This resource is using open-source code maintained in github (see the quick-start-guide section) and available for download from NGC

The NCF model focuses on providing recommendations, also known as collaborative filtering with implicit feedback. The training data for this model should contain binary information about whether a user interacted with a specific item.\nNCF was first described by Xiangnan He, Lizi Liao, Hanwang Zhang, Liqiang Nie, Xia Hu and Tat-Seng Chua in the Neural Collaborative Filtering paper.

The implementation in this repository focuses on the NeuMF instantiation of the NCF architecture.\nWe modified it to use dropout in the FullyConnected layers. This reduces overfitting and increases the final accuracy.\nTraining the other two instantiations of NCF (GMF and MLP) is not supported.

Contrary to the original paper, we benchmark the model on the larger ML-20m dataset\ninstead of using the smaller ML-1m dataset because we think this is more realistic for production type environments.\nHowever, using the ML-1m dataset is also supported.

This model is trained with mixed precision using Tensor Cores on Volta, Turing, and the NVIDIA Ampere GPU architectures.\nTherefore, researchers can get results 2x faster than training without Tensor Cores while experiencing the benefits of mixed precision training.

Model architecture

This model is based mainly on Embedding and FullyConnected layers. The control flow is divided into two branches:

Multi Layer Perceptron (MLP) branch, which transforms the input through FullyConnected layers with ReLU activations, and dropout.
Matrix Factorization (MF) branch, which performs collaborative filtering factorization.\nEach user and each item has two embedding vectors associated with it -- one for the MLP branch and the other for the MF branch.

The outputs from those branches are concatenated and fed to the final FullyConnected layer with sigmoid activation.\nThis can be interpreted as a probability of a user interacting with a given item.

\n \n
\nFigure 1. The architecture of a Neural Collaborative Filtering model. Taken from the Neural Collaborative Filtering paper.\n

Default configuration

The following features were implemented in this model:

Automatic Mixed Precision (AMP)
Data-parallel multi-GPU training and evaluation
Dropout
Gradient accumulation

The following performance optimizations were implemented in this model:

FusedAdam optimizer
Approximate train negative sampling
Caching all the positive training samples in the device memory

Feature support matrix

This model supports the following features:

\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\">

Feature	NCF PyTorch
Automatic Mixed Precision (AMP)	Yes
Multi-GPU training with Distributed Data Parallel (DDP)	Yes
Fused Adam	Yes

Features

Automatic Mixed Precision - This implementation of NCF uses AMP to implement mixed precision training.\nIt allows us to use FP16 training with FP32 master weights by modifying just three lines of code.
Multi-GPU training with Distributed Data Parallel - uses Apex's DDP to implement efficient multi-GPU training with NCCL.
Fused Adam - We use a special implementation of the Adam implementation provided by the Apex package. It fuses some operations for faster weight updates.\nSince NCF is a relatively lightweight model with a large number of parameters, we've observed significant performance improvements from using FusedAdam.

Mixed precision training

Mixed precision is the combined use of different numerical precisions in a computational method.\nMixed precision training offers significant computational speedup by performing\noperations in half-precision format while storing minimal information in single-precision to retain as much information\nas possible in critical parts of the network. Since the introduction of\ntensor cores in Volta,\nand following with both the Turing and Ampere architectures, significant training speedups are experienced by switching\nto mixed precision -- up to 3x overall speedup on the most arithmetically intense model architectures.\nUsing mixed precision training\npreviously required two steps:

Porting the model to use the FP16 data type where appropriate.
Adding loss scaling to preserve small gradient values.

The ability to train deep learning networks with lower precision was introduced in the Pascal architecture and first supported in CUDA 8 in the NVIDIA Deep Learning SDK.

For information about:

How to train using mixed precision, refer to the Mixed Precision Training paper and Training With Mixed Precision documentation.
Techniques used for mixed precision training, refer to the Mixed-Precision Training of Deep Neural Networks blog.
APEX tools for mixed precision training, refer to the NVIDIA Apex: Tools for Easy Mixed-Precision Training in PyTorch.

Enabling mixed precision

Using the Automatic Mixed Precision (AMP) package requires two modifications in the source code.\nThe first one is to initialize the model and the optimizer using the amp.initialize function:

model, optimizer = amp.initialize(model, optimizer, opt_level=\"O2\"\n                                          keep_batchnorm_fp32=False, loss_scale='dynamic')\n

The second one is to use the AMP's loss scaling context manager:

with amp.scale_loss(loss, optimizer) as scaled_loss:\n    scaled_loss.backward()\n

Enabling TF32

TensorFloat-32 (TF32) is the new math mode in NVIDIA A100 GPUs for\nhandling the matrix math, also called tensor operations. TF32 running on Tensor Cores in A100 GPUs can provide up to 10x\nspeedups compared to single-precision floating-point math (FP32) on Volta GPUs.

TF32 Tensor Cores can speed up networks using FP32, typically with no loss of accuracy. It is more robust than FP16 for\nmodels which require a high dynamic range for weights or activations.

For more information, refer to the TensorFloat-32 in the A100 GPU Accelerates AI Training, HPC up to 20x blog post.

TF32 is supported in the NVIDIA Ampere GPU architecture and is enabled by default.

BYO dataset functionality overview

This section describes how you can train the DeepLearningExamples RecSys models on your own datasets without changing the model or data loader and with similar performance to the one published in each repository.\nThis can be achieved thanks to Dataset Feature Specification, which describes how the dataset, data loader and model interact with each other during training, inference and evaluation.\nDataset Feature Specification has a consistent format across all recommendation models in NVIDIA's DeepLearningExamples repository, regardless of dataset file type and the data loader, giving you the flexibility to train RecSys models on your own datasets.

Dataset feature specification

Data flow can be described abstractly:\nInput data consists of a list of rows. Each row has the same number of columns; each column represents a feature.\nThe columns are retrieved from the input files, loaded, aggregated into channels and supplied to the model/training script.

FeatureSpec contains metadata to configure this process and can be divided into three parts:

\n
Specification of how data is organized on disk (source_spec). It describes which feature (from feature_spec) is stored in which file and how files are organized on disk.
\n
\n
Specification of features (feature_spec). Describes a dictionary of features, where key is feature name and values are features' characteristics such as dtype and other metadata (for example, cardinalities for categorical features)
\n
\n
Specification of model's inputs and outputs (channel_spec). Describes a dictionary of model's inputs where keys specify model channel's names and values specify lists of features to be loaded into that channel. Model's channels are groups of data streams to which common model logic is applied, for example categorical/continuous data, user/item ids. Required/available channels depend on the model
\n

The FeatureSpec is a common form of description regardless of underlying dataset format, dataset data loader form and model.

Data flow in NVIDIA Deep Learning Examples recommendation models

The typical data flow is as follows:

S.0. Original dataset is downloaded to a specific folder.
S.1. Original dataset is preprocessed into Intermediary Format. For each model, the preprocessing is done differently, using different tools. The Intermediary Format also varies (for example, for NCF implementation in the PyTorch model, the Intermediary Format is Pytorch tensors in *.pt files.)
S.2. The Preprocessing Step outputs Intermediary Format with dataset split into training and validation/testing parts along with the Dataset Feature Specification yaml file. Metadata in the preprocessing step is automatically calculated.
S.3. Intermediary Format data together with Dataset Feature Specification are fed into training/evaluation scripts. Data loader reads Intermediary Format and feeds the data into the model according to the description in the Dataset Feature Specification.
S.4. The model is trained and evaluated

\n \n
\n

Fig.1. Data flow in Recommender models in NVIDIA Deep Learning Examples repository. Channels of the model are drawn in green.

Example of dataset feature specification

As an example, let's consider a Dataset Feature Specification for a small CSV dataset.

feature_spec:\n  user_gender:\n    dtype: torch.int8\n    cardinality: 3 #M,F,Other\n  user_age: #treated as numeric value\n    dtype: torch.int8\n  user_id:\n    dtype: torch.int32\n    cardinality: 2655\n  item_id:\n    dtype: torch.int32\n    cardinality: 856\n  label:\n    dtype: torch.float32\n\nsource_spec:\n  train:\n    - type: csv\n      features:\n        - user_gender\n        - user_age\n      files:\n        - train_data_0_0.csv\n        - train_data_0_1.csv\n    - type: csv\n      features:\n        - user_id\n        - item_id\n        - label\n      files:\n        - train_data_1.csv\n  test:\n    - type: csv\n      features:\n        - user_id\n        - item_id\n        - label\n        - user_gender\n        - user_age\n        \n      files:\n        - test_data.csv\n\nchannel_spec:\n  numeric_inputs: \n    - user_age\n  categorical_user_inputs: \n    - user_gender\n    - user_id\n  categorical_item_inputs: \n    - item_id\n  label_ch: \n    - label\n

The data contains five features: (user_gender, user_age, user_id, item_id, label). Their data types and necessary metadata are described in the feature specification section.

In the source mapping section, two mappings are provided: one describes the layout of the training data, the other of the testing data. The layout for training data has been chosen arbitrarily to showcase the flexibility.\nThe train mapping consists of two chunks. The first one contains user_gender and user_age, saved as a CSV, and is further broken down into two files. For specifics of the layout, refer to the following example and consult the glossary. The second chunk contains the remaining columns and is saved in a single file. Notice that the order of columns is different in the second chunk - this is alright, as long as the order matches the order in that file (that is, columns in the .csv are also switched)

Let's break down the train source mapping. The table contains example data color-paired to the files containing it.

\n\n

The channel spec describes how the data will be consumed. Four streams will be produced and available to the script/model. The feature specification does not specify what happens further: names of these streams are only lookup constants defined by the model/script.\nBased on this example, we can speculate that the model has three input channels: numeric_inputs, categorical_user_inputs, categorical_item_inputs, and one output channel: label.\nFeature names are internal to the FeatureSpec and can be freely modified.

BYO dataset functionality

In order to train any Recommendation model in NVIDIA Deep Learning Examples one can follow one of three possible ways:

\n
One delivers already preprocessed dataset in the Intermediary Format supported by data loader used by the training script (different models use different data loaders) together with FeatureSpec yaml file describing at least specification of dataset, features and model channels
\n
\n
One uses a transcoding script
\n
\n
One delivers dataset in non-preprocessed form and uses preprocessing scripts that are a part of the model repository. In order to use already existing preprocessing scripts, the format of the dataset needs to match the one of the original datasets. This way, the FeatureSpec file will be generated automatically, but the user will have the same preprocessing as in the original model repository.
\n

Glossary

The Dataset Feature Specification consists of three mandatory and one optional section:

feature_spec provides a base of features that may be referenced in other sections, along with their metadata.\nFormat: dictionary (feature name) => (metadata name => metadata value)

source_spec provides information necessary to extract features from the files that store them.\nFormat: dictionary (mapping name) => (list of chunks)

Mappings are used to represent different versions of the dataset (think: train/validation/test, k-fold splits). A mapping is a list of chunks.
Chunks are subsets of features that are grouped together for saving. For example, some formats may constrain data saved in one file to a single data type. In that case, each data type would correspond to at least one chunk. Another example where this might be used is to reduce file size and enable more parallel loading. Chunk description is a dictionary of three keys:
\n
- type provides information about the format in which the data is stored. Not all formats are supported by all models.
- features is a list of features that are saved in a given chunk. Order of this list may matter: for some formats, it is crucial for assigning read data to the proper feature.
- files is a list of paths to files where the data is saved. For Feature Specification in yaml format, these paths are assumed to be relative to the yaml file's directory (basename). Order of this list matters: It is assumed that rows 1 to i appear in the first file, rows i+1 to j in the next one, etc.
\n

channel_spec determines how features are used. It is a mapping (channel name) => (list of feature names).

