Created
April 4, 2023 06:19
-
-
Save snakers4/0d01b31e26f977a8545fe27e7ad31a1c to your computer and use it in GitHub Desktop.
temp
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| {"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)\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\nimport torch\nfrom torch import nn\nfrom torch.utils.data import DataLoader\nimport torch.optim as optim","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2023-04-02T18:30:02.008625Z","iopub.execute_input":"2023-04-02T18:30:02.009328Z","iopub.status.idle":"2023-04-02T18:30:04.758970Z","shell.execute_reply.started":"2023-04-02T18:30:02.009298Z","shell.execute_reply":"2023-04-02T18:30:04.757803Z"},"trusted":true},"execution_count":1,"outputs":[{"name":"stdout","text":"/kaggle/input/silero-stress-predictor/sample_submission.csv\n/kaggle/input/silero-stress-predictor/train.csv\n/kaggle/input/silero-stress-predictor/test.csv\n","output_type":"stream"}]},{"cell_type":"code","source":"torch.cuda.is_available()","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:04.761262Z","iopub.execute_input":"2023-04-02T18:30:04.761794Z","iopub.status.idle":"2023-04-02T18:30:04.843866Z","shell.execute_reply.started":"2023-04-02T18:30:04.761754Z","shell.execute_reply":"2023-04-02T18:30:04.842113Z"},"trusted":true},"execution_count":2,"outputs":[{"execution_count":2,"output_type":"execute_result","data":{"text/plain":"True"},"metadata":{}}]},{"cell_type":"code","source":"device = 'cuda' if torch.cuda.is_available() else 'cpu'","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:04.845010Z","iopub.execute_input":"2023-04-02T18:30:04.845401Z","iopub.status.idle":"2023-04-02T18:30:04.852435Z","shell.execute_reply.started":"2023-04-02T18:30:04.845363Z","shell.execute_reply":"2023-04-02T18:30:04.851256Z"},"trusted":true},"execution_count":3,"outputs":[]},{"cell_type":"markdown","source":"# Предобработка","metadata":{}},{"cell_type":"code","source":"train = pd.read_csv(\"/kaggle/input/silero-stress-predictor/train.csv\")","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:04.855676Z","iopub.execute_input":"2023-04-02T18:30:04.856258Z","iopub.status.idle":"2023-04-02T18:30:05.033608Z","shell.execute_reply.started":"2023-04-02T18:30:04.856218Z","shell.execute_reply":"2023-04-02T18:30:05.032496Z"},"trusted":true},"execution_count":4,"outputs":[]},{"cell_type":"code","source":"train","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:05.038925Z","iopub.execute_input":"2023-04-02T18:30:05.041482Z","iopub.status.idle":"2023-04-02T18:30:05.072535Z","shell.execute_reply.started":"2023-04-02T18:30:05.041426Z","shell.execute_reply":"2023-04-02T18:30:05.071457Z"},"trusted":true},"execution_count":5,"outputs":[{"execution_count":5,"output_type":"execute_result","data":{"text/plain":" id word stress num_syllables lemma\n0 0 румяной 2 3 румяный\n1 1 цифрами 1 3 цифра\n2 2 слугами 1 3 слуга\n3 3 выбирает 3 4 выбирать\n4 4 управдом 3 3 управдом\n... ... ... ... ... ...\n63433 63433 экзамена 2 4 экзамен\n63434 63434 культурой 2 3 культура\n63435 63435 объемной 2 3 объемный\n63436 63436 участком 2 3 участок\n63437 63437 ташкента 2 3 ташкент\n\n[63438 rows x 5 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></th>\n <th>id</th>\n <th>word</th>\n <th>stress</th>\n <th>num_syllables</th>\n <th>lemma</th>\n </tr>\n </thead>\n <tbody>\n <tr>\n <th>0</th>\n <td>0</td>\n <td>румяной</td>\n <td>2</td>\n <td>3</td>\n <td>румяный</td>\n </tr>\n <tr>\n <th>1</th>\n <td>1</td>\n <td>цифрами</td>\n <td>1</td>\n <td>3</td>\n <td>цифра</td>\n </tr>\n <tr>\n <th>2</th>\n <td>2</td>\n <td>слугами</td>\n <td>1</td>\n <td>3</td>\n <td>слуга</td>\n </tr>\n <tr>\n <th>3</th>\n <td>3</td>\n <td>выбирает</td>\n <td>3</td>\n <td>4</td>\n <td>выбирать</td>\n </tr>\n <tr>\n <th>4</th>\n <td>4</td>\n <td>управдом</td>\n <td>3</td>\n <td>3</td>\n <td>управдом</td>\n </tr>\n <tr>\n <th>...</th>\n <td>...</td>\n <td>...</td>\n <td>...</td>\n <td>...</td>\n <td>...</td>\n </tr>\n <tr>\n <th>63433</th>\n <td>63433</td>\n <td>экзамена</td>\n <td>2</td>\n <td>4</td>\n <td>экзамен</td>\n </tr>\n <tr>\n <th>63434</th>\n <td>63434</td>\n <td>культурой</td>\n <td>2</td>\n <td>3</td>\n <td>культура</td>\n </tr>\n <tr>\n <th>63435</th>\n <td>63435</td>\n <td>объемной</td>\n <td>2</td>\n <td>3</td>\n <td>объемный</td>\n </tr>\n <tr>\n <th>63436</th>\n <td>63436</td>\n <td>участком</td>\n <td>2</td>\n <td>3</td>\n <td>участок</td>\n </tr>\n <tr>\n <th>63437</th>\n <td>63437</td>\n <td>ташкента</td>\n <td>2</td>\n <td>3</td>\n <td>ташкент</td>\n </tr>\n </tbody>\n</table>\n<p>63438 rows × 5 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"def simple_splitter(word):\n word = word.lower()\n vowels = ['а', 'у', 'о', 'ы', 'э', 'я', 'ю', 'ё', 'и', 'е']\n for s in word:\n if s in vowels:\n word = word.replace(s, s + '|')\n vowels.remove(s)\n syl = word.split('|')\n if syl[-1] not in vowels:\n syl[-2] = syl[-2] + syl[-1]\n syl.pop(-1)\n return syl","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:05.077051Z","iopub.execute_input":"2023-04-02T18:30:05.079617Z","iopub.status.idle":"2023-04-02T18:30:05.086623Z","shell.execute_reply.started":"2023-04-02T18:30:05.079572Z","shell.execute_reply":"2023-04-02T18:30:05.085708Z"},"trusted":true},"execution_count":6,"outputs":[]},{"cell_type":"code","source":"# all_words = [['UNK'], ['PAD']]\n# for i in train['word']:\n# all_words.append(simple_splitter(i))","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:05.087970Z","iopub.execute_input":"2023-04-02T18:30:05.089438Z","iopub.status.idle":"2023-04-02T18:30:05.096609Z","shell.execute_reply.started":"2023-04-02T18:30:05.089399Z","shell.execute_reply":"2023-04-02T18:30:05.095210Z"},"trusted":true},"execution_count":7,"outputs":[]},{"cell_type":"code","source":"# from gensim.models import Word2Vec\n\n# W2Vmodel = Word2Vec(sentences=all_words, vector_size=30, window=2, min_count=1, workers=4)","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:05.098659Z","iopub.execute_input":"2023-04-02T18:30:05.099430Z","iopub.status.idle":"2023-04-02T18:30:05.108085Z","shell.execute_reply.started":"2023-04-02T18:30:05.099393Z","shell.execute_reply":"2023-04-02T18:30:05.107140Z"},"trusted":true},"execution_count":8,"outputs":[]},{"cell_type":"code","source":"all_sylls = [['UNK'], ['PAD']]\nfor i in train['word']:\n all_sylls.append(simple_splitter(i))\n \nnp_sylls = np.concatenate(all_sylls)\n\nfull_dict_sylls = dict(zip(set(np_sylls), range(0, len(set(np_sylls))))) ## Словарь всех слогов вида слог:индекс\n\nlen(full_dict_sylls)","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:05.109979Z","iopub.execute_input":"2023-04-02T18:30:05.110746Z","iopub.status.idle":"2023-04-02T18:30:06.040306Z","shell.execute_reply.started":"2023-04-02T18:30:05.110710Z","shell.execute_reply":"2023-04-02T18:30:06.039203Z"},"trusted":true},"execution_count":9,"outputs":[{"execution_count":9,"output_type":"execute_result","data":{"text/plain":"7996"},"metadata":{}}]},{"cell_type":"code","source":"def word2ind(word):\n inds = []\n sylls = simple_splitter(word)\n for s in sylls:\n try:\n inds.append(full_dict_sylls[s])\n except KeyError:\n inds.append(full_dict_sylls['UNK'])\n while len(inds) < 6:\n