Skip to content
Snippets Groups Projects

Repository graph

You can move around the graph by using the arrow keys.
Select Git revision
  • 103-sounds
  • 119-docs-new-perspective
  • 124-additional-state-keys-and-values
  • 144-gkk-exercises
  • 157-show-content-progress
  • 159-reduce-gif-size
  • dev protected
  • ffd-demo-arcade
  • llm-agent
  • main default protected
  • markdown-center-badge-html-demo
  • rl-bachelor-thesis
  • show-number-content
  • voicechat
  • web_version
  • v1.3.0
  • ffday
  • v1.2.0
  • v1.1.3
  • v1.1.0
  • v1.0.0
21 results
Created with Raphaël 2.2.07Aug6231Jul302928272625232219171412111087528Jun25241918171210129May15137230Apr2322191715141211109843227Mar2625212019181514131211109876543129Feb2827262523222120191615141312109876Integrated overcooked-ai into cooperative cuisine.Add environment config selection to GUIAdd random item ground and restricted movement hook-movement code now checks if the player still has some movement time left that is smaller then the server step size, and then uses that movement. Movementspeed is now way more consistent and also a lot fast(the intended speed) because of this.Merge branch '142-restructure-reinforcement-learning-files-2' into 'dev'Inference uses the same configs as the training and instantiates the model load function from the configs.Incorporated the feedback to make more use of the advantages of HydraMerge branch 'dev' into 142-restructure-reinforcement-learning-files-2Update changelog with past updatesMerge branch '139-deterministic-order-generation' into 'dev'Merge remote-tracking branch 'refs/remotes/origin/dev' into 139-deterministic-order-generationRemove debug print statements and improve testsSwitch to default study configurationUpdate deterministic order generators and testsEnhance reach goal functionality with new configurationsAdd target ground type and enhance counter movementAdd deadly ground feature with new ground typesRename variables and update function names for clarityAdd Cooperative Cuisine introductory Jupyter notebookRefactor plate config initialization and add loggerUpdate file paths in package_files functionAdd dynamic package data inclusion for additional directoriesAdd package_data to include additional directoriesAdd new hook and environment configuration filesMerge remote-tracking branch 'origin/rl-bachelor-thesis' into rl-bachelor-thesisfiles are now only read when creating cooking taskStart handle active effects and clean up obsolete objectsRemoved unnecessary importsSepareted base agent from other agents, random and llm agent overwrite manage_tasks methodAgents can now detect when they cannot interact and stop interacting when it is finishedMerge branch 'dev' into 142-restructure-reinforcement-learning-files-2Implemented a steerable version of the environment for the user to test it out in the play_gym.py.Updated prompting strategy. Gpt4o seems to work best, although not good. Actually pretty bad :(Added first version of llm agent with gpt4o-miniAdded first version of llm agent with gpt4o-minifixed rendering not working on GPU clusterupdated config files for MARLBetter logging of item manipulations of playersupdated experiment.yamlImplemented two basic StateConverters and integrated them into the hydra config management