final

generate.py

@@ -0,0 +1,92 @@
+import pandas as pd
+from transformers import BertTokenizer, AdamW, AutoModelForSequenceClassification
+import torch
+import matplotlib.pyplot as plt
+from torch.utils.data import TensorDataset, DataLoader, RandomSampler
+import torch.nn as nn
+from sklearn.utils.class_weight import compute_class_weight
+import numpy as np
+from sklearn.metrics import classification_report
+from sklearn.metrics import accuracy_score, f1_score
+from transformers import BertTokenizerFast, BertForSequenceClassification
+from transformers import Trainer, TrainingArguments
+import csv
+
+class Dataset(torch.utils.data.Dataset):
+    def __init__(self, encodings, labels):
+        self.encodings = encodings
+        self.labels = labels
+
+    def __getitem__(self, idx):
+        item = {k: torch.tensor(v[idx]) for k, v in self.encodings.items()}
+        item["labels"] = torch.tensor([self.labels[idx]])
+        return item
+
+    def __len__(self):
+        return len(self.labels)
+
+def save_tsv_result(path, data):
+  with open(path, "w") as save:
+    writer = csv.writer(save, delimiter='\t', lineterminator='\n')
+    for value in [str(x) for x in data]:
+      writer.writerow([value])
+
+def predictions_for_set(inputs, masks):
+  predictions = []
+  with torch.no_grad():
+      batch_size = 60
+      for i in range(0, len(inputs), batch_size):
+          preds = model(inputs[i: i + batch_size].to(device),
+                        masks[i: i + batch_size].to(device))
+          preds = preds.logits.detach().cpu().numpy()
+          preds = np.argmax(preds, axis=1)
+          predictions += preds.tolist()
+  return predictions
+
+device = torch.device('cuda')
+
+# train_texts = \
+#     pd.read_csv('train/in.tsv.xz', compression='xz', sep='\t',
+#                 header=None, error_bad_lines=False, quoting=3)[0].tolist()
+# train_labels = pd.read_csv(
+#     'train/expected.tsv', sep='\t', header=None, quoting=3)[0].tolist()
+dev_texts = pd.read_csv('dev-0/in.tsv.xz', compression='xz',
+                        sep='\t', header=None, quoting=3)[0].tolist()
+dev_labels = pd.read_csv('dev-0/expected.tsv', sep='\t',
+                         header=None, quoting=3)[0].tolist()
+test_texts = pd.read_csv('test-A/in.tsv.xz', compression='xz', sep='\t',
+                header=None, error_bad_lines=False, quoting=3)[0].tolist()
+
+model_name = "bert-base-uncased-pretrained"
+
+model = BertForSequenceClassification.from_pretrained(
+    model_name, num_labels=len(pd.unique(dev_labels))).to(device)
+max_length = 512
+tokenizer = BertTokenizerFast.from_pretrained(model_name, do_lower_case=True)
+
+# model.load_pretrained(model_path)
+# tokenizer.load_pretrainded(model_path)
+
+# train_encodings = tokenizer(
+#     train_texts, truncation=True, padding=True, max_length=max_length)
+valid_encodings = tokenizer(
+    dev_texts, truncation=True, padding=True, max_length=max_length)
+test_encodings =  tokenizer(
+    test_texts, truncation=True, padding=True, max_length=max_length)
+
+input_ids_val = torch.tensor(valid_encodings.data['input_ids'])
+attention_mask_val = torch.tensor(valid_encodings.data['attention_mask'])
+
+input_ids_test = torch.tensor(test_encodings.data['input_ids'])
+attention_mask_test = torch.tensor(test_encodings.data['attention_mask'])
+
+predictions = predictions_for_set(input_ids_val, attention_mask_val)
+print("Predictions for dev set:")
+print(classification_report(dev_labels, predictions))
+print(accuracy_score(dev_labels, predictions))
+print(f1_score(dev_labels, predictions))
+
+save_tsv_result("dev-0/out.tsv", predictions)
+
+predictions = predictions_for_set(input_ids_test, attention_mask_test)
+save_tsv_result("test-A/out.tsv", predictions)

