Commit
•
5c5b425
1
Parent(s):
ffd58f5
quick fixes
Browse files- massive.py +2 -33
massive.py
CHANGED
|
@@ -174,29 +174,8 @@ class MASSIVE(datasets.GeneratorBasedBuilder):
|
|
| 174 |
features=datasets.Features(
|
| 175 |
{
|
| 176 |
"id": datasets.Value("string"),
|
| 177 |
-
"
|
| 178 |
-
"
|
| 179 |
-
"scenario": datasets.features.ClassLabel(names=_SCENARIOS),
|
| 180 |
-
"intent": datasets.features.ClassLabel(names=_INTENTS),
|
| 181 |
-
"utt": datasets.Value("string"),
|
| 182 |
-
"annot_utt": datasets.Value("string"),
|
| 183 |
-
"worker_id": datasets.Value("string"),
|
| 184 |
-
"slot_method": datasets.Sequence(
|
| 185 |
-
{
|
| 186 |
-
"slot": datasets.Value("string"),
|
| 187 |
-
"method": datasets.Value("string"),
|
| 188 |
-
}
|
| 189 |
-
),
|
| 190 |
-
"judgments": datasets.Sequence(
|
| 191 |
-
{
|
| 192 |
-
"worker_id": datasets.Value("string"),
|
| 193 |
-
"intent_score": datasets.Value("int8"),
|
| 194 |
-
"slots_score": datasets.Value("int8"),
|
| 195 |
-
"grammar_score": datasets.Value("int8"),
|
| 196 |
-
"spelling_score": datasets.Value("int8"),
|
| 197 |
-
"language_identification": datasets.Value("string"),
|
| 198 |
-
}
|
| 199 |
-
),
|
| 200 |
},
|
| 201 |
),
|
| 202 |
supervised_keys=None,
|
|
@@ -239,29 +218,19 @@ class MASSIVE(datasets.GeneratorBasedBuilder):
|
|
| 239 |
]
|
| 240 |
|
| 241 |
def _generate_examples(self, files, split, lang):
|
| 242 |
-
|
| 243 |
filepath = "1.0/data/" + lang + ".jsonl"
|
| 244 |
-
|
| 245 |
logger.info("⏳ Generating examples from = %s", filepath)
|
| 246 |
-
|
| 247 |
for path, f in files:
|
| 248 |
if path == filepath:
|
| 249 |
-
# Read the file
|
| 250 |
lines = f.readlines()
|
| 251 |
-
|
| 252 |
key_ = 0
|
| 253 |
-
|
| 254 |
for line in lines:
|
| 255 |
-
|
| 256 |
data = json.loads(line)
|
| 257 |
-
|
| 258 |
if data["partition"] != split:
|
| 259 |
continue
|
| 260 |
-
|
| 261 |
yield key_, {
|
| 262 |
"id": data["id"],
|
| 263 |
"label": data["scenario"],
|
| 264 |
"text": data["utt"],
|
| 265 |
}
|
| 266 |
-
|
| 267 |
key_ += 1
|
|
|
|
| 174 |
features=datasets.Features(
|
| 175 |
{
|
| 176 |
"id": datasets.Value("string"),
|
| 177 |
+
"label": datasets.features.ClassLabel(names=_SCENARIOS),
|
| 178 |
+
"text": datasets.Value("string"),
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 179 |
},
|
| 180 |
),
|
| 181 |
supervised_keys=None,
|
|
|
|
| 218 |
]
|
| 219 |
|
| 220 |
def _generate_examples(self, files, split, lang):
|
|
|
|
| 221 |
filepath = "1.0/data/" + lang + ".jsonl"
|
|
|
|
| 222 |
logger.info("⏳ Generating examples from = %s", filepath)
|
|
|
|
| 223 |
for path, f in files:
|
| 224 |
if path == filepath:
|
|
|
|
| 225 |
lines = f.readlines()
|
|
|
|
| 226 |
key_ = 0
|
|
|
|
| 227 |
for line in lines:
|
|
|
|
| 228 |
data = json.loads(line)
|
|
|
|
| 229 |
if data["partition"] != split:
|
| 230 |
continue
|
|
|
|
| 231 |
yield key_, {
|
| 232 |
"id": data["id"],
|
| 233 |
"label": data["scenario"],
|
| 234 |
"text": data["utt"],
|
| 235 |
}
|
|
|
|
| 236 |
key_ += 1
|