ParCIS / Ok-TopkLinks
Ok-Topk is a scheme for distributed training with sparse gradients. Ok-Topk integrates a novel sparse allreduce algorithm (less than 6k communication volume which is asymptotically optimal) with the decentralized parallel Stochastic Gradient Descent (SGD) optimizer, and its convergence is proved theoretically and empirically.
☆26Updated 2 years ago
Alternatives and similar repositories for Ok-Topk
Users that are interested in Ok-Topk are comparing it to the libraries listed below
Sorting:
- Hi-Speed DNN Training with Espresso: Unleashing the Full Potential of Gradient Compression with Near-Optimal Usage Strategies (EuroSys '2…☆15Updated last year
- ☆19Updated 3 years ago
- ☆14Updated 3 years ago
- A Cluster-Wide Model Manager to Accelerate DNN Training via Automated Training Warmup☆34Updated 2 years ago
- ☆25Updated last year
- gTop-k S-SGD: A Communication-Efficient Distributed Synchronous SGD Algorithm for Deep Learning☆36Updated 5 years ago
- Artifact for PPoPP22 QGTC: Accelerating Quantized GNN via GPU Tensor Core.☆28Updated 3 years ago
- Artifact for OSDI'23: MGG: Accelerating Graph Neural Networks with Fine-grained intra-kernel Communication-Computation Pipelining on Mult…☆40Updated last year
- ☆9Updated 2 years ago
- ☆23Updated 2 years ago
- [ASPLOS'23] Optimus-CC: Efficient Large NLP Model Training with 3D Parallelism Aware Communication Compression☆6Updated 9 months ago
- Compiler for Dynamic Neural Networks☆46Updated last year
- Cupcake: A Compression Scheduler for Scalable Communication-Efficient Distributed Training (MLSys '23)☆9Updated last year
- SHADE: Enable Fundamental Cacheability for Distributed Deep Learning Training☆33Updated 2 years ago
- ☆79Updated 2 years ago
- ☆50Updated 2 years ago
- Sharing the codebase and steps for artifact evaluation/reproduction for MICRO 2024 paper☆9Updated 9 months ago
- ☆8Updated 3 years ago
- Artifacts for our ASPLOS'23 paper ElasticFlow☆51Updated last year
- ☆26Updated 4 years ago
- We present a set of all-reduce compatible gradient compression algorithms which significantly reduce the communication overhead while mai…☆10Updated 3 years ago
- Artifacts for our SIGCOMM'22 paper Muri☆41Updated last year
- ☆30Updated last year
- Open-source implementation for "Helix: Serving Large Language Models over Heterogeneous GPUs and Network via Max-Flow"☆46Updated 6 months ago
- This serves as a repository for reproducibility of the SC21 paper "In-Depth Analyses of Unified Virtual Memory System for GPU Accelerated…☆32Updated last year
- MAGIS: Memory Optimization via Coordinated Graph Transformation and Scheduling for DNN (ASPLOS'24)☆51Updated last year
- Proteus: A High-Throughput Inference-Serving System with Accuracy Scaling☆12Updated last year
- ☆10Updated 4 years ago
- Source code for the paper: "A Latency-Predictable Multi-Dimensional Optimization Framework forDNN-driven Autonomous Systems"☆22Updated 4 years ago
- ☆53Updated 4 years ago