facebookincubator / AITemplate
AITemplate is a Python framework which renders neural network into high performance CUDA/HIP C++ code. Specialized for FP16 TensorCore (NVIDIA GPU) and MatrixCore (AMD GPU) inference.
☆4,531Updated this week
Related projects: ⓘ
- Transformer related optimization, including BERT, GPT☆5,773Updated 5 months ago
- Accessible large language models via k-bit quantization for PyTorch.☆6,029Updated this week
- A machine learning compiler for GPUs, CPUs, and ML accelerators☆2,577Updated 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…☆1,519Updated 7 months ago
- Foundation Architecture for (M)LLMs☆3,003Updated 5 months ago
- 🚀 Accelerate training and inference of 🤗 Transformers and 🤗 Diffusers with easy to use hardware optimization tools☆2,459Updated this week
- Hackable and optimized Transformers building blocks, supporting a composable construction.☆8,351Updated this week
- Fast and memory-efficient exact attention☆13,401Updated this week
- 🚀 A simple way to launch, train, and use PyTorch models on almost any device and distributed configuration, automatic mixed precision (i…☆7,687Updated this week
- Training and serving large-scale neural networks with auto parallelization.☆3,045Updated 9 months ago
- Supercharge Your Model Training☆5,116Updated this week
- Development repository for the Triton language and compiler☆12,698Updated this week
- PyTorch extensions for high performance and large scale training.☆3,149Updated 2 weeks ago
- A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper and Ada GPUs…☆1,811Updated this week
- 4 bits quantization of LLaMA using GPTQ☆2,979Updated 2 months ago
- Ongoing research training transformer models at scale☆9,949Updated this week
- Simple, safe way to store and distribute tensors☆2,755Updated 2 weeks ago
- A repo for distributed training of language models with Reinforcement Learning via Human Feedback (RLHF)☆4,442Updated 8 months ago
- Running large language models on a single GPU for throughput-oriented scenarios.☆9,133Updated last month
- [ICLR 2024] Fine-tuning LLaMA to follow Instructions within 1 Hour and 1.2M Parameters☆5,691Updated 6 months ago
- MII makes low-latency and high-throughput inference possible, powered by DeepSpeed.☆1,843Updated last week
- Train transformer language models with reinforcement learning.☆9,288Updated this week
- Implementation of the LLaMA language model based on nanoGPT. Supports flash attention, Int8 and GPTQ 4bit quantization, LoRA and LLaMA-Ad…☆5,958Updated last week
- An open-source framework for training large multimodal models.☆3,658Updated 2 weeks ago
- Code for the ICLR 2023 paper "GPTQ: Accurate Post-training Quantization of Generative Pretrained Transformers".☆1,875Updated 5 months ago
- ☆2,635Updated last week
- SOTA low-bit LLM quantization (INT8/FP8/INT4/FP4/NF4) & sparsity; leading model compression techniques on TensorFlow, PyTorch, and ONNX R…☆2,146Updated this week
- [ICLR 2024] Efficient Streaming Language Models with Attention Sinks☆6,537Updated 2 months ago
- QLoRA: Efficient Finetuning of Quantized LLMs☆9,906Updated 3 months ago
- [MLSys 2024 Best Paper Award] AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration☆2,333Updated 2 months ago