nepsa-nepner-inference.ipynb

{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"none","dataSources":[{"sourceId":7068905,"sourceType":"datasetVersion","datasetId":4070644},{"sourceId":7077325,"sourceType":"datasetVersion","datasetId":4076638},{"sourceId":7077374,"sourceType":"datasetVersion","datasetId":4076665}],"dockerImageVersionId":30587,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":false}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"%%capture\n!python3 -m pip install -U transformers\n!python3 -m pip install -U datasets\n!python3 -m pip install -U accelerate\n!python3 -m pip install -U wandb","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from transformers import pipeline\nfrom tabulate import tabulate\nimport pandas as pd\n\nlabel_map = {\n    0: 'GENERAL POSITIVE',\n    1: 'GENERAL NEGATIVE',\n    2: 'PROFANITY',\n    3: 'VIOLENCE'\n}\n\n# Sentiment Analysis\nnepsa_model='/kaggle/input/nepsa-distilbert-model'\ntext_classifier = pipeline(\"text-classification\", model=nepsa_model)\n\n# Named Entity Recognition\nnepner_model='/kaggle/input/nepner-roberta-model/nepsa_ner-roberta-epoch6'\ntoken_classifier = pipeline(\n    \"token-classification\", model=nepner_model, aggregation_strategy=\"simple\"\n)","metadata":{"execution":{"iopub.status.busy":"2023-12-06T02:48:59.729480Z","iopub.execute_input":"2023-12-06T02:48:59.729886Z","iopub.status.idle":"2023-12-06T02:49:29.630735Z","shell.execute_reply.started":"2023-12-06T02:48:59.729845Z","shell.execute_reply":"2023-12-06T02:49:29.629488Z"},"trusted":true},"execution_count":1,"outputs":[{"name":"stderr","text":"/opt/conda/lib/python3.10/site-packages/scipy/__init__.py:146: UserWarning: A NumPy version >=1.16.5 and <1.23.0 is required for this version of SciPy (detected version 1.24.3\n  warnings.warn(f\"A NumPy version >={np_minversion} and <{np_maxversion}\"\n","output_type":"stream"}]},{"cell_type":"markdown","source":"## Perform Offensive Sentiment Classification","metadata":{}},{"cell_type":"code","source":"sentence = \"भ्रश्ट्चारी हरुलाइ टुंडिखेल मा ल्याइ झुन्ड्याएर गोलि ठोक्नु पर्छ ।\"\n\n# The corrupt should be brought to Tundikhel and shot.\n\nresults = text_classifier(sentence)[0]\nprediction_results = []\npred = results['label'].split('_')[1]\n\nprediction_results.append([sentence, pred, label_map[int(pred)]])\n\nprint(\"Sentence:\", sentence)\npd.DataFrame(prediction_results, columns=['Sentences', 'Predicted Label', 'Remarks'])","metadata":{"execution":{"iopub.status.busy":"2023-12-06T02:50:26.795730Z","iopub.execute_input":"2023-12-06T02:50:26.796165Z","iopub.status.idle":"2023-12-06T02:50:26.857557Z","shell.execute_reply.started":"2023-12-06T02:50:26.796131Z","shell.execute_reply":"2023-12-06T02:50:26.856481Z"},"trusted":true},"execution_count":11,"outputs":[{"name":"stdout","text":"Sentence: भ्रश्ट्चारी हरुलाइ टुंडिखेल मा ल्याइ झुन्ड्याएर गोलि ठोक्नु पर्छ ।\n","output_type":"stream"},{"execution_count":11,"output_type":"execute_result","data":{"text/plain":"                                           Sentences Predicted Label   Remarks\n0  भ्रश्ट्चारी हरुलाइ टुंडिखेल मा ल्याइ झुन्ड्याए...               3  VIOLENCE","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Sentences</th>\n      <th>Predicted Label</th>\n      <th>Remarks</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>भ्रश्ट्चारी हरुलाइ टुंडिखेल मा ल्याइ झुन्ड्याए...