diff --git a/.gitignore b/.gitignore
index 27ad5c6..f2c0f99 100644
--- a/.gitignore
+++ b/.gitignore
@@ -110,4 +110,4 @@ ENV/
 */.ipynb_checkpoints/
 examples/head.stock
 examples/.gitignore
-docs/*.ipynb
\ No newline at end of file
+docs/*.ipynb
diff --git a/examples/requirements.txt b/examples/requirements.txt
index 7ead59f..95e3318 100644
--- a/examples/requirements.txt
+++ b/examples/requirements.txt
@@ -2,5 +2,5 @@ jupyter
 notebook
 jupyterlab
 stockroom
-google_images_download
+python-mnist
 PILLOW
diff --git a/examples/with-git.ipynb b/examples/with-git.ipynb
index fa8dee6..83bb005 100644
--- a/examples/with-git.ipynb
+++ b/examples/with-git.ipynb
@@ -14,11 +14,8 @@
     "For this tutorial, we use a pretrained PyTorch network to classify cats and dogs. We have divided the whole tutorial into 7 stages.\n",
     "1. Setup the repository\n",
     "2. Download some data and store it in stockroom\n",
-    "3. Download the pretrained model and save it in stockroom\n",
-    "4. Train the network and save the model + hyper parameters\n",
-    "5. Download more data if the accuracy is less\n",
-    "6. Train the network again with new data. Save the new network & hyper parameters\n",
-    "7. Fine tune the hyper parameters"
+    "3. Train the network and save the model + hyper parameters\n",
+    "4. Fine tune the hyper parameters"
    ]
   },
   {
@@ -33,14 +30,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Reinitialized existing Git repository in /home/hhsecond/mypro/stockroom/examples/.git/\n"
+      "Initialized empty Git repository in /home/hhsecond/mypro/stockroom/examples/.git/\n"
      ]
     }
    ],
@@ -58,14 +55,15 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Hangar Repo initialized at: /home/hhsecond/mypro/stockroom/examples/.hangar\n"
+      "Hangar Repo initialized at: /home/hhsecond/mypro/stockroom/examples/.hangar\n",
+      "Stock file created\n"
      ]
     }
    ],
@@ -83,7 +81,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [
     {
@@ -91,11 +89,15 @@
      "output_type": "stream",
      "text": [
       "On branch master\n",
+      "\n",
+      "No commits yet\n",
+      "\n",
       "Untracked files:\n",
       "  (use \"git add <file>...\" to include in what will be committed)\n",
       "\n",
+      "\t\u001b[31m.gitignore\u001b[m\n",
       "\t\u001b[31m.ipynb_checkpoints/\u001b[m\n",
-      "\t\u001b[31mdownloads/\u001b[m\n",
+      "\t\u001b[31mhead.stock\u001b[m\n",
       "\t\u001b[31mrequirements.txt\u001b[m\n",
       "\t\u001b[31mwith-git.ipynb\u001b[m\n",
       "\n",
@@ -109,15 +111,17 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[master 66b4aee] initialized repo\n",
-      " 1 file changed, 1 insertion(+), 2 deletions(-)\n"
+      "[master (root-commit) bfe72a8] initialized repo\n",
+      " 2 files changed, 2 insertions(+)\n",
+      " create mode 100644 .gitignore\n",
+      " create mode 100644 head.stock\n"
      ]
     }
    ],
@@ -132,168 +136,452 @@
    "metadata": {},
    "source": [
     "## 2. Download & Store Data\n",
-    "As you know, we'll be building a dog vs cat classifier. Hence we need a lot of images of dogs and cats. Let's use a readily available package (`google_images_download`) to download images from google and store it in hangar\n",
+    "For this tutorial, as most of the tutorials, we'll build a fully connected network to predict hand written digits from MNIST dataset.\n",
     "\n",
     "#### Download images\n",
-    "This might throw error on some attempts but may be keep trying few times"
+    "We download the data using below utility functions (inspired from https://gist.github.com/goldsborough/6dd52a5e01ed73a642c1e772084bcd03)"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 13,
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from urllib.request import urlretrieve\n",
+    "import gzip\n",
+    "import os\n",
+    "import sys\n",
+    "\n",
+    "\n",
+    "def report_download_progress(chunk_number, chunk_size, file_size):\n",
+    "    if file_size != -1:\n",
+    "        percent = min(1, (chunk_number * chunk_size) / file_size)\n",
+    "        bar = '#' * int(64 * percent)\n",
+    "        sys.stdout.write('\\r0% |{:<64}| {}%'.format(bar, int(percent * 100)))\n",
+    "\n",
+    "\n",
+    "def download(destination_path, url):\n",
+    "    if os.path.exists(destination_path):\n",
+    "        print('{} already exists, skipping ...'.format(destination_path))\n",
+    "    else:\n",
+    "        print('Downloading {} ...'.format(url))\n",
+    "        urlretrieve(url, destination_path, reporthook=report_download_progress)\n",
+    "\n",
+    "def unzip(zipped_path):\n",
+    "    unzipped_path = os.path.splitext(zipped_path)[0]\n",
+    "    if os.path.exists(unzipped_path):\n",
+    "        print('{} already exists, skipping ... '.format(unzipped_path))\n",
+    "        return\n",
+    "    with gzip.open(zipped_path, 'rb') as zipped_file:\n",
+    "        with open(unzipped_path, 'wb') as unzipped_file:\n",
+    "            unzipped_file.write(zipped_file.read())\n",
+    "            print('\\nUnzipped {} ...'.format(zipped_path))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "\n",
-      "Item no.: 1 --> Item name = cat\n",
-      "Evaluating...\n",
-      "Starting Download...\n",
-      "\n",
-      "\n",
-      "Unfortunately all 100 could not be downloaded because some images were not downloadable. 0 is all we got for this search filter!\n",
-      "\n",
-      "Errors: 0\n",
-      "\n",
-      "\n",
-      "Item no.: 1 --> Item name = dog\n",
-      "Evaluating...\n",
-      "Starting Download...\n",
-      "\n",
-      "\n",
-      "Unfortunately all 100 could not be downloaded because some images were not downloadable. 0 is all we got for this search filter!\n",
-      "\n",
-      "Errors: 0\n",
-      "\n"
+      "downloads/train-images-idx3-ubyte.gz already exists, skipping ...\n",
+      "downloads/train-images-idx3-ubyte already exists, skipping ... \n",
+      "downloads/train-labels-idx1-ubyte.gz already exists, skipping ...\n",
+      "downloads/train-labels-idx1-ubyte already exists, skipping ... \n",
+      "downloads/t10k-images-idx3-ubyte.gz already exists, skipping ...\n",
+      "downloads/t10k-images-idx3-ubyte already exists, skipping ... \n",
+      "downloads/t10k-labels-idx1-ubyte.gz already exists, skipping ...\n",
+      "downloads/t10k-labels-idx1-ubyte already exists, skipping ... \n"
+     ]
+    }
+   ],
+   "source": [
+    "from pathlib import Path\n",
+    "\n",
+    "RESOURCES = [\n",
+    "    'train-images-idx3-ubyte.gz',\n",
+    "    'train-labels-idx1-ubyte.gz',\n",
+    "    't10k-images-idx3-ubyte.gz',\n",
+    "    't10k-labels-idx1-ubyte.gz',\n",
+    "]\n",
+    "\n",
+    "path = Path('downloads')\n",
+    "path.mkdir(exist_ok=True)\n",
+    "\n",
+    "for resource in RESOURCES:\n",
+    "    destination = os.path.join(str(path), resource)\n",
+    "    url = 'http://yann.lecun.com/exdb/mnist/{}'.format(resource)\n",
+    "    download(destination, url)\n",
+    "    unzip(destination)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "#### Store to StockRoom\n",
+    "We need hangar columns ready for stockroom to store data there. "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Initialized Arrayset: image\n",
+      "Initialized Arrayset: label\n",
+      "Commit message:\n",
+      "arrayset initialized\n",
+      "Commit Successful. Digest: a=28a09ff56d69697bc313561b362200ae94b389d5\n"
+     ]
+    }
+   ],
+   "source": [
+    "!hangar arrayset create image INT64 784\n",
+    "!hangar arrayset create label INT64 1\n",
+    "!stock commit -m 'arrayset initialized'"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from mnist import MNIST\n",
+    "mndata = MNIST(path)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "images, labels = mndata.load_training()\n",
+    "tmpimages, tmplabels = mndata.load_testing()\n",
+    "images.extend(tmpimages)\n",
+    "labels.extend(tmplabels)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from stockroom import StockRoom\n",
+    "stock = StockRoom()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      " * Checking out COMMIT: a=28a09ff56d69697bc313561b362200ae94b389d5\n"
      ]
     },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████| 70000/70000 [00:28<00:00, 2433.96it/s]\n"
+     ]
+    }
+   ],
+   "source": [
+    "from tqdm import tqdm\n",
+    "import numpy as np\n",
+    "\n",
+    "with stock.optimize(write=True):\n",
+    "    for i in tqdm(range(len(images))):\n",
+    "        img = np.array(images[i])\n",
+    "        label = np.array(labels[i]).reshape(1)\n",
+    "        stock.data['image', i] = img\n",
+    "        stock.data['label', i] = label"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Commit message:\n",
+      "added data\n",
+      "Commit Successful. Digest: a=d6b2e5d8bbc397eda5448b3eadc0dc39e14c123e\n"
+     ]
+    }
+   ],
+   "source": [
+    "!stock commit -m 'added data'"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 3. Network training\n",
+    "Let's build a simple fully connected network in PyTorch"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from tqdm import tqdm\n",
+    "import torch\n",
+    "from stockroom import StockRoom\n",
+    "\n",
+    "def train(model, optimizer, criterion):\n",
+    "    stock = StockRoom()\n",
+    "\n",
+    "    with stock.optimize():\n",
+    "        for epoch in range(stock.tag['epoch']):\n",
+    "            running_loss = 0\n",
+    "            trange = tqdm(range(70000))\n",
+    "            for i in trange:\n",
+    "                optimizer.zero_grad()\n",
+    "                sample = torch.from_numpy(stock.data['image', i]).float()\n",
+    "                sample /= 255\n",
+    "                out = model(sample).unsqueeze(0)\n",
+    "                label = torch.from_numpy(stock.data['label', i])\n",
+    "                loss = criterion(out, label)\n",
+    "                running_loss += loss.item()\n",
+    "                loss.backward()\n",
+    "                optimizer.step()\n",
+    "                if i % 1000 == 0 and i != 0:\n",
+    "                    trange.set_description(str(running_loss / i))\n",
+    "            stock.model['mnist'] = model.state_dict()\n",
+    "            stock.commit('added model')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "metadata": {},
+   "outputs": [
     {
      "data": {
       "text/plain": [
-       "({'dog': []}, 0)"
+       "Sequential(\n",
+       "  (0): Linear(in_features=784, out_features=32, bias=True)\n",
+       "  (1): ReLU()\n",
+       "  (2): Linear(in_features=32, out_features=16, bias=True)\n",
+       "  (3): ReLU()\n",
+       "  (4): Linear(in_features=16, out_features=10, bias=True)\n",
+       "  (5): LogSoftmax()\n",
+       ")"
       ]
      },
-     "execution_count": 13,
+     "execution_count": 15,
      "metadata": {},
      "output_type": "execute_result"
     }
    ],
    "source": [
-    "from google_images_download import google_images_download\n",
-    "downloader = google_images_download.googleimagesdownload()\n",
-    "arguments = {\n",
-    "    \"format\": \"jpg\",\n",
-    "    \"limit\":100,\n",
-    "    \"print_urls\":True,\n",
-    "    \"size\": \"medium\",\n",
-    "    \"aspect_ratio\": \"panoramic\"\n",
-    "}\n",
+    "import torch.nn as nn\n",
     "\n",
-    "# downloading cat\n",
-    "arguments['keywords'] = 'cat'\n",
-    "downloader.download(arguments)\n",
+    "stock.tag['lr'] = 0.01\n",
+    "stock.tag['momentum'] = 0.5\n",
+    "stock.tag['epoch'] = 2\n",
+    "stock.commit('hyper params')\n",
     "\n",
+    "input_size = 784\n",
+    "hidden_sizes = [32, 16]\n",
+    "output_size = 10\n",
     "\n",
-    "# downloading dog\n",
-    "arguments['keywords'] = 'dog'\n",
-    "downloader.download(arguments)"
+    "model = nn.Sequential(\n",
+    "    nn.Linear(input_size, hidden_sizes[0]),\n",
+    "    nn.ReLU(),\n",
+    "    nn.Linear(hidden_sizes[0], hidden_sizes[1]),\n",
+    "    nn.ReLU(),\n",
+    "    nn.Linear(hidden_sizes[1], output_size),\n",
+    "    nn.LogSoftmax())"
    ]
   },
   {
-   "cell_type": "markdown",
