Using the famous cnn model in Pytorch, we run benchmarks on various gpu.

Remove Backgroud from a video 🧤🎥

A kubernetes operator for managing nvidia MIG instances.

Z-AI V1.4 发行版本

ClusterOps is an enterprise-grade Python library developed and maintained by the Swarms Team to help you manage and execute agents on specific CPUs and GPUs across clusters. This tool enables advanced CPU and GPU selection, dynamic task allocation, and resource monitoring, making it ideal for high-performance distributed computing environments.

Dataset and code for "Coarse-Grained Density Functional Theory Predictions via Deep Kernel Learning"

Implement named entity recognition (NER) using regex and fine-tuned LLM, with a total of 15 categories. The ultimate goal is to apply the model to detect personally identifiable information (PII) in student writing.

Implementing science-related multiple-choice question answering based on LLMs and RAG.

Found out that using A100 and V100 on Vicuna and Llama2 have a different result, while other model such as Falcon doesn't has such question.

Text to Image Gen [ Demo ]

CLI tool to check Oracle Cloud compute shape availability - find GPU capacity (A100, H100) across regions

This competition challenges you to predict which responses users will prefer in a head-to-head battle between chatbots powered by large language models (LLMs).

Analyze a dataset of conversations from the Chatbot Arena, where various LLMs provide responses to user prompts. The goal is to develop a model that enhances chatbot interactions, ensuring they align more closely with human preferences.

Transformer implementation in pytorch trained on NVIDIA A100 in fp16

Performance of CUDA example benchmark code on NVIDIA A100.

Patent-pending K8s DaemonSet that validates GPU nodes via GEMM latency, fail-slow CV detection, and NVLink ring profiling before Slurm resumes them

Prolepsis is a speculative decoding implementation that accelerates LLM inference by 1.30x on an A100. By pairing a small draft model (Qwen 1.7B) with a larger target (Qwen 8B), it shifts generation workloads into a parallel verification pass. A rigorous rejection sampling pipeline guarantees the output distribution is preserved.

An implementation of Speculative RAG exploring latency-quality trade-offs in multi-draft retrieval. Features batched parallel drafting via vLLM and log-probability verifier selection for fast, high-quality QA on a single A100 GPU.

GPU profiling suite & CUDA kernels on A100 80GB — ResNet-50 benchmarks, Nsight Systems profiling, tiled matrix multiplication with shared memory

This competition focuses on developing accurate motion forecasting models using the Argoverse 2 Motion Forecasting Dataset.