</td>\n      <td>3</td>\n      <td>VIOLENCE</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"markdown","source":"","metadata":{}},{"cell_type":"markdown","source":"## Perform Offensive Entity Recognition ","metadata":{}},{"cell_type":"code","source":"results = token_classifier(sentence)\n\nner_results = []\nfor each_entity in results:\n    ner_results.append([each_entity['word'], each_entity['entity_group']])\n\nprint(\"Sentence:\", sentence)\nprint(\"English translation: The corrupt should be brought to Tundikhel and shot.\")\npd.DataFrame(ner_results, columns=['Word', 'Predictions'])","metadata":{"execution":{"iopub.status.busy":"2023-12-06T02:51:03.579153Z","iopub.execute_input":"2023-12-06T02:51:03.579621Z","iopub.status.idle":"2023-12-06T02:51:03.712055Z","shell.execute_reply.started":"2023-12-06T02:51:03.579582Z","shell.execute_reply":"2023-12-06T02:51:03.710925Z"},"trusted":true},"execution_count":13,"outputs":[{"name":"stdout","text":"Sentence: भ्रश्ट्चारी हरुलाइ टुंडिखेल मा ल्याइ झुन्ड्याएर गोलि ठोक्नु पर्छ ।\nEnglish translation: The corrupt should be brought to Tundikhel and shot.\n","output_type":"stream"},{"execution_count":13,"output_type":"execute_result","data":{"text/plain":"               Word Predictions\n0       भ्रश्ट्चारी     GENERAL\n1                टु         LOC\n2            ंडिखेल         ORG\n3  गोलि ठोक्नु पर्छ    VIOLENCE","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Word</th>\n      <th>Predictions</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>भ्रश्ट्चारी</td>\n      <td>GENERAL</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>टु</td>\n      <td>LOC</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>ंडिखेल</td>\n      <td>ORG</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>गोलि ठोक्नु पर्छ</td>\n      <td>VIOLENCE</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"markdown","source":"Our model performs pretty good.","metadata":{}},{"cell_type":"markdown","source":"## Perform testing on curated out of domain test set","metadata":{}},{"cell_type":"code","source":"!wget https://raw.githubusercontent.com/merishnaSuwal/nep-off-langdetect/main/NepSA/dataset/out-of-domain/dataset.txt","metadata":{"execution":{"iopub.status.busy":"2023-12-06T02:49:30.355446Z","iopub.execute_input":"2023-12-06T02:49:30.356101Z","iopub.status.idle":"2023-12-06T02:49:31.661337Z","shell.execute_reply.started":"2023-12-06T02:49:30.356050Z","shell.execute_reply":"2023-12-06T02:49:31.659953Z"},"trusted":true},"execution_count":4,"outputs":[{"name":"stdout","text":"--2023-12-06 02:49:31--  https://raw.githubusercontent.com/merishnaSuwal/nep-off-langdetect/main/NepSA/dataset/out-of-domain/dataset.txt\nResolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.108.133, 185.199.109.133, 185.199.110.133, ...\nConnecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.108.133|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 6950 (6.8K) [text/plain]\nSaving to: ‘dataset.txt’\n\ndataset.txt         100%[===================>]   6.79K  --.