---

license: apache-2.0

base_model: aimeri/spoomplesmaxx-thrasher-24B

library_name: gguf

pipeline_tag: text-generation

tags:

- roleplay

- creative-writing

- chatml

language:

- en

---

spoomplesmaxx-thrasher-24B — i1 GGUF (imatrix)

Weighted/imatrix GGUF quants of spoomplesmaxx-thrasher-24B

("Thrash Metal", mimids 02). Calibrated on ~2M chars sampled from the model's own

training corpus (the imatrix.dat is included if you want to cut your own sizes).

ChatML template embedded — llama.cpp picks it up without ceremony.

Prefer these over the static quants

at 3–4 bit; at 5 bit and above the statics are fine.

spoomplesmaxx-thrasher-24B.i1-IQ3_XXS.gguf   ~8.7 GB   smallest usable
spoomplesmaxx-thrasher-24B.i1-Q3_K_M.gguf    ~11 GB    12GB cards
spoomplesmaxx-thrasher-24B.i1-IQ4_XS.gguf    ~12 GB    quality-per-GB pick
spoomplesmaxx-thrasher-24B.i1-Q4_K_M.gguf    ~14 GB    the 24GB-card sweet spot

Sampler (swept on the full-precision model): temperature 1.0 · min_p 0.05.

A mild repetition_penalty 1.05 eliminated the verbatim-loop tail in our sweep at

the cost of a rare unfinished turn — a reasonable opt-in here. See the main card

for the full story: the token surgery that put ChatML on a Mistral base, the

checkpoint-selection battery, and the measured sampler guidance.

For adults. Stays in character by design; bring your own moderation. Apache 2.0.