Update lm_finetuning.py

lm_finetuning.py

@@ -64,13 +64,11 @@ def get_metrics():
     def compute_metric_search(eval_pred):
         logits, labels = eval_pred
         predictions = np.array([[int(sigmoid(j) > 0.5) for j in i] for i in logits])
-        print(labels.shape, logits.shape, predictions.shape)
         return metric_f1.compute(predictions=predictions, references=labels, average='micro')
 
     def compute_metric_all(eval_pred):
         logits, labels = eval_pred
         predictions = np.array([[int(sigmoid(j) > 0.5) for j in i] for i in logits])
-        print(labels.shape, logits.shape, predictions.shape)
         return {
             'f1': metric_f1.compute(predictions=predictions, references=labels, average='micro')['f1'],
             'f1_macro': metric_f1.compute(predictions=predictions, references=labels, average='macro')['f1'],
@@ -167,7 +165,8 @@ def main():
     model = AutoModelForSequenceClassification.from_pretrained(
         best_model_path,
         num_labels=len(dataset[opt.split_train]['label'][0]),
-        local_files_only=not network)
+        local_files_only=not network,
+        problem_type="multi_label_classification")
     trainer = Trainer(
         model=model,
         args=TrainingArguments(
@@ -179,7 +178,11 @@ def main():
         eval_dataset=tokenized_datasets[opt.split_test],
         compute_metrics=compute_metric_all,
         model_init=lambda x: AutoModelForSequenceClassification.from_pretrained(
-            opt.model, return_dict=True, num_labels=len(dataset[opt.split_train]['label'][0]))
+            opt.model,
+            return_dict=True,
+            num_labels=len(dataset[opt.split_train]['label'][0]),
+            local_files_only=not network,
+            problem_type="multi_label_classification")
     )
     summary_file = pj(opt.output_dir, opt.summary_file)
     if not opt.skip_eval: