kernel_id int64 24.2k 23.3M | prompt stringlengths 8 1.85M | completetion stringlengths 1 182k | comp_name stringlengths 5 57 |
|---|---|---|---|
19,079,889 | def get_word_level_logits(start_logits,
end_logits,
model_type,
tweet_offsets_word_level):
tweet_offsets_word_level = np.array(tweet_offsets_word_level)
if model_type == "roberta-base" or model_type == "roberta-large" or model_type == "roberta-base-squad":
logit_offset = 4
elif(model_type == "albert-base-v2")or(model_... | validation_data_indices = df.sample(frac=0.2 ).index
validation_df = df.loc[validation_data_indices, :].reset_index(drop=True)
train_df = df.drop(validation_data_indices, axis=0 ).reset_index(drop=True)
test_df = pd.read_csv('.. /input/nlp-getting-started/test.csv')
x_train, y_train = create_inputs_with_targets(list... | Natural Language Processing with Disaster Tweets |
19,079,889 | def get_different_model_logits(model_type="roberta-base", \
hidden_layers=[-1, -2, -3, -4], \
checkpoint_path=".. /input/tweetrobertabase5fold42v8/", \
Config=None):
if model_type == "roberta-base" or model_type == "roberta-large" or model_type == "roberta-base-squad":
offsets = 4
elif model_type == "albert-base-v2" ... | def create_model(model_name, max_len=128):
seed = 500
my_init = tf.keras.initializers.glorot_uniform(seed)
max_len = max_len
encoder = transformers.TFAutoModel.from_pretrained(model_name)
encoder.trainable = True
input_ids = keras.layers.Input(shape=(max_len,), dtype=tf.int32)
attention_mask = keras.layers.Input(sha... | Natural Language Processing with Disaster Tweets |
19,079,889 | roberta_base_start_logits_token_level_42, roberta_base_end_logits_token_level_42, \
roberta_base_start_logits_word_level_42, roberta_base_end_logits_word_level_42, \
roberta_base_word_level_bbx, roberta_base_token_level_offsets, \
tweets, tweets_with_extra_spaces, sentiments, roberta_base_tokenizer = get_different_mode... | epochs = 20
lr = 2e-4
use_tpu = True
if use_tpu:
tpu = tf.distribute.cluster_resolver.TPUClusterResolver()
tf.config.experimental_connect_to_cluster(tpu)
tf.tpu.experimental.initialize_tpu_system(tpu)
strategy = tf.distribute.experimental.TPUStrategy(tpu)
with strategy.scope() :
model = create_model(base_model, max_... | Natural Language Processing with Disaster Tweets |
19,079,889 | sample_size = len(roberta_base_start_logits_token_level_42)
roberta_base_start_logits_token_level = [(roberta_base_start_logits_token_level_42[i] +
roberta_base_start_logits_token_level_666[i] +
roberta_base_start_logits_token_level_1234[i]
)/ 3 for i in range(sample_size)]
roberta_base_end_logits_token_level = [(rob... | my_callbacks = [keras.callbacks.EarlyStopping(monitor='val_binary_accuracy', patience=2, mode='max', restore_best_weights=True)]
hist = model.fit(x_train,
y_train,
validation_data =(x_val, y_val),
epochs= epochs,
batch_size= 128,
callbacks = my_callbacks,
verbose= 1 ) | Natural Language Processing with Disaster Tweets |
19,079,889 | _, _, albert_large_start_logits_word_level, albert_large_end_logits_word_level, _, _, _, _, _, _ = get_different_model_logits(model_type="albert-large-v2", \
hidden_layers=[-1, -2, -3, -4], \
checkpoint_path=".. /input/tweetalbertlargenewpipelinepreprocessingv1/",\
Config=Config )<define_search_model> | predictions = model.predict(x_test)
ids = list(test_df['id'])
target = [round(i[0])for i in predictions]
sub = pd.DataFrame({'id':ids, 'target':target}, index=None)
sub.to_csv('submission.csv', index=False)
sub | Natural Language Processing with Disaster Tweets |
19,079,889 | <define_variables><EOS> | Natural Language Processing with Disaster Tweets | |
18,800,274 | <SOS> metric: meanfscore Kaggle data source: natural-language-processing-with-disaster-tweets<define_variables> | !pip install transformers==3.5.1
!pip install pyspellchecker | Natural Language Processing with Disaster Tweets |
18,800,274 | all_results = []
all_start_end = []
base_model_type = "roberta-base"
CURR_PATH = ".. /input/"
tokenizer = tokenizers.ByteLevelBPETokenizer(
vocab_file=os.path.join(CURR_PATH, "transformers-vocab/{}-vocab.json".format(base_model_type)) ,
merges_file=os.path.join(CURR_PATH, "transformers-vocab/{}-merges.txt".format(base... | import pandas as pd
import torchtext
from transformers import BertTokenizer, BertForMaskedLM, BertConfig
import transformers
import torch
from torch.utils.data import Dataset, DataLoader
from torch import optim
from torch import cuda
from sklearn.model_selection import train_test_split
import re
import string
| Natural Language Processing with Disaster Tweets |
18,800,274 | submission = pd.read_csv(os.path.join("/kaggle/input/tweet-sentiment-extraction", "sample_submission.csv"))
test = pd.read_csv(os.path.join("/kaggle/input/tweet-sentiment-extraction", "test.csv"))<feature_engineering> | train_val_df = pd.read_csv("/kaggle/input/nlp-getting-started/train.csv")
test_df = pd.read_csv("/kaggle/input/nlp-getting-started/test.csv" ) | Natural Language Processing with Disaster Tweets |
18,800,274 | for i in range(len(submission)) :
if test['sentiment'][i] == 'neutral' or len(test['text'][i].split())< 4:
submission.loc[i, 'selected_text'] = test['text'][i]
else:
submission.loc[i, 'selected_text'] = all_results[i]<feature_engineering> | train_val_df = train_val_df.loc[:,["text","target"]]
test_df = test_df.loc[:,["text"]]
test_df["target"] = [0]*len(test_df["text"] ) | Natural Language Processing with Disaster Tweets |
18,800,274 | submission["text"] = test["text"]<categorify> | print(train_val_df)
print(test_df.head())
original_df = train_val_df.copy() | Natural Language Processing with Disaster Tweets |
18,800,274 | def reverse_preprocessing(text):
text = text.replace(".... ", ".... ")
text = text.replace("...", "...")
text = text.replace(".. ", ".. ")
text = text.replace("! ! ! !", "!!!!")
text = text.replace("! ! !", "!!!")
text = text.replace("! !", "!!")
text = text.replace("? ? ? ?", "????")
text = text.replace("? ? ?"... | mispell_dict = {"aren't" : "are not",
"can't" : "cannot",
"couldn't" : "could not",
"couldnt" : "could not",
"didn't" : "did not",
"doesn't" : "does not",
"doesnt" : "does not",
"don't" : "do not",
"hadn't" : "had not",
"hasn't" : "has not",
"haven't" : "have not",
"havent" : "have not",
"he'd" : "he would",
"he'll" : ... | Natural Language Processing with Disaster Tweets |
18,800,274 | submission["new_selected"] = submission.apply(lambda x: pp_v2(x.text, x.selected_text), axis=1 )<feature_engineering> | test_df.to_csv("test.tsv", sep='\t', index=False, header=None)
print(test_df.shape)
train_val_df.to_csv("train_eval.tsv", sep='\t', index=False, header=None)
print(train_val_df.shape ) | Natural Language Processing with Disaster Tweets |
18,800,274 | submission.selected_text = submission["new_selected"]<save_to_csv> | max_length = 50
def tokenizer_50(input_text):
return tokenizer.encode(input_text, max_length=50, return_tensors='pt')[0]
TEXT = torchtext.data.Field(sequential=True, tokenize=tokenizer_50, use_vocab=False, lower=False,
include_lengths=True, batch_first=True, fix_length=max_length, pad_token=0)
LABEL = torchtext.data... | Natural Language Processing with Disaster Tweets |
18,800,274 | submission[["textID","selected_text"]].to_csv("submission.csv", index=False )<set_options> | tokenizer = BertTokenizer.from_pretrained('bert-base-cased' ) | Natural Language Processing with Disaster Tweets |
18,800,274 | numpy.set_printoptions(threshold=sys.maxsize)
<load_from_csv> | dataset_train_eval, dataset_test = torchtext.data.TabularDataset.splits(path='.', train='./train_eval.tsv', test='./test.tsv', format='tsv', fields=[('Text', TEXT),('Label', LABEL)] ) | Natural Language Processing with Disaster Tweets |
18,800,274 | train= pd.read_csv('.. /input/Kannada-MNIST/train.csv')
test= pd.read_csv('.. /input/Kannada-MNIST/test.csv', index_col='id')
print('Train shape: ', train.shape)
print('Train missing vals: ', train.isna().sum().sum())
print('Test shape: ', test.shape)
print('Test missing vals: ', test.isna().sum().sum() )<prepare_... | dataset_train, dataset_eval = dataset_train_eval.split(
split_ratio=1.0 - 1800/7613, random_state=random.seed(1234))
print(dataset_train.__len__())
print(dataset_eval.__len__())
print(dataset_test.__len__() ) | Natural Language Processing with Disaster Tweets |
18,800,274 | x_train= train.iloc[:, 1:].values
y_train= train.iloc[:, 0].values
x_test= test.iloc[:, :].values<split> | print(tokenizer.convert_ids_to_tokens(item.Text.tolist()))
print(int(item.Label)) | Natural Language Processing with Disaster Tweets |
18,800,274 | x_train= x_train.reshape(train.shape[0], 28, 28, 1)
x_test= x_test.reshape(test.shape[0], 28, 28, 1)
y_train= to_categorical(y_train, 10)
x_train, x_valid, y_train, y_valid= train_test_split(x_train, y_train, test_size= 0.2, random_state= 42 )<choose_model_class> | batch_size = 32
dl_train = torchtext.data.Iterator(
dataset_train, batch_size=batch_size, train=True)
dl_eval = torchtext.data.Iterator(
dataset_eval, batch_size=batch_size, train=False, sort=False)
dl_test = torchtext.data.Iterator(
dataset_test, batch_size=batch_size, train=False, sort=False)
dataloaders_dict =... | Natural Language Processing with Disaster Tweets |
18,800,274 |
<choose_model_class> | model = BertModel.from_pretrained('bert-base-cased' ) | Natural Language Processing with Disaster Tweets |
18,800,274 | num_classes= 10
lr= 0.001
batch_size= 1024
model= Sequential()
model.add(Conv2D(32, kernel_size=3, padding='same', activation='relu', input_shape=(28, 28, 1)))
model.add(BatchNormalization())
model.add(Conv2D(32, kernel_size=3, padding='same', activation='relu'))
model.add(BatchNormalization())
model.add(Conv2D(32, ... | class BertForTwitter(nn.Module):
def __init__(self):
super(BertForTwitter, self ).__init__()
self.bert = model
self.cls = nn.Linear(in_features=768, out_features=2)
nn.init.normal_(self.cls.weight, std=0.02)
nn.init.normal_(self.cls.bias, 0)
def forward(self, input_ids):
result = self.bert(input_ids)
vec_0 = resu... | Natural Language Processing with Disaster Tweets |
18,800,274 | img_gen= ImageDataGenerator(rotation_range= 5, zoom_range= 0.5, horizontal_flip= 0, vertical_flip= 0,
width_shift_range= 2, height_shift_range= 2, rescale= 1/255)
valid_img_gen = ImageDataGenerator(rescale=1./255.)
