darkmaniac7/Qwen3-14B-abliterated-MNN

Qwen3-14B Abliterated MNN

MNN 4-bit HQQ quantized model for TokForge on-device inference.

Model Info

Parameters	14B
Quantization	4-bit HQQ (quant_block=64)
Size	9.7GB
Source	huihui-ai/Qwen3-14B-abliterated
Backend	MNN OpenCL (GPU-accelerated)
Min RAM	24GB+

Performance (RedMagic SM8850)

Config	Decode Speed	Notes
Baseline (no spec decode)	6.4 tok/s	OpenCL GPU
With Acceleration Pack	10.4 tok/s	+63% (2880MHz)

Speculative Decoding

This model is compatible with the TokForge Acceleration Pack (abliterated Qwen3-0.6B draft model). The abliterated draft provides +5.5% better token acceptance vs censored draft, working with both censored and uncensored targets.

Usage

Download via TokForge app → Models → Roleplay category, or manually place files in the TokForge models directory.

Limitations and Intended Use

• Intended for TokForge / MNN mobile inference.
• Large-model performance is device-sensitive and can require backend or OEM-process-management tuning.
• This repo is a runtime/export artifact, not a standard Transformers release.

Files

• llm.mnn — Model graph
• llm.mnn.weight — Quantized weights
• tokenizer.txt — Tokenizer vocabulary
• llm_config.json — Model configuration
• embeddings_bf16.bin — Embedding table (8B/14B only)

Credits

• Original model: huihui-ai/Qwen3-14B-abliterated
• MNN framework: alibaba/MNN
• TokForge: tokforge.ai

Community

• Website: tokforge.ai
• Discord: Join the Discord