inds.append(full_dict_sylls['PAD'])\n \n if len(inds) > 6: inds = inds[:6]\n \n return inds","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:06.044800Z","iopub.execute_input":"2023-04-02T18:30:06.045590Z","iopub.status.idle":"2023-04-02T18:30:06.052943Z","shell.execute_reply.started":"2023-04-02T18:30:06.045553Z","shell.execute_reply":"2023-04-02T18:30:06.051636Z"},"trusted":true},"execution_count":10,"outputs":[]},{"cell_type":"code","source":"def word2indW2V(word):\n inds = []\n sylls = simple_splitter(word)\n for s in sylls:\n try:\n inds.append(W2Vmodel.wv[s])\n except KeyError:\n inds.append(W2Vmodel.wv['UNK'])\n while len(inds) < 6:\n inds.append(W2Vmodel.wv['PAD'])\n \n return torch.tensor(inds)","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:06.054953Z","iopub.execute_input":"2023-04-02T18:30:06.055892Z","iopub.status.idle":"2023-04-02T18:30:06.069930Z","shell.execute_reply.started":"2023-04-02T18:30:06.055782Z","shell.execute_reply":"2023-04-02T18:30:06.065950Z"},"trusted":true},"execution_count":11,"outputs":[]},{"cell_type":"code","source":"# tenz = torch.zeros((words.shape[0], words.shape[1], 30), dtype=torch.float32)\n# tenz.shape","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:06.073531Z","iopub.execute_input":"2023-04-02T18:30:06.074242Z","iopub.status.idle":"2023-04-02T18:30:06.079871Z","shell.execute_reply.started":"2023-04-02T18:30:06.074204Z","shell.execute_reply":"2023-04-02T18:30:06.078705Z"},"trusted":true},"execution_count":12,"outputs":[]},{"cell_type":"code","source":"# for i, word in enumerate(train['word']):\n# vect = word2indW2V(word)\n# # vect = torch.unsqueeze(vect, dim=0)\n# tenz[i, :] = vect","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:06.082866Z","iopub.execute_input":"2023-04-02T18:30:06.084631Z","iopub.status.idle":"2023-04-02T18:30:06.092892Z","shell.execute_reply.started":"2023-04-02T18:30:06.084594Z","shell.execute_reply":"2023-04-02T18:30:06.089714Z"},"trusted":true},"execution_count":13,"outputs":[]},{"cell_type":"code","source":"trainRxWord = train[['word']].applymap(word2ind)\nwords = pd.DataFrame(trainRxWord['word'].to_list(), columns = ['w1', 'w2', 'w3', 'w4', 'w5', 'w6'])\nwordsNN = torch.tensor(words.values.astype(np.int32))\nwordsNN","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:06.095008Z","iopub.execute_input":"2023-04-02T18:30:06.095570Z","iopub.status.idle":"2023-04-02T18:30:07.035041Z","shell.execute_reply.started":"2023-04-02T18:30:06.095533Z","shell.execute_reply":"2023-04-02T18:30:07.034023Z"},"trusted":true},"execution_count":14,"outputs":[{"execution_count":14,"output_type":"execute_result","data":{"text/plain":"tensor([[5473, 5755, 7914, 6397, 6397, 6397],\n [6934, 1393, 7036, 6397, 6397, 6397],\n [6657, 949, 7036, 6397, 6397, 6397],\n ...,\n [7904, 1615, 2184, 6397, 6397, 6397],\n [ 388, 1260, 266, 6397, 6397, 6397],\n [4196, 3011, 1923, 6397, 6397, 6397]], dtype=torch.int32)"},"metadata":{}}]},{"cell_type":"code","source":"trainRxLemm = train[['lemma']].applymap(word2ind)\nlemms = pd.DataFrame(trainRxLemm['lemma'].to_list(), columns = ['l1', 'l2', 'l3', 'l4', 'l5', 'l6'])\nlemmsNN = torch.tensor(words.values.astype(np.int32))\nlemmsNN","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:07.036436Z","iopub.execute_input":"2023-04-02T18:30:07.036809Z","iopub.status.idle":"2023-04-02T18:30:07.490260Z","shell.execute_reply.started":"2023-04-02T18:30:07.036771Z","shell.execute_reply":"2023-04-02T18:30:07.489263Z"},"trusted":true},"execution_count":15,"outputs":[{"execution_count":15,"output_type":"execute_result","data":{"text/plain":"tensor([[5473, 5755, 7914, 6397, 6397, 6397],\n [6934, 1393, 7036, 6397, 6397, 6397],\n [6657, 949, 7036, 6397, 6397, 6397],\n ...,\n [7904, 1615, 2184, 6397, 6397, 6397],\n [ 388, 1260, 266, 6397, 6397, 6397],\n [4196, 3011, 1923, 6397, 6397, 6397]], dtype=torch.int32)"},"metadata":{}}]},{"cell_type":"code","source":"# trX_NN = torch.tensor(pd.concat([words, lemms , train['num_syllables']], axis=1).values.astype(np.float32))\n# trX_NN.shape","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:07.493480Z","iopub.execute_input":"2023-04-02T18:30:07.493771Z","iopub.status.idle":"2023-04-02T18:30:07.500231Z","shell.execute_reply.started":"2023-04-02T18:30:07.493744Z","shell.execute_reply":"2023-04-02T18:30:07.499140Z"},"trusted":true},"execution_count":16,"outputs":[]},{"cell_type":"code","source":"trainRy = train['stress']\ntrainRy","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:07.501838Z","iopub.execute_input":"2023-04-02T18:30:07.502273Z","iopub.status.idle":"2023-04-02T18:30:07.513594Z","shell.execute_reply.started":"2023-04-02T18:30:07.502236Z","shell.execute_reply":"2023-04-02T18:30:07.512417Z"},"trusted":true},"execution_count":17,"outputs":[{"execution_count":17,"output_type":"execute_result","data":{"text/plain":"0 2\n1 1\n2 1\n3 3\n4 3\n ..\n63433 2\n63434 2\n63435 2\n63436 2\n63437 2\nName: stress, Length: 63438, dtype: int64"},"metadata":{}}]},{"cell_type":"markdown","source":"# ML","metadata":{}},{"cell_type":"markdown","source":"## Нейросеть","metadata":{}},{"cell_type":"code","source":"import torch\nfrom torch import nn\nfrom torch.utils.data import DataLoader\nimport torch.optim as optim","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:30:07.515073Z","iopub.execute_input":"2023-04-02T18:30:07.515554Z","iopub.status.idle":"2023-04-02T18:30:07.523761Z","shell.execute_reply.started":"2023-04-02T18:30:07.515518Z","shell.execute_reply":"2023-04-02T18:30:07.522783Z"},"trusted":true},"execution_count":18,"outputs":[]},{"cell_type":"code","source":"class SimpleNet(nn.Module):\n def __init__(self, embD):\n super().__init__()\n \n self.embD = embD\n \n self.emb = nn.Embedding(num_embeddings=7996, embedding_dim=self.embD)\n \n self.enc = nn.Sequential(\n nn.Conv1d(in_channels=self.embD, out_channels=self.embD + 5, kernel_size=2, padding='same'),\n nn.ReLU(),\n nn.Dropout1d(p=0.5),\n \n nn.Conv1d(in_channels=self.embD + 5, out_channels=self.embD + 10, kernel_size=3, padding='same'),\n nn.ReLU(),\n nn.Dropout1d(p=0.5),\n \n nn.Conv1d(in_channels=self.embD + 10, out_channels=self.embD + 15, kernel_size=4, padding='same'),\n nn.ReLU(),\n nn.Dropout1d(p=0.5),\n \n nn.Conv1d(in_channels=self.embD + 15, out_channels=self.embD + 20, kernel_size=5, padding='same'),\n nn.ReLU(),\n nn.Dropout1d(p=0.5),\n )\n\n self.lin = nn.Linear(in_features=self.embD + 20, out_features=1)\n\n\n def forward(self, x):\n\n x = self.emb(x)\n \n x = torch.permute(x, (0, 2, 1))\n \n x = self.enc(x)\n \n x = torch.permute(x, (0, 2, 1))\n \n x = self.lin(x)\n \n res = torch.squeeze(x, dim=2)\n\n return res","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:26:36.398218Z","iopub.execute_input":"2023-04-02T20:26:36.398881Z","iopub.status.idle":"2023-04-02T20:26:36.412722Z","shell.execute_reply.started":"2023-04-02T20:26:36.398844Z","shell.execute_reply":"2023-04-02T20:26:36.410790Z"},"trusted":true},"execution_count":195,"outputs":[]},{"cell_type":"code","source":"testMod = SimpleNet(embD=30).to(device)","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:26:48.401480Z","iopub.execute_input":"2023-04-02T20:26:48.401839Z","iopub.status.idle":"2023-04-02T20:26:48.412644Z","shell.execute_reply.started":"2023-04-02T20:26:48.401805Z","shell.execute_reply":"2023-04-02T20:26:48.411614Z"},"trusted":true},"execution_count":197,"outputs":[]},{"cell_type":"code","source":"# trainSetLOL = tenz[:5]\n# trainSetLOL.shape","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:40:08.511637Z","iopub.execute_input":"2023-04-02T18:40:08.514023Z","iopub.status.idle":"2023-04-02T18:40:08.520010Z","shell.execute_reply.started":"2023-04-02T18:40:08.513981Z","shell.execute_reply":"2023-04-02T18:40:08.518863Z"},"trusted":true},"execution_count":49,"outputs":[]},{"cell_type":"code","source":"# trainSetLOL = trX_NN[:5]\n# # trainSetLOL = torch.tensor(trainSetLOL.values.astype(np.int32))\n# trainSetLOL","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:40:08.708400Z","iopub.execute_input":"2023-04-02T18:40:08.709075Z","iopub.status.idle":"2023-04-02T18:40:08.714075Z","shell.execute_reply.started":"2023-04-02T18:40:08.709041Z","shell.execute_reply":"2023-04-02T18:40:08.712799Z"},"trusted":true},"execution_count":50,"outputs":[]},{"cell_type":"code","source":"trainSetLOLw = wordsNN[:5].to(device)\ntrainSetLOLl = lemmsNN[:5].to(device)","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:40:08.901320Z","iopub.execute_input":"2023-04-02T18:40:08.901634Z","iopub.status.idle":"2023-04-02T18:40:08.907528Z","shell.execute_reply.started":"2023-04-02T18:40:08.901604Z","shell.execute_reply":"2023-04-02T18:40:08.906360Z"},"trusted":true},"execution_count":51,"outputs":[]},{"cell_type":"code","source":"testMod(trainSetLOLw)","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:26:51.004214Z","iopub.execute_input":"2023-04-02T20:26:51.005143Z","iopub.status.idle":"2023-04-02T20:26:51.017351Z","shell.execute_reply.started":"2023-04-02T20:26:51.005089Z","shell.execute_reply":"2023-04-02T20:26:51.016380Z"},"trusted":true},"execution_count":198,"outputs":[{"execution_count":198,"output_type":"execute_result","data":{"text/plain":"tensor([[-0.0495, -0.1283, -0.1177, -0.0029, -0.0055, -0.0037],\n [ 0.0411, 0.0404, 0.0161, 0.0371, 0.0468, 0.0281],\n [ 0.0249, 0.0228, 0.0543, 0.0550, 0.0291, 0.0340],\n [ 0.0382, 0.0095, -0.0020, 0.0237, -0.0323, 0.0058],\n [ 0.1027, 0.0657, 0.0143, 0.0167, -0.0228, -0.0014]],\n device='cuda:0', grad_fn=<SqueezeBackward1>)"},"metadata":{}}]},{"cell_type":"code","source":"# trainSetX = tenz[:60000]\n# trainSetX.shape","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:40:09.482148Z","iopub.execute_input":"2023-04-02T18:40:09.482624Z","iopub.status.idle":"2023-04-02T18:40:09.488489Z","shell.execute_reply.started":"2023-04-02T18:40:09.482589Z","shell.execute_reply":"2023-04-02T18:40:09.487201Z"},"trusted":true},"execution_count":53,"outputs":[]},{"cell_type":"code","source":"trainSetXw = wordsNN[:60000].to(device)\n# trainSetX = torch.tensor(trainSetX.values.astype(np.int32))\n\ntrainSetXw.shape","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:40:09.663183Z","iopub.execute_input":"2023-04-02T18:40:09.663486Z","iopub.status.idle":"2023-04-02T18:40:09.672244Z","shell.execute_reply.started":"2023-04-02T18:40:09.663457Z","shell.execute_reply":"2023-04-02T18:40:09.671160Z"},"trusted":true},"execution_count":54,"outputs":[{"execution_count":54,"output_type":"execute_result","data":{"text/plain":"torch.Size([60000, 6])"},"metadata":{}}]},{"cell_type":"code","source":"trainSetXl = lemmsNN[:60000].to(device)\n\ntrainSetXl.shape","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:40:09.919704Z","iopub.execute_input":"2023-04-02T18:40:09.920594Z","iopub.status.idle":"2023-04-02T18:40:09.930484Z","shell.execute_reply.started":"2023-04-02T18:40:09.920543Z","shell.execute_reply":"2023-04-02T18:40:09.929353Z"},"trusted":true},"execution_count":55,"outputs":[{"execution_count":55,"output_type":"execute_result","data":{"text/plain":"torch.Size([60000, 6])"},"metadata":{}}]},{"cell_type":"code","source":"trainSetY = train['stress'][:60000] - 1\ntrainSetY = torch.tensor(trainSetY.values.astype(np.float32)).to(device)\n# trainSetY = nn.functional.one_hot(trainSetY.long(), num_classes=6)\n\ntrainSetY.shape","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:40:10.312768Z","iopub.execute_input":"2023-04-02T18:40:10.313332Z","iopub.status.idle":"2023-04-02T18:40:10.323720Z","shell.execute_reply.started":"2023-04-02T18:40:10.313297Z","shell.execute_reply":"2023-04-02T18:40:10.322326Z"},"trusted":true},"execution_count":56,"outputs":[{"execution_count":56,"output_type":"execute_result","data":{"text/plain":"torch.Size([60000])"},"metadata":{}}]},{"cell_type":"code","source":"# testSetX = tenz[60000:]\n\n# testSetY = train['stress'][60000:] - 1\n# testSetY = torch.tensor(testSetY.values.astype(np.float32))","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:40:10.654757Z","iopub.execute_input":"2023-04-02T18:40:10.655636Z","iopub.status.idle":"2023-04-02T18:40:10.660577Z","shell.execute_reply.started":"2023-04-02T18:40:10.655586Z","shell.execute_reply":"2023-04-02T18:40:10.659446Z"},"trusted":true},"execution_count":57,"outputs":[]},{"cell_type":"code","source":"testSetXw = wordsNN[60000:].to(device)\ntestSetXl = lemmsNN[60000:].to(device)\n# testSetX = torch.tensor(testSetX.values.astype(np.int32))\n\ntestSetY = train['stress'][60000:] - 1\ntestSetY = torch.tensor(testSetY.values.astype(np.float32)).to(device)","metadata":{"execution":{"iopub.status.busy":"2023-04-02T18:40:10.942936Z","iopub.execute_input":"2023-04-02T18:40:10.943631Z","iopub.status.idle":"2023-04-02T18:40:10.952316Z","shell.execute_reply.started":"2023-04-02T18:40:10.943593Z","shell.execute_reply":"2023-04-02T18:40:10.951271Z"},"trusted":true},"execution_count":58,"outputs":[]},{"cell_type":"code","source":"train_dataloader = DataLoader(list(zip(trainSetXw, trainSetY)), batch_size=150, shuffle=True)\ntest_dataloader = DataLoader(list(zip(testSetXw, testSetY)), batch_size=150, shuffle=True)","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:27:00.000799Z","iopub.execute_input":"2023-04-02T20:27:00.001184Z","iopub.status.idle":"2023-04-02T20:27:00.600446Z","shell.execute_reply.started":"2023-04-02T20:27:00.001143Z","shell.execute_reply":"2023-04-02T20:27:00.599433Z"},"trusted":true},"execution_count":199,"outputs":[]},{"cell_type":"code","source":"testMod = SimpleNet(embD=300)\nn_epochs = 200\nloss_fn = nn.CrossEntropyLoss()\noptimizer = optim.Adam(testMod.parameters(), weight_decay=0.00001)\nscheduler = optim.lr_scheduler.ExponentialLR(optimizer, gamma=0.999)\ntestMod.train()","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:27:04.353521Z","iopub.execute_input":"2023-04-02T20:27:04.354462Z","iopub.status.idle":"2023-04-02T20:27:04.393673Z","shell.execute_reply.started":"2023-04-02T20:27:04.354408Z","shell.execute_reply":"2023-04-02T20:27:04.392027Z"},"trusted":true},"execution_count":200,"outputs":[{"execution_count":200,"output_type":"execute_result","data":{"text/plain":"SimpleNet(\n (emb): Embedding(7996, 300)\n (enc): Sequential(\n (0): Conv1d(300, 305, kernel_size=(2,), stride=(1,), padding=same)\n (1): ReLU()\n (2): Dropout1d(p=0.5, inplace=False)\n (3): Conv1d(305, 310, kernel_size=(3,), stride=(1,), padding=same)\n (4): ReLU()\n (5): Dropout1d(p=0.5, inplace=False)\n (6): Conv1d(310, 315, kernel_size=(4,), stride=(1,), padding=same)\n (7): ReLU()\n (8): Dropout1d(p=0.5, inplace=False)\n (9): Conv1d(315, 320, kernel_size=(5,), stride=(1,), padding=same)\n (10): ReLU()\n (11): Dropout1d(p=0.5, inplace=False)\n )\n (lin): Linear(in_features=320, out_features=1, bias=True)\n (sumLin): Sequential(\n (0): Linear(in_features=300, out_features=310, bias=True)\n (1): ReLU()\n (2): Dropout(p=0.3, inplace=False)\n (3): Linear(in_features=310, out_features=300, bias=True)\n (4): ReLU()\n )\n)"},"metadata":{}}]},{"cell_type":"code","source":"testMod.to(device)\nhigh_acc = 0.