changed chunk generation
Browse files- ref_seg_ger.py +3 -0
ref_seg_ger.py
CHANGED
|
@@ -215,6 +215,9 @@ class RefSeg(datasets.GeneratorBasedBuilder):
|
|
| 215 |
for i, row in df.iterrows():
|
| 216 |
|
| 217 |
#tokenized_input = row['token'].split(' ')
|
|
|
|
|
|
|
|
|
|
| 218 |
tokenized_input, offsets = zip(*self.TOKENIZER.pre_tokenize_str(row['token']))
|
| 219 |
tokenized_input = list(tokenized_input)
|
| 220 |
if len(tokenized_input) > 1:
|
|
|
|
| 215 |
for i, row in df.iterrows():
|
| 216 |
|
| 217 |
#tokenized_input = row['token'].split(' ')
|
| 218 |
+
print(row['token'])
|
| 219 |
+
print(type(row['token']))
|
| 220 |
+
print(self.TOKENIZER.pre_tokenize_str(row['token']))
|
| 221 |
tokenized_input, offsets = zip(*self.TOKENIZER.pre_tokenize_str(row['token']))
|
| 222 |
tokenized_input = list(tokenized_input)
|
| 223 |
if len(tokenized_input) > 1:
|