-KB/s    in 0s      \n\n2023-12-06 02:49:31 (64.6 MB/s) - ‘dataset.txt’ saved [6950/6950]\n\n","output_type":"stream"}]},{"cell_type":"markdown","source":"## Out-of-domain inference and testing","metadata":{}},{"cell_type":"code","source":"import pandas as pd\n\ndata = pd.read_csv('/kaggle/working/dataset.txt', delimiter='\\t', names=['text', 'label'])\ndata.head()","metadata":{"execution":{"iopub.status.busy":"2023-12-06T02:49:31.662910Z","iopub.execute_input":"2023-12-06T02:49:31.663264Z","iopub.status.idle":"2023-12-06T02:49:31.683767Z","shell.execute_reply.started":"2023-12-06T02:49:31.663232Z","shell.execute_reply":"2023-12-06T02:49:31.682628Z"},"trusted":true},"execution_count":5,"outputs":[{"execution_count":5,"output_type":"execute_result","data":{"text/plain":"                                                text  label\n0  म लाई त यो मावबादी मनै पर्दैन तर गगन ले गरे को...      0\n1                       सुशील पान्दे is awesome ✌✌ ।      0\n2              प्रदिप यादव को कुरा एकदमै सहि छ , , ।      0\n3  जुन दिन साझा पार्टी खराब निस्कन्छ अनि औला ठडाउ...      1\n4            रातो लागाउनी सर को कुरो राम्रो लाग्यो ।      0","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>text</th>\n      <th>label</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>म लाई त यो मावबादी मनै पर्दैन तर गगन ले गरे को...</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>सुशील पान्दे is awesome ✌✌ ।</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>प्रदिप यादव को कुरा एकदमै सहि छ , , ।</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>जुन दिन साझा पार्टी खराब निस्कन्छ अनि औला ठडाउ...</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>रातो लागाउनी सर को कुरो राम्रो लाग्यो ।</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"data['label'].value_counts()","metadata":{"execution":{"iopub.status.busy":"2023-12-06T02:49:31.685347Z","iopub.execute_input":"2023-12-06T02:49:31.685829Z","iopub.status.idle":"2023-12-06T02:49:31.705140Z","shell.execute_reply.started":"2023-12-06T02:49:31.685785Z","shell.execute_reply":"2023-12-06T02:49:31.704004Z"},"trusted":true},"execution_count":6,"outputs":[{"execution_count":6,"output_type":"execute_result","data":{"text/plain":"label\n1    14\n0    13\n2     6\n3     3\nName: count, dtype: int64"},"metadata":{}}]},{"cell_type":"code","source":"sentences = list(data['text'].values)\n\nprediction_results = []\nfor sentence in sentences:\n    results = text_classifier(sentence)[0]\n    pred = results['label'].split('_')[1]\n\n    prediction_results.append([sentence, pred, label_map[int(pred)]])\n\npred_df = pd.DataFrame(prediction_results, columns=['Sentences', 'Predicted Label', 'Remarks'])\npred_df['Predicted Label'] = pred_df['Predicted Label'].astype(int)\npred_df['True_label'] = data['label'].values","metadata":{"execution":{"iopub.status.busy":"2023-12-06T02:49:31.706475Z","iopub.execute_input":"2023-12-06T02:49:31.706894Z","iopub.status.idle":"2023-12-06T02:49:33.135129Z","shell.execute_reply.started":"2023-12-06T02:49:31.706854Z","shell.execute_reply":"2023-12-06T02:49:33.134126Z"},"trusted":true},"execution_count":7,"outputs":[]},{"cell_type":"code","source":"pred_df.head()","metadata":{"execution":{"iopub.status.busy":"2023-12-06T02:49:33.136600Z","iopub.execute_input":"2023-12-06T02:49:33.137669Z","iopub.status.idle":"2023-12-06T02:49:33.156069Z","shell.execute_reply.started":"2023-12-06T02:49:33.137625Z","shell.execute_reply":"2023-12-06T02:49:33.154690Z"},"trusted":true},"execution_count":8,"outputs":[{"execution_count":8,"output_type":"execute_result","data":{"text/plain":"                                            Sentences  Predicted Label  \\\n0   म लाई त यो मावबादी मनै पर्दैन तर गगन ले गरे को...                