Open models extended to 1M context with YaRN and certified needle by needle: Ornith, Gemma 4 uncensored, Qwen3.6 uncensored. MTP speculative decoding grafts, vision, full test harness.
yarn quantization mtp gemma uncensored apple-silicon llm long-context llama-cpp ollama qwen gguf speculative-decoding needle-in-a-haystack 1m-context ornith rope-scaling
-
Updated
Jul 9, 2026 - Python