Uncensored 1M Context: Gemma 4 + Qwen3.6 - a satgeze Collection

satgeze 's Collections

Beyond 1M Context

Uncensored 1M Context: Gemma 4 + Qwen3.6

Ornith 1M Context

Uncensored 1M Context: Gemma 4 + Qwen3.6

updated about 13 hours ago

Uncensored trunks with YaRN 1M context, needle-certified, plus MTP speculative decoding and vision. Full archives on ModelScope.

satgeze/Qwen3.6-35B-Uncensored-HauhauCS-1M-GGUF

Text Generation • 0.4B • Updated about 17 hours ago

Note Flagship: 70/70 needles to 1M, grafted MTP (+34%), vision verified. One file, four capabilities.
satgeze/Gemma4-12B-Uncensored-HauhauCS-1M-GGUF

Text Generation • 12B • Updated about 17 hours ago

Note First Gemma 4 certified needle-perfect at 1M, fits a 32GB card. MTP +51%.
satgeze/Gemma4-26B-A4B-Uncensored-HauhauCS-1M-GGUF

Text Generation • 25B • Updated about 17 hours ago

Note Fastest of the trio (MTP +48%); ~91% recall, honestly documented.
satgeze/Gemma4-31B-Uncensored-HauhauCS-1M-GGUF

Text Generation • 31B • Updated about 17 hours ago

Note Dense 31B, MTP +46%, top rungs in progress.
satgeze/QwenPaw-Flash-9B-heretic-1M-GGUF

Text Generation • 0.5B • Updated about 17 hours ago
satgeze/Gemma4-12B-Uncensored-HauhauCS-1.5M-GGUF

Text Generation • 12B • Updated about 13 hours ago

Note The 1.5M build: certified 10/10 through 1.31M on a 32GB card, factor-tax and 2M ceiling study published.