Cosmos Tokenizer: A suite of image and video tokenizers

Website | NVIDIA News | NVIDIA Blog | Hugging Face | YouTube

Model Overview

Description:

Cosmos Tokenizer is a suite of visual tokenizers for images and videos that delivers various compression rates while maintaining high reconstruction quality. Cosmos Tokenizer can serve as an effective and efficient building block in both diffusion-based and autoregressive models for image and video generation. This model is ready for commercial use.

Our tokenizers come in two types: Continuous (C) and Discrete (D), each with Image (I) and Video (V) variants:

Continuous tokenizers encode visual data into continuous latent embeddings, as shown in latent diffusion models like Stable Diffusion. These embeddings are suitable for models that generate data by sampling from continuous distributions.
Discrete tokenizers encode visual data into discrete latent codes, mapping them into quantized indices, as seen in autoregressive transformers such as VideoPoet. This discretization is required for models that generate data by optimizing the cross-entropy loss, such as the GPT models.

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n

tr:nth-child(odd)]:bg-surface-sunken\">

	Continuous ( C )	Discrete ( D )
Images ( I )	Cosmos-Tokenizer-CI	Cosmos-Tokenizer-DI
Videos ( V )	Cosmos-Tokenizer-CV	Cosmos-Tokenizer-DV

Given an image or a video, Cosmos Tokenizer outputs either continuous latents or discrete tokens. Cosmos Tokenizer achieves spatial compression rates of 8x8 or 16x16 and temporal compression factors of 4x or 8x, resulting in a total compression factor of up to 2048x (=8x16x16). Cosmos Tokenizer delivers 8x more total compression than state-of-the-art (SOTA) methods while simultaneously maintaining higher image quality and running up to 12x faster than the best available SOTA tokenizers.

Model Developer: NVIDIA

Model Versions

This release (v1.0) of Cosmos Tokenizer includes the following tokenizers:

Continuous Tokenizers\n
- Cosmos-1.0-Tokenizer-CV8x8x8 (8x temporal compression, 8x8 spatial compression, 720 short spatial size, 121 frames context)
\n
Discrete Tokenizers\n
- Cosmos-1.0-Tokenizer-DV8x16x16 (8x temporal compression, 16x16 spatial compression, 720 short spatial size, 49 frames context)
\n

The previous release (v0.1) of Cosmos Tokenizer included the following tokenizers:

Continuous Tokenizers\n
- Continuous Image (CI) Tokenizer\n
  - Cosmos-0.1-Tokenizer-CI8x8 (8x8 spatial compression)
  - Cosmos-0.1-Tokenizer-CI16x16 (16x16 spatial compression)
  \n
- Continuous Video (CV) Tokenizer\n
  - Cosmos-0.1-Tokenizer-CV4x8x8 (4x temporal compression, 8x8 spatial compression)
  - Cosmos-0.1-Tokenizer-CV8x8x8 (8x temporal compression, 8x8 spatial compression)
  - Cosmos-0.1-Tokenizer-CV8x16x16 (8x temporal compression, 16x16 spatial compression)
  \n
\n
Discrete Tokenizers\n
- Discrete Image (DI) Tokenizer\n
  - Cosmos-0.1-Tokenizer-DI8x8 (8x8 spatial compression)
  - Cosmos-0.1-Tokenizer-DI16x16 (16x16 spatial compression)
  \n
- Discrete Video (DV) Tokenizer\n
  - Cosmos-0.1-Tokenizer-DV4x8x8 (4x temporal compression, 8x8 spatial compression)
  - Cosmos-1.0-Tokenizer-DV8x8x8 (8x temporal compression, 8x8 spatial compression)
  - Cosmos-1.0-Tokenizer-DV8x16x16 (8x temporal compression, 16x16 spatial compression)
  \n
\n

License/Terms of Use:

NVIDIA Open Model License

Under the NVIDIA Open Model License, NVIDIA confirms:

Models are commercially usable.
You are free to create and distribute Derivative Models.
NVIDIA does not claim ownership to any outputs generated using the Models or Derivative Models.

Model Architecture:

We designed Cosmos Tokenizer using a lightweight and computationally efficient architecture, featuring a temporally causal design. Specifically, we employ causal temporal convolution and causal temporal attention layers to preserve the natural temporal order of video frames, ensuring seamless tokenization of images and videos using a single unified network architecture. The encoder and decoder form a symmetrical pair, which are mirrors of each other. The encoder starts with a 2-level Haar wavelet transform layer, which down-samples inputs by a factor of 4 in both spatial and temporal dimensions. Likewise, the decoder ends with an inverse wavelet transform. We employ the vanilla autoencoder (AE) formulation to model the latent space for continuous tokenizers. For discrete tokenizers, we adopt the Finite-Scalar-Quantization (FSQ) as the latent space quantizer.

$\"image/jpeg\"$

Input/Output Specifications

Encoder

\n
Input
\n
- Type: Images or Videos
- Format: RGB (Red, Green, Blue)
- Properties:\n
  - Resolution: Minimum: 256px (shorter side). Maximum: Up to 4K
  - Video Length: Up to 8 seconds for 1080p videos (bounded by A100 80G GPU memory; higher resolutions will have shorter supported durations)
  \n
\n
\n
Output
\n
- Type: Tokens
- Properties:\n
  - Integer indices ranging from 0 to 63,999
  \n
\n

Decoder

\n
Input
\n
- Type: Tokens
- Properties:\n
  - Integer indices ranging from 0 to 63,999
  \n
\n
\n
Output
\n
- Type Images or Videos (matching input type)
- Format: RGB (Red, Green, Blue)
- Properties:\n
  - Resolution: Same as input resolution
  - Video Length: Same as input video length
  \n
\n

Software Integration (Required For NVIDIA Models Only):

Runtime Engine(s):

Cosmos-Tokenizer

Supported Hardware Microarchitecture Compatibility:

NVIDIA Ampere (e.g., A100)
NVIDIA Hopper (e.g., H100)

Note: We have only tested Cosmos Tokenizer with BF16 precision on Ampere and Hopper GPUs. If you are using older versions of NVIDIA GPUs (e.g., NVIDIA Volta GPUs), you may need to switch to FP32 precision.

Operating System(s):

Linux (We have not tested on other operating systems.)

Usage

Inference Engines:

Cosmos-Tokenizer (PyTorch)

Inference with `Cosmos-Tokenizer` (PyTorch)

Step-1: Installation of `Cosmos-Tokenizer`

Note: Currently, the Cosmos-Tokenizer code is only supported on Linux.

\n
Please clone the Cosmos-Tokenizer from GitHub repo github.com/NVIDIA/Cosmos-Tokenizer.
\n
```
git clone https://github.com/NVIDIA/Cosmos-Tokenizer.git\ncd Cosmos-Tokenizer\n
```
\n

Install dependencies

pip3 install -r requirements.txt\napt-get install -y ffmpeg\n

Preferably, you could build a docker image using our provided Dockerfile.

docker build -t cosmos-docker -f Dockerfile.  \n# You can run the container as:\ndocker run --gpus all -it --rm -v /home/${USER}:/home/${USER} \\\n    --workdir ${PWD} cosmos-docker /bin/bash\n

Step-2: Download Pre-trained Checkpoints

Create a local directory for the pre-trained checkpoints and download the\npre-trained checkpoints from HuggingFace.

, add_to_git_credential=True)\n# You could specify the tokenizers you want to download.\nmodel_names = [\n "Cosmos-1.0-Tokenizer-DV8x16x16",\n]\nfor model_name in model_names:\n hf_repo = "nvidia/" + model_name\n local_dir = "pretrained_ckpts/" + model_name\n os.makedirs(local_dir, exist_ok=True)\n print(f"downloading {model_name} to {local_dir}...")\n snapshot_download(repo_id=hf_repo, local_dir=local_dir)\n\">

from huggingface_hub import login, snapshot_download\nimport os\n# You could get your Hugging Face token from https://huggingface.co/settings/tokens\nlogin(token=<YOUT-HF-TOKEN>, add_to_git_credential=True)\n# You could specify the tokenizers you want to download.\nmodel_names = [\n        \"Cosmos-1.0-Tokenizer-DV8x16x16\",\n]\nfor model_name in model_names:\n    hf_repo = \"nvidia/\" + model_name\n    local_dir = \"pretrained_ckpts/\" + model_name\n    os.makedirs(local_dir, exist_ok=True)\n    print(f\"downloading {model_name} to {local_dir}...\")\n    snapshot_download(repo_id=hf_repo, local_dir=local_dir)\n

Under the ech checkpoint directory pretrained_ckpts/<model-name>, we provide the encoder,\ndecoder and the full autoencoder JIT models.

├── pretrained_ckpts/  \n│   ├── Cosmos-1.0-Tokenizer-DV8x16x16/\n│   │   ├── encoder.jit\n│   │   ├── decoder.jit\n│   │   ├── autoencoder.jit\n│   ...\n

Step-3: Run Inference

You can use the following example commands to encode and decode images or videos. For each, the same command works for both continuous and discrete tokenization. Simply provide the proper JIT-compiled ckpt to checkpoint_enc, checkpoint_dec, or the full autoencoder ckpt to checkpoint.

import torch\nfrom cosmos_tokenizer.video_lib import CausalVideoTokenizer\nmodel_name = \"Cosmos-1.0-Tokenizer-DV8x16x16\"\ninput_tensor = torch.randn(1, 3, 9, 512, 512).to('cuda').to(torch.bfloat16)\nencoder = CausalVideoTokenizer(checkpoint_enc=f'pretrained_ckpts/{model_name}/encoder.jit')\n(indices, codes) = encoder.encode(input_tensor)\ntorch.testing.assert_close(indices.shape, (1, 2, 32, 32))\ntorch.testing.assert_close(codes.shape, (1, 6, 2, 32, 32))\n\n# The input tensor can be reconstructed by the decoder as:\ndecoder = CausalVideoTokenizer(checkpoint_dec=f'pretrained_ckpts/{model_name}/decoder.jit')\nreconstructed_tensor = decoder.decode(indices)\ntorch.testing.assert_close(reconstructed_tensor.shape, input_tensor.shape)\n

The indices will have the shape (1, 2, 32, 32) and contain integral values in the range [1..64K], where the first of the three integral maps represents the first frame.\nThe codes will contain the pre-quantization continuous latent with shape (1, 6, 2, 32, 32), where C=6 represents the number of FSQ levels.

Note: More inference usage commands, including both TorchScript (JIT) and PyTorch Inference APIs on real images and videos, can be found on our GitHub repository github.com/NVIDIA/Cosmos-Tokenizer.

Evaluation

TokenizationPerformance Comparison

We have extensively evaluated the additional Cosmos Tokenizer models on DAVIS video benchmark dataset.

