Tyler Williams's picture

Building on HF

Tyler Williams PRO

unmodeled-tyler

·

https://quantaintellect.com

AI & ML interests

AI research engineer & solo operator of VANTA Research/Quanta Intellect

Recent Activity

liked a model about 19 hours ago

deepseek-ai/DeepSeek-V4-Pro

repliedto witcheer's post about 24 hours ago

updated my MoE offload bench dataset + collection. >>> previous finding: Qwen3.6-35B-A3B via full expert offload on RTX 4060 Ti 8GB + 32GB RAM → 7.4 tok/sec. RAM-ceilinged, disk-bound. >>> new finding: built llama.cpp from source inside WSL2, swept -ncmoe values for partial offload. ``` ncmoe 32, 16K ctx → 29.7 tok/sec ncmoe 30, 16K ctx → 32.0 tok/sec ncmoe 30, 32K ctx → 35.4 tok/sec ncmoe 28, 16K ctx → 16.3 tok/sec (VRAM cliff) ncmoe 30, 65K ctx → 17.4 tok/sec (VRAM cliff) ``` 4.8x faster than full offload. 8GB VRAM cliff is sharp - crossing ~7 GB halves throughput instantly. the hybrid SSM+attention architecture means 32K context is nearly free (KV cache only scales for 10/40 layers). dataset: https://huggingface.co/datasets/witcheer/windows-rtx-4060ti-8gb-moe-offload-bench-2026-05 collection: https://hf.co/collections/witcheer/8gb-vram-local-llms-practitioner-tested

reacted to sergiopaniego's post with 🔥 about 24 hours ago

OpenEnv already ships 🚢 with a ready-to-deploy RLM environment on free HF Spaces Drop "Attention Is All You Need", write code that spawns parallel LLM calls → ✅ correct answer, reward 1.0, in 4.2s Run GRPO (TRL) → model learns to write that search strategy itself test it yourself → https://huggingface.co/spaces/sergiopaniego/repl-env check out OpenEnv → https://github.com/meta-pytorch/OpenEnv

View all activity

Organizations

unmodeled-tyler 's collections 3