Update README.md and refactor code snippets

README.md

@@ -8,6 +8,26 @@ tags:
 ---
 
 # ECG2HRV
-Pipeline for the processing of heart rate data from raw ECG signals (.csv file) towards HRV features.
+Pipeline for the processing of heart rate data from raw ECG signals towards HRV features. For more details see [HUBII](https://hubii.world/)
 
-For more details see [HUBII](https://hubii.world/)
+## How to use
+For importing the model in your project, you can use the following code:
+```python
+# Imports
+from huggingface_hub import hf_hub_download
+import joblib
+
+# Define parameters
+REPO_ID = "HUBII-Platform/ECG2HRV"
+FILENAME = "feature-extractor.joblib"
+
+# Load the model
+model = joblib.load(
+    hf_hub_download(repo_id=REPO_ID, filename=FILENAME)
+)
+```
+Example usage of the model:
+```python
+# TBD: Execute the model with the input data
+result = model.forward(...)
+```

notebooks/{01_Demo_Pipeline_Usage_In_Code.ipynb → 01_Model_Deployment_Research.ipynb}

@@ -3,6 +3,18 @@
   {
    "cell_type": "markdown",
    "source": [
+    "# Introduction\n",
+    "This notebook is used to test different ways to deploy the feature extraction model to the huggingface hub. The goal is to find the best way to deploy the model so that it can be used in the inference API and can be easy accessible for user. In the best way it would also be possible to simply use the huggingface library directly. The following methods will be tested:\n",
+    "\n",
+    "1. [Using ``timm`` to extract features](#Using-timm-to-extract-features) -> ❌\n",
+    "2. [Using ``transformers`` to extract features](#Using-transformers-to-extract-features) --> ❌\n",
+    "    1. [Feature extraction task](#Feature-extraction-task)\n",
+    "    2. [``AutoModel``](#AutoModel)\n",
+    "    3. [Batched feature extraction](#Batched-feature-extraction)\n",
+    "3. [Using simple download](#Using-simple-download) -> ✅\n",
+    "4. [Using custom model](#Using-custom-model) --> 🚧\n",
+    "\n",
+    "**Helpful links and resources**\n",
     "- https://huggingface.co/docs/transformers/custom_models - Alternative creating custom models\n",
     "- https://huggingface.co/templates/feature-extraction - Template for inference API\n",
     "- https://huggingface-widgets.netlify.app/ - Widgets for visualizing models in inference API\n",
@@ -12,6 +24,15 @@
     "collapsed": false
    }
   },
+  {
+   "cell_type": "markdown",
+   "source": [
+    "# Imports"
+   ],
+   "metadata": {
+    "collapsed": false
+   }
+  },
   {
    "cell_type": "code",
    "execution_count": 12,
@@ -32,7 +53,7 @@
   {
    "cell_type": "markdown",
    "source": [
-    "# Using timm to extract features"
+    "# 1. Using ``timm`` to extract features"
    ],
    "metadata": {
     "collapsed": false
@@ -136,7 +157,7 @@
   {
    "cell_type": "markdown",
    "source": [
-    "# Using transformers to extract features"
+    "# 2. Using ``transformers`` to extract features"
    ],
    "metadata": {
     "collapsed": false
@@ -309,7 +330,7 @@
   {
    "cell_type": "markdown",
    "source": [
-    "2. AutoModel"
+    "2. ``AutoModel``"
    ],
    "metadata": {
     "collapsed": false
@@ -369,7 +390,7 @@
   {
    "cell_type": "markdown",
    "source": [
-    "# Using simple download\n",
+    "# 3. Using simple download\n",
     "(See https://huggingface.co/julien-c/wine-quality?structured_data=%7B%7D)"
    ],
    "metadata": {
@@ -378,11 +399,27 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 25,
    "outputs": [],
    "source": [
-    "from huggingface_hub import hf_hub_url, cached_download\n",
-    "import joblib"
+    "from huggingface_hub import hf_hub_download\n",
+    "import joblib\n",
+    "import torch\n",
+    "\n",
+    "from src.model import HR2HRV"
+   ],
+   "metadata": {
+    "collapsed": false,
+    "ExecuteTime": {
+     "end_time": "2024-02-21T11:39:25.775871100Z",
+     "start_time": "2024-02-21T11:39:25.755838Z"
+    }
+   }
+  },
+  {
+   "cell_type": "markdown",
+   "source": [
+    "**Instantiate model and save the model as a joblib file in the huggingface repository**"
    ],
    "metadata": {
     "collapsed": false
@@ -390,15 +427,99 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 27,
    "outputs": [],
    "source": [
+    "# Instantiate model\n",
+    "model = HR2HRV()\n",
+    "# Save\n",
+    "joblib.dump(model, \"..\\ECG2HRV.joblib\")\n",
+    "# Load in notebook\n",
+    "model = joblib.load(\"..\\ECG2HRV.joblib\")"
+   ],
+   "metadata": {
+    "collapsed": false,
+    "ExecuteTime": {
+     "end_time": "2024-02-21T12:01:28.600527100Z",
+     "start_time": "2024-02-21T12:01:28.580278200Z"
+    }
+   }
+  },
+  {
+   "cell_type": "markdown",
+   "source": [
+    "**Test if the model can be loaded from the hub and used**"
+   ],
+   "metadata": {
+    "collapsed": false
+   }
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "outputs": [],
+   "source": [
+    "# Load from hub\n",
     "REPO_ID = \"HUBII-Platform/ECG2HRV\"\n",
     "FILENAME = \"feature-extractor.joblib\"\n",
     "\n",
-    "model = joblib.load(cached_download(\n",
-    "    hf_hub_url(REPO_ID, FILENAME)\n",
-    "))\n"
+    "model = joblib.load(\n",
+    "    hf_hub_download(repo_id=REPO_ID, filename=FILENAME)\n",
+    ")"
+   ],
+   "metadata": {
+    "collapsed": false,
+    "ExecuteTime": {
+     "end_time": "2024-02-21T11:36:52.302912800Z",
+     "start_time": "2024-02-21T11:36:52.145834500Z"
+    }
+   }
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 20,
+   "outputs": [],
+   "source": [
+    "# Create a example tensor input\n",
+    "tensor = torch.tensor([2.0, 3.0, 4.0])"
+   ],
+   "metadata": {
+    "collapsed": false,
+    "ExecuteTime": {
+     "end_time": "2024-02-21T11:36:55.990475100Z",
+     "start_time": "2024-02-21T11:36:55.989181100Z"
+    }
+   }
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 24,
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "{'rmssd': tensor(1.), 'mean_rr': tensor(3.), 'sdnn': tensor(1.), 'mean_hr': tensor(20000.)}\n"
+     ]
+    }
+   ],
+   "source": [
+    "# Run forward pass\n",
+    "output = model.forward(tensor)\n",
+    "print(output)\n"
+   ],
+   "metadata": {
+    "collapsed": false,
+    "ExecuteTime": {
+     "end_time": "2024-02-21T11:37:12.239246200Z",
+     "start_time": "2024-02-21T11:37:12.219556200Z"
+    }
+   }
+  },
+  {
+   "cell_type": "markdown",
+   "source": [
+    "# 4. Using custom model (not tested yet)\n"
    ],
    "metadata": {
     "collapsed": false