It is not compatible with the new pytorch about infersent

[soumith]

PR pytorch/pytorch#2266 should fix this.

[aconneau]

Could you print the full error message?

[aconneau]

also, what is your version of pytorch?

[ylmeng]

I did not have any problem with pytorch 0.1.12, but when I used 2.0 I had the error.

Here is the full error message:
`~/InferSent/encoder/models.py in encode(self, sentences, bsize, tokenize, verbose)
177 if self.use_cuda:
178 batch = batch.cuda()
--> 179 batch = self.forward((batch, lengths[stidx:stidx + bsize])).data.cpu().numpy()
180 embeddings.append(batch)
181 embeddings = np.vstack(embeddings)

~/InferSent/encoder/models.py in forward(self, sent_tuple)
44 # Handling padding in Recurrent Networks
45 sent_packed = nn.utils.rnn.pack_padded_sequence(sent, sent_len)
---> 46 sent_output = self.enc_lstm(sent_packed)[0] #seqlen x batch x 2*nhid
47 sent_output = nn.utils.rnn.pad_packed_sequence(sent_output)[0]
48

/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py in call(self, *input, **kwargs)
222 for hook in self._forward_pre_hooks.values():
223 hook(self, input)
--> 224 result = self.forward(*input, **kwargs)
225 for hook in self._forward_hooks.values():
226 hook_result = hook(self, input, result)

/usr/local/lib/python3.5/dist-packages/torch/nn/modules/rnn.py in forward(self, input, hx)
140 hx = (hx, hx)
141
--> 142 has_flat_weights = list(p.data.data_ptr() for p in self.parameters()) == self._data_ptrs
143 if has_flat_weights:
144 first_data = next(self.parameters()).data

/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py in getattr(self, name)
260 return modules[name]
261 raise AttributeError("'{}' object has no attribute '{}'".format(
--> 262 type(self).name, name))
263
264 def setattr(self, name, value):

AttributeError: 'LSTM' object has no attribute '_data_ptrs'
`