Dataset · Hugging Face

morinoppp/cosmos_results

Name: cosmos_results
Creator: morinoppp
License: mit
Keywords: task_categories:text-to-video, license:mit, size_categories:n<1K, modality:video, library:datasets, library:mlcroissant, region:us, world-model, action-conditioned-video-generation, cosmos-predict2, robotics, physics-simulation

Cosmos Predict2 Benchmark Results Generated videos from the Cosmos-Predict2 action-conditioned world model on the full benchmark suite, covering physics consistency, action following, perturbation sensitivity, and run-action tasks. Models Model Size Checkpoint Cosmos-Predict2 14B iter_000006000/model_ema_bf16.pt Cosmos-Predict2 2B iter_000008000/model_ema_bf16.pt Both models are fine-tuned on GR1 robot data with 29-dimensional joint actions.… See the full description on the dataset page: https://huggingface.co/datasets/morinoppp/cosmos_results.

View on Hugging FaceReport an issue

Author / organization	morinoppp
License (as declared)	mit
Downloads	155 (as reported by Hugging Face)
Likes	0
Created	5 May 2026
Last updated at source	5 May 2026

world-modelaction-conditioned-video-generationcosmos-predict2roboticsphysics-simulation

Embed this entry

<iframe src="https://silytics.si/embed/item/ae5e2d7a-7b9b-4da5-a43a-b9c0986ff17d" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 3 days

155 now (+13 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.