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n

tr:nth-child(odd)]:bg-surface-sunken\">

Tokenizer	Compression Ratio	Height	Num. of Frames	Quantization	PSNR (DAVIS)	SSIM (DAVIS)	rFVD (DAVIS)
VideoGPT	4×4×4	-	-	VQ	32.23	0.850	72.33
OmniTokenizer	4×8×8	-	-	VQ	28.44	0.712	188.60
Cosmos-Tokenizer-DV	4×8×8	720	17	FSQ	32.98	0.818	37.36
Cosmos-Tokenizer-DV	8×8×8	720	17	FSQ	32.11	0.775	100.15
Cosmos-Tokenizer-DV	8×16×16	720	17	FSQ	31.42	0.716	241.52
Cosmos-Tokenizer-DV	8×16×16	720	49	FSQ	31.59	0.719	259.33

We compare with the state-of-the-art discrete video tokenizer, OmniTokenizer.
Evaluation metrics:\n
- Peak Signal-to-Noise Ratio (PSNR)
- Structural Similarity (SSIM)
- Reconstruction Fréchet Video Distance (rFVD)
\n

Runtime Comparison

The following table shows the number of parameters and the averaged encoding and decoding times per image or video frame, measured on a single A100 80GB GPU. For comparison, we also list the parameters and average speeds of prior state-of-the-art tokenizer(s) with the same compression ratio.

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n

tr:nth-child(odd)]:bg-surface-sunken\">

Tokenizer	Resolution	Compression Ratio	Parameters	Time (ms)
OmniTokenizer	720x1280	4×8×8	54M	53.2
Cosmos-DV	720x1280	4×8×8	105M	51.5

Note: We benchmarked the runtime for images under the 8x8 compression and videos under the 4×8×8 compression. Tokenizers with different compression ratios are not included in this comparison.

Ethical Considerations

NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.

For more detailed information on ethical considerations for this model, please see the subcards of Explainability, Bias, Safety & Security, and Privacy below. Please report security vulnerabilities or NVIDIA AI Concerns here.

","resourceVersions","resourceVersionsTotal","hasMoreVersions","versionMetadata","versionNotFound","isLoggedIn","latestVersionPath","/orgs/nvidia/cosmos/models/cosmos-1.0-tokenizer-dv8x16x16/-","version","-","imageReference","cliDownloadCommand","ngc registry model download-version nvidia/cosmos/cosmos-1.0-tokenizer-dv8x16x16:1.0","helmFetchCommand","viewAllVersionsPath","/orgs/nvidia/cosmos/models/cosmos-1.0-tokenizer-dv8x16x16/-/version-history","accessCheckResult",{"_195":196,"_197":26},"hasRelatedCollections",true,"context",{"_86":87,"_88":85},1785756706815,[],[],[228,229,230,231,232,233,234],{"_40":55,"_42":67,"_44":57,"_46":58,"_48":68,"_50":51},{"_40":55,"_42":64,"_44":45,"_46":65,"_48":66,"_50":51},{"_40":55,"_42":62,"_44":45,"_46":65,"_48":235,"_50":51},{"_40":55,"_42":60,"_44":57,"_46":58,"_48":61,"_50":51},{"_40":55,"_42":56,"_44":57,"_46":58,"_48":59,"_50":51},{"_40":52,"_42":53,"_44":45,"_46":47,"_48":54,"_50":51},{"_40":41,"_42":43,"_44":45,"_46":47,"_48":49,"_50":51},49,"nvidia/cosmos/cosmos-1.0-tokenizer-dv8x16x16","nvidia/cosmos/models/cosmos-1.0-tokenizer-dv8x16x16","cosmos-1.0-tokenizer-dv8x16x16","Cosmos Tokenizer is a suite of visual tokenizers for images and videos that delivers various compression rates while maintaining high reconstruction quality.","# **Cosmos Tokenizer**: A suite of image and video tokenizers\n\n[**Website**](https://research.nvidia.com/labs/dir/cosmos-tokenizer) | [**NVIDIA News**](https://blogs.nvidia.com/blog/robot-learning-humanoid-development/) | [**NVIDIA Blog**](https://developer.nvidia.com/blog/state-of-the-art-multimodal-generative-ai-model-development-with-nvidia-nemo/) | [**Hugging Face**](https://huggingface.co/collections/nvidia/cosmos-tokenizer-672b93023add81b66a8ff8e6) | [**YouTube**](https://youtu.be/Soy_myOfWIU)\n\n\n# Model Overview\n\n## Description:\n**Cosmos Tokenizer** is a suite of visual tokenizers for images and videos that delivers various compression rates while maintaining high reconstruction quality. Cosmos Tokenizer can serve as an effective and efficient building block in both diffusion-based and autoregressive models for image and video generation. This model is ready for commercial use.\n\n\nOur tokenizers come in two types: **Continuous** (C) and **Discrete** (D), each with **Image** (I) and **Video** (V) variants:\n* Continuous tokenizers encode visual data into continuous latent embeddings, as shown in latent diffusion models like [Stable Diffusion](https://github.com/CompVis/stable-diffusion). These embeddings are suitable for models that generate data by sampling from continuous distributions.\n* Discrete tokenizers encode visual data into discrete latent codes, mapping them into quantized indices, as seen in autoregressive transformers such as [VideoPoet](https://sites.research.google/videopoet/). This discretization is required for models that generate data by optimizing the cross-entropy loss, such as the GPT models.\n\n\n| | Continuous ( C ) | Discrete ( D ) |\n| ------------------|---------------------|---------------------|\n| **Images ( I )** | Cosmos-Tokenizer-CI | Cosmos-Tokenizer-DI |\n| **Videos ( V )** | Cosmos-Tokenizer-CV | Cosmos-Tokenizer-DV |\n\n\nGiven an image or a video, Cosmos Tokenizer outputs either continuous latents or discrete tokens. Cosmos Tokenizer achieves spatial compression rates of 8x8 or 16x16 and temporal compression factors of 4x or 8x, resulting in a total compression factor of up to 2048x (=8x16x16). Cosmos Tokenizer delivers 8x more total compression than state-of-the-art (SOTA) methods while simultaneously maintaining higher image quality and running up to 12x faster than the best available SOTA tokenizers.\n\n**Model Developer**: NVIDIA\n\n## Model Versions\n\nThis release (v1.0) of Cosmos Tokenizer includes the following tokenizers:\n* **Continuous Tokenizers**\n * [Cosmos-1.0-Tokenizer-CV8x8x8](https://huggingface.co/nvidia/Cosmos-1.0-Tokenizer-CV8x8x8) (8x temporal compression, 8x8 spatial compression, 720 short spatial size, 121 frames context)\n* **Discrete Tokenizers** \n * [Cosmos-1.0-Tokenizer-DV8x16x16](https://huggingface.co/nvidia/Cosmos-1.0-Tokenizer-DV8x16x16) (8x temporal compression, 16x16 spatial compression, 720 short spatial size, 49 frames context)\n\nThe previous release (v0.1) of Cosmos Tokenizer included the following tokenizers:\n* **Continuous Tokenizers**\n * Continuous Image (CI) Tokenizer\n * [Cosmos-0.1-Tokenizer-CI8x8](https://huggingface.co/nvidia/Cosmos-Tokenizer-CI8x8) (8x8 spatial compression)\n * [Cosmos-0.1-Tokenizer-CI16x16](https://huggingface.co/nvidia/Cosmos-Tokenizer-CI16x16) (16x16 spatial compression)\n * Continuous Video (CV) Tokenizer\n * [Cosmos-0.1-Tokenizer-CV4x8x8](https://huggingface.co/nvidia/Cosmos-Tokenizer-CV4x8x8) (4x temporal compression, 8x8 spatial compression)\n * [Cosmos-0.1-Tokenizer-CV8x8x8](https://huggingface.co/nvidia/Cosmos-Tokenizer-CV8x8x8) (8x temporal compression, 8x8 spatial compression)\n * [Cosmos-0.1-Tokenizer-CV8x16x16](https://huggingface.co/nvidia/Cosmos-Tokenizer-CV8x16x16) (8x temporal compression, 16x16 spatial compression)\n* **Discrete Tokenizers**\n * Discrete Image (DI) Tokenizer\n * [Cosmos-0.1-Tokenizer-DI8x8](https://huggingface.co/nvidia/Cosmos-Tokenizer-DI8x8) (8x8 spatial compression)\n * [Cosmos-0.1-Tokenizer-DI16x16](https://huggingface.co/nvidia/Cosmos-Tokenizer-DI16x16) (16x16 spatial compression)\n * Discrete Video (DV) Tokenizer\n * [Cosmos-0.1-Tokenizer-DV4x8x8](https://huggingface.co/nvidia/Cosmos-Tokenizer-DV4x8x8) (4x temporal compression, 8x8 spatial compression)\n * [Cosmos-1.0-Tokenizer-DV8x8x8](https://huggingface.co/nvidia/Cosmos-Tokenizer-DV8x8x8) (8x temporal compression, 8x8 spatial compression)\n * [Cosmos-1.0-Tokenizer-DV8x16x16](https://huggingface.co/nvidia/Cosmos-Tokenizer-DV8x16x16) (8x temporal compression, 16x16 spatial compression)\n\n\n\n### License/Terms of Use:\n[NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/)\n\nUnder the NVIDIA Open Model License, NVIDIA confirms:\n\n* Models are commercially usable.\n* You are free to create and distribute Derivative Models.\n* NVIDIA does not claim ownership to any outputs generated using the Models or Derivative Models.\n\n## Model Architecture:\n\nWe designed Cosmos Tokenizer using a lightweight and computationally efficient architecture, featuring a temporally causal design. Specifically, we employ causal temporal convolution and causal temporal attention layers to preserve the natural temporal order of video frames, ensuring seamless tokenization of images and videos using a single unified network architecture. The encoder and decoder form a symmetrical pair, which are mirrors of each other. The encoder starts with a 2-level [Haar wavelet](https://link.springer.com/book/10.1007/978-3-319-04295-4) transform layer, which down-samples inputs by a factor of 4 in both spatial and temporal dimensions. Likewise, the decoder ends with an inverse wavelet transform. We employ the vanilla autoencoder (AE) formulation to model the latent space for continuous tokenizers. For discrete tokenizers, we adopt the [Finite-Scalar-Quantization](https://openreview.net/forum?id=8ishA3LxN8) (FSQ) as the latent space quantizer.\n\n![image/jpeg](https://cdn-uploads.huggingface.co/production/uploads/638fb8cf2380ffd99caf8c2a/gQH5n9iCEtqZc7uutUwdL.jpeg)\n\n\n\n## Input/Output Specifications\n\n### Encoder\n* **Input**\n * **Type:** Images or Videos\n * **Format:** RGB (Red, Green, Blue)\n * **Properties:**\n * **Resolution:** Minimum: 256px (shorter side). Maximum: Up to 4K\n * **Video Length:** Up to 8 seconds for 1080p videos (bounded by A100 80G GPU memory; higher resolutions will have shorter supported durations)\n\n* **Output**\n * **Type:** Tokens\n * **Properties:**\n * Integer indices ranging from 0 to 63,999\n\n### Decoder\n* **Input**\n * **Type:** Tokens\n * **Properties:**\n * Integer indices ranging from 0 to 63,999\n\n* **Output**\n * **Type** Images or Videos (matching input type)\n * **Format:** RGB (Red, Green, Blue)\n * **Properties:**\n * **Resolution:** Same as input resolution\n * **Video Length:** Same as input video length\n\n## Software Integration (Required For NVIDIA Models Only):\n**Runtime Engine(s):**\n* [Cosmos-Tokenizer](https://github.com/NVIDIA/Cosmos-Tokenizer)\n\n**Supported Hardware Microarchitecture Compatibility:**\n* NVIDIA Ampere (e.g., A100)\n* NVIDIA Hopper (e.g., H100)\n\nNote: We have only tested Cosmos Tokenizer with BF16 precision on Ampere and Hopper GPUs. If you are using older versions of NVIDIA GPUs (e.g., NVIDIA Volta GPUs), you may need to switch to FP32 precision.\n\n\n**Operating System(s):**\n* Linux (We have not tested on other operating systems.)\n\n# Usage\nInference Engines:\n* [Cosmos-Tokenizer](https://github.com/NVIDIA/Cosmos-Tokenizer) (PyTorch)\n\n## Inference with `Cosmos-Tokenizer` (PyTorch)\n### Step-1: Installation of `Cosmos-Tokenizer`\nNote: Currently, the `Cosmos-Tokenizer` code is only supported on Linux.\n\n- Please clone the `Cosmos-Tokenizer` from GitHub repo [github.com/NVIDIA/Cosmos-Tokenizer](https://github.com/NVIDIA/Cosmos-Tokenizer).\n\n ```bash\n git clone https://github.com/NVIDIA/Cosmos-Tokenizer.git\n cd Cosmos-Tokenizer\n ```\n- Install dependencies\n\n ```bash\n pip3 install -r requirements.txt\n apt-get install -y ffmpeg\n ```\n\n- Preferably, you could build a docker image using our provided Dockerfile.\n ```bash\n docker build -t cosmos-docker -f Dockerfile. \n # You can run the container as:\n docker run --gpus all -it --rm -v /home/${USER}:/home/${USER} \\\n --workdir ${PWD} cosmos-docker /bin/bash\n ```\n\n### Step-2: Download Pre-trained Checkpoints\n- Create a local directory for the pre-trained checkpoints and download the\npre-trained checkpoints from HuggingFace.\n\n ```python\n from huggingface_hub import login, snapshot_download\n import os\n # You could get your Hugging Face token from https://huggingface.co/settings/tokens\n login(token=, add_to_git_credential=True)\n # You could specify the tokenizers you want to download.\n model_names = [\n \"Cosmos-1.0-Tokenizer-DV8x16x16\",\n ]\n for model_name in model_names:\n hf_repo = \"nvidia/\" + model_name\n local_dir = \"pretrained_ckpts/\" + model_name\n os.makedirs(local_dir, exist_ok=True)\n print(f\"downloading {model_name} to {local_dir}...