Reporium
GraphWikiTaxonomyStacksInsightsTrendsArchitectureAI-NativeFAQ
Ask anything about the repo library…
Loading repo…
←Library/DALI
Library/DALIForked

NVIDIA/DALI

DALI

A GPU-accelerated library containing highly optimized building blocks and an execution engine for data processing to accelerate deep learning training and inference applications.

View on GitHub↗Upstream NVIDIA/DALI↗

Builder

NVIDIA

NVIDIA

NVIDIA • big-tech

Stars

5,724

Using upstream star count

Forks

669

Using upstream fork count

Open Issues

0

Activity Score

0/100

0 commits in 30d

Created

Jun 1, 2018

Project creation date

README Summary

NVIDIA DALI =========== .. overview-begin-marker-do-not-remove

Community Evaluation

Loading…

AI Dev Skills

Unmapped

Computer Vision PreprocessingData Pipeline OptimizationDeep Learning InfrastructureGPU Computing and CUDA ProgrammingHigh-Performance ComputingMemory Management and OptimizationNeural Network Training AccelerationParallel Data Processing

Tags

Computer Vision PreprocessingData Pipeline OptimizationDeep Learning InfrastructureGPU Computing and CUDA ProgrammingHigh-Performance ComputingMemory Management and OptimizationNeural Network Training AccelerationParallel Data ProcessingBenchmarkingC++Computer VisionCurated ListData EngineeringData ScienceDeep LearningForkedGPU / CUDAInferenceJupyterLLM ServingMLOpsModel OptimizationOpen SourcePose EstimationPyTorchPythonRoadmapSystemsTensorFlowTensorRTTritonTutorial

Taxonomy

AI Trends

High-Performance ML InfrastructureGPU-Accelerated ComputingEfficient Neural Network Training

category

Inference & ServingModel TrainingEvals & BenchmarkingComputer VisionMLOps & InfrastructureDev Tools & AutomationLearning ResourcesData Science & Analytics

Deployment Context

Cloud GPU InstancesOn-premise GPU ClustersSelf-hosted

Modalities

ImageVideoAudio

Skill Areas

GPU Computing and CUDA ProgrammingData Pipeline OptimizationDeep Learning InfrastructureComputer Vision PreprocessingHigh-Performance ComputingMemory Management and OptimizationParallel Data ProcessingNeural Network Training Acceleration

tag

BenchmarkingC++Computer VisionCurated ListData EngineeringData ScienceDeep LearningForkedGPU / CUDAInferenceJupyterLLM ServingMLOpsModel OptimizationOpen SourcePose EstimationPyTorchPythonRoadmapSystemsTensorFlowTensorRTTritonTutorial

Use Cases

Image Data Augmentation for Computer VisionVideo Processing for Deep LearningAudio Preprocessing for Neural NetworksReal-time Data Pipeline AccelerationLarge-scale Training Data ProcessingInference Pipeline Optimization

Recent Activity

Updated 4 months ago

7 Days

0

30 Days

0

90 Days

0

Fix UB in JpegCompressionDistortion: use data() for past-the-end pointer (#6251)

Janusz Lisiecki • Mar 16, 2026

e1fdf02

New ThreadPool + thread pool facade (#6224)

Michał Zientkiewicz • Mar 13, 2026

1621253

Make result of AtScopeExit non-discardable. (#6248)

Michał Zientkiewicz • Mar 11, 2026

f343230

Quality

production
Quality
high
Maturity
production

Categories

Model TrainingPrimaryComputer VisionInference & ServingEvals & BenchmarkingMLOps & InfrastructureDev Tools & AutomationLearning ResourcesData Science & AnalyticsFoundation ModelsML Platform & InfrastructureOther AI / ML

PM Skills

Cost & EfficiencyScale & ReliabilityData & Evaluation

Languages

C++100.0%

Timeline

Project created
Jun 1, 2018
Forked
Mar 14, 2026
Your last push
4 months ago
Upstream last push
2 months ago
Tracked since
Mar 17, 2026

Similar Repos

pgvector cosine similarity · $0

Loading…