Channels are model specific magic constants. In general, data within a channel is processed using the same logic. Example channels: model output (labels), categorical ids, numerical inputs, user data, and item data.

metadata is a catch-all, wildcard section: If there is some information about the saved dataset that does not fit into the other sections, you can store it here.

","resourceVersions",[230,231],"resourceVersionsTotal","hasMoreVersions","versionMetadata",{"_209":28,"_210":-5,"_52":116,"_113":-5,"_117":211,"_119":-5,"_212":115,"_124":16,"_121":34,"_213":-5,"_120":34,"_214":215,"_216":90,"_217":218,"_219":90,"_220":221,"_222":90,"_223":224,"_225":90,"_226":227,"_228":90,"_122":229},"versionNotFound","isLoggedIn","latestVersionPath","/orgs/nvidia/dle/resources/ncf_pyt/-","imageReference","cliDownloadCommand","ngc registry resource download-version nvidia/dle/ncf_pyt:21.04.0","helmFetchCommand","viewAllVersionsPath","/orgs/nvidia/dle/resources/ncf_pyt/21.04.0/version-history","accessCheckResult",{"_208":90},"hasRelatedCollections","context",{"_204":205,"_206":207},1786029669981,[],[],[197,198,199,200,201,202,203],{"_48":62,"_50":73,"_52":64,"_54":65,"_56":74,"_58":59},{"_48":62,"_50":71,"_52":146,"_54":-5,"_56":-5,"_58":59},{"_48":62,"_50":69,"_52":64,"_54":65,"_56":70,"_58":59},{"_48":62,"_50":67,"_52":64,"_54":65,"_56":68,"_58":59},{"_48":62,"_50":63,"_52":64,"_54":65,"_56":66,"_58":59},{"_48":60,"_50":61,"_52":53,"_54":55,"_56":57,"_58":59},{"_48":49,"_50":51,"_52":53,"_54":55,"_56":57,"_58":59},"orgName","nvidia","teamName","dle","allowed","versionId","description",["D",1663857435439],"totalSizeInBytes","malwareScanStatus","releaseNotes","The performance measurements in this document were conducted at the time of publication and may not reflect\nthe performance achieved from NVIDIA's latest software release. For the most up-to-date performance measurements, \ngo to [NVIDIA Data Center Deep Learning Product Performance](https://developer.nvidia.com/deep-learning-performance-training-inference).\n\n### Changelog\n1. January 22, 2018\n * Initial release\n2. May, 2019\n * Lower memory consumption (down from about 18GB to 10GB for batch size 1M on a single NVIDIA Tesla V100). Achieved by using an approximate method for generating negatives for training.\n * Automatic Mixed Precision (AMP) with dynamic loss scaling instead of a custom mixed-precision optimizer.\n * Performance numbers for NVIDIA DGX-2.\n * Data loading code cleanup.\n * Default container updated to PyTorch 19.05-py3.\n * Updated README.md.\n3. June, 2019\n * Updated performance tables.\n * Default container changed to PyTorch 19.06-py3.\n * Caching validation negatives between runs\n4. September, 2019\n * Adjusting for API changes in PyTorch and APEX\n * Checkpoints loading fix\n5. January, 2020\n * DLLogger support added\n6. June, 2020\n * Updated performance tables to include A100 results5. \n7. June, 2021\n * Enhanced BYO dataset functionality - added Feature Specification and transcoding\n * Default container changed to PyTorch 21.04-py3\n * Updated performance and accuracy tables and plots\n * Code cleanup\n\n### Known issues\n \n#### Scaling beyond 8 GPUs\nNeural Collaborative Filtering is a relatively lightweight model that trains quickly with this relatively smaller dataset, ML-20m.\nBecause of that, the high ratio of communication to computation makes it difficult to \nefficiently use more than 8 GPUs. Typically, this is not an issue because when using 8\nGPUs with FP16 precision, the training is sufficiently fast. However, if you'd like to\n scale the training to 16 GPUs and beyond, you might try modifying the model so that \n the communication-computation ratio facilitates better scaling. This could be done, for example,\n by finding hyperparameters that enable using a larger batch size or by reducing the \n number of trainable parameters.\n\n#### Memory usage\n\nIn the default settings, the additional memory beyond 16GB may not be fully utilized.\nThis is because we set the default batch size for the ML-20m dataset to 1M,\nwhich is too small to fill up multiple 32GB GPUs completely.\n1M is the batch size for which we experienced the best convergence on the ML-20m dataset.\nHowever, on other datasets, even faster performance can be possible by finding hyperparameters that work well for larger batches and leverage additional GPU memory.","hasReleaseNotes","setup","The following section lists the requirements in order to start training the Neural Collaborative Filtering model.\n\n### Requirements\nThis repository contains a Dockerfile that extends the PyTorch NGC container and encapsulates some dependencies. \nAside from these dependencies, ensure you have the following components:\n\n- [NVIDIA Docker](https://github.com/NVIDIA/nvidia-docker)\n- PyTorch 21.04-py3+ NGC container\n- Supported GPUs:\n - [NVIDIA Volta architecture](https://www.nvidia.com/en-us/data-center/volta-gpu-architecture/)\n - [NVIDIA Turing architecture](https://www.nvidia.com/en-us/design-visualization/technologies/turing-architecture/)\n - [NVIDIA Ampere architecture](https://www.nvidia.com/en-us/data-center/nvidia-ampere-gpu-architecture/)\n\nFor more information about how to get started with NGC containers, refer to the following sections from the NVIDIA GPU Cloud Documentation and the Deep Learning Documentation:\n- [Getting Started Using NVIDIA GPU Cloud](https://docs.nvidia.com/ngc/ngc-getting-started-guide/index.html)\n- [Accessing And Pulling From The NGC Container Registry](https://docs.nvidia.com/deeplearning/frameworks/user-guide/index.html#accessing_registry)\n\nRunning PyTorch\n\nFor those unable to use the PyTorch NGC container, to set up the required environment or create your own container, refer to the versioned [NVIDIA Container Support Matrix](https://docs.nvidia.com/deeplearning/frameworks/support-matrix/index.html).","hasSetup","quickStartGuide","To train your model using mixed or TF32 precision with Tensor Cores or using FP32, perform the following steps using the default parameters of the NCF model on the ML-20m dataset. For the specifics concerning training and inference, refer to the [Advanced](#advanced) section.\n\n1. Clone the repository.\n```bash\ngit clone https://github.com/NVIDIA/DeepLearningExamples\ncd DeepLearningExamples/PyTorch/Recommendation/NCF\n```\n\n2. Build an NCF PyTorch Docker container.\n\nAfter Docker is set up, you can build the NCF image with:\n```bash\ndocker build . -t nvidia_ncf\n``` \n\n3. Start an interactive session in the NGC container to run preprocessing/training and inference.\n\nThe NCF PyTorch container can be launched with:\n```bash\ndocker run --runtime=nvidia -it --rm --ipc=host -v ${PWD}/data:/data nvidia_ncf bash\n```\n\nThis will launch the container and mount the `./data` directory as a volume to the `/data` directory inside the container.\nAny datasets and experiment results (logs, checkpoints etc.) saved to `/data` will be accessible\nin the `./data` directory on the host. \n\n4. Download and preprocess the data.\n\nDownload the data from https://grouplens.org/datasets/movielens/20m/ and put it in `/data/ml-20m/ml-20m.zip`.\n\nPreprocessing consists of sorting the data, dropping the duplicates, renumbering users and items,\nselecting last interaction of each user to include in the test set,\nthen randomly generating negative test set members (scoring candidates).\nThe preprocessed train and test data is then saved in PyTorch binary format to be loaded just before training.\n\nNote: Preprocessing requires PyTorch and should therefore be run inside the Docker container.\n\nTo preprocess the ML-20m dataset, you can run:\n\n```bash\n./prepare_dataset.sh\n```\n\nNote: This command will return immediately without downloading anything if the data is already present in the `/data` directory.\n\nThis will store the preprocessed training and evaluation data in the `/data` directory so that it can be later\nused to train the model (by passing the appropriate `--data` argument to the `ncf.py` script).\n\n5. Start training.\n\nAfter the Docker container is launched, the training with the default hyperparameters (suitable for a DGX-1V or DGX A100 with 8 GPUs) can be started with:\n\n```bash\npython -m torch.distributed.launch --nproc_per_node=8 --use_env ncf.py --data /data/cache/ml-20m --checkpoint_dir /data/checkpoints/\n```\n\nThis will result in a checkpoint file being written to `/data/checkpoints/model.pth`.\n\n6. Start validation/evaluation.\n\nThe trained model can be evaluated by passing the `--mode` test flag to the `run.sh` script:\n\n```bash\npython -m torch.distributed.launch --nproc_per_node=1 --use_env ncf.py --data /data/cache/ml-20m --mode test --load_checkpoint_path /data/checkpoints/model.pth\n```","hasQuickStartGuide","advanced","The following sections provide greater details of the dataset, running training and inference, and the training results.\n\n### Scripts and sample code\n\nThe `ncf.py` script contains most of the training and validation logic. Data loading and preprocessing code is located in `dataloading.py`.\nThe model architecture is defined in `neumf.py`. Some initial data preprocessing is located in `convert.py`.\nThe logger directory contains simple bookkeeping utilities for storing training results.\nThe `transcode.py` script enables transcoding data from a CSV containing preprocessed data to a format accessible by the model.\n\n### Command-line options\n\nTo view the full list of available options and their descriptions, use the `-h` or `--help` command-line option, for example: \n`python ncf.py --help`\n\nThe following example output is printed when running the sample:\n```\nusage: ncf.py [-h] [--data DATA] [--feature_spec_file FEATURE_SPEC_FILE] [-e EPOCHS] [-b BATCH_SIZE] [--valid_batch_size VALID_BATCH_SIZE] [-f FACTORS]\n [--layers LAYERS [LAYERS ...]] [-n NEGATIVE_SAMPLES] [-l LEARNING_RATE] [-k TOPK] [--seed SEED] [--threshold THRESHOLD] [--beta1 BETA1] [--beta2 BETA2]\n [--eps EPS] [--dropout DROPOUT] [--checkpoint_dir CHECKPOINT_DIR] [--load_checkpoint_path LOAD_CHECKPOINT_PATH] [--mode {train,test}]\n [--grads_accumulated GRADS_ACCUMULATED] [--amp] [--log_path LOG_PATH]\n\nTrain a Neural Collaborative Filtering model\n\noptional arguments:\n -h, --help show this help message and exit\n --data DATA Path to the directory containing the feature specification yaml\n --feature_spec_file FEATURE_SPEC_FILE\n Name of the feature specification file or path relative to the data directory.\n -e EPOCHS, --epochs EPOCHS\n Number of epochs for training\n -b BATCH_SIZE, --batch_size BATCH_SIZE\n Number of examples for each iteration. This will be divided by the number of devices\n --valid_batch_size VALID_BATCH_SIZE\n Number of examples in each validation chunk. This will be the maximum size of a batch on each device.\n -f FACTORS, --factors FACTORS\n Number of predictive factors\n --layers LAYERS [LAYERS ...]\n Sizes of hidden layers for MLP\n -n NEGATIVE_SAMPLES, --negative_samples NEGATIVE_SAMPLES\n Number of negative examples per interaction\n -l LEARNING_RATE, --learning_rate LEARNING_RATE\n Learning rate for optimizer\n -k TOPK, --topk TOPK Rank for test examples to be considered a hit\n --seed SEED, -s SEED Manually set random seed for torch\n --threshold THRESHOLD, -t THRESHOLD\n Stop training early at threshold\n --beta1 BETA1, -b1 BETA1\n Beta1 for Adam\n --beta2 BETA2, -b2 BETA2\n Beta1 for Adam\n --eps EPS Epsilon for Adam\n --dropout DROPOUT Dropout probability, if equal to 0 will not use dropout at all\n --checkpoint_dir CHECKPOINT_DIR\n Path to the directory storing the checkpoint file, passing an empty path disables checkpoint saving\n --load_checkpoint_path LOAD_CHECKPOINT_PATH\n Path to the checkpoint file to be loaded before training/evaluation\n --mode {train,test} Passing \"test\" will only run a single evaluation; otherwise, full training will be performed\n --grads_accumulated GRADS_ACCUMULATED\n Number of gradients to accumulate before performing an optimization step\n --amp Enable mixed precision training\n --log_path LOG_PATH Path for the JSON training log\n\n```\n\n### Getting the data\n\nThe NCF model was trained on the ML-20m dataset.\nFor each user, the interaction with the latest timestamp was included in the test set, and the rest of the examples are used as the training data. \n\nThis repository contains the `./prepare_dataset.sh` script that automatically preprocess the training and validation datasets. \nBy default, the preprocessed data will be placed in `/data/cache`.\n\n#### Dataset guidelines\n\nNCF supports all datasets that include a Feature Specification file and are properly formatted.\nFor details, refer to the [BYO dataset](#byo-dataset) section.\n\n#### ML-1m\n\nTo preprocess and train on the ML-1m dataset run:\n```bash\n./prepare_dataset.sh ml-1m\npython -m torch.distributed.launch --nproc_per_node=8 --use_env ncf.py --data /data/cache/ml-1m\n```\n\n### BYO dataset \n\nThis implementation supports using other datasets thanks to BYO dataset functionality. \nThe BYO dataset functionality allows users to plug in their dataset in a common fashion for all Recommender models \nthat support this functionality. Using BYO dataset functionality, the user does not have to modify the source code of \nthe model thanks to the Feature Specification file. For general information on how BYO dataset works, refer to the \n[BYO dataset overview section](#byo-dataset-functionality-overview).\n\nThere are three ways to plug in user's dataset:\n