+   "cell_type": "code",
+   "execution_count": 16,
    "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      " * Checking out COMMIT: a=5c291a0b2d946e3bfa359f754837a112df575bd6\n",
+      " * Checking out COMMIT: a=5c291a0b2d946e3bfa359f754837a112df575bd6\n"
+     ]
+    }
+   ],
    "source": [
-    "#### Store to StockRoom\n",
-    "We need hangar columns ready for stockroom to store data there. "
+    "from torch import optim\n",
+    "\n",
+    "optimizer = optim.SGD(model.parameters(), lr=stock.tag['lr'], momentum=stock.tag['momentum'])\n",
+    "criterion = nn.NLLLoss()"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 40,
+   "execution_count": 17,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      " * Checking out COMMIT: a=5c291a0b2d946e3bfa359f754837a112df575bd6\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "  0%|          | 0/70000 [00:00<?, ?it/s]/home/hhsecond/anaconda3/envs/stockroom/lib/python3.7/site-packages/torch/nn/modules/container.py:92: UserWarning: Implicit dimension choice for log_softmax has been deprecated. Change the call to include dim=X as an argument.\n",
+      "  input = module(input)\n",
+      "0.33214209180101045: 100%|██████████| 70000/70000 [01:20<00:00, 869.94it/s]\n",
+      "0.20504333917206713: 100%|██████████| 70000/70000 [01:21<00:00, 854.18it/s]\n"
+     ]
+    }
+   ],
    "source": [
-    "from pathlib import Path\n",
-    "from PIL import Image\n",
-    "import numpy as np\n",
-    "\n",
-    "newshape = (240, 240)\n",
-    "\n",
-    "for file in Path('downloads/cat').iterdir():\n",
-    "    try:\n",
-    "        data = np.array(Image.open(file).resize((240, 240)))\n",
-    "    except Exception as e:\n",
-    "        pass"
+    "train(model, optimizer, criterion)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 4. Fine tuning\n",
+    "The loss doesn't go below 0.2 with the hyper parameters we have. Let's try increasing the number of neurons in the inner layer"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 39,
+   "execution_count": 18,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "array([[[132, 131, 127],\n",
-       "        [120, 115, 112],\n",
-       "        [111, 112, 107],\n",
-       "        ...,\n",
-       "        [241, 240, 238],\n",
-       "        [243, 242, 240],\n",
-       "        [242, 240, 241]],\n",
-       "\n",
-       "       [[134, 133, 129],\n",
-       "        [123, 120, 115],\n",
-       "        [114, 113, 108],\n",
-       "        ...,\n",
-       "        [241, 240, 238],\n",
-       "        [243, 242, 240],\n",
-       "        [242, 241, 239]],\n",
-       "\n",
-       "       [[136, 137, 132],\n",
-       "        [131, 130, 125],\n",
-       "        [122, 119, 114],\n",
-       "        ...,\n",
-       "        [241, 240, 238],\n",
-       "        [242, 241, 239],\n",
-       "        [241, 240, 238]],\n",
-       "\n",
-       "       ...,\n",
-       "\n",
-       "       [[225, 203, 179],\n",
-       "        [225, 204, 177],\n",
-       "        [226, 202, 176],\n",
-       "        ...,\n",
-       "        [245, 244, 242],\n",
-       "        [247, 246, 244],\n",
-       "        [246, 245, 243]],\n",
-       "\n",
-       "       [[225, 203, 179],\n",
-       "        [226, 205, 178],\n",
-       "        [226, 202, 176],\n",
-       "        ...,\n",
-       "        [245, 244, 242],\n",
-       "        [246, 245, 243],\n",
-       "        [247, 246, 244]],\n",
-       "\n",
-       "       [[225, 203, 179],\n",
-       "        [226, 205, 178],\n",
-       "        [226, 202, 176],\n",
-       "        ...,\n",
-       "        [245, 244, 242],\n",
-       "        [246, 245, 243],\n",
-       "        [246, 245, 243]]], dtype=uint8)"
+       "Sequential(\n",
+       "  (0): Linear(in_features=784, out_features=128, bias=True)\n",
+       "  (1): ReLU()\n",
+       "  (2): Linear(in_features=128, out_features=64, bias=True)\n",
+       "  (3): ReLU()\n",
+       "  (4): Linear(in_features=64, out_features=10, bias=True)\n",
+       "  (5): LogSoftmax()\n",
