This is a minimalist PyTorch implementation of the label inducer in [1]. For the full codebase used in experiments, refer to the repository at [2].
The code is in Python 3.6 and uses PyTorch version 1.0.1.post2. Tested with Geforce RTX 2080 Ti and CUDA version 10.1.105.
You can get the universal treebank v2.0 at [3] (McDonald et al., 2013), which provides both fine-grained and coarse-grained labels.
python main.py example-model example.words --train --epochs 10 --num_labels 3
python main.py en45-model ${EN45PATH}/en.words --train --num_labels 45 --epochs 5 --cuda --clusters clusters.txt --pred pred.txtOutput logged in file en45-model.log
| epoch 1 | loss -1.08 | 1.55 bits | acc 77.95 | vm 72.51 | time 0:02:38
| epoch 2 | loss -1.55 | 2.23 bits | acc 79.17 | vm 73.25 | time 0:02:41
| epoch 3 | loss -1.73 | 2.50 bits | acc 79.47 | vm 73.45 | time 0:02:42
| epoch 4 | loss -1.86 | 2.69 bits | acc 79.37 | vm 73.44 | time 0:02:38
| epoch 5 | loss -1.96 | 2.83 bits | acc 79.28 | vm 73.35 | time 0:02:41
Training time 0:13:23
=========================================================================================
| Best | acc 79.47 | vm 73.45
=========================================================================================[1] Mutual Information Maximization for Simple and Accurate Part-Of-Speech Induction (Stratos, 2018)
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent