Update readme to include link to full doc and paper

Author: anonymousICLR2019submitter

README.md

@@ -1,9 +1,20 @@
-# Quick Start Guide
+# DeepOBS - A Deep Learning Optimizer Benchmark Suite
 
-## Install Deep OBS
+DeepOBS is a benchmarking suite that drastically simplifies, automates and improves the evaluation of deep learning optimizers.
+
+It can evaluate the performance of new optimizers on a variety of **real-world test problems** and automatically compare them with **realistic baselines**.
+
+The full documentation is available on readthedocs: https://deepobs-iclr.readthedocs.io/
+
+The paper describing DeepOBS is currently under review for ICLR 2019:
+https://openreview.net/forum?id=rJg6ssC5Y7
+
+##  Quick Start Guide
+
+### Install Deep OBS
 	pip install git+https://github.com/anonymousICLR2019submitter/DeepOBS.git
 
-## Download the data
+### Download the data
 	deepobs_prepare_data.sh
 
 This will automatically download, sort and prepare all the datasets (except ImageNet). It can take a while, as it will download roughly 1 GB.
@@ -18,7 +29,7 @@ to run SGD on a simple multi-layer perceptron (with a learning rate of 1e-1 and
 
 Of course, the real value of a benchmark lies in evaluating new optimizers:
 
-## Download and edit a run script
+### Download and edit a run script
 You can download a template run script from there
 
 https://github.com/anonymousICLR2019submitter/DeepOBS/blob/master/scripts/deepobs_run_sgd.py
@@ -35,7 +46,7 @@ Let's assume that we want to benchmark the RMSProp optimizer. Then we only have
 
 Usually the hyperparameters of the optimizers need to be included as well, but for now let's only take the learning rate as a hyperparameter for RMSProp (and if you want change all the 'sgd's in the comments to 'rmsprop'). Let's name this run script now deepobs_run_rmsprop.py
 
-## Run your optimizer
+### Run your optimizer
    You can now run your optimizer on a test problem. Let's try it on a noisy quadratic problem:
 
 	python deepobs_run_rmsprop.py quadratic.noisy_quadratic --num_epochs=100 --lr=1e-1 --bs=128 --pickle --run_name=RMSProp_1e-1/
@@ -66,7 +77,7 @@ Usually the hyperparameters of the optimizers need to be included as well, but f
 	python deepobs_run_rmsprop.py mnist.mnist_mlp --num_epochs=5 --lr=1e-2 --bs=128 --pickle --run_name=RMSProp_1e-2/ --random_seed=44
 
 
-## Plot Results
+### Plot Results
    Now we can plot the results of those two "new" optimizers "RMSProp_1e-1" and "RMSProp_1e-2". Since the performance is always relative, we automatically plot the performance against the most popular optimizers (SGD, Momentum, Adam) with the best settings we found after tuning their hyperparameters. Try out:
 
 	deepobs_plot_results.py --results_dir=results --log
@@ -75,7 +86,7 @@ Usually the hyperparameters of the optimizers need to be included as well, but f
    Additionally it will print out a table summarizing the performances over all test problems (here we only have one or two).
    If you add the option --saveto=save_dir the plots and a color coded table are saved as .png and ready-to-include .tex-files!
 
-## Estimate runtime overhead
+### Estimate runtime overhead
    You can estimate the runtime overhead of the new optimizers compared to SGD like this:
 
 	deepobs_estimate_runtime.py deepobs_run_rmsprop.py --optimizer_arguments=--lr=1e-2