Upload CWE_Predictor.ipynb with huggingface_hub
Browse files- CWE_Predictor.ipynb +57 -32
CWE_Predictor.ipynb
CHANGED
|
@@ -620,7 +620,7 @@
|
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"cell_type": "code",
|
| 623 |
-
"execution_count":
|
| 624 |
"id": "4b087fae",
|
| 625 |
"metadata": {},
|
| 626 |
"outputs": [
|
|
@@ -630,8 +630,8 @@
|
|
| 630 |
"\n",
|
| 631 |
" <div>\n",
|
| 632 |
" \n",
|
| 633 |
-
" <progress value='
|
| 634 |
-
" [
|
| 635 |
" </div>\n",
|
| 636 |
" <table border=\"1\" class=\"dataframe\">\n",
|
| 637 |
" <thead>\n",
|
|
@@ -672,6 +672,20 @@
|
|
| 672 |
" <td>0.720193</td>\n",
|
| 673 |
" <td>0.242404</td>\n",
|
| 674 |
" </tr>\n",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 675 |
" </tbody>\n",
|
| 676 |
"</table><p>"
|
| 677 |
],
|
|
@@ -681,6 +695,16 @@
|
|
| 681 |
},
|
| 682 |
"metadata": {},
|
| 683 |
"output_type": "display_data"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 684 |
}
|
| 685 |
],
|
| 686 |
"source": [
|
|
@@ -689,7 +713,7 @@
|
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"cell_type": "code",
|
| 692 |
-
"execution_count":
|
| 693 |
"id": "48faf17c",
|
| 694 |
"metadata": {},
|
| 695 |
"outputs": [],
|
|
@@ -705,37 +729,31 @@
|
|
| 705 |
},
|
| 706 |
{
|
| 707 |
"cell_type": "code",
|
| 708 |
-
"execution_count":
|
| 709 |
"id": "fcb11390",
|
| 710 |
"metadata": {},
|
| 711 |
"outputs": [
|
| 712 |
{
|
| 713 |
"data": {
|
| 714 |
"application/vnd.jupyter.widget-view+json": {
|
| 715 |
-
"model_id": "
|
| 716 |
"version_major": 2,
|
| 717 |
"version_minor": 0
|
| 718 |
},
|
| 719 |
"text/plain": [
|
| 720 |
-
"
|
| 721 |
]
|
| 722 |
},
|
| 723 |
"metadata": {},
|
| 724 |
"output_type": "display_data"
|
| 725 |
},
|
| 726 |
{
|
| 727 |
-
"
|
| 728 |
-
|
| 729 |
-
|
| 730 |
-
|
| 731 |
-
|
| 732 |
-
|
| 733 |
-
"text/plain": [
|
| 734 |
-
"README.md: 0.00B [00:00, ?B/s]"
|
| 735 |
-
]
|
| 736 |
-
},
|
| 737 |
-
"metadata": {},
|
| 738 |
-
"output_type": "display_data"
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"name": "stdout",
|
|
@@ -759,19 +777,10 @@
|
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"cell_type": "code",
|
| 762 |
-
"execution_count":
|
| 763 |
"id": "33847880",
|
| 764 |
"metadata": {},
|
| 765 |
-
"outputs": [
|
| 766 |
-
{
|
| 767 |
-
"name": "stdout",
|
| 768 |
-
"output_type": "stream",
|
| 769 |
-
"text": [
|
| 770 |
-
"CWE-119\n",
|
| 771 |
-
"CWE-89\n"
|
| 772 |
-
]
|
| 773 |
-
}
|
| 774 |
-
],
|
| 775 |
"source": [
|
| 776 |
"# Load the model\n",
|
| 777 |
"model = AutoModelForSequenceClassification.from_pretrained(\"artifacts/model\")\n",
|
|
@@ -790,10 +799,26 @@
|
|
| 790 |
},
|
| 791 |
{
|
| 792 |
"cell_type": "code",
|
| 793 |
-
"execution_count":
|
| 794 |
"id": "cdeaadbb",
|
| 795 |
"metadata": {},
|
| 796 |
-
"outputs": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 797 |
"source": [
|
| 798 |
"\n",
|
| 799 |
"print(predict_cwe(\"Buffer overflow in POP servers allows remote attackers to gain root access using a long PASS command.\"))\n",
|
|
|
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"cell_type": "code",
|
| 623 |
+
"execution_count": 32,
|
| 624 |
"id": "4b087fae",
|
| 625 |
"metadata": {},
|
| 626 |
"outputs": [
|
|
|
|
| 630 |
"\n",
|
| 631 |
" <div>\n",
|
| 632 |
" \n",
|
| 633 |
+
" <progress value='6978' max='6978' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
|
| 634 |
+
" [6978/6978 1:18:31, Epoch 1/1]\n",
|
| 635 |
" </div>\n",
|
| 636 |
" <table border=\"1\" class=\"dataframe\">\n",
|
