Open Requalvia opened 9 months ago
Hi. I am using code completion task in UniXCoder, and I am wondering if it is possible to contain >=2 masks in one code. For example: `
context = """ def write_json(data,file_path): data = json.<mask0>(data) with open(file_path, 'w') as f: <mask1>.write(data) """ tokens_ids = model.tokenize([context],max_length=512,mode="<encoder-decoder>") source_ids = torch.tensor(tokens_ids).to(device) prediction_ids = model.generate(source_ids, decoder_only=False, beam_size=3, max_length=128) predictions = model.decode(prediction_ids) print([x.replace("<mask0>","").strip() for x in predictions[0]])
` But len(prediction) is 1, so is there a method to get the prediction of \<mask1> ?
Thanks.
Hi. I am using code completion task in UniXCoder, and I am wondering if it is possible to contain >=2 masks in one code. For example: `
` But len(prediction) is 1, so is there a method to get the prediction of \<mask1> ?
Thanks.