1. Provide an unprocessed dataset in a format matching the one used by ml-20m, then use ml-20m's preprocessing. Feature Specification file is then generated automatically.

\nThe required format of the user's dataset is:\n\n* CSV file with three columns: `user_id`, `item_id` and `timestamp`\n* This CSV should contain only the positive examples. The negatives will be sampled during the training and validation.\n\nThe correct torch.tensor dataset files together with the Feature Specification yaml file will be generated automatically by preprocessing script.\n\nThe following example shows how to use this way of plugging a user's dataset:\n\nBuild the NCF image with:\n```bash\ndocker build . -t nvidia_ncf\n```\nLaunch the container with:\n```bash\ndocker run --runtime=nvidia -it --rm --ipc=host -v ${PWD}/data:/data nvidia_ncf bash\n```\n\nInside the container run:\n```bash\n./prepare_dataset.sh like_movielens\n```\nThis will preprocess the `data/like_movielens/ratings.csv` file and save the output in `data/cache/like_movielens`\n\nTo run the training on 1 GPU:\n```bash\npython -m torch.distributed.launch --nproc_per_node=1 --use_env ncf.py --data /data/cache/like_movielens\n```\n\nTo run the training on 8 GPUs\n```bash\npython -m torch.distributed.launch --nproc_per_node=8 --use_env ncf.py --data /data/cache/like_movielens\n```\n\nOne can also add direct support for your dataset in the `prepare_dataset.sh` and `load.py` scripts.\n

\n\n

2. Provide a CSV containing preprocessed data and a simplified Feature Specification yaml file, then transcode the data with `transcode.py` script

\nThis option should be used if the user has their own CSV file with a preprocessed dataset they want to train on.\n\nThe required format of the user's dataset is:\n* CSV files containing the data, already split into train and test sets. \n* Feature Specification yaml file describing the layout of the CSV data\n\nFor an example of a feature specification file and properly formatted CSVs, refer to the `data/csv_conversion` folder.\n\nThe CSV containing the data:\n* should be already split into train and test\n* should contain no header\n* should contain one column per feature, in the order specified by the feature specification file\n* `user_id` and `item_id` should be all non-negative integers of range (0,num_users), (0,num_items) respectively\n* negative examples for the testing set should already be present\n* negative examples for the training set may be already present. By default, the training script samples additional random negative examples (controlled by the '--negative_samples' flag supplied to the `ncf.py` script).\n \nThe Feature Specification yaml file:\n* needs to describe the layout of data in CSV files\n* should contain information about user, item cardinalities. However, if set to `auto`, they will be inferred from the data by the transcoding script.\n\nRefer to `data/csv_conversion/feature_spec.yaml` for an example of the yaml Feature Specification.\n\nThe following example shows how to use this way of plugging user's dataset:\n\nBuild the NCF image with:\n```bash\ndocker build . -t nvidia_ncf\n```\nLaunch the container with:\n```bash\ndocker run --runtime=nvidia -it --rm --ipc=host -v ${PWD}/data:/data nvidia_ncf bash\n```\n\nInside the container run:\n```bash\nmkdir /data/conversion_output\npython transcode.py --path /data/csv_conversion --output /data/conversion_output\n```\n\nThis will convert the data from `data/csv_conversion` and save the output in `data/conversion_output`.\nRefer to `data/csv_conversion/feature_spec.yaml` for an example of the yaml Feature Specification.\n\nTo run the training on 1 GPU:\n```bash\npython -m torch.distributed.launch --nproc_per_node=1 --use_env ncf.py --data /data/conversion_output -k 3\n```\n\nTo run the training on 8 GPUs\n```bash\npython -m torch.distributed.launch --nproc_per_node=8 --use_env ncf.py --data /data/conversion_output -k 3\n```\n\nThe parameter `k` changes the computed metric from HR@10 to HR@3. This is done because the examples are extremely small,\nand hit rate depth may not be longer than lists of candidates.\n

3. Provide a fully preprocessed dataset, saved in torch.tensor files, and a Feature Specification yaml file

\nThis is the option to choose if you want full control over preprocessing and/or want to preprocess data directly to the target format.\n\nYour final output will need to contain a Feature Specification yaml describing data and file layout. For an example feature specification file, refer to `data/ml-20m/feature_spec_template.yaml`\n\nFor details, refer to the [BYO dataset overview section](#byo-dataset-functionality-overview).\n

