jsedoc

REALM - Training data for pre-trained model

1

In your [README](https://github.com/google-research/language/blob/master/language/realm/README.md) you say _We are in the process of releasing the full pre-training corpus and retrieval corpus (Wikipedia) needed for pre-training REALM_ Is this still the plan? Maybe...

Error Loading ELMo

3

I'm just trying to use the ELMo embedding feature and getting the following issue: using a fresh pip install in a virtualenv `(D) jsedoc@****:$ pip install -U pymagnitude Requirement already...

Recommended implementation of multi-turn training

4

In order to incorporate a model like HRED into parlai, we need to get triplets (). I've made a few hack with a custom loader to do this, but I...

for cuda 7.5 please point to https://developer.nvidia.com/cuda-75-downloads-archive

model parameters from paper

1

Would it be possible to share your checkpoints used in your paper? Thanks! and great work!

Hi I copied the pytorch toturial and changed it for MxNet. This is a work in progress, but any comments would be useful. Here's the [link](https://colab.research.google.com/drive/1PpzMl29L-18CK7v3mig-GnE_k6r7n08a).