SymbioticLab / ModelKeeper
A Cluster-Wide Model Manager to Accelerate DNN Training via Automated Training Warmup
☆34Updated last year
Related projects ⓘ
Alternatives and complementary repositories for ModelKeeper
- ☆23Updated last year
- Artifacts for our ASPLOS'23 paper ElasticFlow☆52Updated 6 months ago
- ☆48Updated last year
- A resilient distributed training framework☆85Updated 7 months ago
- ☆19Updated 2 years ago
- Artifacts for our SIGCOMM'22 paper Muri☆40Updated 10 months ago
- HeliosArtifact☆18Updated 2 years ago
- A Deep Learning Cluster Scheduler☆37Updated 3 years ago
- ☆29Updated 4 months ago
- Efficient Interactive LLM Serving with Proxy Model-based Sequence Length Prediction | A tiny model can tell you the verbosity of an LLM (…☆22Updated 5 months ago
- Surrogate-based Hyperparameter Tuning System☆27Updated last year
- A ChatGPT(GPT-3.5) & GPT-4 Workload Trace to Optimize LLM Serving Systems☆132Updated last month
- Official Repo for "LLM-PQ: Serving LLM on Heterogeneous Clusters with Phase-Aware Partition and Adaptive Quantization"☆27Updated 8 months ago
- DISB is a new DNN inference serving benchmark with diverse workloads and models, as well as real-world traces.☆54Updated 3 months ago
- Ok-Topk is a scheme for distributed training with sparse gradients. Ok-Topk integrates a novel sparse allreduce algorithm (less than 6k c…☆23Updated last year
- Lucid: A Non-Intrusive, Scalable and Interpretable Scheduler for Deep Learning Training Jobs☆50Updated last year
- ☆46Updated 5 months ago
- Code for "Apparate: Rethinking Early Exits to Tame Latency-Throughput Tensions in ML Serving" [SOSP '24]☆16Updated last month
- ☆14Updated 5 months ago
- ☆37Updated 3 years ago
- Stateful LLM Serving☆38Updated 3 months ago
- ☆12Updated last year
- Code for "Shockwave: Fair and Efficient Cluster Scheduling for Dynamic Adaptation in Machine Learning" [NSDI '23]☆38Updated last year
- AlpaServe: Statistical Multiplexing with Model Parallelism for Deep Learning Serving (OSDI 23)☆78Updated last year
- ☆18Updated 2 years ago
- ☆51Updated 3 years ago
- Bamboo is a system for running large pipeline-parallel DNNs affordably, reliably, and efficiently using spot instances.☆47Updated last year
- SpotServe: Serving Generative Large Language Models on Preemptible Instances☆101Updated 8 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
- Official resporitory for "IPDPS' 24 QSync: Quantization-Minimized Synchronous Distributed Training Across Hybrid Devices".☆19Updated 8 months ago