\")\n snapshot_download(repo_id=hf_repo, local_dir=local_dir)\n ```\n\n- Under the ech checkpoint directory `pretrained_ckpts/`, we provide the encoder,\ndecoder and the full autoencoder JIT models.\n\n ```bash\n ├── pretrained_ckpts/ \n │ ├── Cosmos-1.0-Tokenizer-DV8x16x16/\n │ │ ├── encoder.jit\n │ │ ├── decoder.jit\n │ │ ├── autoencoder.jit\n │ ...\n ```\n\n### Step-3: Run Inference\nYou can use the following example commands to encode and decode images or videos. For each, the same command works for both continuous and discrete tokenization. Simply provide the proper JIT-compiled ckpt to `checkpoint_enc`, `checkpoint_dec`, or the full autoencoder ckpt to `checkpoint`.\n\n```python\nimport torch\nfrom cosmos_tokenizer.video_lib import CausalVideoTokenizer\nmodel_name = \"Cosmos-1.0-Tokenizer-DV8x16x16\"\ninput_tensor = torch.randn(1, 3, 9, 512, 512).to('cuda').to(torch.bfloat16)\nencoder = CausalVideoTokenizer(checkpoint_enc=f'pretrained_ckpts/{model_name}/encoder.jit')\n(indices, codes) = encoder.encode(input_tensor)\ntorch.testing.assert_close(indices.shape, (1, 2, 32, 32))\ntorch.testing.assert_close(codes.shape, (1, 6, 2, 32, 32))\n\n# The input tensor can be reconstructed by the decoder as:\ndecoder = CausalVideoTokenizer(checkpoint_dec=f'pretrained_ckpts/{model_name}/decoder.jit')\nreconstructed_tensor = decoder.decode(indices)\ntorch.testing.assert_close(reconstructed_tensor.shape, input_tensor.shape)\n```\n\nThe `indices` will have the shape `(1, 2, 32, 32)` and contain integral values in the range `[1..64K]`, where the first of the three integral maps represents the first frame.\nThe `codes` will contain the pre-quantization continuous latent with shape `(1, 6, 2, 32, 32)`, where C=6 represents the number of FSQ levels.\n\n**Note**: More inference usage commands, including both TorchScript (JIT) and PyTorch Inference APIs on real images and videos, can be found on our GitHub repository [github.com/NVIDIA/Cosmos-Tokenizer](https://github.com/NVIDIA/Cosmos-Tokenizer).\n\n\n# Evaluation\n\n## TokenizationPerformance Comparison\nWe have extensively evaluated the additional **Cosmos Tokenizer** models on [DAVIS](https://davischallenge.org/) video benchmark dataset.\n\n| Tokenizer | Compression Ratio | Height | Num. of Frames | Quantization | PSNR (DAVIS) | SSIM (DAVIS) | rFVD (DAVIS) |\n|-----------|------------------|--------------|--------------|--------------|--------------|--------------|--------------|\n| VideoGPT | 4×4×4 | - | - | VQ | 32.23 | **0.850** | 72.33 |\n| OmniTokenizer | 4×8×8 | - | - | VQ | 28.44 | 0.712 | 188.60 |\n| Cosmos-Tokenizer-DV | 4×8×8 | 720 | 17 | FSQ | **32.98** | 0.818 | **37.36** |\n| Cosmos-Tokenizer-DV | 8×8×8 | 720 | 17 | FSQ | 32.11 | 0.775 | 100.15 |\n| Cosmos-Tokenizer-DV | 8×16×16 | 720 | 17 | FSQ | 31.42 | 0.716 | 241.52 |\n| Cosmos-Tokenizer-DV | 8×16×16 | 720 | 49 | FSQ | 31.59 | 0.719 | 259.33 |\n\n\n* We compare with the state-of-the-art discrete video tokenizer, [OmniTokenizer](https://github.com/FoundationVision/OmniTokenizer).\n* Evaluation metrics:\n * Peak Signal-to-Noise Ratio (PSNR)\n * Structural Similarity (SSIM)\n * Reconstruction Fréchet Video Distance (rFVD)\n\n## Runtime Comparison\n\nThe following table shows the number of parameters and the averaged encoding and decoding times per image or video frame, measured on a single A100 80GB GPU. For comparison, we also list the parameters and average speeds of prior state-of-the-art tokenizer(s) with the same compression ratio.\n\n| Tokenizer | Resolution | Compression Ratio | Parameters | Time (ms) |\n|----------------|------------|-------------------|------------|-----------|\n| OmniTokenizer | 720x1280 | 4×8×8 | 54M | 53.2 |\n| Cosmos-DV | 720x1280 | 4×8×8 | 105M | 51.5 |\n\nNote: We benchmarked the runtime for images under the 8x8 compression and videos under the 4×8×8 compression. Tokenizers with different compression ratios are not included in this comparison.\n\n## Ethical Considerations\nNVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse. \n\nFor more detailed information on ethical considerations for this model, please see the subcards of Explainability, Bias, Safety & Security, and Privacy below. Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).","Cosmos-1.0-Tokenizer-DV8x16x16",["D",1736232722098],"1.0","/api/image/nvidia/cosmos/models/cosmos-1.0-tokenizer-dv8x16x16",447945175,"bias","Field | Response\n:---------------------------------------------------------------------------------------------------|:---------------\nParticipation considerations from adversely impacted groups [protected classes](https://www.senate.ca.gov/content/protected-classes) in model design and testing: | None\nMeasures taken to mitigate against unwanted bias: | None","explainability","Field | Response\n:------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------\nIntended Application & Domain: | Tokenization of images and videos\nModel Type: | Auto-Encoder\nIntended Users: | Generative AI developers for image and video generation models\nOutput: | Images/Videos and Latent Tokens\nDescribe how the model works: | Compresses and decompresses visual input (image/video).\nTechnical Limitations: | Due to tokenizer compression limitations, some visual information (such as small text and other structured fine details) may not be reconstructed accurately. The tokenizers may not produce as high of a reconstruction results for videos with low resolution, e.g. less than 320p.\nVerified to have met prescribed NVIDIA quality standards: | Yes\nPerformance Metrics: | Peak Signal-to-Noise Ratio (PSNR), Structural Similarity (SSIM), Reconstruction Fréchet Video Distance (rFVD), Reconstruction Fréchet Inception Distance (rFID), Latency\nPotential Known Risks: | Tokenizer's output can parse all forms of input, including what may be considered toxic, offensive, or indecent.\nLicensing: | [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/)","privacy","Field | Response\n:----------------------------------------------------------------------------------------------------------------------------------|:-----------------------------------------------\nGeneratable or reverse engineerable personal information? | No\nProtected class data used to create this model? | None Known\nWas consent obtained for any personal data used? | None Known\nHow often is dataset reviewed? | Before Release\nIs a mechanism in place to honor data subject right of access or deletion of personal data? | Not Applicable\nIf personal collected for the development of the model, was it collected directly by NVIDIA? | Not Applicable\nIf personal collected for the development of the model by NVIDIA, do you maintain or have access to disclosures made to data subjects? | Not Applicable\nIf personal collected for the development of this AI model, was it minimized to only what was required? | Not Applicable\nIs there provenance for all datasets used in training? | Yes\nDoes data labeling (annotation, metadata) comply with privacy laws? | Yes\nIs data compliant with data subject requests for data correction or removal, if such a request was made? | Not Applicable","safetyAndSecurity","Field | Response\n:---------------------------------------------------|:----------------------------------\nModel Application(s): | Tokenization of images and videos\nDescribe the life critical impact (if present). | None Known\nUse Case Restrictions: | See [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/)\nModel and dataset restrictions: | The Principle of least privilege (PoLP) is applied limiting access for dataset generation and model development. Restrictions enforce dataset access during training, and dataset license constraints adhered to. Model checkpoints are made available on Hugging Face, and may become available on cloud providers' model catalog.",[],[259],[],[],[],{"_92":145,"_146":145,"_147":260,"_149":150},{"_151":150,"_146":145},{"_178":1225,"_180":1226},"ngcApps",[1178],"csaApiBaseUrl","https://www.nvidia.com/ai-agents/csa-agent/dcsm","datadogRum",{"_1159":1160,"_1161":1162,"_1163":1164,"_1165":1166,"_1167":272,"_210":296,"_1168":1169,"_1170":51,"_1171":221,"_1172":221,"_1173":1174,"_1175":221,"_1176":1177},"amplitude",{"_1156":1157,"_809":-7,"_1158":26},"disableJs","environment","prod","featureFlagsBlob","JTdCJTIydGVzdEZsYWclMjIlM0FmYWxzZSUyQyUyMnNob3dTZWFyY2hMYWJlbFNldEZpbHRlcnMlMjIlM0FmYWxzZSUyQyUyMmVuYWJsZUNzYVdpZGdldCUyMiUzQWZhbHNlJTJDJTIyZW5hYmxlUHJvZHVjdEdyb3VwaW5nQnlTdWJzY3JpcHRpb24lMjIlM0FmYWxzZSUyQyUyMmNvbnRhaW5lclNjYW5uaW5nUmVkZXNpZ25NMSUyMiUzQXRydWUlN0Q=","featureFlagsGuarded","isDevLikeEnv","isPlaywright","isMobileDevice","categorizedOrgs",{"_1150":1151,"_1152":1153,"_1154":1155},"currentOrgContext","currentContextRoles",[],"enterpriseOrgs",[],"locale","en","ngcApiOrigin","ngcHelmOrigin","origin","https://catalog.ngc.nvidia.com","usesMocks","theme","year",2026,"1.10.0","defaultSearchSuggestions",{"_1049":377,"_1050":1051,"_1052":1053,"_1054":1055},"announcements",[],"consentState",{"_334":1039,"_1040":221,"_1041":26,"_1042":26,"_1043":1044},"consentConfig",{"_316":317,"_318":319,"_320":321,"_322":323,"_324":325,"_326":327,"_328":329,"_330":331,"_332":333,"_334":335},"showContactAdmin",1785756706704,[],[],[310,311,312,313,314,315],{"_40":55,"_42":67,"_44":63,"_46":-5,"_48":-5,"_50":51},{"_40":55,"_42":64,"_44":63,"_46":-5,"_48":-5,"_50":51},{"_40":55,"_42":60,"_44":63,"_46":-5,"_48":-5,"_50":51},{"_40":55,"_42":56,"_44":63,"_46":-5,"_48":-5,"_50":51},{"_40":52,"_42":53,"_44":63,"_46":-5,"_48":-5,"_50":51},{"_40":41,"_42":43,"_44":45,"_46":47,"_48":54,"_50":51},"mainInfoText","We and our third-party partners (including social media, advertising, and analytics partners) use cookies and other tracking technologies to collect, store, monitor, and process certain information about you when you visit our website. The information collected might relate to you, your preferences, or your device. We use that information to make the site work, analyze performance and traffic on our website, provide a more personalized web experience, and assist in our marketing efforts.

Under certain privacy laws, you have the right to direct us not to \"sell\" or \"share\" your personal information for targeted advertising. To opt-out of the \"sale\" and \"sharing\" of personal information through cookies, you must opt-out of optional cookies using the toggles below. To opt out of the \"sale\" and \"sharing\" of data collected by other means (e.g., online forms) you must also update your data sharing preferences through the NVIDIA Preference Center.

Click on the different category headings below to find out more and change the settings according to your preference. You cannot opt out of Required Cookies as they are deployed to ensure the proper functioning of our website (such as prompting the cookie banner and remembering your settings, etc.). By clicking \"Save and Accept\" or \"Decline All\" at the bottom, you consent to the use of cookies and other tools as described in our Cookie Policy in accordance with your settings and accept our Terms of Service (which contains important waivers). For more information about our privacy practices, please see our Privacy Policy.","mainText","Cookie Settings","alertNoticeText",{"_1031":1032,"_1033":1034,"_1035":1036,"_1037":1038},"bannerAcceptLabel","Accept All","bannerRejectLabel","Reject Optional","bannerSettingsLabel","Manage Settings","preferenceSaveLabel","Save Choices","preferenceRejectLabel","Decline All","alwaysActiveText","Always Active","groups",[336,337,338,339,340],{"_341":721,"_125":722,"_91":723,"_345":221,"_346":16,"_348":724,"_350":725},{"_341":624,"_125":625,"_91":626,"_345":26,"_346":49,"_348":627,"_350":628},{"_341":532,"_125":533,"_91":534,"_345":26,"_346":54,"_348":535,"_350":536},{"_341":352,"_125":353,"_91":354,"_345":26,"_346":355,"_348":356,"_350":357},{"_341":342,"_125":343,"_91":344,"_345":26,"_346":347,"_348":349,"_350":351},"optanonGroupId","C0005","Social Media Cookies","These cookies are set by a range of social media services that we have added to the site to enable you to share our content with your friends and networks. They are capable of tracking your browser across other sites and building up a profile of your interests. This may impact the content and messages you see on other websites you visit. If you do not allow these cookies you may not be able to use or see these sharing tools.","alwaysActive","order",11,"cookies",[],"hosts",[],"C0004","Advertising Cookies","These cookies record your visit to our websites, the pages you have visited and the links you have followed to influence the advertisements that you see on other websites. These cookies and the information they collect may be managed by other companies, including our advertising partners, and may be used to build a profile of your interests and show you relevant advertising on other sites. We and our advertising partners will use this information to make our websites and the advertising displayed on it, more relevant to your interests.",5,[474,475,476,477,478,479,480,481,482,483,484,485,486,487,488,489,490,491,492,493,494,495],[358,359,360,361,362,363,364,365,366,367,368,369,370,371,372,373,374],{"_375":469,"_92":469,"_91":377,"_348":470},{"_375":465,"_92":465,"_91":377,"_348":466},{"_375":460,"_92":460,"_91":377,"_348":461},{"_375":456,"_92":456,"_91":377,"_348":457},{"_375":451,"_92":451,"_91":377,"_348":452},{"_375":446,"_92":446,"_91":377,"_348":447},{"_375":438,"_92":438,"_91":377,"_348":439},{"_375":430,"_92":430,"_91":377,"_348":431},{"_375":425,"_92":425,"_91":377,"_348":426},{"_375":420,"_92":420,"_91":377,"_348":421},{"_375":415,"_92":415,"_91":377,"_348":416},{"_375":410,"_92":410,"_91":377,"_348":411},{"_375":405,"_92":405,"_91":377,"_348":406},{"_375":398,"_92":398,"_91":377,"_348":399},{"_375":393,"_92":393,"_91":377,"_348":394},{"_375":386,"_92":386,"_91":377,"_348":387},{"_375":376,"_92":376,"_91":377,"_348":378},"hostName","amazon-adsystem.com","",[379,380],{"_125":385,"_382":376,"_383":26,"_91":384},{"_125":381,"_382":376,"_383":26,"_91":384},"ad-privacy","host","isSession","This domain is owned by online retailer Amazon and is used as part of its affiliate marketing programme.","ad-id","doubleclick.net",[388,389],{"_125":392,"_382":386,"_383":26,"_91":391},{"_125":390,"_382":386,"_383":26,"_91":391},"IDE","This domain is owned by Doubleclick (Google). The main business activity is: Doubleclick is Googles real time bidding advertising exchange","test_cookie","bing.com",[395],{"_125":396,"_382":393,"_383":26,"_91":397},"MUID","This domain is owned by Mircosoft - it is the site for the search engine Bing.","assets.adobedtm.com",[400,401],{"_125":404,"_382":398,"_383":26,"_91":403},{"_125":402,"_382":398,"_383":26,"_91":403},"kndctr_F207D74D549850760A4C98C6_AdobeOrg_*","If you know anything about this host and its uses please add your contributions here. ","kndctr_F207D74D549850760A4C98C6_AdobeOrg_consent","tiktok.com",[407],{"_125":408,"_382":405,"_383":26,"_91":409},"_ttp","TikTok is a social media platform owned by ByteDance, known for short-form video sharing. The service is used to create and share videos, and it utilizes cookies for various purposes including advertising and analytics.","t.co",[412],{"_125":413,"_382":410,"_383":26,"_91":414},"muc_ads","This domain is owned by Twitter. It is used as a URL shortening service, which allows Twitter to collect data on how links are shared.","snowplow.com",[417],{"_125":418,"_382":415,"_383":26,"_91":419},"sp","Testing","google.com",[422],{"_125":423,"_382":420,"_383":26,"_91":424},"NID","This domain is owned by Google Inc. Although Google is primarily known as a search engine, the company provides a diverse range of products and services. Its main source of revenue however is advertising. Google tracks users extensively both through its own products and sites, and the numerous technologies embedded into many millions of websites around the world. It uses the data gathered from most of these services to profile the interests of web users and sell advertising space to organisations based on such interest profiles as well as aligning adverts to the content on the pages where its customer's adverts appear.","linkedin.com",[427],{"_125":428,"_382":425,"_383":26,"_91":429},"bcookie","This domain is owned by LinkedIn, the business networking platform. It typically acts as a third party host where website owners have placed one of its content sharing buttons in their pages, although its content and services can be embedded in other ways. Although such buttons add functionality to the website they are on, cookies are set regardless of whether or not the visitor has an active Linkedin profile, or agreed to their terms and conditions. For this reason it is classified as a primarily tracking/targeting domain.","twitter.com",[432,433],{"_125":436,"_382":430,"_383":26,"_91":437},{"_125":434,"_382":430,"_383":26,"_91":435},"personalization_id","This domain is owned by Twitter. The main business activity is: Social Networking Services. Where twitter acts as a third party host, it collects data through a range of plug-ins and integrations, that is primarily used for tracking and targeting.","guest_*","This domain is owned by X (formerly Twitter). The main business activity is: Social Networking Services. Where X (formerly Twitter) acts as a third party host, it collects data through a range of plug-ins and integrations, that is primarily used for tracking and targeting.","youtube.com",[440,441],{"_125":444,"_382":438,"_383":26,"_91":445},{"_125":442,"_382":438,"_383":26,"_91":443},"VISITOR_INFO1_LIVE","This cookie is used as a unique identifier to track viewing of videos","__Secure-*","YouTube is a Google owned platform for hosting and sharing videos. YouTube collects user data through videos embedded in websites, which is aggregated with profile data from other Google services in order to display targeted advertising to web visitors across a broad range of their own and other websites.","demdex.net",[448],{"_125":449,"_382":446,"_383":26,"_91":450},"demdex","This cookie helps Adobe Audience Manger perform basic functions such as visitor identification, ID synchronization, segmentation, modeling, reporting, etc.","im-apps.net",[453],{"_125":454,"_382":451,"_383":26,"_91":455},"receive-cookie-deprecation","This domain is owned by im-apps.net which provides information and resources about various instant messaging (IM) applications to help users make informed decisions based on their communication needs.","influ2.com",[458],{"_125":459,"_382":456,"_383":26,"_91":419},"R","6sc.co",[462],{"_125":463,"_382":460,"_383":26,"_91":464},"6suuid","This domain is associated with the company 6Sense Insights, Inc., which provides predictive analytics and account engagement services. It is primarily used for tracking and targeting in marketing automation and advertising.","hushly.com",[467],{"_125":468,"_382":465,"_383":26,"_91":419},"_hly_gvid","t.contentsquare.net",[471],{"_125":472,"_382":469,"_383":221,"_91":473},"_cs_*","This is a session cookie. This is a pattern type cookie where a random 13-digit number is appended to the prefix _cs_.",{"_125":530,"_382":497,"_383":26,"_91":531},{"_125":528,"_382":497,"_383":26,"_91":529},{"_125":526,"_382":497,"_383":26,"_91":527},{"_125":408,"_382":497,"_383":26,"_91":525},{"_125":502,"_382":497,"_383":221,"_91":504},{"_125":524,"_382":497,"_383":26,"_91":377},{"_125":523,"_382":500,"_383":26,"_91":377},{"_125":521,"_382":497,"_383":26,"_91":522},{"_125":519,"_382":497,"_383":26,"_91":520},{"_125":518,"_382":497,"_383":26,"_91":377},{"_125":516,"_382":497,"_383":26,"_91":517},{"_125":502,"_382":515,"_383":221,"_91":504},{"_125":513,"_382":500,"_383":26,"_91":514},{"_125":513,"_382":503,"_383":26,"_91":514},{"_125":472,"_382":497,"_383":26,"_91":473},{"_125":511,"_382":497,"_383":26,"_91":512},{"_125":418,"_382":509,"_383":26,"_91":510},{"_125":507,"_382":497,"_383":26,"_91":508},{"_125":505,"_382":497,"_383":26,"_91":506},{"_125":502,"_382":503,"_383":221,"_91":504},{"_125":499,"_382":500,"_383":26,"_91":501},{"_125":496,"_382":497,"_383":221,"_91":498},"AMCVS_xxxxxAdobeOrg","nvidia.com","This is a pattern type cookie name associated with Adobe Marketing Cloud. It stores a unique visitor identifier, and uses an organisation identifier to allow a company to track users across their domains and services.This is a pattern type cookie name associated with Adobe Marketing Cloud. It stores a unique visitor identifier, and uses an organisation identifier.","_rdt_pn","www.nvidia.com","This cookie name is associated with www.mind.org.uk which is used for marketing purposes.","AMP_*","build.nvidia.com","Supports Amplitude product analytics.","_gcl_au","Used by Google AdSense for experimenting with advertisement efficiency across websites using their services","_mkto_trk","This cookie is associated with an email marketing service provided by Marketo. This tracking cookie allows a website to link visitor behaviour to the recipient of an email marketing campaign, to measure campaign effectiveness.","investor.nvidia.com","This cookie is associated with Eventbrite and is used to deliver content tailored to the end user's interests and improve content creation. This cookie is also used for event-booking purposes.","_gat_gtag_UA_140825383_2","Google Analytics","_rdt_uuid","This cookie is set by Reddit and is used for tracking the views of embedded videos and performance of the advertisements.","catalog.ngc.nvidia.com","_uetsid","This cookie is used by Bing to determine what ads should be shown that may be relevant to the end user perusing the site. ","kndctr_F207D74D549850760A4C98C6_AdobeOrg_bc_session_id","ttcsid_XXXXXXXXXXXXXXXXXXXXCategory","This cookie name is associated with TikTok which is used to track your activity on the platform, including how you interact with their embedded services on other websites.This cookie is associated with TikTok. It helps track your activity on the platform, including how you interact with their embedded services on other websites.","_fbp","Used by Facebook to deliver a series of advertisement products such as real time bidding from third party advertisers","_zitok","_twpid","This cookie is typically associated with the use of Taptap Digital, a company specializing in digital advertising. It is used for tracking and ad personalization purposes.","_tt_enable_cookie","This cookie name is associated with TikTok which is used to collect data about behavior and purchases on the website and to measure the effect of advertising. It is also used to track website activity to help optimize advertising efforts.","AMCV_xxxxxAdobeOrg","This cookie is used to identify a unique visitor for tracking purposes by Adobe Experience Cloud.This is a pattern type cookie name associated with Adobe Marketing Cloud. It stores a unique visitor identifier, and uses an organisation identifier to allow a company to track users across their domains and services.","ttcsid","This cookie name is associated with TikTok which is used to enable TikTok Ads session tracking by assigning a unique session ID to users who click on TikTok ads, allowing for the attribution of user behavior and conversions within the same session.This cookie is associated with TikTok. It helps track your activity on the platform, including how you interact with their embedded services on other websites.","C0003","Personalization Cookies","These cookies collect data about how you have interacted with our website to help us improve your web experience, such as which pages you have visited. These cookies may store a unique ID so that our system will remember you when you return. They may be set by us or by third party providers whose services we have added to our pages. These cookies enable us to provide enhanced website functionality and personalization as well as make the marketing messages we send to you more relevant to your interests. If you do not allow these cookies, then some or all of these services may not function properly.",[566,567,568,569,570,571,572,573,574,575,576,577,578,579,580,581,582,583,584,585,586,587,588,589,590],[537,538,539,540,541],{"_375":451,"_92":451,"_91":377,"_348":561},{"_375":425,"_92":425,"_91":377,"_348":556},{"_375":551,"_92":551,"_91":377,"_348":552},{"_375":546,"_92":546,"_91":377,"_348":547},{"_375":542,"_92":542,"_91":377,"_348":543},"nvidia.partners",[544],{"_125":545,"_382":542,"_383":26,"_91":419},"edge","app.hushly.com",[548],{"_125":549,"_382":546,"_383":26,"_91":550},"*CORS","This domain is associated with Hushly, a platform that provides lead conversion solutions for content marketing by using personalization and content engagement strategies.","github.com",[553],{"_125":554,"_382":551,"_383":26,"_91":555},"_octo","This domain is owned by Github. The main business activity is: Source code repository and collaboration platform",[557,558],{"_125":560,"_382":425,"_383":26,"_91":429},{"_125":559,"_382":425,"_383":26,"_91":429},"lidc","li_gc",[562,563],{"_125":565,"_382":451,"_383":26,"_91":455},{"_125":564,"_382":451,"_383":26,"_91":455},"imid_*","imid",{"_125":591,"_382":500,"_383":26,"_91":593},{"_125":622,"_382":497,"_383":26,"_91":623},{"_125":591,"_382":621,"_383":26,"_91":593},{"_125":619,"_382":497,"_383":221,"_91":620},{"_125":617,"_382":497,"_383":26,"_91":618},{"_125":616,"_382":599,"_383":221,"_91":377},{"_125":591,"_382":595,"_383":26,"_91":593},{"_125":614,"_382":497,"_383":26,"_91":615},{"_125":591,"_382":503,"_383":26,"_91":593},{"_125":602,"_382":500,"_383":26,"_91":377},{"_125":591,"_382":613,"_383":26,"_91":593},{"_125":611,"_382":497,"_383":26,"_91":612},{"_125":608,"_382":609,"_383":26,"_91":610},{"_125":606,"_382":607,"_383":26,"_91":377},{"_125":591,"_382":603,"_383":26,"_91":593},{"_125":605,"_382":497,"_383":26,"_91":377},{"_125":591,"_382":604,"_383":26,"_91":593},{"_125":602,"_382":603,"_383":26,"_91":377},{"_125":601,"_382":599,"_383":221,"_91":377},{"_125":591,"_382":600,"_383":26,"_91":593},{"_125":591,"_382":599,"_383":26,"_91":593},{"_125":591,"_382":598,"_383":26,"_91":593},{"_125":597,"_382":497,"_383":221,"_91":593},{"_125":594,"_382":595,"_383":221,"_91":596},{"_125":591,"_382":592,"_383":26,"_91":593},"_gd_*","info.nvidia.com","This cookie contains no identifying information and is used for diagnostic purposes by Impact Radius, a service we use to track signup commissions for our affiliate program. We have categorized it under Functionality because it is necessary for the operation of our affiliate program.","nv_country_code","developer.nvidia.com","This cookie is likely used to store the country code of the visitor to provide localized content and possibly for legal compliance with regional data privacy laws.","_gdxxxxxxxxxxxxx","nvidianews.nvidia.com","docs.api.nvidia.com","research.nvidia.com","readme_library","wcmmode","ai-podcast.nvidia.com","resources.nvidia.com","geo_locator","openedx-language-preference","learn.nvidia.com","doxygen_width","raytracing-docs.nvidia.com","This cookie is likely used to remember the width preferences of a sidebar or interface element within a documentation platform like Doxygen.","_sfid_eb78","This cookie is categorized as Functionality Cookie.","crowdsource.nvidia.com","mbox","Cookie set by Adobe SiteCatalyst analytics software - used for measuring the performance of page content using A/B split testing","readme_language","vid","This cookie name 'vid' is used for tracking the search request made by the user on the 'Find a practitioner' page.","at_check","Used by Adobe Target to check if cookies are enabled/supported on the browser","blogs.nvidia.com","nvHeadless","This cookie is likely used to track if a user is using the site in a 'headless' non-GUI mode, which could be for accessibility reasons or for automated processes.","C0002","Performance Cookies","These cookies are used to provide quantitative measures of our website visitors, such as the number of times you visit, time on page, your mouse movements, scrolling, clicks and keystroke activity on the websites; other browsing, search, or product research behavior; and what brought you to our site. These cookies may store a unique ID so that our system will remember you when you return. Information collected with these cookies is used to measure and find ways to improve website performance.",[649,650,651,652,653,654,655,656,657,658,659,660,661,662,663,664,665,666,667,668,669,670,671,672,673,674,675,676,677],[629,630,631,632],{"_375":644,"_92":644,"_91":377,"_348":645},{"_375":639,"_92":639,"_91":377,"_348":640},{"_375":438,"_92":438,"_91":377,"_348":636},{"_375":551,"_92":551,"_91":377,"_348":633},[634],{"_125":635,"_382":551,"_383":221,"_91":555},"_gh_sess",[637],{"_125":638,"_382":438,"_383":221,"_91":445},"YSC","bizible.com",[641],{"_125":642,"_382":639,"_383":26,"_91":643},"_BUID","Bizible is a marketing attribution software that tracks the performance of marketing efforts and connects them to revenue, primarily used for B2B environments.","a.usbrowserspeed.com",[646],{"_125":647,"_382":644,"_383":26,"_91":648},"tuid","This domain is owned by Ookla which provides internet speed testing services through its Speedtest application.",{"_125":719,"_382":497,"_383":26,"_91":720},{"_125":704,"_382":718,"_383":221,"_91":705},{"_125":687,"_382":500,"_383":26,"_91":612},{"_125":687,"_382":600,"_383":26,"_91":612},{"_125":716,"_382":497,"_383":26,"_91":717},{"_125":704,"_382":503,"_383":221,"_91":705},{"_125":714,"_382":497,"_383":26,"_91":715},{"_125":704,"_382":500,"_383":221,"_91":705},{"_125":712,"_382":599,"_383":26,"_91":713},{"_125":704,"_382":621,"_383":221,"_91":705},{"_125":684,"_382":500,"_383":26,"_91":686},{"_125":710,"_382":497,"_383":221,"_91":711},{"_125":708,"_382":497,"_383":26,"_91":709},{"_125":706,"_382":497,"_383":221,"_91":707},{"_125":704,"_382":595,"_383":221,"_91":705},{"_125":702,"_382":497,"_383":26,"_91":703},{"_125":700,"_382":497,"_383":26,"_91":701},{"_125":698,"_382":497,"_383":221,"_91":699},{"_125":696,"_382":604,"_383":26,"_91":697},{"_125":684,"_382":595,"_383":26,"_91":686},{"_125":694,"_382":497,"_383":26,"_91":695},{"_125":692,"_382":497,"_383":221,"_91":693},{"_125":690,"_382":497,"_383":26,"_91":691},{"_125":688,"_382":497,"_383":26,"_91":689},{"_125":687,"_382":497,"_383":26,"_91":612},{"_125":684,"_382":685,"_383":26,"_91":686},{"_125":682,"_382":497,"_383":26,"_91":683},{"_125":680,"_382":497,"_383":221,"_91":681},{"_125":678,"_382":497,"_383":26,"_91":679},"_ga_xxxxxxxxxx","Used by Google Analytics to identify and track an individual session with your device.","s_ppvl","This cookie is set by Adobe Analytics to store information on the percentage of the page displayed to the end user.","coveo_visitorId","This cookie name is associated with Coveo which is used to store a unique user ID for marketing purposes and to track the number of distinct users who submitted a query or clicked an item, counting them only once even if they did the same query or clicked the same item over the course of many visits.","_dd_s","marketplace.nvidia.com","Datadog Cookie used to group all events generated from a unique user session across multiple pages. It contains the current session ID, whether the session is excluded due to sampling, and the expiration date of the session. The cookie is extended for an extra 15 minutes every time the user interacts with the website, up to the maximum user session duration (4 hours).","nvweb_A","_ga","This cookie name is associated with Google Universal Analytics - which is a significant update to Google's more commonly used analytics service. This cookie is used to distinguish unique users by assigning a randomly generated number as a client identifier. It is included in each page request in a site and used to calculate visitor, session and campaign data for the sites analytics reports. By default it is set to expire after 2 years, although this is customisable by website owners.","_uetvid","This is a cookie utilised by Microsoft Bing Ads and is a tracking cookie. It allows us to engage with a user that has previously visited our website.","tp","This cookie is associated with Adobe Analytics site tracking. It registers if the website's scroll-depth detection is active. This function remembers how far the user has scrolled on the current session across the website's sub-pages.","_hly_vid","This cookie _hly_vid is used to record website and server performance with each visit to the site. The data is aggregated and analyzed; therefore, it remains anonymous upon use.","_session_id","This cookie creates an interim session ID used as an in-session user ID.\n","s_cc","This cookie is associated with the Adobe Site Catalyst. It determines whether cookies are enabled in the web browser.","s_ecid","Adobe Analytics cookie containing the the Adobe Experience Cloud 1st party visitor id if the Analytics AMCV visitor id is not found.","_hjSessionUser_xxxxxx","Hotjar cookie that is set when a user first lands on a page with the Hotjar script. It is used to persist the Hotjar User ID, unique to that site on the browser. This ensures that behavior in subsequent visits to the same site will be attributed to the same user ID.","_hjHasCachedUserAttributes","This cookie name is associated with Hotjar which is used to know whether the data set in _hjUserAttributes Local Storage item is up to date or not.","s_ppv","This cookie is associated with Adobe Analytics site tracking and is used to measure a user's scroll activity to see how much of a page they view before moving to another page.","_gid","This cookie name is associated with Google Universal Analytics. This appears to be a new cookie and as of Spring 2017 no information is available from Google. It appears to store and update a unique value for each page visited.","_hjTLDTest","When the Hotjar script executes we try to determine the most generic cookie path we should use, instead of the page hostname. This is done so that cookies can be shared across subdomains (where applicable). To determine this, we try to store the _hjTLDTest cookie for different URL substring alternatives until it fails. After this check, the cookie is removed.","ekfls-2025-03-27","This cookie name is associated with readme.io which is used to help maintain data for the API reference support page.","s_getNewRepeat","This cookie is used to determine whether a visitor is a new visitor or a repeat visitor to the site, and is used for analytical purposes.","_hly_sid","This cookie is likely associated with website analytics, as the prefix '_hly_' suggests a connection to Holistics, a business intelligence and analytics platform.","docs.nvidia.com","_hjSession_xxxxxx","A cookie that holds the current session data. This ensues that subsequent requests within the session window will be attributed to the same Hotjar session.","C0001","Required Cookies","These cookies enable core functionality such as security, network management, and accessibility. These cookies are required for the site to function and cannot be turned off.",[866,867,868,869,870,871,872,873,874,875,876,877,878,879,880,881,882,883,884,885,886,887,888,889,890,891,892,893,894,895,896,897,898,899,900,901,902,903,904,905,906,907,908,909,910,911,912,913,914,915,916,917,918,919,920,921,922,923,924,925,926,927,928,929,930,931,932,933,934,935,936,937,938],[726,727,728,729,730,731,732,733,734,735,736,737,738,739,740,741,742,743,744,745,746,747,748,749,750],{"_375":863,"_92":863,"_91":377,"_348":864},{"_375":857,"_92":857,"_91":377,"_348":858},{"_375":854,"_92":854,"_91":377,"_348":855},{"_375":842,"_92":842,"_91":377,"_348":843},{"_375":839,"_92":839,"_91":377,"_348":840},{"_375":836,"_92":836,"_91":377,"_348":837},{"_375":831,"_92":831,"_91":377,"_348":832},{"_375":438,"_92":438,"_91":377,"_348":828},{"_375":825,"_92":825,"_91":377,"_348":826},{"_375":430,"_92":430,"_91":377,"_348":823},{"_375":425,"_92":425,"_91":377,"_348":821},{"_375":816,"_92":816,"_91":377,"_348":817},{"_375":811,"_92":811,"_91":377,"_348":812},{"_375":806,"_92":806,"_91":377,"_348":807},{"_375":803,"_92":803,"_91":377,"_348":804},{"_375":551,"_92":551,"_91":377,"_348":800},{"_375":410,"_92":410,"_91":377,"_348":797},{"_375":546,"_92":546,"_91":377,"_348":790},{"_375":542,"_92":542,"_91":377,"_348":784},{"_375":778,"_92":778,"_91":377,"_348":779},{"_375":773,"_92":773,"_91":377,"_348":774},{"_375":768,"_92":768,"_91":377,"_348":769},{"_375":764,"_92":764,"_91":377,"_348":765},{"_375":759,"_92":759,"_91":377,"_348":760},{"_375":751,"_92":751,"_91":377,"_348":752},"polyfill.io",[753,754],{"_125":757,"_382":751,"_383":26,"_91":758},{"_125":755,"_382":751,"_383":26,"_91":756},"cf_ob_info","This cookie is associated with sites using CloudFlare. It is used to improve page load times and to override any security restrictions based on the visitors IP address. It does not contain any user identification information.","cf_use_ob","This cookie is associated with sites using CloudFlare. It is used to improve page load times and to override any security restrictions based on the visitors IP address. It does not contain any user identification information. ","www.recaptcha.net",[761],{"_125":762,"_382":759,"_383":26,"_91":763},"_GRECAPTCHA","This domain is used by Google reCAPTCHA, which provides spam protection services to websites by detecting if a user is a human or a bot.","api2.hcaptcha.com",[766],{"_125":767,"_382":764,"_383":26,"_91":419},"__cflb","instagram.com",[770],{"_125":771,"_382":768,"_383":26,"_91":772},"csrftoken","Instagram is a social networking service owned by Facebook, Inc. It's designed for sharing photos and videos from a smartphone. The service also uses cookies for various purposes, including analytics, personalization, and advertising.","readme.io",[775],{"_125":776,"_382":773,"_383":26,"_91":777},"__cf_bm","The __cf_bm cookie is a cookie necessary to support Cloudflare Bot Management, currently in private beta. As part of our bot management service, this cookie helps manage incoming traffic that matches criteria associated with bots.","nvidiacorp.us-5.evergage.com",[780,781],{"_125":783,"_382":778,"_383":26,"_91":419},{"_125":782,"_382":778,"_383":26,"_91":419},"AWSALBTG","AWSALBTGCORS",[785,786],{"_125":788,"_382":542,"_383":26,"_91":789},{"_125":787,"_382":542,"_383":26,"_91":419},"ak_bmsc","bm_mi","This domain is likely associated with NVIDIA's partner program, which could involve marketing, sales, and support services for partners.",[791,792,793],{"_125":796,"_382":546,"_383":26,"_91":419},{"_125":795,"_382":546,"_383":221,"_91":419},{"_125":794,"_382":546,"_383":26,"_91":419},"AWSALBCORS","JSESSIONID","AWSALB",[798],{"_125":776,"_382":410,"_383":26,"_91":799},"This is a CloudFoundry cookie",[801],{"_125":802,"_382":551,"_383":26,"_91":555},"logged_in","hcaptcha.com",[805],{"_125":776,"_382":803,"_383":26,"_91":799},"tracking.intentsify.io",[808],{"_125":809,"_382":806,"_383":26,"_91":810},"userId","Used to track usage of the embedded video player.","static.reo.dev",[813],{"_125":814,"_382":811,"_383":26,"_91":815},"cookietest","Common cookie name could have a number of different origins. Where this is first party and a session cookie, its most likely to do with checking to see if the browser is set to block or allow cookies.","www.datadoghq-browser-agent.com",[818],{"_125":819,"_382":816,"_383":26,"_91":820},"dd_cookie_test_","dd_cookie_test",[822],{"_125":776,"_382":425,"_383":26,"_91":799},[824],{"_125":776,"_382":430,"_383":26,"_91":799},"supabase.co",[827],{"_125":776,"_382":825,"_383":26,"_91":799},[829],{"_125":830,"_382":438,"_383":26,"_91":445},"VISITOR_PRIVACY_METADATA","dosa-prd.azurewebsites.net",[833],{"_125":834,"_382":831,"_383":26,"_91":835},"access_token","This cookie is used to obtain the permissions for users of the portal to allow for API access in order to retrieve information required to perform portal functions.","w.hcaptcha.com",[838],{"_125":776,"_382":836,"_383":26,"_91":799},"cdn.readme.io",[841],{"_125":776,"_382":839,"_383":26,"_91":799},"queue-it.net",[844,845,846,847,848],{"_125":853,"_382":842,"_383":26,"_91":377},{"_125":852,"_382":842,"_383":26,"_91":377},{"_125":851,"_382":842,"_383":26,"_91":377},{"_125":850,"_382":842,"_383":26,"_91":377},{"_125":849,"_382":842,"_383":26,"_91":377},"Queue-it-{queueId}","Queue-it-{dmonvidia }________{eventId}","Queue-it-token","queueitsoftblock_{queueId}","Queue-it-{nvidia}________{eventId}","x.com",[856],{"_125":776,"_382":854,"_383":26,"_91":777},"api.hcaptcha.com",[859,860],{"_125":862,"_382":857,"_383":26,"_91":419},{"_125":767,"_382":857,"_383":26,"_91":861},"This domain is associated with hCaptcha, a service that provides CAPTCHA and reCAPTCHA alternatives for user verification and bot protection.","hmt_id","vimeo.com",[865],{"_125":776,"_382":863,"_383":26,"_91":799},{"_125":999,"_382":497,"_383":26,"_91":1030},{"_125":1028,"_382":592,"_383":221,"_91":1029},{"_125":1021,"_382":990,"_383":221,"_91":1022},{"_125":1027,"_382":685,"_383":221,"_91":377},{"_125":1025,"_382":497,"_383":26,"_91":1026},{"_125":783,"_382":1011,"_383":26,"_91":950},{"_125":958,"_382":1020,"_383":221,"_91":960},{"_125":782,"_382":949,"_383":26,"_91":950},{"_125":961,"_382":1020,"_383":26,"_91":962},{"_125":1023,"_382":497,"_383":221,"_91":1024},{"_125":1021,"_382":997,"_383":221,"_91":1022},{"_125":955,"_382":1020,"_383":26,"_91":957},{"_125":1018,"_382":497,"_383":26,"_91":1019},{"_125":782,"_382":941,"_383":26,"_91":950},{"_125":946,"_382":1017,"_383":26,"_91":948},{"_125":794,"_382":947,"_383":26,"_91":963},{"_125":796,"_382":947,"_383":26,"_91":1014},{"_125":961,"_382":956,"_383":26,"_91":962},{"_125":1015,"_382":595,"_383":26,"_91":1016},{"_125":796,"_382":621,"_383":26,"_91":1014},{"_125":1012,"_382":497,"_383":26,"_91":1013},{"_125":788,"_382":621,"_383":26,"_91":1007},{"_125":984,"_382":600,"_383":221,"_91":377},{"_125":782,"_382":1011,"_383":26,"_91":950},{"_125":1009,"_382":949,"_383":221,"_91":1010},{"_125":973,"_382":497,"_383":26,"_91":964},{"_125":946,"_382":497,"_383":26,"_91":948},{"_125":1008,"_382":509,"_383":26,"_91":377},{"_125":788,"_382":497,"_383":26,"_91":1007},{"_125":796,"_382":941,"_383":26,"_91":1006},{"_125":1004,"_382":604,"_383":221,"_91":1005},{"_125":1002,"_382":497,"_383":221,"_91":1003},{"_125":1001,"_382":497,"_383":26,"_91":377},{"_125":999,"_382":947,"_383":26,"_91":1000},{"_125":998,"_382":509,"_383":26,"_91":377},{"_125":989,"_382":997,"_383":221,"_91":991},{"_125":995,"_382":600,"_383":26,"_91":996},{"_125":993,"_382":685,"_383":26,"_91":994},{"_125":992,"_382":503,"_383":26,"_91":377},{"_125":989,"_382":990,"_383":221,"_91":991},{"_125":987,"_382":956,"_383":221,"_91":988},{"_125":985,"_382":497,"_383":26,"_91":986},{"_125":984,"_382":500,"_383":221,"_91":377},{"_125":955,"_382":959,"_383":26,"_91":957},{"_125":787,"_382":974,"_383":26,"_91":964},{"_125":982,"_382":500,"_383":26,"_91":983},{"_125":980,"_382":497,"_383":26,"_91":981},{"_125":978,"_382":497,"_383":26,"_91":979},{"_125":977,"_382":497,"_383":26,"_91":377},{"_125":819,"_382":500,"_383":26,"_91":976},{"_125":973,"_382":621,"_383":26,"_91":964},{"_125":776,"_382":592,"_383":26,"_91":975},{"_125":973,"_382":974,"_383":26,"_91":964},{"_125":972,"_382":509,"_383":26,"_91":377},{"_125":971,"_382":497,"_383":26,"_91":944},{"_125":969,"_382":956,"_383":26,"_91":970},{"_125":782,"_382":967,"_383":26,"_91":968},{"_125":966,"_382":509,"_383":26,"_91":377},{"_125":965,"_382":497,"_383":26,"_91":377},{"_125":958,"_382":956,"_383":221,"_91":960},{"_125":787,"_382":621,"_383":26,"_91":964},{"_125":794,"_382":621,"_383":26,"_91":963},{"_125":961,"_382":959,"_383":26,"_91":962},{"_125":958,"_382":959,"_383":221,"_91":960},{"_125":955,"_382":956,"_383":26,"_91":957},{"_125":953,"_382":949,"_383":221,"_91":954},{"_125":951,"_382":607,"_383":26,"_91":952},{"_125":783,"_382":949,"_383":26,"_91":950},{"_125":946,"_382":947,"_383":26,"_91":948},{"_125":776,"_382":599,"_383":26,"_91":945},{"_125":943,"_382":497,"_383":26,"_91":944},{"_125":794,"_382":941,"_383":26,"_91":942},{"_125":939,"_382":497,"_383":26,"_91":940},"LoginHints","Keeps users logged in across NVIDIA domains.","api.ngc.nvidia.com","This cookie is managed by AWS and is used for load balancing.","bm_so","This cookie name is used to load the site as quickly as possible and to recognize bots using Akamai.","The __cf_bm cookie is a cookie necessary to support Cloudflare Bot Management, currently in private beta. As part of our bot management service, this cookie helps manage incoming traffic that matches criteria associated with bots.\nThis is a CloudFoundry cookie","OptanonConsent","nvid.nvidia.com","This cookie is set by the cookie compliance solution from OneTrust. It stores information about the categories of cookies the site uses and whether visitors have given or withdrawn consent for the use of each category. This enables site owners to prevent cookies in each category from being set in the users browser, when consent is not given. The cookie has a normal lifespan of one year, so that returning visitors to the site will have their preferences remembered. It contains no information that can identify the site visitor.","events.nvidia.com","This cookie is associated with the Amazon Web Services Elastic Load Balancing functionality for routing client request on the server.Used to attribute commission to affiliates when you arrive at the website from an affiliate referral link. It is set when you click on one of our links and used to let the advertiser and us know the website from which you came.","sessionid","This is a very generic cookie name that may have different purposes on different sites, but generally it will be some kind of anonymous session identifier.","_csrf","This cookie is used to prevent Cross-Site Request Forgery (CSRF) attacks, ensuring that each request made to the server is from the user's browser and is intentional.","LSKey-c$CookieConsentPolicy","enterprise-support.nvidia.com","Used to apply end-user cookie consent preferences set by our client-side utility.Custom- Short Duration. Unique ID of an Enquiry- used to validate access to case, and load exisiting data.\n","renderCtx","academicgrants.nvidia.com","Used to store site parameters in the session for reuse across requests by a single client for functionality and performance reasons.","CookieConsentPolicy","Used to apply end-user cookie consent preferences set by our client-side utility.","This cookie is managed by AWS and is used for load balancing.Used to attribute commission to affiliates when you arrive at the website from an affiliate referral link. It is set when you click on one of our links and used to let the advertiser and us know the website from which you came.","This cookie is associated with Akamai and is used to differentiate between traffic from humans and bots.","nv_co","__uxq412__ses.88c1","learn.learn.nvidia.com","This cookie is associated with the Amazon Web Services Elastic Load Balancing functionality for routing client request on the server.","pctrk","Used to count page views by unauthenticated users against license usage.","bm_s","bpazaws52gukakzc__ctrl0_ctl36_uccaptcha","bm_sv","store.nvidia.com","The __cf_bm cookie is a cookie necessary to support Cloudflare Bot Management, currently in private beta. As part of our bot management service, this cookie helps manage incoming traffic that matches criteria associated with bots.\n","This cookie name is used by DataDog to determine if cookies are supported on the website.dd_cookie_test","AMP_TLDTEST_xxxxxxxx","SESSlsd","This cookie is likely to be a session identifier used to maintain user session state across page requests.","mboxEdgeCluster","This cookie name is associated with Adobe Target which is used to communicate the correct edge cluster to at.js so that Target profiles can stay in sync as users navigate across a site.","_rdt_em","This cookie name is associated with Reddit which is used to manage session state and user interactions.","megaNavCollapsed","_abck","The _abck cookie is associated with Akamai's Bot Manager, which is used to prevent fraud and abuse by safeguarding online interactions and detecting irregularities that could suggest malicious activity, such as bot-driven attacks or hacking attempts.","idccsrf","This cookie name is associated with Salesforce which is used to track CrossSiteRequestForgery validation for certain single sign-on (SSO) flows.","ApplicationGatewayAffinity","partners.nvidia.com","This cookie is used to maintain a user session by the server in a load balanced environment, to ensure the user has a consistent experience.","AMP_TEST_xxxxxxxx","bm_lso","This cookie name is associated with Akamai which is used to load the site as quickly as possible and to recognize bots.","SSESS###########","This cookie is likely to be a session cookie used to maintain an anonymized user session by the server in a secure environment, often seen in sites built with the Drupal content management system.","nvdeveloper.nvidia.com","__uxq412__id.88c1","OptanonAlertBoxClosed","This cookie is set by websites using certain versions of the cookie law compliance solution from OneTrust. It is set after visitors have seen a cookie information notice and in some cases only when they actively close the notice down. It enables the website not to show the message more than once to a user. The cookie has a normal lifespan of one year and contains no personal information.","QueueITAccepted-SDFrts345E-V3_{eventId}","TEST_AMCV_COOKIE_WRITE","This cookie is used to test whether cookies are enabled in the browser. It is a temporary cookie used to determine if the browser accepts cookies.","XSRF-TOKEN","This cookie is written to help with site security in preventing Cross-Site Request Forgery attacks.This cookie helps with site security in preventing Cross-Site Request Forgery attacks.","AWS ELB application load balancer","This cookie is associated with Akamai. It is used to analyse traffic to determine if it is automated traffic generated by bots or a human user.","bpazaws52gukakzc__ctrl0_ctl78_list_ctl09_systemcaptcha","PHPSESSID","PHP session cookie associated with embedded content from this domain.\n","base.manager.learn.nvidia.com","bm_ss","This cookie is used to help protect websites from automated bot traffic.","AWS ELB application load balancerUsed to attribute commission to affiliates when you arrive at the website from an affiliate referral link. It is set when you click on one of our links and used to let the advertiser and us know the website from which you came.","_devzone_session","This cookie is likely used to maintain a user session on the website by the server. It is common for session identifiers to be used to keep users logged in and track their interaction with the site during a session.","docs.ngc.nvidia.com","bm_sz","This cookie is used by Akamai to optimize site performance and security. It is typically associated with the delivery of web content and services.","vcalliance.nvidia.com","ASP.NET_SessionId","General purpose platform session cookie, used by sites written with Microsoft .NET based technologies. Usually used to maintain an anonymised user session by the server.","com.adobe.alloy.getTld","This cookie name is associated with the Adobe WebSDK which is used to determine the top-most domain that can accept cookies.","SESSls","This cookie is likely to be a session identifier, used to maintain a user's session state across page requests.","c_code","BIGipServerx","This cookie name is associated with the BIG-IP product suite from company F5. Usually associated with managing sessions on load balanced servers, to ensure user requests are routed consistently to the correct server. The common root is BIGipServer most commonly followed by a domain name, usually the one that it is hosted on, but not always.","This cookie is set by websites using certain versions of the cookie law compliance solution from OneTrust. It is set after visitors have seen a cookie information notice and in some cases only when they actively close the notice down. It enables the website not to show the message more than once to a user. The cookie has a one year lifespan and contains no personal information.","default","

