ELMoForManyLangs
ELMoForManyLangs copied to clipboard
HIT-SCIR
Pre-trained ELMo Representations for Many Languages
I am exporting Embedder to ONNX format. But the size of output model is so large(~7G) that I doubt there are many duplications of weights. How to export Embedder correctly?...
With the addition of signature checks in `override v5.0.0` (https://github.com/mkorpela/overrides/releases/tag/5.0.0), importing the `ELMoForManyLangs` package will fail as mentioned in issue #95. I propose a simple fix by simply disabling the...
It is not clear how to use the trained model to reproduce the results of the ConLL 2018 Shared Task. (neither ho to performe training for the same task) Following...
python -m elmoformanylangs.biLM train \ --train_path data/en.raw \ --config_path elmoformanylangs/configs/cnn_50_100_512_4096_sample.json \ --model output/en \ --optimizer adam \ --lr 0.001 \ --lr_decay 0.8 \ --max_epoch 10 \ --max_sent_len 20 \ --max_vocab_size...
Hi, thank you so much for your great and well-explained work; do you have any idea how we con get the document embedding using Embedder ? and result like (num_documents,...
能否让一组语句的向量能够维度一致，便于拼接？ 
Hi, I am using ELMo for Japanese. Here is my code: ``` from elmoformanylangs import Embedder e = Embedder('/Users/tanh/Desktop/alt/JapaneseElmo') if __name__ == '__main__': sents = [ ['今'], ['今'], ['潮水', '退']...
Gigaword Chinese V5 里面的中文好像没有分好词？ 想问一下那个simplified Chinese model 的segmentation 是怎么做的? 因为我download 下来的embedding 好像是word based的？
请问介绍中提到的简体中文的预训练语料xinhua proportion of Chinese gigawords-v5可以从哪里获得？ 非常感谢~
