nn - bigrams

.gitignore

@@ -9,4 +9,5 @@
 geval
 *in.tsv
 train_file.txt
-model.arpa
+model.arpa
+model*.bin

run.py

@@ -1,66 +1,144 @@
 #%%
+# importy
+from torchtext.vocab import build_vocab_from_iterator
+from torch.utils.data import DataLoader
+import torch
 import pandas as pd
+import regex as re
 import csv
-import os
+import itertools
-import kenlm
+from os.path import exists
-from collections import Counter, defaultdict
-from math import log10
 
+vocab_size = 30000
+embed_size = 150
 #%%
+# funkcje pomocnicze
 def clean(text):
-    text = str(text).lower().strip().replace("’", "'").replace('\\n', " ").replace("'t", " not").replace("'s", " is").replace("'ll", " will").replace("'m", " am").replace("'ve", " have").replace(",", "").replace("-", "").replace(".", "").replace("'", "".replace("”", "").replace(">", ""))
+    text = str(text).strip().lower()
+    text = re.sub("’|>|<|\.|\\|\"|”|-|,|\*|:|\/", "", text)
+    text = text.replace('\\n', " ").replace("'t", " not").replace("'s", " is").replace("'ll", " will").replace("'m", " am").replace("'ve", " have")
+    text = text.replace("'", "")
     return text
 
+def get_words_from_line(line):
+    line = line.rstrip()
+    yield '<s>'
+    for m in re.finditer(r'[\p{L}0-9\*]+|\p{P}+', line):
+        yield m.group(0).lower()
+    yield '</s>'
+
+
+def get_word_lines_from_data(d):
+    for line in d:
+        yield get_words_from_line(line)
+
+#%%
+class Model(torch.nn.Module):
+    def __init__(self, vocabulary_size, embedding_size):
+        super(Model, self).__init__()
+        self.model = torch.nn.Sequential(
+            torch.nn.Embedding(vocabulary_size, embedding_size),
+            torch.nn.Linear(embedding_size, vocabulary_size),
+            torch.nn.Softmax()
+        )
+
+    def forward(self, x):
+        return self.model(x)
+
+#%%
+class Trigrams(torch.utils.data.IterableDataset):
+    def __init__(self, data, vocabulary_size):
+        self.vocab = build_vocab_from_iterator(
+           get_word_lines_from_data(data),
+           max_tokens = vocabulary_size,
+           specials = ['<unk>'])
+        self.vocab.set_default_index(self.vocab['<unk>'])
+        self.vocabulary_size = vocabulary_size
+        self.data = data
+
+    @staticmethod
+    def look_ahead_iterator(gen):
+        w1 = None
+        for item in gen:
+            if w1 is not None:
+                yield (w1, item)
+            w1 = item
+
+    def __iter__(self):
+       return self.look_ahead_iterator(
+           (self.vocab[t] for t in itertools.chain.from_iterable(get_word_lines_from_data(self.data))))
+
+
+#%%
+# ładowanie danych treningowych
 train_in = pd.read_csv("train/in.tsv.xz", sep='\t', header=None, encoding="UTF-8", on_bad_lines="skip", quoting=csv.QUOTE_NONE, nrows=300000)[[6, 7]]
 train_expected = pd.read_csv("train/expected.tsv", sep='\t', header=None, encoding="UTF-8", on_bad_lines="skip", quoting=csv.QUOTE_NONE, nrows=300000)
-data = pd.concat([train_in, train_expected], axis=1)
+train_data = pd.concat([train_in, train_expected], axis=1)
-data = data[6] + data[0] + data[7]
+train_data = train_data[6] + train_data[0] + train_data[7]
-data = data.apply(clean)
+train_data = train_data.apply(clean)
-
+train_dataset = Trigrams(train_data, vocab_size)
-if not os.path.isfile('train_file.txt'):
-    with open("train_file.txt", "w+") as f:
-        for text in data:
-            f.write(text + "\n")
 
 #%%
-#get_ipython().system('../kenlm/build/bin/lmplz -o 4 < train_file.txt > model.arpa --skip_symbols')
+# trenowanie/wczytywanie modelu
-model = kenlm.Model("model.arpa")
+device = 'cuda' if torch.cuda.is_available() else 'cpu'
+model = Model(vocab_size, embed_size).to(device)
+if(not exists('model1.bin')):
+    data = DataLoader(train_dataset, batch_size=200)
+    optimizer = torch.optim.Adam(model.parameters())
+    criterion = torch.nn.NLLLoss()
+
+    model.train()
+    step = 0
+    for x, y in data:
+        x = x.to(device)
+        y = y.to(device)
+        optimizer.zero_grad()
+        ypredicted = model(x)
+        loss = criterion(torch.log(ypredicted), y)
+        if step % 100 == 0:
+            print(step, loss)
+        step += 1
+        loss.backward()
+        optimizer.step()
+
+    torch.save(model.state_dict(), 'model1.bin')
+else:
+    model.load_state_dict(torch.load('model1.bin'))
+
 #%%
-import nltk
+vocab = train_dataset.vocab
+
+def predict(tokens):
+    ixs = torch.tensor(vocab.forward(tokens)).to(device)
+    out = model(ixs)
+    top = torch.topk(out[0], 10)
+    top_indices = top.indices.tolist()
+    top_probs = top.values.tolist()
+    top_words = vocab.lookup_tokens(top_indices)
+    result = ""
+    for word, prob in list(zip(top_words, top_probs)):
+                result += f"{word}:{prob} "
+    result  += f':0.01'
+    return result
+
 from nltk import word_tokenize
-nltk.download('punkt')
+def predict_file(result_path, data):
-
-most_common = defaultdict(lambda: 0)
-for text in data:
-    words = word_tokenize(text)
-    if "d" in words:
-        words.remove("d")
-    for w in words:
-        most_common[w] += 1
-most_common = Counter(most_common).most_common(8000)
-
-#%%
-def predict(path, result_path):
-    data = pd.read_csv(path, sep='\t', header=None, encoding="UTF-8", on_bad_lines="skip", quoting=csv.QUOTE_NONE)
     with open(result_path, "w+", encoding="UTF-8") as f:
-        for i, row in data.iterrows():
+        for row in data:
             result = {}
-            before = word_tokenize(clean(str(row[6])))[-3:]
+            before = word_tokenize(clean(str(row)))[-1:]
-            if(len(before) < 2):
+            if(len(before) < 1):
                 result = "a:0.2 the:0.2 to:0.2 of:0.1 and:0.1 of:0.1 :0.1"
             else:
-                for w in most_common:
+                result = predict(before)
-                    word = w[0]
-                    prob = model.score(" ".join(before + [word]))
-                    result[word] = prob
-                predictions = dict(Counter(result).most_common(12))
-                result = ""
-                for word, prob in predictions.items():
-                    result += f"{word}:{prob} "
-            result  += f':{log10(0.99)}'
             f.write(result + "\n")
             print(result)
 
+#%%
+dev_data = pd.read_csv("dev-0/in.tsv.xz", sep='\t', header=None, quoting=csv.QUOTE_NONE)[6]
+dev_data = dev_data.apply(clean)
+predict_file("dev-0/out.tsv", dev_data)
 
-predict("dev-0/in.tsv.xz", "dev-0/out.tsv")
+test_data = pd.read_csv("test-A/in.tsv.xz", sep='\t', header=None, quoting=csv.QUOTE_NONE)[6]
-predict("test-A/in.tsv.xz", "test-A/out.tsv")
+test_data = test_data.apply(clean)
+predict_file("test-A/out.tsv", test_data)