initial commit

sroie2019.py

@@ -0,0 +1,89 @@
+import os
+import gin
+import torch
+import datasets
+
+from pathlib import Path
+from torch.utils.data import Dataset
+
+from datasets import load_dataset, Features, Value, ClassLabel, DownloadConfig
+
+_DESCRIPTION = """\
+"""
+
+_CITATION = """\
+"""
+
+# it could be file or url path
+_TRAIN_DOWNLOAD_URL = "train.txt"
+_VAL_DOWNLOAD_URL = "val.txt"
+
+CLASS_NAMES = ["company", "date", "address", "total", "O"]
+
+
+class CustomTokenDataset(datasets.GeneratorBasedBuilder):
+    """CustomTokenDataset dataset."""
+
+    def _info(self):
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=datasets.Features(
+                {
+                    "id": datasets.Value("string"),
+                    "tokens": datasets.Sequence(datasets.Value("string")),
+                    "ner_tags": datasets.Sequence(
+                        datasets.features.ClassLabel(names=sorted(list(CLASS_NAMES)))
+                    ),
+                }
+            ),
+            supervised_keys=None,
+            homepage="",
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        """Returns SplitGenerators."""
+        urls_to_download = {
+            "train": _TRAIN_DOWNLOAD_URL,
+            "val": _VAL_DOWNLOAD_URL,
+        }
+        downloaded_files = dl_manager.download_and_extract(urls_to_download)
+
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={"filepath": downloaded_files["train"]},
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.VALIDATION,
+                gen_kwargs={"filepath": downloaded_files["val"]},
+            ),
+        ]
+
+    def _generate_examples(self, filepath):
+        with open(filepath, encoding="utf-8") as f:
+            guid = 0
+            tokens = []
+            ner_tags = []
+            for line in f:
+                if line == "" or line == "\n":
+                    if tokens:
+                        yield guid, {
+                            "id": str(guid),
+                            "tokens": tokens,
+                            "ner_tags": ner_tags,
+                        }
+                        guid += 1
+                        tokens = []
+                        ner_tags = []
+                else:
+                    # CustomDataset tokens are space separated
+                    splits = line.split(" ")
+                    tokens.append(splits[0])
+                    ner_tags.append(splits[1].rstrip())
+            # last example
+            yield guid, {
+                "id": str(guid),
+                "tokens": tokens,
+                "ner_tags": ner_tags,
+            }

