test
Browse files
parallel_train.jsonl → parallel_train.json
RENAMED
|
File without changes
|
run_translation_t5_flax.py
CHANGED
|
@@ -215,17 +215,7 @@ class DataTrainingArguments:
|
|
| 215 |
|
| 216 |
|
| 217 |
summarization_name_mapping = {
|
| 218 |
-
"
|
| 219 |
-
"big_patent": ("description", "abstract"),
|
| 220 |
-
"cnn_dailymail": ("article", "highlights"),
|
| 221 |
-
"orange_sum": ("text", "summary"),
|
| 222 |
-
"pn_summary": ("article", "summary"),
|
| 223 |
-
"psc": ("extract_text", "summary_text"),
|
| 224 |
-
"samsum": ("dialogue", "summary"),
|
| 225 |
-
"thaisum": ("body", "summary"),
|
| 226 |
-
"xglue": ("news_body", "news_title"),
|
| 227 |
-
"xsum": ("document", "summary"),
|
| 228 |
-
"wiki_summary": ("article", "highlights"),
|
| 229 |
}
|
| 230 |
|
| 231 |
|
|
@@ -407,7 +397,7 @@ def main():
|
|
| 407 |
return
|
| 408 |
|
| 409 |
# Get the column names for input/target.
|
| 410 |
-
dataset_columns = summarization_name_mapping.get(
|
| 411 |
if data_args.text_column is None:
|
| 412 |
text_column = dataset_columns[0] if dataset_columns is not None else column_names[0]
|
| 413 |
else:
|
|
|
|
| 215 |
|
| 216 |
|
| 217 |
summarization_name_mapping = {
|
| 218 |
+
"translation": ("nb", "nn"),
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 219 |
}
|
| 220 |
|
| 221 |
|
|
|
|
| 397 |
return
|
| 398 |
|
| 399 |
# Get the column names for input/target.
|
| 400 |
+
dataset_columns = summarization_name_mapping.get("translation", None)
|
| 401 |
if data_args.text_column is None:
|
| 402 |
text_column = dataset_columns[0] if dataset_columns is not None else column_names[0]
|
| 403 |
else:
|