QUASAR QAT Models High-fidelity low-bit QAT checkpoints for Qwen, Gemma, and Muse, trained with QUASAR and shipped in deployment-ready formats. QUASAR-QAT/Qwen3.8-27B-QUASAR-NVFP4 Image-Text-to-Text • 28B • Updated about 10 hours ago • 33.9k • 119 QUASAR: Lowering the Loss Floor of Quantization-Aware Training with Loss-Aware Reconstruction Paper • 2608.13966 • Published 28 days ago • 4
QUASAR-QAT/Qwen3.8-27B-QUASAR-NVFP4 Image-Text-to-Text • 28B • Updated about 10 hours ago • 33.9k • 119
QUASAR: Lowering the Loss Floor of Quantization-Aware Training with Loss-Aware Reconstruction Paper • 2608.13966 • Published 28 days ago • 4
QUASAR QAT Models High-fidelity low-bit QAT checkpoints for Qwen, Gemma, and Muse, trained with QUASAR and shipped in deployment-ready formats. QUASAR-QAT/Qwen3.8-27B-QUASAR-NVFP4 Image-Text-to-Text • 28B • Updated about 10 hours ago • 33.9k • 119 QUASAR: Lowering the Loss Floor of Quantization-Aware Training with Loss-Aware Reconstruction Paper • 2608.13966 • Published 28 days ago • 4
QUASAR-QAT/Qwen3.8-27B-QUASAR-NVFP4 Image-Text-to-Text • 28B • Updated about 10 hours ago • 33.9k • 119
QUASAR: Lowering the Loss Floor of Quantization-Aware Training with Loss-Aware Reconstruction Paper • 2608.13966 • Published 28 days ago • 4
QUASAR-QAT/Qwen3.8-27B-QUASAR-NVFP4 Image-Text-to-Text • 28B • Updated about 10 hours ago • 33.9k • 119