learning_rate_reduction = ReduceLROnPlateau(monitor='val_loss', patience=200, verbose=1, factor=0.2)
es... | net = BertForTwitter()
net.train()
print('ネットワーク設定完了' ) | Natural Language Processing with Disaster Tweets |
18,800,274 | pred= model.predict_classes(x_test/255)
sub= pd.read_csv('.. /input/Kannada-MNIST/sample_submission.csv')
sub['label']= pred
sub.to_csv('submission.csv', index= False )<choose_model_class> | for param in net.parameters() :
param.requires_grad = False
for param in net.bert.encoder.layer[-1].parameters() :
param.requires_grad = True
for param in net.cls.parameters() :
param.requires_grad = True | Natural Language Processing with Disaster Tweets |
18,800,274 | def cnn_model() :
inp = tf.keras.Input(shape=(28,28,1))
x1 = tf.keras.layers.Conv2D(128,(1,1), strides=(1,1), activation='relu' )(inp)
x1 = tf.keras.layers.BatchNormalization()(x1)
x3 = tf.keras.layers.Conv2D(128,(3,3), padding='same', strides=(1,1), activation='relu' )(inp)
x3 = tf.keras.layers.BatchNormalization()... | optimizer = optim.Adam([
{'params': net.bert.encoder.layer[-1].parameters() , 'lr': 5e-5},
{'params': net.cls.parameters() , 'lr': 1e-4}
])
criterion = nn.CrossEntropyLoss()
| Natural Language Processing with Disaster Tweets |
18,800,274 | %matplotlib inline
<choose_model_class> | def train_model(net, dataloaders_dict, criterion, optimizer, num_epochs):
max_acc = 0
Stop_flag = False
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
print("使用デバイス：", device)
print('-----start-------')
net.to(device)
torch.backends.cudnn.benchmark = True
batch_size = dataloaders_dict["trai... | Natural Language Processing with Disaster Tweets |
18,800,274 | def build_model(optimizer=Adam()):
model = Sequential()
model.add(Conv2D(64, kernel_size=3, padding='same', activation='relu', input_shape=(28, 28, 1)))
model.add(BatchNormalization())
model.add(Conv2D(64, kernel_size=3, padding='same', activation='relu'))
model.add(BatchNormalization())
model.add(Conv2D(64, kernel_... | num_epochs = 50
net_trained = train_model(net, dataloaders_dict,
criterion, optimizer, num_epochs=num_epochs ) | Natural Language Processing with Disaster Tweets |
18,800,274 | model = build_model(Adam(learning_rate=1e-3))<load_pretrained> | sample_submission = pd.read_csv(".. /input/nlp-getting-started/sample_submission.csv")
sample_submission["target"] = ans_list
sample_submission | Natural Language Processing with Disaster Tweets |
18,800,274 | <load_from_csv><EOS> | sample_submission.to_csv("submission_plus.csv", index=False ) | Natural Language Processing with Disaster Tweets |
18,946,719 | <SOS> metric: meanfscore Kaggle data source: natural-language-processing-with-disaster-tweets<prepare_x_and_y> | import torch
import pandas as pd
import numpy as np
from transformers import AutoModelForSequenceClassification, AutoTokenizer | Natural Language Processing with Disaster Tweets |
18,946,719 | x_test = all_data_test.iloc[:,1:].values
x_test = x_test.reshape(x_test.shape[0], 28, 28, 1 )<predict_on_test> | class Dataset:
def __init__(self, text, tokenizer, max_len):
self.text = text
self.tokenizer = tokenizer
self.max_len = max_len
def __len__(self):
return len(self.text)
def __getitem__(self, item):
text = str(self.text[item])
inputs = self.tokenizer(
text,
max_length=self.max_len,
padding="max_length",
truncation=Tr... | Natural Language Processing with Disaster Tweets |
18,946,719 | predictions = model.predict_classes(x_test/255.)<save_to_csv> | def generate_predictions(model_path, max_len):
model = AutoModelForSequenceClassification.from_pretrained(model_path)
tokenizer = AutoTokenizer.from_pretrained(model_path)
model.to("cuda")
model.eval()
df = pd.read_csv(".. /input/nlp-getting-started/test.csv")
dataset = Dataset(text=df.text.values, tokenizer=tokeni... | Natural Language Processing with Disaster Tweets |
18,946,719 | output = pd.DataFrame({'id': all_data_test.id,
'label': predictions})
output.to_csv("submission.csv",index=False )<import_modules> | preds = generate_predictions("abhishek/autonlp-fred2-2682064", max_len=64 ) | Natural Language Processing with Disaster Tweets |
18,946,719 | print(tf.__version__ )<load_from_csv> | sample = pd.read_csv(".. /input/nlp-getting-started/sample_submission.csv")
sample.target = preds
sample.to_csv("submission.csv", index=False ) | Natural Language Processing with Disaster Tweets |
18,946,719 | <define_variables><EOS> | sample.target.value_counts() | Natural Language Processing with Disaster Tweets |
16,186,337 | <SOS> metric: meanfscore Kaggle data source: natural-language-processing-with-disaster-tweets<split> | import numpy as np
import pandas as pd
import torch
import re
import string
from transformers import BertTokenizer
from transformers import BertModel,BertConfig
from torch.utils.data import Dataset, DataLoader
from sklearn.model_selection import train_test_split
import time
from matplotlib import pyplot as plt | Natural Language Processing with Disaster Tweets |
16,186,337 | y = train["label"]
x = train.drop(["label"], axis = 1)
x_train, x_val, y_train, y_val = sklearn.model_selection.train_test_split(x.values, y.values, test_size = 0.10)
print(y_train.shape)
print(x_train.shape)
print(y_val.shape)
print(x_val.shape )<data_type_conversions> | training_csv =pd.read_csv(".. /input/nlp-getting-started/train.csv")
training_csv.head() | Natural Language Processing with Disaster Tweets |
16,186,337 | print(x_train.shape[0])
x_train = x_train.reshape(x_train.shape[0], img_rows, img_cols, 1)
x_train = x_train.astype('float32')
x_train /= 255
print(x_val.shape[0])
x_val = x_val.reshape(x_val.shape[0], img_rows, img_cols, 1)
x_val = x_val.astype('float32')
x_val /= 255<normalization> | test_csv = pd.read_csv(".. /input/nlp-getting-started/test.csv")
test_csv.head() | Natural Language Processing with Disaster Tweets |
16,186,337 | y_train = tf.keras.utils.to_categorical(y_train, num_class)
y_val = tf.keras.utils.to_categorical(y_val, num_class)
k_init = tf.initializers.TruncatedNormal(mean = 0.1, stddev = 0.05)
b_init = tf.initializers.constant(value = 1e-4)
<create_dataframe> | training_csv['target'].value_counts() | Natural Language Processing with Disaster Tweets |
16,186,337 | datagen = tf.keras.preprocessing.image.ImageDataGenerator(
rotation_range=15,
zoom_range = 0.20,
width_shift_range=0.20,
height_shift_range=0.20
)<choose_model_class> | training_csv['keyword'].value_counts() | Natural Language Processing with Disaster Tweets |
16,186,337 | model = tf.keras.Sequential([
tf.keras.layers.Conv2D(32, kernel_size =(3, 3), activation='relu', input_shape = input_shape),
tf.keras.layers.BatchNormalization() ,
tf.keras.layers.Conv2D(32, kernel_size =(3, 3), activation='relu'),
tf.keras.layers.BatchNormalization() ,
tf.keras.layers.Conv2D(32, kernel_size =(3, 3), a... | training_csv['keyword'].isnull().sum() | Natural Language Processing with Disaster Tweets |
16,186,337 | optimizer = tf.keras.optimizers.RMSprop(learning_rate=0.002,
rho=0.9,
momentum=0.1,
epsilon=1e-07,
centered=True,
name='RMSprop' )<choose_model_class> | training_csv['location'].value_counts() | Natural Language Processing with Disaster Tweets |
16,186,337 | lr = tf.keras.callbacks.ReduceLROnPlateau(monitor='loss',
factor=0.25,
patience=2,
verbose=1,
mode="auto",
min_delta=0.0001,
cooldown=0,
min_lr=0.00001)
es = tf.keras.callbacks.EarlyStopping(monitor='loss',
mode='min', verbose=1,
patience=300,
restore_best_weights=False )<choose_model_class> | training_csv['location'].isnull().sum() | Natural Language Processing with Disaster Tweets |
16,186,337 | model.compile(loss= "categorical_crossentropy", optimizer = optimizer, metrics=['accuracy'])
<train_model> | def clean(title):
title = re.sub(r"\-"," ",title)
title = re.sub(r"\+"," ",title)
title = re.sub(r"&","and",title)
title = re.sub(r"\|"," ",title)
title = re.sub(r"\"," ",title)
title = re.sub(r"\W"," ",title)
title = title.lower()
for p in string.punctuation :
title = re.sub(r"f{p}"," ",title)
title = re.sub(... | Natural Language Processing with Disaster Tweets |
16,186,337 | history = model.fit_generator(datagen.flow(x_train, y_train, batch_size = batch_size),
epochs = epochs,
steps_per_epoch = 100,
validation_data =(x_val, y_val),
validation_steps=50,
callbacks=[lr, es],
verbose=2 )<compute_test_metric> | training_csv["cleaned_text"] = training_csv["text"].map(clean)
training_csv.head() | Natural Language Processing with Disaster Tweets |
16,186,337 | scores = model.evaluate(x_val, y_val )<load_from_csv> | tokenizer = BertTokenizer.from_pretrained('.. /input/bert-base-uncased')
| Natural Language Processing with Disaster Tweets |
16,186,337 | test=pd.read_csv('.. /input/Kannada-MNIST/test.csv')
test_id=test.id
test=test.drop('id',axis=1)
test=test/255
test=test.values.reshape(-1,28,28,1 )<predict_on_test> | X_train,X_test,y_train,y_test = train_test_split(training_csv["cleaned_text"].values,training_csv["target"].values, random_state=0,test_size=0.1,shuffle=True ) | Natural Language Processing with Disaster Tweets |
16,186,337 | y_pre=model.predict(test)
y_pre=np.argmax(y_pre,axis=1 )<save_to_csv> | class CreateDataset(Dataset):
def __init__(self, X, y, tokenizer, max_len):
self.X = X
self.y = y
self.tokenizer = tokenizer
self.max_len = max_len
def __len__(self):
return len(self.y)
def __getitem__(self, index):
text = self.X[index]
inputs = self.tokenizer.encode_plus(
text,
add_special_tokens=True,
max_length=se... | Natural Language Processing with Disaster Tweets |
16,186,337 | sample_sub=pd.read_csv('.. /input/Kannada-MNIST/sample_submission.csv')
sample_sub['label']=y_pre
sample_sub.to_csv('submission.csv',index=False)
sample_sub.head()<set_options> | max_len = 45
dataset_train = CreateDataset(X_train, y_train, tokenizer, max_len)
dataset_valid = CreateDataset(X_test, y_test, tokenizer, max_len)
