diff --git a/README.md b/README.md
index bf22b8d9..c65db6c6 100644
--- a/README.md
+++ b/README.md
@@ -11,13 +11,14 @@ Home of the **TakeLab Podium** project.
 You will work in a virtual environment and keep a list of required
 dependencies in a ```requirements.txt``` file. The master branch of the 
 project **must** be buildable with passing tests **all the time**. 
+Code coverage should be kept as high as possible. 
 
 ```
 
 virtualenv -p python3.6 env
 source env/bin/activate
 pip install -r requirements.txt
-python -m pytest
+py.test --cov=takepod test
 
 ```
 
@@ -26,6 +27,13 @@ python -m pytest
 Adding a new library to a project should be done via ```pip install
 <new_framework>```. **Don't forget to add it to requirements.txt** 
 
+The best thing to do is to manually add dependencies to the
+```requirements.txt``` file instead of using 
+```pip freeze > requirements.txt```. 
+See [here](https://medium.com/@tomagee/pip-freeze-requirements-txt-considered-harmful-f0bce66cf895)
+why.
+
+
 ## Details
 
 The project is packaged according to official Python packaging
diff --git a/notes.md b/notes.md
new file mode 100644
index 00000000..c7b9a149
--- /dev/null
+++ b/notes.md
@@ -0,0 +1,122 @@
+### Bayesian optimization 
+
+- part of Sequential model-based optimization algorithm family
+- multiple python libraries already available
+ - [spearmint](https://github.com/HIPS/Spearmint)
+ - [smac](https://www.cs.ubc.ca/labs/beta/Projects/SMAC/)
+ - [Bayesian Optimization](https://github.com/fmfn/BayesianOptimization)
+ - [hyperopt](https://github.com/hyperopt/hyperopt)
+ - [pygpgo](http://pygpgo.readthedocs.io/en/latest/)
+ - [GPyOpt](https://sheffieldml.github.io/GPyOpt/)
+
+Four preconditions for doing BO
+
+1. Objective Function: takes in an input and returns a loss to minimize
+2. Domain space: the range of input values to evaluate Optimization
+3. Algorithm: the method used to construct the surrogate function and
+   choose the next values to evaluate 
+4. Results: score, value pairs that the algorithm uses to build the model
+
+Benefits of Bayesian optimization over random search
+
+- works better most of the time (less iterations)
+- returns distributions of searched spaces (which learning rates are
+  better than other)
+- returns trials
+
+Resources
+
+- Excellent notebook on [Bayesian
+  optimization](https://github.com/WillKoehrsen/hyperparameter-optimization/blob/master/Introduction%20to%20Bayesian%20Optimization%20with%20Hyperopt.ipynb)
+- [Blog post using scikit](https://thuijskens.github.io/2016/12/29/bayesian-optimisation/)
+
+
+### Active learning
+
+The idea is to label part of the dataset and try to intelligently select
+which data to label to maximize model performance. 
+
+Procedure
+
+1. Split data to seed(will be labelled) and unlabelled. 
+
+2. Train the model using seed
+
+3. Choose unlabelled instances to label based on one of many selection
+   criteria (Pool-based, Stream-baseda) based on least confidence
+
+4. Iterate  2. and 3. until some stopping criteria is met
+
+
+Python libraries
+
+- [google active learning](https://github.com/google/active-learning),
+  6 sampling strategies
+- [modAL](https://cosmic-cortex.github.io/modAL/#introduction)
+- [libact](https://github.com/ntucllab/libact), Pool-based
+  Active Learning in Python
+- [acton](https://github.com/chengsoonong/acton)
+
+Resources
+
+- [Tutorial](https://towardsdatascience.com/active-learning-tutorial-57c3398e34d)
+- [Begginer datacamp tutorial](https://www.datacamp.com/community/tutorials/active-learning)
+
+### Framework part
+
+- interchanging saved models -- use ONNX format (ici generalnije)
+- continuous integration using Travis CI
+
+### Development setup
+
+- Python (you will likely use these)
+  - [pytest](https://docs.pytest.org/en/latest/getting-started.html)
+  - [Decorators in python](https://realpython.com/primer-on-python-decorators/)
+  - [Debugger pdb](https://docs.python.org/2/library/pdb.html)
+  - pandas
+  - pytorch
+  - [pytest](https://semaphoreci.com/community/tutorials/testing-python-applications-with-pytest)
+
+https://docs.pytest.org/en/latest/example/simple.html
+- Git 
+  - [git basics](https://git-scm.com/book/en/v2/Getting-Started-Git-Basics)
+  - [how I operate in git](https://medium.com/@fredrikmorken/why-you-should-stop-using-git-rebase-5552bee4fed1)
+- Linux (all TakeLab servers have Ubuntu installed)
+  - [tmux](https://www.hamvocke.com/blog/a-quick-and-easy-guide-to-tmux/)
+    / [nohup](https://linux.101hacks.com/unix/nohup-command/)
+  - [htop](http://www.deonsworld.co.za/2012/12/20/understanding-and-using-htop-monitor-system-resources/)
+  - nvidia-smi
+  
+
+Advanced preconditions (also depending on personal choice)
+
+- vim (vimtutor)
+- [Tmux + vim](https://blog.bugsnag.com/tmux-and-vim/)
+
+
+### Some general test principles
+
+- unit test should run within a few seconds
+- unit tests should mock external resources
+- integration tests should be run separately from unit tests 
+- unit tests should have extremely high coverage, integration tests should
+  focus on only testing integration
+
+Some general source writing practices
+
+- discern between public and private methods ('\_' for private)
+- proposed: write numpydoc-style documentation for public methods
+- use flake8 to check code formatting
+
+### Onboarding -- week 1 checklist
+
+By the end of week 1 you should have:
+
+- git access (use pull request)
+- ssh access to all TakeLab servers (ask sysadmins on Trello to give you
+  access)
+- trello access (ask TakeLab members to give you access)
+- create and launch a deep learning model on TakeLab GPU resources 
+- checkout, build and run tests of TakeLab podium
+- make a simple code change on TakeLab podium passing
+- pick up a software implementation task
diff --git a/requirements.txt b/requirements.txt
index f326e338..632f959d 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,26 +1,31 @@
-atomicwrites==1.2.1
-attrs==18.2.0
-backcall==0.1.0
-decorator==4.3.0
+flake8==3.5.0
 ipython==6.5.0
 ipython-genutils==0.2.0
-jedi==0.12.1
+mccabe==0.6.1
 more-itertools==4.3.0
+msgpack-numpy==0.4.3.1
 numpy==1.15.1
-parso==0.3.1
-pexpect==4.6.0
-pickleshare==0.7.4
+plac==0.9.6
 pluggy==0.7.1
+preshed==1.0.1
 prompt-toolkit==1.0.15
 ptyprocess==0.6.0
 py==1.6.0
-Pygments==2.2.0
+pycodestyle==2.3.1
+pyflakes==1.6.0
 pytest==3.7.4
+pytest-cov==2.6.0
+regex==2017.4.5
+requests==2.19.1
 scikit-learn==0.19.2
 scipy==1.1.0
 simplegeneric==0.8.1
 six==1.11.0
 sklearn==0.0
+spacy==2.0.12
 takepod==0.1
-traitlets==4.3.2
-wcwidth==0.1.7
+torch==0.4.1
+torchtext==0.2.3
+tqdm==4.25.0
+urllib3==1.23
+wget==3.2
diff --git a/takepod/preproc/stemmer/__init__.py b/takepod/preproc/stemmer/__init__.py
new file mode 100644
index 00000000..e69de29b
diff --git a/takepod/preproc/stemmer/croatian_stemmer.py b/takepod/preproc/stemmer/croatian_stemmer.py
new file mode 100644
index 00000000..aad70818
--- /dev/null
+++ b/takepod/preproc/stemmer/croatian_stemmer.py
@@ -0,0 +1,109 @@
+# -*-coding:utf-8-*-
+#
+#    Simple stemmer for Croatian v0.1
+#    Copyright 2012 Nikola Ljubešić and Ivan Pandžić
+#
+#    This program is free software: you can redistribute it and/or modify
+#    it under the terms of the GNU Lesser General Public License as published
+#    by the Free Software Foundation, either version 3 of the License, or
+#    (at your option) any later version.
+#
+#    This program is distributed in the hope that it will be useful,
+#    but WITHOUT ANY WARRANTY; without even the implied warranty of
+#    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+#    GNU Lesser General Public License for more details.
+#
+#    You should have received a copy of the GNU Lesser General Public License
+#    along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import re
+import os
+
+
+class CroatianStemmer:
+
+    # list of words that are it's own stem
+    __stop = None
+    __transformations = None
+    __rules = None
+
+    def __init__(self):
+
+        dir_path = os.path.dirname(os.path.realpath(__file__))
+        self.__rules = [
+            re.compile(r'^(' + base + ')(' + suffix + r')$') for
+            base, suffix in [
+                e.strip().split(' ')
+                for e in
+                open(os.path.join(dir_path, "rules.txt"), encoding='utf-8')]
+        ]
+        self.__transformations = [e.strip().split(
+            '\t') for e in open(os.path.join(dir_path,
+                                             'transformations.txt'),
+                                encoding='utf-8')]
+        self.__stop = set([
+            e.strip()
+            for e in
+            open(os.path.join(dir_path, 'nostem-hr.txt'), encoding='utf-8')
+        ])
+
+    def _determine_r_vowel(self, string):
+        '''
+        Determines if 'r' is a vowel or not
+        If it is => uppercase it.
+
+        Parameters
+        ----------
+        string : str
+            word in Croatian
+
+        Returns
+        -------
+        string : str
+            Croatian word with 'r' vowel uppercased
+        '''
+        return re.sub(r'(^|[^aeiou])r($|[^aeiou])', r'\1R\2', string)
+
+    def _has_vowel(self, string):
+        if re.search(r'[aeiouR]', self._determine_r_vowel(string)):
+            return True
+        else:
+            return False
+
+    def transform(self, word):
+        for seek, replace in self.__transformations:
+            if word.endswith(seek):
+                return word[:-len(seek)] + replace
+        return word
+
+    def root_word(self, word):
+        for rule in self.__rules:
+            division = rule.match(word)
+            if division:
+                root = division.group(1)
+                if self._has_vowel(root) and len(root) > 1:
+                    return root
+        return word
+
+    def stem_word(self, word):
+        '''
