VoidStare's picture
Upload folder using huggingface_hub
410fb42 verified
---
base_model:
- PocketDoc/Dans-PersonalityEngine-V1.2.0-24b
- SicariusSicariiStuff/Redemption_Wind_24B
tags:
- merge
- mergekit
- lazymergekit
- PocketDoc/Dans-PersonalityEngine-V1.2.0-24b
- SicariusSicariiStuff/Redemption_Wind_24B
---
# WinterEngine-24B-Instruct
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
</head>
<div class="winter-container">
<div class="winter-case">
<div class="winter-inner-case">
<div class="winter-bezel">
<div class="terminal-screen">
<div style="text-align: center;">
<h2 style="color: #8ecae6; font-size: 32px;">WinterEngine-24B-Instruct</h2>
<pre class="code-block" style="display: inline-block; text-align: middle; white-space: pre; color: #ffffff;">
❄ ❄
❄ ❄
❄ ❄❄❄ ❄
❄❄❄❄❄❄❄❄❄
❄ ❄❄❄ ❄
❄ ❄
❄ ❄
</pre>
</div>
<h3 style="color: #8ecae6;">Key Details</h3>
<pre class="code-block" style="color: #ffffff; background: linear-gradient(135deg, #219ebc, #8ecae6);">
BASE MODEL: mistralai/Mistral-Small-24B-Base-2501
LICENSE: apache-2.0
LANGUAGE: English
CONTEXT LENGTH: 32768 tokens</pre>
<h3 style="color: #8ecae6;">Recommended Settings</h3>
<pre class="code-block" style="color: #ffffff; background: linear-gradient(135deg, #219ebc, #8ecae6);">
TEMPERATURE: 1.2
MIN_P: 0.05
(Everything Else Neutral MEME Samplers Too.)
</pre>
<h3 style="color: #8ecae6;">Prompting Format</h3>
<pre class="code-block" style="color: #ffffff; background: linear-gradient(135deg, #219ebc, #8ecae6);">
<|im_start|>system
system prompt<|im_end|>
<|im_start|>user
Hello, WinterEngine!<|im_end|>
<|im_start|>assistant
Hello! How can I help you today?<|im_end|></pre>
<h3 style="color: #8ecae6;">Quants</h3>
<pre class="code-block"style="color: #ffffff; background: linear-gradient(135deg, #219ebc, #8ecae6);">
I-mat: https://huggingface.co/mradermacher/WindEngine-24B-Instruct-i1-GGUF
Normal: https://huggingface.co/mradermacher/WindEngine-24B-Instruct-GGUF
<h2>Big Thanks to mradermacher for the Quants.</h2></pre>
<h3 style="color: #8ecae6;">Story</h3>
<pre class="code-block" style="color: #ffffff; background: linear-gradient(135deg, #219ebc, #8ecae6);">
You can ignore this if you want, but I just wanted to share something.
I was trying to create a model that follows prompts well, stays uncensored, and brings a lot of creativity β€” especially with roleplay capabilities.
Started out using the base 24B Instruct model β€” it was decent, but felt a bit dry and overly censored.
So, I began testing and merging different models.
Then found PersonalityEngine 24B, which followed instructions well and had solid roleplay potential, though it felt a little bland.
Discovered Redemption Winds β€” much better at roleplay, but not as strong when it came to following instructions. After trying three different model merges, this pairing turned out to be the best combination.
[The result? A model that follows instructions, excels at roleplay, and β€” for my single folks out there β€” works great for AI girlfriend roleplay, too.] </pre>
</div>
</div>
</div>
</div>
</div>
<style>
@import url('https://fonts.googleapis.com/css2?family=Fira+Code&display=swap');
.winter-container { background-color: #edf6f9; padding: 20px; border-radius: 20px; }
.winter-case { border: 2px solid #8ecae6; padding: 10px; }
.terminal-screen { background-color: #023047; color: #ffb703; padding: 15px; border-radius: 15px; font-family: 'Fira Code', monospace; }
.code-block { background: #219ebc; padding: 10px; border-radius: 10px; }
</style>
[LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
* [PocketDoc/Dans-PersonalityEngine-V1.2.0-24b](https://huggingface.co/PocketDoc/Dans-PersonalityEngine-V1.2.0-24b)
* [SicariusSicariiStuff/Redemption_Wind_24B](https://huggingface.co/SicariusSicariiStuff/Redemption_Wind_24B)
## 🧩 Configuration
```yaml
slices:
- sources:
- model: PocketDoc/Dans-PersonalityEngine-V1.2.0-24b
layer_range: [0, 40]
- model: SicariusSicariiStuff/Redemption_Wind_24B
layer_range: [0, 40]
merge_method: slerp
base_model: PocketDoc/Dans-PersonalityEngine-V1.2.0-24b
parameters:
t:
- filter: self_attn
value: [0, 0.5, 0.3, 0.7, 1]
- filter: mlp
value: [1, 0.5, 0.7, 0.3, 0]
- value: 0.5
dtype: bfloat16
```