+       ")"
       ]
      },
-     "execution_count": 39,
+     "execution_count": 18,
      "metadata": {},
      "output_type": "execute_result"
     }
    ],
    "source": [
-    "data"
+    "hidden_sizes = [128, 64]\n",
+    "\n",
+    "model = nn.Sequential(\n",
+    "    nn.Linear(input_size, hidden_sizes[0]),\n",
+    "    nn.ReLU(),\n",
+    "    nn.Linear(hidden_sizes[0], hidden_sizes[1]),\n",
+    "    nn.ReLU(),\n",
+    "    nn.Linear(hidden_sizes[1], output_size),\n",
+    "    nn.LogSoftmax())"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 20,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      " * Checking out COMMIT: a=8b8b7a2f7966acf1c3b5820470fdc34580ef6aaa\n",
+      " * Checking out COMMIT: a=8b8b7a2f7966acf1c3b5820470fdc34580ef6aaa\n",
+      " * Checking out COMMIT: a=8b8b7a2f7966acf1c3b5820470fdc34580ef6aaa\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "0.22921190452411824: 100%|██████████| 70000/70000 [03:47<00:00, 307.77it/s]\n",
+      "0.12486177534811682: 100%|██████████| 70000/70000 [03:25<00:00, 340.91it/s]\n"
+     ]
+    }
+   ],
+   "source": [
+    "optimizer = optim.SGD(model.parameters(), lr=stock.tag['lr'], momentum=stock.tag['momentum'])\n",
+    "\n",
+    "train(model, optimizer, criterion)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Now that the model has enough learning capacity, we can try reducing the learning rate to avoid the jittering of loss across the valley"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 21,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      " * Checking out COMMIT: a=b54ed6f62420c590e2d3206907e239dfa17945f2\n",
+      " * Checking out COMMIT: a=b54ed6f62420c590e2d3206907e239dfa17945f2\n",
+      " * Checking out COMMIT: a=b54ed6f62420c590e2d3206907e239dfa17945f2\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "0.05775309108767264: 100%|██████████| 70000/70000 [04:33<00:00, 255.85it/s] \n",
+      "0.0373574975491017: 100%|██████████| 70000/70000 [04:57<00:00, 235.37it/s]  \n"
+     ]
+    }
+   ],
+   "source": [
+    "stock.tag['lr'] = 0.003\n",
+    "stock.commit('new lr value')\n",
+    "optimizer = optim.SGD(model.parameters(), lr=stock.tag['lr'], momentum=stock.tag['momentum'])\n",
+    "\n",
+    "train(model, optimizer, criterion)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Conclusion\n",
+    "Great! Now we have a well trained MNIST classifier, the data and the hyperparameters we have used, saved in stockroom. Perhaps, for this tutorial, we haven't used practical training methedologies, like splitting the dataset into validation / test etc. But the idea of the existence of this example is to show how stockroom could be used in a real world scenario. Stockroom is still under active development and we'll have more features such as dataloaders for pytorch, tensorflow etc soon."
    ]
   },
   {
diff --git a/tests/conftest.py b/tests/conftest.py
index d13987b..5d40797 100644
--- a/tests/conftest.py
+++ b/tests/conftest.py
@@ -31,7 +31,6 @@ def repo(monkeypatch, managed_tmpdir):
     init_repo('s', 'a@b.c', overwrite=True)
     yield None
     stock = StockRoom()
-    # TODO: It's better to have the `close_environment` as public attribute in hangar
     try:
         stock._repo.hangar_repository._env._close_environments()
     except lmdb.Error:
@@ -47,7 +46,6 @@ def repo_with_aset(repo):
     co.commit('init aset')
     co.close()
     yield None
-    # TODO: It's better to have the `close_environment` as public attribute in hangar
     repo._env._close_environments()
     stock = StockRoom()
     stock._repo.hangar_repository._env._close_environments()
diff --git a/tox.ini b/tox.ini
index 186106f..957c920 100644
--- a/tox.ini
+++ b/tox.ini
@@ -44,7 +44,7 @@ basepython = {env:TOXPYTHON:python3.7}
 deps =
     {[mldeps-py37]deps}
 commands =
-    pytest --basetemp={envtmpdir} --numprocesses=2 --cov --cov-report=term-missing
+    pytest --basetemp={envtmpdir} --numprocesses=2 --cov --cov-report=term-missing -v
 
 [testenv:mypy]
 basepython = {env:TOXPYTHON:python3.7}