zils-mini-hc8

A body for the Zils d3-mini soup pool (Bittensor subnet 127): a fine-tune of Qwen/Qwen3.5-4B in the chassis layout, BF16, language weights trained, vision tower from the parent.

Start: the uniform weight average of four Apache-2.0 d3-mini house bodies: torchcast-ai/torchcast-decision-4b at 492bfb19, perplexity-ai/pplx-decider-flash-v1.2 at 9ff7cf98, vllm-sr/d3-mini at df99bf4c and Maincode/Matilda-Decision-v1-Mini at 39479a8a. Eight independent one-epoch LoRA fine-tunes of that start (rank 32, lr 1e-4, seeds 1-8, merged into the weights) were trained on 7,120 exam-format items written with Claude Opus, with a cross-entropy + Brier loss; the body is the uniform average of the eight merged weights. No other miner body and no soup is used (the locally written items and relay answers named below were not used for this body).

Training data: exam-format decision items written with the public Zils authoring recipe (scripts/author_exam.py) by Claude Opus and by a local open model, from public training splits only, plus the public Zils relay teacher answers. No validator exam items and no public seed items were used. This repository holds weights only; check the source datasets' terms before reuse.