ELMoForManyLangs
ELMoForManyLangs copied to clipboard
HIT-SCIR
Pre-trained ELMo Representations for Many Languages
Is it possible to use tensorflow's [hub.Module](https://www.tensorflow.org/hub/api_docs/python/hub/Module) to load your pretrained models programmatically instead of using your [embedder ](https://github.com/HIT-SCIR/ELMoForManyLangs#use-elmoformanylangs-programmatically)
Can you please inform me what do you mean by the parameter 'n_samples' in the configuration file? If you can provide a description of the parameters of the configuration file...
hi，我尝试了使用多GPU训练，在model和optimizer加入了 model = nn.DataParallel(model, device_ids=gpu_ids) optimizer = nn.DataParallel(optimizer, device_ids=gpu_ids) 但是遇到了 terminate called after throwing an instance of 'std::runtime_error' what(): cuda runtime error (59) : device-side assert triggered at /pytorch/aten/src/THC/generic/THCStorage.c:184 错误...
at line 111 of biLM.py bow_id, eow_id, oov_id, pad_id = char2id.get('eow', None), char2id.get('bow', None), char2id.get(oov, None), char2id.get(pad, None) should change to bow_id, eow_id, oov_id, pad_id =char2id.get('bow', None), char2id.get('eow', None), char2id.get(oov,...
Thanks for this work! Could you please make available the weights and options file (in .hdf5 format), like how the allennlp pre-trained model works?
Hi, I can't get how to use elmo embedding instead of word2vec in code. Can you guide me or link example code? Thanks. ``` embedding = torch.from_numpy(embedding).float() # size self.embedding...
According to [[Peters et al., 2018](https://arxiv.org/pdf/1802.05365.pdf)], ELMo is a **task** specific combination of the intermediate layer representations in the biLM. The computation of ELMo embeddings is done by a **task**...