0   \n1                        सुशील पान्दे is awesome ✌✌ ।                0   \n2               प्रदिप यादव को कुरा एकदमै सहि छ , , ।                0   \n3   जुन दिन साझा पार्टी खराब निस्कन्छ अनि औला ठडाउ...                0   \n4             रातो लागाउनी सर को कुरो राम्रो लाग्यो ।                0   \n5   धमला अरु को ( बोलाए को गेस्ट को ) कुरा सुन्न न...                1   \n6   यो मुजि केपि को चम्चा लाइ नि गोलि खाना मन लायो...                2   \n7           थुक्क पोखरेल भन्ने माहाभ्रष्ट को कुतर्क ।                1   \n8   आफ्नो गरिमामय पद को पनि ख्याल नगरी बोल्ने व्यत...                1   \n9               यो सति ले सरापे को देश मा भलो हुन्न ।                1   \n10  यो कस्तो पत्रु मान्छे , नालायक धिक्कार होसा ते...                2   \n11  यो जाठो पुलिस प्रसासन नै गतिलो छैन सब मिलिमतो ...                2   \n12  यी सबै भारत को कुकुर हरु हुन अब जन्ता मिलेर फा...                2   \n13  रवि दाई पृथ्वी को सतह हुन , तिन करोड नेपाली oz...                1   \n14  देस र जन्ता को लागि सोच्ने महान सपूत रबि सर ला...                0   \n15                   भारत को दलाल नै हुन दुई जना नै ।                1   \n16  सागर सर तपाई मा देखिएकाे राष्ट्रभक्ति लाई सलाम...                0   \n17  चिनी आँउदा खुसी हुने भर्ती आउँदा रिस उठ्ने को ...                1   \n18  यस्ता पत्रकार लाई हटाउनु पर्छ यो के सि पनि गफड...                2   \n19  भ्रष्टाचार ले चुर्लुम्मै डुबे को देश नेपाल थुई...                2   \n20  कुनै पनि कारयलये मा धुस मागे भने चाक मा आगो लग...                3   \n21  कुरो बुझ्दै जादा राम बहादुर बम्जन को आश्रम मा ...                3   \n22              यस्ता नेता भय पो हून्छ देस मा बिकास ।                1   \n23  केन्द्र देखि जिल्ला जिल्ला सम्म वर्दीधारी गुन्...                1   \n24  सारै खुसी लाग्यो यथार्थ बोल्नु भयो निर्दोस रबि...                1   \n25  बिना रक्सी , मासु , पेट्रोल र पैशा को समर्थन र...                1   \n26  यदि यस्तै पारा हो भने त्यों दिन आउन बेर लाग्दै...                1   \n27  म हान्छु गोली अब आयेर सबै चोरहरुलाइ याद गर्दै ...                3   \n28  100% मा 95 रबी दाई को जनता हून् अरू चही दूसमन ...                1   \n29  कुन घाँस हो प्रश्न सोध्ने दोषि ले न्याए पाउनु ...                1   \n30  गगन थापा धेरै राम्रो नेता हुन इन् लाई म सलुट ग...                0   \n31  यो कुकुर पत्रकार चुतिया बोल्ने शैली सोध्ने प्र...                2   \n32                    धेरै देखियो भांडतन्त्र का खेल ।                1   \n33  अब केहि हुन्छ नेपाल मा , नया यूग को सुरूवात भै...                0   \n34  अब को केही दिन पछि हाम्रो रबि दाजु लाई नेपाल क...                0   \n35  यो सोम भन्ने खाते दलाल लाई जहा भेट्यो तेही काल...                2   \n\n             Remarks  True_label  \n0   GENERAL POSITIVE           0  \n1   GENERAL POSITIVE           0  \n2   GENERAL POSITIVE           0  \n3   GENERAL POSITIVE           1  \n4   GENERAL POSITIVE           0  \n5   GENERAL NEGATIVE           1  \n6          PROFANITY           2  \n7   GENERAL NEGATIVE           1  \n8   GENERAL NEGATIVE           1  \n9   GENERAL NEGATIVE           1  \n10         PROFANITY           2  \n11         PROFANITY           2  \n12         PROFANITY           2  \n13  GENERAL NEGATIVE           