genomics-project-starter-code.ipynb

{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.7.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load\n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nimport gc\n\n# Input data files are available in the read-only \"../input/\" directory\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n\n# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-12-18T07:44:40.723668Z","iopub.execute_input":"2022-12-18T07:44:40.724363Z","iopub.status.idle":"2022-12-18T07:44:40.763704Z","shell.execute_reply.started":"2022-12-18T07:44:40.724256Z","shell.execute_reply":"2022-12-18T07:44:40.762537Z"},"trusted":true},"execution_count":1,"outputs":[{"name":"stdout","text":"/kaggle/input/testx/cite_test_x.npy\n/kaggle/input/open-problems-multimodal/sample_submission.csv\n/kaggle/input/open-problems-multimodal/train_cite_targets.h5\n/kaggle/input/open-problems-multimodal/metadata_cite_day_2_donor_27678.csv\n/kaggle/input/open-problems-multimodal/test_multi_inputs.h5\n/kaggle/input/open-problems-multimodal/evaluation_ids.csv\n/kaggle/input/open-problems-multimodal/train_cite_inputs.h5\n/kaggle/input/open-problems-multimodal/train_multi_targets.h5\n/kaggle/input/open-problems-multimodal/train_multi_inputs.h5\n/kaggle/input/open-problems-multimodal/metadata.csv\n/kaggle/input/open-problems-multimodal/test_cite_inputs_day_2_donor_27678.h5\n/kaggle/input/open-problems-multimodal/test_cite_inputs.h5\n/kaggle/input/dataset/y_train.npy\n/kaggle/input/dataset/y_test.npy\n/kaggle/input/dataset/x_test.npy\n/kaggle/input/dataset/x_train.npy\n/kaggle/input/trainx/cite_train_x.npy\n","output_type":"stream"}]},{"cell_type":"markdown","source":"### Chromatin accessibility input and targets\nmultiome-->ATEC->Chromatin accesibility","metadata":{}},{"cell_type":"code","source":"hdf = pd.HDFStore('/kaggle/input/open-problems-multimodal/train_multi_inputs.h5',\n                  mode='r')\ndf_multi_input = hdf.get(hdf.keys()[0])\ndf_multi_input.head()","metadata":{"execution":{"iopub.status.busy":"2022-12-18T03:28:18.668115Z","iopub.execute_input":"2022-12-18T03:28:18.668499Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"gc.collect()\nhdf = pd.HDFStore('/kaggle/input/open-problems-multimodal/train_multi_targets.h5',\n                  mode='r')\ndf_multi_output = hdf.get(hdf.keys()[0])\ndf_multi_output.head()","metadata":{"execution":{"iopub.status.busy":"2022-12-18T03:31:40.936753Z","iopub.execute_input":"2022-12-18T03:31:40.937527Z","iopub.status.idle":"2022-12-18T03:31:41.160957Z","shell.execute_reply.started":"2022-12-18T03:31:40.937488Z","shell.execute_reply":"2022-12-18T03:31:41.159916Z"},"trusted":true},"execution_count":2,"outputs":[]},{"cell_type":"markdown","source":"### Gene expressioninput and targets\nCITESeq-->given RNA gene expression->Predict Protien levels","metadata":{}},{"cell_type":"code","source":"hdf = pd.HDFStore('/kaggle/input/open-problems-multimodal/train_cite_inputs.h5',\n                  mode='r')\ndf_cite_input = hdf.get(hdf.keys()[0])\ndf_cite_input.head()","metadata":{"execution":{"iopub.status.busy":"2022-12-18T03:45:02.521866Z","iopub.execute_input":"2022-12-18T03:45:02.522241Z","iopub.status.idle":"2022-12-18T03:45:50.602318Z","shell.execute_reply.started":"2022-12-18T03:45:02.522210Z","shell.execute_reply":"2022-12-18T03:45:50.601187Z"},"trusted":true},"execution_count":3,"outputs":[{"execution_count":3,"output_type":"execute_result","data":{"text/plain":"gene_id       ENSG00000121410_A1BG  ENSG00000268895_A1BG-AS1  \\\ncell_id                                                        \n45006fe3e4c8                   0.0                       0.0   \nd02759a80ba2                   0.0                       0.0   \nc016c6b0efa5                   0.0                       0.0   \nba7f733a4f75                   0.0                       0.0   \nfbcf2443ffb2                   0.0                       0.0   \n\ngene_id       ENSG00000175899_A2M  ENSG00000245105_A2M-AS1  \\\ncell_id                                                      \n45006fe3e4c8                  0.0                      0.0   \nd02759a80ba2                  0.0                      0.0   \nc016c6b0efa5                  0.0                      0.0   \nba7f733a4f75                  0.0                      0.0   \nfbcf2443ffb2                  0.0                      0.0   \n\ngene_id       ENSG00000166535_A2ML1  ENSG00000128274_A4GALT  \\\ncell_id                                                       \n45006fe3e4c8                    0.0                0.000000   \nd02759a80ba2                    0.0                0.000000   \nc016c6b0efa5                    0.0                3.847321   \nba7f733a4f75                    0.0                0.000000   \nfbcf2443ffb2                    0.0                0.000000   \n\ngene_id       ENSG00000094914_AAAS  ENSG00000081760_AACS  \\\ncell_id                                                    \n45006fe3e4c8              0.000000              0.000000   \nd02759a80ba2              0.000000              0.000000   \nc016c6b0efa5              0.000000              3.847321   \nba7f733a4f75              3.436846              3.436846   \nfbcf2443ffb2              0.000000              4.196826   \n\ngene_id       ENSG00000109576_AADAT  ENSG00000103591_AAGAB  ...  \\\ncell_id                                                     ...   \n45006fe3e4c8               0.000000                    0.0  ...   \nd02759a80ba2               0.000000                    0.0  ...   \nc016c6b0efa5               3.847321                    0.0  ...   \nba7f733a4f75               0.000000                    0.0  ...   \nfbcf2443ffb2               0.000000                    0.0  ...   \n\ngene_id       ENSG00000153975_ZUP1  ENSG00000086827_ZW10  \\\ncell_id                                                    \n45006fe3e4c8              0.000000              0.000000   \nd02759a80ba2              0.000000              0.000000   \nc016c6b0efa5              0.000000              0.000000   \nba7f733a4f75              3.436846              0.000000   \nfbcf2443ffb2              0.000000              4.196826   \n\ngene_id       ENSG00000174442_ZWILCH  ENSG00000122952_ZWINT  \\\ncell_id                                                       \n45006fe3e4c8                0.000000               0.000000   \nd02759a80ba2                0.000000               4.039545   \nc016c6b0efa5                3.847321               4.529743   \nba7f733a4f75                4.113780               5.020215   \nfbcf2443ffb2                4.196826               4.196826   \n\ngene_id       ENSG00000198205_ZXDA  ENSG00000198455_ZXDB  \\\ncell_id                                                    \n45006fe3e4c8                   0.0                   0.0   \nd02759a80ba2                   0.0                   0.0   \nc016c6b0efa5                   0.0                   0.0   \nba7f733a4f75                   0.0                   0.0   \nfbcf2443ffb2                   0.0                   0.0   \n\ngene_id       ENSG00000070476_ZXDC  ENSG00000162378_ZYG11B  \\\ncell_id                                                      \n45006fe3e4c8               0.00000                0.000000   \nd02759a80ba2               0.00000                0.000000   \nc016c6b0efa5               0.00000                3.847321   \nba7f733a4f75               0.00000                3.436846   \nfbcf2443ffb2               3.51861                4.196826   \n\ngene_id       ENSG00000159840_ZYX  ENSG00000074755_ZZEF1  \ncell_id                                                   \n45006fe3e4c8             4.090185                    0.0  \nd02759a80ba2             0.000000                    0.0  \nc016c6b0efa5             3.847321                    0.0  \nba7f733a4f75             4.113780                    0.0  \nfbcf2443ffb2             3.518610                    0.0  \n\n[5 rows x 22050 columns]","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>gene_id</th>\n      <th>ENSG00000121410_A1BG</th>\n      <th>ENSG00000268895_A1BG-AS1</th>\n      <th>ENSG00000175899_A2M</th>\n      <th>ENSG00000245105_A2M-AS1</th>\n      <th>ENSG00000166535_A2ML1</th>\n      <th>ENSG00000128274_A4GALT</th>\n      <th>ENSG00000094914_AAAS</th>\n      <th>ENSG00000081760_AACS</th>\n      <th>ENSG00000109576_AADAT</th>\n      <th>ENSG00000103591_AAGAB</th>\n      <th>...</th>\n      <th>ENSG00000153975_ZUP1</th>\n      <th>ENSG00000086827_ZW10</th>\n      <th>ENSG00000174442_ZWILCH</th>\n      <th>ENSG00000122952_ZWINT</th>\n      <th>ENSG00000198205_ZXDA</th>\n      <th>ENSG00000198455_ZXDB</th>\n      <th>ENSG00000070476_ZXDC</th>\n      <th>ENSG00000162378_ZYG11B</th>\n      <th>ENSG00000159840_ZYX</th>\n      <th>ENSG00000074755_ZZEF1</th>\n    </tr>\n    <tr>\n      <th>cell_id</th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>45006fe3e4c8</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.00000</td>\n      <td>0.000000</td>\n      <td>4.090185</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>d02759a80ba2</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>4.039545</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.00000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>c016c6b0efa5</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>3.847321</td>\n      <td>0.000000</td>\n      <td>3.847321</td>\n      <td>3.847321</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>3.847321</td>\n      <td>4.529743</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.00000</td>\n      <td>3.847321</td>\n      <td>3.847321</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>ba7f733a4f75</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.000000</td>\n      <td>3.436846</td>\n      <td>3.436846</td>\n      <td>0.000000</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>3.436846</td>\n      <td>0.000000</td>\n      <td>4.113780</td>\n      <td>5.020215</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.00000</td>\n      <td>3.436846</td>\n      <td>4.113780</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>fbcf2443ffb2</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>4.196826</td>\n      <td>0.000000</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0.000000</td>\n      <td>4.196826</td>\n      <td>4.196826</td>\n      <td>4.196826</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>3.51861</td>\n      <td>4.196826</td>\n      <td>3.518610</td>\n      <td>0.0</td>\n    </tr>\n  </tbody>\n</table>\n<p>5 rows × 22050 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"gc.collect()\nhdf = pd.HDFStore('/kaggle/input/open-problems-multimodal/train_cite_targets.h5',\n                  mode='r')\ndf_cite_output = hdf.get(hdf.keys()[0])\ndf_cite_output.head()","metadata":{"execution":{"iopub.status.busy":"2022-12-18T03:50:03.814565Z","iopub.execute_input":"2022-12-18T03:50:03.814965Z","iopub.status.idle":"2022-12-18T03:50:04.701226Z","shell.execute_reply.started":"2022-12-18T03:50:03.814929Z","shell.execute_reply":"2022-12-18T03:50:04.700280Z"},"trusted":true},"execution_count":3,"outputs":[{"execution_count":3,"output_type":"execute_result","data":{"text/plain":"gene_id           CD86     CD274     CD270     CD155     CD112      CD47  \\\ncell_id                                                                    \n45006fe3e4c8  1.167804  0.622530  0.106959  0.324989  3.331674  6.426002   \nd02759a80ba2  0.818970  0.506009  1.078682  6.848758  3.524885  5.279456   \nc016c6b0efa5 -0.356703 -0.422261 -0.824493  1.137495  0.518924  7.221962   \nba7f733a4f75 -1.201507  0.149115  2.022468  6.021595  7.258670  2.792436   \nfbcf2443ffb2 -0.100404  0.697461  0.625836 -0.298404  1.369898  3.254521   \n\ngene_id            CD48      CD40     CD154      CD52  ...      CD94  \\\ncell_id                                                ...             \n45006fe3e4c8   1.480766 -0.728392 -0.468851 -0.073285  ... -0.448390   \nd02759a80ba2   4.930438  2.069372  0.333652 -0.468088  ...  0.323613   \nc016c6b0efa5  -0.375034  1.738071  0.142919 -0.971460  ...  1.348692   \nba7f733a4f75  21.708519 -0.137913  1.649969 -0.754680  ...  1.504426   \nfbcf2443ffb2  -1.659380  0.643531  0.902710  1.291877  ...  0.777023   \n\ngene_id           CD162     CD85j      CD23     CD328     HLA-E      CD82  \\\ncell_id                                                                     \n45006fe3e4c8   3.220174 -0.533004  0.674956 -0.006187  0.682148  1.398105   \nd02759a80ba2   8.407108  0.131301  0.047607 -0.243628  0.547864  1.832587   \nc016c6b0efa5   4.888579 -0.279483 -0.131097 -0.177604 -0.689188  9.013709   \nba7f733a4f75  12.391979  0.511394  0.587863 -0.752638  1.714851  3.893782   \nfbcf2443ffb2   6.496499  0.279898 -0.841950 -0.869419  0.675092  5.259685   \n\ngene_id          CD101      CD88     CD224  \ncell_id                                     \n45006fe3e4c8  0.414292  1.780314  0.548070  \nd02759a80ba2  0.982308  2.736507  2.184063  \nc016c6b0efa5 -1.182975  3.958148  2.868600  \nba7f733a4f75  1.799661  1.537249  4.407671  \nfbcf2443ffb2 -0.835379  9.631781  1.765445  \n\n[5 rows x 140 columns]","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>gene_id</th>\n      <th>CD86</th>\n      <th>CD274</th>\n      <th>CD270</th>\n      <th>CD155</th>\n      <th>CD112</th>\n      <th>CD47</th>\n      <th>CD48</th>\n      <th>CD40</th>\n      <th>CD154</th>\n      <th>CD52</th>\n      <th>...</th>\n      <th>CD94</th>\n      <th>CD162</th>\n      <th>CD85j</th>\n      <th>CD23</th>\n      <th>CD328</th>\n      <th>HLA-E</th>\n      <th>CD82</th>\n      <th>CD101</th>\n      <th>CD88</th>\n      <th>CD224</th>\n    </tr>\n    <tr>\n      <th>cell_id</th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>45006fe3e4c8</th>\n      <td>1.167804</td>\n      <td>0.622530</td>\n      <td>0.106959</td>\n      <td>0.324989</td>\n      <td>3.331674</td>\n      <td>6.426002</td>\n      <td>1.480766</td>\n      <td>-0.728392</td>\n      <td>-0.468851</td>\n      <td>-0.073285</td>\n      <td>...</td>\n      <td>-0.448390</td>\n      <td>3.220174</td>\n      <td>-0.533004</td>\n      <td>0.674956</td>\n      <td>-0.006187</td>\n      <td>0.682148</td>\n      <td>1.398105</td>\n      <td>0.414292</td>\n      <td>1.780314</td>\n      <td>0.548070</td>\n    </tr>\n    <tr>\n      <th>d02759a80ba2</th>\n      <td>0.818970</td>\n      <td>0.506009</td>\n      <td>1.078682</td>\n      <td>6.848758</td>\n      <td>3.524885</td>\n      <td>5.279456</td>\n      <td>4.930438</td>\n      <td>2.069372</td>\n      <td>0.333652</td>\n      <td>-0.468088</td>\n      <td>...</td>\n      <td>0.323613</td>\n      <td>8.407108</td>\n      <td>0.131301</td>\n      <td>0.047607</td>\n      <td>-0.243628</td>\n      <td>0.547864</td>\n      <td>1.832587</td>\n      <td>0.982308</td>\n      <td>2.736507</td>\n      <td>2.184063</td>\n    </tr>\n    <tr>\n      <th>c016c6b0efa5</th>\n      <td>-0.356703</td>\n      <td>-0.422261</td>\n      <td>-0.824493</td>\n      <td>1.137495</td>\n      <td>0.518924</td>\n      <td>7.221962</td>\n      <td>-0.375034</td>\n      <td>1.738071</td>\n      <td>0.142919</td>\n      <td>-0.971460</td>\n      <td>...</td>\n      <td>1.348692</td>\n      <td>4.888579</td>\n      <td>-0.279483</td>\n      <td>-0.131097</td>\n      <td>-0.177604</td>\n      <td>-0.689188</td>\n      <td>9.013709</td>\n      <td>-1.182975</td>\n      <td>3.958148</td>\n      <td>2.868600</td>\n    </tr>\n    <tr>\n      <th>ba7f733a4f75</th>\n      <td>-1.201507</td>\n      <td>0.149115</td>\n      <td>2.022468</td>\n      <td>6.021595</td>\n      <td>7.258670</td>\n      <td>2.792436</td>\n      <td>21.708519</td>\n      <td>-0.137913</td>\n      <td>1.649969</td>\n      <td>-0.754680</td>\n      <td>...</td>\n      <td>1.504426</td>\n      <td>12.391979</td>\n      <td>0.511394</td>\n      <td>0.587863</td>\n      <td>-0.752638</td>\n      <td>1.714851</td>\n      <td>3.893782</td>\n      <td>1.799661</td>\n      <td>1.537249</td>\n      <td>4.407671</td>\n    </tr>\n    <tr>\n      <th>fbcf2443ffb2</th>\n      <td>-0.100404</td>\n      <td>0.697461</td>\n      <td>0.625836</td>\n      <td>-0.298404</td>\n      <td>1.369898</td>\n      <td>3.254521</td>\n      <td>-1.659380</td>\n      <td>0.643531</td>\n      <td>0.902710</td>\n      <td>1.291877</td>\n      <td>...</td>\n      <td>0.777023</td>\n      <td>6.496499</td>\n      <td>0.279898</td>\n      <td>-0.841950</td>\n      <td>-0.869419</td>\n      <td>0.675092</td>\n      <td>5.259685</td>\n      <td>-0.835379</td>\n      <td>9.631781</td>\n      <td>1.765445</td>\n    </tr>\n  </tbody>\n</table>\n<p>5 rows × 140 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"df_cite_output.shape","metadata":{"execution":{"iopub.status.busy":"2022-12-18T03:54:08.801062Z","iopub.execute_input":"2022-12-18T03:54:08.801443Z","iopub.status.idle":"2022-12-18T03:54:08.809369Z","shell.execute_reply.started":"2022-12-18T03:54:08.801414Z","shell.execute_reply":"2022-12-18T03:54:08.808294Z"},"trusted":true},"execution_count":4,"outputs":[{"execution_count":4,"output_type":"execute_result","data":{"text/plain":"(70988, 140)"},"metadata":{}}]},{"cell_type":"code","source":"df_sample_submission = pd.read_csv('/kaggle/input/open-problems-multimodal/sample_submission.csv')\ndf_sample_submission.head(3)","metadata":{"execution":{"iopub.status.busy":"2022-12-18T04:04:37.130869Z","iopub.execute_input":"2022-12-18T04:04:37.132017Z","iopub.status.idle":"2022-12-18T04:04:57.273132Z","shell.execute_reply.started":"2022-12-18T04:04:37.131965Z","shell.execute_reply":"2022-12-18T04:04:57.271843Z"},"trusted":true},"execution_count":5,"outputs":[{"execution_count":5,"output_type":"execute_result","data":{"text/plain":"   row_id  target\n0       0     0.0\n1       1     0.0\n2       2     0.0","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>row_id</th>\n      <th>target</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>2</td>\n      <td>0.0</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"df_sample_submission.shape","metadata":{"execution":{"iopub.status.busy":"2022-12-18T04:04:57.705602Z","iopub.execute_input":"2022-12-18T04:04:57.706283Z","iopub.status.idle":"2022-12-18T04:04:57.714177Z","shell.execute_reply.started":"2022-12-18T04:04:57.706208Z","shell.execute_reply":"2022-12-18T04:04:57.713112Z"},"trusted":true},"execution_count":6,"outputs":[{"execution_count":6,"output_type":"execute_result","data":{"text/plain":"(65744180, 2)"},"metadata":{}}]},{"cell_type":"code","source":"df1 = pd.read_csv('/kaggle/input/open-problems-multimodal/metadata.csv')\ndf1.head()","metadata":{"execution":{"iopub.status.busy":"2022-12-17T09:26:18.885859Z","iopub.execute_input":"2022-12-17T09:26:18.886483Z","iopub.status.idle":"2022-12-17T09:26:19.173376Z","shell.execute_reply.started":"2022-12-17T09:26:18.886434Z","shell.execute_reply":"2022-12-17T09:26:19.172496Z"},"trusted":true},"execution_count":6,"outputs":[{"execution_count":6,"output_type":"execute_result","data":{"text/plain":"        cell_id  day  donor cell_type technology\n0  c2150f55becb    2  27678       HSC    citeseq\n1  65b7edf8a4da    2  27678       HSC    citeseq\n2  c1b26cb1057b    2  27678      EryP    citeseq\n3  917168fa6f83    2  27678      NeuP    citeseq\n4  2b29feeca86d    2  27678      EryP    citeseq","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>cell_id</th>\n      <th>day</th>\n      <th>donor</th>\n      <th>cell_type</th>\n      <th>technology</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>c2150f55becb</td>\n      <td>2</td>\n      <td>27678</td>\n      <td>HSC</td>\n      <td>citeseq</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>65b7edf8a4da</td>\n      <td>2</td>\n      <td>27678</td>\n      <td>HSC</td>\n      <td>citeseq</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>c1b26cb1057b</td>\n      <td>2</td>\n      <td>27678</td>\n      <td>EryP</td>\n      <td>citeseq</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>917168fa6f83</td>\n      <td>2</td>\n      <td>27678</td>\n      <td>NeuP</td>\n      <td>citeseq</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>2b29feeca86d</td>\n      <td>2</td>\n      <td>27678</td>\n      <td>EryP</td>\n      <td>citeseq</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"df1.cell_type.unique()","metadata":{"execution":{"iopub.status.busy":"2022-12-17T09:26:19.174970Z","iopub.execute_input":"2022-12-17T09:26:19.175352Z","iopub.status.idle":"2022-12-17T09:26:19.201706Z","shell.execute_reply.started":"2022-12-17T09:26:19.175317Z","shell.execute_reply":"2022-12-17T09:26:19.200848Z"},"trusted":true},"execution_count":7,"outputs":[{"execution_count":7,"output_type":"execute_result","data":{"text/plain":"array(['HSC', 'EryP', 'NeuP', 'MasP', 'MkP', 'BP', 'MoP', 'hidden'],\n      dtype=object)"},"metadata":{}}]},{"cell_type":"code","source":"df_cell_cite = df1[df1.technology==\"citeseq\"]\ndf_cell_multi = df1[df1.technology==\"multiome\"]","metadata":{"execution":{"iopub.status.busy":"2022-12-11T03:02:50.861012Z","iopub.execute_input":"2022-12-11T03:02:50.861427Z","iopub.status.idle":"2022-12-11T03:02:50.915348Z","shell.execute_reply.started":"2022-12-11T03:02:50.861392Z","shell.execute_reply":"2022-12-11T03:02:50.914391Z"},"trusted":true},"execution_count":12,"outputs":[]},{"cell_type":"code","source":"df_cite_cell_dist = df_cell_cite.set_index(\"cell_id\")[[\"day\",\"donor\"]].value_counts().to_frame()\\\n                .sort_values(\"day\").reset_index()\\\n                .rename(columns={0:\"# cells\"})\ndf_cite_cell_dist","metadata":{"execution":{"iopub.status.busy":"2022-12-11T03:03:16.478217Z","iopub.execute_input":"2022-12-11T03:03:16.479136Z","iopub.status.idle":"2022-12-11T03:03:16.509111Z","shell.execute_reply.started":"2022-12-11T03:03:16.479098Z","shell.execute_reply":"2022-12-11T03:03:16.508161Z"},"trusted":true},"execution_count":13,"outputs":[{"execution_count":13,"output_type":"execute_result","data":{"text/plain":"    day  donor  # cells\n0     2  31800     8395\n1     2  27678     7476\n2     2  32606     7476\n3     2  13176     6071\n4     3  13176     7643\n5     3  32606     6999\n6     3  27678     6488\n7     3  31800     6259\n8     4  31800    10149\n9     4  32606     9511\n10    4  13176     8485\n11    4  27678     7832\n12    7  32606     7254\n13    7  13176     7195\n14    7  27678     6247\n15    7  31800     6171","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>day</th>\n      <th>donor</th>\n      <th># cells</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>2</td>\n      <td>31800</td>\n      <td>8395</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>2</td>\n      <td>27678</td>\n      <td>7476</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>2</td>\n      <td>32606</td>\n      <td>7476</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>2</td>\n      <td>13176</td>\n      <td>6071</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>3</td>\n      <td>13176</td>\n      <td>7643</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>3</td>\n      <td>32606</td>\n      <td>6999</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>3</td>\n      <td>27678</td>\n      <td>6488</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>3</td>\n      <td>31800</td>\n      <td>6259</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>4</td>\n      <td>31800</td>\n      <td>10149</td>\n    </tr>\n    <tr>\n      <th>9</th>\n      <td>4</td>\n      <td>32606</td>\n      <td>9511</td>\n    </tr>\n    <tr>\n      <th>10</th>\n      <td>4</td>\n      <td>13176</td>\n      <td>8485</td>\n    </tr>\n    <tr>\n      <th>11</th>\n      <td>4</td>\n      <td>27678</td>\n      <td>7832</td>\n    </tr>\n    <tr>\n      <th>12</th>\n      <td>7</td>\n      <td>32606</td>\n      <td>7254</td>\n    </tr>\n    <tr>\n      <th>13</th>\n      <td>7</td>\n      <td>13176</td>\n      <td>7195</td>\n    </tr>\n    <tr>\n      <th>14</th>\n      <td>7</td>\n      <td>27678</td>\n      <td>6247</td>\n    </tr>\n    <tr>\n      <th>15</th>\n      <td>7</td>\n      <td>31800</td>\n      <td>6171</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"df_multi_cell_dist = df_cell_multi.set_index(\"cell_id\")[[\"day\",\"donor\"]].value_counts().to_frame()\\\n                .sort_values(\"day\").reset_index()\\\n                .rename(columns={0:\"# cells\"})\ndf_multi_cell_dist","metadata":{"execution":{"iopub.status.busy":"2022-12-11T03:08:50.119421Z","iopub.execute_input":"2022-12-11T03:08:50.119781Z","iopub.status.idle":"2022-12-11T03:08:50.148434Z","shell.execute_reply.started":"2022-12-11T03:08:50.119750Z","shell.execute_reply":"2022-12-11T03:08:50.147327Z"},"trusted":true},"execution_count":14,"outputs":[{"execution_count":14,"output_type":"execute_result","data":{"text/plain":"    day  donor  # cells\n0     2  13176     9558\n1     2  27678     8921\n2     2  32606     7264\n3     2  31800     7089\n4     3  31800    11306\n5     3  13176     8965\n6     3  32606     8544\n7     3  27678     7950\n8     4  32606    10933\n9     4  31800    10794\n10    4  13176     9407\n11    7  27678     7291\n12    7  32606     6960\n13    7  31800     7656\n14    7  13176     7466\n15   10  32606     7418\n16   10  27678     8790\n17   10  31800     7144\n18   10  13176     8421","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>day</th>\n      <th>donor</th>\n      <th># cells</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>2</td>\n      <td>13176</td>\n      <td>9558</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>2</td>\n      <td>27678</td>\n      <td>8921</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>2</td>\n      <td>32606</td>\n      <td>7264</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>2</td>\n      <td>31800</td>\n      <td>7089</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>3</td>\n      <td>31800</td>\n      <td>11306</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>3</td>\n      <td>13176</td>\n      <td>8965</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>3</td>\n      <td>32606</td>\n      <td>8544</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>3</td>\n      <td>27678</td>\n      <td>7950</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>4</td>\n      <td>32606</td>\n      <td>10933</td>\n    </tr>\n    <tr>\n      <th>9</th>\n      <td>4</td>\n      <td>31800</td>\n      <td>10794</td>\n    </tr>\n    <tr>\n      <th>10</th>\n      <td>4</td>\n      <td>13176</td>\n      <td>9407</td>\n    </tr>\n    <tr>\n      <th>11</th>\n      <td>7</td>\n      <td>27678</td>\n      <td>7291</td>\n    </tr>\n    <tr>\n      <th>12</th>\n      <td>7</td>\n      <td>32606</td>\n      <td>6960</td>\n    </tr>\n    <tr>\n      <th>13</th>\n      <td>7</td>\n      <td>31800</td>\n      <td>7656</td>\n    </tr>\n    <tr>\n      <th>14</th>\n      <td>7</td>\n      <td>13176</td>\n      <td>7466</td>\n    </tr>\n    <tr>\n      <th>15</th>\n      <td>10</td>\n      <td>32606</td>\n      <td>7418</td>\n    </tr>\n    <tr>\n      <th>16</th>\n      <td>10</td>\n      <td>27678</td>\n      <td>8790</td>\n    </tr>\n    <tr>\n      <th>17</th>\n      <td>10</td>\n      <td>31800</td>\n      <td>7144</td>\n    </tr>\n    <tr>\n      <th>18</th>\n      <td>10</td>\n      <td>13176</td>\n      <td>8421</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"fig, axs = plt.subplots(1,2,figsize=(12,6))\n\nsns.barplot(data=df_cite_cell_dist, x=\"day\",hue=\"donor\",y=\"# cells\", ax=axs[0])\naxs[0].set_title(\"Number of cells measured with CITEseq\")\n\n\nsns.barplot(data=df_multi_cell_dist, x=\"day\",hue=\"donor\",y=\"# cells\", ax=axs[1])\naxs[1].set_title(\"Number of cells measured with Multiome\")\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2022-12-11T03:10:56.019346Z","iopub.execute_input":"2022-12-11T03:10:56.019712Z","iopub.status.idle":"2022-12-11T03:10:56.464047Z","shell.execute_reply.started":"2022-12-11T03:10:56.019682Z","shell.execute_reply":"2022-12-11T03:10:56.462955Z"},"trusted":true},"execution_count":17,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 864x432 with 2 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":"import os, gc, pickle\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport seaborn as sns\nimport numpy as np\nfrom colorama import Fore, Back, Style\nfrom matplotlib.ticker import MaxNLocator\nimport warnings\nwarnings.simplefilter('ignore')\nfrom sklearn.base import BaseEstimator, TransformerMixin\nfrom sklearn.model_selection import KFold\nfrom sklearn.preprocessing import StandardScaler, scale\nfrom sklearn.decomposition import PCA\nfrom sklearn.dummy import DummyRegressor\nfrom sklearn.pipeline import make_pipeline, Pipeline\nfrom sklearn.linear_model import Ridge\nfrom sklearn.linear_model import Lasso\nfrom sklearn.multioutput import MultiOutputRegressor\nimport lightgbm as lgb\nfrom sklearn.compose import TransformedTargetRegressor\nfrom sklearn.metrics import mean_squared_error","metadata":{"execution":{"iopub.status.busy":"2022-12-18T07:45:07.060172Z","iopub.execute_input":"2022-12-18T07:45:07.060641Z","iopub.status.idle":"2022-12-18T07:45:10.199842Z","shell.execute_reply.started":"2022-12-18T07:45:07.060604Z","shell.execute_reply":"2022-12-18T07:45:10.198923Z"},"trusted":true},"execution_count":2,"outputs":[{"output_type":"display_data","data":{"text/plain":"<IPython.core.display.HTML object>","text/html":"<style type='text/css'>\n.datatable table.frame { margin-bottom: 0; }\n.datatable table.frame thead { border-bottom: none; }\n.datatable table.frame tr.coltypes td {  color: #FFFFFF;  line-height: 6px;  padding: 0 0.5em;}\n.datatable .bool    { background: #DDDD99; }\n.datatable .object  { background: #565656; }\n.datatable .int     { background: #5D9E5D; }\n.datatable .float   { background: #4040CC; }\n.datatable .str     { background: #CC4040; }\n.datatable .time    { background: #40CC40; }\n.datatable .row_index {  background: var(--jp-border-color3);  border-right: 1px solid var(--jp-border-color0);  color: var(--jp-ui-font-color3);  font-size: 9px;}\n.datatable .frame tbody td { text-align: left; }\n.datatable .frame tr.coltypes .row_index {  background: var(--jp-border-color0);}\n.datatable th:nth-child(2) { padding-left: 12px; }\n.datatable .hellipsis {  color: var(--jp-cell-editor-border-color);}\n.datatable .vellipsis {  background: var(--jp-layout-color0);  color: var(--jp-cell-editor-border-color);}\n.datatable .na {  color: var(--jp-cell-editor-border-color);  font-size: 80%;}\n.datatable .sp {  opacity: 0.25;}\n.datatable .footer { font-size: 9px; }\n.datatable .frame_dimensions {  background: var(--jp-border-color3);  border-top: 1px solid var(--jp-border-color0);  color: var(--jp-ui-font-color3);  display: inline-block;  opacity: 0.6;  padding: 1px 10px 1px 5px;}\n.datatable .frame thead tr.colnames {  background-image: url('data:image/png;base64,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');  background-repeat: repeat-x;  background-size: 14px;  height: 28px;}\n</style>\n"},"metadata":{}}]},{"cell_type":"code","source":"DATA_DIR = \"/kaggle/input/open-problems-multimodal/\"\nFP_CELL_METADATA = os.path.join(DATA_DIR,\"metadata.csv\")\n\nFP_CITE_TRAIN_INPUTS = os.path.join(DATA_DIR,\"train_cite_inputs.h5\")\nFP_CITE_TRAIN_TARGETS = os.path.join(DATA_DIR,\"train_cite_targets.h5\")\nFP_CITE_TEST_INPUTS = os.path.join(DATA_DIR,\"test_cite_inputs.h5\")\n\nFP_MULTIOME_TRAIN_INPUTS = os.path.join(DATA_DIR,\"train_multi_inputs.h5\")\nFP_MULTIOME_TRAIN_TARGETS = os.path.join(DATA_DIR,\"train_multi_targets.h5\")\nFP_MULTIOME_TEST_INPUTS = os.path.join(DATA_DIR,\"test_multi_inputs.h5\")\n\nFP_SUBMISSION = os.path.join(DATA_DIR,\"sample_submission.csv\")\nFP_EVALUATION_IDS = os.path.join(DATA_DIR,\"evaluation_ids.csv\")","metadata":{"execution":{"iopub.status.busy":"2022-12-18T03:09:41.305958Z","iopub.execute_input":"2022-12-18T03:09:41.306247Z","iopub.status.idle":"2022-12-18T03:09:41.313143Z","shell.execute_reply.started":"2022-12-18T03:09:41.306221Z","shell.execute_reply":"2022-12-18T03:09:41.312162Z"},"trusted":true},"execution_count":6,"outputs":[]},{"cell_type":"code","source":"!pip install --quiet tables","metadata":{"execution":{"iopub.status.busy":"2022-12-17T04:02:01.227796Z","iopub.execute_input":"2022-12-17T04:02:01.228183Z","iopub.status.idle":"2022-12-17T04:02:11.468350Z","shell.execute_reply.started":"2022-12-17T04:02:01.228146Z","shell.execute_reply":"2022-12-17T04:02:11.467063Z"},"trusted":true},"execution_count":3,"outputs":[{"name":"stdout","text":"\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n\u001b[0m","output_type":"stream"}]},{"cell_type":"code","source":"df_cell = pd.read_csv(FP_CELL_METADATA)\ndf_cell_cite = df_cell[df_cell.technology==\"citeseq\"]\ndf_cell_multi = df_cell[df_cell.technology==\"multiome\"]\ndf_cell_cite.shape, df_cell_multi.shape","metadata":{"execution":{"iopub.status.busy":"2022-12-17T09:25:16.979739Z","iopub.execute_input":"2022-12-17T09:25:16.982346Z","iopub.status.idle":"2022-12-17T09:25:17.501344Z","shell.execute_reply.started":"2022-12-17T09:25:16.982309Z","shell.execute_reply":"2022-12-17T09:25:17.500338Z"},"trusted":true},"execution_count":4,"outputs":[{"execution_count":4,"output_type":"execute_result","data":{"text/plain":"((119651, 5), (161877, 5))"},"metadata":{}}]},{"cell_type":"code","source":"pd.read_hdf(FP_MULTIOME_TEST_INPUTS).values","metadata":{"execution":{"iopub.status.busy":"2022-12-17T09:23:03.092436Z","iopub.execute_input":"2022-12-17T09:23:03.092884Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"%%time\n# Preprocessing\ncol_start = 10000\n\nclass PreprocessCiteseq(BaseEstimator, TransformerMixin):\n    columns_to_use = 13000\n    \n    @staticmethod\n    def take_column_subset(X):\n        return X[:,-(PreprocessCiteseq.columns_to_use+col_start):-col_start]\n    \n    def transform(self, X):\n        print('orignal shape', X.shape)\n        X = X[:,~self.all_zero_columns]\n        print('after removing zeros', X.shape)\n        X = PreprocessCiteseq.take_column_subset(X) # use only a part of the columns\n        print('after subsetting', X.shape)\n        gc.collect()\n\n        X = self.pca.transform(X)\n        print('after PCA', X.shape)\n        return X\n\n    def fit_transform(self, X):\n        gc.collect()\n        print('orignal shape', X.shape)\n        self.all_zero_columns = (X == 0).all(axis=0)\n        X = X[:,~self.all_zero_columns]\n        print('after removing zeros', X.shape)\n        X = PreprocessCiteseq.take_column_subset(X) # use only a part of the columns\n        print('after subsetting', X.shape)\n        gc.collect()\n\n        self.pca = PCA(n_components=240, copy=False, random_state=1)\n        X = self.pca.fit_transform(X)\n#         plt.plot(self.pca.explained_variance_ratio_.cumsum())\n#         plt.title(\"Cumulative explained variance ratio\")\n#         plt.gca().xaxis.set_major_locator(MaxNLocator(integer=True))\n#         plt.xlabel('PCA component')\n#         plt.ylabel('Cumulative explained variance ratio')\n#         plt.show()\n        print(X.shape)\n        return X\n\ngc.collect()\npreprocessor = PreprocessCiteseq()\n\ncite_train_x = None\ncite_train_x = preprocessor.fit_transform(pd.read_hdf(FP_MULTIOME_TEST_INPUTS).values)\n\ncite_train_y = pd.read_hdf(FP_CITE_TRAIN_TARGETS).values\nprint(cite_train_y.shape)","metadata":{"execution":{"iopub.status.busy":"2022-12-17T09:23:55.942370Z","iopub.execute_input":"2022-12-17T09:23:55.942875Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"params = {\n     'learning_rate': 0.1, \n     'metric': 'mae', \n     \"seed\": 42,\n    'reg_alpha': 0.0014, \n    'reg_lambda': 0.2, \n    'colsample_bytree': 0.8, \n    'subsample': 0.5, \n    'max_depth': 10, \n    'num_leaves': 722, \n    'min_child_samples': 83, \n    }\n\nmodel = MultiOutputRegressor(lgb.LGBMRegressor(**params, n_estimators=1000))\n\nmodel.fit(cite_train_x, cite_train_y)\n\ny_va_pred = model.predict(cite_train_x)\nmse = mean_squared_error(cite_train_y, y_va_pred)\nprint(mse)\ndel cite_train_x, cite_train_y\ngc.collect()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import os, gc, pickle\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport seaborn as sns\nimport numpy as np\nfrom colorama import Fore, Back, Style\nfrom matplotlib.ticker import MaxNLocator\nfrom tqdm import tqdm\n\nfrom sklearn.model_selection import KFold\nfrom sklearn.metrics import mean_squared_error\n\nimport torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nfrom torch.optim import AdamW\nfrom torch.optim import lr_scheduler\nfrom torch.utils.data import DataLoader, Dataset\n\ndevice = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\nprint(device)\n\nDATA_DIR = \"/kaggle/input/open-problems-multimodal/\"\nFP_CELL_METADATA = os.path.join(DATA_DIR,\"metadata.csv\")\n\nFP_CITE_TRAIN_INPUTS = os.path.join(DATA_DIR,\"train_cite_inputs.h5\")\nFP_CITE_TRAIN_TARGETS = os.path.join(DATA_DIR,\"train_cite_targets.h5\")\nFP_CITE_TEST_INPUTS = os.path.join(DATA_DIR,\"test_cite_inputs.h5\")\n\nFP_MULTIOME_TRAIN_INPUTS = os.path.join(DATA_DIR,\"train_multi_inputs.h5\")\nFP_MULTIOME_TRAIN_TARGETS = os.path.join(DATA_DIR,\"train_multi_targets.h5\")\nFP_MULTIOME_TEST_INPUTS = os.path.join(DATA_DIR,\"test_multi_inputs.h5\")\n\nFP_SUBMISSION = os.path.join(DATA_DIR,\"sample_submission.csv\")\nFP_EVALUATION_IDS = os.path.join(DATA_DIR,\"evaluation_ids.csv\")","metadata":{"execution":{"iopub.status.busy":"2022-12-18T07:45:12.455991Z","iopub.execute_input":"2022-12-18T07:45:12.456379Z","iopub.status.idle":"2022-12-18T07:45:13.914438Z","shell.execute_reply.started":"2022-12-18T07:45:12.456349Z","shell.execute_reply":"2022-12-18T07:45:13.912537Z"},"trusted":true},"execution_count":3,"outputs":[{"name":"stdout","text":"cuda\n","output_type":"stream"}]},{"cell_type":"code","source":"cite_train_x = np.load(\"/kaggle/input/trainx/cite_train_x.npy\")\nprint(cite_train_x.shape)\n\n\ncite_train_y = pd.read_hdf(FP_CITE_TRAIN_TARGETS).values\nprint(cite_train_y.shape)","metadata":{"execution":{"iopub.status.busy":"2022-12-18T05:33:14.843040Z","iopub.execute_input":"2022-12-18T05:33:14.843736Z","iopub.status.idle":"2022-12-18T05:33:15.608382Z","shell.execute_reply.started":"2022-12-18T05:33:14.843701Z","shell.execute_reply":"2022-12-18T05:33:15.606422Z"},"trusted":true},"execution_count":5,"outputs":[{"name":"stdout","text":"(70988, 240)\n(70988, 140)\n","output_type":"stream"}]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"RNA to protien encoder decoder","metadata":{}},{"cell_type":"code","source":"class CFG:\n    tr_batch_size = 16 # 16\n    va_batch_size = 128 # 32\n    \n    optimizer = \"AdamW\"\n    lr = 1e-5\n    weight_decay = 0.1\n    betas = (0.9, 0.999)\n    epochs = 15\n    ","metadata":{"execution":{"iopub.status.busy":"2022-12-18T07:45:22.071135Z","iopub.execute_input":"2022-12-18T07:45:22.071754Z","iopub.status.idle":"2022-12-18T07:45:22.078426Z","shell.execute_reply.started":"2022-12-18T07:45:22.071721Z","shell.execute_reply":"2022-12-18T07:45:22.076262Z"},"trusted":true},"execution_count":4,"outputs":[]},{"cell_type":"code","source":"class CtieseqDataset(Dataset):\n    \"\"\"\n    Train, Validation or Test dataset for CITEseq samples\n    Prepares data for simple vector to vector NN\n    \"\"\"\n    def __init__(self, X, y=None):\n        self.train = False \n        if y is not None:\n            self.train = True\n        self.X = X\n        self.y = y\n            \n    def __len__(self):\n        return len(self.X)\n    \n    def __getitem__(self, idx):\n        X = self.X[idx]\n        \n        if self.train:\n            y = self.y[idx]\n            return {\n                \"X\" : torch.tensor(X).to(device),\n                \"y\" : torch.tensor(y).to(device)\n            }\n        else:\n            return {\n                \"X\" : torch.tensor(X).to(device)\n            }","metadata":{"execution":{"iopub.status.busy":"2022-12-18T07:45:22.500665Z","iopub.execute_input":"2022-12-18T07:45:22.501213Z","iopub.status.idle":"2022-12-18T07:45:22.508864Z","shell.execute_reply.started":"2022-12-18T07:45:22.501177Z","shell.execute_reply":"2022-12-18T07:45:22.507964Z"},"trusted":true},"execution_count":5,"outputs":[]},{"cell_type":"code","source":"def criterion(outputs, labels):\n    \"\"\" MSE Loss function\"\"\"\n    return nn.MSELoss()(outputs, labels)\n\ndef correlation_score(y_true, y_pred):\n    \"\"\"\n    Scores the predictions according to the competition rules. \n    It is assumed that the predictions are not constant.\n    Returns the average of each sample's Pearson correlation coefficient\n    \"\"\"\n    \n    if type(y_true) == pd.DataFrame: y_true = y_true.values\n    if type(y_pred) == pd.DataFrame: y_pred = y_pred.values\n    corrsum = 0\n    for i in range(len(y_true)):\n        corrsum += np.corrcoef(y_true[i], y_pred[i])[1, 0]\n    return corrsum / len(y_true)\n\ndef get_optimizer(model, lr, weight_decay, betas):\n    \"\"\" Gets AdamW optimizer \"\"\"\n    param_optimizer = list(model.named_parameters())\n    no_decay = ['bias', 'LayerNorm.bias', 'LayerNorm.weight']\n    optimizer_grouped_parameters = [\n        {'params': [p for n, p in param_optimizer if not any(\n            nd in n for nd in no_decay)], 'weight_decay': weight_decay},\n        {'params': [p for n, p in param_optimizer if any(\n            nd in n for nd in no_decay)], 'weight_decay': 0.0}\n    ]\n    optimizer = AdamW(optimizer_grouped_parameters,\n                      lr=lr,\n                      weight_decay=weight_decay,\n                      betas=betas,\n                     )\n    return optimizer\n\ndef get_scheduler(optimizer, T_max=300):\n    \"\"\" Gets Consine scheduler \"\"\"\n    scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(optimizer,\n                                                           T_max=T_max)\n    return scheduler","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:12:20.479800Z","iopub.execute_input":"2022-12-18T08:12:20.480181Z","iopub.status.idle":"2022-12-18T08:12:20.491108Z","shell.execute_reply.started":"2022-12-18T08:12:20.480148Z","shell.execute_reply":"2022-12-18T08:12:20.490066Z"},"trusted":true},"execution_count":17,"outputs":[]},{"cell_type":"code","source":"class FCBlock(nn.Module):\n    \"\"\"\n    A Pytorch Block for a fully connected Layer\n    Includes Linear, Activation Function, and Dropout\n    \"\"\"\n    def __init__(self, input_dim, hidden_dim, dropout):\n        super().__init__()\n        self.fc = nn.Linear(input_dim, hidden_dim)\n        self.dropout = nn.Dropout(dropout)\n        \n    def forward(self, x):\n        x = self.fc(x)\n        x = F.selu(x)\n        x = self.dropout(x)\n        return x\n\nclass Encoder(nn.Module):\n    \"\"\"\n    Encoder module to generate embeddings of a RNA vector\n    \"\"\"\n    def __init__(self):\n        super().__init__()\n        self.l0 = FCBlock(240, 120, 0.05)\n        self.l1 = FCBlock(120, 60, 0.05)\n        self.l2 = FCBlock(60, 30, 0.05)\n        \n    def forward(self, x):\n        x = self.l0(x)\n        x = self.l1(x)\n        x = self.l2(x)\n        return x\n    \nclass Decoder(nn.Module):\n    \"\"\"\n    Decoder module to extract Protein sequences from RNA embeddings\n    \"\"\"\n    def __init__(self):\n        super().__init__()\n        self.l0 = FCBlock(30, 70, 0.05)\n        self.l1 = FCBlock(70, 100, 0.05)\n        self.l2 = FCBlock(100, 140, 0.05)\n        \n    def forward(self, x):\n        x = self.l0(x)\n        x = self.l1(x)\n        x = self.l2(x)\n        return x\n    \nclass CtieseqModel(nn.Module):\n    \"\"\"\n    Wrapper for the Encoder and Decoder modules\n    Converts RNA sequence to Protein sequence\n    \"\"\"\n    def __init__(self):\n        super().__init__()\n        self.encoder = Encoder()\n        self.decoder = Decoder()\n        \n    def forward(self, x):\n        embeddings = self.encoder(x)\n        outputs = self.decoder(embeddings)\n        return outputs","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:12:22.917127Z","iopub.execute_input":"2022-12-18T08:12:22.917479Z","iopub.status.idle":"2022-12-18T08:12:22.929842Z","shell.execute_reply.started":"2022-12-18T08:12:22.917447Z","shell.execute_reply":"2022-12-18T08:12:22.928845Z"},"trusted":true},"execution_count":18,"outputs":[]},{"cell_type":"code","source":"class AverageMeter(object):\n    \"\"\"Computes and stores the average and current value\"\"\"\n    def __init__(self):\n        self.reset()\n    def reset(self):\n        self.val = 0\n        self.avg = 0\n        self.sum = 0\n        self.count = 0\n    def update(self, val, n=1):\n        self.val = val\n        self.sum += val * n\n        self.count += n\n        self.avg = self.sum / self.count","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:12:23.852326Z","iopub.execute_input":"2022-12-18T08:12:23.852708Z","iopub.status.idle":"2022-12-18T08:12:23.859800Z","shell.execute_reply.started":"2022-12-18T08:12:23.852674Z","shell.execute_reply":"2022-12-18T08:12:23.858660Z"},"trusted":true},"execution_count":19,"outputs":[]},{"cell_type":"code","source":"def train_one_epoch(model, optimizer, scheduler, dataloader, device, epoch):\n    \"\"\" Trains one epoch and returns loss \"\"\"\n    model.train()\n    \n    losses = AverageMeter()\n    corr = AverageMeter()\n    \n    bar = tqdm(enumerate(dataloader), total=len(dataloader))\n    for step, data in bar:\n        X, y = data[\"X\"], data[\"y\"]\n        \n        batch_size = X.size(0)\n\n        outputs = model(X)\n\n        n = outputs.size(0)\n        loss = criterion(outputs, y)\n        losses.update(loss.item(), n)\n        loss.backward()\n        \n        outputs = outputs.detach().cpu().numpy()\n        y = y.detach().cpu().numpy()\n        corr_score = correlation_score(y, outputs)\n        corr.update(corr_score, n)\n        \n        optimizer.step()\n        optimizer.zero_grad()\n\n        if scheduler is not None:\n            scheduler.step()\n        \n        bar.set_postfix(Epoch=epoch, Train_Loss=losses.avg, Corr=corr.avg,\n                        LR=optimizer.param_groups[0]['lr'])\n    gc.collect()\n    \n    return losses.avg","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:12:25.194949Z","iopub.execute_input":"2022-12-18T08:12:25.195703Z","iopub.status.idle":"2022-12-18T08:12:25.205476Z","shell.execute_reply.started":"2022-12-18T08:12:25.195659Z","shell.execute_reply":"2022-12-18T08:12:25.204240Z"},"trusted":true},"execution_count":20,"outputs":[]},{"cell_type":"code","source":"@torch.no_grad()\ndef valid_one_epoch(model, optimizer, dataloader, device, epoch):\n    \"\"\" Evaluates one epoch and returns loss \"\"\"\n    model.eval()\n    \n    losses = AverageMeter()\n    corr = AverageMeter()\n    \n    bar = tqdm(enumerate(dataloader), total=len(dataloader))\n    for step, data in bar:        \n        X, y = data[\"X\"], data[\"y\"]\n        \n        batch_size = X.size(0)\n\n        outputs = model(X)\n        \n        n = outputs.size(0)\n        loss = criterion(outputs, y)\n        losses.update(loss.item(), n)\n        \n        outputs = outputs.detach().cpu().numpy()\n        y = y.detach().cpu().numpy()\n        corr_score = correlation_score(y, outputs)\n        corr.update(corr_score, n)\n        \n        bar.set_postfix(Epoch=epoch, Valid_Loss=losses.avg, Corr=corr.avg,\n                        LR=optimizer.param_groups[0]['lr'])   \n    \n    gc.collect()\n    \n    return losses.avg","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:12:26.720466Z","iopub.execute_input":"2022-12-18T08:12:26.721148Z","iopub.status.idle":"2022-12-18T08:12:26.729564Z","shell.execute_reply.started":"2022-12-18T08:12:26.721111Z","shell.execute_reply":"2022-12-18T08:12:26.728372Z"},"trusted":true},"execution_count":21,"outputs":[]},{"cell_type":"code","source":"def train_one_fold(model, \n                   optimizer, \n                   scheduler, \n                   train_loader, \n                   valid_loader, \n                   fold):\n    \"\"\" Trains and saves a full fold of a pytorch model \"\"\"\n    best_epoch_loss = np.inf\n    model.to(device)\n\n    for epoch in range(CFG.epochs):\n        gc.collect()\n        train_epoch_loss = train_one_epoch(model, \n                                           optimizer, \n                                           scheduler, \n                                           dataloader=train_loader, \n                                           device=device, \n                                           epoch=epoch)\n\n        val_epoch_loss = valid_one_epoch(model,\n                                         optimizer, \n                                         valid_loader, \n                                         device=device, epoch=epoch)\n        torch.save({\n            'epoch': epoch,\n            'model_state_dict': model.state_dict(),\n            'optimizer_state_dict': optimizer.state_dict(),\n            'loss': val_epoch_loss,\n            }, '/kaggle/working/latest_model_training_stage')\n        \n        if val_epoch_loss <= best_epoch_loss:\n            print(f\"Validation Loss Improved ({best_epoch_loss} ---> {val_epoch_loss})\")\n            best_epoch_loss = val_epoch_loss\n            torch.save(model.state_dict(), f\"model_f{fold}.bin\")\n            \n    print(\"Best Loss: {:.4f}\".format(best_epoch_loss))","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:12:28.256980Z","iopub.execute_input":"2022-12-18T08:12:28.257692Z","iopub.status.idle":"2022-12-18T08:12:28.267215Z","shell.execute_reply.started":"2022-12-18T08:12:28.257652Z","shell.execute_reply":"2022-12-18T08:12:28.265931Z"},"trusted":true},"execution_count":22,"outputs":[]},{"cell_type":"code","source":"kf = KFold(n_splits=5, shuffle=True, random_state=42)\nscore_list = []\nfor fold, (idx_tr, idx_va) in enumerate(kf.split(cite_train_x)):\n    print(f\"\\nfold = {fold}\")\n    X_tr = cite_train_x[idx_tr] \n    y_tr = cite_train_y[idx_tr]\n    \n    X_va = cite_train_x[idx_va]\n    y_va = cite_train_y[idx_va]\n    \n    ds_tr = CtieseqDataset(X_tr, y_tr)\n    ds_va = CtieseqDataset(X_tr, y_tr)\n    dl_tr = DataLoader(ds_tr, batch_size=CFG.tr_batch_size, shuffle=True)\n    dl_va = DataLoader(ds_va, batch_size=CFG.va_batch_size, shuffle=False)\n    \n    model = CtieseqModel()\n    optimizer = get_optimizer(model, CFG.lr, CFG.weight_decay, CFG.betas)\n    scheduler = get_scheduler(optimizer) \n\n    train_one_fold(model, optimizer, scheduler, dl_tr, dl_va, fold)\n    ","metadata":{"execution":{"iopub.status.busy":"2022-12-17T05:22:57.497136Z","iopub.execute_input":"2022-12-17T05:22:57.497970Z"},"trusted":true},"execution_count":null,"outputs":[{"name":"stdout","text":"\nfold = 0\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.34it/s, Corr=0.297, Epoch=0, LR=9.33e-6, Train_Loss=13.7] \n100%|██████████| 444/444 [00:10<00:00, 42.24it/s, Corr=0.683, Epoch=0, LR=9.33e-6, Valid_Loss=9.81]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (inf ---> 9.80943159622969)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:29<00:00, 118.63it/s, Corr=0.726, Epoch=1, LR=7.5e-6, Train_Loss=6.6]   \n100%|██████████| 444/444 [00:11<00:00, 40.06it/s, Corr=0.804, Epoch=1, LR=7.5e-6, Valid_Loss=4.71]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (9.80943159622969 ---> 4.711339220586724)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.29it/s, Corr=0.771, Epoch=2, LR=5e-6, Train_Loss=5.19]    \n100%|██████████| 444/444 [00:10<00:00, 41.08it/s, Corr=0.816, Epoch=2, LR=5e-6, Valid_Loss=4.37]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.711339220586724 ---> 4.372307234453596)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.36it/s, Corr=0.786, Epoch=3, LR=2.5e-6, Train_Loss=4.89]  \n100%|██████████| 444/444 [00:11<00:00, 39.73it/s, Corr=0.833, Epoch=3, LR=2.5e-6, Valid_Loss=4.06]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.372307234453596 ---> 4.055687797302553)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.57it/s, Corr=0.8, Epoch=4, LR=6.7e-7, Train_Loss=4.64]    \n100%|██████████| 444/444 [00:11<00:00, 40.26it/s, Corr=0.843, Epoch=4, LR=6.7e-7, Valid_Loss=3.84]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.055687797302553 ---> 3.8448395148531023)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.55it/s, Corr=0.806, Epoch=5, LR=0, Train_Loss=4.51]       \n100%|██████████| 444/444 [00:10<00:00, 41.32it/s, Corr=0.846, Epoch=5, LR=0, Valid_Loss=3.76]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.8448395148531023 ---> 3.758324454027617)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:31<00:00, 112.05it/s, Corr=0.809, Epoch=6, LR=6.7e-7, Train_Loss=4.43]  \n100%|██████████| 444/444 [00:10<00:00, 40.39it/s, Corr=0.848, Epoch=6, LR=6.7e-7, Valid_Loss=3.7] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.758324454027617 ---> 3.7034477350849944)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:32<00:00, 107.92it/s, Corr=0.811, Epoch=7, LR=2.5e-6, Train_Loss=4.39]  \n100%|██████████| 444/444 [00:10<00:00, 41.29it/s, Corr=0.85, Epoch=7, LR=2.5e-6, Valid_Loss=3.65] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.7034477350849944 ---> 3.654522175351422)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.89it/s, Corr=0.813, Epoch=8, LR=5e-6, Train_Loss=4.35]    \n100%|██████████| 444/444 [00:10<00:00, 41.39it/s, Corr=0.852, Epoch=8, LR=5e-6, Valid_Loss=3.61]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.654522175351422 ---> 3.605671771283426)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.27it/s, Corr=0.815, Epoch=9, LR=7.5e-6, Train_Loss=4.28]  \n100%|██████████| 444/444 [00:11<00:00, 38.93it/s, Corr=0.854, Epoch=9, LR=7.5e-6, Valid_Loss=3.56]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.605671771283426 ---> 3.5552296275340205)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.36it/s, Corr=0.818, Epoch=10, LR=9.33e-6, Train_Loss=4.23] \n100%|██████████| 444/444 [00:10<00:00, 41.38it/s, Corr=0.857, Epoch=10, LR=9.33e-6, Valid_Loss=3.49]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.5552296275340205 ---> 3.4921957119604006)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.08it/s, Corr=0.821, Epoch=11, LR=1e-5, Train_Loss=4.17]   \n100%|██████████| 444/444 [00:10<00:00, 41.41it/s, Corr=0.859, Epoch=11, LR=1e-5, Valid_Loss=3.44]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.4921957119604006 ---> 3.4354745950764345)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.59it/s, Corr=0.823, Epoch=12, LR=9.33e-6, Train_Loss=4.12] \n100%|██████████| 444/444 [00:10<00:00, 41.06it/s, Corr=0.861, Epoch=12, LR=9.33e-6, Valid_Loss=3.39]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.4354745950764345 ---> 3.3886562388470987)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 114.86it/s, Corr=0.825, Epoch=13, LR=7.5e-6, Train_Loss=4.07]  \n100%|██████████| 444/444 [00:10<00:00, 41.48it/s, Corr=0.863, Epoch=13, LR=7.5e-6, Valid_Loss=3.35]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.3886562388470987 ---> 3.3473781834313683)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.79it/s, Corr=0.826, Epoch=14, LR=5e-6, Train_Loss=4.04]    \n100%|██████████| 444/444 [00:10<00:00, 40.64it/s, Corr=0.864, Epoch=14, LR=5e-6, Valid_Loss=3.32]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.3473781834313683 ---> 3.316426324718413)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.82it/s, Corr=0.828, Epoch=15, LR=2.5e-6, Train_Loss=4.01]  \n100%|██████████| 444/444 [00:10<00:00, 40.50it/s, Corr=0.865, Epoch=15, LR=2.5e-6, Valid_Loss=3.28]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.316426324718413 ---> 3.2843621533822587)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.65it/s, Corr=0.829, Epoch=16, LR=6.7e-7, Train_Loss=3.97]  \n100%|██████████| 444/444 [00:11<00:00, 39.78it/s, Corr=0.866, Epoch=16, LR=6.7e-7, Valid_Loss=3.25]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.2843621533822587 ---> 3.2495191248732356)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.78it/s, Corr=0.83, Epoch=17, LR=0, Train_Loss=3.94]        \n100%|██████████| 444/444 [00:10<00:00, 43.88it/s, Corr=0.868, Epoch=17, LR=0, Valid_Loss=3.21]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.2495191248732356 ---> 3.214154979111674)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.26it/s, Corr=0.832, Epoch=18, LR=6.7e-7, Train_Loss=3.91]  \n100%|██████████| 444/444 [00:10<00:00, 43.84it/s, Corr=0.869, Epoch=18, LR=6.7e-7, Valid_Loss=3.19]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.214154979111674 ---> 3.191789592229726)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:31<00:00, 114.13it/s, Corr=0.833, Epoch=19, LR=2.5e-6, Train_Loss=3.89]  \n100%|██████████| 444/444 [00:10<00:00, 41.67it/s, Corr=0.87, Epoch=19, LR=2.5e-6, Valid_Loss=3.16] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.191789592229726 ---> 3.161293193701603)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.76it/s, Corr=0.834, Epoch=20, LR=5e-6, Train_Loss=3.86]    \n100%|██████████| 444/444 [00:10<00:00, 43.47it/s, Corr=0.871, Epoch=20, LR=5e-6, Valid_Loss=3.13]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.161293193701603 ---> 3.131065457008699)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:29<00:00, 118.56it/s, Corr=0.835, Epoch=21, LR=7.5e-6, Train_Loss=3.83]  \n100%|██████████| 444/444 [00:10<00:00, 42.97it/s, Corr=0.872, Epoch=21, LR=7.5e-6, Valid_Loss=3.11]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.131065457008699 ---> 3.1089307445983563)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:31<00:00, 114.50it/s, Corr=0.836, Epoch=22, LR=9.33e-6, Train_Loss=3.81] \n100%|██████████| 444/444 [00:10<00:00, 40.91it/s, Corr=0.873, Epoch=22, LR=9.33e-6, Valid_Loss=3.09]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.1089307445983563 ---> 3.0874448871713307)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.59it/s, Corr=0.836, Epoch=23, LR=1e-5, Train_Loss=3.79]    \n100%|██████████| 444/444 [00:10<00:00, 43.88it/s, Corr=0.873, Epoch=23, LR=1e-5, Valid_Loss=3.07]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.0874448871713307 ---> 3.0696187536824358)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.21it/s, Corr=0.837, Epoch=24, LR=9.33e-6, Train_Loss=3.77] \n100%|██████████| 444/444 [00:10<00:00, 42.04it/s, Corr=0.874, Epoch=24, LR=9.33e-6, Valid_Loss=3.05]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.0696187536824358 ---> 3.053767490210636)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 118.18it/s, Corr=0.838, Epoch=25, LR=7.5e-6, Train_Loss=3.75]  \n100%|██████████| 444/444 [00:11<00:00, 39.57it/s, Corr=0.875, Epoch=25, LR=7.5e-6, Valid_Loss=3.03]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.053767490210636 ---> 3.0286192437884125)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.26it/s, Corr=0.838, Epoch=26, LR=5e-6, Train_Loss=3.74]    \n100%|██████████| 444/444 [00:10<00:00, 41.13it/s, Corr=0.875, Epoch=26, LR=5e-6, Valid_Loss=3.01]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.0286192437884125 ---> 3.0142181038541636)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:31<00:00, 113.23it/s, Corr=0.839, Epoch=27, LR=2.5e-6, Train_Loss=3.72]  \n100%|██████████| 444/444 [00:10<00:00, 42.66it/s, Corr=0.876, Epoch=27, LR=2.5e-6, Valid_Loss=3]   \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.0142181038541636 ---> 2.9998788413793234)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.62it/s, Corr=0.84, Epoch=28, LR=6.7e-7, Train_Loss=3.71]  \n100%|██████████| 444/444 [00:10<00:00, 40.62it/s, Corr=0.876, Epoch=28, LR=6.7e-7, Valid_Loss=2.99]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.9998788413793234 ---> 2.9861281684385803)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.81it/s, Corr=0.84, Epoch=29, LR=0, Train_Loss=3.69]        \n100%|██████████| 444/444 [00:11<00:00, 39.67it/s, Corr=0.877, Epoch=29, LR=0, Valid_Loss=2.97]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.9861281684385803 ---> 2.967899898173713)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.03it/s, Corr=0.841, Epoch=30, LR=6.7e-7, Train_Loss=3.67]  \n100%|██████████| 444/444 [00:10<00:00, 41.46it/s, Corr=0.878, Epoch=30, LR=6.7e-7, Valid_Loss=2.95]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.967899898173713 ---> 2.9544261694584204)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:31<00:00, 114.30it/s, Corr=0.842, Epoch=31, LR=2.5e-6, Train_Loss=3.66]  \n100%|██████████| 444/444 [00:10<00:00, 43.46it/s, Corr=0.878, Epoch=31, LR=2.5e-6, Valid_Loss=2.94]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.9544261694584204 ---> 2.9391314369499906)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.59it/s, Corr=0.842, Epoch=32, LR=5e-6, Train_Loss=3.65]    \n100%|██████████| 444/444 [00:11<00:00, 40.30it/s, Corr=0.879, Epoch=32, LR=5e-6, Valid_Loss=2.93]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.9391314369499906 ---> 2.9266456103740515)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.80it/s, Corr=0.843, Epoch=33, LR=7.5e-6, Train_Loss=3.64]  \n100%|██████████| 444/444 [00:10<00:00, 41.38it/s, Corr=0.879, Epoch=33, LR=7.5e-6, Valid_Loss=2.91]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.9266456103740515 ---> 2.9128124478291952)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.72it/s, Corr=0.843, Epoch=34, LR=9.33e-6, Train_Loss=3.62] \n100%|██████████| 444/444 [00:09<00:00, 45.45it/s, Corr=0.88, Epoch=34, LR=9.33e-6, Valid_Loss=2.91] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.9128124478291952 ---> 2.9116368096192593)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 114.66it/s, Corr=0.844, Epoch=35, LR=1e-5, Train_Loss=3.61]    \n100%|██████████| 444/444 [00:10<00:00, 40.67it/s, Corr=0.88, Epoch=35, LR=1e-5, Valid_Loss=2.89] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.9116368096192593 ---> 2.8895060273832285)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.77it/s, Corr=0.844, Epoch=36, LR=9.33e-6, Train_Loss=3.61] \n100%|██████████| 444/444 [00:10<00:00, 42.97it/s, Corr=0.88, Epoch=36, LR=9.33e-6, Valid_Loss=2.89] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8895060273832285 ---> 2.8871891371778546)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:31<00:00, 113.69it/s, Corr=0.844, Epoch=37, LR=7.5e-6, Train_Loss=3.6]   \n100%|██████████| 444/444 [00:10<00:00, 42.41it/s, Corr=0.881, Epoch=37, LR=7.5e-6, Valid_Loss=2.88]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8871891371778546 ---> 2.8780717921941332)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.92it/s, Corr=0.845, Epoch=38, LR=5e-6, Train_Loss=3.59]    \n100%|██████████| 444/444 [00:11<00:00, 39.13it/s, Corr=0.881, Epoch=38, LR=5e-6, Valid_Loss=2.87]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8780717921941332 ---> 2.8715993832274)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.90it/s, Corr=0.845, Epoch=39, LR=2.5e-6, Train_Loss=3.58]  \n100%|██████████| 444/444 [00:10<00:00, 42.43it/s, Corr=0.881, Epoch=39, LR=2.5e-6, Valid_Loss=2.86]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8715993832274 ---> 2.861205612351555)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.32it/s, Corr=0.845, Epoch=40, LR=6.7e-7, Train_Loss=3.57]  \n100%|██████████| 444/444 [00:10<00:00, 43.20it/s, Corr=0.881, Epoch=40, LR=6.7e-7, Valid_Loss=2.85]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.861205612351555 ---> 2.8517929797484056)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.85it/s, Corr=0.846, Epoch=41, LR=0, Train_Loss=3.56]       \n100%|██████████| 444/444 [00:10<00:00, 41.47it/s, Corr=0.882, Epoch=41, LR=0, Valid_Loss=2.85]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8517929797484056 ---> 2.8453857619948404)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.70it/s, Corr=0.846, Epoch=42, LR=6.7e-7, Train_Loss=3.56]  \n100%|██████████| 444/444 [00:10<00:00, 40.69it/s, Corr=0.882, Epoch=42, LR=6.7e-7, Valid_Loss=2.84]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8453857619948404 ---> 2.837199746834012)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:29<00:00, 118.71it/s, Corr=0.846, Epoch=43, LR=2.5e-6, Train_Loss=3.56]  \n100%|██████████| 444/444 [00:10<00:00, 42.80it/s, Corr=0.882, Epoch=43, LR=2.5e-6, Valid_Loss=2.84]\n100%|██████████| 3550/3550 [00:30<00:00, 117.40it/s, Corr=0.846, Epoch=44, LR=5e-6, Train_Loss=3.54]    \n100%|██████████| 444/444 [00:10<00:00, 40.73it/s, Corr=0.882, Epoch=44, LR=5e-6, Valid_Loss=2.83]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.837199746834012 ---> 2.830822434092013)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.30it/s, Corr=0.847, Epoch=45, LR=7.5e-6, Train_Loss=3.54]  \n100%|██████████| 444/444 [00:10<00:00, 41.11it/s, Corr=0.883, Epoch=45, LR=7.5e-6, Valid_Loss=2.83]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.830822434092013 ---> 2.8274979677181844)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.47it/s, Corr=0.846, Epoch=46, LR=9.33e-6, Train_Loss=3.54] \n100%|██████████| 444/444 [00:10<00:00, 43.53it/s, Corr=0.883, Epoch=46, LR=9.33e-6, Valid_Loss=2.82]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8274979677181844 ---> 2.8199223370929065)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.45it/s, Corr=0.847, Epoch=47, LR=1e-5, Train_Loss=3.53]    \n100%|██████████| 444/444 [00:10<00:00, 41.82it/s, Corr=0.883, Epoch=47, LR=1e-5, Valid_Loss=2.82]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8199223370929065 ---> 2.81746477158814)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.07it/s, Corr=0.847, Epoch=48, LR=9.33e-6, Train_Loss=3.52] \n100%|██████████| 444/444 [00:11<00:00, 39.18it/s, Corr=0.883, Epoch=48, LR=9.33e-6, Valid_Loss=2.81]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.81746477158814 ---> 2.809659725882088)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.57it/s, Corr=0.847, Epoch=49, LR=7.5e-6, Train_Loss=3.52]  \n100%|██████████| 444/444 [00:10<00:00, 41.70it/s, Corr=0.883, Epoch=49, LR=7.5e-6, Valid_Loss=2.8] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.809659725882088 ---> 2.803507777346883)\nBest Loss: 2.8035\n\nfold = 1\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:29<00:00, 118.40it/s, Corr=0.288, Epoch=0, LR=9.33e-6, Train_Loss=13.4]  \n100%|██████████| 444/444 [00:10<00:00, 43.42it/s, Corr=0.713, Epoch=0, LR=9.33e-6, Valid_Loss=9.04]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (inf ---> 9.040361483178103)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:29<00:00, 119.24it/s, Corr=0.73, Epoch=1, LR=7.5e-6, Train_Loss=6.39]   \n100%|██████████| 444/444 [00:10<00:00, 41.22it/s, Corr=0.8, Epoch=1, LR=7.5e-6, Valid_Loss=4.75]  \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (9.040361483178103 ---> 4.752309166417405)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 118.07it/s, Corr=0.766, Epoch=2, LR=5e-6, Train_Loss=5.25]    \n100%|██████████| 444/444 [00:10<00:00, 43.76it/s, Corr=0.807, Epoch=2, LR=5e-6, Valid_Loss=4.51]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.752309166417405 ---> 4.507984515490887)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.82it/s, Corr=0.774, Epoch=3, LR=2.5e-6, Train_Loss=5.08]  \n100%|██████████| 444/444 [00:10<00:00, 43.92it/s, Corr=0.817, Epoch=3, LR=2.5e-6, Valid_Loss=4.32]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.507984515490887 ---> 4.319451452095044)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:29<00:00, 118.48it/s, Corr=0.788, Epoch=4, LR=6.7e-7, Train_Loss=4.83]  \n100%|██████████| 444/444 [00:10<00:00, 44.30it/s, Corr=0.834, Epoch=4, LR=6.7e-7, Valid_Loss=4.01]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.319451452095044 ---> 4.010569823252145)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 118.18it/s, Corr=0.802, Epoch=5, LR=0, Train_Loss=4.58]       \n100%|██████████| 444/444 [00:10<00:00, 40.49it/s, Corr=0.844, Epoch=5, LR=0, Valid_Loss=3.79]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.010569823252145 ---> 3.7885308997944684)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.36it/s, Corr=0.808, Epoch=6, LR=6.7e-7, Train_Loss=4.44]  \n100%|██████████| 444/444 [00:10<00:00, 43.11it/s, Corr=0.848, Epoch=6, LR=6.7e-7, Valid_Loss=3.7] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.7885308997944684 ---> 3.6951307058292366)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.77it/s, Corr=0.811, Epoch=7, LR=2.5e-6, Train_Loss=4.37]  \n100%|██████████| 444/444 [00:10<00:00, 43.52it/s, Corr=0.85, Epoch=7, LR=2.5e-6, Valid_Loss=3.64] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.6951307058292366 ---> 3.641700591258462)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:29<00:00, 119.34it/s, Corr=0.813, Epoch=8, LR=5e-6, Train_Loss=4.33]    \n100%|██████████| 444/444 [00:11<00:00, 39.37it/s, Corr=0.852, Epoch=8, LR=5e-6, Valid_Loss=3.59]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.641700591258462 ---> 3.589469149045949)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 115.04it/s, Corr=0.815, Epoch=9, LR=7.5e-6, Train_Loss=4.27]  \n100%|██████████| 444/444 [00:10<00:00, 42.29it/s, Corr=0.854, Epoch=9, LR=7.5e-6, Valid_Loss=3.54]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.589469149045949 ---> 3.5403937946681636)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.60it/s, Corr=0.818, Epoch=10, LR=9.33e-6, Train_Loss=4.22] \n100%|██████████| 444/444 [00:10<00:00, 42.28it/s, Corr=0.856, Epoch=10, LR=9.33e-6, Valid_Loss=3.49]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.5403937946681636 ---> 3.4851058288073116)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:31<00:00, 114.30it/s, Corr=0.82, Epoch=11, LR=1e-5, Train_Loss=4.17]    \n100%|██████████| 444/444 [00:11<00:00, 39.41it/s, Corr=0.858, Epoch=11, LR=1e-5, Valid_Loss=3.43]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.4851058288073116 ---> 3.4314339277378627)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 114.86it/s, Corr=0.822, Epoch=12, LR=9.33e-6, Train_Loss=4.12] \n100%|██████████| 444/444 [00:10<00:00, 42.41it/s, Corr=0.86, Epoch=12, LR=9.33e-6, Valid_Loss=3.39] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.4314339277378627 ---> 3.3937613032489287)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 117.42it/s, Corr=0.824, Epoch=13, LR=7.5e-6, Train_Loss=4.08]  \n100%|██████████| 444/444 [00:10<00:00, 42.95it/s, Corr=0.861, Epoch=13, LR=7.5e-6, Valid_Loss=3.36]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.3937613032489287 ---> 3.3624762570780335)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 3550/3550 [00:30<00:00, 116.89it/s, Corr=0.825, Epoch=14, LR=5e-6, Train_Loss=4.05]    \n100%|██████████| 444/444 [00:10<00:00, 42.00it/s, Corr=0.863, Epoch=14, LR=5e-6, Valid_Loss=3.33]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.3624762570780335 ---> 3.331233226927596)\n","output_type":"stream"},{"name":"stderr","text":" 54%|█████▍    | 1920/3550 [00:15<00:16, 97.17it/s, Corr=0.826, Epoch=15, LR=9.76e-6, Train_Loss=4.03]  ","output_type":"stream"}]},{"cell_type":"code","source":"PATH = '/kaggle/input/model-latest/latest_model_training_stage'\n# PATH = '/kaggle/working/model_f0.bin'\n# torch.save(model.state_dict(), PATH)","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:11:13.384832Z","iopub.execute_input":"2022-12-18T08:11:13.385200Z","iopub.status.idle":"2022-12-18T08:11:13.390476Z","shell.execute_reply.started":"2022-12-18T08:11:13.385168Z","shell.execute_reply":"2022-12-18T08:11:13.389257Z"},"trusted":true},"execution_count":14,"outputs":[]},{"cell_type":"code","source":"","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def predict(fold):\n    preds = list()\n    ds = CtieseqDataset(cite_test_x)\n    dl = DataLoader(ds, batch_size=32, shuffle=False)\n    model = CtieseqModel().to(device)\n    model.load_state_dict(torch.load(f\"model_f{fold}.bin\"))\n    model.eval()\n\n    bar = tqdm(enumerate(dl), total=len(dl))\n    for step, data in bar:        \n        X = data[\"X\"]\n\n        batch_size = X.size(0)\n\n        outputs = model(X.cuda())\n        preds.append(outputs.detach().cpu().numpy())\n    test_pred = np.concatenate(preds)\n    return test_pred\n\ntest_preds = np.array([predict(0), predict(1), predict(2)])\ntest_preds = np.mean(test_preds, axis=0)\nprint(test_preds.shape)","metadata":{"execution":{"iopub.status.busy":"2022-12-18T06:33:28.255340Z","iopub.execute_input":"2022-12-18T06:33:28.255685Z","iopub.status.idle":"2022-12-18T06:33:28.282241Z","shell.execute_reply.started":"2022-12-18T06:33:28.255657Z","shell.execute_reply":"2022-12-18T06:33:28.280673Z"},"collapsed":true,"jupyter":{"outputs_hidden":true},"trusted":true},"execution_count":25,"outputs":[{"traceback":["\u001b[0;31m---------------------------------------------------------------------------\u001b[0m","\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)","\u001b[0;32m/tmp/ipykernel_23/353110774.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m     18\u001b[0m     \u001b[0;32mreturn\u001b[0m \u001b[0mtest_pred\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     19\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 20\u001b[0;31m \u001b[0mtest_preds\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0marray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpredict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpredict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     21\u001b[0m \u001b[0mtest_preds\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmean\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtest_preds\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     22\u001b[0m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtest_preds\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/tmp/ipykernel_23/353110774.py\u001b[0m in \u001b[0;36mpredict\u001b[0;34m(fold)\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mpredict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfold\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      2\u001b[0m     \u001b[0mpreds\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m     \u001b[0mds\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mCtieseqDataset\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcite_test_x\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      4\u001b[0m     \u001b[0mdl\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mDataLoader\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mds\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_size\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m32\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mshuffle\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m     \u001b[0mmodel\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mCtieseqModel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdevice\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;31mNameError\u001b[0m: name 'cite_test_x' is not defined"],"ename":"NameError","evalue":"name 'cite_test_x' is not defined","output_type":"error"}]},{"cell_type":"markdown","source":"Making train Test splits","metadata":{"execution":{"iopub.status.busy":"2022-12-17T08:56:27.700804Z","iopub.execute_input":"2022-12-17T08:56:27.701478Z","iopub.status.idle":"2022-12-17T08:56:27.741026Z","shell.execute_reply.started":"2022-12-17T08:56:27.701435Z","shell.execute_reply":"2022-12-17T08:56:27.739320Z"}}},{"cell_type":"code","source":"hdf = pd.HDFStore('/kaggle/input/open-problems-multimodal/train_cite_inputs.h5', mode='r')\nhdf.keys()\ndf1 = hdf.get(hdf.keys()[0])\ndf1.head()","metadata":{"execution":{"iopub.status.busy":"2022-12-17T09:36:02.834584Z","iopub.execute_input":"2022-12-17T09:36:02.834952Z","iopub.status.idle":"2022-12-17T09:36:54.312855Z","shell.execute_reply.started":"2022-12-17T09:36:02.834913Z","shell.execute_reply":"2022-12-17T09:36:54.311893Z"},"trusted":true},"execution_count":4,"outputs":[{"execution_count":4,"output_type":"execute_result","data":{"text/plain":"gene_id       ENSG00000121410_A1BG  ENSG00000268895_A1BG-AS1  \\\ncell_id                                                        \n45006fe3e4c8                   0.0                       0.0   \nd02759a80ba2                   0.0                       0.0   \nc016c6b0efa5                   0.0                       0.0   \nba7f733a4f75                   0.0                       0.0   \nfbcf2443ffb2                   0.0                       0.0   \n\ngene_id       ENSG00000175899_A2M  ENSG00000245105_A2M-AS1  \\\ncell_id                                                      \n45006fe3e4c8                  0.0                      0.0   \nd02759a80ba2                  0.0                      0.0   \nc016c6b0efa5                  0.0                      0.0   \nba7f733a4f75                  0.0                      0.0   \nfbcf2443ffb2                  0.0                      0.0   \n\ngene_id       ENSG00000166535_A2ML1  ENSG00000128274_A4GALT  \\\ncell_id                                                       \n45006fe3e4c8                    0.0                0.000000   \nd02759a80ba2                    0.0                0.000000   \nc016c6b0efa5                    0.0                3.847321   \nba7f733a4f75                    0.0                0.000000   \nfbcf2443ffb2                    0.0                0.000000   \n\ngene_id       ENSG00000094914_AAAS  ENSG00000081760_AACS  \\\ncell_id                                                    \n45006fe3e4c8              0.000000              0.000000   \nd02759a80ba2              0.000000              0.000000   \nc016c6b0efa5              0.000000              3.847321   \nba7f733a4f75              3.436846              3.436846   \nfbcf2443ffb2              0.000000              4.196826   \n\ngene_id       ENSG00000109576_AADAT  ENSG00000103591_AAGAB  ...  \\\ncell_id                                                     ...   \n45006fe3e4c8               0.000000                    0.0  ...   \nd02759a80ba2               0.000000                    0.0  ...   \nc016c6b0efa5               3.847321                    0.0  ...   \nba7f733a4f75               0.000000                    0.0  ...   \nfbcf2443ffb2               0.000000                    0.0  ...   \n\ngene_id       ENSG00000153975_ZUP1  ENSG00000086827_ZW10  \\\ncell_id                                                    \n45006fe3e4c8              0.000000              0.000000   \nd02759a80ba2              0.000000              0.000000   \nc016c6b0efa5              0.000000              0.000000   \nba7f733a4f75              3.436846              0.000000   \nfbcf2443ffb2              0.000000              4.196826   \n\ngene_id       ENSG00000174442_ZWILCH  ENSG00000122952_ZWINT  \\\ncell_id                                                       \n45006fe3e4c8                0.000000               0.000000   \nd02759a80ba2                0.000000               4.039545   \nc016c6b0efa5                3.847321               4.529743   \nba7f733a4f75                4.113780               5.020215   \nfbcf2443ffb2                4.196826               4.196826   \n\ngene_id       ENSG00000198205_ZXDA  ENSG00000198455_ZXDB  \\\ncell_id                                                    \n45006fe3e4c8                   0.0                   0.0   \nd02759a80ba2                   0.0                   0.0   \nc016c6b0efa5                   0.0                   0.0   \nba7f733a4f75                   0.0                   0.0   \nfbcf2443ffb2                   0.0                   0.0   \n\ngene_id       ENSG00000070476_ZXDC  ENSG00000162378_ZYG11B  \\\ncell_id                                                      \n45006fe3e4c8               0.00000                0.000000   \nd02759a80ba2               0.00000                0.000000   \nc016c6b0efa5               0.00000                3.847321   \nba7f733a4f75               0.00000                3.436846   \nfbcf2443ffb2               3.51861                4.196826   \n\ngene_id       ENSG00000159840_ZYX  ENSG00000074755_ZZEF1  \ncell_id                                                   \n45006fe3e4c8             4.090185                    0.0  \nd02759a80ba2             0.000000                    0.0  \nc016c6b0efa5             3.847321                    0.0  \nba7f733a4f75             4.113780                    0.0  \nfbcf2443ffb2             3.518610                    0.0  \n\n[5 rows x 22050 columns]","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>gene_id</th>\n      <th>ENSG00000121410_A1BG</th>\n      <th>ENSG00000268895_A1BG-AS1</th>\n      <th>ENSG00000175899_A2M</th>\n      <th>ENSG00000245105_A2M-AS1</th>\n      <th>ENSG00000166535_A2ML1</th>\n      <th>ENSG00000128274_A4GALT</th>\n      <th>ENSG00000094914_AAAS</th>\n      <th>ENSG00000081760_AACS</th>\n      <th>ENSG00000109576_AADAT</th>\n      <th>ENSG00000103591_AAGAB</th>\n      <th>...</th>\n      <th>ENSG00000153975_ZUP1</th>\n      <th>ENSG00000086827_ZW10</th>\n      <th>ENSG00000174442_ZWILCH</th>\n      <th>ENSG00000122952_ZWINT</th>\n      <th>ENSG00000198205_ZXDA</th>\n      <th>ENSG00000198455_ZXDB</th>\n      <th>ENSG00000070476_ZXDC</th>\n      <th>ENSG00000162378_ZYG11B</th>\n      <th>ENSG00000159840_ZYX</th>\n      <th>ENSG00000074755_ZZEF1</th>\n    </tr>\n    <tr>\n      <th>cell_id</th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>45006fe3e4c8</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.00000</td>\n      <td>0.000000</td>\n      <td>4.090185</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>d02759a80ba2</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>4.039545</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.00000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>c016c6b0efa5</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>3.847321</td>\n      <td>0.000000</td>\n      <td>3.847321</td>\n      <td>3.847321</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>3.847321</td>\n      <td>4.529743</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.00000</td>\n      <td>3.847321</td>\n      <td>3.847321</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>ba7f733a4f75</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.000000</td>\n      <td>3.436846</td>\n      <td>3.436846</td>\n      <td>0.000000</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>3.436846</td>\n      <td>0.000000</td>\n      <td>4.113780</td>\n      <td>5.020215</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.00000</td>\n      <td>3.436846</td>\n      <td>4.113780</td>\n      <td>0.0</td>\n    </tr>\n    <tr>\n      <th>fbcf2443ffb2</th>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>4.196826</td>\n      <td>0.000000</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0.000000</td>\n      <td>4.196826</td>\n      <td>4.196826</td>\n      <td>4.196826</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>3.51861</td>\n      <td>4.196826</td>\n      <td>3.518610</td>\n      <td>0.0</td>\n    </tr>\n  </tbody>\n</table>\n<p>5 rows × 22050 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"hdf = pd.HDFStore('/kaggle/input/open-problems-multimodal/train_cite_targets.h5', mode='r')\nhdf.keys()\ndf1 = hdf.get(hdf.keys()[0])\ndf1.head()","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"gc.collect()\nX = np.load('/kaggle/input/trainx/cite_train_x.npy', mmap_mode=None, allow_pickle=False, fix_imports=True, encoding='ASCII')\nX.shape","metadata":{"execution":{"iopub.status.busy":"2022-12-17T13:28:34.249920Z","iopub.execute_input":"2022-12-17T13:28:34.250336Z","iopub.status.idle":"2022-12-17T13:28:34.398348Z","shell.execute_reply.started":"2022-12-17T13:28:34.250303Z","shell.execute_reply":"2022-12-17T13:28:34.397274Z"},"trusted":true},"execution_count":7,"outputs":[{"execution_count":7,"output_type":"execute_result","data":{"text/plain":"(70988, 240)"},"metadata":{}}]},{"cell_type":"code","source":"gc.collect()\nY = pd.read_hdf(FP_CITE_TRAIN_TARGETS).values\nY.shape","metadata":{"execution":{"iopub.status.busy":"2022-12-17T13:28:34.748829Z","iopub.execute_input":"2022-12-17T13:28:34.749677Z","iopub.status.idle":"2022-12-17T13:28:35.634382Z","shell.execute_reply.started":"2022-12-17T13:28:34.749634Z","shell.execute_reply":"2022-12-17T13:28:35.633277Z"},"trusted":true},"execution_count":8,"outputs":[{"execution_count":8,"output_type":"execute_result","data":{"text/plain":"(70988, 140)"},"metadata":{}}]},{"cell_type":"code","source":"from sklearn.model_selection import train_test_split\nx_train, x_test, y_train, y_test = train_test_split(X,Y ,\n                                   random_state=0, \n                                   test_size=0.25, \n                                   shuffle=True)","metadata":{"execution":{"iopub.status.busy":"2022-12-17T13:28:36.889125Z","iopub.execute_input":"2022-12-17T13:28:36.889637Z","iopub.status.idle":"2022-12-17T13:28:37.020182Z","shell.execute_reply.started":"2022-12-17T13:28:36.889593Z","shell.execute_reply":"2022-12-17T13:28:37.019024Z"},"trusted":true},"execution_count":9,"outputs":[]},{"cell_type":"code","source":"x_train = np.load('/kaggle/input/dataset/x_train.npy')\nx_test = np.load('/kaggle/input/dataset/x_test.npy')\ny_train = np.load('/kaggle/input/dataset/y_train.npy')\ny_test = np.load('/kaggle/input/dataset/y_test.npy')","metadata":{"execution":{"iopub.status.busy":"2022-12-18T07:48:36.423701Z","iopub.execute_input":"2022-12-18T07:48:36.424074Z","iopub.status.idle":"2022-12-18T07:48:37.844984Z","shell.execute_reply.started":"2022-12-18T07:48:36.424040Z","shell.execute_reply":"2022-12-18T07:48:37.843989Z"},"trusted":true},"execution_count":12,"outputs":[]},{"cell_type":"code","source":"print(x_train.shape)\nprint(y_train.shape)\nprint(x_test.shape)\nprint(y_test.shape)","metadata":{"execution":{"iopub.status.busy":"2022-12-18T07:48:38.014146Z","iopub.execute_input":"2022-12-18T07:48:38.014460Z","iopub.status.idle":"2022-12-18T07:48:38.020240Z","shell.execute_reply.started":"2022-12-18T07:48:38.014434Z","shell.execute_reply":"2022-12-18T07:48:38.018815Z"},"trusted":true},"execution_count":13,"outputs":[{"name":"stdout","text":"(53241, 240)\n(53241, 140)\n(17747, 240)\n(17747, 140)\n","output_type":"stream"}]},{"cell_type":"code","source":"# np.save('/kaggle/working/x_train.npy', x_train)\n# np.save('/kaggle/working/x_test.npy', x_test)\n# np.save('/kaggle/working/y_train.npy', y_train)\n# np.save('/kaggle/working/y_test.npy', y_test)","metadata":{"execution":{"iopub.status.busy":"2022-12-17T09:49:18.376235Z","iopub.execute_input":"2022-12-17T09:49:18.376906Z","iopub.status.idle":"2022-12-17T09:49:18.469788Z","shell.execute_reply.started":"2022-12-17T09:49:18.376870Z","shell.execute_reply":"2022-12-17T09:49:18.468783Z"},"trusted":true},"execution_count":22,"outputs":[]},{"cell_type":"markdown","source":"### retraining the model again","metadata":{}},{"cell_type":"code","source":"kf = KFold(n_splits=3, shuffle=True, random_state=42)\nscore_list = []\nfor fold, (idx_tr, idx_va) in enumerate(kf.split(x_train)):\n    print(f\"\\nfold = {fold}\")\n    X_tr = x_train[idx_tr] \n    y_tr = y_train[idx_tr]\n    \n    X_va = x_train[idx_va]\n    y_va = y_train[idx_va]\n    \n    ds_tr = CtieseqDataset(X_tr, y_tr)\n    ds_va = CtieseqDataset(X_tr, y_tr)\n    dl_tr = DataLoader(ds_tr, batch_size=CFG.tr_batch_size, shuffle=True)\n    dl_va = DataLoader(ds_va, batch_size=CFG.va_batch_size, shuffle=False)\n    \n    model = CtieseqModel()\n    optimizer = get_optimizer(model, CFG.lr, CFG.weight_decay, CFG.betas)\n    scheduler = get_scheduler(optimizer) \n\n    train_one_fold(model, optimizer, scheduler, dl_tr, dl_va, fold)\n    ","metadata":{"execution":{"iopub.status.busy":"2022-12-18T07:14:14.572823Z","iopub.execute_input":"2022-12-18T07:14:14.573196Z"},"trusted":true},"execution_count":null,"outputs":[{"name":"stdout","text":"\nfold = 0\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 138.29it/s, Corr=0.15, Epoch=0, LR=3.41e-6, Train_Loss=14.6]   \n100%|██████████| 278/278 [00:05<00:00, 48.83it/s, Corr=0.431, Epoch=0, LR=3.41e-6, Valid_Loss=12.8]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (inf ---> 12.83895823537326)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 132.75it/s, Corr=0.595, Epoch=1, LR=1.02e-6, Train_Loss=9.52] \n100%|██████████| 278/278 [00:05<00:00, 48.00it/s, Corr=0.748, Epoch=1, LR=1.02e-6, Valid_Loss=6.38]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (12.83895823537326 ---> 6.3782445658037155)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 137.51it/s, Corr=0.738, Epoch=2, LR=9.14e-6, Train_Loss=5.93] \n100%|██████████| 278/278 [00:06<00:00, 41.72it/s, Corr=0.799, Epoch=2, LR=9.14e-6, Valid_Loss=4.77]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (6.3782445658037155 ---> 4.766546098361212)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 135.89it/s, Corr=0.766, Epoch=3, LR=6.34e-6, Train_Loss=5.29] \n100%|██████████| 278/278 [00:05<00:00, 48.93it/s, Corr=0.809, Epoch=3, LR=6.34e-6, Valid_Loss=4.52]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.766546098361212 ---> 4.52016820744769)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 133.59it/s, Corr=0.774, Epoch=4, LR=6.85e-9, Train_Loss=5.12] \n100%|██████████| 278/278 [00:05<00:00, 49.04it/s, Corr=0.816, Epoch=4, LR=6.85e-9, Valid_Loss=4.36]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.52016820744769 ---> 4.364926959357678)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 133.64it/s, Corr=0.782, Epoch=5, LR=6.84e-6, Train_Loss=4.96] \n100%|██████████| 278/278 [00:05<00:00, 50.81it/s, Corr=0.826, Epoch=5, LR=6.84e-6, Valid_Loss=4.18]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.364926959357678 ---> 4.175362367377978)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 137.78it/s, Corr=0.792, Epoch=6, LR=8.82e-6, Train_Loss=4.78] \n100%|██████████| 278/278 [00:05<00:00, 46.39it/s, Corr=0.835, Epoch=6, LR=8.82e-6, Valid_Loss=4]   \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.175362367377978 ---> 3.997922256294221)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 137.16it/s, Corr=0.8, Epoch=7, LR=7.23e-7, Train_Loss=4.63]   \n100%|██████████| 278/278 [00:05<00:00, 51.81it/s, Corr=0.842, Epoch=7, LR=7.23e-7, Valid_Loss=3.86]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.997922256294221 ---> 3.8571279291946223)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 133.39it/s, Corr=0.805, Epoch=8, LR=3.91e-6, Train_Loss=4.52] \n100%|██████████| 278/278 [00:05<00:00, 51.35it/s, Corr=0.845, Epoch=8, LR=3.91e-6, Valid_Loss=3.77]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.8571279291946223 ---> 3.7715219666522986)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 136.20it/s, Corr=0.807, Epoch=9, LR=9.97e-6, Train_Loss=4.46] \n100%|██████████| 278/278 [00:05<00:00, 47.93it/s, Corr=0.846, Epoch=9, LR=9.97e-6, Valid_Loss=3.73]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.7715219666522986 ---> 3.7306974628592715)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 135.89it/s, Corr=0.808, Epoch=10, LR=2.92e-6, Train_Loss=4.42] \n100%|██████████| 278/278 [00:05<00:00, 49.88it/s, Corr=0.847, Epoch=10, LR=2.92e-6, Valid_Loss=3.69]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.7306974628592715 ---> 3.6941722232522833)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 131.92it/s, Corr=0.81, Epoch=11, LR=1.36e-6, Train_Loss=4.39] \n100%|██████████| 278/278 [00:05<00:00, 51.32it/s, Corr=0.848, Epoch=11, LR=1.36e-6, Valid_Loss=3.67]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.6941722232522833 ---> 3.6654699692935977)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 131.27it/s, Corr=0.811, Epoch=12, LR=9.41e-6, Train_Loss=4.37]\n100%|██████████| 278/278 [00:05<00:00, 50.62it/s, Corr=0.849, Epoch=12, LR=9.41e-6, Valid_Loss=3.64]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.6654699692935977 ---> 3.6388515536869894)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 136.29it/s, Corr=0.812, Epoch=13, LR=5.83e-6, Train_Loss=4.34] \n100%|██████████| 278/278 [00:05<00:00, 50.65it/s, Corr=0.85, Epoch=13, LR=5.83e-6, Valid_Loss=3.62]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.6388515536869894 ---> 3.6180819644654996)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 133.59it/s, Corr=0.813, Epoch=14, LR=6.16e-8, Train_Loss=4.31] \n100%|██████████| 278/278 [00:05<00:00, 50.77it/s, Corr=0.851, Epoch=14, LR=6.16e-8, Valid_Loss=3.58]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.6180819644654996 ---> 3.584818953828838)\nBest Loss: 3.5848\n\nfold = 1\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 133.96it/s, Corr=0.201, Epoch=0, LR=3.41e-6, Train_Loss=14.5] \n100%|██████████| 278/278 [00:05<00:00, 49.93it/s, Corr=0.492, Epoch=0, LR=3.41e-6, Valid_Loss=13] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (inf ---> 12.978757527846245)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 135.24it/s, Corr=0.638, Epoch=1, LR=1.02e-6, Train_Loss=9.63] \n100%|██████████| 278/278 [00:06<00:00, 45.41it/s, Corr=0.773, Epoch=1, LR=1.02e-6, Valid_Loss=6.26]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (12.978757527846245 ---> 6.262137840853723)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:15<00:00, 138.98it/s, Corr=0.751, Epoch=2, LR=9.14e-6, Train_Loss=5.79] \n100%|██████████| 278/278 [00:05<00:00, 46.59it/s, Corr=0.805, Epoch=2, LR=9.14e-6, Valid_Loss=4.67]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (6.262137840853723 ---> 4.6733767648693565)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 136.42it/s, Corr=0.77, Epoch=3, LR=6.34e-6, Train_Loss=5.21]  \n100%|██████████| 278/278 [00:05<00:00, 48.24it/s, Corr=0.813, Epoch=3, LR=6.34e-6, Valid_Loss=4.45]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.6733767648693565 ---> 4.448255320364316)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 133.90it/s, Corr=0.779, Epoch=4, LR=6.85e-9, Train_Loss=5.02] \n100%|██████████| 278/278 [00:05<00:00, 47.67it/s, Corr=0.822, Epoch=4, LR=6.85e-9, Valid_Loss=4.25]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.448255320364316 ---> 4.247789290963331)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 137.22it/s, Corr=0.789, Epoch=5, LR=6.84e-6, Train_Loss=4.82] \n100%|██████████| 278/278 [00:05<00:00, 47.11it/s, Corr=0.834, Epoch=5, LR=6.84e-6, Valid_Loss=4.02]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.247789290963331 ---> 4.024350032595747)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 135.45it/s, Corr=0.799, Epoch=6, LR=8.82e-6, Train_Loss=4.64] \n100%|██████████| 278/278 [00:05<00:00, 46.78it/s, Corr=0.841, Epoch=6, LR=8.82e-6, Valid_Loss=3.87]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (4.024350032595747 ---> 3.868015586258479)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 136.71it/s, Corr=0.805, Epoch=7, LR=7.23e-7, Train_Loss=4.52] \n100%|██████████| 278/278 [00:05<00:00, 48.40it/s, Corr=0.845, Epoch=7, LR=7.23e-7, Valid_Loss=3.77]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.868015586258479 ---> 3.7714141571750814)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:15<00:00, 138.74it/s, Corr=0.807, Epoch=8, LR=3.91e-6, Train_Loss=4.46] \n100%|██████████| 278/278 [00:05<00:00, 47.38it/s, Corr=0.847, Epoch=8, LR=3.91e-6, Valid_Loss=3.72]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.7714141571750814 ---> 3.718973805335177)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:15<00:00, 139.77it/s, Corr=0.809, Epoch=9, LR=9.97e-6, Train_Loss=4.41] \n100%|██████████| 278/278 [00:05<00:00, 49.66it/s, Corr=0.847, Epoch=9, LR=9.97e-6, Valid_Loss=3.69]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.718973805335177 ---> 3.6916139647746653)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 136.11it/s, Corr=0.809, Epoch=10, LR=2.92e-6, Train_Loss=4.39] \n100%|██████████| 278/278 [00:05<00:00, 46.89it/s, Corr=0.848, Epoch=10, LR=2.92e-6, Valid_Loss=3.66]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.6916139647746653 ---> 3.664633258870738)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 136.28it/s, Corr=0.811, Epoch=11, LR=1.36e-6, Train_Loss=4.36] \n100%|██████████| 278/278 [00:05<00:00, 49.84it/s, Corr=0.849, Epoch=11, LR=1.36e-6, Valid_Loss=3.65]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.664633258870738 ---> 3.6480704582583208)\n","output_type":"stream"},{"name":"stderr","text":" 83%|████████▎ | 1846/2219 [00:13<00:02, 141.01it/s, Corr=0.811, Epoch=12, LR=7.9e-8, Train_Loss=4.35]  ","output_type":"stream"}]},{"cell_type":"code","source":"model = CtieseqModel()\noptimizer = get_optimizer(model, CFG.lr, CFG.weight_decay, CFG.betas)\nscheduler = get_scheduler(optimizer) \ncheckpoint = torch.load(PATH)\nmodel.load_state_dict(checkpoint['model_state_dict'])\noptimizer.load_state_dict(checkpoint['optimizer_state_dict'])\nepoch = checkpoint['epoch']\nloss = checkpoint['loss']\n\n# model.eval()\n# - or -\nkf = KFold(n_splits=3, shuffle=True, random_state=42)\nscore_list = []\nfor fold, (idx_tr, idx_va) in enumerate(kf.split(x_train)):\n    print(f\"\\nfold = {fold}\")\n    X_tr = x_train[idx_tr] \n    y_tr = y_train[idx_tr]\n    \n    X_va = x_train[idx_va]\n    y_va = y_train[idx_va]\n    \n    ds_tr = CtieseqDataset(X_tr, y_tr)\n    ds_va = CtieseqDataset(X_tr, y_tr)\n    dl_tr = DataLoader(ds_tr, batch_size=CFG.tr_batch_size, shuffle=True)\n    dl_va = DataLoader(ds_va, batch_size=CFG.va_batch_size, shuffle=False)\n    \n#     model = CtieseqModel()\n    optimizer = get_optimizer(model, CFG.lr, CFG.weight_decay, CFG.betas)\n#     scheduler = get_scheduler(optimizer) \n\n    train_one_fold(model, optimizer, scheduler, dl_tr, dl_va, fold)\n    ","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:15:19.266443Z","iopub.execute_input":"2022-12-18T08:15:19.266793Z","iopub.status.idle":"2022-12-18T08:33:08.926971Z","shell.execute_reply.started":"2022-12-18T08:15:19.266762Z","shell.execute_reply":"2022-12-18T08:33:08.925786Z"},"trusted":true},"execution_count":25,"outputs":[{"name":"stdout","text":"\nfold = 0\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 130.44it/s, Corr=0.816, Epoch=0, LR=1e-5, Train_Loss=4.25]\n100%|██████████| 278/278 [00:05<00:00, 48.56it/s, Corr=0.855, Epoch=0, LR=1e-5, Valid_Loss=3.52]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (inf ---> 3.5201383831602167)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 126.52it/s, Corr=0.819, Epoch=1, LR=1e-5, Train_Loss=4.19]\n100%|██████████| 278/278 [00:06<00:00, 45.72it/s, Corr=0.857, Epoch=1, LR=1e-5, Valid_Loss=3.45]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.5201383831602167 ---> 3.4489724093869403)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 127.42it/s, Corr=0.822, Epoch=2, LR=1e-5, Train_Loss=4.13]\n100%|██████████| 278/278 [00:06<00:00, 46.19it/s, Corr=0.86, Epoch=2, LR=1e-5, Valid_Loss=3.39] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.4489724093869403 ---> 3.394337240364099)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 130.48it/s, Corr=0.824, Epoch=3, LR=1e-5, Train_Loss=4.08]\n100%|██████████| 278/278 [00:06<00:00, 43.02it/s, Corr=0.862, Epoch=3, LR=1e-5, Valid_Loss=3.35]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.394337240364099 ---> 3.346235619737523)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 131.51it/s, Corr=0.826, Epoch=4, LR=1e-5, Train_Loss=4.03]\n100%|██████████| 278/278 [00:06<00:00, 45.24it/s, Corr=0.863, Epoch=4, LR=1e-5, Valid_Loss=3.31]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.346235619737523 ---> 3.3101916173185946)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 127.00it/s, Corr=0.827, Epoch=5, LR=1e-5, Train_Loss=4]   \n100%|██████████| 278/278 [00:05<00:00, 49.54it/s, Corr=0.865, Epoch=5, LR=1e-5, Valid_Loss=3.27]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.3101916173185946 ---> 3.2722702248301876)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 127.30it/s, Corr=0.829, Epoch=6, LR=1e-5, Train_Loss=3.96]\n100%|██████████| 278/278 [00:05<00:00, 47.49it/s, Corr=0.866, Epoch=6, LR=1e-5, Valid_Loss=3.24]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.2722702248301876 ---> 3.237046741905928)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 131.47it/s, Corr=0.831, Epoch=7, LR=1e-5, Train_Loss=3.92]\n100%|██████████| 278/278 [00:06<00:00, 43.96it/s, Corr=0.868, Epoch=7, LR=1e-5, Valid_Loss=3.2] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.237046741905928 ---> 3.1961116357918233)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 131.88it/s, Corr=0.832, Epoch=8, LR=1e-5, Train_Loss=3.88]\n100%|██████████| 278/278 [00:05<00:00, 49.50it/s, Corr=0.869, Epoch=8, LR=1e-5, Valid_Loss=3.16]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.1961116357918233 ---> 3.160769662622224)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 125.36it/s, Corr=0.834, Epoch=9, LR=1e-5, Train_Loss=3.85]\n100%|██████████| 278/278 [00:05<00:00, 47.17it/s, Corr=0.871, Epoch=9, LR=1e-5, Valid_Loss=3.12]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.160769662622224 ---> 3.1246472954407487)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 128.08it/s, Corr=0.835, Epoch=10, LR=1e-5, Train_Loss=3.81]\n100%|██████████| 278/278 [00:06<00:00, 44.69it/s, Corr=0.872, Epoch=10, LR=1e-5, Valid_Loss=3.1] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.1246472954407487 ---> 3.0995513850509964)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 133.24it/s, Corr=0.836, Epoch=11, LR=1e-5, Train_Loss=3.79]\n100%|██████████| 278/278 [00:06<00:00, 41.00it/s, Corr=0.873, Epoch=11, LR=1e-5, Valid_Loss=3.07]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.0995513850509964 ---> 3.0716420037045022)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 129.10it/s, Corr=0.837, Epoch=12, LR=1e-5, Train_Loss=3.77]\n100%|██████████| 278/278 [00:05<00:00, 46.69it/s, Corr=0.874, Epoch=12, LR=1e-5, Valid_Loss=3.05]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.0716420037045022 ---> 3.051570672239795)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 124.27it/s, Corr=0.838, Epoch=13, LR=1e-5, Train_Loss=3.75]\n100%|██████████| 278/278 [00:05<00:00, 47.02it/s, Corr=0.875, Epoch=13, LR=1e-5, Valid_Loss=3.02]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.051570672239795 ---> 3.022599682174092)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 126.34it/s, Corr=0.839, Epoch=14, LR=1e-5, Train_Loss=3.73]\n100%|██████████| 278/278 [00:05<00:00, 46.79it/s, Corr=0.876, Epoch=14, LR=1e-5, Valid_Loss=3]   \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (3.022599682174092 ---> 2.9997779636266984)\nBest Loss: 2.9998\n\nfold = 1\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 128.17it/s, Corr=0.839, Epoch=0, LR=1e-5, Train_Loss=3.7] \n100%|██████████| 278/278 [00:06<00:00, 42.66it/s, Corr=0.876, Epoch=0, LR=1e-5, Valid_Loss=2.99]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (inf ---> 2.9865812448848836)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 131.38it/s, Corr=0.84, Epoch=1, LR=1e-5, Train_Loss=3.69] \n100%|██████████| 278/278 [00:06<00:00, 44.08it/s, Corr=0.877, Epoch=1, LR=1e-5, Valid_Loss=2.97]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.9865812448848836 ---> 2.9662930634939726)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 129.18it/s, Corr=0.841, Epoch=2, LR=1e-5, Train_Loss=3.67]\n100%|██████████| 278/278 [00:06<00:00, 45.83it/s, Corr=0.878, Epoch=2, LR=1e-5, Valid_Loss=2.95]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.9662930634939726 ---> 2.953176336516298)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 127.97it/s, Corr=0.842, Epoch=3, LR=1e-5, Train_Loss=3.66]\n100%|██████████| 278/278 [00:05<00:00, 47.51it/s, Corr=0.878, Epoch=3, LR=1e-5, Valid_Loss=2.93]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.953176336516298 ---> 2.931879094239859)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 130.34it/s, Corr=0.842, Epoch=4, LR=1e-5, Train_Loss=3.64]\n100%|██████████| 278/278 [00:06<00:00, 45.25it/s, Corr=0.879, Epoch=4, LR=1e-5, Valid_Loss=2.92]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.931879094239859 ---> 2.9151735245466113)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 128.70it/s, Corr=0.843, Epoch=5, LR=1e-5, Train_Loss=3.62]\n100%|██████████| 278/278 [00:05<00:00, 49.40it/s, Corr=0.88, Epoch=5, LR=1e-5, Valid_Loss=2.9]  \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.9151735245466113 ---> 2.9038276106241745)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 129.83it/s, Corr=0.844, Epoch=6, LR=1e-5, Train_Loss=3.6] \n100%|██████████| 278/278 [00:05<00:00, 48.78it/s, Corr=0.88, Epoch=6, LR=1e-5, Valid_Loss=2.89]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.9038276106241745 ---> 2.8891454338523617)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 128.96it/s, Corr=0.844, Epoch=7, LR=1e-5, Train_Loss=3.59]\n100%|██████████| 278/278 [00:06<00:00, 46.17it/s, Corr=0.881, Epoch=7, LR=1e-5, Valid_Loss=2.88]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8891454338523617 ---> 2.876198659287363)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 129.08it/s, Corr=0.845, Epoch=8, LR=1e-5, Train_Loss=3.57]\n100%|██████████| 278/278 [00:06<00:00, 43.58it/s, Corr=0.881, Epoch=8, LR=1e-5, Valid_Loss=2.87]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.876198659287363 ---> 2.8659036220372776)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 135.09it/s, Corr=0.845, Epoch=9, LR=1e-5, Train_Loss=3.58]\n100%|██████████| 278/278 [00:05<00:00, 47.03it/s, Corr=0.881, Epoch=9, LR=1e-5, Valid_Loss=2.85]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8659036220372776 ---> 2.8517033933431133)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 129.06it/s, Corr=0.845, Epoch=10, LR=1e-5, Train_Loss=3.56]\n100%|██████████| 278/278 [00:05<00:00, 46.57it/s, Corr=0.882, Epoch=10, LR=1e-5, Valid_Loss=2.85]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8517033933431133 ---> 2.8514139427093275)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 130.11it/s, Corr=0.846, Epoch=11, LR=1e-5, Train_Loss=3.55]\n100%|██████████| 278/278 [00:05<00:00, 47.98it/s, Corr=0.882, Epoch=11, LR=1e-5, Valid_Loss=2.84]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8514139427093275 ---> 2.8410741686787397)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 130.98it/s, Corr=0.846, Epoch=12, LR=1e-5, Train_Loss=3.55]\n100%|██████████| 278/278 [00:06<00:00, 45.99it/s, Corr=0.882, Epoch=12, LR=1e-5, Valid_Loss=2.83]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8410741686787397 ---> 2.830136027948456)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 131.21it/s, Corr=0.846, Epoch=13, LR=1e-5, Train_Loss=3.54]\n100%|██████████| 278/278 [00:05<00:00, 46.98it/s, Corr=0.882, Epoch=13, LR=1e-5, Valid_Loss=2.83]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.830136027948456 ---> 2.830004717831196)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 126.53it/s, Corr=0.847, Epoch=14, LR=1e-5, Train_Loss=3.52]\n100%|██████████| 278/278 [00:06<00:00, 45.59it/s, Corr=0.883, Epoch=14, LR=1e-5, Valid_Loss=2.82]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.830004717831196 ---> 2.8182893142355203)\nBest Loss: 2.8183\n\nfold = 2\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 126.86it/s, Corr=0.846, Epoch=0, LR=1e-5, Train_Loss=3.54]\n100%|██████████| 278/278 [00:06<00:00, 45.37it/s, Corr=0.883, Epoch=0, LR=1e-5, Valid_Loss=2.81]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (inf ---> 2.8141085621658672)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:16<00:00, 130.71it/s, Corr=0.847, Epoch=1, LR=1e-5, Train_Loss=3.52]\n100%|██████████| 278/278 [00:06<00:00, 41.56it/s, Corr=0.883, Epoch=1, LR=1e-5, Valid_Loss=2.8] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8141085621658672 ---> 2.80401967127585)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 128.66it/s, Corr=0.847, Epoch=2, LR=1e-5, Train_Loss=3.52]\n100%|██████████| 278/278 [00:05<00:00, 48.17it/s, Corr=0.883, Epoch=2, LR=1e-5, Valid_Loss=2.8] \n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.80401967127585 ---> 2.8016615429145957)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 125.16it/s, Corr=0.847, Epoch=3, LR=1e-5, Train_Loss=3.51]\n100%|██████████| 278/278 [00:06<00:00, 45.84it/s, Corr=0.884, Epoch=3, LR=1e-5, Valid_Loss=2.79]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.8016615429145957 ---> 2.7914881018804496)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 125.26it/s, Corr=0.847, Epoch=4, LR=1e-5, Train_Loss=3.5] \n100%|██████████| 278/278 [00:05<00:00, 46.54it/s, Corr=0.884, Epoch=4, LR=1e-5, Valid_Loss=2.79]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.7914881018804496 ---> 2.7878379646325344)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 125.91it/s, Corr=0.848, Epoch=5, LR=1e-5, Train_Loss=3.5] \n100%|██████████| 278/278 [00:06<00:00, 42.62it/s, Corr=0.884, Epoch=5, LR=1e-5, Valid_Loss=2.78]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.7878379646325344 ---> 2.781149545360459)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 129.61it/s, Corr=0.848, Epoch=6, LR=1e-5, Train_Loss=3.49]\n100%|██████████| 278/278 [00:06<00:00, 43.33it/s, Corr=0.884, Epoch=6, LR=1e-5, Valid_Loss=2.78]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.781149545360459 ---> 2.7785036006215367)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 125.15it/s, Corr=0.849, Epoch=7, LR=1e-5, Train_Loss=3.47]\n100%|██████████| 278/278 [00:05<00:00, 46.85it/s, Corr=0.884, Epoch=7, LR=1e-5, Valid_Loss=2.76]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.7785036006215367 ---> 2.7646475696171438)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:18<00:00, 123.09it/s, Corr=0.849, Epoch=8, LR=1e-5, Train_Loss=3.48]\n100%|██████████| 278/278 [00:06<00:00, 45.13it/s, Corr=0.885, Epoch=8, LR=1e-5, Valid_Loss=2.76]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.7646475696171438 ---> 2.762658892171164)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 128.09it/s, Corr=0.849, Epoch=9, LR=1e-5, Train_Loss=3.47]\n100%|██████████| 278/278 [00:06<00:00, 42.60it/s, Corr=0.885, Epoch=9, LR=1e-5, Valid_Loss=2.76]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.762658892171164 ---> 2.7563959464171357)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 129.64it/s, Corr=0.849, Epoch=10, LR=1e-5, Train_Loss=3.46]\n100%|██████████| 278/278 [00:06<00:00, 46.16it/s, Corr=0.885, Epoch=10, LR=1e-5, Valid_Loss=2.75]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.7563959464171357 ---> 2.74729930167602)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 127.96it/s, Corr=0.849, Epoch=11, LR=1e-5, Train_Loss=3.47]\n100%|██████████| 278/278 [00:06<00:00, 44.29it/s, Corr=0.885, Epoch=11, LR=1e-5, Valid_Loss=2.74]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.74729930167602 ---> 2.741973216872071)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 127.81it/s, Corr=0.85, Epoch=12, LR=1e-5, Train_Loss=3.46] \n100%|██████████| 278/278 [00:06<00:00, 43.87it/s, Corr=0.886, Epoch=12, LR=1e-5, Valid_Loss=2.74]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.741973216872071 ---> 2.7388547011051014)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 127.75it/s, Corr=0.85, Epoch=13, LR=1e-5, Train_Loss=3.43]\n100%|██████████| 278/278 [00:06<00:00, 44.34it/s, Corr=0.886, Epoch=13, LR=1e-5, Valid_Loss=2.73]\n","output_type":"stream"},{"name":"stdout","text":"Validation Loss Improved (2.7388547011051014 ---> 2.7268215952345694)\n","output_type":"stream"},{"name":"stderr","text":"100%|██████████| 2219/2219 [00:17<00:00, 129.17it/s, Corr=0.851, Epoch=14, LR=1e-5, Train_Loss=3.43]\n100%|██████████| 278/278 [00:06<00:00, 45.89it/s, Corr=0.886, Epoch=14, LR=1e-5, Valid_Loss=2.73]\n","output_type":"stream"},{"name":"stdout","text":"Best Loss: 2.7268\n","output_type":"stream"}]},{"cell_type":"code","source":"cite_test_x = torch.Tensor(x_test)\ny_pred = model(cite_test_x.cuda())\ny_pred_num = y_pred.cpu().detach().numpy()\ny_pred_num","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:35:47.344652Z","iopub.execute_input":"2022-12-18T08:35:47.345054Z","iopub.status.idle":"2022-12-18T08:35:47.364193Z","shell.execute_reply.started":"2022-12-18T08:35:47.344994Z","shell.execute_reply":"2022-12-18T08:35:47.363050Z"},"trusted":true},"execution_count":26,"outputs":[{"execution_count":26,"output_type":"execute_result","data":{"text/plain":"array([[0.2702591 , 0.30914834, 0.6942065 , ..., 0.29321927, 1.2497057 ,\n        2.5605533 ],\n       [0.25708705, 0.31852576, 0.65744984, ..., 0.34980628, 1.0983958 ,\n        2.0967472 ],\n       [0.25524384, 0.43038815, 0.7939347 , ..., 0.44670647, 2.213581  ,\n        2.523274  ],\n       ...,\n       [0.24247661, 0.43957323, 0.90434927, ..., 0.432444  , 6.7587986 ,\n        3.0721874 ],\n       [0.11118694, 0.26383644, 0.86775   , ..., 0.33342928, 2.7379787 ,\n        3.0038    ],\n       [1.5717447 , 0.5863046 , 1.2715551 , ..., 2.1690924 , 3.688074  ,\n        4.036354  ]], dtype=float32)"},"metadata":{}}]},{"cell_type":"code","source":"y_test","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:35:48.199039Z","iopub.execute_input":"2022-12-18T08:35:48.199425Z","iopub.status.idle":"2022-12-18T08:35:48.208300Z","shell.execute_reply.started":"2022-12-18T08:35:48.199395Z","shell.execute_reply":"2022-12-18T08:35:48.206969Z"},"trusted":true},"execution_count":27,"outputs":[{"execution_count":27,"output_type":"execute_result","data":{"text/plain":"array([[-0.48483855, -0.19155058,  0.19444361, ...,  1.5171883 ,\n         0.79830027,  2.1234303 ],\n       [-0.07841358,  0.61742836,  0.5715745 , ...,  1.4949981 ,\n         2.3034525 ,  4.738177  ],\n       [ 0.643268  , -0.03656904,  0.07300317, ...,  0.8929054 ,\n         1.0072782 ,  4.1010137 ],\n       ...,\n       [ 1.909388  ,  1.3529774 ,  0.22883272, ..., -0.0540573 ,\n         8.120423  ,  2.192377  ],\n       [-0.24512935,  1.35951   , -0.6800245 , ...,  1.2471745 ,\n         4.5448756 ,  0.36025345],\n       [-1.6982182 , -1.5857863 ,  0.31762958, ..., -1.2038312 ,\n         6.146836  ,  2.386191  ]], dtype=float32)"},"metadata":{}}]},{"cell_type":"code","source":"np.save('/kaggle/working/y_predict.npy', y_pred_num)","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:52:47.792153Z","iopub.execute_input":"2022-12-18T08:52:47.792505Z","iopub.status.idle":"2022-12-18T08:52:47.808240Z","shell.execute_reply.started":"2022-12-18T08:52:47.792473Z","shell.execute_reply":"2022-12-18T08:52:47.807380Z"},"trusted":true},"execution_count":35,"outputs":[]},{"cell_type":"code","source":"from sklearn.metrics import r2_score\nr2 = r2_score(y_test, y_pred_num)\nr2","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:35:51.459880Z","iopub.execute_input":"2022-12-18T08:35:51.460633Z","iopub.status.idle":"2022-12-18T08:35:51.487532Z","shell.execute_reply.started":"2022-12-18T08:35:51.460594Z","shell.execute_reply":"2022-12-18T08:35:51.486288Z"},"trusted":true},"execution_count":28,"outputs":[{"execution_count":28,"output_type":"execute_result","data":{"text/plain":"0.1924584074151504"},"metadata":{}}]},{"cell_type":"code","source":"MSE = np.square(np.subtract(y_pred_num,y_test)).mean()\nMSE","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:36:41.363108Z","iopub.execute_input":"2022-12-18T08:36:41.363474Z","iopub.status.idle":"2022-12-18T08:36:41.377362Z","shell.execute_reply.started":"2022-12-18T08:36:41.363442Z","shell.execute_reply":"2022-12-18T08:36:41.375977Z"},"trusted":true},"execution_count":29,"outputs":[{"execution_count":29,"output_type":"execute_result","data":{"text/plain":"2.8004546"},"metadata":{}}]},{"cell_type":"code","source":"import math\nprint(math.sqrt(MSE))","metadata":{"execution":{"iopub.status.busy":"2022-12-18T08:36:47.137606Z","iopub.execute_input":"2022-12-18T08:36:47.137964Z","iopub.status.idle":"2022-12-18T08:36:47.143577Z","shell.execute_reply.started":"2022-12-18T08:36:47.137933Z","shell.execute_reply":"2022-12-18T08:36:47.142491Z"},"trusted":true},"execution_count":30,"outputs":[{"name":"stdout","text":"1.6734558902303434\n","output_type":"stream"}]},{"cell_type":"code","source":"np.std(y_test)","metadata":{"execution":{"iopub.status.busy":"2022-12-17T11:44:44.504739Z","iopub.execute_input":"2022-12-17T11:44:44.505335Z","iopub.status.idle":"2022-12-17T11:44:44.523333Z","shell.execute_reply.started":"2022-12-17T11:44:44.505291Z","shell.execute_reply":"2022-12-17T11:44:44.522106Z"},"trusted":true},"execution_count":53,"outputs":[{"execution_count":53,"output_type":"execute_result","data":{"text/plain":"3.440138"},"metadata":{}}]},{"cell_type":"code","source":"np.mean(y_test)","metadata":{"execution":{"iopub.status.busy":"2022-12-17T11:45:12.909997Z","iopub.execute_input":"2022-12-17T11:45:12.910723Z","iopub.status.idle":"2022-12-17T11:45:12.922682Z","shell.execute_reply.started":"2022-12-17T11:45:12.910681Z","shell.execute_reply":"2022-12-17T11:45:12.921699Z"},"trusted":true},"execution_count":54,"outputs":[{"execution_count":54,"output_type":"execute_result","data":{"text/plain":"1.8930663"},"metadata":{}}]},{"cell_type":"code","source":"from sklearn.metrics import precision_score, accuracy_score\nprint(accuracy_score(y_test, y_pred_num))\n# print(precision_score(y_test, y_pred_num))\n","metadata":{"execution":{"iopub.status.busy":"2022-12-17T11:50:19.848656Z","iopub.execute_input":"2022-12-17T11:50:19.849101Z","iopub.status.idle":"2022-12-17T11:50:20.397495Z","shell.execute_reply.started":"2022-12-17T11:50:19.849062Z","shell.execute_reply":"2022-12-17T11:50:20.395709Z"},"trusted":true},"execution_count":60,"outputs":[{"traceback":["\u001b[0;31m---------------------------------------------------------------------------\u001b[0m","\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)","\u001b[0;32m/tmp/ipykernel_23/1755366801.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0msklearn\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmetrics\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mprecision_score\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maccuracy_score\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0maccuracy_score\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_test\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_pred_num\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      3\u001b[0m \u001b[0;31m# print(precision_score(y_test, y_pred_num))\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/sklearn/metrics/_classification.py\u001b[0m in \u001b[0;36maccuracy_score\u001b[0;34m(y_true, y_pred, normalize, sample_weight)\u001b[0m\n\u001b[1;32m    209\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    210\u001b[0m     \u001b[0;31m# Compute accuracy for each possible representation\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 211\u001b[0;31m     \u001b[0my_type\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_pred\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_check_targets\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_pred\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    212\u001b[0m     \u001b[0mcheck_consistent_length\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_pred\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msample_weight\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    213\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0my_type\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstartswith\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"multilabel\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/sklearn/metrics/_classification.py\u001b[0m in \u001b[0;36m_check_targets\u001b[0;34m(y_true, y_pred)\u001b[0m\n\u001b[1;32m    102\u001b[0m     \u001b[0;31m# No metrics support \"multiclass-multioutput\" format\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    103\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0my_type\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32min\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m\"binary\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"multiclass\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"multilabel-indicator\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 104\u001b[0;31m         \u001b[0;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"{0} is not supported\"\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_type\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    105\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    106\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0my_type\u001b[0m \u001b[0;32min\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m\"binary\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"multiclass\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;31mValueError\u001b[0m: continuous-multioutput is not supported"],"ename":"ValueError","evalue":"continuous-multioutput is not supported","output_type":"error"}]},{"cell_type":"code","source":"from sklearn.metrics import confusion_matrix\nprint(confusion_matrix(y_test, y_pred_num))\n","metadata":{"execution":{"iopub.status.busy":"2022-12-17T11:52:47.717117Z","iopub.execute_input":"2022-12-17T11:52:47.717532Z","iopub.status.idle":"2022-12-17T11:52:48.231382Z","shell.execute_reply.started":"2022-12-17T11:52:47.717499Z","shell.execute_reply":"2022-12-17T11:52:48.229821Z"},"trusted":true},"execution_count":62,"outputs":[{"traceback":["\u001b[0;31m---------------------------------------------------------------------------\u001b[0m","\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)","\u001b[0;32m/tmp/ipykernel_23/1877939405.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0msklearn\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmetrics\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mconfusion_matrix\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mconfusion_matrix\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_test\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_pred_num\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/sklearn/metrics/_classification.py\u001b[0m in \u001b[0;36mconfusion_matrix\u001b[0;34m(y_true, y_pred, labels, sample_weight, normalize)\u001b[0m\n\u001b[1;32m    305\u001b[0m     \u001b[0;34m(\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    306\u001b[0m     \"\"\"\n\u001b[0;32m--> 307\u001b[0;31m     \u001b[0my_type\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_pred\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_check_targets\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_pred\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    308\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0my_type\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32min\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;34m\"binary\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"multiclass\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    309\u001b[0m         \u001b[0;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"%s is not supported\"\u001b[0m \u001b[0;34m%\u001b[0m \u001b[0my_type\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/sklearn/metrics/_classification.py\u001b[0m in \u001b[0;36m_check_targets\u001b[0;34m(y_true, y_pred)\u001b[0m\n\u001b[1;32m    102\u001b[0m     \u001b[0;31m# No metrics support \"multiclass-multioutput\" format\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    103\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0my_type\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32min\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m\"binary\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"multiclass\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"multilabel-indicator\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 104\u001b[0;31m         \u001b[0;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"{0} is not supported\"\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_type\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    105\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    106\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0my_type\u001b[0m \u001b[0;32min\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m\"binary\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"multiclass\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;31mValueError\u001b[0m: continuous-multioutput is not supported"],"ename":"ValueError","evalue":"continuous-multioutput is not supported","output_type":"error"}]},{"cell_type":"code","source":"train_cite_test_x = torch.Tensor(x_train)\ny_pred_train = model(train_cite_test_x.cuda())\ny_pred_train_num = y_pred_train.cpu().detach().numpy()\ny_pred_train_num","metadata":{"execution":{"iopub.status.busy":"2022-12-17T13:29:02.583674Z","iopub.execute_input":"2022-12-17T13:29:02.584062Z","iopub.status.idle":"2022-12-17T13:29:02.603732Z","shell.execute_reply.started":"2022-12-17T13:29:02.584028Z","shell.execute_reply":"2022-12-17T13:29:02.601568Z"},"trusted":true},"execution_count":11,"outputs":[{"traceback":["\u001b[0;31m---------------------------------------------------------------------------\u001b[0m","\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)","\u001b[0;32m/tmp/ipykernel_23/1296686739.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0mtrain_cite_test_x\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mTensor\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx_train\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m \u001b[0my_pred_train\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtrain_cite_test_x\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcuda\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      3\u001b[0m \u001b[0my_pred_train_num\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0my_pred_train\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcpu\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdetach\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnumpy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0my_pred_train_num\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;31mNameError\u001b[0m: name 'model' is not defined"],"ename":"NameError","evalue":"name 'model' is not defined","output_type":"error"}]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}