NVIDIA uses cookies to improve your experience on our web site. We and our third-party partners also use cookies and other tools to collect and record information you provide as well as information about your interactions with our websites for performance improvement, analytics, and to assist in marketing efforts. By clicking \"Accept All\", you consent to our use of cookies and other tools as described in our Cookie Policy. You can manage your cookie settings by clicking on \"Manage Settings.\" By continuing to use this site or by clicking one of the buttons below, you agree to our Terms of Service (which contains important waivers). Please see our Privacy Policy for more information on our privacy practices.

","gpc","

We have detected the Global Privacy Control (GPC) signal and recorded your rejection of all optional cookies on this site for this browser. You can manage your cookie settings by clicking on \"Manage Settings\". Please see our Cookie Policy for more information. To opt out of non-cookie personal information \"sales\" / \"sharing\" for targeted advertising purposes, please visit the NVIDIA Preference Center. Please see our Privacy Policy for more information on our privacy practices.

","gpcLoggedIn","

We have detected the Global Privacy Control Signal (GPC) and have opted you out of all optional cookies on this browser. You can manage your cookie settings by clicking on \"Manage Settings\". Please see our Cookie Policy for more information. We have also opted you out of \"sharing\"/\"sales\" of personal information outside of cookies. You can manage these settings in the NVIDIA NVIDIA Preference Center. Please see our Privacy Policy for more information.

","gpcLoggedInOverridden","

We have detected the Global Privacy Control Signal (GPC) and have opted you out of all optional cookies on this browser. You can manage your cookie settings by clicking on \"Manage Settings\". Please see our Cookie Policy for more information. We have also opted you out of \"sharing\"/\"sales\" of personal information outside of cookies which overrides at least one of your previous settings. You can manage them in the NVIDIA Preference Center. Please see our Privacy Policy for more information.

",{"_721":16,"_624":51,"_532":51,"_352":51,"_342":51},"hasConsented","gpcApplied","showBanner","geolocation",{"_1045":1046,"_1047":1048},"country","US","state","OR","query","categories",[1068,1069,1070,1071,1072],"optionUrls",[1056,1057,1058,1059,1060,1061,1062,1063,1064,1065,1066,1067],"total",12,"/orgs/nim/deepseek-ai/containers/deepseek-r1/-","/orgs/nim/nvidia/containers/llama-3.1-nemotron-70b-instruct/-","/orgs/nvidia/-/containers/pytorch/-","/orgs/nvidia/omniverse/collections/kit/-","/orgs/nvidia/-/collections/deepstream_sdk/-","/orgs/nvidia/omniverse/collections/kit-appstreaming-collection/-","/orgs/nvidia/research/models/stylegan3/-","/orgs/nvidia/tao/models/peoplenet/-","/orgs/nvidia/tao/models/trafficcamnet/-","/orgs/nvidia/riva/resources/riva_quickstart/-","/orgs/nvidia/-/helm-charts/gpu-operator/-","/orgs/nvidia/-/helm-charts/k8s-nim-operator/-",{"_83":152,"_1073":1133,"_1075":51},{"_83":71,"_1073":1109,"_1075":49},{"_83":154,"_1073":1093,"_1075":1094},{"_83":155,"_1073":1086,"_1075":1087},{"_83":153,"_1073":1074,"_1075":1076},"items",[1077,1078],"startIndex",10,{"_83":153,"_92":1083,"_86":87,"_88":-7,"_84":1084,"_96":97,"_119":1085,"_1082":1066},{"_83":153,"_92":1079,"_86":87,"_88":-7,"_84":1080,"_96":97,"_119":1081,"_1082":1067},"NVIDIA NIM Operator","k8s-nim-operator",[],"url","GPU Operator","gpu-operator",[],[1088],9,{"_83":155,"_92":1089,"_86":87,"_88":1090,"_84":1091,"_96":97,"_119":1092,"_1082":1065},"Riva Skills Quick Start","riva","riva_quickstart",[],[1095,1096,1097],6,{"_83":154,"_92":1105,"_86":87,"_88":1106,"_84":1107,"_96":97,"_119":1108,"_1082":1062},{"_83":154,"_92":1102,"_86":87,"_88":1099,"_84":1103,"_96":97,"_119":1104,"_1082":1063},{"_83":154,"_92":1098,"_86":87,"_88":1099,"_84":1100,"_96":97,"_119":1101,"_1082":1064},"TrafficCamNet","tao","trafficcamnet",[],"PeopleNet","peoplenet",[],"StyleGAN3 pretrained models","research","stylegan3",[],[1110,1111,1112],{"_83":71,"_92":1130,"_86":87,"_88":1114,"_84":1131,"_96":97,"_119":1132,"_1082":1059},{"_83":71,"_92":1127,"_86":87,"_88":-7,"_84":1128,"_96":97,"_119":1129,"_1082":1060},{"_83":71,"_92":1113,"_86":87,"_88":1114,"_84":1115,"_96":97,"_119":1116,"_1082":1061},"Omniverse Kit App Streaming","omniverse","kit-appstreaming-collection",[1117,1118,1119],{"_125":126,"_92":127,"_89":128,"_102":129,"_130":1126},{"_125":1114,"_92":1123,"_89":1124,"_102":1121,"_130":1125},{"_125":1120,"_92":136,"_89":137,"_102":1121,"_130":1122},"omniverse-dev","https://store-images.s-microsoft.com/image/apps.33936.6f1bcf42-55d0-45d1-b4b6-f08c68a50bbe.e91ca707-db64-49d6-9761-4781590a9685.7ba64fc5-7c44-4864-a5e6-7bde05126935",[132],"NVIDIA Omniverse","NVIDIA Omniverse is a platform of APIs, services, and software development kits (SDKs) that enable developers to build generative AI-enabled tools and applications.",[132,133],[132,133,134],"DeepStream SDK","deepstream_sdk",[],"Omniverse Kit (FB)","kit",[1117,1118,1119],[1134,1135,1136],{"_83":152,"_92":1146,"_86":1141,"_88":1147,"_84":1148,"_96":97,"_119":1149,"_1082":1056},{"_83":152,"_92":1140,"_86":1141,"_88":87,"_84":1142,"_96":97,"_119":1143,"_1082":1057},{"_83":152,"_92":1137,"_86":87,"_88":-7,"_84":1138,"_96":97,"_119":1139,"_1082":1058},"PyTorch","pytorch",[],"Llama-3.1-Nemotron-70B-Instruct","nim","llama-3.1-nemotron-70b-instruct",[1144,1117],{"_125":135,"_92":136,"_89":137,"_102":138,"_130":1145},[132],"DeepSeek-R1","deepseek-ai","deepseek-r1",[1144,1117],"ownerOrgs",[],"memberOrgs",[],"externalOrgs",[],"apiKey","2b44f86fea50e90ffd2db29ee2028419","consentGranted","applicationId","4c857e4f-5ffa-45fa-b75e-af773c22f0ad","clientToken","pubb6927db33dae60ba5031a186cde0242a","site","datadoghq.com","service","catalog-ui","env","sessionSampleRate",100,"sessionReplaySampleRate","compressIntakeRequests","storeContextsAcrossPages","defaultPrivacyLevel","mask-user-input","trackViewsManually","trackingConsent","not-granted",{"_1179":1180,"_1181":291,"_1182":26,"_1183":1184,"_1073":1185,"_1186":1187,"_1188":1180,"_1189":26,"_1190":26,"_1191":26,"_1192":1193},"appName","ngc-catalog","appOrigin","hideOnPrivateOrg","icon","ObjectsModules2",[1194,1195,1196,1197,1198,1199,1200],"menuIcon","editor-swatches","productName","requiresAllScopes","requiresLogin","requiresOrgLevelScopesOnly","title","Catalog",{"_1192":1222,"_42":1223,"_1186":1224},{"_1215":1216,"_1192":1217,"_42":1218,"_1188":126},{"_1192":1213,"_42":1214,"_1186":1187},{"_1192":1210,"_42":1211,"_1186":1212},{"_1192":1207,"_42":1208,"_1186":1209},{"_1192":1204,"_42":1205,"_1186":1206},{"_1192":1201,"_42":1202,"_1186":1203},"Resources","/resources","misc-library","Models","/models","misc-neural-network","Helm Charts","/helm-charts","misc-ship-wheel","Containers","/containers","3d-cube","Collections","/collections","access",[1219],"Enterprise Catalog","/enterprise",[1220,1221],"NVAIE","READ","Explore Catalog","/","common-home",{"_183":184,"_185":186,"_187":188,"_189":190,"_191":192,"_193":194},{"_182":26}]

Cosmos Tokenizer: A suite of image and video tokenizers

Model Overview

Description:

Model Versions

License/Terms of Use:

Model Architecture:

Input/Output Specifications

Encoder

Decoder

Software Integration (Required For NVIDIA Models Only):

Usage

Inference with Cosmos-Tokenizer (PyTorch)

Step-1: Installation of Cosmos-Tokenizer

Step-2: Download Pre-trained Checkpoints

Step-3: Run Inference

Evaluation

TokenizationPerformance Comparison

Runtime Comparison

Ethical Considerations

Inference with `Cosmos-Tokenizer` (PyTorch)

Step-1: Installation of `Cosmos-Tokenizer`