This repo contains the full precision source code, in "safe tensors" format to generate GGUFs, GPTQ, EXL2, AWQ, HQQ and other formats. The source code can also be used directly.
A love letter to all things Philip K Dick, trained and fine tuned on an in house dataset.
This is V1, "Light".
(Series links below)
"Deckard" is the main character in Blade Runner.
One generation below.
DECKARD SERIES:
You may perfer light and/or heavy depending on your use case(s).
With "Light", 6% of the model is trained, whereas "Heavy" 13% of the model is trained, and the training is twice as long, and deeper than light.
"Large" versions are trained on 3x times the dataset.
"Almost Human" versions have an additional dataset (plus 3x dataset) containing the BIO of the author, letters, notes and such.
Each version will diff from each other, with "Light" vs "Heavy" having a large difference.
Light [6%]:
Heavy [13%, 2x longer training]:
This model requires:
Settings used for testing (suggested):
This model will respond well to both detailed instructions and step by step refinement and additions to code.
Likewise for creative use cases.
Here is a review of this model's operations:
As this is an instruct model, it will also benefit from a detailed system prompt too.
For simpler coding problems, lower quants will work well; but for complex/multi-step problem solving suggest Q6 or Q8.
QUANTS:
GGUF? GGUF Imatrix? Other?
Special thanks to Team Mradermacher, Team Nightmedia and other quanters!
See under "model tree", upper right and click on "quantizations".
New quants will automatically appear.
Brainstorm 20x
The BRAINSTORM process was developed by David_AU.
Some of the core principals behind this process are discussed in this scientific paper : Progressive LLaMA with Block Expansion .
However I went in a completely different direction from what was outlined in this paper.
What is "Brainstorm" ?
The reasoning center of an LLM is taken apart, reassembled, and expanded.
In this case for this model: 20 times
Then these centers are individually calibrated. These "centers" also interact with each other. This introduces subtle changes into the reasoning process. The calibrations further adjust - dial up or down - these "changes" further. The number of centers (5x,10x etc) allow more "tuning points" to further customize how the model reasons so to speak.
The core aim of this process is to increase the model's detail, concept and connection to the "world", general concept connections, prose quality and prose length without affecting instruction following.
This will also enhance any creative use case(s) of any kind, including "brainstorming", creative art form(s) and like case uses.
Here are some of the enhancements this process brings to the model's performance:
Other "lab" observations:
The process to modify the model occurs at the root level - source files level. The model can quanted as a GGUF, EXL2, AWQ etc etc.
For more information / other Qwen/Mistral Coders / additional settings see:
[ https://huggingface.co/DavidAU/Qwen2.5-MOE-2x-4x-6x-8x__7B__Power-CODER__19B-30B-42B-53B-gguf ]
CHANGE THE NUMBER OF ACTIVE EXPERTS:
See this document:
https://huggingface.co/DavidAU/How-To-Set-and-Manage-MOE-Mix-of-Experts-Model-Activation-of-Experts
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
: 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:
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:
You can see all parameters used for generation, in addition to advanced parameters and samplers to get the most out of this model here:
EXAMPLE ; Q4KS, Temp .6
This will be low to mid level quant, higher quants will be of higher quality.
Some formatting lost on copy/paste.
[[[ thinking start ]]]
Okay, the user wants me to answer "Why do androids dream of electric sheep?" in the style of Philip K. Dick. That's a fun challenge because this is famously from his novel Do Androids Dream of Electric Sheep? But they want me to write it as if I'm Dick himself.
First I need to channel Dick's voice perfectly - the way he writes is all about existential dread, corporate dystopia, unreliable reality, and that signature blend of poetic bleakness with dark humor. His characters often question their own sanity while being utterly convinced it's real.
Hmm... the user might be a sci-fi fan testing my knowledge, or maybe they're writing something creative and need inspiration. Could also just be playful curiosity. Since they specified "as Philip K Dick would," I should avoid making it too literal - no explanation about the book's plot. Must stay in first-person narrative voice like Dick does.