\neps_no_imp = 0\nfor ep in range(n_epochs):\n print(f\"Epoch {ep + 1}\")\n last_loss = 0.\n run_loss = 0.\n \n \n for i, data in enumerate(train_dataloader):\n optimizer.zero_grad()\n \n Xb, Yb = data\n \n ypred = testMod(Xb)\n \n loss = loss_fn(ypred, Yb.long())\n \n \n loss.backward()\n optimizer.step()\n \n \n run_loss += loss.item()\n \n if i % 100 == 99:\n last_loss = run_loss / 100\n print(f\"batch {i + 1}, loss {last_loss}\")\n run_loss = 0\n \n ## test check\n \n testMod.eval()\n y_pred = testMod(testSetXw)\n acc = (y_pred.argmax(1) == testSetY).float().mean()\n acc = float(acc)\n print(\"Model accuracy: %.2f%%\" % (acc*100))\n \n if acc > high_acc:\n print(f\"New highest accuracy: {acc}\")\n high_acc = acc\n torch.save(testMod.state_dict(), \"bestModel.pt\")\n eps_no_imp = 0\n else:\n eps_no_imp += 1\n \n if eps_no_imp == 20:\n print(f\"No improvement in accuracy. Stopping early\\n Best accuracy -> {high_acc}\")\n break\n \n testMod.train()\n \n ## end of check\n \n scheduler.step()\n print(f\"End of epoch {ep + 1}\\n-------------------------\")\n## 86.41","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:27:17.385019Z","iopub.execute_input":"2023-04-02T20:27:17.385704Z","iopub.status.idle":"2023-04-02T20:31:22.733795Z","shell.execute_reply.started":"2023-04-02T20:27:17.385669Z","shell.execute_reply":"2023-04-02T20:31:22.732707Z"},"trusted":true},"execution_count":201,"outputs":[{"name":"stdout","text":"Epoch 1\nbatch 100, loss 0.9018569391965866\nbatch 200, loss 0.7573528355360031\nbatch 300, loss 0.701995764374733\nbatch 400, loss 0.6737083846330643\nModel accuracy: 71.76%\nNew highest accuracy: 0.7175683975219727\nEnd of epoch 1\n-------------------------\nEpoch 2\nbatch 100, loss 0.6252882352471352\nbatch 200, loss 0.5996284347772598\nbatch 300, loss 0.6007369866967202\nbatch 400, loss 0.5757922276854515\nModel accuracy: 77.72%\nNew highest accuracy: 0.7771960496902466\nEnd of epoch 2\n-------------------------\nEpoch 3\nbatch 100, loss 0.5255890455842018\nbatch 200, loss 0.5299744012951851\nbatch 300, loss 0.5148718211054802\nbatch 400, loss 0.5161626166105271\nModel accuracy: 79.64%\nNew highest accuracy: 0.7963932752609253\nEnd of epoch 3\n-------------------------\nEpoch 4\nbatch 100, loss 0.4652759745717049\nbatch 200, loss 0.46547737628221514\nbatch 300, loss 0.46803209722042083\nbatch 400, loss 0.4587903994321823\nModel accuracy: 81.56%\nNew highest accuracy: 0.815590500831604\nEnd of epoch 4\n-------------------------\nEpoch 5\nbatch 100, loss 0.4154364788532257\nbatch 200, loss 0.42744381457567215\nbatch 300, loss 0.430217262506485\nbatch 400, loss 0.4258602997660637\nModel accuracy: 82.98%\nNew highest accuracy: 0.8298429846763611\nEnd of epoch 5\n-------------------------\nEpoch 6\nbatch 100, loss 0.37280248835682867\nbatch 200, loss 0.3994229930639267\nbatch 300, loss 0.3890038043260574\nbatch 400, loss 0.4004691049456596\nModel accuracy: 83.94%\nNew highest accuracy: 0.8394415974617004\nEnd of epoch 6\n-------------------------\nEpoch 7\nbatch 100, loss 0.3488214275240898\nbatch 200, loss 0.3608837823569775\nbatch 300, loss 0.3567549601197243\nbatch 400, loss 0.37874603986740113\nModel accuracy: 83.97%\nNew highest accuracy: 0.8397324681282043\nEnd of epoch 7\n-------------------------\nEpoch 8\nbatch 100, loss 0.32347454741597176\nbatch 200, loss 0.34386890560388567\nbatch 300, loss 0.3557080373167992\nbatch 400, loss 0.35328751638531686\nModel accuracy: 84.21%\nNew highest accuracy: 0.8420593738555908\nEnd of epoch 8\n-------------------------\nEpoch 9\nbatch 100, loss 0.308316530585289\nbatch 200, loss 0.31053061842918395\nbatch 300, loss 0.33456177726387976\nbatch 400, loss 0.34341296076774597\nModel accuracy: 84.38%\nNew highest accuracy: 0.8438045978546143\nEnd of epoch 9\n-------------------------\nEpoch 10\nbatch 100, loss 0.29412034586071967\nbatch 200, loss 0.30991573989391324\nbatch 300, loss 0.3167182119190693\nbatch 400, loss 0.32487204879522324\nModel accuracy: 85.31%\nNew highest accuracy: 0.8531123399734497\nEnd of epoch 10\n-------------------------\nEpoch 11\nbatch 100, loss 0.2885857158899307\nbatch 200, loss 0.2882878114283085\nbatch 300, loss 0.30791873157024385\nbatch 400, loss 0.3103193962574005\nModel accuracy: 85.14%\nEnd of epoch 11\n-------------------------\nEpoch 12\nbatch 100, loss 0.2718512934446335\nbatch 200, loss 0.2921528285741806\nbatch 300, loss 0.29038614735007284\nbatch 400, loss 0.2997064162790775\nModel accuracy: 85.89%\nNew highest accuracy: 0.8589296340942383\nEnd of epoch 12\n-------------------------\nEpoch 13\nbatch 100, loss 0.2612572342157364\nbatch 200, loss 0.27173383444547655\nbatch 300, loss 0.2815230955183506\nbatch 400, loss 0.29423739537596705\nModel accuracy: 85.95%\nNew highest accuracy: 0.8595113754272461\nEnd of epoch 13\n-------------------------\nEpoch 14\nbatch 100, loss 0.24721902281045913\nbatch 200, loss 0.27125407800078394\nbatch 300, loss 0.26293970689177515\nbatch 400, loss 0.27878078982234\nModel accuracy: 85.92%\nEnd of epoch 14\n-------------------------\nEpoch 15\nbatch 100, loss 0.24093874424695969\nbatch 200, loss 0.25243014097213745\nbatch 300, loss 0.25922741830348967\nbatch 400, loss 0.27497716680169104\nModel accuracy: 86.01%\nNew highest accuracy: 0.8600931167602539\nEnd of epoch 15\n-------------------------\nEpoch 16\nbatch 100, loss 0.231727212369442\nbatch 200, loss 0.2542976912856102\nbatch 300, loss 0.25651106528937817\nbatch 400, loss 0.26031990617513656\nModel accuracy: 86.36%\nNew highest accuracy: 0.863583505153656\nEnd of epoch 16\n-------------------------\nEpoch 17\nbatch 100, loss 0.2280552652478218\nbatch 200, loss 0.24291264697909354\nbatch 300, loss 0.25091256856918337\nbatch 400, loss 0.25288270488381387\nModel accuracy: 86.15%\nEnd of epoch 17\n-------------------------\nEpoch 18\nbatch 100, loss 0.21470939636230468\nbatch 200, loss 0.2422358785569668\nbatch 300, loss 0.2407553720474243\nbatch 400, loss 0.24177830010652543\nModel accuracy: 85.92%\nEnd of epoch 18\n-------------------------\nEpoch 19\nbatch 100, loss 0.21143778443336486\nbatch 200, loss 0.21626549579203128\nbatch 300, loss 0.24130339480936527\nbatch 400, loss 0.2430872191488743\nModel accuracy: 86.18%\nEnd of epoch 19\n-------------------------\nEpoch 20\nbatch 100, loss 0.20455520279705525\nbatch 200, loss 0.21826332837343215\nbatch 300, loss 0.2296634064614773\nbatch 400, loss 0.2398214316368103\nModel accuracy: 86.56%\nNew highest accuracy: 0.8656195998191833\nEnd of epoch 20\n-------------------------\nEpoch 21\nbatch 100, loss 0.20347689174115657\nbatch 200, loss 0.21140942648053168\nbatch 300, loss 0.2108816261589527\nbatch 400, loss 0.22860585793852806\nModel