Sergey Sokolov

[MODEL] BiDAF and QANet use common data pipeline

4

## Description ## This is a PR with current results of merging BiDAF and QANet together. Unfortunately, there was something missing / wrong in the merge as I couldn't get...

What is a good value for nSamples in EstimateBetweenness?

1

Hi there, I managed to run in parallel EstimateBetweenness algorithm on my graph of 350M nodes and 5.5B edges with only nSamples=5. Any value significantly larger gives me OOM even...

Closeness centrality for a huge graph

1

Hi there, I am using 2.12:66565565-SNAPSHOT version of the Sparkling, which is compatible with 2.12 version of Scala. I have a single csv file of nodes and ~265M edges (4.5...

[QUESTION] How to get the latest Spark 3.1 version?

5

Hi team, I see that the Spark3.1 branch is updated regularly, and I wonder what is the latest Spark 3.1 compatible version which I could use? Is there is a...

Assemble BiDAF model and add training script

1

## Description ## The BiDAF model is assembled and training script can be run for multiple epochs. Need to fix everything that is marked with `TODO:` in code. [X] Reduce...

## Description ## Better responsibility separation. `_read_data()` now just only reads the json into memory, not parses it. It allows to get the json as-is later, if needed. I use...

## Description ## I add bidaf output layer, which output is batch_size of a tuple of start_index and end_index. ## Checklist ## ### Essentials ### - [X] Changes are complete...

## Description ## Add modeling layer of BiDAF: just 2-layer bi-LSTM and mimic data input from attention flow. ## Checklist ## ### Essentials ### - [X] Changes are complete (i.e....

## Description ## BiDAF embedding is done and tested using original SQuAD dataset ## Checklist ## ### Essentials ### - [X] Changes are complete (i.e. I finished coding on this...

Hi, I would like to use googleVis to display Intervals chart (https://developers.google.com/chart/interactive/docs/gallery/intervals). My particular use case I want to display standard deviation from the line. Thank you.