| 637 |
" <thead>\n",
|
|
|
|
| 672 |
" <td>0.720193</td>\n",
|
| 673 |
" <td>0.242404</td>\n",
|
| 674 |
" </tr>\n",
|
| 675 |
+
" <tr>\n",
|
| 676 |
+
" <td>5000</td>\n",
|
| 677 |
+
" <td>1.110300</td>\n",
|
| 678 |
+
" <td>1.111053</td>\n",
|
| 679 |
+
" <td>0.722934</td>\n",
|
| 680 |
+
" <td>0.244389</td>\n",
|
| 681 |
+
" </tr>\n",
|
| 682 |
+
" <tr>\n",
|
| 683 |
+
" <td>6000</td>\n",
|
| 684 |
+
" <td>1.134700</td>\n",
|
| 685 |
+
" <td>1.082806</td>\n",
|
| 686 |
+
" <td>0.727207</td>\n",
|
| 687 |
+
" <td>0.251264</td>\n",
|
| 688 |
+
" </tr>\n",
|
| 689 |
" </tbody>\n",
|
| 690 |
"</table><p>"
|
| 691 |
],
|
|
|
|
| 695 |
},
|
| 696 |
"metadata": {},
|
| 697 |
"output_type": "display_data"
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"data": {
|
| 701 |
+
"text/plain": [
|
| 702 |
+
"TrainOutput(global_step=6978, training_loss=1.1011348515535433, metrics={'train_runtime': 4712.2885, 'train_samples_per_second': 23.691, 'train_steps_per_second': 1.481, 'total_flos': 2912105519756448.0, 'train_loss': 1.1011348515535433, 'epoch': 1.0})"
|
| 703 |
+
]
|
| 704 |
+
},
|
| 705 |
+
"execution_count": 32,
|
| 706 |
+
"metadata": {},
|
| 707 |
+
"output_type": "execute_result"
|
| 708 |
}
|
| 709 |
],
|
| 710 |
"source": [
|
|
|
|
| 713 |
},
|
| 714 |
{
|
| 715 |
"cell_type": "code",
|
| 716 |
+
"execution_count": 33,
|
| 717 |
"id": "48faf17c",
|
| 718 |
"metadata": {},
|
| 719 |
"outputs": [],
|
|
|
|
| 729 |
},
|
| 730 |
{
|
| 731 |
"cell_type": "code",
|
| 732 |
+
"execution_count": 34,
|
| 733 |
"id": "fcb11390",
|
| 734 |
"metadata": {},
|
| 735 |
"outputs": [
|
| 736 |
{
|
| 737 |
"data": {
|
| 738 |
"application/vnd.jupyter.widget-view+json": {
|
| 739 |
+
"model_id": "a33b852d1c594a69974bb9c3d30c014a",
|
| 740 |
"version_major": 2,
|
| 741 |
"version_minor": 0
|
| 742 |
},
|
| 743 |
"text/plain": [
|
| 744 |
+
"README.md: 0.00B [00:00, ?B/s]"
|
| 745 |
]
|
| 746 |
},
|
| 747 |
"metadata": {},
|
| 748 |
"output_type": "display_data"
|
| 749 |
},
|
| 750 |
{
|
| 751 |
+
"name": "stderr",
|
| 752 |
+
"output_type": "stream",
|
| 753 |
+
"text": [
|
| 754 |
+
"No files have been modified since last commit. Skipping to prevent empty commit.\n",
|
| 755 |
+
"No files have been modified since last commit. Skipping to prevent empty commit.\n"
|
| 756 |
+
]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 757 |
},
|
| 758 |
{
|
| 759 |
"name": "stdout",
|
|
|
|
| 777 |
},
|
| 778 |
{
|
| 779 |
"cell_type": "code",
|
| 780 |
+
"execution_count": 35,
|
| 781 |
"id": "33847880",
|
| 782 |
"metadata": {},
|
| 783 |
+
"outputs": [],
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 784 |
"source": [
|
| 785 |
"# Load the model\n",
|
| 786 |
"model = AutoModelForSequenceClassification.from_pretrained(\"artifacts/model\")\n",
|
|
|
|
| 799 |
},
|
| 800 |
{
|
| 801 |
"cell_type": "code",
|
| 802 |
+
"execution_count": 36,
|
| 803 |
"id": "cdeaadbb",
|
| 804 |
"metadata": {},
|
| 805 |
+
"outputs": [
|
| 806 |
+
{
|
| 807 |
+
"name": "stdout",
|
| 808 |
+
"output_type": "stream",
|
| 809 |
+
"text": [
|
| 810 |
+
"CWE-119\n",
|
| 811 |
+
"CWE-89\n",
|
| 812 |
+
"CWE-79\n",
|
| 813 |
+
"CWE-287\n",
|
| 814 |
+
"CWE-22\n",
|
| 815 |
+
"CWE-190\n",
|
| 816 |
+
"CWE-401\n",
|
| 817 |
+
"CWE-77\n",
|
| 818 |
+
"CWE-326\n"
|
| 819 |
+
]
|
| 820 |
+
}
|
| 821 |
+
],
|
| 822 |
"source": [
|
| 823 |
"\n",
|
| 824 |
"print(predict_cwe(\"Buffer overflow in POP servers allows remote attackers to gain root access using a long PASS command.\"))\n",
|