accuracy: 86.97%\nNew highest accuracy: 0.8696917295455933\nEnd of epoch 21\n-------------------------\nEpoch 22\nbatch 100, loss 0.1942390874028206\nbatch 200, loss 0.2099410466849804\nbatch 300, loss 0.2186720285564661\nbatch 400, loss 0.22552408687770367\nModel accuracy: 86.33%\nEnd of epoch 22\n-------------------------\nEpoch 23\nbatch 100, loss 0.18661766462028026\nbatch 200, loss 0.20858853705227376\nbatch 300, loss 0.20480633154511452\nbatch 400, loss 0.21510949850082398\nModel accuracy: 86.74%\nEnd of epoch 23\n-------------------------\nEpoch 24\nbatch 100, loss 0.17789838716387749\nbatch 200, loss 0.19294073104858397\nbatch 300, loss 0.20428663805127145\nbatch 400, loss 0.21487028874456882\nModel accuracy: 86.53%\nEnd of epoch 24\n-------------------------\nEpoch 25\nbatch 100, loss 0.1770741441845894\nbatch 200, loss 0.19298666790127755\nbatch 300, loss 0.19812761060893536\nbatch 400, loss 0.20529725089669226\nModel accuracy: 87.09%\nNew highest accuracy: 0.8708552122116089\nEnd of epoch 25\n-------------------------\nEpoch 26\nbatch 100, loss 0.17795370914041997\nbatch 200, loss 0.18348632037639617\nbatch 300, loss 0.19024646796286107\nbatch 400, loss 0.205008674710989\nModel accuracy: 86.24%\nEnd of epoch 26\n-------------------------\nEpoch 27\nbatch 100, loss 0.16751805812120438\nbatch 200, loss 0.1890131201595068\nbatch 300, loss 0.1924343678355217\nbatch 400, loss 0.20168838880956172\nModel accuracy: 86.91%\nEnd of epoch 27\n-------------------------\nEpoch 28\nbatch 100, loss 0.169182011783123\nbatch 200, loss 0.17790321715176105\nbatch 300, loss 0.1879586534947157\nbatch 400, loss 0.18292610198259354\nModel accuracy: 86.45%\nEnd of epoch 28\n-------------------------\nEpoch 29\nbatch 100, loss 0.1662400298565626\nbatch 200, loss 0.17537716835737227\nbatch 300, loss 0.17895046889781951\nbatch 400, loss 0.18686383858323097\nModel accuracy: 86.97%\nEnd of epoch 29\n-------------------------\nEpoch 30\nbatch 100, loss 0.16220371913164855\nbatch 200, loss 0.1733390137553215\nbatch 300, loss 0.1783679297566414\nbatch 400, loss 0.18232472941279412\nModel accuracy: 86.50%\nEnd of epoch 30\n-------------------------\nEpoch 31\nbatch 100, loss 0.16122045144438743\nbatch 200, loss 0.16560416162014008\nbatch 300, loss 0.17328639417886735\nbatch 400, loss 0.18381614863872528\nModel accuracy: 86.45%\nEnd of epoch 31\n-------------------------\nEpoch 32\nbatch 100, loss 0.15512956656515597\nbatch 200, loss 0.16755689322948455\nbatch 300, loss 0.16981551706790923\nbatch 400, loss 0.17737352006137372\nModel accuracy: 86.68%\nEnd of epoch 32\n-------------------------\nEpoch 33\nbatch 100, loss 0.15085506230592727\nbatch 200, loss 0.16914329260587693\nbatch 300, loss 0.16748904056847094\nbatch 400, loss 0.1761363573372364\nModel accuracy: 86.77%\nEnd of epoch 33\n-------------------------\nEpoch 34\nbatch 100, loss 0.1573239029198885\nbatch 200, loss 0.1623350628465414\nbatch 300, loss 0.1593898731470108\nbatch 400, loss 0.18208031319081783\nModel accuracy: 86.53%\nEnd of epoch 34\n-------------------------\nEpoch 35\nbatch 100, loss 0.14681999817490576\nbatch 200, loss 0.15299119487404822\nbatch 300, loss 0.15951380953192712\nbatch 400, loss 0.1789644856005907\nModel accuracy: 87.14%\nNew highest accuracy: 0.8714368939399719\nEnd of epoch 35\n-------------------------\nEpoch 36\nbatch 100, loss 0.13780128542333842\nbatch 200, loss 0.15104680247604846\nbatch 300, loss 0.16281914472579956\nbatch 400, loss 0.16415032714605332\nModel accuracy: 86.56%\nEnd of epoch 36\n-------------------------\nEpoch 37\nbatch 100, loss 0.14140414912253618\nbatch 200, loss 0.14879427693784236\nbatch 300, loss 0.15043172776699065\nbatch 400, loss 0.16124424047768116\nModel accuracy: 86.65%\nEnd of epoch 37\n-------------------------\nEpoch 38\nbatch 100, loss 0.13963917940855025\nbatch 200, loss 0.14696975976228713\nbatch 300, loss 0.1498751526325941\nbatch 400, loss 0.16724546425044537\nModel accuracy: 86.50%\nEnd of epoch 38\n-------------------------\nEpoch 39\nbatch 100, loss 0.14469791296869516\nbatch 200, loss 0.14637224573642016\nbatch 300, loss 0.1623458718508482\nbatch 400, loss 0.15424846913665533\nModel accuracy: 86.68%\nEnd of epoch 39\n-------------------------\nEpoch 40\nbatch 100, loss 0.13476959645748138\nbatch 200, loss 0.14432808820158244\nbatch 300, loss 0.1522339789569378\nbatch 400, loss 0.15014691196382046\nModel accuracy: 86.07%\nEnd of epoch 40\n-------------------------\nEpoch 41\nbatch 100, loss 0.140174614302814\nbatch 200, loss 0.1508945333212614\nbatch 300, loss 0.15196941532194613\nbatch 400, loss 0.1483056141063571\nModel accuracy: 87.23%\nNew highest accuracy: 0.8723095059394836\nEnd of epoch 41\n-------------------------\nEpoch 42\nbatch 100, loss 0.13751666147261857\nbatch 200, loss 0.139865167029202\nbatch 300, loss 0.1380647499859333\nbatch 400, loss 0.1476748574525118\nModel accuracy: 86.97%\nEnd of epoch 42\n-------------------------\nEpoch 43\nbatch 100, loss 0.13006325714290143\nbatch 200, loss 0.1424691466242075\nbatch 300, loss 0.1392911682650447\nbatch 400, loss 0.14881628528237342\nModel accuracy: 86.91%\nEnd of epoch 43\n-------------------------\nEpoch 44\nbatch 100, loss 0.1238890501484275\nbatch 200, loss 0.137397523149848\nbatch 300, loss 0.13997289538383484\nbatch 400, loss 0.15086022712290287\nModel accuracy: 86.62%\nEnd of epoch 44\n-------------------------\nEpoch 45\nbatch 100, loss 0.1351698400080204\nbatch 200, loss 0.13414906188845635\nbatch 300, loss 0.1385753569379449\nbatch 400, loss 0.1407256292924285\nModel accuracy: 87.00%\nEnd of epoch 45\n-------------------------\nEpoch 46\nbatch 100, loss 0.12339191194623708\nbatch 200, loss 0.13611533541232348\nbatch 300, loss 0.14306778989732266\nbatch 400, loss 0.1465745335072279\nModel accuracy: 86.39%\nEnd of epoch 46\n-------------------------\nEpoch 47\nbatch 100, loss 0.1267332899570465\nbatch 200, loss 0.13155565716326237\nbatch 300, loss 0.13074469082057477\nbatch 400, loss 0.14163539059460162\nModel accuracy: 86.79%\nEnd of epoch 47\n-------------------------\nEpoch 48\nbatch 100, loss 0.12001773077994585\nbatch 200, loss 0.13302614204585553\nbatch 300, loss 0.14182888820767403\nbatch 400, loss 0.13658559188246727\nModel accuracy: 86.56%\nEnd of epoch 48\n-------------------------\nEpoch 49\nbatch 100, loss 0.12255903270095586\nbatch 200, loss 0.12362348977476359\nbatch 300, loss 0.1345749046653509\nbatch 400, loss 0.139198206178844\nModel accuracy: 87.03%\nEnd of epoch 49\n-------------------------\nEpoch 50\nbatch 100, loss 0.11813947290182114\nbatch 200, loss 0.13258680544793605\nbatch 300, loss 0.1362661897391081\nbatch 400, loss 0.14307982981204986\nModel accuracy: 86.77%\nEnd of epoch 50\n-------------------------\nEpoch 51\nbatch 100, loss 0.11312606774270534\nbatch 200, loss 0.13347773265093565\nbatch 300, loss 0.1291982262581587\nbatch 400, loss 0.1381356829404831\nModel accuracy: 86.62%\nEnd of epoch 51\n-------------------------\nEpoch 52\nbatch 100, loss 0.11617817420512438\nbatch 200, loss 0.12462763722985983\nbatch 300, loss 0.13041483018547295\nbatch 400, loss 0.13594958994537593\nModel accuracy: 86.39%\nEnd of epoch 52\n-------------------------\nEpoch 53\nbatch 100, loss 0.11495684150606394\nbatch 200, loss 0.12233315281569958\nbatch 300, loss 0.12879078544676303\nbatch 400, loss 0.13400998651981355\nModel accuracy: 86.74%\nEnd of epoch 53\n-------------------------\nEpoch 54\nbatch 100, loss 0.11334789831191301\nbatch 200, loss 0.12377498582005501\nbatch 300, loss 0.13191846806555987\nbatch 400, loss 0.1272725586593151\nModel accuracy: 87.11%\nEnd of epoch 54\n-------------------------\nEpoch 55\nbatch 100, loss 0.11219710145145655\nbatch 200, loss 0.11760351460427046\nbatch 300, loss 0.13237097259610892\nbatch 400, loss 0.13496288113296032\nModel accuracy: 87.26%\nNew highest accuracy: 0.8726003766059875\nEnd of epoch 55\n-------------------------\nEpoch 56\nbatch 100, loss 0.11166935112327338\nbatch 200, loss 0.11971858706325293\nbatch 300, loss 0.1202097450196743\nbatch 400, loss 0.12764150727540255\nModel accuracy: 86.36%\nEnd of epoch 56\n-------------------------\nEpoch 57\nbatch 100, loss 0.10257240902632475\nbatch 200, loss 0.11765343647450209\nbatch 300, loss 0.11703212883323431\nbatch 400, loss 0.12566446259617806\nModel accuracy: 87.52%\nNew highest accuracy: 0.8752182126045227\nEnd of epoch 57\n-------------------------\nEpoch 58\nbatch 100, loss 0.10790527418255806\nbatch 200, loss 0.12074193950742483\nbatch 300, loss 0.12469904065132141\nbatch 400, loss 0.1278570229932666\nModel accuracy: 86.45%\nEnd of epoch 58\n-------------------------\nEpoch 59\nbatch 100, loss 0.10504985239356757\nbatch 200, loss 0.11492413766682148\nbatch 300, loss 0.12118184957653284\nbatch 400, loss 0.13197467107325792\nModel accuracy: 86.82%\nEnd of epoch 59\n-------------------------\nEpoch 60\nbatch 100, loss 0.10825933616608381\nbatch 200, loss 0.11681649874895811\nbatch 300, loss 0.11647801648825407\nbatch 400, loss 0.13008976448327303\nModel accuracy: 87.20%\nEnd of epoch 60\n-------------------------\nEpoch 61\nbatch 100, loss 0.10180965416133404\nbatch 200, loss 0.11671413134783507\nbatch 300, loss 0.11545869991183282\nbatch 400, loss 0.12630713261663915\nModel accuracy: 87.58%\nNew highest accuracy: 0.8757998943328857\nEnd of epoch 61\n-------------------------\nEpoch 62\nbatch 100, loss 0.10995391804724931\nbatch 200, loss 0.10986792344599962\nbatch 300, loss 0.11852037061005831\nbatch 400, loss 0.12597699865698814\nModel accuracy: 87.55%\nEnd of epoch 62\n-------------------------\nEpoch 63\nbatch 100, loss 0.10668433804064989\nbatch 200, loss 0.1098334052413702\nbatch 300, loss 0.12044183690100908\nbatch 400, loss 0.1158076674491167\nModel accuracy: 87.43%\nEnd of epoch 63\n-------------------------\nEpoch 64\nbatch 100, loss 0.10601182598620654\nbatch 200, loss 0.1097069663926959\nbatch 300, loss 0.11127791944891215\nbatch 400, loss 0.12495768345892429\nModel accuracy: 87.49%\nEnd of epoch 64\n-------------------------\nEpoch 65\nbatch 100, loss 0.10043533198535443\nbatch 200, loss 0.110718375146389\nbatch 300, loss 0.11771305467933417\nbatch 400, loss 0.12339903887361288\nModel accuracy: 87.09%\nEnd of epoch 65\n-------------------------\nEpoch 66\nbatch 100, loss 0.09939460391178727\nbatch 200, loss 0.1144513813778758\nbatch 300, loss 0.1066908597573638\nbatch 400, loss 0.11124557465314865\nModel accuracy: 87.32%\nEnd of epoch 66\n-------------------------\nEpoch 67\nbatch 100, loss 0.10762397821992636\nbatch 200, loss 0.10650279968976975\nbatch 300, loss 0.10767663531005382\nbatch 400, loss 0.11479169704020023\nModel accuracy: 87.14%\nEnd of epoch 67\n-------------------------\nEpoch 68\nbatch 100, loss 0.10010240273550153\nbatch 200, loss 0.10536838762462139\nbatch 300, loss 0.11181689091026784\nbatch 400, loss 0.12067786559462547\nModel accuracy: 87.58%\nEnd of epoch 68\n-------------------------\nEpoch 69\nbatch 100, loss 0.0952544462122023\nbatch 200, loss 0.11155086569488049\nbatch 300, loss 0.11455501675605774\nbatch 400, loss 0.11072668544948101\nModel accuracy: 86.88%\nEnd of epoch 69\n-------------------------\nEpoch 70\nbatch 100, loss 0.0936788029782474\nbatch 200, loss 0.10011235978454351\nbatch 300, loss 0.10793911900371313\nbatch 400, loss 0.11547569010406733\nModel accuracy: 87.11%\nEnd of epoch 70\n-------------------------\nEpoch 71\nbatch 100, loss 0.10421726990491152\nbatch 200, loss 0.10555256571620703\nbatch 300, loss 0.11177353248000145\nbatch 400, loss 0.1156128654256463\nModel accuracy: 87.17%\nEnd of epoch 71\n-------------------------\nEpoch 72\nbatch 100, loss 0.1049131329357624\nbatch 200, loss 0.10579334639012813\nbatch 300, loss 0.1120777241513133\nbatch 400, loss 0.1117202722094953\nModel accuracy: 87.75%\nNew highest accuracy: 0.8775451183319092\nEnd of epoch 72\n-------------------------\nEpoch 73\nbatch 100, loss 0.09633252434432507\nbatch 200, loss 0.10547576978802681\nbatch 300, loss 0.1073480474203825\nbatch 400, loss 0.11320575729012489\nModel accuracy: 87.49%\nEnd of epoch 73\n-------------------------\nEpoch 74\nbatch 100, loss 0.09525072108954191\nbatch 200, loss 0.10154122976586223\nbatch 300, loss 0.10131758511066437\nbatch 400, loss 0.11673280103132129\nModel accuracy: 87.70%\nEnd of epoch 74\n-------------------------\nEpoch 75\nbatch 100, loss 0.09710830971598625\nbatch 200, loss 0.09735006069764494\nbatch 300, loss 0.10233162555843592\nbatch 400, loss 0.11405609011650085\nModel accuracy: 87.32%\nEnd of epoch 75\n-------------------------\nEpoch 76\nbatch 100, loss 0.09434206038713455\nbatch 200, loss 0.10505881786346435\nbatch 300, loss 0.11231484092772007\nbatch 400, loss 0.10525997526943684\nModel accuracy: 86.65%\nEnd of epoch 76\n-------------------------\nEpoch 77\nbatch 100, loss 0.0943560291454196\nbatch 200, loss 0.10224988359957933\nbatch 300, loss 0.10538167864084244\nbatch 400, loss 0.096354298312217\nModel accuracy: 86.97%\nEnd of epoch 77\n-------------------------\nEpoch 78\nbatch 100, loss 0.09480333685874939\nbatch 200, loss 0.0944455861672759\nbatch 300, loss 0.10802934501320123\nbatch 400, loss 0.10488686054944991\nModel accuracy: 87.23%\nEnd of epoch 78\n-------------------------\nEpoch 79\nbatch 100, loss 0.0941369539871812\nbatch 200, loss 0.09605370812118054\nbatch 300, loss 0.10392073981463909\nbatch 400, loss 0.11097734656184911\nModel accuracy: 87.26%\nEnd of epoch 79\n-------------------------\nEpoch 80\nbatch 100, loss 0.08992297133430839\nbatch 200, loss 0.09761941272765398\nbatch 300, loss 0.10265288643538951\nbatch 400, loss 0.10450071014463902\nModel accuracy: 87.32%\nEnd of epoch 80\n-------------------------\nEpoch 81\nbatch 100, loss 0.09164213659241796\nbatch 200, loss 0.09774349711835384\nbatch 300, loss 0.09652327008545399\nbatch 400, loss 0.11365355037152768\nModel accuracy: 87.14%\nEnd of epoch 81\n-------------------------\nEpoch 82\nbatch 100, loss 0.09872948873788118\nbatch 200, loss 0.09633034691214562\nbatch 300, loss 0.10720755336806179\nbatch 400, loss 0.10877235535532236\nModel accuracy: 87.29%\nEnd of epoch 82\n-------------------------\nEpoch 83\nbatch 100, loss 0.08615791471675038\nbatch 200, loss 0.0969615351781249\nbatch 300, loss 0.10451982267200947\nbatch 400, loss 0.10347757913172245\nModel accuracy: 87.35%\nEnd of epoch 83\n-------------------------\nEpoch 84\nbatch 100, loss 0.09008787389844657\nbatch 200, loss 0.09931891463696957\nbatch 300, loss 0.10183713119477034\nbatch 400, loss 