LLaDA2.0 Uni: Unifying Multimodal Understanding and Generation with Diffusion Large Language Model FP8 Quantized Version of LLaDA2.0 Uni [📑 Technical Report]   [🌐 Github] AGI Research Center, Inclusion AI Overview This is the FP8 quantized version of LLaDA2.0 Uni, featuring block wise FP8 quantization of MoE expert weights. This reduces GPU memory usage by ~48% for model loading while preserving output quality. Quantization Details Method : Block wise FP8 (float8 e4m3fn) with per block scale factors Block size : 128×128 Quantized layers : MoE routed expert weights (gate proj, up proj, down proj) Kept in BF16 : Embeddings, lm head, attention projections, shared experts, layer norms, routing gates Memory Comparison Variant Model Loading T2I Peak Understanding Peak Edit Peak BF16 62.9 GB 35.3 GB 33.2 GB 41.7 GB FP8 32.5 GB 35.3 GB 33.3 GB 41.8 GB Note : FP8 halves the static model weight memory (~30 GB saved at load time). Peak inference memory is similar because activations dominate during generation. Quick Start Model Capabilities Same as the base LLaDA2.0 Uni model: 🖼️ Text to Image Generation 🔍 Image Understanding ✏️ Image Editing ⚡ Sprint Acceleration ⚠️ License This pro…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy