Search Results for "gpu speed"
Sort By:
25 projects for "gpu speed" with 2 filters applied:
-
Ship Agents FasterGemini Enterprise Agent Platform lets you rapidly build, scale, govern and optimize production-ready agents grounded in your organization's data. The platform enables developers to build custom or pre-built agents for virtually any use case. New customers get $300 in free credits.
-
Our Free Plans just got better! | Auth0You asked, we delivered! Auth0 is excited to expand our Free and Paid plans to include more options so you can focus on building, deploying, and scaling applications without having to worry about your security. Auth0 now, thank yourself later.
-
1
GPU Hot
Real-time NVIDIA GPU dashboard
...The dashboard collects and displays a wide range of performance metrics including temperature, memory usage, power consumption, clock speeds, fan speed, and active processes. It can scale from monitoring a single GPU workstation to large distributed environments with dozens or even hundreds of GPUs by running lightweight containers on each node and aggregating the data centrally. -
2
llmfit
157 models, 30 providers, one command to find what runs on hardware
llmfit is a terminal-based utility that helps developers determine which large language models can realistically run on their local hardware by analyzing system resources and model requirements. The tool automatically detects CPU, RAM, GPU, and VRAM specifications, then ranks available models based on performance factors such as speed, quality, and memory fit. It provides both an interactive terminal user interface and a traditional CLI mode, enabling flexible workflows for different user preferences. llmfit also supports advanced configurations including multi-GPU setups, mixture-of-experts architectures, and dynamic quantization recommendations. ... -
3
how-to-optim-algorithm-in-cuda
How to optimize some algorithm in cuda
...The repository also contains extensive learning notes that summarize CUDA programming concepts, GPU architecture details, and performance engineering strategies. -
4
LightLLM
LightLLM is a Python-based LLM (Large Language Model) inference
LightLLM is a high-performance inference and serving framework designed specifically for large language models, focusing on lightweight architecture, scalability, and efficient deployment. The framework enables developers to run and serve modern language models with significantly improved speed and resource efficiency compared to many traditional inference systems. Built primarily in Python, the project integrates optimization techniques and ideas from several leading open-source implementations, including FasterTransformer, vLLM, and FlashAttention, to accelerate token generation and reduce latency. LightLLM is designed to handle large-scale model workloads in production environments, supporting efficient batching and GPU utilization for fast inference across multiple requests. ... -
Stop Cyber Threats with VM-Series Next-Gen Firewall on AzureGain integrated visibility across all traffic in a single pass. Deploy Palo Alto Networks VM-Series to determine application identity and content while automating security policy updates via rich APIs.
-
5
LuxTTS
A high-quality rapid TTS voice cloning model
LuxTTS is an open-source text-to-speech (TTS) system focused on delivering high-quality, rapid voice synthesis and voice cloning that runs extremely fast and efficiently on consumer hardware. It implements a lightweight architecture based on ZipVoice and optimized sampling techniques so that it can generate speech at speeds up to roughly 150 times real-time on a single GPU and faster than real-time on CPU, all while producing audio at high fidelity with 48 kHz quality. The project supports... -
6
Flash-MoE
Running a big model on a small laptop
...It focuses on accelerating routing and computation by leveraging optimized kernels and memory management techniques, allowing models to dynamically select specialized sub-networks during inference. The project aims to reduce the computational cost typically associated with MoE systems while maintaining or improving performance. It likely includes support for GPU acceleration and parallel processing, enabling it to handle large-scale workloads effectively. The architecture emphasizes speed and efficiency, making it suitable for both research and production environments where performance is critical. It may also provide tools for benchmarking and tuning model behavior. Overall, flash-moe represents a technical advancement in making MoE models more practical and deployable. -
7
clip-retrieval
Easily compute clip embeddings and build a clip retrieval system
clip-retrieval is an open-source toolkit designed to build large-scale semantic search systems for images and text by leveraging CLIP embeddings to enable multimodal retrieval. It allows developers to compute embeddings for both images and text efficiently and then index them for fast similarity search across massive datasets. The system is optimized for performance and scalability, capable of processing tens or even hundreds of millions of embeddings using GPU acceleration. It includes... -
8
Depth Pro
Sharp Monocular Metric Depth in Less Than a Second
...Unlike many prior approaches, it does not require camera intrinsics or extra metadata, yet still outputs metric depth suitable for downstream 3D tasks. Apple highlights both accuracy and speed: the model can synthesize a ~2.25-megapixel depth map in around 0.3 seconds on a standard GPU, enabling near real-time applications. The repo and research page emphasize boundary fidelity and crisp geometry, addressing a common weakness in monocular depth where edges can blur. Community integrations (e.g., inference wrappers and UI nodes) have sprung up around the model, reflecting practical interest in video, AR, and generative pipelines. ... -
9
FlashMLA
FlashMLA: Efficient Multi-head Latent Attention Kernels
FlashMLA is a high-performance decoding kernel library designed especially for Multi-Head Latent Attention (MLA) workloads, targeting NVIDIA Hopper GPU architectures. It provides optimized kernels for MLA decoding, including support for variable-length sequences, helping reduce latency and increase throughput in model inference systems using that attention style. The library supports both BF16 and FP16 data types, and includes a paged KV cache implementation with a block size of 64 to... -
MongoDB Atlas runs apps anywhereMongoDB Atlas gives you the freedom to build and run modern applications anywhere—across AWS, Azure, and Google Cloud. With global availability in over 115 regions, Atlas lets you deploy close to your users, meet compliance needs, and scale with confidence across any geography.