\n\n#### Channel definitions and requirements for NCF-PyT feature specifications.\n\nThis model defines three channels, each accepting a single feature:\n\n- user_ch\n- item_ch\n- label_ch\n\nThe training script expects two mappings:\n\n- train\n- test\n\nAs this NeuMF implementation computes list ranking metrics, the testing set actually consists of lists of candidates.\nUsually, all entries in a list share the same user id, although this is not mandatory.\nAll entries from a given list must appear consecutively in the testing set.\nList boundaries are not marked in the testing set.\nAll lists must have the same length. This length must be set by the metadata:test_samples_per_series parameter in the Feature Specification yaml file.\n\n#### BYO dataset constraints for the model\n\nThere are the following constraints of BYO dataset functionality for this model:\n1. The performance of the model depends on the dataset size. Generally, the model should scale better for datasets containing more data points. For a smaller dataset, you might experience slower performance than the one reported for ml-20m\n2. As this implementation keeps the training and testing data in GPU VRAM, supported dataset size is limited by the GPU VRAM size.\n3. Using other datasets might require tuning some hyperparameters (for example, learning rate, beta1 and beta2) to reach desired accuracy.\n4. The transcoding script uses pandas, and the user's dataset needs to fit into the system memory\n\n### Training process\nThe name of the training script is `ncf.py`. Because of the multi-GPU support, it should always be run with the torch distributed launcher like this:\n```bash\npython -m torch.distributed.launch --nproc_per_node= --use_env ncf.py --data [other_parameters]\n```\n\nThe main result of the training are checkpoints stored by default in `/data/checkpoints/`. This location can be controlled\nby the `--checkpoint_dir` command-line argument.\n\nThe validation metric is Hit Rate at 10 (HR@10) with 100 test negative samples. This means that for each positive sample in \nthe test set, 100 negatives are sampled. All resulting 101 samples are then scored by the model. If the true positive sample is\namong the 10 samples with the highest scores we have a \"hit,\" and the metric is equal to 1; otherwise, it's equal to 0.\nThe HR@10 metric is the number of hits in the entire test set divided by the number of samples in the test set. \n\n### Inference process\n\nInference can be launched with the same script used for training by passing the `--mode test` flag:\n```bash\npython -m torch.distributed.launch --nproc_per_node= --use_env ncf.py --data --mode test [other_parameters]\n```\n\nThe script will then:\n* Load the checkpoint from the directory specified by the `--checkpoint_dir` directory\n* Run inference on the test dataset\n* Compute and print the validation metric","hasAdvanced","performance","The performance measurements in this document were conducted at the time of publication and may not reflect the performance achieved from NVIDIA's latest software release. For the most up-to-date performance measurements, go to [NVIDIA Data Center Deep Learning Product Performance](https://developer.nvidia.com/deep-learning-performance-training-inference).\n\n### Benchmarking\n\n#### Training performance benchmark\n\nNCF training on NVIDIA DGX systems is very fast; therefore, in order to measure train and validation throughput, you can simply run the full training job with: \n```bash\n./prepare_dataset.sh\npython -m torch.distributed.launch --nproc_per_node=8 --use_env ncf.py --data /data/cache/ml-20m --epochs 5\n```\n\nAt the end of the script, a line reporting the best train throughput is printed.\n\n#### Inference performance benchmark\n\nValidation throughput can be measured by running the full training job with:\n```bash\n./prepare_dataset.sh\npython -m torch.distributed.launch --nproc_per_node=8 --use_env ncf.py --data /data/cache/ml-20m --epochs 5\n```\n\nThe best validation throughput is reported to the standard output. \n\n### Results\n\nThe following sections provide details on how we achieved our performance and accuracy in training and inference. \n\n#### Training accuracy results\n\n##### Training accuracy: NVIDIA DGX A100(8x A100 40GB)\nOur results were obtained by following the steps in the Quick Start Guide in the PyTorch 21.04-py3 NGC container on NVIDIA DGX A100 (8x A100 40GB) GPUs.\n\nThe following table lists the best hit rate at 10 for DGX A100 with 8 A100 40GB GPUs. It also shows the time to reach this HR@10.\nResults are averages across 20 random seeds.\n\n| GPUs | Batch size / GPU | Accuracy - TF32 | Accuracy - mixed precision | Time to train - TF32 | Time to train - mixed precision | Time to train speedup (TF32 to mixed precision) |\n|-------:|-------------------:|------------------:|-----------------------------:|-----------------------:|----------------------------------:|--------------------------------------------------:|\n| 1 | 1048576 | 0.958925 | 0.958892 | 140.771 | 94.2386 | 1.49 |\n| 8 | 131072 | 0.958938 | 0.959089 | 30.0928 | 23.7362 | 1.27 |\n\n##### Training accuracy: NVIDIA DGX-1 (8x V100 16GB)\n\nOur results were obtained by following the steps in the Quick Start Guide in the PyTorch 21.04-py3 NGC container on NVIDIA DGX-1 with 8x V100 16GB GPUs.\n\nThe following table lists the best hit rate at 10 for DGX-1 with 8 V100 16GB GPUs. It also shows the time to reach this HR@10.\nResults are averages across 20 random seeds.\nThe training time was measured excluding data downloading, preprocessing, validation data generation and library initialization times.\n\n| GPUs | Batch size / GPU | Accuracy - FP32 | Accuracy - mixed precision | Time to train - FP32 | Time to train - mixed precision | Time to train speedup (FP32 to mixed precision) |\n|-------:|-------------------:|------------------:|-----------------------------:|-----------------------:|----------------------------------:|--------------------------------------------------:|\n| 1 | 1048576 | 0.958857 | 0.958815 | 302.443 | 145.423 | 2.08 |\n| 8 | 131072 | 0.958768 | 0.959052 | 53.7044 | 34.2503 | 1.57 |\n\nTo reproduce this result, start the NCF Docker container interactively and run:\n```bash\n./prepare_dataset.sh\npython -m torch.distributed.launch --nproc_per_node=8 --use_env ncf.py --data /data/cache/ml-20m\n```\n\n##### Training accuracy: NVIDIA DGX-1 (8x V100 32GB)\n\nOur results were obtained by following the steps in the Quick Start Guide in the PyTorch 21.04-py3 NGC container on NVIDIA DGX-1 with 8x V100 32GB GPUs.\n\nThe following table lists the best hit rate at 10 for DGX-1 with 8 V100 32GB GPUs. It also shows the time to reach this HR@10.\nResults are averages across 20 random seeds.\nThe training time was measured excluding data downloading, preprocessing, validation data generation and library initialization times.\n\n| GPUs | Batch size / GPU | Accuracy - FP32 | Accuracy - mixed precision | Time to train - FP32 | Time to train - mixed precision | Time to train speedup (FP32 to mixed precision) |\n|-------:|-------------------:|------------------:|-----------------------------:|-----------------------:|----------------------------------:|--------------------------------------------------:|\n| 1 | 1048576 | 0.958992 | 0.959002 | 310.467 | 153.616 | 2.02 |\n| 8 | 131072 | 0.95871 | 0.958925 | 55.716 | 36.3384 | 1.53 |\n\nTo reproduce this result, start the NCF Docker container interactively and run:\n```bash\n./prepare_dataset.sh\npython -m torch.distributed.launch --nproc_per_node=8 --use_env ncf.py --data /data/cache/ml-20m\n```\n\n##### Training accuracy: NVIDIA DGX-2 (16x V100 32GB)\n\nOur results were obtained by following the steps in the Quick Start Guide in the PyTorch 21.04-py3 NGC container on NVIDIA DGX-2 with 16x V100 32GB GPUs.\n\nThe following table lists the best hit rate at 10 for DGX-2 with 16 V100 32GB GPUs. It also shows the time to reach this HR@10.\nResults are averages across 20 random seeds.\nThe training time was measured excluding data downloading, preprocessing, validation data generation and library initialization times.\n\n| GPUs | Batch size / GPU | Accuracy - FP32 | Accuracy - mixed precision | Time to train - FP32 | Time to train - mixed precision | Time to train speedup (FP32 to mixed precision) |\n|-------:|-------------------:|------------------:|-----------------------------:|-----------------------:|----------------------------------:|--------------------------------------------------:|\n| 1 | 1048576 | 0.958756 | 0.958833 | 289.004 | 143.61 | 2.01 |\n| 8 | 131072 | 0.958864 | 0.958806 | 52.1788 | 33.7456 | 1.55 |\n| 16 | 65536 | 0.958905 | 0.958893 | 37.7075 | 27.174 | 1.39 |\n\nTo reproduce this result, start the NCF Docker container interactively and run:\n```bash\n./prepare_dataset.sh\npython -m torch.distributed.launch --nproc_per_node=16 --use_env ncf.py --data /data/cache/ml-20m\n```\n\n##### Influence of AMP on accuracy\n\nThe box plots below show the best accuracy achieved in each run.\nTwenty experiments were performed for each configuration.\n\n![hr_boxplot](https://raw.githubusercontent.com/NVIDIA/DeepLearningExamples/master/PyTorch/Recommendation/NCF//img/box_plots.png)\n\n\n##### Training validation curves\n\nThe plots below show the validation accuracy over the course of training.\nOne sample curve is shown for each configuration.\n\n![validation_accuracy](https://raw.githubusercontent.com/NVIDIA/DeepLearningExamples/master/PyTorch/Recommendation/NCF//img/val_curves.png)\n\n\n#### Training performance results\n\nResults are averages over 20 runs for each configuration.\n\n##### Training performance: NVIDIA DGX A100 (8x A100 40GB)\nOur results were obtained by following the steps in the Quick Start Guide in the PyTorch 21.04-py3 NGC container on NVIDIA DGX A100 (8x A100 40GB) GPUs. Performance numbers (in items per second) were averaged over an entire training epoch.\n\n| GPUs | Batch size / GPU | Throughput - TF32 (samples/s) | Throughput - mixed precision (samples/s) | Throughput speedup (TF32 to mixed precision) | Strong scaling - TF32 | Strong scaling - mixed precision |\n|-------:|-------------------:|:--------------------------------|:-------------------------------------------|-----------------------------------------------:|------------------------:|-----------------------------------:|\n| 1 | 1048576 | 22.59M | 34.08M | 0.66 | 1 | 1 |\n| 8 | 131072 | 110.16M | 142.90M | 0.77 | 4.88 | 4.19 |\n\n##### Training performance: NVIDIA DGX-1 (8x V100 16GB)\n\nOur results were obtained by following the steps in the Quick Start Guide in the PyTorch 21.04-py3 NGC container on NVIDIA DGX-1 with 8x V100 16GB GPUs. \n\nThe following table shows the best training throughput:\n\n| GPUs | Batch size / GPU | Throughput - FP32 (samples/s) | Throughput - mixed precision (samples/s) | Throughput speedup (FP32 to mixed precision) | Strong scaling - FP32 | Strong scaling - mixed precision |\n|-------:|-------------------:|:--------------------------------|:-------------------------------------------|-----------------------------------------------:|------------------------:|-----------------------------------:|\n| 1 | 1048576 | 10.42M | 21.84M | 0.48 | 1 | 1 |\n| 8 | 131072 | 60.03M | 95.95M | 0.63 | 5.76 | 4.39 |\n\n##### Training performance: NVIDIA DGX-1 (8x V100 32GB)\n\nOur results were obtained by following the steps in the Quick Start Guide in the PyTorch 21.04-py3 NGC container on NVIDIA DGX-1 with 8x V100 32GB GPUs. \n\nThe following table shows the best training throughput:\n\n| GPUs | Batch size / GPU | Throughput - FP32 (samples/s) | Throughput - mixed precision (samples/s) | Throughput speedup (FP32 to mixed precision) | Strong scaling - FP32 | Strong scaling - mixed precision |\n|-------:|-------------------:|:--------------------------------|:-------------------------------------------|-----------------------------------------------:|------------------------:|-----------------------------------:|\n| 1 | 1048576 | 10.14M | 20.65M | 0.49 | 1 | 1 |\n| 8 | 131072 | 58.50M | 91.77M | 0.64 | 5.77 | 4.44 |\n\n##### Training performance: NVIDIA DGX-2 (16x V100 32GB)\n\nOur results were obtained by following the steps in the Quick Start Guide in the PyTorch 21.04-py3 NGC container on NVIDIA DGX-2 with 16x V100 32GB GPUs. \n\nThe following table shows the best training throughput:\n\n| GPUs | Batch size / GPU | Throughput - FP32 (samples/s) | Throughput - mixed precision (samples/s) | Throughput speedup (FP32 to mixed precision) | Strong scaling - FP32 | Strong scaling - mixed precision |\n|-------:|-------------------:|:--------------------------------|:-------------------------------------------|-----------------------------------------------:|------------------------:|-----------------------------------:|\n| 1 | 1048576 | 10.90M | 22.16M | 0.49 | 1 | 1 |\n| 8 | 131072 | 62.16M | 98.56M | 0.63 | 5.7 | 4.45 |\n| 16 | 65536 | 92.20M | 134.91M | 0.68 | 8.46 | 6.09 |\n\n#### Inference performance results\n\n##### Inference performance: NVIDIA DGX A100 (1x A100 40GB)\n\nOur results were obtained by running the `inference.py` script in the PyTorch 21.04 NGC container on NVIDIA DGX A100 with 1x A100 GPU.\n\nTF32\n\n| Batch size | Throughput Avg | Latency Avg | Latency 90% | Latency 95% | Latency 99% |\n|-------------:|-----------------:|--------------:|--------------:|--------------:|--------------:|\n| 1024 | 2.96198e+06 | 0.000346 | 0.00037 | 0.000374 | 0.000383 |\n| 4096 | 1.16823e+07 | 0.000351 | 0.000375 | 0.000382 | 0.000389 |\n| 16384 | 4.01876e+07 | 0.000408 | 0.000442 | 0.000443 | 0.000445 |\n| 65536 | 5.06161e+07 | 0.001295 | 0.001319 | 0.001321 | 0.001324 |\n| 262144 | 5.62193e+07 | 0.004663 | 0.004655 | 0.00466 | 0.005091 |\n| 1048576 | 5.74678e+07 | 0.018246 | 0.018258 | 0.018261 | 0.018276 |\n\nFP16\n\n| Batch size | Throughput Avg | Latency Avg | Latency 90% | Latency 95% | Latency 99% |\n|-------------:|-----------------:|--------------:|--------------:|--------------:|--------------:|\n| 1024 | 2.9068e+06 | 0.000352 | 0.000379 | 0.000383 | 0.000401 |\n| 4096 | 1.1149e+07 | 0.000367 | 0.000394 | 0.000396 | 0.000402 |\n| 16384 | 4.46873e+07 | 0.000367 | 0.000391 | 0.000397 | 0.000406 |\n| 65536 | 7.15357e+07 | 0.000916 | 0.001064 | 0.001068 | 0.001071 |\n| 262144 | 8.02216e+07 | 0.003268 | 0.00327 | 0.003272 | 0.00338 |\n| 1048576 | 8.27085e+07 | 0.012678 | 0.012685 | 0.012688 | 0.012809 |\n\n##### Inference performance: NVIDIA DGX-1 (1x V100 16GB)\n\nOur results were obtained by running the `inference.py` script in the PyTorch 21.04 NGC container on NVIDIA DGX-1 with 1x V100 16GB GPU.\n\nFP32\n\n| Batch size | Throughput Avg | Latency Avg | Latency 90% | Latency 95% | Latency 99% |\n|-------------:|-----------------:|--------------:|--------------:|--------------:|--------------:|\n| 1024 | 1.91315e+06 | 0.000535 | 0.000557 | 0.000565 | 0.000589 |\n| 4096 | 7.4782e+06 | 0.000548 | 0.000566 | 0.000577 | 0.000718 |\n| 16384 | 2.15241e+07 | 0.000761 | 0.000783 | 0.000791 | 0.000842 |\n| 65536 | 2.77005e+07 | 0.002366 | 0.00242 | 0.002431 | 0.002435 |\n| 262144 | 2.95251e+07 | 0.008879 | 0.008888 | 0.008895 | 0.008932 |\n| 1048576 | 2.92491e+07 | 0.03585 | 0.03603 | 0.036078 | 0.036144 |\n\nFP16\n\n| Batch size | Throughput Avg | Latency Avg | Latency 90% | Latency 95% | Latency 99% |\n|-------------:|-----------------:|--------------:|--------------:|--------------:|--------------:|\n| 1024 | 2.00172e+06 | 0.000512 | 0.000538 | 0.000546 | 0.000577 |\n| 4096 | 8.08797e+06 | 0.000506 | 0.000519 | 0.000535 | 0.000569 |\n| 16384 | 3.22482e+07 | 0.000508 | 0.000516 | 0.000519 | 0.000557 |\n| 65536 | 5.20587e+07 | 0.001259 | 0.001265 | 0.001267 | 0.001278 |\n| 262144 | 5.66404e+07 | 0.004628 | 0.004636 | 0.004638 | 0.004642 |\n| 1048576 | 5.66507e+07 | 0.018509 | 0.018547 | 0.018556 | 0.018583 |","hasPerformance",[],{"_85":233,"_112":234,"_113":-5,"_121":34,"_120":34,"_122":232,"_216":90},{"_85":28,"_112":28,"_113":-5,"_121":34,"_120":34,"_122":232,"_216":90},[],"-","Latest (21.04.0)","nvidia/dle/ncf_pyt","key","nvidia/dle/resources/ncf_pyt","resourceId","ncf_pyt","shortDescription","The NCF model focuses on providing recommendations. This is a modified implementation with improved overfitting and better accuracy.","This resource is using open-source code maintained in github (see the quick-start-guide section) and available for download from [NGC](https://catalog.ngc.nvidia.com/orgs/nvidia/teams/dle/resources/ncf-pyt)\n\n\nThe NCF model focuses on providing recommendations, also known as collaborative filtering with implicit feedback. The training data for this model should contain binary information about whether a user interacted with a specific item.\nNCF was first described by Xiangnan He, Lizi Liao, Hanwang Zhang, Liqiang Nie, Xia Hu and Tat-Seng Chua in the [Neural Collaborative Filtering paper](https://arxiv.org/abs/1708.05031).\n\nThe implementation in this repository focuses on the NeuMF instantiation of the NCF architecture.\nWe modified it to use dropout in the FullyConnected layers. This reduces overfitting and increases the final accuracy.\nTraining the other two instantiations of NCF (GMF and MLP) is not supported. \n \nContrary to the original paper, we benchmark the model on the larger [ML-20m dataset](https://grouplens.org/datasets/movielens/20m/)\ninstead of using the smaller [ML-1m](https://grouplens.org/datasets/movielens/1m/) dataset because we think this is more realistic for production type environments.\nHowever, using the ML-1m dataset is also supported.\n\nThis model is trained with mixed precision using Tensor Cores on Volta, Turing, and the NVIDIA Ampere GPU architectures.\nTherefore, researchers can get results 2x faster than training without Tensor Cores while experiencing the benefits of mixed precision training.\n\n### Model architecture\n\nThis model is based mainly on Embedding and FullyConnected layers. The control flow is divided into two branches:\n* Multi Layer Perceptron (MLP) branch, which transforms the input through FullyConnected layers with ReLU activations, and dropout.\n* Matrix Factorization (MF) branch, which performs collaborative filtering factorization.\nEach user and each item has two embedding vectors associated with it -- one for the MLP branch and the other for the MF branch.\n\nThe outputs from those branches are concatenated and fed to the final FullyConnected layer with sigmoid activation.\nThis can be interpreted as a probability of a user interacting with a given item.\n\n

