-
-
Notifications
You must be signed in to change notification settings - Fork 440
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
21 changed files
with
1,493 additions
and
660 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Large diffs are not rendered by default.
Oops, something went wrong.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,270 @@ | ||
{ | ||
"cells": [ | ||
{ | ||
"cell_type": "markdown", | ||
"metadata": { | ||
"pycharm": { | ||
"name": "#%% md\n" | ||
} | ||
}, | ||
"source": [ | ||
"# Multi-Label Classification Benchmark\n", | ||
"\n", | ||
"- Kashgari: 2.0.0\n", | ||
"- TensorFlow: 2.0.0\n" | ||
] | ||
}, | ||
{ | ||
"cell_type": "markdown", | ||
"source": [ | ||
"## Macros" | ||
], | ||
"metadata": { | ||
"collapsed": false, | ||
"pycharm": { | ||
"name": "#%% md\n" | ||
} | ||
} | ||
}, | ||
{ | ||
"cell_type": "code", | ||
"execution_count": null, | ||
"outputs": [], | ||
"source": [ | ||
"CORPUS_PATH = '/Users/brikerman/Downloads/jigsaw-toxic-comment-classification-challenge/train.csv'" | ||
], | ||
"metadata": { | ||
"collapsed": false, | ||
"pycharm": { | ||
"name": "#%%\n" | ||
} | ||
} | ||
}, | ||
{ | ||
"cell_type": "code", | ||
"execution_count": null, | ||
"outputs": [], | ||
"source": [ | ||
"from kashgari.corpus import JigsawToxicCommentCorpus\n", | ||
"corpus = JigsawToxicCommentCorpus(CORPUS_PATH)\n", | ||
"\n", | ||
"train_x, train_y = corpus.load_data()\n", | ||
"test_x, test_y = corpus.load_data('test')\n", | ||
"valid_x, valid_y = corpus.load_data('valid')" | ||
], | ||
"metadata": { | ||
"collapsed": false, | ||
"pycharm": { | ||
"name": "#%%\n" | ||
} | ||
} | ||
}, | ||
{ | ||
"cell_type": "code", | ||
"execution_count": 9, | ||
"metadata": {}, | ||
"outputs": [ | ||
{ | ||
"name": "stderr", | ||
"output_type": "stream", | ||
"text": [ | ||
"Preparing text vocab dict: 100%|██████████| 3510/3510 [00:00<00:00, 43976.22it/s]\n", | ||
"INFO:root:------ Build vocab dict finished, Top 10 token ------\n", | ||
"INFO:root:Token: [PAD] -> 0\n", | ||
"INFO:root:Token: [UNK] -> 1\n", | ||
"INFO:root:Token: [BOS] -> 2\n", | ||
"INFO:root:Token: [EOS] -> 3\n", | ||
"INFO:root:Token: . -> 4\n", | ||
"INFO:root:Token: the -> 5\n", | ||
"INFO:root:Token: , -> 6\n", | ||
"INFO:root:Token: \" -> 7\n", | ||
"INFO:root:Token: to -> 8\n", | ||
"INFO:root:Token: i -> 9\n", | ||
"INFO:root:------ Build vocab dict finished, Top 10 token ------\n", | ||
"Preparing classification label vocab dict: 100%|██████████| 3510/3510 [00:00<00:00, 938723.91it/s]\n", | ||
"Calculating sequence length: 100%|██████████| 3510/3510 [00:00<00:00, 871846.92it/s]\n", | ||
"WARNING:root:Calculated sequence length = 282\n" | ||
] | ||
}, | ||
{ | ||
"name": "stdout", | ||
"output_type": "stream", | ||
"text": [ | ||
"Model: \"model_7\"\n", | ||
"_________________________________________________________________\n", | ||
"Layer (type) Output Shape Param # \n", | ||
"=================================================================\n", | ||
"input (InputLayer) [(None, None)] 0 \n", | ||
"_________________________________________________________________\n", | ||
"layer_embedding (Embedding) (None, None, 100) 666700 \n", | ||
"_________________________________________________________________\n", | ||
"bidirectional_3 (Bidirection (None, 256) 234496 \n", | ||
"_________________________________________________________________\n", | ||
"dense_3 (Dense) (None, 6) 1542 \n", | ||
"_________________________________________________________________\n", | ||
"activation_3 (Activation) (None, 6) 0 \n", | ||
"=================================================================\n", | ||
"Total params: 902,738\n", | ||
"Trainable params: 902,738\n", | ||
"Non-trainable params: 0\n", | ||
"_________________________________________________________________\n", | ||
"Train for 54 steps, validate for 11 steps\n", | ||
"Epoch 1/2\n", | ||
"53/54 [============================>.] - ETA: 0s - loss: 0.2300 - accuracy: 0.9547" | ||
] | ||
}, | ||
{ | ||
"name": "stderr", | ||
"output_type": "stream", | ||
"text": [ | ||
"WARNING:root:Sequence length is None, will use the max length of the samples, which is 1038\n" | ||
] | ||
}, | ||
{ | ||
"name": "stdout", | ||
"output_type": "stream", | ||
"text": [ | ||
" precision recall f1-score support\n", | ||
" toxic 0.0000 0.0000 0.0000 81\n", | ||
" obscene 0.0000 0.0000 0.0000 44\n", | ||
" insult 0.0000 0.0000 0.0000 46\n", | ||
" identity_hate 0.0000 0.0000 0.0000 10\n", | ||
" severe_toxic 0.0000 0.0000 0.0000 8\n", | ||
" threat 0.0000 0.0000 0.0000 4\n", | ||
" macro avg 0.0000 0.0000 0.0000 193\n", | ||
"\n", | ||
"\n", | ||
"epoch: 0 precision: 0.000000, recall: 0.000000, f1-score: 0.000000\n", | ||
"54/54 [==============================] - 46s 855ms/step - loss: 0.2279 - accuracy: 0.9551 - val_loss: 0.1627 - val_accuracy: 0.9567\n" | ||
] | ||
}, | ||
{ | ||
"name": "stderr", | ||
"output_type": "stream", | ||
"text": [ | ||
"/Users/brikerman/Desktop/python/Kashgari2/venv/lib/python3.7/site-packages/sklearn/metrics/_classification.py:1272: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n", | ||
" _warn_prf(average, modifier, msg_start, len(result))\n" | ||
] | ||
}, | ||
{ | ||
"name": "stdout", | ||
"output_type": "stream", | ||
"text": [ | ||
"Epoch 2/2\n", | ||
"53/54 [============================>.] - ETA: 0s - loss: 0.1440 - accuracy: 0.9619 precision recall f1-score support\n", | ||
" toxic 0.0000 0.0000 0.0000 81\n", | ||
" obscene 0.0000 0.0000 0.0000 44\n", | ||
" insult 0.0000 0.0000 0.0000 46\n", | ||
" identity_hate 0.0000 0.0000 0.0000 10\n", | ||
" severe_toxic 0.0000 0.0000 0.0000 8\n", | ||
" threat 0.0000 0.0000 0.0000 4\n", | ||
" macro avg 0.0000 0.0000 0.0000 193\n", | ||
"\n", | ||
"\n", | ||
"epoch: 1 precision: 0.000000, recall: 0.000000, f1-score: 0.000000\n", | ||
"54/54 [==============================] - 40s 736ms/step - loss: 0.1422 - accuracy: 0.9624 - val_loss: 0.1673 - val_accuracy: 0.9573\n" | ||
] | ||
} | ||
], | ||
"source": [ | ||
"import logging\n", | ||
"from kashgari.tasks.classification import BiLSTM_Model\n", | ||
"from kashgari.callbacks import EvalCallBack\n", | ||
"\n", | ||
"logging.basicConfig(level='DEBUG')\n", | ||
"\n", | ||
"model = BiLSTM_Model(multi_label=True)\n", | ||
"\n", | ||
"eval_callback = EvalCallBack(model,\n", | ||
" test_x,\n", | ||
" test_y,\n", | ||
" step=1)\n", | ||
"\n", | ||
"x = model.fit(train_x, train_y, valid_x, valid_y,\n", | ||
" epochs=2,\n", | ||
" callbacks=[eval_callback])" | ||
] | ||
}, | ||
{ | ||
"cell_type": "code", | ||
"execution_count": 13, | ||
"metadata": { | ||
"pycharm": { | ||
"name": "#%%\n" | ||
} | ||
}, | ||
"outputs": [ | ||
{ | ||
"name": "stdout", | ||
"output_type": "stream", | ||
"text": [ | ||
" precision recall f1-score support\n", | ||
" toxic 0.0000 0.0000 0.0000 351\n", | ||
" obscene 0.0000 0.0000 0.0000 182\n", | ||
" insult 0.0000 0.0000 0.0000 181\n", | ||
" identity_hate 0.0000 0.0000 0.0000 34\n", | ||
" severe_toxic 0.0000 0.0000 0.0000 33\n", | ||
" threat 0.0000 0.0000 0.0000 13\n", | ||
" macro avg 0.0000 0.0000 0.0000 794\n", | ||
"\n" | ||
] | ||
}, | ||
{ | ||
"name": "stderr", | ||
"output_type": "stream", | ||
"text": [ | ||
"/Users/brikerman/Desktop/python/Kashgari2/venv/lib/python3.7/site-packages/sklearn/metrics/_classification.py:1272: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n", | ||
" _warn_prf(average, modifier, msg_start, len(result))\n" | ||
] | ||
}, | ||
{ | ||
"data": { | ||
"text/plain": [ | ||
"{'precision': 0.0,\n", | ||
" 'recall': 0.0,\n", | ||
" 'f1-score': 0.0,\n", | ||
" 'support': 794,\n", | ||
" 'detail': {'toxic': {'precision': 0.0,\n", | ||
" 'recall': 0.0,\n", | ||
" 'f1': 0.0,\n", | ||
" 'support': 351},\n", | ||
" 'obscene': {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'support': 182},\n", | ||
" 'insult': {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'support': 181},\n", | ||
" 'identity_hate': {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'support': 34},\n", | ||
" 'severe_toxic': {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'support': 33},\n", | ||
" 'threat': {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'support': 13}}}" | ||
] | ||
}, | ||
"execution_count": 13, | ||
"metadata": {}, | ||
"output_type": "execute_result" | ||
} | ||
], | ||
"source": [ | ||
"model.evaluate(train_x, train_y)\n", | ||
"\n" | ||
] | ||
} | ||
], | ||
"metadata": { | ||
"kernelspec": { | ||
"name": "python37464bitvenvvenv1fddca7c786445709a8f03a12aa91dfc", | ||
"language": "python", | ||
"display_name": "Python 3.7.4 64-bit ('venv': venv)" | ||
}, | ||
"language_info": { | ||
"codemirror_mode": { | ||
"name": "ipython", | ||
"version": 3 | ||
}, | ||
"file_extension": ".py", | ||
"mimetype": "text/x-python", | ||
"name": "python", | ||
"nbconvert_exporter": "python", | ||
"pygments_lexer": "ipython3", | ||
"version": "3.7.4" | ||
} | ||
}, | ||
"nbformat": 4, | ||
"nbformat_minor": 4 | ||
} |
Oops, something went wrong.