Flash Attention v2 for Pascal (P100, GTX 1080) and Volta (V100) GPUs — bringing O(N) memory attention to hardware the official flash-attn doesn't support.
☆30Nov 3, 2026Updated last week
Alternatives and similar repositories for flash-attention-legacy
Users that are interested in flash-attention-legacy are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- BROKEN REPO. DO NOT USE UNDER ANY CIRCUMSTANCES☆20Oct 13, 2026Updated 3 weeks ago
- 限制问题区域显存 62906MiB / 65024MiB 支持驱动 610.57.04 ，修改显卡显示名称为NVIDIA Tesla A100☆12Aug 30, 2026Updated 2 months ago
- Sol-Attn (arXiv 2607.24027) sparse acceleration for MiniMax H3 on V100 — single-node ComfyUI plugin (embedded FP16Safe + keep-or-drop ker…☆33Aug 29, 2026Updated 2 months ago
- A standalone, modular prompt generator for ComfyUI. Generate detailed, natural language prompts for characters and spaceships with extens…☆14Aug 3, 2026Updated 3 months ago
- Implementation of FlashAttention-2 for Nvidia Tesla V100 / Titan V☆218Nov 5, 2026Updated last week
- Managed hosting for WordPress and PHP on Cloudways • AdManaged hosting for WordPress, Magento, Laravel, or PHP apps, on multiple cloud providers. Deploy in minutes on Cloudways by DigitalOcean.
- ☆37Nov 4, 2026Updated last week
- dsv41-flash-pp5-170hx &pp6 Supports the original (full/unquantized) model. Prefill 7,000 t/s, decode 45-75-125 t/s. dspark. Requires 200 …☆13Nov 2, 2026Updated last week
- Hindsight Memory for Open WebUI☆20May 27, 2026Updated 5 months ago
- DeepSeek-V4.1-Flash (764B) on 4×CMP 170HX: PP4 + shadow KV + EXL3 2bpw + DSpark + 512k ctx + native vision — patches, recipes, acceptance…☆15Sep 29, 2026Updated last month
- A LD_PRELOAD library to share CUDA VRAM weights across multiple processes using IPC.☆12Jun 30, 2026Updated 4 months ago
- Implementation of "FlashPreill: Instantaneous Pattern Discovery and Thresholding for Ultra-Fast Long-Context Prefilling"☆57Apr 27, 2026Updated 6 months ago
- Qwen3.8-Flash-Next (W4A16) on 2x CMP 170HX: 110 tok/s decode, 7K prefill, 800 tok/s @ 23 concurrent - runs on 38 GB home RAM, no pinned m…☆20Sep 18, 2026Updated last month
- Demonstration of MoE distributed training using various techniques☆11Oct 31, 2024Updated 2 years ago
- VGGT implemented in C++☆14Mar 31, 2026Updated 7 months ago
- Virtual machines for every use case on DigitalOcean • AdGet dependable uptime with 99.99% SLA, simple security tools, and predictable monthly pricing with DigitalOcean's virtual machines, called Droplets.
- This repository contains a Automatic1111 Extension allows users to select and apply different styles to their inputs using SDXL 1.0.☆14Aug 17, 2023Updated 3 years ago
- Works in Progress and Experiments for the Innova-2 Flex XCKU15P-based Board☆18Apr 4, 2024Updated 2 years ago
- 中文汉化☆16Dec 25, 2023Updated 2 years ago
- Tiny vi-like text editor in only 100 lines of C++ code!☆14Aug 12, 2022Updated 4 years ago
- A proxy that hosts multiple single-model runners such as LLama.cpp and vLLM☆12May 30, 2025Updated last year
- MiniMax Music 3 的中文曲风预设与作词工具节点包 · 选曲风即出歌，无需手写英文 Caption☆18Aug 22, 2026Updated 2 months ago
- Multi-GPU prefill acceleration for llama.cpp☆16Oct 23, 2026Updated 2 weeks ago
- FlashPrefill V2: Block-Sparse Prefill Attention for Long-Context LLM Serving☆33Aug 28, 2026Updated 2 months ago
- ☆12Feb 19, 2025Updated last year
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- SGLang for cool dudes using V100☆85Updated this week
- A powerful OCR node for ComfyUI that integrates the DeepSeek-OCR model from Hugging Face.☆17Oct 21, 2025Updated last year
- 神棍☆16May 1, 2026Updated 6 months ago
- High-performance single-GPU inference for selected model checkpoints and GPUs.☆264Sep 14, 2026Updated last month
- A set of utilities to turn Dataclasses into useful configuration managers.☆11Mar 27, 2024Updated 2 years ago
- This project adds high-quality Text-to-Speech capabilities to ComfyUI using the Orpheus TTS model. Create natural-sounding voices with em…☆12May 3, 2025Updated last year
- Your personal AI team that gets things done. Shared memory. Any model. Local or cloud. Run. Rest. Dream. 🐨☆35Updated this week
- Speed by Simplicity: A Single-Stream Architecture for Fast Audio-Video Generative Foundation Model☆10Apr 2, 2026Updated 7 months ago
- Testing a cheap and fast AliExpress 1.9" IPS Display with ESP32 Wemos lite and TFT_eSPI by Bodmer☆11Jun 28, 2024Updated 2 years ago
- Managed hosting for WordPress and PHP on Cloudways • AdManaged hosting for WordPress, Magento, Laravel, or PHP apps, on multiple cloud providers. Deploy in minutes on Cloudways by DigitalOcean.
- Dockerized environment to ease odoo development☆18Nov 5, 2026Updated last week
- GuardianBridge keeps messaging and coordination working when cellular or internet is unavailable. By combining a Meshtastic mesh network …☆25Feb 16, 2026Updated 8 months ago
- Flash-attention-class memory efficiency for GPUs without flash attention☆12Mar 28, 2026Updated 7 months ago
- ☆19Oct 29, 2026Updated 2 weeks ago
- vLLM fork for Tesla V100☆15May 31, 2026Updated 5 months ago
- Recover usable memory on DGX Spark with a 64 KiB kernel: an opt-in NVIDIA UVM page-table packing patch, build tools, and measured validat…☆39Oct 16, 2026Updated 3 weeks ago
- Custom HIP inference engine for AMD MI50/MI60 (gfx906)☆19Oct 4, 2026Updated last month
Are these results useful?