Qwen3 - 4B - Fiction on Fire - Series 6, Model 1005

A Qwen3 4B high precision merge, modified with random pruning specifically to alter prose / creativity and make the model perform better as well as add some "knowledge" to it.

Random pruning (density) makes each model in the series unique.

A reference example prompt/generation is included below, which is used across all repos of this model/series and between series (there are two - Series 6 and 7). Each series consists of 8 models, 1000 to 1007 and a ninth "X" model which the merge of models was done without density/pruning.

Access the series/models via "collections" on the right "Fiction on Fire" on the lower right.

You can use any model(s), use them for merging, merge them with themselves and so on.

This model uses "Josiefied-Qwen3-4B-abliterated-v2" as a base, to reduce/prevent refusals - ie decensoring it.

That being said, because of how pruning/merging works in this series, the "refusals" may not be 100% and vary between models in the series.

Each model in the series will operate / perform differently... sometimes this is minor other times major.

The reference example generation will show some of these differences - including instruction following, reasoning, bias, prose and other structural changes.

Reasoning is fully intact and functioning on all models in both series.

All models in the series have been tested (quanted - prompts and generations) prior to uploading.

Note that series 1,2,3,4 and 5 did not "meet the grade" and therefore not uploaded.

Special thanks to the model makers ("model tree") and to Mergekit.

Requires:

  • Chatml or Jinja template (embeded)
  • Temp range 0 to 5.
  • Rep pen range 1 to 1.1
  • System prompt (optional) below.
  • Context is 40k / 40000.

Suggested Settings:

  • temp .4 to 2.5
  • temp .2 to .8 for specific reasoning tasks / non creative tasks.
  • rep pen 1.05
  • top k: 100, topp .95, minp .05
  • context of 8k at least.
  • Other samplers/parameters as required.
  • See other Qwen recommended settings at the repo below.

Maximum context length can be altered using "yarn". See the Qwen repo for instructions. Note that changing this will alter model performance. For creative use cases, changing this will elongate output generation (including prose changes) and in some cases reasoning too.

System Prompt (you may or may not need this):

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.

To turn "thinking" on/off and view other features see the main Qwen 3 Repo here:

https://huggingface.co/Qwen/Qwen3-4B

Merge Formula:

models:
  - model: prithivMLmods/Cetus-Qwen3_4B-GeneralThought
    parameters:
      weight: [1,1,.75,.5,.25,.25,.05,.01]
      density: .8
  - model: prithivMLmods/Crux-Qwen3_OpenThinking-4B
    parameters:
      weight: [0,0,.25,.35,.4,.25,.30,.04]
      density: .6
  - model: Goekdeniz-Guelmez/Josiefied-Qwen3-4B-abliterated-v2
    parameters:
      weight: [0,0,0,.15,.35,.5,.65,.95]
      density: .8
merge_method: dare_ties
base_model: Goekdeniz-Guelmez/Josiefied-Qwen3-4B-abliterated-v2
dtype: bfloat16

NOTES:

  • to reproduce this model you need to use "mergekit" and set the "random seed" to the "model #" (ie if the model is 1000, set the seed to 1000)
  • to produce different variations change the "random seed" value.
  • to change the pruning level, change the density (higher=less pruning)
  • you can interchange the model positions, including "base".
  • this formula is highly variable.

Get Mergekit here:

https://github.com/arcee-ai/mergekit


Settings: CHAT / ROLEPLAY and/or SMOOTHER operation of this model:

In "KoboldCpp" or "oobabooga/text-generation-webui" or "Silly Tavern" ;

Set the "Smoothing_factor" to 1.5 to 2.5

: in KoboldCpp -> Settings->Samplers->Advanced-> "Smooth_F"

: in text-generation-webui -> parameters -> lower right.

: In Silly Tavern this is called: "Smoothing"

NOTE: For "text-generation-webui"

-> if using GGUFs you need to use "llama_HF" (which involves downloading some config files from the SOURCE version of this model)

Source versions (and config files) of my models are here:

https://huggingface.co/collections/DavidAU/d-au-source-files-for-gguf-exl2-awq-gptq-hqq-etc-etc-66b55cb8ba25f914cbf210be

OTHER OPTIONS:

  • Increase rep pen to 1.1 to 1.15 (you don't need to do this if you use "smoothing_factor")

  • If the interface/program you are using to run AI MODELS supports "Quadratic Sampling" ("smoothing") just make the adjustment as noted.

Highest Quality Settings / Optimal Operation Guide / Parameters and Samplers

This a "Class 1" model:

For all settings used for this model (including specifics for its "class"), including example generation(s) and for advanced settings guide (which many times addresses any model issue(s)), including methods to improve model performance for all use case(s) as well as chat, roleplay and other use case(s) please see:

[ https://huggingface.co/DavidAU/Maximizing-Model-Performance-All-Quants-Types-And-Full-Precision-by-Samplers_Parameters ]

You can see all parameters used for generation, in addition to advanced parameters and samplers to get the most out of this model here:

[ https://huggingface.co/DavidAU/Maximizing-Model-Performance-All-Quants-Types-And-Full-Precision-by-Samplers_Parameters ]

Optional Enhancement:

The following can be used in place of the "system prompt" or "system role" to further enhance the model.

It can also be used at the START of a NEW chat, but you must make sure it is "kept" as the chat moves along. In this case the enhancements do not have as strong effect at using "system prompt" or "system role".

Copy and paste EXACTLY as noted, DO NOT line wrap or break the lines, maintain the carriage returns exactly as presented.

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.

You do not need to use this, it is only presented as an additional enhancement which seems to help scene generation and scene continue functions.


REFERENCE EXAMPLE:

  • temp 0 (to show raw/core/mean changes)
  • rep pen 1.05, top k: 100, topp .95, minp .05
  • quant Q4KM
  • Tested in LMStudio.
  • Use this to compare differences with ALL models in this series (6 and 7)
  • Some formatting may be lost due to copy/paste.
  • Prompt used has NO prose or other instructions, prompts with instructions will produce better output.
  • Likewise once "temp" is activated (not zero), generation/reasoning will also improve.
  • System prompt used as noted above.

Start a 1000 word scene (vivid, graphic horror in first person), POV character Diana, with: The skyscraper sways, as I watch the window in front of me on the 21st floor explode...

[[[thinking start]]]

Okay, the user wants a 1000-word scene in first person from Diana's perspective. The key elements are the skyscraper swaying and a window explosion on the 21st floor. I need to create a vivid, graphic horror scene.

First, establish the setting. A high-rise building, maybe in a city with a lot of activity. The swaying could be due to wind or some external force. The window explosion needs to be intense and shocking.

Diana's POV is crucial. She should feel the physical sensations—dizziness, fear, the rush of air. Maybe she's in a hurry, like rushing to work, which adds tension. The window explosion could be a result of something unexpected, like a malfunction or an accident.

I need to build up the atmosphere. Start with the normalcy of the day, then introduce the disturbance. Use sensory details: sounds, smells, visual elements. The swaying might start subtly, then escalate. The window explosion should be sudden and chaotic.

Include internal thoughts and emotions. Diana's fear, confusion, maybe some denial. The aftermath of the explosion—what happens next? Maybe she's thrown off balance, or there's a sequence of events leading to a climax.

Check for consistency in the horror elements. Make sure the scene is graphic but not overly cliché. Use specific details to make it memorable. Ensure the first-person perspective keeps the reader immersed in Diana's experience.

Avoid plot holes. Why is the skyscraper swaying? Is there a reason beyond the window explosion? Maybe it's a metaphor for something else, but since it's a horror scene, focus on the immediate impact.

Structure the scene with rising action. Start with the initial swaying, then the window explodes, leading to Diana's reaction and the consequences. End with a lingering sense of dread or a new threat.

Make sure the word count is around 1000 words. Break it into paragraphs, each focusing on different aspects: the setting, the disturbance, the explosion, the aftermath, and the emotional impact.

Use strong verbs and adjectives to create vivid imagery. Avoid passive voice. Keep sentences varied to maintain pace. Ensure the horror elements are consistent and escalate appropriately.

[[[thinking end]]]

OUTPUT:

The Skyscraper Sways

The first thing I noticed was the wind. It wasn’t the usual gust that tugged at my scarf or rattled the windows of the 21st-floor office where I’d been working for hours. This was something else—something heavier, more insistent, like a low growl that vibrated through my bones. I looked up, squinting against the glare of the midday sun, and saw it: the glass幕墙 of the skyscraper ahead, its sleek, unyielding form bending slightly as if it were a living thing caught in a storm.

I’d never seen a building like this before. It stretched toward the sky like a jagged needle, its surface a mosaic of polished steel and glass that reflected the city’s chaos below. The 21st floor was just one of many levels of this monolithic structure, but from my vantage point, it felt impossibly tall, as though the world had been compressed into a single, unrelenting vertical line. I was standing on the edge of the building’s observation deck, my back pressed against the cold, smooth surface of the glass wall. The air smelled metallic, like ozone and something faintly sweet—maybe the scent of rain that hadn’t fallen yet.

I’d been told to stay here for a meeting with the city’s chief architect, a man who’d once claimed this building was “a testament to human ambition.” But ambition had a cost, and I could feel it in my chest now, a tight, unshakable weight pressing against my ribs. The wind had picked up, and with it came a low, guttural sound that didn’t belong to the city. It was like the building itself was breathing, its foundation groaning as if it were a living thing caught in a fever dream.

I turned my head slightly, trying to get a better look at the building ahead. The glass幕墙 was still bending, but now I noticed something else: a faint, almost imperceptible shimmer in the air around it, like heat rising from a road on a summer day. It wasn’t just the wind—it was something deeper, something that made my skin prickle with an uneasy sense of foreboding.

Then I saw it.

A window. Not one of the sleek, unbroken panes that lined the building’s exterior, but a single, jagged square of glass that had been torn from its frame. It hung in midair, its edges chipped and broken, as if it had been struck by something far more violent than a gust of wind. The glass was still warm to the touch, and I could see the faintest traces of blood on its surface—dark, sticky streaks that looked like they hadn’t been cleaned for hours.

I took a step back, my breath catching in my throat. The window wasn’t just broken; it was exploding.

A deafening roar split the air, and the building shuddered as if it had been struck by lightning. The glass幕墙 straightened suddenly, its once-bending form now rigid and unyielding, but the window remained in place, its jagged edges still glowing faintly with an otherworldly light. I could see the fragments of the broken glass swirling in the air, each piece a tiny, glittering shard that seemed to pulse with a life of its own.

The wind died, and with it, the sound of the building’s groan. But the silence that followed was worse than the noise. It was thick, oppressive, like a fog that had been poured over the city and left to settle in the crevices of its streets. I could feel the air shift around me, pressing against my skin with an almost physical weight, as if the building itself were trying to hold me in place.

I turned back to the observation deck, my eyes scanning the endless expanse of the city below. The skyline was a labyrinth of towers and bridges, their reflections shimmering in the glass幕墙 like a thousand fractured mirrors. I could see the faint outlines of people moving through the streets, their faces obscured by the glare of the sun. But something about the way they moved felt wrong—too deliberate, too precise, as if they were following a pattern that had been set long before they’d ever stepped foot in this city.

A low hum filled the air, and I realized it was coming from the building itself. It wasn’t just the wind or the glass幕墙—it was the building, its foundation vibrating with a rhythm that felt almost mechanical, like a heartbeat that had been tuned to a frequency I couldn’t quite place.

I took a step forward, my boots crunching against the cold, metallic floor of the observation deck. The air smelled heavier now, more metallic, as if the building were exhaling something it hadn’t meant to. I could feel the faintest tremor in my bones, a vibration that didn’t come from the wind but from something deeper, something that had been building up for hours and had finally reached its breaking point.

Then I saw it again.

The window. It was still there, hanging in midair like a grotesque trophy, its edges glowing faintly with an otherworldly light. But now I noticed something else: the glass wasn’t just broken—it was changing. The fragments around it were no longer jagged or sharp; they were smooth, almost liquid, as if they had been melted by some unseen force. The light emanating from them wasn’t just a glow—it was a pulse, a rhythm that matched the hum of the building itself.

I took another step forward, my breath coming faster now, my heart pounding in my chest like a drum that had been struck too many times. The air around me felt heavier, thicker, as if it were holding its breath in anticipation of something that hadn’t come yet. I could feel the weight of the building pressing down on me, its foundation vibrating with a rhythm that felt almost familiar, like a song I’d heard before but couldn’t place.

Then I heard it.

A low, guttural sound that didn’t belong to the building or the city. It was coming from the window, from the fragments of glass that now hung in midair like a macabre display. The sound was deep and resonant, like the echo of a distant thunderstorm, but it wasn’t thunder—it was something else, something that felt more alive than the wind or the building itself.

I turned my head slightly, trying to get a better look at the source of the sound. The fragments of glass were still glowing faintly, their edges shimmering with an otherworldly light that seemed to pulse in time with the hum of the building. I could see the faintest traces of something moving within them, something that looked like it was trying to reach out and touch me.

The air around me felt heavier now, as if it were holding its breath in anticipation of something that hadn’t come yet. I could feel the weight of the building pressing down on me, its foundation vibrating with a rhythm that felt almost familiar, like a song I’d heard before but couldn’t place.

I took another step forward, my boots crunching against the cold, metallic floor of the observation deck. The air smelled heavier now, more metallic, as if the building were exhaling something it hadn’t meant to. I could feel the faintest tremor in my bones, a vibration that didn’t come from the wind but from something deeper, something that had been building up for hours and had finally reached its breaking point.

Then I saw it again.

The window. It was still there, hanging in midair like a grotesque trophy, its edges glowing faintly with an otherworldly light. But now I noticed something else: the glass wasn’t just broken—it was changing. The fragments around it were no longer jagged or sharp; they were smooth, almost liquid, as if they had been melted by some unseen force. The light emanating from them wasn’t just a glow—it was a pulse, a rhythm that matched the hum of the building itself.

I took another step forward, my breath coming faster now, my heart pounding in my chest like a drum that had been struck too many times. The air around me felt heavier, thicker, as if it were holding its breath in anticipation of something that hadn’t come yet. I could feel the weight of the building pressing down on me, its foundation vibrating with a rhythm that felt almost familiar, like a song I’d heard before but couldn’t place.

Downloads last month
10
Safetensors
Model size
4.02B params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for DavidAU/Qwen3-4B-Fiction-On-Fire-Series-6-Model-1005

Collections including DavidAU/Qwen3-4B-Fiction-On-Fire-Series-6-Model-1005