0  \n14  GENERAL POSITIVE           0  \n15  GENERAL NEGATIVE           1  \n16  GENERAL POSITIVE           0  \n17  GENERAL NEGATIVE           1  \n18         PROFANITY           1  \n19         PROFANITY           3  \n20          VIOLENCE           3  \n21          VIOLENCE           1  \n22  GENERAL NEGATIVE           0  \n23  GENERAL NEGATIVE           1  \n24  GENERAL NEGATIVE           0  \n25  GENERAL NEGATIVE           0  \n26  GENERAL NEGATIVE           1  \n27          VIOLENCE           2  \n28  GENERAL NEGATIVE           1  \n29  GENERAL NEGATIVE           1  \n30  GENERAL POSITIVE           0  \n31         PROFANITY           2  \n32  GENERAL NEGATIVE           1  \n33  GENERAL POSITIVE           0  \n34  GENERAL POSITIVE           0  \n35         PROFANITY           3  ","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Sentences</th>\n      <th>Predicted Label</th>\n      <th>Remarks</th>\n      <th>True_label</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>म लाई त यो मावबादी मनै पर्दैन तर गगन ले गरे को...</td>\n      <td>0</td>\n      <td>GENERAL POSITIVE</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>सुशील पान्दे is awesome ✌✌ ।</td>\n      <td>0</td>\n      <td>GENERAL POSITIVE</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>प्रदिप यादव को कुरा एकदमै सहि छ , , ।</td>\n      <td>0</td>\n      <td>GENERAL POSITIVE</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>जुन दिन साझा पार्टी खराब निस्कन्छ अनि औला ठडाउ...</td>\n      <td>0</td>\n      <td>GENERAL POSITIVE</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>रातो लागाउनी सर को कुरो राम्रो लाग्यो ।</td>\n      <td>0</td>\n      <td>GENERAL POSITIVE</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>धमला अरु को ( बोलाए को गेस्ट को ) कुरा सुन्न न...</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>यो मुजि केपि को चम्चा लाइ नि गोलि खाना मन लायो...</td>\n      <td>2</td>\n      <td>PROFANITY</td>\n      <td>2</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>थुक्क पोखरेल भन्ने माहाभ्रष्ट को कुतर्क ।</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>आफ्नो गरिमामय पद को पनि ख्याल नगरी बोल्ने व्यत...</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>9</th>\n      <td>यो सति ले सरापे को देश मा भलो हुन्न ।</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>10</th>\n      <td>यो कस्तो पत्रु मान्छे , नालायक धिक्कार होसा ते...</td>\n      <td>2</td>\n      <td>PROFANITY</td>\n      <td>2</td>\n    </tr>\n    <tr>\n      <th>11</th>\n      <td>यो जाठो पुलिस प्रसासन नै गतिलो छैन सब मिलिमतो ...</td>\n      <td>2</td>\n      <td>PROFANITY</td>\n      <td>2</td>\n    </tr>\n    <tr>\n      <th>12</th>\n      <td>यी सबै भारत को कुकुर हरु हुन अब जन्ता मिलेर फा...</td>\n      <td>2</td>\n      <td>PROFANITY</td>\n      <td>2</td>\n    </tr>\n    <tr>\n      <th>13</th>\n      <td>रवि दाई पृथ्वी को सतह हुन , तिन करोड नेपाली oz...</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>14</th>\n      <td>देस र जन्ता को लागि सोच्ने महान सपूत रबि सर ला...</td>\n      <td>0</td>\n      <td>GENERAL POSITIVE</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>15</th>\n      <td>भारत को दलाल नै हुन दुई जना नै ।</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>16</th>\n      <td>सागर सर तपाई मा देखिएकाे राष्ट्रभक्ति लाई सलाम...</td>\n      <td>0</td>\n      <td>GENERAL POSITIVE</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>17</th>\n      <td>चिनी आँउदा खुसी हुने भर्ती आउँदा रिस उठ्ने को ...</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>18</th>\n      <td>यस्ता पत्रकार लाई हटाउनु पर्छ यो के सि पनि गफड...</td>\n      <td>2</td>\n      <td>PROFANITY</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>19</th>\n      <td>भ्रष्टाचार ले चुर्लुम्मै डुबे को देश नेपाल थुई...</td>\n      <td>2</td>\n      <td>PROFANITY</td>\n      <td>3</td>\n    </tr>\n    <tr>\n      <th>20</th>\n      <td>कुनै पनि कारयलये मा धुस मागे भने चाक मा आगो लग...</td>\n      <td>3</td>\n      <td>VIOLENCE</td>\n      <td>3</td>\n    </tr>\n    <tr>\n      <th>21</th>\n      <td>कुरो बुझ्दै जादा राम बहादुर बम्जन को आश्रम मा ...</td>\n      <td>3</td>\n      <td>VIOLENCE</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>22</th>\n      <td>यस्ता नेता भय पो हून्छ देस मा बिकास ।</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>23</th>\n      <td>केन्द्र देखि जिल्ला जिल्ला सम्म वर्दीधारी गुन्...</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>24</th>\n      <td>सारै खुसी लाग्यो यथार्थ बोल्नु भयो निर्दोस रबि...</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>25</th>\n      <td>बिना रक्सी , मासु , पेट्रोल र पैशा को समर्थन र...</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>26</th>\n      <td>यदि यस्तै पारा हो भने त्यों दिन आउन बेर लाग्दै...</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>27</th>\n      <td>म हान्छु गोली अब आयेर सबै चोरहरुलाइ याद गर्दै ...</td>\n      <td>3</td>\n      <td>VIOLENCE</td>\n      <td>2</td>\n    </tr>\n    <tr>\n      <th>28</th>\n      <td>100% मा 95 रबी दाई को जनता हून् अरू चही दूसमन ...</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>29</th>\n      <td>कुन घाँस हो प्रश्न सोध्ने दोषि ले न्याए पाउनु ...</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>30</th>\n      <td>गगन थापा धेरै राम्रो नेता हुन इन् लाई म सलुट ग...</td>\n      <td>0</td>\n      <td>GENERAL POSITIVE</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>31</th>\n      <td>यो कुकुर पत्रकार चुतिया बोल्ने शैली सोध्ने प्र...</td>\n      <td>2</td>\n      <td>PROFANITY</td>\n      <td>2</td>\n    </tr>\n    <tr>\n      <th>32</th>\n      <td>धेरै देखियो भांडतन्त्र का खेल ।</td>\n      <td>1</td>\n      <td>GENERAL NEGATIVE</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>33</th>\n      <td>अब केहि हुन्छ नेपाल मा , नया यूग को सुरूवात भै...</td>\n      <td>0</td>\n      <td>GENERAL POSITIVE</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>34</th>\n      <td>अब को केही दिन पछि हाम्रो रबि दाजु लाई नेपाल क...</td>\n      <td>0</td>\n      <td>GENERAL POSITIVE</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>35</th>\n      <td>यो सोम भन्ने खाते दलाल लाई जहा भेट्यो तेही काल...</td>\n      <td>2</td>\n      <td>PROFANITY</td>\n      <td>3</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"markdown","source":"## Clasification report and analysis","metadata":{}},{"cell_type":"code","source":"from sklearn.metrics import classification_report, confusion_matrix\n\n# Assuming predictions and true_labels are your model's predictions and true labels\npredictions = pred_df['Predicted Label'].values  # Replace with your actual prediction logic\ntrue_labels = pred_df['True_label'].values  # Replace with your true labels\n\n# Classification Report\nclass_report = classification_report(true_labels, predictions)\nprint(\"Classification Report:\")\nprint(class_report)\n\n# Confusion Matrix\nconf_matrix = confusion_matrix(true_labels, predictions)\nprint(\"Confusion Matrix:\")\nprint(conf_matrix)\n\n# Individual misclassified instances\nmisclassified_indices = [i for i in range(len(true_labels)) if true_labels[i] != predictions[i]]\nprint(\"Misclassified Instances:\")\nfor idx in misclassified_indices:\n    print(f\"Actual: {true_labels[idx]}, Predicted: {predictions[idx]}\")\n","metadata":{"execution":{"iopub.status.busy":"2023-12-06T02:49:33.157844Z","iopub.execute_input":"2023-12-06T02:49:33.158601Z","iopub.status.idle":"2023-12-06T02:49:33.184378Z","shell.execute_reply.started":"2023-12-06T02:49:33.158553Z","shell.execute_reply":"2023-12-06T02:49:33.183348Z"},"trusted":true},"execution_count":9,"outputs":[{"name":"stdout","text":"Classification Report:\n              precision    recall  f1-score   support\n\n           0       0.90      0.69      0.78        13\n           1       0.73      0.79      0.76        14\n           2       0.62      0.83      0.71         6\n           3       0.33      0.33      0.33         3\n\n    accuracy                           0.72        36\n   macro avg       0.65      0.66      0.65        36\nweighted avg       0.74      0.72      0.72        36\n\nConfusion Matrix:\n[[ 9  4  0  0]\n [ 1 11  1  1]\n [ 0  0  5  1]\n [ 0  0  2  1]]\nMisclassified Instances:\nActual: 1, Predicted: 0\nActual: 0, Predicted: 1\nActual: 1, Predicted: 2\nActual: 3, Predicted: 2\nActual: 1, Predicted: 3\nActual: 0, Predicted: 1\nActual: 0, Predicted: 1\nActual: 0, Predicted: 1\nActual: 2, Predicted: 3\nActual: 3, Predicted: 2\n","output_type":"stream"}]},{"cell_type":"code","source":"import seaborn as sns\nimport matplotlib.pyplot as plt\n# Compute confusion matrix\ncm = confusion_matrix(true_labels, predictions)\n\n# Plot confusion matrix\nplt.figure(figsize=(8, 6))\nsns.heatmap(cm, annot=True, fmt=\"d\", cmap=\"Blues\", xticklabels=[\"0\", \"1\", \"2\", \"3\"], yticklabels=[\"0\", \"1\", \"2\", \"3\"])\nplt.title(\"Confusion Matrix\")\nplt.xlabel(\"Predicted Label\")\nplt.ylabel(\"True Label\")\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2023-12-06T02:49:33.188393Z","iopub.execute_input":"2023-12-06T02:49:33.189328Z","iopub.status.idle":"2023-12-06T02:49:33.721910Z","shell.execute_reply.started":"2023-12-06T02:49:33.189287Z","shell.execute_reply":"2023-12-06T02:49:33.720880Z"},"trusted":true},"execution_count":10,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 800x600 with 2 Axes>","image/png":"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"},"metadata":{}}]},{"cell_type":"markdown","source":"### Extra text for inference\n\nसुशील जि धन्यवाद जन्ता को आवाज बि ले को मा ।\n\nThank you Sushil ji for being the voice of the people.\n\n---\n\nयेस्ता मानब अधिकार कर्मि को काम छइन ... ।\n\nThis human rights activist is of no use.\n\n---\n\nयो खाते अधिवक्ता दिनेश त्रिपाठी को अावज सुन्न पनि मन पर्दैन ।\n\nI don't even like to hear the voice of this account advocate Dinesh Tripathi.\n\n---\n\nभ्रश्ट्चारी हरुलाइ टुंडिखेल मा ल्याइ झुन्ड्याएर गोलि ठोक्नु पर्छ ।\n\nThe corrupt should be brought to Tundikhel and shot.\n","metadata":{}}]}