Muse Glimmer Collection Muse Glimmer 30B: multimodal agentic model for local deployment. BF16 weights, GGUF k-quants, ExecuTorch builds, DFlash drafter. • 4 items • Updated Aug 10 • 131
view article Article Token Pooling-Aware Training for Multi-Vector Embedding Models tomaarsen • 11 days ago • 39
view article Article Decision Models in Hugging Face +2 ariG23498, lysandre, pcuenq, tomaarsen • 12 days ago • 109
view post Post 6,327 Qwen3.8-Flash-Next GGUFs now run faster with less thinking and less token use.Unsloth Dynamic V3.1 GGUFs bring thinking & token use to match 95% of full precision Qwen3.8-Flash.Results: 50% fewer tokens + 15% higher accuracy vs same size quants.GGUF: unsloth/Qwen3.8-Flash-Next-GGUFAnalysis/Guide: https://unsloth.ai/docs/models/qwen3.8-next See translation 3 replies · 🔥 11 11 ❤️ 4 4 🚀 4 4 + Reply
Running on CPU Upgrade 813 The Ultimate MoE Training Handbook 🧩 813 A hands-on guide to training Mixture-of-Experts models
HauhauCS/Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive Image-Text-to-Text • 35B • Updated Apr 17 • 1.47M • 4.22k
Muse Spark Open Collection Muse Spark Open model. A 816B parameter model with 42B Active parameters. BF16 and NVFP4 Weights. • 2 items • Updated Oct 15 • 177
view article Article Welcome Muse Spark Open: open-weight frontier multimodal mixture-of-experts model from Meta +6 pcuenq, merve, burtenshaw, guarin, qgallouedec, alvarobartt, Serveurperso, Jundot • Oct 15 • 74
Encoder comparison after equivalent retrieval fine-tuning Collection ModernBERT, Ettin and UltraBERT: 54 retrieval models, frozen data and results. Validation-selected models first. Matched training, one seed. • 59 items • Updated 12 days ago • 4