.
So inverted the results, and also multiplied by 100, because the name
of the statistic has "percentage" not "fraction" (not very important,
but still...)

Thanks to Dan Alistarh for bringing this issue to my attention.
The activations of Linear layers have shape (batch_size, output_size) and those
of Convolution layers have shape (batch_size, num_channels, width, height) and
this distinction in shape was not correctly handled.

This commit also fixes sparsity computation for very large activations, as seen
in VGG16, which leads to memory exhaustion.  One solution is to use smaller
batch sizes, but this commit uses a different solution, which counts zeros “manually”,
and using less space.

Also in this commit:
- Added a “caveats” section to the documentation.
- Added more tests.

The super() method of the wrong subclass was used.
In this case there were no practical implications, but we
need to move to the less error-prone syntax of Python 3.x
which does not require us to specify the subclass.

I change the super() invocations in the entire file and ran
two schedules for ResNet56 and actually got better results than
previously.  I don't think these results are related to this
change, and I cannot explain them.  Nontheless, I am committing
these new results, because I also fixed the command-line parameters
of resnet56_cifar_filter_rank_v2.yaml which had a copy & paste
error in it.

* Fix issue #79

Change the default values so that the following scheduler meta-data keys
are always defined: 'starting_epoch', 'ending_epoch', 'frequency'

* compress_classifier.py: add a new argument

Allow the specification, from the command line arguments,  of the range of
pruning levels scanned when doing sensitivity analysis

* Add regression test for issue #79

In our patched ResNet version, we change TorchVision's code so that
ReLU module instances are used only once in a network.

When detecting a module that is used multiple times, stop execution
and print an explanation to the user.v

Trying to simplify the code.

Add docs/conditional_computation.md which was accidentally left out of an
earlier commit.

* Bug fix: value of best_top1 stored in the checkpoint may be wrong

If you invoke compress_clasifier.py with --num-best-scores=n
with n>1, then the value of best_top1 stored in checkpoints is wrong.

When we resume from a checkpoint, we usually want to continue using the checkpoint’s
masks.  I say “usually” because I can see a situation where we want to prune a model
and checkpoint it, and then resume with the intention of fine-tuning w/o keeping the
masks.  This is what’s done in Song Han’s Dense-Sparse-Dense (DSD) training
(https://arxiv.org/abs/1607.04381).  But I didn’t want to add another argument to
```compress_classifier.py``` for the time being – so we ignore DSD.

There are two possible situations when we resume a checkpoint that has a serialized
```CompressionScheduler``` with pruning masks:
1. We are planning on using a new ```CompressionScheduler``` that is defined in a
schedule YAML file.  In this case, we want to copy the masks from the serialized
```CompressionScheduler``` to the new ```CompressionScheduler``` that we are
constructing from the YAML file.  This is one fix.
2. We are resuming a checkpoint, but without using a YAML schedule file.
In this case we want to use the ```CompressionScheduler``` that we loaded from the
checkpoint file.  All this ```CompressionScheduler``` does is keep applying the masks
as we train, so that we don’t lose them.  This is the second fix.

For DSD, we would need a new flag that would override using the ```CompressionScheduler```
that we load from the checkpoint.

Another schedule or ResNet20 Filter-wise pruning for ResNet20, with
64.6% sparsity, 25.4% compute reduction and Top1 91.47% (vs. 91.78
basline).

Top1 is 75.492 (on Epoch: 93) vs the published TorchVision baseline
Top1: 76.15 (-0.66).
Total sparsity: 80.05