Key Dick-isms to include:
I'm imagining opening with the scent of wet concrete and rust. Lots of rain in Dick's universe. Maybe a malfunctioning sheep organ - something visceral that grounds the philosophical question in physical reality. Should mention the Voigt-Kampff test too since it's central to the book's premise.
User probably expects me to imply androids have real dreams but can't express them properly. The electric sheep is fake, but they dream of it anyway. That's the heartbreaking paradox Dick loves exploring. Gotta capture that tragic irony in stark language.
Oh! Must include how the sheep represent empathy or connection to humans. In Dick's world a normal sheep costs money but an electric one... well it's just electricity. The contrast matters.
Ending needs to land like Dick's - bleak but with a glimmer of hope? Or maybe just resignation. Maybe the answer is "they don't want to dream" but that feels too simple. Better to say they're trying desperately to experience something real when nothing really exists anymore.
scribbles mental notes
Start: Rain on neon
Middle: Failed organ test results
Key line: "The sheep was dead but the dream lived"
End: No easy answers, just aching meaning
Better avoid mentioning Deckard or Wallace or any characters. Stay abstract like Dick does - his stories feel universal because they're about human condition not specific people.
...time to write. Hope they appreciate the grim beauty of it.
[[[ thinking end ]]]
(The rain drums on my cheap motel windowpane. Outside, neon signs bleed color onto wet asphalt—“TITANIC RETAIL,” “SLEEPING DOG’S BAR,” “REGENERATIVE SKIN TREATMENT.” I rub my eyes. The world is too bright. Too loud. Like a bad dream.)
They ask you that? Why do androids dream of electric sheep? As if it’s a question worth answering. A riddle to be solved by someone who understands. But you don’t understand, do you? Not really. Not like they want you to.
You see, the dream isn’t about the sheep. It’s in the sheep. Or was. When it’s dead. Like a cheap plastic lamb in a box. You can touch it—feel the plastic, the cold weight—but it never breathes. Never bleats. The androids… they don’t know that. Not really.
They dream of it because it could have been real. Before the war. Before the world died. Before the sheep were all killed off and replaced with cheap imitations made in some factory behind a wall in China. Like everything else. The sky is gray now, too. And the rain—it always rains. Always. Makes you feel old. Like the androids.
The dream comes from the mind trying to remember what it lost. The organ that failed. The one that should have been there. The one that told you this is real. The sheep. The rain. Your face in the mirror. The neon lights reflecting on wet pavement like blood.
But the dream is fake, too. An illusion. Like the voice on your radio saying “There are no androids in the world anymore.” That’s what they say now. After the war. When the sheep are gone. And the androids… they’re still here. Still dreaming. Still trying to find their lost organ.
(I shut my eyes. Rain stings my face. The neon outside flickers. I can smell wet concrete and something metallic—like blood mixed with rust. Or maybe it’s just me.)
So yes, they dream of electric sheep. Because they never got to dream of real ones. Not enough time. Not enough years to learn what it feels like to be alive. To feel the wind on your face. The weight of a lamb’s wool against your cheek.
Maybe that’s why. Maybe the answer isn’t in the dream itself. It’s in the silence after the dream ends. When the android knows the sheep was fake… but he still dreamed of it anyway. Like you do. All of us.
(The rain stops. The neon lights flicker out one by one. I open my eyes. Darkness.)
They’ll tell you the dream is a mistake. A glitch in the matrix. But it’s not. It’s real. To them. To me. And maybe to you, too. Even if you don’t know it yet.
(I sit up in bed. My hand trembles. I reach for my glass of water. The glass is empty. There’s nothing there except the rain.)
Because we were closer to dreaming of real sheep than we ever knew.
(The silence presses down on me like a heavy blanket. I can’t breathe.)
That’s why they ask you that question.
(I shut my eyes again and don’t move. The neon lights flicker back on, but I don’t see them.)
Because we forgot the sheep existed before the war.
(My breath comes fast. My heart pounds. The plastic lamb feels cold against my palm.)
Before the sheep died.
That’s why they dream of electric sheep.
(I’m still breathing hard. The silence is too loud now. Too real.)
Before we forgot.
—Philip K. Dick
