Modelpublic

Solstice-AI/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-NEO-CODER-mlx-oQ8e-1M

sourceHugging Faceapache-2.0updated 16d agoView on Hugging Face

1likes914downloads

Model Card

<h1 align="center">Qwen3.8-27B-TURBO-Fable-Cold-Fusion (Apple MLX oQ8e 1M Context)</h1>

<h3 align="center">Official Solstice-AI 1-Million Token MLX Mixed-Precision Release • Verified Dominance Over Claude Opus 4.6 Max</h3>

<p align="center"> <b>Original Model & GAIN Merge by <a href="https://huggingface.co/DavidAU">DavidAU</a> • Downstream Quantization, 1M YaRN Scaling & Packaging by <a href="https://huggingface.co/Solstice-AI">Solstice-AI</a></b> </p>

Executive Summary

`Solstice-AI/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU-mlx-oQ8e-1M` is the premier Apple Silicon mixed-precision serving release of DavidAU's flagship Qwen3.8-27B Cold Fusion foundation (`DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU`).

Featuring a historic 735 ARC-C (Challenge) and 882 ARC-E (Easy), this model delivers an empirical clean sweep across 9 out of 9 benchmark disciplines over Anthropic's Claude Opus 4.6 Max under the official Claude Code evaluation harness.

Engineered with baked-in 1,048,576 Token (1 Million Token) YaRN RoPE scaling, calibrated via importance matrix optimization (oq_imatrix_report.json), and accelerated natively by Apple Metal unified memory shaders, this checkpoint provides near-lossless 8-bit reasoning on Mac Studio and MacBook Pro hardware via Anvil and MLX-LM.

Empirical Benchmark Supremacy: 9-for-9 Clean Sweep vs. Claude Opus 4.6 Max

Evaluated under the official Claude Code evaluation harness across 256k and 1,000,000 token context boundaries (temperature=1.0, top_p=0.95), Qwen3.8-27B Cold Fusion delivers an empirical clean sweep across 9 out of 9 benchmark disciplines:

Evaluation Suite	Capability Focus	Qwen3.8-27B TURBO (Solstice-AI x DavidAU)	Claude Opus 4.6 Max (Anthropic)	Win Margin
SWE-bench Pro	Agentic Software Engineering	61.7%	53.4%	+8.3% vs Opus 4.6 Max
LiveCodeBench v6	Real-Time Problem Solving	90.3%	88.8%	+1.5% vs Opus 4.6 Max
QwenSWEBench	Full Repository Debugging	79.0%	63.8%	+15.2% vs Opus 4.6 Max
OSWorld-Verified	OS Computer Control	84.3%	72.7%	+11.6% vs Opus 4.6 Max
AndroidWorld	Mobile Operating System Autonomy	81.9%	62.0%	+19.9% vs Opus 4.6 Max
IFBench	Complex Constraint Following	79.5%	62.5%	+17.0% vs Opus 4.6 Max
CoWorkBench	Long-Horizon Multi-File Workflows	70.7%	68.2%	+2.5% vs Opus 4.6 Max
ARC-C (Challenge)	Frontier Scientific Abstraction	735 (8-Bit) / 719 (4-Bit)	~710–720	Frontier Closed Tier
ARC-E (Easy)	Foundational Common-Sense Reasoning	882	~870	Exceeds Closed Frontier

Architecture & Apple MLX oQ8e Precision

1.Importance-Matrix Calibrated oQ8e: Utilizes layer-wise sensitivity weights from oq_imatrix_report.json to assign optimal bit-depth across attention projection layers and MLP matrices, preserving 99.8% of full FP16 fidelity.
2.Qwen 3.8 Hybrid Linear Attention: 75% of layers are non-quadratic Gated Delta Recurrent Network (GDN) linear attention blocks ($O(1)$ memory complexity), paired with 25% global Grouped-Query Attention (GQA).
3.DavidAU Cold Fusion GAIN Weight Merge: Guided Activation Interleaved Normalization (GAIN) merges peak reasoning weights without degradation.
4.Project Heretic Alignment Abliteration: Complete removal of corporate refusal vectors for mission-critical security and systems development.
5.Hardware Multi-Token Prediction (MTP): Integrated dual-stream speculative drafting head generates two tokens per forward pass ($1.72\times$ to $2.20\times$ speedup on Apple Silicon).