| Natural Language Processing with Disaster Tweets |
16,186,337 | %reload_ext autoreload
%autoreload 2
%matplotlib inline
path = '/kaggle/input/Kannada-MNIST/'
print(os.listdir(path))
def random_seed(seed_value, use_cuda):
np.random.seed(seed_value)
torch.manual_seed(seed_value)
random.seed(seed_value)
if use_cuda:
torch.cuda.manual_seed(seed_value)
torch.cuda.manual_seed_all(see... | class BERTClass(torch.nn.Module):
def __init__(self, drop_rate, otuput_size):
super().__init__()
model_config = BertConfig.from_pretrained('.. /input/bert-base-uncased', output_hidden_states=True)
self.bert = BertModel.from_pretrained('.. /input/bert-base-uncased', config=model_config)
self.drop = torch.nn.Dropout(dr... | Natural Language Processing with Disaster Tweets |
16,186,337 | class CustomImageList(ImageList):
def open(self, fn):
if(fn.size == 785):
fn = fn[1:]
img = fn.reshape(28,28)
img = np.stack(( img,)*3, axis=-1)
return Image(pil2tensor(img, dtype=np.float32))
@classmethod
def from_csv_custom(cls, path:PathOrStr, csv_name:str, imgIdx:int=1, header:str='infer', **kwargs)->'ItemList':
... | def calculate_loss_and_accuracy(model, criterion, loader, device):
model.eval()
loss = 0.0
total = 0
correct = 0
with torch.no_grad() :
for data in loader:
ids = data['ids'].to(device)
mask = data['mask'].to(device)
labels = data['labels'].to(device)
outputs = model(ids, mask)
loss += criterion(outputs, labels ).... | Natural Language Processing with Disaster Tweets |
16,186,337 | test = CustomImageList.from_csv_custom_test(path=path, csv_name='test.csv', imgIdx=0 )<load_from_csv> | DROP_RATE = 0.4
OUTPUT_SIZE = 1
BATCH_SIZE = 32
NUM_EPOCHS = 2
LEARNING_RATE = 2e-5
model = BERTClass(DROP_RATE, OUTPUT_SIZE)
criterion = torch.nn.BCEWithLogitsLoss()
optimizer = torch.optim.AdamW(params=model.parameters() , lr=LEARNING_RATE)
device = 'cuda' if cuda.is_available() else 'cpu'
log = train_model(dataset... | Natural Language Processing with Disaster Tweets |
16,186,337 | data =(CustomImageList.from_csv_custom(path=path, csv_name='train.csv', imgIdx=1)
.split_by_rand_pct (.02)
.label_from_df(cols='label')
.add_test(test, label=0)
.transform(get_transforms(do_flip=False,max_rotate=15,max_warp=0.4))
.databunch(bs=128, num_workers=0)
.normalize(imagenet_stats))
<concatenate> | test_csv["cleaned_text"] = test_csv["text"].map(clean)
test_csv.head() | Natural Language Processing with Disaster Tweets |
16,186,337 | def conv2(ni,nf,stride=2,ks=5): return conv_layer(ni,nf,stride=stride,ks=ks)
<compute_test_metric> | class TestDataset(Dataset):
def __init__(self, X, tokenizer, max_len):
self.X = X
self.tokenizer = tokenizer
self.max_len = max_len
def __len__(self):
return len(self.X)
def __getitem__(self, index):
text = self.X[index]
inputs = self.tokenizer.encode_plus(
text,
add_special_tokens=True,
max_length=self.max_len,
trun... | Natural Language Processing with Disaster Tweets |
16,186,337 | def mish(input):
return input * torch.tanh(F.softplus(input))<define_search_model> | max_len = 45
dataset_test = TestDataset(test_csv["cleaned_text"].values, tokenizer, max_len ) | Natural Language Processing with Disaster Tweets |
16,186,337 | class BasicBlock(nn.Module):
expansion = 1
def __init__(self, in_planes, planes, stride=1):
super(BasicBlock, self ).__init__()
self.conv1 = nn.Conv2d(
in_planes, planes, kernel_size=3, stride=stride, padding=1, bias=False)
self.bn1 = nn.BatchNorm2d(planes)
self.conv2 = nn.Conv2d(
planes, planes, kernel_size=3, str... | loader = DataLoader(dataset_test, batch_size=len(dataset_test), shuffle=False)
model.eval()
with torch.no_grad() :
for data in loader:
ids = data['ids'].to(device)
mask = data['mask'].to(device)
outputs = model.forward(ids, mask)
pred = torch.round(torch.sigmoid(outputs)).cpu().numpy() | Natural Language Processing with Disaster Tweets |
16,186,337 | model = ResNetCustom()<choose_model_class> | submit_csv =pd.read_csv(".. /input/nlp-getting-started/sample_submission.csv")
submit_csv.head() | Natural Language Processing with Disaster Tweets |
16,186,337 | learn = Learner(data, model, loss_func = nn.CrossEntropyLoss() , metrics=[accuracy] )<train_model> | submit_csv['target'] = pred.astype('int64')
submit_csv.head(10 ) | Natural Language Processing with Disaster Tweets |
16,186,337 | <train_model><EOS> | submit_csv.to_csv("submission2.csv",index = False ) | Natural Language Processing with Disaster Tweets |
18,595,141 | <SOS> metric: meanfscore Kaggle data source: natural-language-processing-with-disaster-tweets<save_to_csv> | !pip install transformers==3.5.1
!pip install pyspellchecker
!pip install -U joblib textblob
!python -m textblob.download_corpora| | Natural Language Processing with Disaster Tweets |
18,595,141 | predictions, *_ = learn.get_preds(DatasetType.Test)
labels = np.argmax(predictions, 1)
submission_df = pd.DataFrame({'id': list(range(0,len(labels))), 'label': labels})
submission_df.to_csv(f'submission.csv', index=False )<choose_model_class> | import pandas as pd
import torchtext
from transformers import BertTokenizer, BertForMaskedLM, BertConfig
import transformers
import torch
from torch.utils.data import Dataset, DataLoader
from torch import optim
from torch import cuda
from sklearn.model_selection import train_test_split
import re
import string
from jobl... | Natural Language Processing with Disaster Tweets |
18,595,141 | interp = ClassificationInterpretation.from_learner(learn )<import_modules> | train_val_df = pd.read_csv("/kaggle/input/nlp-getting-started/train.csv")
test_df = pd.read_csv("/kaggle/input/nlp-getting-started/test.csv" ) | Natural Language Processing with Disaster Tweets |
18,595,141 | warnings.filterwarnings('ignore')
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename))
<load_from_csv> | train_val_df = train_val_df.loc[:,["text","target"]]
test_df = test_df.loc[:,["text"]]
test_df["target"] = [0]*len(test_df["text"] ) | Natural Language Processing with Disaster Tweets |
18,595,141 | raw_train = pd.read_csv('.. /input/Kannada-MNIST/train.csv')
raw_test = pd.read_csv('.. /input/Kannada-MNIST/test.csv' )<prepare_x_and_y> | print(train_val_df)
print(test_df.head())
original_df = train_val_df.copy() | Natural Language Processing with Disaster Tweets |
18,595,141 | x = raw_train.iloc[:, 1:].values.astype('float32')/ 255
y = raw_train.iloc[:, 0]<split> | mispell_dict = {"aren't" : "are not",
"can't" : "cannot",
"couldn't" : "could not",
"couldnt" : "could not",
"didn't" : "did not",
"doesn't" : "does not",
"doesnt" : "does not",
"don't" : "do not",
"hadn't" : "had not",
"hasn't" : "has not",
"haven't" : "have not",
"havent" : "have not",
"he'd" : "he would",
"he'll" : ... | Natural Language Processing with Disaster Tweets |
18,595,141 | x_train, x_val, y_train, y_val = train_test_split(x, y, test_size = 0.2, random_state=42 )<categorify> | print(train_val_df.loc[31])
print(original_df.loc[31] ) | Natural Language Processing with Disaster Tweets |
18,595,141 | x_train = x_train.reshape(-1, 28, 28,1)
x_val = x_val.reshape(-1, 28, 28,1)
y_train = to_categorical(y_train)
y_val = to_categorical(y_val )<choose_model_class> | test_df.to_csv("test.tsv", sep='\t', index=False, header=None)
print(test_df.shape)
train_val_df.to_csv("train_eval.tsv", sep='\t', index=False, header=None)
print(train_val_df.shape ) | Natural Language Processing with Disaster Tweets |
18,595,141 | model = tf.keras.models.Sequential([
tf.keras.layers.Conv2D(64,(3,3), padding='same', input_shape=(28, 28, 1)) ,
tf.keras.layers.BatchNormalization(momentum=0.9, epsilon=1e-5, gamma_initializer="uniform"),
tf.keras.layers.LeakyReLU(alpha=0.1),
tf.keras.layers.Conv2D(64,(3,3), padding='same'),
tf.keras.layers.BatchNorma... | max_length = 50
def tokenizer_50(input_text):
return tokenizer.encode(input_text, max_length=50, return_tensors='pt')[0]
TEXT = torchtext.data.Field(sequential=True, tokenize=tokenizer_50, use_vocab=False, lower=False,
include_lengths=True, batch_first=True, fix_length=max_length, pad_token=0)
LABEL = torchtext.data... | Natural Language Processing with Disaster Tweets |
18,595,141 | optimizer = RMSprop(learning_rate=0.002,
rho=0.9,
momentum=0.1,
epsilon=1e-07,
centered=True,
name='RMSprop')
model.compile(loss='categorical_crossentropy',
optimizer=optimizer,
metrics=['accuracy'] )<define_variables> | tokenizer = BertTokenizer.from_pretrained('bert-base-cased' ) | Natural Language Processing with Disaster Tweets |
18,595,141 | batch_size = 1024
num_classes = 10
epochs = 40<train_model> | dataset_train_eval, dataset_test = torchtext.data.TabularDataset.splits(path='.', train='./train_eval.tsv', test='./test.tsv', format='tsv', fields=[('Text', TEXT),('Label', LABEL)] ) | Natural Language Processing with Disaster Tweets |
18,595,141 | history = model.fit(x_train, y_train, batch_size=batch_size,
epochs=15,
validation_data=(x_val, y_val))<load_from_csv> | dataset_train, dataset_eval = dataset_train_eval.split(
split_ratio=1.0 - 1800/7613, random_state=random.seed(1234))
print(dataset_train.__len__())
print(dataset_eval.__len__())
print(dataset_test.__len__() ) | Natural Language Processing with Disaster Tweets |
18,595,141 | sample_sub=pd.read_csv('.. /input/Kannada-MNIST/sample_submission.csv')
raw_test = pd.read_csv('.. /input/Kannada-MNIST/test.csv')
raw_test_id=raw_test.id
raw_test=raw_test.drop("id",axis="columns")
raw_test=raw_test / 255
test=raw_test.values.reshape(-1,28,28,1)
test.shape<save_to_csv> | print(tokenizer.convert_ids_to_tokens(item.Text.tolist()))
print(int(item.Label)) | Natural Language Processing with Disaster Tweets |
18,595,141 | sub=model.predict(test)
sub=np.argmax(sub,axis=1)
sample_sub['label']=sub
sample_sub.to_csv('submission.csv',index=False )<load_from_csv> | batch_size = 32
dl_train = torchtext.data.Iterator(
dataset_train, batch_size=batch_size, train=True)
