stephenroller Follow

🗨️

Stephen Roller stephenroller

🗨️

Follow

NLP researcher working on chatbots. Tech lead for facebookresearch/ParlAI.

498 followers · 43 following

Achievements

BetaSend feedback

Achievements

BetaSend feedback

Highlights

Pro
7 discussions answered

Block or Report

Block or report stephenroller

Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse

Pinned

facebookresearch/metaseq Public

Repo for external large-scale work

Python 3.3k 319
facebookresearch/ParlAI Public

A framework for training and evaluating AI models on a variety of openly available dialogue datasets.

Python 8.9k 1.9k

860 contributions in the last year

Learn how we count contributions

Less More

NEW! View your contributions in 3D, VR and IRL!

Activity overview

Contributed to facebookresearch/ParlAI, facebookresearch/metaseq, facebookresearch/hypernymysuite and 9 other repositories

Contribution activity

June 2022

Created 8 commits in 2 repositories

Opened 5 pull requests in 2 repositories

facebookresearch/metaseq 4 merged

[api] More temperature woes Jun 22
[api] Support temp=0. Return errors as json. Jun 21
[api] Fix singular positive logit Jun 21
[api] Simplify launch config. Jun 17

facebookresearch/ParlAI 1 merged

Update README.md Jun 21

Reviewed 18 pull requests in 2 repositories

facebookresearch/metaseq 11 pull requests

[Community] Add OPT-Alpa integration Jun 24
Updating codeowners to recent active users Jun 23
Update CODEOWNERS Jun 23
OPT-66B Release Jun 22
updating API launch instruction, and adding the supporting script Jun 21
[api] Fix singular positive logit Jun 21
Singleton checkpoint needs to include decoder.version for single-ton checkpoint to run correctly Jun 21
Fix Beam Search Jun 17
fix off by one in API Jun 16
Fix progress bar exception when both tensorboard and wandb are turned on Jun 13
Add instructions to download the GPT2 asset files Jun 1

facebookresearch/ParlAI 7 pull requests

trying prev version Jun 18
Add poetry install instructions in crowdsourcing doc Jun 17
director link: oops link was wrong somehow Jun 17
[lint] Fix failure to provide red checks for lint Jun 17
minor typo in docstring Jun 17
Add TorchScriptable transformer classifier and subword BPE tokenizer Jun 2
chore: Set permissions for GitHub actions Jun 1

Opened 2 issues in 1 repository

facebookresearch/metaseq 2 open

--arch transformer_lm and --tensor-init-on-gpu are incompatible Jun 23
Add support for subshards Jun 21

24 contributions in private repositories Jun 6 – Jun 24