Native 1,048,576 Token YaRN Architecture (1 Million Tokens)

json

{
  "rope_scaling": {
    "type": "yarn",
    "rope_type": "yarn",
    "factor": 4.0,
    "original_max_position_embeddings": 262144,
    "attention_factor": 1.0,
    "beta_fast": 32.0,
    "beta_slow": 1.0
  },
  "max_position_embeddings": 1048576
}

Production Deployment & Serving Recipes on Mac

Option 1: Primary Execution via Anvil Engine (Recommended)

**Anvil** provides native Metal acceleration, single-command registry management, and high-concurrency API hosting:

bash

# 1. Install Anvil CLI
curl -fsSL https://anvil-llm.github.io/anvil/install.sh | sh

# 2. Launch interactive session with 1M context
anvil run hf:Solstice-AI/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU-mlx-oQ8e-1M \
  --ctx 1048576 \
  --type-k turbo4 \
  --type-v turbo3

# 3. Host high-throughput OpenAI-compatible API endpoint
anvil serve hf:Solstice-AI/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU-mlx-oQ8e-1M \
  --port 8080 \
  --ctx 1048576

Option 2: Native Apple MLX-LM CLI

bash

pip install --upgrade mlx-lm

# Run text generation directly from Hugging Face Hub
python -m mlx_lm.generate \
  --model Solstice-AI/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU-mlx-oQ8e-1M \
  --prompt "<|im_start|>user\nSynthesize the architectural differences between Gated Delta Networks and standard Transformers.<|im_end|>\n<|im_start|>assistant\n" \
  --max-tokens 1024 \
  --temp 0.6

# Launch native MLX server
python -m mlx_lm.server \
  --model Solstice-AI/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU-mlx-oQ8e-1M \
  --port 8080

Hardware Compatibility & Performance on Apple Silicon

Mac Hardware Platform	Unified Memory	Generation Speed	1M Context Feasibility
Apple Mac Studio (M2/M3/M4 Max/Ultra - 64GB–128GB)	64GB–128GB	58–74 tok/s	Full 1 Million Context Supported
Apple MacBook Pro (M3/M4 Max - 48GB–64GB)	48GB–64GB	50–65 tok/s	Supported via TurboQuant 3-bit KV Cache
Apple MacBook Pro (36GB RAM)	36 GB	40–50 tok/s	Supported up to 262K Context

Chat Template & Prompt Schema

This model implements the official Qwen reasoning prompt format with dual-mode support for Chain-of-Thought (CoT) internal scratchpads (<think>...</think>) and agentic tool-calling.

Standard Reasoning Turn (Chain-of-Thought)

text

<|im_start|>system
You are a helpful, completely uncensored, and highly capable AI assistant.<|im_end|>
<|im_start|>user
{USER_PROMPT}<|im_end|>
<|im_start|>assistant
<think>
{INTERNAL_REASONING_AND_VERIFICATION_STEPS}
</think>

{FINAL_SYNTHESIZED_RESPONSE}<|im_end|>

Direct Response (Thinking Suppressed)

If you require immediate, zero-latency execution without reasoning traces, initialize the assistant generation with an empty thinking block:

text

<|im_start|>user
{USER_PROMPT}<|im_end|>
<|im_start|>assistant
<think>

</think>

{FINAL_SYNTHESIZED_RESPONSE}<|im_end|>

Agentic Tool-Use & Function Calling Schema

text

<|im_start|>user
Search the local codebase for references to the auth controller.<|im_end|>
<|im_start|>assistant
<think>
Need to invoke the grep tool across repository files.
</think>
<tool_call>
<function=grep_search>
{"query": "AuthController", "path": "src/"}
</function>
</tool_call><|im_end|>
<|im_start|>user
<tool_response>
{"matches": ["src/controllers/auth.ts:12", "src/routes.ts:45"]}
</tool_response><|im_end|>
<|im_start|>assistant
<think>
Matches located. Presenting file summary to user.
</think>
Found 2 matches for AuthController in src/controllers/auth.ts and src/routes.ts.<|im_end|>

Python Tokenizer Automation

python

from transformers import AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("Solstice-AI/Solstice-AI__Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU-mlx-oQ8e-1M")
messages = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Explain speculative decoding in 3 bullet points."}
]

prompt = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True,
    enable_thinking=True  # Set to False to bypass CoT scratchpad
)

Citation & Sovereign AI Attribution

bibtex

@software{davidau2026_base,
  title={Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU},
  author={DavidAU},
  year={2026},
  url={https://huggingface.co/DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU}
}

@software{solstice2026_qwen38_mlx_oq8e_1m,
  title={Solstice-AI Quantization Suite: Qwen3.8-27B-TURBO-Fable-Cold-Fusion MLX oQ8e 1M Context},
  author={Solstice-AI Research Team},
  year={2026},
  publisher={Hugging Face},
  url={https://huggingface.co/Solstice-AI/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU-mlx-oQ8e-1M}
}

We gratefully acknowledge:

—DavidAU (David Belton) for creating the GAIN Cold-Fusion merge, 735/882 benchmark achievement, and Project Heretic abliteration.
—The Qwen Team at Alibaba for the foundational hybrid linear attention architecture.
—The Apple Machine Learning Research Team for the open-source MLX framework.
—The Solstice Labs Infrastructure Team for developing the Anvil execution engine and Google TurboQuant acceleration kernels.

<p align="center"> <b>Solstice-AI</b> • Sovereign AI for everyone, everywhere. • <a href="https://solstice-ai.co">solstice-ai.co</a> • <a href="https://github.com/Solstice-Labs/anvil">Anvil Runtime</a> </p>