Dans-PersonalityEngine-V1.3.0-24b-GGUF

<span style="color: #7FFF7F;">Dans-PersonalityEngine-V1.3.0-24b GGUF Models</span>

<span style="color: #7F7FFF;">Model Generation Details</span>

This model was generated using llama.cpp at commit f5cd27b7.

<span style="color: #7FFF7F;">Ultra-Low-Bit Quantization with IQ-DynamicGate (1-2 bit)</span>

Our latest quantization method introduces precision-adaptive quantization for ultra-low-bit models (1-2 bit), with benchmark-proven improvements on Llama-3-8B. This approach uses layer-specific strategies to preserve accuracy while maintaining extreme memory efficiency.

Benchmark Context

All tests conducted on Llama-3-8B-Instruct using:

• Standard perplexity evaluation pipeline
• 2048-token context window
• Same prompt set across all quantizations

Method

• Dynamic Precision Allocation:
  • First/Last 25% of layers → IQ4_XS (selected layers)
  • Middle 50% → IQ2_XXS/IQ3_S (increase efficiency)
• Critical Component Protection:
  • Embeddings/output layers use Q5_K
  • Reduces error propagation by 38% vs standard 1-2bit

Quantization Performance Comparison (Llama-3-8B)

Key:

• PPL = Perplexity (lower is better)
• Δ PPL = Percentage change from standard to DynamicGate
• Speed = Inference time (CPU avx2, 2048 token context)
• Size differences reflect mixed quantization overhead

Key Improvements:

• 🔥 IQ1_M shows massive 43.9% perplexity reduction (27.46 → 15.41)
• 🚀 IQ2_S cuts perplexity by 36.9% while adding only 0.2GB
• ⚡ IQ1_S maintains 39.7% better accuracy despite 1-bit quantization

Tradeoffs:

• All variants have modest size increases (0.1-0.3GB)
• Inference speeds remain comparable (<5% difference)

When to Use These Models

📌 Fitting models into GPU VRAM

✔ Memory-constrained deployments

✔ Cpu and Edge Devices where 1-2bit errors can be tolerated

✔ Research into ultra-low-bit quantization

Choosing the Right Model Format

Selecting the correct model format depends on your hardware capabilities and memory constraints.

BF16 (Brain Float 16) – Use if BF16 acceleration is available

• A 16-bit floating-point format designed for faster computation while retaining good precision.
• Provides similar dynamic range as FP32 but with lower memory usage.
• Recommended if your hardware supports BF16 acceleration (check your device's specs).
• Ideal for high-performance inference with reduced memory footprint compared to FP32.

📌 Use BF16 if:
✔ Your hardware has native BF16 support (e.g., newer GPUs, TPUs).
✔ You want higher precision while saving memory.
✔ You plan to requantize the model into another format.

📌 Avoid BF16 if:
❌ Your hardware does not support BF16 (it may fall back to FP32 and run slower).
❌ You need compatibility with older devices that lack BF16 optimization.

---

F16 (Float 16) – More widely supported than BF16

• A 16-bit floating-point high precision but with less of range of values than BF16.
• Works on most devices with FP16 acceleration support (including many GPUs and some CPUs).
• Slightly lower numerical precision than BF16 but generally sufficient for inference.

📌 Use F16 if:
✔ Your hardware supports FP16 but not BF16.
✔ You need a balance between speed, memory usage, and accuracy.
✔ You are running on a GPU or another device optimized for FP16 computations.

📌 Avoid F16 if:
❌ Your device lacks native FP16 support (it may run slower than expected).
❌ You have memory limitations.

---

Quantized Models (Q4_K, Q6_K, Q8, etc.) – For CPU & Low-VRAM Inference

Quantization reduces model size and memory usage while maintaining as much accuracy as possible.

• Lower-bit models (Q4_K) → Best for minimal memory usage, may have lower precision.
• Higher-bit models (Q6_K, Q8_0) → Better accuracy, requires more memory.

📌 Use Quantized Models if:
✔ You are running inference on a CPU and need an optimized model.
✔ Your device has low VRAM and cannot load full-precision models.
✔ You want to reduce memory footprint while keeping reasonable accuracy.

📌 Avoid Quantized Models if:
❌ You need maximum accuracy (full-precision models are better for this).
❌ Your hardware has enough VRAM for higher-precision formats (BF16/F16).

