Datasets:

tweettemposhift
/

tweet_temporal_shift

asahi417 commited on Oct 23, 2023

Commit

1b182d0

1 Parent(s): b9ac9ba

init

Files changed (1) hide show

experiments/model_finetuning_topic.py CHANGED Viewed

@@ -81,7 +81,7 @@ def main(
     tokenized_datasets = dataset.map(
         lambda x: tokenizer(x["text"], padding="max_length", truncation=True, max_length=256), batched=True
     )
-    tokenized_datasets.rename_column("gold_label_list", "label")
     metric_accuracy = evaluate.load("accuracy", "multilabel")
     metric_f1 = evaluate.load("f1", "multilabel")

     tokenized_datasets = dataset.map(
         lambda x: tokenizer(x["text"], padding="max_length", truncation=True, max_length=256), batched=True
     )
+    tokenized_datasets = tokenized_datasets.rename_column("gold_label_list", "label")
     metric_accuracy = evaluate.load("accuracy", "multilabel")
     metric_f1 = evaluate.load("f1", "multilabel")