Downloads · 30 days
421
24% of all-time downloads
DavidAU/Llama-3.3-8B-Thinking-Gemini-Flash-11000x-128k
Llama-3.3-8B-Thinking-Gemini-Flash-11000x-128k is a text generation model from DavidAU. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as apache-2.0.
<small<B style="font-family: 'Arial Black', sans-serif; font-weight: 900; color: f1c40f; text-decoration: none; text-shadow: 1px 1px 0 d35400, 2px 2px 0 d35400, 3px 3px 0 d35400, 4px 4px 0 d35400, 5px 5px 0 d35400, 6p…
Downloads · 30 days
421
24% of all-time downloads
All-time downloads
1.8K
Public
Parameters
8B
16.1 GB on disk
Likes
7
Trending 1
Click a slice to open those files.
.safetensors16.1 GB · 100%
From the Hugging Face model README
<small><B style="font-family: 'Arial Black', sans-serif; font-weight: 900; color: #f1c40f; text-decoration: none; text-shadow: 1px 1px 0 #d35400, 2px 2px 0 #d35400, 3px 3px 0 #d35400, 4px 4px 0 #d35400, 5px 5px 0 #d35400, 6px 6px 7px rgba(0,0,0,0.4);">[DavidAU]</B> <a href="https://huggingface.co/DavidAU" target="_blank" style="padding-left:5px">FOLLOW</a> | <a href="https://huggingface.co/DavidAU" target="_blank">ALL MODELS</a> | <a href="https://huggingface.co/DavidAU/collections" target="_blank">COLLECTIONS</a> | <a href="https://huggingface.co/models?sort=trending&search=DavidAU" target="_blank">TRENDING</a> | <a href="https://huggingface.co/DavidAU/support" target="_blank">FAQs</a> | <a href="https://huggingface.co/DavidAU/support" target="_blank">SUPPORT</a> | <a href="https://buymeacoffee.com/davidbelton" target="_blank">Buy Me A Coffee (😁)</a></small>
<H2>Llama-3.3-8B-Thinking-Gemini-Flash-11000x-128k</H2> <img src="matrix-neo-reloaded-fight.gif" style="float:right; width:300px; height:300px; padding:10px;">What madness is this?
Someone found "Llama3.3-8B" source (never publicly released) in the "wild", then it was adjusted back to 128k and then I added my own special madness by training the model with Unsloth (3 epochs) and Gemini-2.5-flash-11000x (11000 samples) dataset.
This has created an Thinking model (128k context, Llama 3.3 model) which REASONS / THINKS like Gemini, but in a Llama 3.3 8B model.
Note this tuning was not only to create an thinking model, but ALSO update some the model's core knowledge / root training.
1 example generation at the bottom of this page.
Some "Llama" issues may still exist including out-of-date data.
<B>Thinking (will activate automatically) prompts - examples: </B>
Explain orbital mechanics including detailed math and examples.
Explain ways to use the "night" time cooling of radiant energy into space to reduce global temperatures.
Think Deeply: 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.
Think deeply: Tell me a horror story.
[sometimes you don't need the "think deeply"]
EXTENDED:
Come up with a plan to write:
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.
[ you may need to add after the "plan" appears: "Use the outline now, to write the scene"]
Certain phrases/words will automatically activate thinking:
<B>Instruct (thinking may NOT activate) prompts - examples:</B>
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.
Tell me a horror story.
<B>SETTINGS (suggested):</B>
Temp .7, rep pen 1.05, topp: .95, minp .05, topk: 40
Min context window: 4k, but suggest 8k+.
NO system prompt [thinking tags will self generate in MOST cases].
NOTE:
Temp can range from .1 to 2.5 or higher. Temp will NOT affect "thinking activation" in most cases.
Strongly suggest Q4KS or higher // IQ3_M (Imatrix) or higher to avoid reasoning/reasoning activation issues.
SYSTEM PROMPT to "force" thinking (can lead to odd / interesting "output"):
A user will ask you to solve a task. You should first draft your thinking process (inner monologue) until you have derived the final answer. Afterwards, write a self-contained summary of your thoughts (i.e. your summary should be succinct but contain all the critical steps you needed to reach the conclusion). You should use Markdown and Latex to format your response. Write both your thoughts and summary in the same language as the task posed by the user.
Your thinking process must follow the template below:
<think>
Your thoughts or/and draft, like working through an exercise on scratch paper. Be as casual and as long as you want until you are confident to generate a correct answer.