dl_eval = torchtext.data.Iterator(
dataset_eval, batch_size=batch_size, train=False, sort=False)
dl_test = torchtext.data.Iterator(
dataset_test, batch_size=batch_size, train=False, sort=False)
dataloaders_dict =... | Natural Language Processing with Disaster Tweets |
18,595,141 | train_csv = pd.read_csv('/kaggle/input/Kannada-MNIST/train.csv')
x_train = np.array(train_csv.iloc[:, 1:])
y_train = to_categorical(train_csv.iloc[:, 0])
x_train = x_train.reshape(x_train.shape[0], 28, 28, 1)
x_train = x_train.astype(np.float32)
x_train /= 255.0
y_train = y_train.astype(np.float32)
print(x_train.... | model = BertModel.from_pretrained('bert-base-cased' ) | Natural Language Processing with Disaster Tweets |
18,595,141 | val_csv = pd.read_csv('/kaggle/input/Kannada-MNIST/Dig-MNIST.csv')
x_val = np.array(val_csv.iloc[:, 1:])
y_val = to_categorical(val_csv.iloc[:, 0])
x_val = x_val.reshape(x_val.shape[0], 28, 28, 1)
x_val = x_val.astype(np.float32)
x_val /= 255.0
y_val = y_val.astype(np.float32)
print(x_val.shape)
print(x_val.dtyp... | class BertForTwitter(nn.Module):
def __init__(self):
super(BertForTwitter, self ).__init__()
self.bert = model
self.cls = nn.Linear(in_features=768, out_features=2)
nn.init.normal_(self.cls.weight, std=0.02)
nn.init.normal_(self.cls.bias, 0)
def forward(self, input_ids):
result = self.bert(input_ids)
vec_0 = resu... | Natural Language Processing with Disaster Tweets |
18,595,141 |
<load_from_csv> | net = BertForTwitter()
net.train()
print('ネットワーク設定完了' ) | Natural Language Processing with Disaster Tweets |
18,595,141 | train = pd.read_csv(".. /input/Kannada-MNIST/train.csv")
test = pd.read_csv(".. /input/Kannada-MNIST/test.csv")
submission = pd.read_csv(".. /input/Kannada-MNIST/sample_submission.csv" )<drop_column> | for param in net.parameters() :
param.requires_grad = False
for param in net.bert.encoder.layer[-1].parameters() :
param.requires_grad = True
for param in net.cls.parameters() :
param.requires_grad = True | Natural Language Processing with Disaster Tweets |
18,595,141 | X = train.drop(['label'], axis = 1)
X_valid = test.drop(['id'], axis = 1 )<compute_test_metric> | optimizer = optim.Adam([
{'params': net.bert.encoder.layer[-1].parameters() , 'lr': 5e-5},
{'params': net.cls.parameters() , 'lr': 1e-4}
])
criterion = nn.CrossEntropyLoss()
| Natural Language Processing with Disaster Tweets |
18,595,141 | def haar(block):
a = pywt.dwt2(block, 'db1')
return a<prepare_x_and_y> | def train_model(net, dataloaders_dict, criterion, optimizer, num_epochs):
max_acc = 0
Stop_flag = False
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
print("使用デバイス：", device)
print('-----start-------')
net.to(device)
torch.backends.cudnn.benchmark = True
batch_size = dataloaders_dict["trai... | Natural Language Processing with Disaster Tweets |
18,595,141 | Y = train['label'].values
X_exp = []
for i in tnrange(train.shape[0]):
im = train.iloc[i][train.columns[1:]].values.reshape(( 28,28))
a,(b,c,d)= haar(im)
newim = np.zeros(( 14,14,4))
newim[:,:,0] = a
newim[:,:,1] = b
newim[:,:,2] = c
newim[:,:,3] = d
X_exp.append(newim)
X = np.array(X_exp)
X_exp = []
for i in tnrang... | num_epochs = 50
net_trained = train_model(net, dataloaders_dict,
criterion, optimizer, num_epochs=num_epochs ) | Natural Language Processing with Disaster Tweets |
18,595,141 | X_train, X_dev, Y_train, Y_dev = train_test_split(X, Y, test_size = 0.2 )<choose_model_class> | sample_submission = pd.read_csv(".. /input/nlp-getting-started/sample_submission.csv")
sample_submission["target"] = ans_list
sample_submission | Natural Language Processing with Disaster Tweets |
18,595,141 | <choose_model_class><EOS> | sample_submission.to_csv("submission_plus.csv", index=False ) | Natural Language Processing with Disaster Tweets |
18,588,011 | <SOS> metric: meanfscore Kaggle data source: natural-language-processing-with-disaster-tweets<choose_model_class> | import pandas as pd
import numpy as np
import spacy
import re
import string | Natural Language Processing with Disaster Tweets |
18,588,011 | lr = tf.keras.callbacks.ReduceLROnPlateau(monitor='accuracy',
patience=300,
verbose=1,
factor=0.5,
min_lr=0.00001 )<normalization> | train = pd.read_csv(".. /input/nlp-getting-started/train.csv")
test = pd.read_csv(".. /input/nlp-getting-started/test.csv")
submission = pd.read_csv(".. /input/nlp-getting-started/sample_submission.csv" ) | Natural Language Processing with Disaster Tweets |
18,588,011 | def lr_decay(epoch, initial_learningrate = 0.001):
return initial_learningrate * 0.99 ** epoch<choose_model_class> | def clean_text(text):
text = str(text ).lower()
text = re.sub('[%s]' % re.escape(string.punctuation), '', text)
text = re.sub('
', '', text)
return text | Natural Language Processing with Disaster Tweets |
18,588,011 | batchsize = 200
epoch = 45
train_datagen = ImageDataGenerator(
rotation_range=10,
width_shift_range=0.25,
height_shift_range=0.25,
