# NV-CLIP (Commercial Foundation Model) ## Model Overview NV-CLIP is a multimodal embeddings model for image and text. Trained on 700M proprietary images, NV-CLIP is the NVIDIA commercial version of OpenAI’s CLIP (Contrastive Language-Image Pre-Training) model. NV-CLIP can be applied to various areas such as multimodal search, zero-shot image classification, and downstream computer vision tasks such as object detection and more. ## Getting Started with NV-CLIP NIM microservice Deploying and integrating NV-CLIP NIM microservice is straightforward and based on industry standard APIs. See the NV-CLIP NIM microservice documentation to get started. ## Applications - Multimodal search: Enable accurate image and text search to quickly search database of images and videos. - Zero-shot and few-shot inference: Classify images without re-training or fine-tuning. - Downstream vision tasks: Use the embeddings to enable downstream complex vision AI tasks such as segmentation, detection, VLMs and more. ## References: - Radford, Alec, et al. "Learning transferable visual models from natural language supervision." International conference on machine learning. PMLR, 2021. ## Model Architecture: **Architecture Type:** Transformer-Based
NV-CLIP as a backbone can be used towards various downstream tasks such as classification, detection, segmentation and text based image retrieval. ## Input: **Input Type(s):** Images, Texts
**Input Format(s):** List of Red, Green, Blue (RGB) Images or Strings
**Other Properties Related to Input:**
Channel Ordering of the Input: NCHW, where N = Batch Size, C = number of channels (3), H = Height of images (224), W = Width of the images (224) ## Output: **Output Type(s):** Float tensor
**Output Format:** 3D Tensor
**Other Properties Related to Output:**
The output of this model is an embedding of an input image or text of size 1024 for ViT-H variant. **Supported Operating System(s):**
* Linux
## Model Version(s): - **nv_clip_224_vit_h_trainable_v1.0** - NV-CLIP ViT-H with 224 resolution is foundation model and is trainable. ## Using this Model These models need to be used with NVIDIA hardware and software. These models can only be used with NV-CLIP NIM microservice. The primary use case for these models is getting feature embeddings from images and text. These embeddings can then be used for curation, clustering, zero-shot or few-shot downstream tasks such as classification. These embeddings can also be used towards text and image-based image ## Training Dataset: **Data Collection Method by dataset:**
* Automated
**Labeling Method by dataset:**
* Automated
**Properties:**
| Dataset | No. of Images | |--|--| |NV Internal Data| 700M | ## Evaluation Dataset: **Link:** [https://www.image-net.org/](https://www.image-net.org/) **Data Collection Method by dataset:**
* Unknown
**Labeling Method by dataset:**
* Unknown
**Properties:**
50,000 validation images from [ImageNet dataset](https://www.image-net.org/download.php) #### Methodology and KPI The performance of zero shot accuracy of NV-CLIP on ImageNet validation dataset. | model | top-1 Accuracy | | ----------------------- | -------------- | | ViT-H-224 | 0.7786 | ## Ethical Considerations: ### Bias, Safety & Security, and Privacy 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 supporting 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 Model Card++ Explainability, Bias, Safety & Security, and Privacy Subcards [here](https://catalog.ngc.nvidia.com/orgs/nvidia/teams/tao/models/nvclip_vit/bias). Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/). ### Special Training Data Considerations The model was trained on publicly available data, which may contain toxic language and societal biases. Therefore, the model may amplify those biases, such as associating certain genders with specific social stereotypes. #### Governing Terms The NIM container is governed by the NVIDIA AI Enterprise Software License Agreement | NVIDIA; and the use of this model is governed by the ai-foundation-models-community-license.pdf (nvidia.com). **You are responsible for ensuring that your use of NVIDIA AI Foundation Models complies with all applicable laws.**