init

experiments/model_predict_ner.py

@@ -96,7 +96,6 @@ class NER:
                 })
                 prob = torch.softmax(output['logits'], dim=-1).cpu().detach().float().tolist()
                 print(prob)
-                print(prob.shape)
                 pred = torch.max(prob, dim=-1)[1].cpu().detach().int().tolist()
                 preds += [[self.id_to_label[_p] for _p in p] for p in pred]
         return [{"prediction": p, "label": i} for p, i in zip(preds, labels)]