pretrain dataset
Browse files
scripts/prepare_pretrain_dataset.py
CHANGED
|
@@ -85,7 +85,7 @@ def batch_iterator(name=None):
|
|
| 85 |
dataset = load_dataset(name, split=split)
|
| 86 |
|
| 87 |
for row in dataset:
|
| 88 |
-
yield row['question'] + '? ' + row['answer'] + '. ' + row['passage']
|
| 89 |
|
| 90 |
del dataset
|
| 91 |
gc.collect()
|
|
|
|
| 85 |
dataset = load_dataset(name, split=split)
|
| 86 |
|
| 87 |
for row in dataset:
|
| 88 |
+
yield row['question'] + '? ' + str(row['answer']) + '. ' + row['passage']
|
| 89 |
|
| 90 |
del dataset
|
| 91 |
gc.collect()
|