Spaces:
Sleeping
Sleeping
Varun Wadhwa
commited on
Logs
Browse files
app.py
CHANGED
@@ -86,10 +86,11 @@ def align_labels_with_tokens(label, word_ids):
|
|
86 |
aligned_label_ids.append(-100)
|
87 |
elif word_idx != previous_word_idx: # Only label the first token of a given word.
|
88 |
if label[word_idx].startswith("B-"):
|
89 |
-
print(
|
90 |
-
print(label[word_idx])
|
91 |
label[word_idx] = label[word_idx].replace("B-", "I-")
|
92 |
-
|
|
|
93 |
else:
|
94 |
aligned_label_ids.append(-100)
|
95 |
previous_word_idx = word_idx
|
@@ -112,7 +113,7 @@ def tokenize_function(examples):
|
|
112 |
print("Printing partial input with tokenized output")
|
113 |
print(inputs.tokens()[:1000])
|
114 |
print(inputs.word_ids()[:1000])
|
115 |
-
print(new_labels[
|
116 |
inputs["labels"] = new_labels
|
117 |
return inputs
|
118 |
|
|
|
86 |
aligned_label_ids.append(-100)
|
87 |
elif word_idx != previous_word_idx: # Only label the first token of a given word.
|
88 |
if label[word_idx].startswith("B-"):
|
89 |
+
print("B vs I")
|
90 |
+
print(label2id[label[word_idx]])
|
91 |
label[word_idx] = label[word_idx].replace("B-", "I-")
|
92 |
+
print(label2id[label[word_idx]])
|
93 |
+
aligned_label_ids.append(label2id[label[word_idx]])
|
94 |
else:
|
95 |
aligned_label_ids.append(-100)
|
96 |
previous_word_idx = word_idx
|
|
|
113 |
print("Printing partial input with tokenized output")
|
114 |
print(inputs.tokens()[:1000])
|
115 |
print(inputs.word_ids()[:1000])
|
116 |
+
print(new_labels[0])
|
117 |
inputs["labels"] = new_labels
|
118 |
return inputs
|
119 |
|