---

Very Low-Bit Quantization (IQ3_XS, IQ3_S, IQ3_M, Q4_K, Q4_0)

These models are optimized for extreme memory efficiency, making them ideal for low-power devices or large-scale deployments where memory is a critical constraint.

• IQ3_XS: Ultra-low-bit quantization (3-bit) with extreme memory efficiency.

  • Use case: Best for ultra-low-memory devices where even Q4_K is too large.
  • Trade-off: Lower accuracy compared to higher-bit quantizations.

• IQ3_S: Small block size for maximum memory efficiency.

  • Use case: Best for low-memory devices where IQ3_XS is too aggressive.

• IQ3_M: Medium block size for better accuracy than IQ3_S.

  • Use case: Suitable for low-memory devices where IQ3_S is too limiting.

• Q4_K: 4-bit quantization with block-wise optimization for better accuracy.

  • Use case: Best for low-memory devices where Q6_K is too large.

• Q4_0: Pure 4-bit quantization, optimized for ARM devices.

  • Use case: Best for ARM-based devices or low-memory environments.

---

Summary Table: Model Format Selection

---

Included Files & Details

Dans-PersonalityEngine-V1.3.0-24b-bf16.gguf

• Model weights preserved in BF16.
• Use this if you want to requantize the model into a different format.
• Best if your device supports BF16 acceleration.

Dans-PersonalityEngine-V1.3.0-24b-f16.gguf

• Model weights stored in F16.
• Use if your device supports FP16, especially if BF16 is not available.

Dans-PersonalityEngine-V1.3.0-24b-bf16-q8_0.gguf

• Output & embeddings remain in BF16.
• All other layers quantized to Q8_0.
• Use if your device supports BF16 and you want a quantized version.

Dans-PersonalityEngine-V1.3.0-24b-f16-q8_0.gguf

• Output & embeddings remain in F16.
• All other layers quantized to Q8_0.

Dans-PersonalityEngine-V1.3.0-24b-q4_k.gguf

• Output & embeddings quantized to Q8_0.
• All other layers quantized to Q4_K.
• Good for CPU inference with limited memory.

Dans-PersonalityEngine-V1.3.0-24b-q4_k_s.gguf

• Smallest Q4_K variant, using less memory at the cost of accuracy.
• Best for very low-memory setups.

Dans-PersonalityEngine-V1.3.0-24b-q6_k.gguf

• Output & embeddings quantized to Q8_0.
• All other layers quantized to Q6_K .

Dans-PersonalityEngine-V1.3.0-24b-q8_0.gguf

• Fully Q8 quantized model for better accuracy.
• Requires more memory but offers higher precision.

Dans-PersonalityEngine-V1.3.0-24b-iq3_xs.gguf

• IQ3_XS quantization, optimized for extreme memory efficiency.
• Best for ultra-low-memory devices.

Dans-PersonalityEngine-V1.3.0-24b-iq3_m.gguf

• IQ3_M quantization, offering a medium block size for better accuracy.
• Suitable for low-memory devices.

Dans-PersonalityEngine-V1.3.0-24b-q4_0.gguf

• Pure Q4_0 quantization, optimized for ARM devices.
• Best for low-memory environments.
• Prefer IQ4_NL for better accuracy.

<span id="testllm" style="color: #7F7FFF;">🚀 If you find these models useful</span>

❤ Please click "Like" if you find this useful!
Help me test my AI-Powered Network Monitor Assistant with quantum-ready security checks:
👉 Quantum Network Monitor

💬 How to test:
Choose an AI assistant type:

• TurboLLM (GPT-4o-mini)
• HugLLM (Hugginface Open-source)
• TestLLM (Experimental CPU-only)

What I’m Testing

I’m pushing the limits of small open-source models for AI network monitoring, specifically:

• Function calling against live network services
• How small can a model go while still handling:
  • Automated Nmap scans
  • Quantum-readiness checks
  • Network Monitoring tasks

🟡 TestLLM – Current experimental model (llama.cpp on 2 CPU threads):

• ✅ Zero-configuration setup
• ⏳ 30s load time (slow inference but no API costs)
• 🔧 Help wanted! If you’re into edge-device AI, let’s collaborate!