train.txt

@@ -0,0 +1,294 @@
+TAN O
+WOON O
+YANN O
+BOOK company
+TA company
+.K(TAMAN company
+DAYA) company
+SDN company
+BND company
+789417-W O
+NO.53 address
+55 address
+57 address
+& address
+59 address
+ address
+JALAN address
+SAGU address
+18 address
+TAMAN address
+DAYA address
+81100 address
+JOHOR address
+BAHRU address
+JOHOR. address
+DOCUMENT O
+NO O
+: O
+TD01167104 O
+DATE: O
+25/12/2018 date
+8:13:39 date
+PM date
+CASHIER: O
+MANIS O
+MEMBER: O
+CASH O
+BILL O
+CODE/DESC O
+PRICE O
+DISC O
+AMOUNT O
+QTY O
+RM O
+9556939040116 O
+KF O
+MODELLING O
+CLAY O
+KIDDY O
+FISH O
+1 O
+PC O
+* O
+9.000 total
+0.00 O
+9.00 total
+TOTAL: O
+ROUR O
+DING O
+ADJUSTMENT: O
+ROUND O
+D O
+TOTAL O
+(RM): O
+CASH O
+10.00 O
+CHANGE O
+1.00 O
+GOODS O
+SOLD O
+ARE O
+NOT O
+RETURNABLE O
+OR O
+EXCHANGEABLE O
+*** O
+THANK O
+YOU O
+PLEASE O
+COME O
+AGAIN O
+! O
+
+TAN O
+WOON O
+YANN O
+INDAH company
+GIFT company
+& company
+HOME company
+DECO company
+27 address
+JALAN address
+DEDAP address
+13 address
+TAMAN address
+JOHOR address
+JAYA address
+81100 address
+JOHOR address
+BAHRU address
+JOHOR. address
+TEL:07-3507405 O
+FAX:07-3558160 O
+RECEIPT O
+19/10/2018 date
+20:49:59 date
+#01 date
+CASHIER: O
+CN O
+LOCATION/SP: O
+05 O
+/0531 O
+MB: O
+MO26588 O
+ROOM O
+NO: O
+01 O
+050100035279 O
+DESC/ITEM O
+QTY O
+PRICE O
+AMT(RM) O
+ST-PRIVILEGE O
+CARD/GD O
+INDAH O
+88888 O
+1 O
+10.00 O
+GF-TABLE O
+LAMP/STITCH O
+<I> O
+62483 O
+55.90 O
+@DISC O
+10.00% O
+-5.59 O
+#TOTAL O
+QTY O
+2 O
+TOTAL O
+AMT................. O
+RM O
+60.31 O
+ROUNDING O
+ADJ............ O
+-0.01 O
+RM O
+60.30 total
+CASH.................... O
+RM O
+70.30 O
+CHANGE.................. O
+RM O
+THANK O
+YOU O
+! O
+PLEASE O
+COME O
+AGAIN O
+! O
+GOODS O
+SOLD O
+ARE O
+NOT O
+RETURNABLE O
+THANK O
+YOU O
+! O
+FLEASE O
+COME O
+AOSIN O
+! O
+DEALING O
+IN O
+WHOLESALE O
+AND O
+RETAIL. O
+
+TAN O
+WOON O
+YANN O
+MR company
+D.T.Y. company
+(JOHOR) company
+SDN company
+BHD company
+(CO.REG O
+: O
+933109-X) O
+LOT address
+1851-A address
+& address
+1851-B address
+ address
+JALAN address
+KPB address
+6 address
+KAWASAN address
+PERINDUSTRIAN address
+BALAKONG address
+43300 address
+SERI address
+KEMBANGAN address
+ address
+SELANGOR address
+(MR address
+DIY address
+TESCO address
+TERBAU) address
+-INVOICE- O
+CHOPPING O
+BOARD O
+35.5X25.5CM O
+803M# O
+EZ10HD05 O
+- O
+24 O
+8970669 O
+1 O
+X O
+19.00 O
+AIR O
+PRESSURE O
+SPRAYER O
+SX-575-1 O
+1.5L O
+HC03-7 O
+- O
+15 O
+9066468 O
+8.02 O
+WAXCO O
+WINDSHILED O
+CLEANER O
+120ML O
+WA14-3A O
+- O
+48 O
+9557031100236 O
+3.02 O
+BOPP O
+TAPE O
+48MM*100M O
+CLEAR O
+FZ-04 O
+- O
+36 O
+6935818350846 O
+3.88 O
+ITEM(S) O
+: O
+4 O
+QTY(S) O
+: O
+4 O
+TOTAL O
+RM O
+33.92 O
+ROUNDING O
+ADJUSTMENT O
+-RM O
+0.02 O
+TOTAL O
+ROUNDED O
+RM total
+33.90 total
+CASH O
+RM O
+50.00 O
+CHANGE O
+RM O
+16.10 O
+12-01-19 date
+21:13 date
+SH01 date
+ZK09 date
+T4 O
+R000027830 O
+OPERATOR O
+TRAINEE O
+CASHIER O
+EXCHANGE O
+ARE O
+ALLOWED O
+WITHIN O
+7 O
+DAYS O
+WITH O
+RECEIPT. O
+STRICTLY O
+NO O
+CASH O
+REFUND. O

val.txt

@@ -0,0 +1,86 @@
+RESTORAN company
+WAN company
+SHENG company
+002043319-W O
+NO.2 address
+ address
+JALAN address
+TEMENGGUNG address
+19/9 address
+SEKSYEN address
+9 address
+ address
+BANDAR address
+MAHKOTA address
+CHERAS address
+43200 address
+CHERAS address
+ address
+SELANGOR address
+GST O
+REG O
+NO: O
+001335787520 O
+TAX O
+INVOICE O
+INV O
+NO. O
+: O
+1085405 O
+CASHIER: O
+THANDAR O
+DATE O
+: date
+09-04-2018 date
+13:16:21 date
+DESCRIPTION O
+QTY O
+U.PRICE O
+TOTAL O
+TAX O
+TEH O
+(B) O
+1 O
+X O
+2.20 O
+SR O
+CHAM O
+(B) O
+HERBAL O
+TEA O
+2 O
+X O
+1.70 O
+3.40 O
+TAKE O
+AWAY O
+4 O
+X O
+0.20 O
+0.80 O
+TOTAL O
+QTY: O
+8 O
+TOTAL O
+(EXCLUDING O
+GST): O
+8.11 O
+GST O
+PAYABLE O
+(6%): O
+0.49 O
+TOTAL O
+(INCLUSIVE O
+OF O
+GST): O
+8.60 total
+TOTAL O
+: O
+CASH O
+: O
+GST O
+SUMMARY O
+AMOUNT(RM) O
+TAX(RM) O
+(@ O
+6%) O