Reverse Dependencies of flash-attn

The following projects have a declared dependency on flash-attn:

• achatbot — An open source chat bot for voice (and multimodal) assistants
• airoboros — Updated and improved implementation of the self-instruct system.
• alpaca-farm — no summary
• archai — Platform for Neural Architecture Search
• audio2chat — Generate chat data from multi-speaker audio files
• autoawq — AutoAWQ implements the AWQ algorithm for 4-bit quantization with a 2x speedup during inference.
• autodistill-florence-2 — Use Florence 2 to auto-label data for use in training fine-tuned object detection models.
• axolotl — LLM Trainer
• bisheng-pybackend-libs — libraries for bisheng rt pybackend
• borzoi-pytorch — The Borzoi model from Linder et al., but in Pytorch
• cehrgpt — CEHR-GPT: Generating Electronic Health Records with Chronological Patient Timelines
• cornstarch — A multimodal model training toolkit
• ctopt — Tool used for determining optimal number of cell type clusters in spatial transcriptomics samples
• enhance-long — The tool integrates the lowest cost and highest performance methods to enhance the ability of large languages to process long contexts
• epiagent — Foundation model for single-cell epigenomic data.
• esm-efficient — Efficient Evolutionary Scale Modeling: Efficient and simplified implementation of protein language model for inference and training.
• eyefocus — Stay focused!
• faesm — no summary
• fastckpt — A fast gradient checkpointing strategy for training with memory-efficient attention (e.g., FlashAttention).
• fms-acceleration-peft — FMS Acceleration for PeFT
• fms-hf-tuning — FMS HF Tuning
• formless — Handwritten + image OCR.
• fschat — An open platform for training, serving, and evaluating large language model based chatbots.
• GeneSplice — GeneSplice Model, Ultra-Long Rage Genomic Expression Modelling
• geniusrise-audio — audio bolts for geniusrise
• geniusrise-text — Text bolts for geniusrise
• geniusrise-vision — Huggingface bolts for geniusrise
• genomeocean — A Python library for GenomeOcean inference and fine-tuning.
• grounded-ai — A Python package for evaluating LLM application outputs.
• hip-attn — HiP Attention
• Hippogriff — no summary
• instructlab-training — Training Library
• lazyllm — A Low-code Development Tool For Building Multi-agent LLMs Applications.
• lazyllm-beta — A Low-code Development Tool For Building Multi-agent LLMs Applications.
• lbster — Language models for Biological Sequence Transformation and Evolutionary Representation.
• llm-engines — A unified inference engine for large language models (LLMs) including open-source models (VLLM, SGLang, Together) and commercial models (OpenAI, Mistral, Claude).
• llm-foundry — LLM Foundry
• llm-optimized-inference — no summary
• lmetric — Large Model Metrics
• lmwrapper — An object-oriented wrapper around language models with caching, batching, and more.
• maestro — Streamline the fine-tuning process for vision-language models like PaliGemma 2, Florence-2, and Qwen2.5-VL.
• modalities — Modalities, a PyTorch-native framework for distributed and reproducible foundation model training.
• mqa — Multi Query Attention package
• mteb — Massive Text Embedding Benchmark
• nanotron — Minimalistic Large Language Model Training and Finetuning
• nextai-star — An open platform for training, serving, and evaluating large language model based chatbots by next ai
• oat-llm — Online AlignmenT (OAT) for LLMs.
• ochat — An efficient framework for training and serving top-tier, open-source conversational LLMs.
• olive-ai — Olive: Simplify ML Model Finetuning, Conversion, Quantization, and Optimization for CPUs, GPUs and NPUs.
• openrlhf — A Ray-based High-performance RLHF framework.
• optimum-benchmark — Optimum-Benchmark is a unified multi-backend utility for benchmarking Transformers, Timm, Diffusers and Sentence-Transformers with full support of Optimum's hardware optimizations & quantization schemes.
• rankify — A Comprehensive Python Toolkit for Retrieval, Re-Ranking, and Retrieval-Augmented Generation
• rerankers — A unified API for various document re-ranking models.
• salt-ml — Multimodal and Multiclass Machine Learning for High Energy Physics
• seb — Scandinavian Embedding Benchmark
• slideflow-noncommercial — Non-commercial extensions and tools for Slideflow.
• spchat — Chat utility library by Spiral.AI
• speechless — LLM based agents with proactive interactions, long-term memory, external tool integration, and local deployment capabilities.
• speechtoolkit — ML for Speech presents SpeechToolkit, a unified, all-in-one toolkit for TTS, ASR, VC, & other models.
• stripedhyena — Model and inference code for beyond Transformer architectures
• swebench — The official SWE-bench package - a benchmark for evaluating LMs on software engineering
• tracllm — A context tracing tool for LLM
• unsloth — 2-5X faster LLM finetuning
• v-clip-server — Embed images and sentences into fixed-length vectors via CLIP
• verl — verl: Volcano Engine Reinforcement Learning for LLM
• vllm-tgis-adapter — vLLM adapter for a TGIS-compatible grpc server
• xfuser — A Scalable Inference Engine for Diffusion Transformers (DiTs) on Multiple Computing Devices
• xllm — Simple & Cutting Edge LLM Finetuning
• yoflo — YO-FLO: A proof-of-concept in using advanced vision models as a YOLO alternative.
• yunchang — a package for long context attention
• zebra-qa — ZEBRA: Zero-Shot Example-Based Retrieval Augmentation for Commonsense Question Answering
• zonos — Text-to-speech