Igor Butylenkov's picture
πŸ—οΈ Building on HF

Igor Butylenkov

end000

AI & ML interests

Local LLM inference optimization β€” MoE, MTP/speculative decoding, quantization, CUDA/llama.cpp Performing low-level vivisection of GPU kernels, dissecting code down to PTX, SASS, and machine instructions. πŸͺΏπŸ”₯

Recent Activity

liked a model 1 day ago
unsloth/Inkling-2-Mini-GGUF
liked a model 1 day ago
thinkingmachines/Inkling-2-Mini
liked a model 4 days ago
maxwell-model/Maxwell-V1-Flash
View all activity

Organizations

None yet