eloialonso/iris

Transformers are Sample-Efficient World Models. ICLR 2023, notable top 5%.

PythonJupyter NotebookShellmachine-learningresearchreinforcement-learningdeep-learningtransformersartificial-intelligenceatariworld-models
This is stars and forks stats for /eloialonso/iris repository. As of 03 May, 2024 this repository has 689 stars and 61 forks.

Transformers are Sample-Efficient World Models (IRIS) Transformers are Sample-Efficient World Models Vincent Micheli*, Eloi Alonso*, François Fleuret * Denotes equal contribution IRIS agent after 100k environment steps, i.e. two hours of real-time experience tl;dr IRIS is a data-efficient agent trained over millions of imagined trajectories in a world model. The world model is composed of a discrete autoencoder and an autoregressive Transformer. Our approach casts dynamics learning as a sequence...
Read on GithubGithub Stats Page
repotechsstarsweeklyforksweekly
impira/docqueryPythonJavaScriptJupyter Notebook1.5k+61000
riverscuomo/new-albumsPython1310210
zhongxinghong/PKUAutoElectivePythonDockerfile61102360
sammchardy/python-binancePython5.5k02.2k0
rbbrdckybk/ai-art-generatorPythonCSS60201260
tibordp/aluminaRustJavaScriptMakefile138080
dotfiles/dotfiles.github.comSassHTMLPython1.4k01370
efchatz/HTTP3-attacksShellDockerfile620200
KISS1996/trexminerShell2250350
lacework/helm-chartsMustacheSmartyShell110270