\n \n
\nFigure 1. The architecture of a Neural Collaborative Filtering model. Taken from the Neural Collaborative Filtering paper.\n

\n\n### Default configuration\n\nThe following features were implemented in this model:\n * Automatic Mixed Precision (AMP)\n * Data-parallel multi-GPU training and evaluation \n * Dropout\n * Gradient accumulation\n\nThe following performance optimizations were implemented in this model:\n * FusedAdam optimizer\n * Approximate train negative sampling\n * Caching all the positive training samples in the device memory\n\n### Feature support matrix\nThis model supports the following features:\n\n| **Feature** | **NCF PyTorch** | \n|:---:|:--------:|\n| Automatic Mixed Precision (AMP) | Yes |\n| Multi-GPU training with Distributed Data Parallel (DDP) | Yes |\n| Fused Adam | Yes |\n\n#### Features\n\n* Automatic Mixed Precision - This implementation of NCF uses AMP to implement mixed precision training.\nIt allows us to use FP16 training with FP32 master weights by modifying just three lines of code. \n* Multi-GPU training with Distributed Data Parallel - uses Apex's DDP to implement efficient multi-GPU training with NCCL.\n* Fused Adam - We use a special implementation of the Adam implementation provided by the Apex package. It fuses some operations for faster weight updates.\nSince NCF is a relatively lightweight model with a large number of parameters, we've observed significant performance improvements from using FusedAdam.\n\n### Mixed precision training\n\nMixed precision is the combined use of different numerical precisions in a computational method. \n[Mixed precision](https://arxiv.org/abs/1710.03740) training offers significant computational speedup by performing \noperations in half-precision format while storing minimal information in single-precision to retain as much information\nas possible in critical parts of the network. Since the introduction of \n[tensor cores](https://developer.nvidia.com/tensor-cores) in Volta, \nand following with both the Turing and Ampere architectures, significant training speedups are experienced by switching \nto mixed precision -- up to 3x overall speedup on the most arithmetically intense model architectures. \nUsing [mixed precision training](https://docs.nvidia.com/deeplearning/performance/mixed-precision-training/index.html) \npreviously required two steps:\n1. Porting the model to use the FP16 data type where appropriate.\n2. Adding loss scaling to preserve small gradient values.\n\nThe ability to train deep learning networks with lower precision was introduced in the Pascal architecture and first supported in [CUDA 8](https://devblogs.nvidia.com/parallelforall/tag/fp16/) in the NVIDIA Deep Learning SDK.\n\nFor information about:\n- How to train using mixed precision, refer to the [Mixed Precision Training](https://arxiv.org/abs/1710.03740) paper and [Training With Mixed Precision](https://docs.nvidia.com/deeplearning/sdk/mixed-precision-training/index.html) documentation.\n- Techniques used for mixed precision training, refer to the [Mixed-Precision Training of Deep Neural Networks](https://devblogs.nvidia.com/mixed-precision-training-deep-neural-networks/) blog.\n- APEX tools for mixed precision training, refer to the [NVIDIA Apex: Tools for Easy Mixed-Precision Training in PyTorch](https://devblogs.nvidia.com/apex-pytorch-easy-mixed-precision-training/).\n\n#### Enabling mixed precision\n\nUsing the Automatic Mixed Precision (AMP) package requires two modifications in the source code.\nThe first one is to initialize the model and the optimizer using the `amp.initialize` function:\n```python\nmodel, optimizer = amp.initialize(model, optimizer, opt_level=\"O2\"\n keep_batchnorm_fp32=False, loss_scale='dynamic')\n```\n\nThe second one is to use the AMP's loss scaling context manager:\n```python\nwith amp.scale_loss(loss, optimizer) as scaled_loss:\n scaled_loss.backward()\n```\n\n#### Enabling TF32\n\nTensorFloat-32 (TF32) is the new math mode in [NVIDIA A100](https://www.nvidia.com/en-us/data-center/a100/) GPUs for \nhandling the matrix math, also called tensor operations. TF32 running on Tensor Cores in A100 GPUs can provide up to 10x \nspeedups compared to single-precision floating-point math (FP32) on Volta GPUs. \n\nTF32 Tensor Cores can speed up networks using FP32, typically with no loss of accuracy. It is more robust than FP16 for \nmodels which require a high dynamic range for weights or activations.\n\nFor more information, refer to the [TensorFloat-32 in the A100 GPU Accelerates AI Training, HPC up to 20x](https://blogs.nvidia.com/blog/2020/05/14/tensorfloat-32-precision-format/) blog post.\n\nTF32 is supported in the NVIDIA Ampere GPU architecture and is enabled by default.\n\n### BYO dataset functionality overview\n\nThis section describes how you can train the DeepLearningExamples RecSys models on your own datasets without changing the model or data loader and with similar performance to the one published in each repository.\nThis can be achieved thanks to Dataset Feature Specification, which describes how the dataset, data loader and model interact with each other during training, inference and evaluation.\nDataset Feature Specification has a consistent format across all recommendation models in NVIDIA's DeepLearningExamples repository, regardless of dataset file type and the data loader, giving you the flexibility to train RecSys models on your own datasets.\n\n- [Dataset Feature Specification](#dataset-feature-specification)\n- [Data Flow in Recommendation Models in DeepLearning examples](#data-flow-in-recommendation-models-in-deeplearning-examples)\n- [Example of Dataset Feature Specification](#example-of-dataset-feature-specification)\n- [BYO dataset functionality](#byo-dataset-functionality)\n- [Glossary](#glossary)\n\n#### Dataset feature specification\n\nData flow can be described abstractly:\nInput data consists of a list of rows. Each row has the same number of columns; each column represents a feature.\nThe columns are retrieved from the input files, loaded, aggregated into channels and supplied to the model/training script. \n\nFeatureSpec contains metadata to configure this process and can be divided into three parts:\n\n* Specification of how data is organized on disk (source_spec). It describes which feature (from feature_spec) is stored in which file and how files are organized on disk.\n\n* Specification of features (feature_spec). Describes a dictionary of features, where key is feature name and values are features' characteristics such as dtype and other metadata (for example, cardinalities for categorical features)\n\n* Specification of model's inputs and outputs (channel_spec). Describes a dictionary of model's inputs where keys specify model channel's names and values specify lists of features to be loaded into that channel. Model's channels are groups of data streams to which common model logic is applied, for example categorical/continuous data, user/item ids. Required/available channels depend on the model\n\nThe FeatureSpec is a common form of description regardless of underlying dataset format, dataset data loader form and model. \n\n#### Data flow in NVIDIA Deep Learning Examples recommendation models\n\nThe typical data flow is as follows:\n* S.0. Original dataset is downloaded to a specific folder.\n* S.1. Original dataset is preprocessed into Intermediary Format. For each model, the preprocessing is done differently, using different tools. The Intermediary Format also varies (for example, for NCF implementation in the PyTorch model, the Intermediary Format is Pytorch tensors in *.pt files.)\n* S.2. The Preprocessing Step outputs Intermediary Format with dataset split into training and validation/testing parts along with the Dataset Feature Specification yaml file. Metadata in the preprocessing step is automatically calculated.\n* S.3. Intermediary Format data together with Dataset Feature Specification are fed into training/evaluation scripts. Data loader reads Intermediary Format and feeds the data into the model according to the description in the Dataset Feature Specification.\n* S.4. The model is trained and evaluated\n\n

\n \n
\n\nFig.1. Data flow in Recommender models in NVIDIA Deep Learning Examples repository. Channels of the model are drawn in green.\n

