hku-systems / vpipe
☆26Updated last year
Related projects ⓘ
Alternatives and complementary repositories for vpipe
- An Efficient Pipelined Data Parallel Approach for Training Large Model☆70Updated 3 years ago
- ☆72Updated last year
- Chimera: Efficiently Training Large-Scale Neural Networks with Bidirectional Pipelines.☆44Updated 11 months ago
- ☆65Updated 3 years ago
- REEF is a GPU-accelerated DNN inference serving system that enables instant kernel preemption and biased concurrent execution in GPU sche…☆85Updated last year
- AlpaServe: Statistical Multiplexing with Model Parallelism for Deep Learning Serving (OSDI 23)☆79Updated last year
- Artifacts for our ASPLOS'23 paper ElasticFlow☆52Updated 6 months ago
- A ChatGPT(GPT-3.5) & GPT-4 Workload Trace to Optimize LLM Serving Systems☆126Updated 3 weeks ago
- ☆41Updated last year
- Boost hardware utilization for ML training workloads via Inter-model Horizontal Fusion☆32Updated 5 months ago
- Compiler for Dynamic Neural Networks☆43Updated 11 months ago
- LLM serving cluster simulator☆74Updated 6 months ago
- ☆37Updated 3 years ago
- An experimental parallel training platform☆52Updated 7 months ago
- nnScaler: Compiling DNN models for Parallel Training☆62Updated 2 weeks ago
- Synthesizer for optimal collective communication algorithms☆98Updated 7 months ago
- PipeSwitch: Fast Pipelined Context Switching for Deep Learning Applications☆124Updated 2 years ago
- ☆51Updated 3 years ago
- ☆38Updated 4 years ago
- ☆47Updated last year
- ☆89Updated 2 years ago
- DISB is a new DNN inference serving benchmark with diverse workloads and models, as well as real-world traces.☆54Updated 2 months ago
- A baseline repository of Auto-Parallelism in Training Neural Networks☆142Updated 2 years ago
- ☆41Updated 6 months ago
- ☆9Updated 2 years ago
- An interference-aware scheduler for fine-grained GPU sharing☆108Updated 5 months ago
- Curated collection of papers in machine learning systems☆156Updated last month
- Artifacts for our SIGCOMM'22 paper Muri☆40Updated 10 months ago
- Bamboo is a system for running large pipeline-parallel DNNs affordably, reliably, and efficiently using spot instances.☆46Updated last year
- Artifact for PPoPP22 QGTC: Accelerating Quantized GNN via GPU Tensor Core.☆27Updated 2 years ago