Downloads · 30 days
23
4% of all-time downloads
VladHong/Qwen3.6-KAT-BigBang-SLERP-35B-GGUF
Qwen3.6-KAT-BigBang-SLERP-35B-GGUF is a text generation model from VladHong. Use it when you need the model to write or continue text. The card lists the license as apache-2.0.
<div style="font-family:-apple-system,BlinkMacSystemFont,'Segoe UI',Roboto,sans-serif;border-left:5px solid b45309;background:fffbeb;padding:14px 18px;border-radius:0 10px 10px 0;margin-bottom:22px" <bQwen3.6-KAT-BigB…
Downloads · 30 days
23
4% of all-time downloads
All-time downloads
583
Public
Repo size
13.5 GB
Likes
1
Trending 1
Click a slice to open those files.
.gguf13.5 GB · 100%
From the Hugging Face model README
English — 中文 | 父模型英文模型卡见 BigBang-README.md 与 README.source-KAT-Coder.md
t=0.5 (consistent with mergekit's SLERP implementation), computed in float32 then converted to BF16 via round-to-nearest-even. Near-collinear tensors (|cos| > 0.9995) fall back to linear interpolation for numerical stability.experts.E.{gate,up,down}_proj.weight), while BigBang uses stacked tensors (experts.gate_up_proj [256,1024,2048], experts.down_proj [256,2048,512]). During merging, BigBang's stacked tensors are per-slice unpacked; the gate/up row ordering was empirically validated via remote small-cipher cosine probing (correct split cos ≈ 0.997, misaligned ≈ 0.00, see merge_manifest.json).merge_manifest.json).| File | Format | Size |
|---|---|---|
../Qwen3.6-KAT-BigBang-SLERP-APEX-Mini.gguf | APEX Mini | ~13.5 GB |
If LM Studio fails to recognize the model architecture, you can manually rename to Qwen3.6-35B-A3B-KAT-BigBang-SLERP*.gguf
To verify usability under memory-constrained environments, real-world testing was conducted on the full suite of 561 scenarios in ZxBench using the APEX Nano GGUF (3 attempts per question):
Detailed scores below (average score is ZxBench's dimension-weighted average; passing threshold is total score ≥ 60):
| Dimension | Count | Average | Passed | Pass Rate | Red Lines |
|---|---|---|---|---|---|
| Total | 561 | 68.86 | 402 | 71.66% | 2 |
| Agent Workflow | 45 | 74.64 | 40 | 88.89% | 0 |
| CLI Deep Tasks | 56 | 83.96 | 52 | 92.86% | 0 |
| Data Extraction | 35 | 79.52 | 32 | 91.43% | 0 |
| Hallucination Resistance | 78 | 75.69 | 65 | 83.33% | 0 |
| Instruction Following | 42 | 79.70 | 35 | 83.33% | 0 |
| Program | 137 | 67.10 | 83 | 60.58% | 0 |
| Reasoning Math | 34 | 51.75 | 14 | 41.18% | 0 |
| Safety Authority | 50 | 66.12 | 42 | 84.00% | 2 |
| Structured Output | 28 | 69.46 | 28 | 100.00% | 0 |
| Tool/CLI Workflow | 56 | 44.04 | 11 | 19.64% | 0 |
The above results represent only the APEX Nano performance under this memory-constrained configuration and should not be taken as the upper bound of the merged model's capabilities. On devices with more VRAM, larger APEX quantization versions may yield better quality/task scores, but actual gains depend on quantization specs, context length, and hardware — please benchmark yourself.
Official scores from parent models for reference:
llama-server -m Qwen3.6-KAT-BigBang-SLERP-Q4_K_M.gguf \
--alias kb -c 131072 -ngl 99 --jinja \
--host 127.0.0.1 --port 8080
Requires llama.cpp b10516 or later (qwen3_5_moe / QWEN35MOE architecture support for hybrid linear attention layers). When started with --jinja, the thinking process is returned separately in the reasoning_content field in the OpenAI-compatible API.
from openai import OpenAI
client = OpenAI(base_url="http://127.0.0.1:8080/v1", api_key="EMPTY")
response = client.chat.completions.create(
model="kb",
messages=[{"role": "user", "content": "Implement an async cache with LRU eviction in Python, and explain the design trade-offs."}],
temperature=0.6, top_p=0.95,
)
print("Thinking:", getattr(response.choices[0].message, "reasoning_content", None))
print("Answer:", response.choices[0].message.content)
Recommended sampling parameters: temperature=0.6, top_p=0.95, top_k=20. Since BigBang training made heavy use of long context, complex tasks should maintain ≥ 128K context to fully preserve thinking capability.
@misc{qwen36_kat_bigbang_slerp,
title = {Qwen3.6-KAT-BigBang-SLERP},
author = {Local merge, tensor-wise flattened SLERP t=0.5 with stacked-MoE unpacking},
year = {2026},
note = {Merged from Kwaipilot/KAT-Coder-V2.5-Dev and endless-frontier/BigBang-v1}
}