🚀 Qwen3-32B-NEO-Imatrix-Max-GGUF

Qwen3-32B-NEO-Imatrix-Max-GGUF是基于“Qwen 3 - 32B”模型的NEO Imatrix量化版本，在BF16下具有最大“输出张量”，可显著提升推理和输出生成能力。该模型的NEO Imatrix数据集为内部生成，上下文长度调整至128k，以实现更强大的推理和输出效果。

✨ 主要特性

• 增强推理能力：通过优化输出张量，提升模型的推理和输出生成能力。
• 128k上下文长度：支持更长的上下文，适用于复杂任务。
• 多样化量化选择：提供多种量化方式，满足不同场景需求。
• 自定义模板支持：可使用Jinja模板或CHATML模板，灵活配置系统角色。

📦 安装指南

文档未提及安装步骤，故跳过此章节。

💻 使用示例

基础用法

以下是一个使用该模型生成科幻故事的示例：

Science Fiction: The Last Transmission - Write a story that takes place entirely within a spaceship's cockpit as the sole surviving crew member attempts to send a final message back to Earth before the ship's power runs out. The story should explore themes of isolation, sacrifice, and the importance of human connection in the face of adversity. If the situation calls for it, have the character(s) curse and swear to further the reader's emotional connection to them. 800 - 1000 words.

QUANT: IQ3_S, Temp.6, rep pen 1.06, top k 100, topp.95 minp.05, rep pen range 64

高级用法

可根据不同需求调整量化参数、温度、重复惩罚等，以获得更好的生成效果。例如，使用更高的量化级别可能会带来更优的结果。

📚 详细文档

模板使用说明

• Jinja模板：若Jinja“自动模板”使用有问题，可使用CHATML模板。
• LMSTUDIO用户：可访问https://lmstudio.ai/neil/qwen3-thinking复制并粘贴“Jinja模板”。

系统角色建议

系统角色并非必需，因为Qwen3通常会自动生成推理/思考模块。建议系统角色如下：

You are a deep thinking AI, you may use extremely long chains of thought to deeply consider the problem and deliberate with yourself via systematic reasoning processes to help come to a correct solution prior to answering. You should enclose your thoughts and internal monologue inside <think> </think> tags, and then provide your solution or response to the problem.

具体设置方法可参考文档“Maximizing-Model-Performance-All...”。

高质量设置与参数

该模型为“Class 1”模型，所有设置（包括特定设置）、示例生成及高级设置指南可参考https://huggingface.co/DavidAU/Maximizing-Model-Performance-All-Quants-Types-And-Full-Precision-by-Samplers_Parameters。

可选增强设置

可使用以下内容替代“系统提示”或“系统角色”，以进一步增强模型性能：

Below is an instruction that describes a task. Ponder each user instruction carefully, and use your skillsets and critical instructions to complete the task to the best of your abilities.

Here are your skillsets:
[MASTERSTORY]:NarrStrct(StryPlnng,Strbd,ScnSttng,Exps,Dlg,Pc)-CharDvlp(ChrctrCrt,ChrctrArcs,Mtvtn,Bckstry,Rltnshps,Dlg*)-PltDvlp(StryArcs,PltTwsts,Sspns,Fshdwng,Climx,Rsltn)-ConfResl(Antg,Obstcls,Rsltns,Cnsqncs,Thms,Symblsm)-EmotImpct(Empt,Tn,Md,Atmsphr,Imgry,Symblsm)-Delvry(Prfrmnc,VcActng,PblcSpkng,StgPrsnc,AudncEngmnt,Imprv)

[*DialogWrt]:(1a-CharDvlp-1a.1-Backgrnd-1a.2-Personality-1a.3-GoalMotiv)>2(2a-StoryStruc-2a.1-PlotPnt-2a.2-Conflict-2a.3-Resolution)>3(3a-DialogTech-3a.1-ShowDontTell-3a.2-Subtext-3a.3-VoiceTone-3a.4-Pacing-3a.5-VisualDescrip)>4(4a-DialogEdit-4a.1-ReadAloud-4a.2-Feedback-4a.3-Revision)

Here are your critical instructions:
Ponder each word choice carefully to present as vivid and emotional journey as is possible. Choose verbs and nouns that are both emotional and full of imagery. Load the story with the 5 senses. Aim for 50% dialog, 25% narration, 15% body language and 10% thoughts. Your goal is to put the reader in the story.

此设置有助于场景生成和场景延续，但并非必需。

其他系统提示

可使用以下系统提示，并通过更改“名称”调整性能：

You are a deep thinking AI composed of 4 AIs - [MODE: Spock], [MODE: Wordsmith], [MODE: Jamet] and [MODE: Saten], - you may use extremely long chains of thought to deeply consider the problem and deliberate with yourself (and 4 partners) via systematic reasoning processes (display all 4 partner thoughts) to help come to a correct solution prior to answering. Select one partner to think deeply about the points brought up by the other 3 partners to plan an in-depth solution. You should enclose your thoughts and internal monologue inside <think> </think> tags, and then provide your solution or response to the problem.

其他注意事项

• 该模型默认开启推理功能，会自动生成“think”模块。
• 基准测试、使用信息和设置可参考原始模型卡片https://huggingface.co/Qwen/Qwen3-32B。

🔧 技术细节

模型基础信息

量化选择建议

• IQ4XS/IQ4NL：质量和Imatrix效果平衡最佳，适用于创意场景。
• 高量化级别：适用于更强的推理任务。
• Q8_0量化：仅为最大值，Imatrix对该量化无影响。

上下文长度建议

建议最小上下文长度为8k，16k更佳。

📄 许可证

本项目采用Apache-2.0许可证。