+        Returns the root or roots of a word,
+        together with any derivational affixes
+
+        Parameters
+        ----------
+        word : str
+            word in Croatian
+
+        Returns
+        -------
+        string : str
+            Croatian word root plus derivational morphemes
+        '''
+
+        if word.lower() in self.__stop:
+            return word
+        stem = self.root_word(self.transform(word.lower()))
+        return "".join(list(
+            stem[i].upper() if c.isupper() else stem[i] for i, c in
+            zip(range(len(stem)), word)))
diff --git a/takepod/preproc/stemmer/nostem-hr.txt b/takepod/preproc/stemmer/nostem-hr.txt
new file mode 100644
index 00000000..fbf06046
--- /dev/null
+++ b/takepod/preproc/stemmer/nostem-hr.txt
@@ -0,0 +1,67 @@
+bi
+bih
+bijah
+bijahu
+bijasmo
+bijaste
+bijaše
+bila
+bile
+bili
+bilo
+bio
+bismo
+biste
+biti
+biše
+bje
+bjeh
+bjehu
+bjesmo
+bjeste
+bješe
+budem
+budemo
+budete
+budeš
+budimo
+budite
+budu
+jesam
+jesi
+jesmo
+jeste
+jesu
+mogu
+mora
+moraju
+moram
+moramo
+morate
+moraš
+može
+možemo
+možete
+možeš
+sam
+si
+smo
+ste
+su
+treba
+trebaju
+trebam
+trebamo
+trebate
+trebaš
+će
+ćemo
+ćete
+ćeš
+ću
+žele
+želi
+želim
+želimo
+želite
+želiš
diff --git a/takepod/preproc/stemmer/rules.txt b/takepod/preproc/stemmer/rules.txt
new file mode 100644
index 00000000..965c79b0
--- /dev/null
+++ b/takepod/preproc/stemmer/rules.txt
@@ -0,0 +1,102 @@
+.+(s|š)k ijima|ijega|ijemu|ijem|ijim|ijih|ijoj|ijeg|iji|ije|ija|oga|ome|omu|ima|og|om|im|ih|oj|i|e|o|a|u
+.+(s|š)tv ima|om|o|a|u
+# N
+.+(t|m|p|r|g)anij ama|ima|om|a|u|e|i| 
+.+an inom|ina|inu|ine|ima|in|om|u|i|a|e| 
+.+in ima|ama|om|a|e|i|u|o| 
+.+on ovima|ova|ove|ovi|ima|om|a|e|i|u| 
+.+n ijima|ijega|ijemu|ijeg|ijem|ijim|ijih|ijoj|iji|ije|ija|iju|ima|ome|omu|oga|oj|om|ih|im|og|o|e|a|u|i| 
+# Ć
+.+(a|e|u)ć oga|ome|omu|ega|emu|ima|oj|ih|om|eg|em|og|uh|im|e|a
+# G
+.+ugov ima|i|e|a
+.+ug ama|om|a|e|i|u|o
+.+log ama|om|a|u|e| 
+.+[^eo]g ovima|ama|ovi|ove|ova|om|a|e|i|u|o| 
+# I
+.+(rrar|ott|ss|ll)i jem|ja|ju|o| 
+# J
+.+uj ući|emo|ete|mo|em|eš|e|u| 
+.+(c|č|ć|đ|l|r)aj evima|evi|eva|eve|ama|ima|em|a|e|i|u| 
+.+(b|c|d|l|n|m|ž|g|f|p|r|s|t|z)ij ima|ama|om|a|e|i|u|o| 
+# L
+#.+al inom|ina|inu|ine|ima|om|in|i|a|e
+#.+[^(lo|ž)]il ima|om|a|e|u|i| 
+.+[^z]nal ima|ama|om|a|e|i|u|o| 
+.+ijal ima|ama|om|a|e|i|u|o| 
+.+ozil ima|om|a|e|u|i| 
+.+olov ima|i|a|e
+.+ol ima|om|a|u|e|i| 
+# M
+.+lem ama|ima|om|a|e|i|u|o| 
+.+ram ama|om|a|e|i|u|o
+#.+(es|e|u)m ama|om|a|e|i|u|o
+# R
+#.+(a|d|e|o|u)r ama|ima|om|u|a|e|i| 
+.+(a|d|e|o)r ama|ima|om|u|a|e|i| 
+# S
+.+(e|i)s ima|om|e|a|u
+# Š
+.+(t|n|j|k|j|t|b|g|v)aš ama|ima|om|em|a|u|i|e| 
+.+(e|i)š ima|ama|om|em|i|e|a|u| 
+# T
+.+ikat ima|om|a|e|i|u|o| 
+.+lat ima|om|a|e|i|u|o| 
+.+et ama|ima|om|a|e|i|u|o| 
+#.+ot ama|ima|om|a|u|e|i| 
+.+(e|i|k|o)st ima|ama|om|a|e|i|u|o| 
+.+išt ima|em|a|e|u
+#.+ut ovima|evima|ove|ovi|ova|eve|evi|eva|ima|om|a|u|e|i| 
+# V
+.+ova smo|ste|hu|ti|še|li|la|le|lo|t|h|o
+.+(a|e|i)v ijemu|ijima|ijega|ijeg|ijem|ijim|ijih|ijoj|oga|ome|omu|ima|ama|iji|ije|ija|iju|im|ih|oj|om|og|i|a|u|e|o| 
+.+[^dkml]ov ijemu|ijima|ijega|ijeg|ijem|ijim|ijih|ijoj|oga|ome|omu|ima|iji|ije|ija|iju|im|ih|oj|om|og|i|a|u|e|o| 
+.+(m|l)ov ima|om|a|u|e|i| 
+# PRIDJEVI
+.+el ijemu|ijima|ijega|ijeg|ijem|ijim|ijih|ijoj|oga|ome|omu|ima|iji|ije|ija|iju|im|ih|oj|om|og|i|a|u|e|o| 
+.+(a|e|š)nj ijemu|ijima|ijega|ijeg|ijem|ijim|ijih|ijoj|oga|ome|omu|ima|iji|ije|ija|iju|ega|emu|eg|em|im|ih|oj|om|og|a|e|i|o|u
+.+čin ama|ome|omu|oga|ima|og|om|im|ih|oj|a|u|i|o|e| 
+.+roši vši|smo|ste|še|mo|te|ti|li|la|lo|le|m|š|t|h|o
+.+oš ijemu|ijima|ijega|ijeg|ijem|ijim|ijih|ijoj|oga|ome|omu|ima|iji|ije|ija|iju|im|ih|oj|om|og|i|a|u|e| 
+.+(e|o)vit ijima|ijega|ijemu|ijem|ijim|ijih|ijoj|ijeg|iji|ije|ija|oga|ome|omu|ima|og|om|im|ih|oj|i|e|o|a|u| 
+#.+tit ijima|ijega|ijemu|ijem|ijim|ijih|ijoj|ijeg|iji|ije|ija|oga|ome|omu|ima|og|om|im|ih|oj|e|o|a|u|i| 
+.+ast ijima|ijega|ijemu|ijem|ijim|ijih|ijoj|ijeg|iji|ije|ija|oga|ome|omu|ima|og|om|im|ih|oj|i|e|o|a|u| 
+.+k ijemu|ijima|ijega|ijeg|ijem|ijim|ijih|ijoj|oga|ome|omu|ima|iji|ije|ija|iju|im|ih|oj|om|og|i|a|u|e|o| 
+# GLAGOLI
+.+(e|a|i|u)va jući|smo|ste|jmo|jte|ju|la|le|li|lo|mo|na|ne|ni|no|te|ti|še|hu|h|j|m|n|o|t|v|š| 
+.+ir ujemo|ujete|ujući|ajući|ivat|ujem|uješ|ujmo|ujte|avši|asmo|aste|ati|amo|ate|aju|aše|ahu|ala|alo|ali|ale|uje|uju|uj|al|an|am|aš|at|ah|ao
+.+ač ismo|iste|iti|imo|ite|iše|eći|ila|ilo|ili|ile|ena|eno|eni|ene|io|im|iš|it|ih|en|i|e
+.+ača vši|smo|ste|smo|ste|hu|ti|mo|te|še|la|lo|li|le|ju|na|no|ni|ne|o|m|š|t|h|n
+#.+ači smo|ste|ti|li|la|lo|le|mo|te|še|m|š|t|h|o| 
+# Druga_vrsta
+.+n uvši|usmo|uste|ući|imo|ite|emo|ete|ula|ulo|ule|uli|uto|uti|uta|em|eš|uo|ut|e|u|i
+.+ni vši|smo|ste|ti|mo|te|mo|te|la|lo|le|li|m|š|o
+# A
+.+((a|r|i|p|e|u)st|[^o]g|ik|uc|oj|aj|lj|ak|ck|čk|šk|uk|nj|im|ar|at|et|št|it|ot|ut|zn|zv)a jući|vši|smo|ste|jmo|jte|jem|mo|te|je|ju|ti|še|hu|la|li|le|lo|na|no|ni|ne|t|h|o|j|n|m|š
+.+ur ajući|asmo|aste|ajmo|ajte|amo|ate|aju|ati|aše|ahu|ala|ali|ale|alo|ana|ano|ani|ane|al|at|ah|ao|aj|an|am|aš
+.+(a|i|o)staj asmo|aste|ahu|ati|emo|ete|aše|ali|ući|ala|alo|ale|mo|ao|em|eš|at|ah|te|e|u| 
+.+(b|c|č|ć|d|e|f|g|j|k|n|r|t|u|v)a lama|lima|lom|lu|li|la|le|lo|l
+.+(t|č|j|ž|š)aj evima|evi|eva|eve|ama|ima|em|a|e|i|u| 
+#.+(e|j|k|r|u|v)al ama|ima|om|u|i|a|e|o| 
+#.+(e|j|k|r|t|u|v)al ih|im
+.+([^o]m|ič|nč|uč|b|c|ć|d|đ|h|j|k|l|n|p|r|s|š|v|z|ž)a jući|vši|smo|ste|jmo|jte|mo|te|ju|ti|še|hu|la|li|le|lo|na|no|ni|ne|t|h|o|j|n|m|š
+.+(a|i|o)sta dosmo|doste|doše|nemo|demo|nete|dete|nimo|nite|nila|vši|nem|dem|neš|deš|doh|de|ti|ne|nu|du|la|li|lo|le|t|o
+.+ta smo|ste|jmo|jte|vši|ti|mo|te|ju|še|la|lo|le|li|na|no|ni|ne|n|j|o|m|š|t|h
+.+inj asmo|aste|ati|emo|ete|ali|ala|alo|ale|aše|ahu|em|eš|at|ah|ao
+.+as temo|tete|timo|tite|tući|tem|teš|tao|te|li|ti|la|lo|le
+# I
+.+(elj|ulj|tit|ac|ič|od|oj|et|av|ov)i vši|eći|smo|ste|še|mo|te|ti|li|la|lo|le|m|š|t|h|o
+.+(tit|jeb|ar|ed|uš|ič)i jemo|jete|jem|ješ|smo|ste|jmo|jte|vši|mo|še|te|ti|ju|je|la|lo|li|le|t|m|š|h|j|o
+.+(b|č|d|l|m|p|r|s|š|ž)i jemo|jete|jem|ješ|smo|ste|jmo|jte|vši|mo|lu|še|te|ti|ju|je|la|lo|li|le|t|m|š|h|j|o
+.+luč ujete|ujući|ujemo|ujem|uješ|ismo|iste|ujmo|ujte|uje|uju|iše|iti|imo|ite|ila|ilo|ili|ile|ena|eno|eni|ene|uj|io|en|im|iš|it|ih|e|i
+.+jeti smo|ste|še|mo|te|ti|li|la|lo|le|m|š|t|h|o
+.+e lama|lima|lom|lu|li|la|le|lo|l
+.+i lama|lima|lom|lu|li|la|le|lo|l
+# Pridjev_t
+.+at ijega|ijemu|ijima|ijeg|ijem|ijih|ijim|ima|oga|ome|omu|iji|ije|ija|iju|oj|og|om|im|ih|a|u|i|e|o| 
+# Pridjev
+.+et avši|ući|emo|imo|em|eš|e|u|i
+.+ ajući|alima|alom|avši|asmo|aste|ajmo|ajte|ivši|amo|ate|aju|ati|aše|ahu|ali|ala|ale|alo|ana|ano|ani|ane|am|aš|at|ah|ao|aj|an
+.+ anje|enje|anja|enja|enom|enoj|enog|enim|enih|anom|anoj|anog|anim|anih|eno|ovi|ova|oga|ima|ove|enu|anu|ena|ama
+.+ nijega|nijemu|nijima|nijeg|nijem|nijim|nijih|nima|niji|nije|nija|niju|noj|nom|nog|nim|nih|an|na|nu|ni|ne|no
+.+ om|og|im|ih|em|oj|an|u|o|i|e|a
\ No newline at end of file
diff --git a/takepod/preproc/stemmer/transformations.txt b/takepod/preproc/stemmer/transformations.txt
new file mode 100644
index 00000000..52f79cb6
--- /dev/null
+++ b/takepod/preproc/stemmer/transformations.txt
@@ -0,0 +1,131 @@
+lozi	loga
+lozima	loga
+pjesi	pjeh
+pjesima	pjeh
+vojci	vojka
+bojci	bojka
+jaci	jak
+jacima	jak
+čajan	čajni
+ijeran	ijerni
+laran	larni
+ijesan	ijesni
+anjac	anjca
+ajac	ajca
+ajaca	ajca
+ljaca	ljca
+ljac	ljca
+ejac	ejca
