NVIDIA / TransformerEngine
A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper and Ada GPUs, to provide better performance with lower memory utilization in both training and inference.
☆1,936Updated this week
Related projects ⓘ
Alternatives and complementary repositories for TransformerEngine
- [ICML 2023] SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models☆1,242Updated 3 months ago
- FlashInfer: Kernel Library for LLM Serving☆1,395Updated this week
- Ongoing research training transformer language models at scale, including: BERT & GPT-2☆1,884Updated 3 weeks ago
- MII makes low-latency and high-throughput inference possible, powered by DeepSpeed.☆1,891Updated this week
- [MLSys 2024 Best Paper Award] AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration☆2,498Updated 3 weeks ago
- Code for the ICLR 2023 paper "GPTQ: Accurate Post-training Quantization of Generative Pretrained Transformers".☆1,922Updated 7 months ago
- Pipeline Parallelism for PyTorch☆725Updated 2 months ago
- FlexFlow Serve: Low-Latency, High-Performance LLM Serving☆1,698Updated this week
- Kernl lets you run PyTorch transformer models several times faster on GPU with a single line of code, and is designed to be easily hackab…☆1,530Updated 8 months ago
- Transformer related optimization, including BERT, GPT☆5,871Updated 7 months ago
- TensorRT Model Optimizer is a unified library of state-of-the-art model optimization techniques such as quantization, pruning, distillati…☆534Updated this week
- PyTorch extensions for high performance and large scale training.☆3,187Updated 2 months ago
- [ICML 2024] Break the Sequential Dependency of LLM Inference Using Lookahead Decoding☆1,141Updated 3 weeks ago
- Ongoing research training transformer language models at scale, including: BERT & GPT-2☆1,334Updated 7 months ago
- A pytorch quantization backend for optimum☆818Updated this week
- The Triton TensorRT-LLM Backend☆698Updated this week
- Microsoft Automatic Mixed Precision Library☆522Updated last month
- AutoAWQ implements the AWQ algorithm for 4-bit quantization with a 2x speedup during inference. Documentation:☆1,743Updated last month
- Tutel MoE: An Optimized Mixture-of-Experts Implementation☆728Updated last week
- Minimalistic large language model 3D-parallelism training☆1,220Updated this week
- FP16xINT4 LLM inference kernel that can achieve near-ideal ~4x speedups up to medium batchsizes of 16-32 tokens.☆611Updated 2 months ago
- Serving multiple LoRA finetuned LLM as one☆979Updated 6 months ago
- A Python-level JIT compiler designed to make unmodified PyTorch programs faster.☆1,010Updated 6 months ago
- 🚀 Accelerate training and inference of 🤗 Transformers and 🤗 Diffusers with easy to use hardware optimization tools☆2,556Updated this week
- Medusa: Simple Framework for Accelerating LLM Generation with Multiple Decoding Heads☆2,295Updated 4 months ago
- PyTorch native quantization and sparsity for training and inference☆1,541Updated this week
- Tile primitives for speedy kernels☆1,629Updated this week
- LightLLM is a Python-based LLM (Large Language Model) inference and serving framework, notable for its lightweight design, easy scalabili…☆2,580Updated this week
- Flash Attention in ~100 lines of CUDA (forward pass only)☆609Updated 7 months ago