From c3b85be126b75b4b0a1015b5dbce340bb11a880e Mon Sep 17 00:00:00 2001 From: Yao Xu Date: Sun, 7 Jun 2026 10:04:43 -0700 Subject: [PATCH] Add Cosmos 3 README redirect --- cosmos_reason1_utils/README.md | 15 +++++++++++++++ examples/benchmark/README.md | 15 +++++++++++++++ examples/post_training/README.md | 15 +++++++++++++++ examples/post_training_hf/README.md | 15 +++++++++++++++ examples/post_training_llava/README.md | 15 +++++++++++++++ examples/video_critic/README.md | 15 +++++++++++++++ prompts/README.md | 15 +++++++++++++++ 7 files changed, 105 insertions(+) diff --git a/cosmos_reason1_utils/README.md b/cosmos_reason1_utils/README.md index 89d581f..3be5501 100644 --- a/cosmos_reason1_utils/README.md +++ b/cosmos_reason1_utils/README.md @@ -1,3 +1,18 @@ +> [!IMPORTANT] +> ## 🚀 [Cosmos 3 Has Arrived](https://github.com/NVIDIA/Cosmos) +> +> Cosmos 3 is NVIDIA's next-generation foundation model platform for Physical AI. Compared with Cosmos-Reason1, Cosmos 3 delivers substantially stronger physical reasoning capabilities while extending beyond reasoning to support world prediction, simulation, transfer, and action generation within a single unified model. +> +> Rather than relying on separate models for reasoning, prediction, transfer, and policy learning, a single Cosmos 3 model can understand the world, reason about physical interactions, predict future outcomes, transform observations across domains, and generate actions for embodied agents. This unified architecture enables stronger performance across a broad range of Physical AI applications, including robotics, autonomous vehicles, and smart spaces. +> +> This repository is no longer under active development and will receive only limited maintenance updates. Future model releases, features, documentation, and community support will be focused on Cosmos 3. +> +> 👉 Visit the new Cosmos home: https://github.com/NVIDIA/Cosmos +> +> There you will find the latest Cosmos 3 models, technical reports, tutorials, benchmarks, and ecosystem updates. +> +> Thank you for your support of Cosmos-Reason1. We encourage all users to migrate to Cosmos 3 for the latest state-of-the-art Physical AI capabilities. + # Cosmos-Reason1 Utilities Inference utilities for Cosmos-Reason1. diff --git a/examples/benchmark/README.md b/examples/benchmark/README.md index 2aba595..166190c 100644 --- a/examples/benchmark/README.md +++ b/examples/benchmark/README.md @@ -1,3 +1,18 @@ +> [!IMPORTANT] +> ## 🚀 [Cosmos 3 Has Arrived](https://github.com/NVIDIA/Cosmos) +> +> Cosmos 3 is NVIDIA's next-generation foundation model platform for Physical AI. Compared with Cosmos-Reason1, Cosmos 3 delivers substantially stronger physical reasoning capabilities while extending beyond reasoning to support world prediction, simulation, transfer, and action generation within a single unified model. +> +> Rather than relying on separate models for reasoning, prediction, transfer, and policy learning, a single Cosmos 3 model can understand the world, reason about physical interactions, predict future outcomes, transform observations across domains, and generate actions for embodied agents. This unified architecture enables stronger performance across a broad range of Physical AI applications, including robotics, autonomous vehicles, and smart spaces. +> +> This repository is no longer under active development and will receive only limited maintenance updates. Future model releases, features, documentation, and community support will be focused on Cosmos 3. +> +> 👉 Visit the new Cosmos home: https://github.com/NVIDIA/Cosmos +> +> There you will find the latest Cosmos 3 models, technical reports, tutorials, benchmarks, and ecosystem updates. +> +> Thank you for your support of Cosmos-Reason1. We encourage all users to migrate to Cosmos 3 for the latest state-of-the-art Physical AI capabilities. + # Cosmos-Reason1 Benchmark Example This guide provides instructions for evaluating models on the [Cosmos-Reason1 Benchmark](https://huggingface.co/datasets/nvidia/Cosmos-Reason1-Benchmark) diff --git a/examples/post_training/README.md b/examples/post_training/README.md index d5c529f..b1e254d 100644 --- a/examples/post_training/README.md +++ b/examples/post_training/README.md @@ -1,3 +1,18 @@ +> [!IMPORTANT] +> ## 🚀 [Cosmos 3 Has Arrived](https://github.com/NVIDIA/Cosmos) +> +> Cosmos 3 is NVIDIA's next-generation foundation model platform for Physical AI. Compared with Cosmos-Reason1, Cosmos 3 delivers substantially stronger physical reasoning capabilities while extending beyond reasoning to support world prediction, simulation, transfer, and action generation within a single unified model. +> +> Rather than relying on separate models for reasoning, prediction, transfer, and policy learning, a single Cosmos 3 model can understand the world, reason about physical interactions, predict future outcomes, transform observations across domains, and generate actions for embodied agents. This unified architecture enables stronger performance across a broad range of Physical AI applications, including robotics, autonomous vehicles, and smart spaces. +> +> This repository is no longer under active development and will receive only limited maintenance updates. Future model releases, features, documentation, and community support will be focused on Cosmos 3. +> +> 👉 Visit the new Cosmos home: https://github.com/NVIDIA/Cosmos +> +> There you will find the latest Cosmos 3 models, technical reports, tutorials, benchmarks, and ecosystem updates. +> +> Thank you for your support of Cosmos-Reason1. We encourage all users to migrate to Cosmos 3 for the latest state-of-the-art Physical AI capabilities. + # Cosmos-Reason1 Post-Training Example This guide provides instructions for post-training Cosmos-Reason1 on the [SFT](https://huggingface.co/datasets/nvidia/Cosmos-Reason1-SFT-Dataset)/[RL](https://huggingface.co/datasets/nvidia/Cosmos-Reason1-RL-Dataset) datasets using [cosmos-rl](https://github.com/nvidia-cosmos/cosmos-rl). diff --git a/examples/post_training_hf/README.md b/examples/post_training_hf/README.md index 46705e1..81c3935 100644 --- a/examples/post_training_hf/README.md +++ b/examples/post_training_hf/README.md @@ -1,3 +1,18 @@ +> [!IMPORTANT] +> ## 🚀 [Cosmos 3 Has Arrived](https://github.com/NVIDIA/Cosmos) +> +> Cosmos 3 is NVIDIA's next-generation foundation model platform for Physical AI. Compared with Cosmos-Reason1, Cosmos 3 delivers substantially stronger physical reasoning capabilities while extending beyond reasoning to support world prediction, simulation, transfer, and action generation within a single unified model. +> +> Rather than relying on separate models for reasoning, prediction, transfer, and policy learning, a single Cosmos 3 model can understand the world, reason about physical interactions, predict future outcomes, transform observations across domains, and generate actions for embodied agents. This unified architecture enables stronger performance across a broad range of Physical AI applications, including robotics, autonomous vehicles, and smart spaces. +> +> This repository is no longer under active development and will receive only limited maintenance updates. Future model releases, features, documentation, and community support will be focused on Cosmos 3. +> +> 👉 Visit the new Cosmos home: https://github.com/NVIDIA/Cosmos +> +> There you will find the latest Cosmos 3 models, technical reports, tutorials, benchmarks, and ecosystem updates. +> +> Thank you for your support of Cosmos-Reason1. We encourage all users to migrate to Cosmos 3 for the latest state-of-the-art Physical AI capabilities. + # Cosmos-Reason1 Post-Training Hugging Face Example This package provides a minimal Cosmos-Reason1 post-training example using the [Hugging Face datasets](https://huggingface.co/docs/datasets/en/index) format. You should first read the full post-training example, see [Cosmos-Reason1 Post-Training Full](../post_training/README.md). diff --git a/examples/post_training_llava/README.md b/examples/post_training_llava/README.md index 1754c17..202eed3 100644 --- a/examples/post_training_llava/README.md +++ b/examples/post_training_llava/README.md @@ -1,3 +1,18 @@ +> [!IMPORTANT] +> ## 🚀 [Cosmos 3 Has Arrived](https://github.com/NVIDIA/Cosmos) +> +> Cosmos 3 is NVIDIA's next-generation foundation model platform for Physical AI. Compared with Cosmos-Reason1, Cosmos 3 delivers substantially stronger physical reasoning capabilities while extending beyond reasoning to support world prediction, simulation, transfer, and action generation within a single unified model. +> +> Rather than relying on separate models for reasoning, prediction, transfer, and policy learning, a single Cosmos 3 model can understand the world, reason about physical interactions, predict future outcomes, transform observations across domains, and generate actions for embodied agents. This unified architecture enables stronger performance across a broad range of Physical AI applications, including robotics, autonomous vehicles, and smart spaces. +> +> This repository is no longer under active development and will receive only limited maintenance updates. Future model releases, features, documentation, and community support will be focused on Cosmos 3. +> +> 👉 Visit the new Cosmos home: https://github.com/NVIDIA/Cosmos +> +> There you will find the latest Cosmos 3 models, technical reports, tutorials, benchmarks, and ecosystem updates. +> +> Thank you for your support of Cosmos-Reason1. We encourage all users to migrate to Cosmos 3 for the latest state-of-the-art Physical AI capabilities. + # Cosmos-Reason1 Post-Training Llava Example This package provides a minimal Cosmos-Reason1 post-training example using the [Llava datasets](https://github.com/haotian-liu/LLaVA/blob/main/docs/Finetune_Custom_Data.md) format. You should first read the full post-training example, see [Cosmos-Reason1 Post-Training Full](../post_training/README.md). diff --git a/examples/video_critic/README.md b/examples/video_critic/README.md index 1f3351d..6970561 100644 --- a/examples/video_critic/README.md +++ b/examples/video_critic/README.md @@ -1,3 +1,18 @@ +> [!IMPORTANT] +> ## 🚀 [Cosmos 3 Has Arrived](https://github.com/NVIDIA/Cosmos) +> +> Cosmos 3 is NVIDIA's next-generation foundation model platform for Physical AI. Compared with Cosmos-Reason1, Cosmos 3 delivers substantially stronger physical reasoning capabilities while extending beyond reasoning to support world prediction, simulation, transfer, and action generation within a single unified model. +> +> Rather than relying on separate models for reasoning, prediction, transfer, and policy learning, a single Cosmos 3 model can understand the world, reason about physical interactions, predict future outcomes, transform observations across domains, and generate actions for embodied agents. This unified architecture enables stronger performance across a broad range of Physical AI applications, including robotics, autonomous vehicles, and smart spaces. +> +> This repository is no longer under active development and will receive only limited maintenance updates. Future model releases, features, documentation, and community support will be focused on Cosmos 3. +> +> 👉 Visit the new Cosmos home: https://github.com/NVIDIA/Cosmos +> +> There you will find the latest Cosmos 3 models, technical reports, tutorials, benchmarks, and ecosystem updates. +> +> Thank you for your support of Cosmos-Reason1. We encourage all users to migrate to Cosmos 3 for the latest state-of-the-art Physical AI capabilities. + # Using Cosmos-Reason1 as Video Critic for Rejection Sampling

diff --git a/prompts/README.md b/prompts/README.md index 3caa094..4d5afb1 100644 --- a/prompts/README.md +++ b/prompts/README.md @@ -1,3 +1,18 @@ +> [!IMPORTANT] +> ## 🚀 [Cosmos 3 Has Arrived](https://github.com/NVIDIA/Cosmos) +> +> Cosmos 3 is NVIDIA's next-generation foundation model platform for Physical AI. Compared with Cosmos-Reason1, Cosmos 3 delivers substantially stronger physical reasoning capabilities while extending beyond reasoning to support world prediction, simulation, transfer, and action generation within a single unified model. +> +> Rather than relying on separate models for reasoning, prediction, transfer, and policy learning, a single Cosmos 3 model can understand the world, reason about physical interactions, predict future outcomes, transform observations across domains, and generate actions for embodied agents. This unified architecture enables stronger performance across a broad range of Physical AI applications, including robotics, autonomous vehicles, and smart spaces. +> +> This repository is no longer under active development and will receive only limited maintenance updates. Future model releases, features, documentation, and community support will be focused on Cosmos 3. +> +> 👉 Visit the new Cosmos home: https://github.com/NVIDIA/Cosmos +> +> There you will find the latest Cosmos 3 models, technical reports, tutorials, benchmarks, and ecosystem updates. +> +> Thank you for your support of Cosmos-Reason1. We encourage all users to migrate to Cosmos 3 for the latest state-of-the-art Physical AI capabilities. + # Prompts We provide a set of task-specific prompt templates that are known to work well with Cosmos-Reason1: