{"cells":[{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"import numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"train = pd.read_csv('/kaggle/input/jigsaw-multilingual-toxic-comment-classification/jigsaw-toxic-comment-train.csv')\ntrain1 = pd.read_csv('/kaggle/input/jigsaw-multilingual-toxic-comment-classification/jigsaw-unintended-bias-train.csv')\n\nvalid = pd.read_csv('/kaggle/input/jigsaw-multilingual-toxic-comment-classification/validation.csv')\n#valid = pd.read_csv('/kaggle/input/jigsaw-multilingual-toxic-test-translated/jigsaw_miltilingual_valid_translated.csv')\n\n#test = pd.read_csv('/kaggle/input/jigsaw-multilingual-toxic-comment-classification/test.csv')\ntest = pd.read_csv('/kaggle/input/jigsaw-multilingual-toxic-test-translated/jigsaw_miltilingual_test_translated.csv')\n\n#subm = pd.read_csv('/kaggle/input/jigsaw-multilingual-toxic-comment-classification/sample_submission.csv')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"test","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"import spacy \nspacy.prefer_gpu() # or spacy.require_gpu()\n\nnlp = spacy.load(\"en_core_web_sm\")\n\n# Create an empty model\n#nlp = spacy.blank(\"en\")\n\n# Create the TextCategorizer with exclusive classes and \"bow\" architecture\ntextcat = nlp.create_pipe(\n              \"textcat\",\n              config={\n                \"exclusive_classes\": True,\n                \"architecture\": \"bow\"})\n\n# Add the TextCategorizer to the empty model\nnlp.add_pipe(textcat)\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Add labels to text classifier\ntextcat.add_label(\"toxic\")\ntextcat.add_label(\"neutral\")\n#train=train[:1000]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_texts = train['comment_text'].values\ntrain_labels = [{'cats': {'toxic': label == 1,\n                          'neutral': label == 0}} \n                for label in train['toxic']]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_data = list(zip(train_texts, train_labels))\ntrain_data[:3]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"from spacy.util import minibatch\n\nspacy.util.fix_random_seed(1)\noptimizer = nlp.begin_training(n_threads=4)\n\n# Create the batch generator with batch size = 8\nbatches = minibatch(train_data, size=12)\n# Iterate through minibatches\nfor batch in batches:\n    # Each batch is a list of (text, label) but we need to\n    # send separate lists for texts and labels to update().\n    # This is a quick way to split a list of tuples into lists\n    texts, labels = zip(*batch)\n    nlp.update(texts, labels, sgd=optimizer)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"import random\n\nrandom.seed(1)\nspacy.util.fix_random_seed(1)\noptimizer = nlp.begin_training(n_threads=4)\n\nlosses = {}\nfor epoch in range(3):\n    random.shuffle(train_data)\n    # Create the batch generator with batch size = 8\n    batches = minibatch(train_data, size=500)\n    # Iterate through minibatches\n    for batch in batches:\n        # Each batch is a list of (text, label) but we need to\n        # send separate lists for texts and labels to update().\n        # This is a quick way to split a list of tuples into lists\n        texts, labels = zip(*batch)\n        nlp.update(texts, labels, sgd=optimizer, losses=losses)\n    print(losses)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"texts = [\"Are you ready for the tea party????? It's gonna be wild\",\n         \"URGENT Reply to this message for GUARANTEED FREE TEA\" ]\ndocs = [nlp.tokenizer(text) for text in texts]\n    \n# Use textcat to get the scores for each doc\ntextcat = nlp.get_pipe('textcat')\nscores, _ = textcat.predict(docs)\n\nprint(scores)\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"docs = [nlp.tokenizer(text) for text in test.translated]\n    \n# Use textcat to get the scores for each doc\ntextcat = nlp.get_pipe('textcat')\nscores, _ = textcat.predict(docs)\n\nprint(scores)\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"scores.shape,test.shape,scores[:,0].shape","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submit=pd.DataFrame( test.id )\nsubmit['toxic']=pd.DataFrame(scores[:,0].reshape(-1,1))\nsubmit.to_csv('submission.csv',index=False)\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submit","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":4}