lambsea/Qwen3.6-27B-AEON-Ultimate-Uncensored-UD-GGUF
Upload README.md with huggingface_hub
Add YAML metadata to model card
v2: Qwen3.6-27B-AEON-UD-IQ4_XS.gguf (SSM/attn at source precision, clean calibration)
v2: Qwen3.6-27B-AEON-UD-Q5_K_M.gguf (SSM/attn at source precision, clean calibration)
v2: Qwen3.6-27B-AEON-UD-Q6_K.gguf (SSM/attn at source precision, clean calibration)
v2: Qwen3.6-27B-AEON-UD-Q8_0.gguf (SSM/attn at source precision, clean calibration)
v2: updated imatrix (clean calibration, 65k context, GPU-native)
Update model card: GPU-native pipeline, SSM preservation, clean calibration
Update usage: YaRN 512k, q8_0 KV, unified KV, 5 slots, draft-mtp
Update usage: YaRN 512k, turbo4 KV, unified KV, 5 parallel slots, draft-mtp
Update README: add preserve_thinking kwarg, HF/GitHub URLs, chat template notes
Upload Qwen3.6-27B-AEON-UD-Q8_0.gguf with huggingface_hub
Upload Qwen3.6-27B-AEON-UD-Q6_K.gguf with huggingface_hub
Upload Qwen3.6-27B-AEON-UD-Q5_K_M.gguf with huggingface_hub
Upload Qwen3.6-27B-AEON-UD-IQ4_XS.gguf with huggingface_hub
Upload Qwen3.6-27B-AEON-mmproj-F16.gguf with huggingface_hub
Upload imatrix_merged.dat with huggingface_hub
Upload README.md with huggingface_hub
initial commit