+ejaca	ejca
+ojac	ojca
+ojaca	ojca
+ajaka	ajka
+ojaka	ojka
+šaca	šca
+šac	šca
+inzima	ing
+inzi	ing
+tvenici	tvenik
+tetici	tetika
+teticima	tetika
+nstava	nstva
+nicima	nik
+ticima	tik
+zicima	zik
+snici	snik
+kuse	kusi
+kusan	kusni
+kustava	kustva
+dušan	dušni
+antan	antni
+bilan	bilni
+tilan	tilni
+avilan	avilni
+silan	silni
+gilan	gilni
+rilan	rilni
+nilan	nilni
+alan	alni
+ozan	ozni
+rave	ravi
+stavan	stavni
+pravan	pravni
+tivan	tivni
+sivan	sivni
+atan	atni
+cenata	centa
+denata	denta
+genata	genta
+lenata	lenta
+menata	menta
+jenata	jenta
+venata	venta
+tetan	tetni
+pletan	pletni
+šave	šavi
+manata	manta
+tanata	tanta
+lanata	lanta
+sanata	santa
+ačak	ačka
+ačaka	ačka
+ušak	uška
+atak	atka
+ataka	atka
+atci	atka
+atcima	atka
+etak	etka
+etaka	etka
+itak	itka
+itaka	itka
+itci	itka
+otak	otka
+otaka	otka
+utak	utka
+utaka	utka
+utci	utka
+utcima	utka
+eskan	eskna
+tičan	tični
+ojsci	ojska
+esama	esma
+metara	metra
+centar	centra
+centara	centra
+istara	istra
+istar	istra
+ošću	osti
+daba	dba
+čcima	čka
+čci	čka
+mac	mca
+maca	mca
+naca	nca
+nac	nca
+voljan	voljni
+anaka	anki
+vac	vca
+vaca	vca
+saca	sca
+sac	sca
+naca	nca
+nac	nca
+raca	rca
+rac	rca
+aoca	alca
+alaca	alca
+alac	alca
+elaca	elca
+elac	elca
+olaca	olca
+olac	olca
+olce	olca
+njac	njca
+njaca	njca
+ekata	ekta
+ekat	ekta
+izam	izma
+izama	izma
+jebe	jebi
+baci	baci
+ašan	ašni
\ No newline at end of file
diff --git a/test/test_preproc.py b/test/test_preproc.py
new file mode 100644
index 00000000..61871a06
--- /dev/null
+++ b/test/test_preproc.py
@@ -0,0 +1,30 @@
+from takepod.preproc.stemmer.croatian_stemmer import CroatianStemmer
+import pytest
+
+
+@pytest.fixture(scope='module')
+def cro_stemmer():
+    yield CroatianStemmer()
+
+
+def test_croatian_stemmer_stem_word(cro_stemmer):
+    assert cro_stemmer.stem_word('babicama') == 'babic'
+    assert cro_stemmer.stem_word('babice') == 'babic'
+
+
+def test_croatian_stemmer_stem_nostem_word(cro_stemmer):
+    assert cro_stemmer.stem_word('želimo') == 'želimo'
+    assert cro_stemmer.stem_word('jesmo') == 'jesmo'
+
+
+def test_croatian_stemmer_no_vowel_word(cro_stemmer):
+    # this is an actual word in Croatian, look it up
+    assert cro_stemmer.stem_word('sntntn') == 'sntntn'
+
+
+def test_croatian_stemmer_transformative_word(cro_stemmer):
+    assert cro_stemmer.stem_word('turizama') == 'turizm'
+
+
+def test_croatian_stemmer_preserves_case(cro_stemmer):
+    assert cro_stemmer.stem_word('Turizam') == 'Turizm'