SamsungLabs / FastFlow
FastFlow is a system that automatically detects CPU bottlenecks in deep learning training pipelines and resolves the bottlenecks with data pipeline offloading to remote resources .
☆26Updated last year
Alternatives and similar repositories for FastFlow:
Users that are interested in FastFlow are comparing it to the libraries listed below
- [ACM EuroSys '23] Fast and Efficient Model Serving Using Multi-GPUs with Direct-Host-Access☆57Updated last year
- ☆24Updated last year
- ☆49Updated 2 years ago
- ☆47Updated 3 months ago
- Hi-Speed DNN Training with Espresso: Unleashing the Full Potential of Gradient Compression with Near-Optimal Usage Strategies (EuroSys '2…☆15Updated last year
- ☆23Updated 2 years ago
- Proteus: A High-Throughput Inference-Serving System with Accuracy Scaling☆10Updated last year
- Artifacts for our ASPLOS'23 paper ElasticFlow☆52Updated 10 months ago
- ☆23Updated last year
- Multi-Instance-GPU profiling tool☆57Updated last year
- ☆53Updated 4 years ago
- ☆35Updated 4 years ago
- ☆37Updated 3 years ago
- DISB is a new DNN inference serving benchmark with diverse workloads and models, as well as real-world traces.☆53Updated 7 months ago
- Artifacts for our SIGCOMM'22 paper Muri☆41Updated last year
- ☆40Updated 8 months ago
- SHADE: Enable Fundamental Cacheability for Distributed Deep Learning Training☆31Updated 2 years ago
- A Cluster-Wide Model Manager to Accelerate DNN Training via Automated Training Warmup☆34Updated 2 years ago
- An interference-aware scheduler for fine-grained GPU sharing☆129Updated 2 months ago
- Repository for MLCommons Chakra schema and tools☆39Updated last year
- Ok-Topk is a scheme for distributed training with sparse gradients. Ok-Topk integrates a novel sparse allreduce algorithm (less than 6k c…☆25Updated 2 years ago
- A GPU-accelerated DNN inference serving system that supports instant kernel preemption and biased concurrent execution in GPU scheduling.☆42Updated 2 years ago
- Dynamic resources changes for multi-dimensional parallelism training☆24Updated 4 months ago
- Bamboo is a system for running large pipeline-parallel DNNs affordably, reliably, and efficiently using spot instances.☆49Updated 2 years ago
- Compiler for Dynamic Neural Networks☆45Updated last year
- AlpaServe: Statistical Multiplexing with Model Parallelism for Deep Learning Serving (OSDI 23)☆82Updated last year
- A ChatGPT(GPT-3.5) & GPT-4 Workload Trace to Optimize LLM Serving Systems☆156Updated 5 months ago
- SOTA Learning-augmented Systems☆35Updated 2 years ago
- Helios Traces from SenseTime☆53Updated 2 years ago
- Lucid: A Non-Intrusive, Scalable and Interpretable Scheduler for Deep Learning Training Jobs☆53Updated last year