sergioburdisso commited on
Commit
df5a112
1 Parent(s): 3f1f61a

Update dialog2flow-dataset.py

Browse files
Files changed (1) hide show
  1. dialog2flow-dataset.py +18 -15
dialog2flow-dataset.py CHANGED
@@ -7,6 +7,7 @@
7
 
8
  #!/usr/bin/env python3
9
  import os
 
10
  import json
11
  import datasets
12
  from datasets import (GeneratorBasedBuilder,
@@ -35,7 +36,7 @@ _CITATION = """
35
  }
36
  """
37
 
38
- DATASETS_PRETRAIN = ["dialog-acts", "slots", "dialog-actions"]
39
  DATASETS_DS = {
40
  'ABCD': ['test', 'train', 'val'],
41
  'BiTOD': ['test', 'train', 'val'],
@@ -56,7 +57,7 @@ DATASETS_DS = {
56
  'Taskmaster2': ['train'],
57
  'Taskmaster3': ['test', 'train', 'val'],
58
  'WOZ2_0': ['test', 'train', 'val'],
59
- # 'SimJointGEN': ['test', 'train', 'val'],
60
  }
61
  DATASETS = list(DATASETS_DS.keys()) + DATASETS_PRETRAIN
62
  SPLIT2NAME = {
@@ -101,7 +102,7 @@ class Dialog2FlowBuilder(GeneratorBasedBuilder):
101
  url="https://github.com/idiap/dialog2flow",
102
  ))
103
 
104
- DEFAULT_CONFIG_NAME = "dialog-actions"
105
 
106
  def _info(self):
107
  if self.config.name in DATASETS_PRETRAIN:
@@ -134,11 +135,10 @@ class Dialog2FlowBuilder(GeneratorBasedBuilder):
134
 
135
  def _split_generators(self, dl_manager):
136
  if self.config.name in DATASETS_PRETRAIN:
137
- # TODO
138
  file_path = dl_manager.download({
139
- "train": "train.csv", # full
140
- "val": "eval.csv", # few shot subset
141
- "test": "test.csv", # SpokenWOZ
142
  })
143
  splits = [
144
  SplitGenerator(
@@ -155,13 +155,13 @@ class Dialog2FlowBuilder(GeneratorBasedBuilder):
155
  "split": datasets.Split.VALIDATION,
156
  },
157
  ),
158
- SplitGenerator(
159
- name=datasets.Split.TEST,
160
- gen_kwargs={
161
- "file_path": file_path["test"],
162
- "split": datasets.Split.TEST,
163
- },
164
- )
165
  ]
166
  else:
167
  splits = []
@@ -195,4 +195,7 @@ class Dialog2FlowBuilder(GeneratorBasedBuilder):
195
  for dial_id, dial in data:
196
  yield dial_id, {"dialog": dial}
197
  else:
198
- pass
 
 
 
 
7
 
8
  #!/usr/bin/env python3
9
  import os
10
+ import csv
11
  import json
12
  import datasets
13
  from datasets import (GeneratorBasedBuilder,
 
36
  }
37
  """
38
 
39
+ DATASETS_PRETRAIN = ["dialog-acts", "slots", "dialog-acts+slots"]
40
  DATASETS_DS = {
41
  'ABCD': ['test', 'train', 'val'],
42
  'BiTOD': ['test', 'train', 'val'],
 
57
  'Taskmaster2': ['train'],
58
  'Taskmaster3': ['test', 'train', 'val'],
59
  'WOZ2_0': ['test', 'train', 'val'],
60
+ 'SimJointGEN': ['test', 'train', 'val'],
61
  }
62
  DATASETS = list(DATASETS_DS.keys()) + DATASETS_PRETRAIN
63
  SPLIT2NAME = {
 
102
  url="https://github.com/idiap/dialog2flow",
103
  ))
104
 
105
+ DEFAULT_CONFIG_NAME = "dialog-acts+slots"
106
 
107
  def _info(self):
108
  if self.config.name in DATASETS_PRETRAIN:
 
135
 
136
  def _split_generators(self, dl_manager):
137
  if self.config.name in DATASETS_PRETRAIN:
 
138
  file_path = dl_manager.download({
139
+ "train": f"{self.config.name}.csv", # full
140
+ "val": "val.csv", # few shot subset
141
+ # "test": "test.csv", # TODO: use SpokenWOZ
142
  })
143
  splits = [
144
  SplitGenerator(
 
155
  "split": datasets.Split.VALIDATION,
156
  },
157
  ),
158
+ # SplitGenerator( # TODO: use SpokenWOZ
159
+ # name=datasets.Split.TEST,
160
+ # gen_kwargs={
161
+ # "file_path": file_path["test"],
162
+ # "split": datasets.Split.TEST,
163
+ # },
164
+ # )
165
  ]
166
  else:
167
  splits = []
 
195
  for dial_id, dial in data:
196
  yield dial_id, {"dialog": dial}
197
  else:
198
+ with open(file_path, newline='') as csvfile:
199
+ csvreader = csv.reader(csvfile)
200
+ for ix, row in enumerate(csvreader):
201
+ yield ix, {"utterance": row[0], "label": row[1]}