Created by: weak-fox

执行python train.py --data-root=ljspeech --use-gpu --preset=../../dv3.single_frame/config.json --train-seq2seq-only --checkpoint=model后报错

Traceback (most recent call last): File "train.py", line 225, in model = make_deepvoice3_from_hparams(hparams) File "train.py", line 110, in make_deepvoice3_from_hparams hparams.key_projection, hparams.value_projection) File "/home/aistudio/work/DeepVoice3/deepvoice3_paddle/builder.py", line 75, in deepvoice3 use_decoder_state_for_postnet_input, "float32") File "/home/aistudio/work/DeepVoice3/deepvoice3_paddle/deepvoice3.py", line 1189, in init window_range, key_projection, value_projection, dropout, dtype) File "/home/aistudio/work/DeepVoice3/deepvoice3_paddle/deepvoice3.py", line 1385, in init dtype=dtype) File "/home/aistudio/work/DeepVoice3/deepvoice3_paddle/deepvoice3.py", line 132, in init dtype=dtype) File "/home/aistudio/work/DeepVoice3/deepvoice3_paddle/modules.py", line 119, in Embedding dtype=dtype) File "/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/dygraph/nn.py", line 1309, in init assert self._is_sparse is True and self._is_distributed is False AssertionError