0.10709619529545307\nModel accuracy: 87.46%\nEnd of epoch 84\n-------------------------\nEpoch 85\nbatch 100, loss 0.08623319059610367\nbatch 200, loss 0.0981818121112883\nbatch 300, loss 0.09549403263255954\nbatch 400, loss 0.10199526280164718\nModel accuracy: 87.32%\nEnd of epoch 85\n-------------------------\nEpoch 86\nbatch 100, loss 0.08751846086233854\nbatch 200, loss 0.10378085581585765\nbatch 300, loss 0.09046246636658907\nbatch 400, loss 0.09905910003930331\nModel accuracy: 87.67%\nEnd of epoch 86\n-------------------------\nEpoch 87\nbatch 100, loss 0.08956490609794855\nbatch 200, loss 0.09870004137977957\nbatch 300, loss 0.09708821553736925\nbatch 400, loss 0.10178605049848556\nModel accuracy: 87.09%\nEnd of epoch 87\n-------------------------\nEpoch 88\nbatch 100, loss 0.08944101119413972\nbatch 200, loss 0.0936494743078947\nbatch 300, loss 0.09616516340523958\nbatch 400, loss 0.10267581447958946\nModel accuracy: 87.67%\nEnd of epoch 88\n-------------------------\nEpoch 89\nbatch 100, loss 0.0885039086639881\nbatch 200, loss 0.09610162362456322\nbatch 300, loss 0.09747009973973036\nbatch 400, loss 0.10595266478136182\nModel accuracy: 87.41%\nEnd of epoch 89\n-------------------------\nEpoch 90\nbatch 100, loss 0.0770477294549346\nbatch 200, loss 0.099730328284204\nbatch 300, loss 0.09587829411029816\nbatch 400, loss 0.09603298252448439\nModel accuracy: 87.78%\nNew highest accuracy: 0.8778359889984131\nEnd of epoch 90\n-------------------------\nEpoch 91\nbatch 100, loss 0.08408609050326049\nbatch 200, loss 0.09370174456387759\nbatch 300, loss 0.0966075786575675\nbatch 400, loss 0.09802194258198142\nModel accuracy: 87.41%\nEnd of epoch 91\n-------------------------\nEpoch 92\nbatch 100, loss 0.08376560196280479\nbatch 200, loss 0.09021430224180221\nbatch 300, loss 0.09928188558667898\nbatch 400, loss 0.09303318437188864\nModel accuracy: 87.52%\nEnd of epoch 92\n-------------------------\nEpoch 93\nbatch 100, loss 0.0883931003883481\nbatch 200, loss 0.08799420539289712\nbatch 300, loss 0.0944927965849638\nbatch 400, loss 0.0985324920155108\nModel accuracy: 87.64%\nEnd of epoch 93\n-------------------------\nEpoch 94\nbatch 100, loss 0.08059066124260425\nbatch 200, loss 0.08742215245962143\nbatch 300, loss 0.08818385323509574\nbatch 400, loss 0.09771269951015711\nModel accuracy: 87.75%\nEnd of epoch 94\n-------------------------\nEpoch 95\nbatch 100, loss 0.08617315959185362\nbatch 200, loss 0.09023557052016258\nbatch 300, loss 0.08764048373326659\nbatch 400, loss 0.09584249943494796\nModel accuracy: 87.26%\nEnd of epoch 95\n-------------------------\nEpoch 96\nbatch 100, loss 0.08249812372028828\nbatch 200, loss 0.08964972715824843\nbatch 300, loss 0.08927668014541268\nbatch 400, loss 0.09504545725882053\nModel accuracy: 87.09%\nEnd of epoch 96\n-------------------------\nEpoch 97\nbatch 100, loss 0.08657911265268922\nbatch 200, loss 0.08752006120979786\nbatch 300, loss 0.09831780819222331\nbatch 400, loss 0.09034885132685304\nModel accuracy: 87.29%\nEnd of epoch 97\n-------------------------\nEpoch 98\nbatch 100, loss 0.08336411898955703\nbatch 200, loss 0.08658858956769108\nbatch 300, loss 0.09285338137298822\nbatch 400, loss 0.09450048698112369\nModel accuracy: 86.94%\nEnd of epoch 98\n-------------------------\nEpoch 99\nbatch 100, loss 0.08213733604177832\nbatch 200, loss 0.09120012558996678\nbatch 300, loss 0.08984984505921602\nbatch 400, loss 0.10017645239830017\nModel accuracy: 87.11%\nEnd of epoch 99\n-------------------------\nEpoch 100\nbatch 100, loss 0.08137811250984668\nbatch 200, loss 0.08539615284651518\nbatch 300, loss 0.09137881260365248\nbatch 400, loss 0.09100246150046587\nModel accuracy: 87.46%\nEnd of epoch 100\n-------------------------\nEpoch 101\nbatch 100, loss 0.07930496525019408\nbatch 200, loss 0.08400334604084492\nbatch 300, loss 0.0855117904394865\nbatch 400, loss 0.09232061143964529\nModel accuracy: 87.38%\nEnd of epoch 101\n-------------------------\nEpoch 102\nbatch 100, loss 0.08289765767753124\nbatch 200, loss 0.08815545243211091\nbatch 300, loss 0.08829482391476631\nbatch 400, loss 0.09039539087563753\nModel accuracy: 87.23%\nEnd of epoch 102\n-------------------------\nEpoch 103\nbatch 100, loss 0.08119489477947354\nbatch 200, loss 0.08984118312597275\nbatch 300, loss 0.09681978680193425\nbatch 400, loss 0.08952074136584998\nModel accuracy: 86.85%\nEnd of epoch 103\n-------------------------\nEpoch 104\nbatch 100, loss 0.08111560147255659\nbatch 200, loss 0.0833469331637025\nbatch 300, loss 0.088416967689991\nbatch 400, loss 0.09031455351039767\nModel accuracy: 87.38%\nEnd of epoch 104\n-------------------------\nEpoch 105\nbatch 100, loss 0.07392966020852328\nbatch 200, loss 0.08815220635384322\nbatch 300, loss 0.08357393940910697\nbatch 400, loss 0.09532751824706792\nModel accuracy: 87.35%\nEnd of epoch 105\n-------------------------\nEpoch 106\nbatch 100, loss 0.07947631476446987\nbatch 200, loss 0.09054830428212882\nbatch 300, loss 0.0889788007363677\nbatch 400, loss 0.08650729153305292\nModel accuracy: 87.78%\nEnd of epoch 106\n-------------------------\nEpoch 107\nbatch 100, loss 0.08217606570571662\nbatch 200, loss 0.08046228423714638\nbatch 300, loss 0.08151343079283833\nbatch 400, loss 0.08912567749619484\nModel accuracy: 87.43%\nEnd of epoch 107\n-------------------------\nEpoch 108\nbatch 100, loss 0.08707285387441516\nbatch 200, loss 0.0808964579179883\nbatch 300, loss 0.08648431923240424\nbatch 400, loss 0.08845541972666979\nModel accuracy: 87.09%\nEnd of epoch 108\n-------------------------\nEpoch 109\nbatch 100, loss 0.07416098069399596\nbatch 200, loss 0.08108017161488533\nbatch 300, loss 0.08783580781891942\nbatch 400, loss 0.08868230246007443\nModel accuracy: 86.91%\nEnd of epoch 109\n-------------------------\nEpoch 110\nbatch 100, loss 0.08106791527941823\nbatch 200, loss 0.07800445534288883\nbatch 300, loss 0.08207052554003895\nbatch 400, loss 0.09356683790683747\nModel accuracy: 87.23%\nNo improvement in accuracy. Stopping early\n Best accuracy -> 0.8778359889984131\n","output_type":"stream"}]},{"cell_type":"code","source":"final_model = SimpleNet(embD=300).to(device)\nfinal_model.load_state_dict(torch.load(\"/kaggle/working/bestModel.pt\"))\nfinal_model.eval()\ny_pred = final_model(testSetXw, testSetXl)\ny_pred = torch.argmax(y_pred, dim=1)\nacc = (y_pred == testSetY).float().mean()\nacc = float(acc)\nprint(\"Final model accuracy: %.2f%%\" % (acc*100))","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:22:06.463712Z","iopub.execute_input":"2023-04-02T20:22:06.464069Z","iopub.status.idle":"2023-04-02T20:22:06.529205Z","shell.execute_reply.started":"2023-04-02T20:22:06.464037Z","shell.execute_reply":"2023-04-02T20:22:06.528197Z"},"trusted":true},"execution_count":170,"outputs":[{"name":"stdout","text":"Final model accuracy: 88.19%\n","output_type":"stream"}]},{"cell_type":"markdown","source":"best 80","metadata":{}},{"cell_type":"markdown","source":"## Submission","metadata":{}},{"cell_type":"code","source":"test = pd.read_csv(\"/kaggle/input/silero-stress-predictor/test.csv\")\ntest","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:22:09.524935Z","iopub.execute_input":"2023-04-02T20:22:09.525309Z","iopub.status.idle":"2023-04-02T20:22:09.577452Z","shell.execute_reply.started":"2023-04-02T20:22:09.525276Z","shell.execute_reply":"2023-04-02T20:22:09.576393Z"},"trusted":true},"execution_count":171,"outputs":[{"execution_count":171,"output_type":"execute_result","data":{"text/plain":" id word num_syllables lemma\n0 0 эпилепсия 5 эпилепсия\n1 1 относящейся 5 относиться\n2 2 размышлениями 6 размышление\n3 3 модемы 3 модем\n4 4 солнц 1 солнце\n... ... ... ... ...\n29955 29955 донбасса 3 донбасс\n29956 29956 обложка 3 обложка\n29957 29957 правителя 4 правитель\n29958 29958 шерстяной 3 шерстяной\n29959 29959 оптимизации 6 оптимизация\n\n[29960 rows x 4 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></th>\n <th>id</th>\n <th>word</th>\n <th>num_syllables</th>\n <th>lemma</th>\n </tr>\n </thead>\n <tbody>\n <tr>\n <th>0</th>\n <td>0</td>\n <td>эпилепсия</td>\n <td>5</td>\n <td>эпилепсия</td>\n </tr>\n <tr>\n <th>1</th>\n <td>1</td>\n <td>относящейся</td>\n <td>5</td>\n <td>относиться</td>\n </tr>\n <tr>\n <th>2</th>\n <td>2</td>\n <td>размышлениями</td>\n <td>6</td>\n <td>размышление</td>\n </tr>\n <tr>\n <th>3</th>\n <td>3</td>\n <td>модемы</td>\n <td>3</td>\n <td>модем</td>\n </tr>\n <tr>\n <th>4</th>\n <td>4</td>\n <td>солнц</td>\n <td>1</td>\n <td>солнце</td>\n </tr>\n <tr>\n <th>...</th>\n <td>...</td>\n <td>...</td>\n <td>...</td>\n <td>...</td>\n </tr>\n <tr>\n <th>29955</th>\n <td>29955</td>\n <td>донбасса</td>\n <td>3</td>\n <td>донбасс</td>\n </tr>\n <tr>\n <th>29956</th>\n <td>29956</td>\n <td>обложка</td>\n <td>3</td>\n <td>обложка</td>\n </tr>\n <tr>\n <th>29957</th>\n <td>29957</td>\n <td>правителя</td>\n <td>4</td>\n <td>правитель</td>\n </tr>\n <tr>\n <th>29958</th>\n <td>29958</td>\n <td>шерстяной</td>\n <td>3</td>\n <td>шерстяной</td>\n </tr>\n <tr>\n <th>29959</th>\n <td>29959</td>\n <td>оптимизации</td>\n <td>6</td>\n <td>оптимизация</td>\n </tr>\n </tbody>\n</table>\n<p>29960 rows × 4 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"testRxW = test[['word']].applymap(word2ind)","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:24:20.985540Z","iopub.execute_input":"2023-04-02T20:24:20.986457Z","iopub.status.idle":"2023-04-02T20:24:21.352155Z","shell.execute_reply.started":"2023-04-02T20:24:20.986403Z","shell.execute_reply":"2023-04-02T20:24:21.351151Z"},"trusted":true},"execution_count":184,"outputs":[]},{"cell_type":"code","source":"wordsTS = pd.DataFrame(testRxW['word'].to_list(), columns = ['w1', 'w2', 'w3', 'w4', 'w5', 'w6'])\n\n# test_readyXGB = pd.concat([lemmasTS, wordsTS, test['num_syllables']], axis = 1)\n# test_readyXGB","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:24:35.214064Z","iopub.execute_input":"2023-04-02T20:24:35.214567Z","iopub.status.idle":"2023-04-02T20:24:35.303057Z","shell.execute_reply.started":"2023-04-02T20:24:35.214525Z","shell.execute_reply":"2023-04-02T20:24:35.302069Z"},"trusted":true},"execution_count":186,"outputs":[]},{"cell_type":"code","source":"SubTenzW = torch.tensor(wordsTS.values.astype(np.int32)).to(device)","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:24:53.802897Z","iopub.execute_input":"2023-04-02T20:24:53.803294Z","iopub.status.idle":"2023-04-02T20:24:53.810269Z","shell.execute_reply.started":"2023-04-02T20:24:53.803260Z","shell.execute_reply":"2023-04-02T20:24:53.809243Z"},"trusted":true},"execution_count":187,"outputs":[]},{"cell_type":"code","source":"final_model.eval()\ny_pred = final_model(SubTenzW)\ny_pred = torch.argmax(y_pred, dim=1)\ny_pred","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:25:09.670327Z","iopub.execute_input":"2023-04-02T20:25:09.670910Z","iopub.status.idle":"2023-04-02T20:25:09.781638Z","shell.execute_reply.started":"2023-04-02T20:25:09.670874Z","shell.execute_reply":"2023-04-02T20:25:09.780716Z"},"trusted":true},"execution_count":188,"outputs":[{"execution_count":188,"output_type":"execute_result","data":{"text/plain":"tensor([2, 2, 2, ..., 1, 2, 3], device='cuda:0')"},"metadata":{}}]},{"cell_type":"code","source":"submPred = y_pred + 1","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:25:10.817185Z","iopub.execute_input":"2023-04-02T20:25:10.817555Z","iopub.status.idle":"2023-04-02T20:25:10.823092Z","shell.execute_reply.started":"2023-04-02T20:25:10.817521Z","shell.execute_reply":"2023-04-02T20:25:10.821839Z"},"trusted":true},"execution_count":189,"outputs":[]},{"cell_type":"code","source":"submPred","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:25:11.449844Z","iopub.execute_input":"2023-04-02T20:25:11.450548Z","iopub.status.idle":"2023-04-02T20:25:11.460170Z","shell.execute_reply.started":"2023-04-02T20:25:11.450508Z","shell.execute_reply":"2023-04-02T20:25:11.458954Z"},"trusted":true},"execution_count":190,"outputs":[{"execution_count":190,"output_type":"execute_result","data":{"text/plain":"tensor([3, 3, 3, ..., 2, 3, 4], device='cuda:0')"},"metadata":{}}]},{"cell_type":"code","source":"submSER = pd.Series(submPred.cpu(), name='stress')\nsubmDF = pd.concat([test['id'], submSER], axis=1)","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:25:12.172001Z","iopub.execute_input":"2023-04-02T20:25:12.174437Z","iopub.status.idle":"2023-04-02T20:25:12.181093Z","shell.execute_reply.started":"2023-04-02T20:25:12.174388Z","shell.execute_reply":"2023-04-02T20:25:12.180086Z"},"trusted":true},"execution_count":191,"outputs":[]},{"cell_type":"code","source":"submDF","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:25:12.899001Z","iopub.execute_input":"2023-04-02T20:25:12.899674Z","iopub.status.idle":"2023-04-02T20:25:12.912010Z","shell.execute_reply.started":"2023-04-02T20:25:12.899638Z","shell.execute_reply":"2023-04-02T20:25:12.910792Z"},"trusted":true},"execution_count":192,"outputs":[{"execution_count":192,"output_type":"execute_result","data":{"text/plain":" id stress\n0 0 3\n1 1 3\n2 2 3\n3 3 2\n4 4 1\n... ... ...\n29955 29955 2\n29956 29956 2\n29957 29957 2\n29958 29958 3\n29959 29959 4\n\n[29960 rows x 2 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></th>\n <th>id</th>\n <th>stress</th>\n </tr>\n </thead>\n <tbody>\n <tr>\n <th>0</th>\n <td>0</td>\n <td>3</td>\n </tr>\n <tr>\n <th>1</th>\n <td>1</td>\n <td>3</td>\n </tr>\n <tr>\n <th>2</th>\n <td>2</td>\n <td>3</td>\n </tr>\n <tr>\n <th>3</th>\n <td>3</td>\n <td>2</td>\n </tr>\n <tr>\n <th>4</th>\n <td>4</td>\n <td>1</td>\n </tr>\n <tr>\n <th>...</th>\n <td>...</td>\n <td>...</td>\n </tr>\n <tr>\n <th>29955</th>\n <td>29955</td>\n <td>2</td>\n </tr>\n <tr>\n <th>29956</th>\n <td>29956</td>\n <td>2</td>\n </tr>\n <tr>\n <th>29957</th>\n <td>29957</td>\n <td>2</td>\n </tr>\n <tr>\n <th>29958</th>\n <td>29958</td>\n <td>3</td>\n </tr>\n <tr>\n <th>29959</th>\n <td>29959</td>\n <td>4</td>\n </tr>\n </tbody>\n</table>\n<p>29960 rows × 2 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"submDF.to_csv(\"submission_IvanovNN10.csv\", index=False)","metadata":{"execution":{"iopub.status.busy":"2023-04-02T20:25:19.205310Z","iopub.execute_input":"2023-04-02T20:25:19.205682Z","iopub.status.idle":"2023-04-02T20:25:19.238917Z","shell.execute_reply.started":"2023-04-02T20:25:19.205648Z","shell.execute_reply":"2023-04-02T20:25:19.237948Z"},"trusted":true},"execution_count":194,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment