Quick note on sudden performance loss when running GGUFs

<!-- SC_OFF --><div class="md"><p>Wrote a deep dive on <strong>FlashAttention-4 (03/05/2026)</strong> that's relevant for anyone thinking about inference performance.</p>

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1t4uj9h/mtp_on_strix_halo_with_llamacpp_pr_22673/"> <img alt="MTP on strix halo with llama.cpp (PR #22673)"

<!-- SC_OFF --><div class="md"><blockquote> <p><strong>WARNING: wait before download from HF: I just realised my upload of the new versions with the additional fix in the chat

<!-- SC_OFF --><div class="md"><p>Hey everyone, I've been working on getting Multi-Token Prediction (MTP) working with quantized GGUFs for Qwen3-27B and the results are pretty

<!-- SC_OFF --><div class="md"><p><a href="https://huggingface.co/unsloth/gemma-4-E2B-it-GGUF">https://huggingface.co/unsloth/gemma-4-E2B-it-GGUF</a></p> <p><a

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1srhtvg/where_we_are_in_a_year_everything_has_changed/"> <img alt="Where we are. In a year, everything has

<!-- SC_OFF --><div class="md"><p>After burning way too much money on GPT-4 API calls, I finally went all-in on local models with Ollama. The progress in 2026 is insane

<!-- SC_OFF --><div class="md"><p>All ggufs were broken, resulting in bad outputs, especially at long context.</p> <p>Anyway, it is fixed now: <a