Spaces:
Paused
Paused
Update train.py
Browse files
train.py
CHANGED
|
@@ -18,7 +18,7 @@ OUTPUT_REPO = "smallama"
|
|
| 18 |
PUSH_TO_HUB = True
|
| 19 |
|
| 20 |
def load_data():
|
| 21 |
-
dataset = load_dataset(INPUT_DATASET, split="train").select(range(int(2e+4)))
|
| 22 |
return dataset
|
| 23 |
|
| 24 |
def create_tokenizer(training_corpus):
|
|
|
|
| 18 |
PUSH_TO_HUB = True
|
| 19 |
|
| 20 |
def load_data():
|
| 21 |
+
dataset = load_dataset(INPUT_DATASET, split="train").select(range(int(2e+4)))
|
| 22 |
return dataset
|
| 23 |
|
| 24 |
def create_tokenizer(training_corpus):
|