🏗️ Building on HF
Lynn
nerkyor
AI & ML interests
Post-training Qwen models for coding and efficient reasoning. Full quantization tiers plus speculative decoding (MTP / DFlash2) weights for llama.cpp, vLLM, SGLang and NInfer.
Recent Activity
updated a model about 8 hours ago
nerkyor/Qwen3.8-27B-Coder390-EfficientThink-Opus5.5-GPT6Astra-Grok4.7-DSV4Pro-K3-SFT-RLOO-GGUF-NInfer new activity about 10 hours ago
nerkyor/Qwen3.8-27B-Coder390-EfficientThink-Opus5.5-GPT6Astra-Grok4.7-DSV4Pro-K3-SFT-RLOO-MTP-DFlash2:The name is misleading new activity about 10 hours ago
nerkyor/Qwen3.8-27B-Coder390-EfficientThink-Opus5.5-GPT6Astra-Grok4.7-DSV4Pro-K3-SFT-RLOO-MTP-DFlash2:Q2/Q3 what doas that mean?Organizations
None yet