\n\n#### Example of dataset feature specification\n\nAs an example, let's consider a Dataset Feature Specification for a small CSV dataset.\n\n```yaml\nfeature_spec:\n user_gender:\n dtype: torch.int8\n cardinality: 3 #M,F,Other\n user_age: #treated as numeric value\n dtype: torch.int8\n user_id:\n dtype: torch.int32\n cardinality: 2655\n item_id:\n dtype: torch.int32\n cardinality: 856\n label:\n dtype: torch.float32\n\nsource_spec:\n train:\n - type: csv\n features:\n - user_gender\n - user_age\n files:\n - train_data_0_0.csv\n - train_data_0_1.csv\n - type: csv\n features:\n - user_id\n - item_id\n - label\n files:\n - train_data_1.csv\n test:\n - type: csv\n features:\n - user_id\n - item_id\n - label\n - user_gender\n - user_age\n \n files:\n - test_data.csv\n\nchannel_spec:\n numeric_inputs: \n - user_age\n categorical_user_inputs: \n - user_gender\n - user_id\n categorical_item_inputs: \n - item_id\n label_ch: \n - label\n```\n\nThe data contains five features: (user_gender, user_age, user_id, item_id, label). Their data types and necessary metadata are described in the feature specification section.\n\nIn the source mapping section, two mappings are provided: one describes the layout of the training data, the other of the testing data. The layout for training data has been chosen arbitrarily to showcase the flexibility.\nThe train mapping consists of two chunks. The first one contains user_gender and user_age, saved as a CSV, and is further broken down into two files. For specifics of the layout, refer to the following example and consult the glossary. The second chunk contains the remaining columns and is saved in a single file. Notice that the order of columns is different in the second chunk - this is alright, as long as the order matches the order in that file (that is, columns in the .csv are also switched)\n\nLet's break down the train source mapping. The table contains example data color-paired to the files containing it.\n\n

