Release v0.7.0

[josephbirkner]

Ravestate v0.7.0

Features

• [x] 🤠 Wilder Wildtalk: We added two new transformer architectures (ConvAI GPT, OpenAI GPT2). All Wildtalk is now running distributed, with auto-started local servers. (#105, #130)
• [x] 👁 VisionIO: Ravestate plugs into face_oracle to start and end conversations based on recognized faces. (#120)
• [x] 🏄 Raveboard! Ravestate has a new (beta) UI, which exposes internal interactions between Spikes and Activations. It also offers a neat chat window. (#90, #91, #92, #93, #95, #138)
• [x] ravestate_ros1: Ravestate may now be integrated with rospy (ROS Melodic). In combination with ravestate_ros2, we got rid of the need for rosbridge. (#116, #129)
• [x] 🐳 Refactored docker image, now includes Neo4j, redis, face_oracle and pre-downloaded Wildtalk/Spacy models. Docker-compose services automate all the path/port/device mappings.
• [x] 🎠 Tests with ROS and Neo4j run in Docker on Travis. 80% coverage secured! (#98. #120)
• [x] 🎁 Improved docs with syntax-highlighting and more examples (but way to go) (#100, #123, #129)
• [x] 🤔 ravestate_nlp: Uber-robust yes/no detection in ravestate_nlp (#82, #107, #132)
• [x] ❓ravestate_nlp: Much better question detection. (#106)
• [x] 👨‍👦‍👦ravestate_nlp: Triple.match_either_lemma(...) can now match subtree tokens of triple elements. (#109)
• [x] 🦑 Module dependencies: Modules may be split across multiple files, and declare explicit dependencies to other Module objects. (#111)

Inference Dataset

As part of this release, we have built an inference dataset which may be used in the future to improve... inference (#101).

• The dataset can be found here
• Some analysis here: #96

This release was built with the help of the following contributors:

@Toseban @emlozin @nbasargin @l-laura @missxa @ec-m @umursen @theycallmefm @NeginsCode

Special thanks go to the SS19 Luigi Team!🍦❤️

[codecov-io]

Codecov Report

Merging #127 into master will decrease coverage by 10.42%. The diff coverage is 26.25%.

@@             Coverage Diff             @@
##           master     #127       +/-   ##
===========================================
- Coverage   74.87%   64.44%   -10.43%     
===========================================
  Files          56       67       +11     
  Lines        3495     4239      +744     
===========================================
+ Hits         2617     2732      +115     
- Misses        878     1507      +629

Impacted Files	Coverage Δ
modules/ravestate_nlp/extract_triples.py	97.05% <ø> (-0.17%)	:arrow_down:
modules/ravestate_wildtalk/server.py	0% <0%> (ø)
modules/ravestate_wildtalk/parlai_backend.py	0% <0%> (ø)
modules/ravestate_wildtalk/__init__.py	0% <0%> (-92.31%)	:arrow_down:
modules/ravestate_akinator/__init__.py	44.44% <0%> (-1.46%)	:arrow_down:
modules/raveboard/ui_context.py	0% <0%> (ø)
modules/ravestate_ros1/__init__.py	0% <0%> (ø)
modules/ravestate_ros1/ros1_properties.py	0% <0%> (ø)
modules/ravestate_wildtalk/gpt2_backend.py	0% <0%> (ø)
modules/ravestate_sendpics/__init__.py	0% <0%> (ø)	:arrow_up:
... and 43 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update d233f9b...3d20937. Read the comment docs.

[josephbirkner]

Thanks for the update!

That gif is probably larger than the whole code I wrote in this repo + the compiled output 😆

GIFs are mandatory 🤷‍♂

[josephbirkner]

To all reviewers: You may easily test this branch via pip install -U ravestate and python3 -m raveboard ravestate_wildtalk.