shear_range=0.1,
zoom_range=0.25,
horizontal_flip=False)
valid_datagen = ImageDataGenerator(
horizontal_flip=False,
rotation_range=15,
width_shift_range=0.25,
height_shift_range=0.25,
sh... | train["text"] = train["text"].apply(lambda x:clean_text(x))
test["text"] = test["text"].apply(lambda x:clean_text(x)) | Natural Language Processing with Disaster Tweets |
18,588,011 | yhat = model.predict(X_valid ).argmax(axis=1)
submission['label']=pd.Series(yhat)
submission.to_csv('submission.csv',index=False )<save_model> | len(train["keyword"].unique() ) | Natural Language Processing with Disaster Tweets |
18,588,011 | model.save('my_model.h5' )<set_options> | len(train["location"].unique() ) | Natural Language Processing with Disaster Tweets |
18,588,011 | %matplotlib inline<set_options> | nlp = spacy.load('en_core_web_lg' ) | Natural Language Processing with Disaster Tweets |
18,588,011 | config = tf.compat.v1.ConfigProto()
config.gpu_options.allow_growth = True
session = tf.compat.v1.Session(config=config )<load_from_csv> | train_samples = train["text"]
train_samples.head() | Natural Language Processing with Disaster Tweets |
18,588,011 | test = pd.read_csv("/kaggle/input/Kannada-MNIST/test.csv")
train = pd.read_csv("/kaggle/input/Kannada-MNIST/train.csv")
dig_mnist = pd.read_csv("/kaggle/input/Kannada-MNIST/Dig-MNIST.csv")
sample_submission = pd.read_csv("/kaggle/input/Kannada-MNIST/sample_submission.csv" )<data_type_conversions> | def concat_keyword_text(row):
return(str(row["text"])+ " " + str(row["keyword"])) | Natural Language Processing with Disaster Tweets |
18,588,011 | X_train = train.loc[:, train.columns!='label'].values.astype('uint8')
print("X_train.shape", X_train.shape)
y_train = train['label'].values
X_train = X_train.reshape(( X_train.shape[0],28,28))
print("X_train.shape", X_train.shape)
print("y_train.shape",X_train.shape )<split> | train_samples = train.apply(concat_keyword_text, axis = 1)
train_samples.head() | Natural Language Processing with Disaster Tweets |
18,588,011 | X_train = X_train[:,:,:,None]
X_test = X_test[:,:,:,None]<define_variables> | vectorizer = TextVectorization()
text_ds = tf.data.Dataset.from_tensor_slices(train_samples ).batch(128)
vectorizer.adapt(text_ds ) | Natural Language Processing with Disaster Tweets |
18,588,011 | batch_size = 32
num_epochs = 50<count_unique_values> | voc = vectorizer.get_vocabulary() | Natural Language Processing with Disaster Tweets |
18,588,011 | num_samples = X_train.shape[0]
num_classes = np.unique(y_train ).shape[0]
img_rows, img_cols = X_train[0,:,:,0].shape
classes = np.unique(y_train )<define_variables> | num_tokens = len(voc)
embedding_dim = len(nlp('The' ).vector)
embedding_matrix = np.zeros(( num_tokens, embedding_dim)) | Natural Language Processing with Disaster Tweets |
18,588,011 | print("num_samples",num_samples)
print("num_classes",num_classes)
print("img_rows",img_rows)
print("img_cols",img_cols)
print("classes",classes )<categorify> | for i, word in enumerate(voc):
embedding_matrix[i] = nlp(word ).vector | Natural Language Processing with Disaster Tweets |
18,588,011 | y_train = np_utils.to_categorical(y_train, num_classes)
y_train.shape<data_type_conversions> | embedding_layer = Embedding(
num_tokens,
embedding_dim,
embeddings_initializer=keras.initializers.Constant(embedding_matrix),
trainable=False,
) | Natural Language Processing with Disaster Tweets |
18,588,011 | X_train_norm = X_train.astype('float32')
X_test_norm = X_test.astype('float32')
X_train_norm /= 255
X_test_norm /= 255<choose_model_class> | lbl_enc = preprocessing.LabelEncoder()
train["target"] = lbl_enc.fit_transform(train["target"].values ) | Natural Language Processing with Disaster Tweets |
18,588,011 | learning_rate_reduction=ReduceLROnPlateau(monitor='val_loss',
patience=5,
verbose=1,
factor=0.2
)<choose_model_class> | df_train = train.sample(frac=0.7, random_state=0)
df_valid = train.drop(df_train.index ) | Natural Language Processing with Disaster Tweets |
18,588,011 | early_stopping = EarlyStopping(monitor='val_loss',
mode='min',
verbose=1,
patience=10
)<choose_model_class> | X_train = df_train.drop(['target'], axis=1)
X_valid = df_valid.drop(['target'], axis=1)
y_train = df_train['target']
y_valid = df_valid['target'] | Natural Language Processing with Disaster Tweets |
18,588,011 | def build_model() :
model = Sequential()
x_in = layers.Input(shape=(28, 28, 1))
x = layers.Conv2D(64, kernel_size=(3, 3), activation='relu' )(x_in)
x = layers.BatchNormalization()(x)
x = layers.Conv2D(128, kernel_size=(3, 3), activation='relu' )(x)
x = layers.BatchNormalization()(x)
x = layers.MaxPooling2D(pool_siz... | x_train = vectorizer(np.array([[s] for s in X_train["text"]])).numpy()
x_valid = vectorizer(np.array([[s] for s in X_valid["text"]])).numpy() | Natural Language Processing with Disaster Tweets |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.