\n\n

\n\nThe channel spec describes how the data will be consumed. Four streams will be produced and available to the script/model. The feature specification does not specify what happens further: names of these streams are only lookup constants defined by the model/script.\nBased on this example, we can speculate that the model has three input channels: numeric_inputs, categorical_user_inputs, categorical_item_inputs, and one output channel: label.\nFeature names are internal to the FeatureSpec and can be freely modified.\n\n#### BYO dataset functionality\n\nIn order to train any Recommendation model in NVIDIA Deep Learning Examples one can follow one of three possible ways:\n* One delivers already preprocessed dataset in the Intermediary Format supported by data loader used by the training script (different models use different data loaders) together with FeatureSpec yaml file describing at least specification of dataset, features and model channels\n\n* One uses a transcoding script\n\n* One delivers dataset in non-preprocessed form and uses preprocessing scripts that are a part of the model repository. In order to use already existing preprocessing scripts, the format of the dataset needs to match the one of the original datasets. This way, the FeatureSpec file will be generated automatically, but the user will have the same preprocessing as in the original model repository.\n\n#### Glossary\n\nThe Dataset Feature Specification consists of three mandatory and one optional section:\n\nfeature_spec provides a base of features that may be referenced in other sections, along with their metadata.\n\tFormat: dictionary (feature name) => (metadata name => metadata value)
\n\nsource_spec provides information necessary to extract features from the files that store them. \n\tFormat: dictionary (mapping name) => (list of chunks)
\n\n* Mappings are used to represent different versions of the dataset (think: train/validation/test, k-fold splits). A mapping is a list of chunks.
\n* Chunks are subsets of features that are grouped together for saving. For example, some formats may constrain data saved in one file to a single data type. In that case, each data type would correspond to at least one chunk. Another example where this might be used is to reduce file size and enable more parallel loading. Chunk description is a dictionary of three keys:
\n * type provides information about the format in which the data is stored. Not all formats are supported by all models.
\n * features is a list of features that are saved in a given chunk. Order of this list may matter: for some formats, it is crucial for assigning read data to the proper feature.
\n * files is a list of paths to files where the data is saved. For Feature Specification in yaml format, these paths are assumed to be relative to the yaml file's directory (basename). Order of this list matters: It is assumed that rows 1 to i appear in the first file, rows i+1 to j in the next one, etc.
\n\nchannel_spec determines how features are used. It is a mapping (channel name) => (list of feature names). \n\nChannels are model specific magic constants. In general, data within a channel is processed using the same logic. Example channels: model output (labels), categorical ids, numerical inputs, user data, and item data.\n\nmetadata is a catch-all, wildcard section: If there is some information about the saved dataset that does not fit into the other sections, you can store it here.","displayName","NCF for PyTorch",["D",1680632767239],"publisher","NVIDIA Deep Learning Examples","latestVersion","collectionCounts","imageUrl","/api/image/nvidia/dle/resources/ncf_pyt","rawImageUrl","https://assets.nvidiagrid.net/ngc/logos/Deep-Learning-Examples.png","canGuestDownload","latestVersionSizeInBytes","licenseTerms","isMultinodeEnabled","isMultiArchitecture","bias","explainability","privacy","safetyAndSecurity","featureLabels",[],"displayLabels",[274],"policyLabels",[],"systemLabels",[],"products",[],"hasNimMetadata",{"_243":275,"_276":277,"_278":279,"_280":87},"Recommendation","value","uscs_recommedation","filter",{"_281":282,"_276":277},"category","field","application",{"_151":1268,"_153":1269},"ngcApps",[1221],"csaApiBaseUrl","https://www.nvidia.com/ai-agents/csa-agent/dcsm","datadogRum",{"_1202":1203,"_1204":1205,"_1206":1207,"_1208":1209,"_1210":294,"_109":322,"_1211":1212,"_1213":59,"_1214":90,"_1215":90,"_1216":1217,"_1218":90,"_1219":1220},"amplitude",{"_1199":1200,"_836":-7,"_1201":34},"disableJs","environment","prod","featureFlagsBlob","JTdCJTIydGVzdEZsYWclMjIlM0FmYWxzZSUyQyUyMnNob3dTZWFyY2hMYWJlbFNldEZpbHRlcnMlMjIlM0FmYWxzZSUyQyUyMmVuYWJsZUNzYVdpZGdldCUyMiUzQWZhbHNlJTJDJTIyZW5hYmxlUHJvZHVjdEdyb3VwaW5nQnlTdWJzY3JpcHRpb24lMjIlM0FmYWxzZSUyQyUyMmNvbnRhaW5lclNjYW5uaW5nUmVkZXNpZ25NMiUyMiUzQWZhbHNlJTdE","featureFlagsGuarded","isDevLikeEnv","isPlaywright","isMobileDevice","categorizedOrgs",{"_1193":1194,"_1195":1196,"_1197":1198},"currentOrgContext","currentContextRoles",[],"enterpriseOrgs",[],"manageableExternalUserGroups",[],"locale","en","ngcApiOrigin","ngcHelmOrigin","origin","https://catalog.ngc.nvidia.com","usesMocks","theme","year",2026,"now",1786029669921,"1.11.0","defaultSearchSuggestions",{"_1076":404,"_1077":1078,"_1079":1080,"_1081":1082},"announcements",[],"consentState",{"_361":1066,"_1067":90,"_1068":34,"_1069":34,"_1070":1071},"consentConfig",{"_343":344,"_345":346,"_347":348,"_349":350,"_351":352,"_353":354,"_355":356,"_357":358,"_359":360,"_361":362},"showContactAdmin",1786029669923,[],[],[336,337,338,339,340,341],{"_48":62,"_50":73,"_52":146,"_54":-5,"_56":-5,"_58":59},{"_48":62,"_50":69,"_52":146,"_54":-5,"_56":-5,"_58":59},{"_48":62,"_50":67,"_52":146,"_54":-5,"_56":-5,"_58":59},{"_48":62,"_50":63,"_52":146,"_54":-5,"_56":-5,"_58":59},{"_48":60,"_50":61,"_52":53,"_54":55,"_56":342,"_58":59},{"_48":49,"_50":51,"_52":146,"_54":-5,"_56":-5,"_58":59},4,"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",{"_1058":1059,"_1060":1061,"_1062":1063,"_1064":1065},"bannerAcceptLabel","Accept All","bannerRejectLabel","Reject Optional","bannerSettingsLabel","Manage Settings","preferenceSaveLabel","Save Choices","preferenceRejectLabel","Decline All","alwaysActiveText","Always Active","groups",[363,364,365,366,367],{"_368":748,"_112":749,"_210":750,"_372":90,"_373":16,"_375":751,"_377":752},{"_368":651,"_112":652,"_210":653,"_372":34,"_373":57,"_375":654,"_377":655},{"_368":559,"_112":560,"_210":561,"_372":34,"_373":342,"_375":562,"_377":563},{"_368":379,"_112":380,"_210":381,"_372":34,"_373":382,"_375":383,"_377":384},{"_368":369,"_112":370,"_210":371,"_372":34,"_373":374,"_375":376,"_377":378},"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,[501,502,503,504,505,506,507,508,509,510,511,512,513,514,515,516,517,518,519,520,521,522],[385,386,387,388,389,390,391,392,393,394,395,396,397,398,399,400,401],{"_402":496,"_243":496,"_210":404,"_375":497},{"_402":492,"_243":492,"_210":404,"_375":493},{"_402":487,"_243":487,"_210":404,"_375":488},{"_402":483,"_243":483,"_210":404,"_375":484},{"_402":478,"_243":478,"_210":404,"_375":479},{"_402":473,"_243":473,"_210":404,"_375":474},{"_402":465,"_243":465,"_210":404,"_375":466},{"_402":457,"_243":457,"_210":404,"_375":458},{"_402":452,"_243":452,"_210":404,"_375":453},{"_402":447,"_243":447,"_210":404,"_375":448},{"_402":442,"_243":442,"_210":404,"_375":443},{"_402":437,"_243":437,"_210":404,"_375":438},{"_402":432,"_243":432,"_210":404,"_375":433},{"_402":425,"_243":425,"_210":404,"_375":426},{"_402":420,"_243":420,"_210":404,"_375":421},{"_402":413,"_243":413,"_210":404,"_375":414},{"_402":403,"_243":403,"_210":404,"_375":405},"hostName","amazon-adsystem.com","",[406,407],{"_112":412,"_409":403,"_410":34,"_210":411},{"_112":408,"_409":403,"_410":34,"_210":411},"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",[415,416],{"_112":419,"_409":413,"_410":34,"_210":418},{"_112":417,"_409":413,"_410":34,"_210":418},"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",[422],{"_112":423,"_409":420,"_410":34,"_210":424},"MUID","This domain is owned by Mircosoft - it is the site for the search engine Bing.","assets.adobedtm.com",[427,428],{"_112":431,"_409":425,"_410":34,"_210":430},{"_112":429,"_409":425,"_410":34,"_210":430},"kndctr_F207D74D549850760A4C98C6_AdobeOrg_*","If you know anything about this host and its uses please add your contributions here. ","kndctr_F207D74D549850760A4C98C6_AdobeOrg_consent","tiktok.com",[434],{"_112":435,"_409":432,"_410":34,"_210":436},"_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",[439],{"_112":440,"_409":437,"_410":34,"_210":441},"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",[444],{"_112":445,"_409":442,"_410":34,"_210":446},"sp","Testing","google.com",[449],{"_112":450,"_409":447,"_410":34,"_210":451},"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",[454],{"_112":455,"_409":452,"_410":34,"_210":456},"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",[459,460],{"_112":463,"_409":457,"_410":34,"_210":464},{"_112":461,"_409":457,"_410":34,"_210":462},"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",[467,468],{"_112":471,"_409":465,"_410":34,"_210":472},{"_112":469,"_409":465,"_410":34,"_210":470},"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",[475],{"_112":476,"_409":473,"_410":34,"_210":477},"demdex","This cookie helps Adobe Audience Manger perform basic functions such as visitor identification, ID synchronization, segmentation, modeling, reporting, etc.","im-apps.net",[480],{"_112":481,"_409":478,"_410":34,"_210":482},"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",[485],{"_112":486,"_409":483,"_410":34,"_210":446},"R","6sc.co",[489],{"_112":490,"_409":487,"_410":34,"_210":491},"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",[494],{"_112":495,"_409":492,"_410":34,"_210":446},"_hly_gvid","t.contentsquare.net",[498],{"_112":499,"_409":496,"_410":90,"_210":500},"_cs_*","This is a session cookie. This is a pattern type cookie where a random 13-digit number is appended to the prefix _cs_.",{"_112":557,"_409":524,"_410":34,"_210":558},{"_112":555,"_409":524,"_410":34,"_210":556},{"_112":553,"_409":524,"_410":34,"_210":554},{"_112":435,"_409":524,"_410":34,"_210":552},{"_112":529,"_409":524,"_410":90,"_210":531},{"_112":551,"_409":524,"_410":34,"_210":404},{"_112":550,"_409":527,"_410":34,"_210":404},{"_112":548,"_409":524,"_410":34,"_210":549},{"_112":546,"_409":524,"_410":34,"_210":547},{"_112":545,"_409":524,"_410":34,"_210":404},{"_112":543,"_409":524,"_410":34,"_210":544},{"_112":529,"_409":542,"_410":90,"_210":531},{"_112":540,"_409":527,"_410":34,"_210":541},{"_112":540,"_409":530,"_410":34,"_210":541},{"_112":499,"_409":524,"_410":34,"_210":500},{"_112":538,"_409":524,"_410":34,"_210":539},{"_112":445,"_409":536,"_410":34,"_210":537},{"_112":534,"_409":524,"_410":34,"_210":535},{"_112":532,"_409":524,"_410":34,"_210":533},{"_112":529,"_409":530,"_410":90,"_210":531},{"_112":526,"_409":527,"_410":34,"_210":528},{"_112":523,"_409":524,"_410":90,"_210":525},"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.",[593,594,595,596,597,598,599,600,601,602,603,604,605,606,607,608,609,610,611,612,613,614,615,616,617],[564,565,566,567,568],{"_402":478,"_243":478,"_210":404,"_375":588},{"_402":452,"_243":452,"_210":404,"_375":583},{"_402":578,"_243":578,"_210":404,"_375":579},{"_402":573,"_243":573,"_210":404,"_375":574},{"_402":569,"_243":569,"_210":404,"_375":570},"nvidia.partners",[571],{"_112":572,"_409":569,"_410":34,"_210":446},"edge","app.hushly.com",[575],{"_112":576,"_409":573,"_410":34,"_210":577},"*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",[580],{"_112":581,"_409":578,"_410":34,"_210":582},"_octo","This domain is owned by Github. The main business activity is: Source code repository and collaboration platform",[584,585],{"_112":587,"_409":452,"_410":34,"_210":456},{"_112":586,"_409":452,"_410":34,"_210":456},"lidc","li_gc",[589,590],{"_112":592,"_409":478,"_410":34,"_210":482},{"_112":591,"_409":478,"_410":34,"_210":482},"imid_*","imid",{"_112":618,"_409":527,"_410":34,"_210":620},{"_112":649,"_409":524,"_410":34,"_210":650},{"_112":618,"_409":648,"_410":34,"_210":620},{"_112":646,"_409":524,"_410":90,"_210":647},{"_112":644,"_409":524,"_410":34,"_210":645},{"_112":643,"_409":626,"_410":90,"_210":404},{"_112":618,"_409":622,"_410":34,"_210":620},{"_112":641,"_409":524,"_410":34,"_210":642},{"_112":618,"_409":530,"_410":34,"_210":620},{"_112":629,"_409":527,"_410":34,"_210":404},{"_112":618,"_409":640,"_410":34,"_210":620},{"_112":638,"_409":524,"_410":34,"_210":639},{"_112":635,"_409":636,"_410":34,"_210":637},{"_112":633,"_409":634,"_410":34,"_210":404},{"_112":618,"_409":630,"_410":34,"_210":620},{"_112":632,"_409":524,"_410":34,"_210":404},{"_112":618,"_409":631,"_410":34,"_210":620},{"_112":629,"_409":630,"_410":34,"_210":404},{"_112":628,"_409":626,"_410":90,"_210":404},{"_112":618,"_409":627,"_410":34,"_210":620},{"_112":618,"_409":626,"_410":34,"_210":620},{"_112":618,"_409":625,"_410":34,"_210":620},{"_112":624,"_409":524,"_410":90,"_210":620},{"_112":621,"_409":622,"_410":90,"_210":623},{"_112":618,"_409":619,"_410":34,"_210":620},"_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.",[676,677,678,679,680,681,682,683,684,685,686,687,688,689,690,691,692,693,694,695,696,697,698,699,700,701,702,703,704],[656,657,658,659],{"_402":671,"_243":671,"_210":404,"_375":672},{"_402":666,"_243":666,"_210":404,"_375":667},{"_402":465,"_243":465,"_210":404,"_375":663},{"_402":578,"_243":578,"_210":404,"_375":660},[661],{"_112":662,"_409":578,"_410":90,"_210":582},"_gh_sess",[664],{"_112":665,"_409":465,"_410":90,"_210":472},"YSC","bizible.com",[668],{"_112":669,"_409":666,"_410":34,"_210":670},"_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",[673],{"_112":674,"_409":671,"_410":34,"_210":675},"tuid","This domain is owned by Ookla which provides internet speed testing services through its Speedtest application.",{"_112":746,"_409":524,"_410":34,"_210":747},{"_112":731,"_409":745,"_410":90,"_210":732},{"_112":714,"_409":527,"_410":34,"_210":639},{"_112":714,"_409":627,"_410":34,"_210":639},{"_112":743,"_409":524,"_410":34,"_210":744},{"_112":731,"_409":530,"_410":90,"_210":732},{"_112":741,"_409":524,"_410":34,"_210":742},{"_112":731,"_409":527,"_410":90,"_210":732},{"_112":739,"_409":626,"_410":34,"_210":740},{"_112":731,"_409":648,"_410":90,"_210":732},{"_112":711,"_409":527,"_410":34,"_210":713},{"_112":737,"_409":524,"_410":90,"_210":738},{"_112":735,"_409":524,"_410":34,"_210":736},{"_112":733,"_409":524,"_410":90,"_210":734},{"_112":731,"_409":622,"_410":90,"_210":732},{"_112":729,"_409":524,"_410":34,"_210":730},{"_112":727,"_409":524,"_410":34,"_210":728},{"_112":725,"_409":524,"_410":90,"_210":726},{"_112":723,"_409":631,"_410":34,"_210":724},{"_112":711,"_409":622,"_410":34,"_210":713},{"_112":721,"_409":524,"_410":34,"_210":722},{"_112":719,"_409":524,"_410":90,"_210":720},{"_112":717,"_409":524,"_410":34,"_210":718},{"_112":715,"_409":524,"_410":34,"_210":716},{"_112":714,"_409":524,"_410":34,"_210":639},{"_112":711,"_409":712,"_410":34,"_210":713},{"_112":709,"_409":524,"_410":34,"_210":710},{"_112":707,"_409":524,"_410":90,"_210":708},{"_112":705,"_409":524,"_410":34,"_210":706},"_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.",[893,894,895,896,897,898,899,900,901,902,903,904,905,906,907,908,909,910,911,912,913,914,915,916,917,918,919,920,921,922,923,924,925,926,927,928,929,930,931,932,933,934,935,936,937,938,939,940,941,942,943,944,945,946,947,948,949,950,951,952,953,954,955,956,957,958,959,960,961,962,963,964,965],[753,754,755,756,757,758,759,760,761,762,763,764,765,766,767,768,769,770,771,772,773,774,775,776,777],{"_402":890,"_243":890,"_210":404,"_375":891},{"_402":884,"_243":884,"_210":404,"_375":885},{"_402":881,"_243":881,"_210":404,"_375":882},{"_402":869,"_243":869,"_210":404,"_375":870},{"_402":866,"_243":866,"_210":404,"_375":867},{"_402":863,"_243":863,"_210":404,"_375":864},{"_402":858,"_243":858,"_210":404,"_375":859},{"_402":465,"_243":465,"_210":404,"_375":855},{"_402":852,"_243":852,"_210":404,"_375":853},{"_402":457,"_243":457,"_210":404,"_375":850},{"_402":452,"_243":452,"_210":404,"_375":848},{"_402":843,"_243":843,"_210":404,"_375":844},{"_402":838,"_243":838,"_210":404,"_375":839},{"_402":833,"_243":833,"_210":404,"_375":834},{"_402":830,"_243":830,"_210":404,"_375":831},{"_402":578,"_243":578,"_210":404,"_375":827},{"_402":437,"_243":437,"_210":404,"_375":824},{"_402":573,"_243":573,"_210":404,"_375":817},{"_402":569,"_243":569,"_210":404,"_375":811},{"_402":805,"_243":805,"_210":404,"_375":806},{"_402":800,"_243":800,"_210":404,"_375":801},{"_402":795,"_243":795,"_210":404,"_375":796},{"_402":791,"_243":791,"_210":404,"_375":792},{"_402":786,"_243":786,"_210":404,"_375":787},{"_402":778,"_243":778,"_210":404,"_375":779},"polyfill.io",[780,781],{"_112":784,"_409":778,"_410":34,"_210":785},{"_112":782,"_409":778,"_410":34,"_210":783},"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",[788],{"_112":789,"_409":786,"_410":34,"_210":790},"_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",[793],{"_112":794,"_409":791,"_410":34,"_210":446},"__cflb","instagram.com",[797],{"_112":798,"_409":795,"_410":34,"_210":799},"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",[802],{"_112":803,"_409":800,"_410":34,"_210":804},"__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",[807,808],{"_112":810,"_409":805,"_410":34,"_210":446},{"_112":809,"_409":805,"_410":34,"_210":446},"AWSALBTG","AWSALBTGCORS",[812,813],{"_112":815,"_409":569,"_410":34,"_210":816},{"_112":814,"_409":569,"_410":34,"_210":446},"ak_bmsc","bm_mi","This domain is likely associated with NVIDIA's partner program, which could involve marketing, sales, and support services for partners.",[818,819,820],{"_112":823,"_409":573,"_410":34,"_210":446},{"_112":822,"_409":573,"_410":90,"_210":446},{"_112":821,"_409":573,"_410":34,"_210":446},"AWSALBCORS","JSESSIONID","AWSALB",[825],{"_112":803,"_409":437,"_410":34,"_210":826},"This is a CloudFoundry cookie",[828],{"_112":829,"_409":578,"_410":34,"_210":582},"logged_in","hcaptcha.com",[832],{"_112":803,"_409":830,"_410":34,"_210":826},"tracking.intentsify.io",[835],{"_112":836,"_409":833,"_410":34,"_210":837},"userId","Used to track usage of the embedded video player.","static.reo.dev",[840],{"_112":841,"_409":838,"_410":34,"_210":842},"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",[845],{"_112":846,"_409":843,"_410":34,"_210":847},"dd_cookie_test_","dd_cookie_test",[849],{"_112":803,"_409":452,"_410":34,"_210":826},[851],{"_112":803,"_409":457,"_410":34,"_210":826},"supabase.co",[854],{"_112":803,"_409":852,"_410":34,"_210":826},[856],{"_112":857,"_409":465,"_410":34,"_210":472},"VISITOR_PRIVACY_METADATA","dosa-prd.azurewebsites.net",[860],{"_112":861,"_409":858,"_410":34,"_210":862},"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",[865],{"_112":803,"_409":863,"_410":34,"_210":826},"cdn.readme.io",[868],{"_112":803,"_409":866,"_410":34,"_210":826},"queue-it.net",[871,872,873,874,875],{"_112":880,"_409":869,"_410":34,"_210":404},{"_112":879,"_409":869,"_410":34,"_210":404},{"_112":878,"_409":869,"_410":34,"_210":404},{"_112":877,"_409":869,"_410":34,"_210":404},{"_112":876,"_409":869,"_410":34,"_210":404},"Queue-it-{queueId}","Queue-it-{dmonvidia }________{eventId}","Queue-it-token","queueitsoftblock_{queueId}","Queue-it-{nvidia}________{eventId}","x.com",[883],{"_112":803,"_409":881,"_410":34,"_210":804},"api.hcaptcha.com",[886,887],{"_112":889,"_409":884,"_410":34,"_210":446},{"_112":794,"_409":884,"_410":34,"_210":888},"This domain is associated with hCaptcha, a service that provides CAPTCHA and reCAPTCHA alternatives for user verification and bot protection.","hmt_id","vimeo.com",[892],{"_112":803,"_409":890,"_410":34,"_210":826},{"_112":1026,"_409":524,"_410":34,"_210":1057},{"_112":1055,"_409":619,"_410":90,"_210":1056},{"_112":1048,"_409":1017,"_410":90,"_210":1049},{"_112":1054,"_409":712,"_410":90,"_210":404},{"_112":1052,"_409":524,"_410":34,"_210":1053},{"_112":810,"_409":1038,"_410":34,"_210":977},{"_112":985,"_409":1047,"_410":90,"_210":987},{"_112":809,"_409":976,"_410":34,"_210":977},{"_112":988,"_409":1047,"_410":34,"_210":989},{"_112":1050,"_409":524,"_410":90,"_210":1051},{"_112":1048,"_409":1024,"_410":90,"_210":1049},{"_112":982,"_409":1047,"_410":34,"_210":984},{"_112":1045,"_409":524,"_410":34,"_210":1046},{"_112":809,"_409":968,"_410":34,"_210":977},{"_112":973,"_409":1044,"_410":34,"_210":975},{"_112":821,"_409":974,"_410":34,"_210":990},{"_112":823,"_409":974,"_410":34,"_210":1041},{"_112":988,"_409":983,"_410":34,"_210":989},{"_112":1042,"_409":622,"_410":34,"_210":1043},{"_112":823,"_409":648,"_410":34,"_210":1041},{"_112":1039,"_409":524,"_410":34,"_210":1040},{"_112":815,"_409":648,"_410":34,"_210":1034},{"_112":1011,"_409":627,"_410":90,"_210":404},{"_112":809,"_409":1038,"_410":34,"_210":977},{"_112":1036,"_409":976,"_410":90,"_210":1037},{"_112":1000,"_409":524,"_410":34,"_210":991},{"_112":973,"_409":524,"_410":34,"_210":975},{"_112":1035,"_409":536,"_410":34,"_210":404},{"_112":815,"_409":524,"_410":34,"_210":1034},{"_112":823,"_409":968,"_410":34,"_210":1033},{"_112":1031,"_409":631,"_410":90,"_210":1032},{"_112":1029,"_409":524,"_410":90,"_210":1030},{"_112":1028,"_409":524,"_410":34,"_210":404},{"_112":1026,"_409":974,"_410":34,"_210":1027},{"_112":1025,"_409":536,"_410":34,"_210":404},{"_112":1016,"_409":1024,"_410":90,"_210":1018},{"_112":1022,"_409":627,"_410":34,"_210":1023},{"_112":1020,"_409":712,"_410":34,"_210":1021},{"_112":1019,"_409":530,"_410":34,"_210":404},{"_112":1016,"_409":1017,"_410":90,"_210":1018},{"_112":1014,"_409":983,"_410":90,"_210":1015},{"_112":1012,"_409":524,"_410":34,"_210":1013},{"_112":1011,"_409":527,"_410":90,"_210":404},{"_112":982,"_409":986,"_410":34,"_210":984},{"_112":814,"_409":1001,"_410":34,"_210":991},{"_112":1009,"_409":527,"_410":34,"_210":1010},{"_112":1007,"_409":524,"_410":34,"_210":1008},{"_112":1005,"_409":524,"_410":34,"_210":1006},{"_112":1004,"_409":524,"_410":34,"_210":404},{"_112":846,"_409":527,"_410":34,"_210":1003},{"_112":1000,"_409":648,"_410":34,"_210":991},{"_112":803,"_409":619,"_410":34,"_210":1002},{"_112":1000,"_409":1001,"_410":34,"_210":991},{"_112":999,"_409":536,"_410":34,"_210":404},{"_112":998,"_409":524,"_410":34,"_210":971},{"_112":996,"_409":983,"_410":34,"_210":997},{"_112":809,"_409":994,"_410":34,"_210":995},{"_112":993,"_409":536,"_410":34,"_210":404},{"_112":992,"_409":524,"_410":34,"_210":404},{"_112":985,"_409":983,"_410":90,"_210":987},{"_112":814,"_409":648,"_410":34,"_210":991},{"_112":821,"_409":648,"_410":34,"_210":990},{"_112":988,"_409":986,"_410":34,"_210":989},{"_112":985,"_409":986,"_410":90,"_210":987},{"_112":982,"_409":983,"_410":34,"_210":984},{"_112":980,"_409":976,"_410":90,"_210":981},{"_112":978,"_409":634,"_410":34,"_210":979},{"_112":810,"_409":976,"_410":34,"_210":977},{"_112":973,"_409":974,"_410":34,"_210":975},{"_112":803,"_409":626,"_410":34,"_210":972},{"_112":970,"_409":524,"_410":34,"_210":971},{"_112":821,"_409":968,"_410":34,"_210":969},{"_112":966,"_409":524,"_410":34,"_210":967},"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.

