diff --git a/README.md b/README.md
index 9b80205..66967ff 100644
--- a/README.md
+++ b/README.md
@@ -1,22 +1,23 @@
 # TIDIGITS recipe
 
-This repository contains a _recipe_ for training an ASR system using the [TIDIGITS database](https://catalog.ldc.upenn.edu/LDC93S10).
+This repository contains a _recipe_ for training an automatic speech recognition (ASR) system using the [TIDIGITS database](https://catalog.ldc.upenn.edu/LDC93S10).
 The recipe is entirely Julia-flavoured and uses following packages (among others):
 * [Flux](https://github.com/FluxML/Flux.jl) as ML library
 * [FiniteStateTransducers](https://github.com/idiap/FiniteStateTransducers.jl) for WFST compositions
 * [HMMGradients](https://github.com/idiap/HMMGradients.jl) for maximum likelihood training
 
-Currently the training runs only on CPU and employs a simple greedy decoder. Stay tuned for more!
+Currently the training runs only on CPU and employs a simple greedy decoder.
 
-### Installation
+## Installation
 
-Set in your environment the path `TIDIGITS_PATH=\your\path\to\tidigits`.
-If you're using SGE set the command flags in `CPU_CMD`, i.e. the queue options.
+Run `julia --project -e 'using Pkg; Pkg.instantiate()'` to install all the dependencies.
+For the live demo install [sox](http://sox.sourceforge.net/).
 
-This can be done e.g. by running `source env.sh` before lunching Julia, where `env.sh` is a script that export these variables. 
-Alternatively, the environment variables can be specified [directly in the REPL](https://docs.julialang.org/en/v1/manual/environment-variables/). 
+## Live Demo
 
-Run `julia --project -e 'using Pkg; Pkg.instantiate()'` to install all the dependencies.
+Open a Julia terminal with `julia --project` and type `include("demo.jl")` to try out the ASR with your own voice. A model trained with configuration `2b` (see below) is already present in this repository. 
+
+## Training
 
 ### Configuration
 
@@ -26,11 +27,18 @@ This folder must contain the following files:
 * `feat_conf.jl` for feature extraction
 * `model_conf.jl` for model and optimisation parameters (hyperparameters)
 A couple of setups are present in this repository for reference in the folder `conf`.
+Currently a TDNN/ConvNet is used as acoustic model.
 
 ### Data preparation
 
+Set in your shell environment the path `TIDIGITS_PATH=\your\path\to\tidigits`.
+If you're using SGE set the command flags in `CPU_CMD`, i.e. the queue options.
+
+This can be done e.g. by running `source env.sh` before lunching Julia, where `env.sh` is a script that export these variables. 
+Alternatively, the environment variables can be specified [directly in the REPL](https://docs.julialang.org/en/v1/manual/environment-variables/). 
+
 Run `julia --project prepare_data.jl --conf 2a` to extract feature and prepare training data using the configuration `2a`.
-Features and transctiptions will be saved in the folder `data/uuid/`.
+Features and transcriptions will be saved in the folder `data/uuid/`.
 Here `uuid` is linked to `feat_conf.jl` file, meaning that if you create a new `model_conf.jl` without modifying feature extraction you don't need to run data preparation twice. 
 If SGE grid is available add the flag `--nj N` to split the work into `N` jobs.
 
@@ -51,7 +59,3 @@ Modify the `conf` by changing the default in the `ArgParse` table.
 ### Evaluation
 
 Run the script `eval.jl` to calculate Word Error Rates (WER) and Phone Error Rate (PER).
-
-### Demo
-
-A live demo can be used by running `demo.jl` (requires [sox](http://sox.sourceforge.net/) to be installed in your system).
diff --git a/models/2a/best_modely_final.bson b/models/2a/best_modely_final.bson
new file mode 100644
index 0000000..c6ffc90
Binary files /dev/null and b/models/2a/best_modely_final.bson differ