main.py

@@ -0,0 +1,90 @@
+import torch
+from transformers.file_utils import is_torch_available
+from transformers import BertTokenizerFast, BertForSequenceClassification
+from transformers import Trainer, TrainingArguments
+import numpy as np
+import random
+from sklearn.metrics import accuracy_score
+import pandas as pd
+
+
+class Dataset(torch.utils.data.Dataset):
+    def __init__(self, encodings, labels):
+        self.encodings = encodings
+        self.labels = labels
+
+    def __getitem__(self, idx):
+        item = {k: torch.tensor(v[idx]) for k, v in self.encodings.items()}
+        item["labels"] = torch.tensor([self.labels[idx]])
+        return item
+
+    def __len__(self):
+        return len(self.labels)
+
+
+def set_seed(seed: int):
+    random.seed(seed)
+    np.random.seed(seed)
+    if is_torch_available():
+        torch.manual_seed(seed)
+        torch.cuda.manual_seed_all(seed)
+
+def compute_metrics(pred):
+    labels = pred.label_ids
+    preds = pred.predictions.argmax(-1)
+    acc = accuracy_score(labels, preds)
+    return {
+        'accuracy': acc,
+    }
+
+set_seed(1)
+
+train_texts = \
+    pd.read_csv('train/in.tsv.xz', compression='xz', sep='\t', header=None, error_bad_lines=False, quoting=3)[0].tolist()[:25000]
+train_labels = pd.read_csv('train/expected.tsv', sep='\t', header=None, quoting=3)[0].tolist()[:25000]
+dev_texts = pd.read_csv('dev-0/in.tsv.xz', compression='xz', sep='\t', header=None, quoting=3)[0].tolist()[:1000]
+dev_labels = pd.read_csv('dev-0/expected.tsv', sep='\t', header=None, quoting=3)[0].tolist()[:1000]
+# test_texts = pd.read_table('test-A/in.tsv.xz', compression='xz', sep='\t', header=None, quoting=3)
+
+model_name = "bert-base-uncased"
+max_length = 25
+tokenizer = BertTokenizerFast.from_pretrained(model_name, do_lower_case=True)
+
+train_encodings = tokenizer(train_texts, truncation=True, padding=True, max_length=max_length)
+valid_encodings = tokenizer(dev_texts, truncation=True, padding=True, max_length=max_length)
+
+train_dataset = Dataset(train_encodings, train_labels)
+valid_dataset = Dataset(valid_encodings, dev_labels)
+
+model = BertForSequenceClassification.from_pretrained(
+    model_name, num_labels=len(pd.unique(train_labels))).to("cuda")
+
+training_args = TrainingArguments(
+    output_dir='./results',          # output directory
+    num_train_epochs=1,              # total number of training epochs
+    per_device_train_batch_size=60,  # batch size per device during training
+    per_device_eval_batch_size=60,   # batch size for evaluation
+    warmup_steps=100,                # number of warmup steps for learning rate scheduler
+    weight_decay=0.01,               # strength of weight decay
+    logging_dir='./logs',            # directory for storing logs
+    load_best_model_at_end=True,     # load the best model when finished training (default metric is loss)
+    # but you can specify `metric_for_best_model` argument to change to accuracy or other metric
+    logging_steps=200,               # log & save weights each logging_steps
+    evaluation_strategy="steps",     # evaluate each `logging_steps`
+)
+
+trainer = Trainer(
+    model=model,                         # the instantiated Transformers model to be trained
+    args=training_args,                  # training arguments, defined above
+    train_dataset=train_dataset,         # training dataset
+    eval_dataset=valid_dataset,          # evaluation dataset
+    compute_metrics=compute_metrics,     # the callback that computes metrics of interest
+)
+
+trainer.train()
+
+trainer.evaluate()
+
+model_path = "bert-base-uncased-pretrained"
+model.save_pretrained(model_path)
+tokenizer.save_pretrained(model_path)