-
10
CosyVoice
Multi-lingual large voice generation model, providing inference
CosyVoice is a multilingual large voice generation model that offers a full-stack solution for training, inference, and deployment of high-quality TTS systems. The model supports multiple languages, including Chinese, English, Japanese, Korean, and a range of Chinese dialects such as Cantonese, Sichuanese, Shanghainese, Tianjinese, and Wuhanese. It is designed for zero-shot voice cloning and cross-lingual or mix-lingual scenarios, so a single reference voice can be used to synthesize speech... -
11
ort
Fast ML inference & training for ONNX models in Rust
ort is a high-performance Rust library that provides bindings to ONNX Runtime, enabling developers to run machine learning inference and training workflows directly within Rust applications using the standardized ONNX model format. It is designed to bridge the gap between modern machine learning frameworks and systems programming by offering a safe, ergonomic API for executing models originally built in ecosystems like PyTorch, TensorFlow, or scikit-learn. The library emphasizes speed and... -
12
CUDA Agent
Large-Scale Agentic RL for High-Performance CUDA Kernel Generation
CUDA Agent is a research-driven agentic reinforcement learning system designed to automatically generate and optimize high-performance CUDA kernels for GPU workloads. The project addresses the long-standing challenge that efficient CUDA programming typically requires deep hardware expertise by training an autonomous coding agent capable of iterative improvement through execution feedback. Its architecture combines large-scale data synthesis, a skill-augmented CUDA development environment,... -
13
nanochat
The best ChatGPT that $100 can buy
...The repository stitches together every stage of the lifecycle: tokenizer training, pretraining a Transformer on a large web corpus, mid-training on dialogue and multiple-choice tasks, supervised fine-tuning, optional reinforcement learning for alignment, and finally efficient inference with caching. Its north star is approachability and speed: you can boot a fresh GPU box and drive the whole pipeline via a single script, producing a usable chat model in hours and a clear markdown report of what happened. The code is written to be read—concise training loops, transparent configs, and minimal wrappers—so you can audit each step, tweak it, and rerun without getting lost in framework indirection. -
14
DeepSpeed
Deep learning optimization library: makes distributed training easy
DeepSpeed is an easy-to-use deep learning optimization software suite that enables unprecedented scale and speed for Deep Learning Training and Inference. With DeepSpeed you can: 1. Train/Inference dense or sparse models with billions or trillions of parameters 2. Achieve excellent system throughput and efficiently scale to thousands of GPUs 3. Train/Inference on resource constrained GPU systems 4. Achieve unprecedented low latency and high throughput for inference 5. ... -
15
Mooncake
Mooncake is the serving platform for Kimi
...Its architecture centers on a high-performance transfer engine that provides unified data transfer across different storage and networking technologies. This engine enables efficient movement of tensors and model data across heterogeneous environments such as GPU memory, system memory, and distributed storage systems. Mooncake also introduces distributed key-value cache storage that allows inference systems to reuse previously computed attention states, significantly improving throughput in large-scale deployments. The system supports advanced networking technologies such as RDMA and NVMe over Fabric, enabling high-speed communication across clusters. -
16
gpu_poor
Calculate token/s & GPU memory requirement for any LLM
gpu_poor is an open-source tool designed to help developers determine whether their hardware is capable of running a specific large language model and to estimate the performance they can expect from it. The project focuses on calculating GPU memory requirements and predicted inference speed for different models, hardware configurations, and quantization strategies. By analyzing factors such as model size, context length, batch size, and GPU specifications, the system estimates how much VRAM will be required and how fast tokens can be generated during inference. The tool also provides a detailed breakdown of where GPU memory is allocated, including model weights, KV cache, activations, and other runtime overhead. ... -
17
OptiMate
Libraries for optimizing AI models, inference speed, and GPU usage
...Another component, Nos, targets infrastructure optimization by improving GPU utilization in Kubernetes clusters through dynamic partitioning and elastic resource quotas. -
18
Point-E
Point cloud diffusion for 3D model synthesis
point-e is the official repository for Point-E, a generative model developed by OpenAI that produces 3D point clouds from textual (or image) prompts. Its principal advantage is speed: it can generate 3D assets in just 1–2 minutes on a single GPU, which is significantly faster than many competing text-to-3D models. The model works via a two-stage diffusion approach: first, it uses a text → image diffusion network to produce a synthetic 2D view consistent with the prompt; then a second diffusion model converts that image into a 3D point cloud. ... -
19
YOLOv4-large
Scaled-YOLOv4: Scaling Cross Stage Partial Network
YOLOv4-large is an open-source implementation of the Scaled-YOLOv4 object detection architecture, designed to improve both the accuracy and scalability of real-time computer vision models. The project provides a PyTorch implementation of the Scaled-YOLOv4 framework, which extends the original YOLOv4 architecture using Cross Stage Partial (CSP) networks and new scaling techniques. Unlike earlier object detection systems that only scale depth or width, this architecture scales multiple aspects... -
20
HiFi-GAN
Generative Adversarial Networks for Efficient and High Fidelity Speech
...It introduces a generator architecture tailored to model the periodic structure of speech and a set of discriminators that focus on different scales and periods of the waveform to better capture naturalness. The model targets a sweet spot between sample quality and generation speed, outperforming many previous GAN vocoders while being far faster than typical autoregressive models. In experiments on LJSpeech, HiFi-GAN was shown to achieve mean opinion scores close to human recordings while synthesizing 22.05 kHz audio up to ~168× faster than real time on an NVIDIA V100 GPU. A smaller configuration trades a bit of quality for even higher speed and can run more than 13× faster than real time on CPU, making it suitable for deployment scenarios without powerful GPUs. -
21
maskrcnn-benchmark
Fast, modular reference implementation of Instance Segmentation
...The framework integrates critical components—region proposal networks (RPNs), RoIAlign layers, mask heads, and backbone architectures such as ResNet and FPN—optimized for both accuracy and speed. It supports multi-GPU distributed training, mixed precision, and custom data loaders for new datasets. Built as a reference implementation, it became a foundation for the next-generation Detectron2, yet remains widely used for research needing a stable, reproducible environment. Visualization tools, model zoo checkpoints, and benchmark scripts make it easy to replicate state-of-the-art results or fine-tune models for custom tasks. -
22
OpenSeq2Seq
Toolkit for efficient experimentation with Speech Recognition
OpenSeq2Seq is a TensorFlow-based toolkit for efficient experimentation with sequence-to-sequence models across speech and NLP tasks. Its core goal is to give researchers a flexible, modular framework for building and training encoder–decoder architectures while fully leveraging distributed and mixed-precision training. The toolkit includes ready-made models for neural machine translation, automatic speech recognition, speech synthesis, language modeling, and additional NLP tasks such as... -
23
gpt-oss-120b
OpenAI’s open-weight 120B model optimized for reasoning and tooling
GPT-OSS-120B is a powerful open-weight language model by OpenAI, optimized for high-level reasoning, tool use, and agentic tasks. With 117B total parameters and 5.1B active parameters, it’s designed to fit on a single H100 GPU using native MXFP4 quantization. The model supports fine-tuning, chain-of-thought reasoning, and structured outputs, making it ideal for complex workflows. It operates in OpenAI’s Harmony response format and can be deployed via Transformers, vLLM, Ollama, LM Studio, and PyTorch. Developers can control the reasoning level (low, medium, high) to balance speed and depth depending on the task. ... -
24
DiffusionGemma
NVFP4 DiffusionGemma model for fast multimodal text generation
...Built on the Gemma 4 26B A4B Mixture-of-Experts architecture, it has 25.2B total parameters and 3.8B active parameters, balancing capability with efficient inference. Its diffusion-based generation produces tokens in parallel 256-token blocks, enabling very high-speed output, with reported generation above 1,100 tokens per second on NVIDIA Hopper H100 in FP8. The model supports a 256K-token context window, configurable thinking mode, native function calling, structured JSON output, and multilingual inference across 35+ languages. The NVFP4 quantization reduces weights and activations from 16-bit to 4-bit, lowering disk size and GPU memory needs for vLLM deployment. -
25
granite-timeseries-ttm-r2
Tiny pre-trained IBM model for multivariate time series forecasting
granite-timeseries-ttm-r2 is part of IBM’s TinyTimeMixers (TTM) series—compact, pre-trained models for multivariate time series forecasting. Unlike massive foundation models, TTM models are designed to be lightweight yet powerful, with only ~805K parameters, enabling high performance even on CPU or single-GPU machines. The r2 version is pre-trained on ~700M samples (r2.1 expands to ~1B), delivering up to 15% better accuracy than the r1 version. TTM supports both zero-shot and fine-tuned...
- Previous
- You're on page 1
- Next
