This is a merge of pre-trained language models created using mergekit, with all 15 native Multi-Token Prediction (mtp.*) speculative decoding tensors preserved and registered in model.safetensors.index.json.
Standard Qwen3.8-27B variants often suffer from circular <think> loops, mid-generation coordinate/logic drift, or unfinished multi-system code on complex single-shot prompts—while losing their native MTP speculative decoding head when merged naively. The goal was to combine three complementary coding and reasoning fine-tunes into a single high-accuracy model without sacrificing native MTP speed.
Qwen3.8-27B (even at xhigh thinking) and Qwen 3.8 Flash Next frequently fail or misalign physics and graphics.<think> tokens than heavily pruned fine-tunes, but spends that budget on clean, top-down architectural planning rather than repetitive self-doubt.mtp.* head inside the checkpoint enables 40–60 tokens/sec on HTML/code generation (and 25–40 tokens/sec on general text) in llama.cpp (Q8_0 imatrix, BF16 calibration PPL = 3.4195 ± 0.02407) with zero separate draft model required.This model was merged using the DARE-TIES (dare_ties) merge method using Qwen/Qwen3.8-27B as a base.
The following models were included in the merge:
density: 0.75, weight: 0.60)<think> structure, accurate 2D physics/spatial math, and polished visual design out of the box.density: 0.45, weight: 0.25)Qwen3.8-27B.density: 0.35, weight: 0.15)<think> Jinja template into tokenizer_config.json.mtp.* tensors from Qwen/Qwen3.8-27B into model-mtp-head.safetensors and registered them in model.safetensors.index.json.The following YAML configuration was used to produce this model:
merge_method: dare_ties
base_model: Qwen/Qwen3.8-27B # Base model acts as the neutral anchor
models:
- model: Jackrong/Qwopus3.8-27B-Coder-Flash
parameters:
density: 0.75
weight: 0.60
- model: ukisai/Swift-1.5-Qwen3.8-27b
parameters:
density: 0.45
weight: 0.25
- model: bottlecapai/ThinkingCap-Qwen3.8-27B
parameters:
density: 0.35
weight: 0.15
parameters:
normalize: true
int8_mask: true
dtype: bfloat16
model-00014-of-00015.safetensors 4.0 GBmodel-00013-of-00015.safetensors 4.0 GBmodel-00003-of-00015.safetensors 4.0 GBmodel-00008-of-00015.safetensors 4.0 GBmodel-00006-of-00015.safetensors 4.0 GBmodel-00012-of-00015.safetensors 4.0 GBmodel-00011-of-00015.safetensors 4.0 GBmodel-00009-of-00015.safetensors 4.0 GBmodel-00007-of-00015.safetensors 4.0 GBmodel-00004-of-00015.safetensors 4.0 GBmodel-00002-of-00015.safetensors 3.9 GBmodel-00005-of-00015.safetensors 3.9 GBmodel-00010-of-00015.safetensors 3.8 GBmodel-00001-of-00015.safetensors 2.5 GBmodel-mtp-head.safetensors 849 MBmodel-00015-of-00015.safetensors 718 MB