</think>
Special thanks to:
https://huggingface.co/allura-forge/Llama-3.3-8B-Instruct (found the model!)
https://huggingface.co/shb777/Llama-3.3-8B-Instruct-128K (adjusted to 128k, and other fixes)
https://huggingface.co/datasets/TeichAI/claude-4.5-opus-high-reasoning-250x (for the F..ing amazing dataset)
and Unsloth for making tuning too easy:
https://github.com/unslothai/unsloth
Details on the "madness":
https://www.reddit.com/r/LocalLLaMA/comments/1pz7bmv/llama338binstruct/
https://www.reddit.com/r/LocalLLaMA/comments/1q06ddc/update_on_the_llama_33_8b_situation/
<B>Settings: CHAT / ROLEPLAY and/or SMOOTHER operation of this model:</B>
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.
<B>Highest Quality Settings / Optimal Operation Guide / Parameters and Samplers</B>
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 , Lmstudio, "suggested settings"
NOTE: Some loss of formatting.
[[[thinking start]]]
Analyzing Cooling Mechanisms
I'm currently focused on how we can leverage the natural cooling of radiant energy during nighttime hours. My goal is to identify practical methods for enhancing this process to mitigate global warming. I'm exploring specific techniques and strategies.
Evaluating Radiative Cooling Methods
Now, I'm digging into how different materials and surfaces can be used to enhance radiative cooling. I'm taking a close look at the spectral properties of various materials, especially their emissivity in infrared wavelengths. The goal is to find the best ways to maximize heat loss during nighttime hours.
Synthesizing Practical Applications
I've just finished a thorough brainstorming session on practical applications. I'm now prioritizing strategies that are most impactful, like large-scale urban planning. I'm focusing on how to integrate reflective surfaces and high-emissivity materials into existing infrastructure to maximize the cooling effect during nighttime hours. The ultimate goal is to devise actionable steps.
Prioritizing Strategic Implementation
I've been drilling down further into the specifics of implementation, particularly the "how." I am now considering the feasibility of each strategy, keeping in mind factors like cost, technology availability, and potential for widespread adoption. My focus is on creating a clear, actionable roadmap.
Developing an Actionable Roadmap
The work continues to center around building out a practical implementation guide. Currently, I'm focusing on refining the "how" behind each strategy, with a strong emphasis on feasibility. I am also exploring the potential for integrating multiple strategies into cohesive solutions, ensuring that they complement each other effectively.
Developing the Comprehensive Plan
I'm now deep in the process of building out the comprehensive plan. My focus is still on the practical side, specifically how to translate theoretical concepts into actionable steps. I've broken down my thoughts into distinct sections: surface design, urban planning, and global-scale strategies. I am refining each section, focusing on clarity and feasibility.
[[[thinking end]]]
OUTPUT:
The "night time cooling of radiant energy into space" is a natural process where the Earth's surface radiates heat back out into the vastness of space, primarily in the infrared (IR) spectrum. This process is essential for regulating Earth's temperature, but it is naturally more efficient during the night due to several factors:
To leverage this natural process and reduce global temperatures, strategies often revolve around enhancing the Earth's ability to radiate heat more efficiently into space, especially during the cooler nighttime hours. While "night time cooling" isn't a direct cause of temperature reduction (it's a natural phenomenon itself), strategies that enhance the efficacy of this natural process contribute to it.
Here are ways to use and enhance this natural cooling mechanism:
The most straightforward way to enhance radiative cooling is by designing surfaces and buildings that emit infrared radiation more efficiently than standard materials.
Structural Designs:
Cities are massive heat sinks, and their surfaces often have high thermal emissivity due to dark materials like asphalt and concrete.
Agricultural land surfaces represent a significant portion of the Earth's surface.
These are very large-scale, highly complex, and often ethically contentious strategies that aim to change the atmospheric properties to allow for greater radiative cooling.
While not directly "night time cooling," a hypothetical array of large space-based reflectors could potentially increase the Earth's overall albedo by reflecting more solar radiation back into space, thereby preventing it from warming the atmosphere and the surface in the first place. This would indirectly lead to a greater ability for the Earth to radiatively cool at night, as there would be less excess heat generated during the day.
Key Considerations & Challenges:
In summary, leveraging the "night time cooling of radiant energy into space" involves designing surfaces and landscapes to radiate heat more efficiently into the vast IR sink of space, especially during the nighttime hours when atmospheric absorption is less of an issue. This can be achieved through material science innovations, urban planning, and even large-scale agricultural practices, always keeping in mind the complex interplay with Earth's atmosphere and the ethical implications of global climate engineering.