",{"_748":16,"_651":59,"_559":59,"_379":59,"_369":59},"hasConsented","gpcApplied","showBanner","geolocation",{"_1072":1073,"_1074":1075},"country","US","state","OR","query","categories",[1095,1096,1097,1098,1099],"optionUrls",[1083,1084,1085,1086,1087,1088,1089,1090,1091,1092,1093,1094],"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/-",{"_20":1173,"_1101":1174,"_1103":59},{"_20":1138,"_1101":1139,"_1103":57},{"_20":1121,"_1101":1122,"_1103":1123},{"_20":21,"_1101":1114,"_1103":1115},{"_20":1100,"_1101":1102,"_1103":19},"helm-charts","items",[1104,1105],"startIndex",{"_20":1100,"_243":1111,"_204":205,"_206":-7,"_238":1112,"_246":1108,"_271":1113,"_1110":1093},{"_20":1100,"_243":1106,"_204":205,"_206":-7,"_238":1107,"_246":1108,"_271":1109,"_1110":1094},"NVIDIA NIM Operator","k8s-nim-operator","NVIDIA",[],"url","GPU Operator","gpu-operator",[],[1116],9,{"_20":21,"_243":1117,"_204":205,"_206":1118,"_238":1119,"_246":1108,"_271":1120,"_1110":1092},"Riva Skills Quick Start","riva","riva_quickstart",[],"models",[1124,1125,1126],6,{"_20":1121,"_243":1134,"_204":205,"_206":1135,"_238":1136,"_246":1108,"_271":1137,"_1110":1089},{"_20":1121,"_243":1131,"_204":205,"_206":1128,"_238":1132,"_246":1108,"_271":1133,"_1110":1090},{"_20":1121,"_243":1127,"_204":205,"_206":1128,"_238":1129,"_246":1108,"_271":1130,"_1110":1091},"TrafficCamNet","tao","trafficcamnet",[],"PeopleNet","peoplenet",[],"StyleGAN3 pretrained models","research","stylegan3",[],"collections",[1140,1141,1142],{"_20":1138,"_243":1170,"_204":205,"_206":1144,"_238":1171,"_246":1108,"_271":1172,"_1110":1086},{"_20":1138,"_243":1167,"_204":205,"_206":-7,"_238":1168,"_246":1108,"_271":1169,"_1110":1087},{"_20":1138,"_243":1143,"_204":205,"_206":1144,"_238":1145,"_246":1108,"_271":1146,"_1110":1088},"Omniverse Kit App Streaming","omniverse","kit-appstreaming-collection",[1147,1148,1149],{"_112":1161,"_243":1162,"_240":1163,"_250":1164,"_1154":1165},{"_112":1144,"_243":1157,"_240":1158,"_250":1153,"_1154":1159},{"_112":1150,"_243":1151,"_240":1152,"_250":1153,"_1154":1155},"omniverse-dev","NVIDIA Developer Program","Join the Developer Program for access to free tools, support, and tech resources.","https://store-images.s-microsoft.com/image/apps.33936.6f1bcf42-55d0-45d1-b4b6-f08c68a50bbe.e91ca707-db64-49d6-9761-4781590a9685.7ba64fc5-7c44-4864-a5e6-7bde05126935","supportedFeatures",[1156],"UNIFIED_CATALOG","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.",[1156,1160],"SHAREABLE_TO_USER_GROUPS","nv-ai-enterprise","NVIDIA AI Enterprise","Accelerate your AI agent development","https://assets.ngc.nvidia.com/products/catalog/images/product-NVAIE.jpg",[1156,1160,1166],"NGC_CONSOLE","DeepStream SDK","deepstream_sdk",[],"Omniverse Kit (FB)","kit",[1147,1148,1149],"containers",[1175,1176,1177],{"_20":1173,"_243":1189,"_204":1182,"_206":1190,"_238":1191,"_246":1108,"_271":1192,"_1110":1083},{"_20":1173,"_243":1181,"_204":1182,"_206":205,"_238":1183,"_246":1108,"_271":1184,"_1110":1084},{"_20":1173,"_243":1178,"_204":205,"_206":-7,"_238":1179,"_246":1108,"_271":1180,"_1110":1085},"PyTorch","pytorch",[],"Llama-3.1-Nemotron-70B-Instruct","nim","llama-3.1-nemotron-70b-instruct",[1185,1147],{"_112":1186,"_243":1151,"_240":1152,"_250":1187,"_1154":1188},"nim-dev","https://assets.ngc.nvidia.com/products/catalog/images/product-nim-dev.jpg",[1156],"DeepSeek-R1","deepseek-ai","deepseek-r1",[1185,1147],"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",{"_1222":1223,"_1224":315,"_1225":34,"_1226":1227,"_1101":1228,"_1229":1230,"_1231":1223,"_1232":34,"_1233":34,"_1234":34,"_1235":1236},"appName","ngc-catalog","appOrigin","hideOnPrivateOrg","icon","ObjectsModules2",[1237,1238,1239,1240,1241,1242,1243],"menuIcon","editor-swatches","productName","requiresAllScopes","requiresLogin","requiresOrgLevelScopesOnly","title","Catalog",{"_1235":1265,"_50":1266,"_1229":1267},{"_1258":1259,"_1235":1260,"_50":1261,"_1231":1161},{"_1235":1256,"_50":1257,"_1229":1230},{"_1235":1253,"_50":1254,"_1229":1255},{"_1235":1250,"_50":1251,"_1229":1252},{"_1235":1247,"_50":1248,"_1229":1249},{"_1235":1244,"_50":1245,"_1229":1246},"Resources","/resources","misc-library","Models","/models","misc-neural-network","Helm Charts","/helm-charts","misc-ship-wheel","Containers","/containers","3d-cube","Collections","/collections","access",[1262],"Enterprise Catalog","/enterprise",[1263,1264],"NVAIE","READ","Explore Catalog","/","common-home",{"_156":157,"_158":159,"_160":161,"_162":163,"_164":165,"_166":167},{"_155":34}]