fine-tuning 0.001

.gitignore

@@ -6,3 +6,6 @@
 *.o
 .DS_Store
 .token
+
+.ipynb_checkpoints/
+geval

fine-tuning.ipynb

@@ -0,0 +1,144 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "binding-contribution",
+   "metadata": {},
+   "source": [
+    "## Alpha  fine-tuning"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "colored-chaos",
+   "metadata": {},
+   "source": [
+    "Fine-tuning parametru wygładzania alpha. Tylko 10000 linii ze zbioru trenującego, aby oszczędzić czas. <br>\n",
+    "Ostatecznie trening na całym zbiorze z wybranym parametrem."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "environmental-session",
+   "metadata": {},
+   "source": [
+    "#### Alpha 0.001"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "coordinated-ratio",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!py run.py 0.001"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "revolutionary-designation",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!./geval -t dev-0"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "molecular-civilian",
+   "metadata": {},
+   "source": [
+    "#### Alpha 0.005"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "coral-silver",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!py run.py 0.005"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "powered-webcam",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!./geval -t dev-0"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "broadband-heating",
+   "metadata": {},
+   "source": [
+    "#### Alpha 0.01"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "moderate-bearing",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!py run.py 0.01"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "guilty-committee",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!./geval -t dev-0"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "blank-coordinate",
+   "metadata": {},
+   "source": [
+    "| alpha | dev-0 score |\n",
+    "|-------|-------------|\n",
+    "| 0.001 |   |\n",
+    "| 0.005 |   |\n",
+    "| 0.01 |   |\n",
+    "| 0.05 |   |\n",
+    "| 0.1 |   |\n",
+    "| 0.2 |   |\n",
+    "| 0.3 |   |\n",
+    "| 0.4 |   |\n",
+    "| 0.5 |   |"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.2"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

run.py

@@ -61,7 +61,7 @@ if len(sys.argv) != 2:
     print('Wrong number of arguments. Expected 1 - alpha smoothing parameter.')
     quit()
 else:
-    alpha = sys.argv[1]
+    alpha = float(sys.argv[1])
 
 # load training data
 train_data = pd.read_csv('train/in.tsv.xz', sep='\t', error_bad_lines=False, warn_bad_lines=False, header=None, quoting=csv.QUOTE_NONE)