End-to-end recipes for optimizing diffusion models with torchao and diffusers (inference and FP8 training).
-
Updated
Jan 8, 2026 - Python
#
torchao
Here are 11 public repositories matching this topic...
A survey of modern quantization formats (e.g., MXFP8, NVFP4) and inference optimization tools (e.g., TorchAO, GemLite), illustrated through the example of Llama-3.1 inference.
-
Updated
Nov 11, 2025 - Python
FLUX.1-dev on AMD Radeon consumer GPUs — fast, low-VRAM, and shippable. Backport patches + benchmarks for torchao + diffusers group_offload on ROCm.
-
Updated
Apr 19, 2026 - Python
Flux.2-Klein-Small-Decoder-Only is an experimental, high-performance image generation and editing application built to exclusively utilize the FLUX.2-klein-4B model paired with the specialized FLUX.2-small-decoder Variational Autoencoder (VAE).
python transformers torch full-stack autoencoder vae gradio torchvision fastapi huggingface diffusion-models diffusers torchao flux-2-klein small-decoder
-
Updated
May 18, 2026 - Python
Deploy AI models with an API through quantization and containerization.
flux ai cog pre-commit torch pytest quantization replicate peft huggingface diffusers bitsandbytes unsloth torchao smollm3 pruna
-
Updated
Jan 17, 2026 - Python
Run FLUX.1-dev on AMD Radeon GPUs using ROCm with backport patches, optimized scripts, and support for low-VRAM configurations.
react docker flux amd self-hosted pytorch music-generation text-to-image rocm int8 ai-music fastapi diffusers generative-ai ollama rdna3 torchao ace-step group-offload
-
Updated
Jun 7, 2026 - Python
Block-scaled FP8 / FP4 / INT4 tensor primitive with Triton scaled-matmul at FP32 parity on H100. NumPy / PyTorch / MLX / JAX backends.
machine-learning deep-learning numpy pytorch triton quantization mlx jax low-precision int4 fp8 fp4 torchao nvfp4 transformer-engine mxfp8
-
Updated
May 24, 2026 - Python
Identity-preserving image-to-video generation: vision-grounded prompt simplification via Qwen3-VL, Lightning LoRA 4-step inference, and SAM3-masked DINOv3 candidate reranking for fluid 720p video from a single reference image.
computer-vision sam pytorch video-generation image-to-video motion-synthesis diffusion-models text-to-video huggingface-transformers prompt-engineering generative-ai dinov2 video-diffusion identity-preservation torchao lora-fine-tuning wan2 cvpr2026
-
Updated
Apr 6, 2026 - Python
Measuring what makes a VLA fast enough to run on the robot: a 5.9x CUDA-graph win, four experiments on why low-bit doesn't, a budget-driven deploy-compiler, and a runtime safety supervisor. Live demo: hf.co/spaces/LaelaZ/embodied-efficiency
-
Updated
Jun 7, 2026 - Python
This repository contains code for benchmarking ModernBERT, RoBERTa, and OPT-350m on multi-class emotion classification using 8-bit quantization, backbone freezing, and LoRA-based PEFT.
scikit-learn python3 pytorch scipy lora quantization emotion-analysis plotly-express huggingface-transformers int8-quantization supervised-finetuning llm-inference llm-fine-tuning roberta-base peft-fine-tuning-llm torchao modernbert facebook-opt cochrans-q-test
-
Updated
Jun 1, 2026 - Python
Improve this page
Add a description, image, and links to the torchao topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the torchao topic, visit your repo's landing page and select "manage topics."