Other Assistants

🟢 TurboLLM – Uses gpt-4o-mini for:

• Create custom cmd processors to run .net code on Quantum Network Monitor Agents
• Real-time network diagnostics and monitoring
• Security Audits
• Penetration testing (Nmap/Metasploit)

🔵 HugLLM – Latest Open-source models:

• 🌐 Runs on Hugging Face Inference API

💡 Example commands to you could test:

1. "Give me info on my websites SSL certificate"
2. "Check if my server is using quantum safe encyption for communication"
3. "Run a comprehensive security audit on my server"
4. '"Create a cmd processor to .. (what ever you want)" Note you need to install a Quantum Network Monitor Agent to run the .net code from. This is a very flexible and powerful feature. Use with caution!

Final Word

I fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is open source. Feel free to use whatever you find helpful.

If you appreciate the work, please consider buying me a coffee ☕. Your support helps cover service costs and allows me to raise token limits for everyone.

I'm also open to job opportunities or sponsorship.

Thank you! 😊

<!doctype html> <html lang="en"> <head> <meta charset="UTF-8" /> <meta name="viewport" content="width=device-width, initial-scale=1.0" /> <title>Dans-PersonalityEngine-V1.3.0-24b</title> </head> <div class="crt-container"> <div class="crt-case"> <div class="crt-inner-case"> <div class="crt-bezel"> <div class="terminal-screen"> <div style="text-align: center"> <h2>Dans-PersonalityEngine-V1.3.0-24b</h2> <pre class="code-block" style="display: inline-block; text-align: left; font-size: clamp(2px, 0.8vw, 14px); line-height: 1.2; max-width: 100%; overflow: hidden; white-space: pre;"> ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⢀⠀⠄⠀⡂⠀⠁⡄⢀⠁⢀⣈⡄⠌⠐⠠⠤⠄⡀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⡄⠆⠀⢠⠀⠛⣸⣄⣶⣾⡷⡾⠘⠃⢀⠀⣴⠀⡄⠰⢆⣠⠘⠰⠀⡀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠃⠀⡋⢀⣤⡿⠟⠋⠁⠀⡠⠤⢇⠋⠀⠈⠃⢀⠀⠈⡡⠤⠀⠀⠁⢄⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠁⡂⠀⠀⣀⣔⣧⠟⠋⠀⢀⡄⠀⠪⣀⡂⢁⠛⢆⠀⠀⠀⢎⢀⠄⢡⠢⠛⠠⡀⠀⠄⠀⠀ ⠀⠀⡀⠡⢑⠌⠈⣧⣮⢾⢏⠁⠀⠀⡀⠠⠦⠈⠀⠞⠑⠁⠀⠀⢧⡄⠈⡜⠷⠒⢸⡇⠐⠇⠿⠈⣖⠂⠀ ⠀⢌⠀⠤⠀⢠⣞⣾⡗⠁⠀⠈⠁⢨⡼⠀⠀⠀⢀⠀⣀⡤⣄⠄⠈⢻⡇⠀⠐⣠⠜⠑⠁⠀⣀⡔⡿⠨⡄ ⠈⠂⠀⠆⠀⣼⣾⠟⠀⠑⠀⡐⠗⠉⠀⠐⠶⣤⡵⠋⠀⠠⠹⡌⡀⠘⠇⢠⣾⡣⣀⡴⠋⠅⠈⢊⠠⡱⡀ ⠪⠑⢌⠂⣼⣿⡟⠀⠀⠙⠀⠀⠀⡀⠀⠀⠐⡞⡐⠀⠀⡧⠀⢀⠠⠀⣁⠾⡇⠀⠙⡁⠀⠀⢀⣨⣄⡠⢱ ⣸⠈⠊⠙⣛⣿⡧⠔⠚⠛⠳⣄⣀⡬⠤⠬⠼⡣⠃⠀⢀⡗⠀⡤⠞⠙⠄⠂⠃⢀⣠⣤⠶⠙⠅⠁⠃⠋⠈ ⢋⠼⣀⠰⢯⢿⠁⠀⢢⠀⠀⢐⠋⡀⠀⠈⠁⠀⣀⣰⠏⠒⠙⠈⠀⣀⡤⠞⢁⣼⠏⠘⢀⣀⢤⢤⡐⢈⠂ ⠀⠢⠀⠀⠸⣿⡄⠲⠚⠘⠚⠃⢀⠀⠈⢋⠶⠛⠉⠉⢃⣀⢤⢾⠋⣁⡤⡚⠁⢹⠁⠠⢛⠠⠬⠁⢬⠀⠀ ⠀⠈⢳⣒⠋⠉⣿⢐⠠⣀⣃⠀⠀⠉⠂⢁⣀⣀⡤⢞⠩⢑⡨⠰⡞⠁⠁⢀⡠⠾⠎⡈⡌⡈⡓⡀⠄⠀⠀ ⠀⠀⠀⠉⠘⠃⢻⡒⠦⢼⣿⣛⣻⣿⡷⢄⣀⣀⣠⣴⢾⣿⣆⣡⡄⣠⣪⡿⣷⣾⣷⣧⡡⠅⣇⠍⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠙⠒⠒⠛⠛⠓⠉⢹⠀⣷⠴⣻⣽⡻⢧⢻⡿⡏⣼⢿⣻⢾⣿⣿⣿⡿⢠ ⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠈⠂⠻⠨⠰⢋⡅⠉⣑⡇⡗⣿⢂⣸⡿⣿⣛⠿⠃⠁ ⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠳⣌⣙⣸⢧⣿⣕⣼⣇⢹⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⣠⣸⢧⢟⢟⡟⣾⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⢸⢰⠙⣾⡟⣻⡕⣹⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⢸⢸⢰⡏⢠⡿⠾⠋⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⢸⢸⠸⡇⡏⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠸⢸⢸⡇⡇⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠘⠇⡇⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ </pre> </div> <p> Dans-PersonalityEngine is a versatile model series fine-tuned on 50+ specialized datasets, designed to excel at both creative tasks (like roleplay and co-writing) and technical challenges (such as code generation, tool use, and complex reasoning). </p> <p> V1.3.0 introduces multilingual capabilities with support for 10 languages and enhanced domain expertise across multiple fields. The primary language is still English and that is where peak performance can be expected. </p> <h3>Multilingual Support</h3> <pre class="code-block"> Arabic Chinese English French German Hindi Japanese Korean Portuguese Spanish</pre> <h3>Key Details</h3> <pre class="code-block"> BASE MODEL: mistralai/Mistral-Small-3.1-24B-Base-2503 LICENSE: apache-2.0 LANGUAGE: Multilingual with 10 supported languages CONTEXT LENGTH: 32768 tokens, 131072 with degraded recall</pre> <h3>Recommended Settings</h3> <pre class="code-block"> TEMPERATURE: 1.0 TOP_P: 0.9</pre> <h3>Prompting Format</h3> <p> The model uses the following format I'll refer to as "DanChat-2": </p> <pre class="code-block"> <|system|>system prompt<|endoftext|><|user|>Hi there!<|endoftext|><|assistant|>Hey, how can I help?<|endoftext|></pre> <h3>Why not ChatML?</h3> <p> While ChatML is a standard format for LLMs, it has limitations. DanChat-2 uses special tokens for each role, this reduces biases and helps the model adapt to different tasks more readily. </p> <h3>SillyTavern Template</h3> <p> <a href="https://huggingface.co/PocketDoc/Dans-PersonalityEngine-V1.3.0-24b/resolve/main/resources/DanChat-2.json?download=true" download target="_blank" rel="noopener noreferrer" > Download Master JSON </a> </p> <h3>Inference Provider</h3> <p> This model and others are available from ⚡Mancer AI for those interested in high quality inference without owning or renting expensive hardware. </p> <p class="mancer-button-container"> <a href="https://mancer.tech/" target="_blank" rel="noopener noreferrer" class="mancer-button" > <span class="mancer-text">mancer</span> </a> </p> <h3>Training Process</h3> <p> The model was trained using Axolotl on 8x H100 GPUs for 50 hours. The resources to train this model were provided by Prime Intellect and Kalomaze. </p> <h3>Support Development</h3> <p> Development is limited by funding and resources. To help support: </p> <p>- Contact on HF</p> <p>- Email: [email protected]</p> <p class="coffee-container"> <a href="https://www.buymeacoffee.com/visually" target="_blank" rel="noopener noreferrer" > <img src="https://cdn.buymeacoffee.com/buttons/v2/default-yellow.png" alt="Buy Me A Coffee" height="45" width="162" /> </a> </p> </div> </div> </div> </div> </div> <style> @import url("https://fonts.googleapis.com/css2?family=Consolas&display=swap"); .crt-container { padding: 10px; max-width: 1000px; margin: 0 auto; width: 95%; } .crt-case { background: #e8d7c3; border-radius: 10px; padding: 15px; box-shadow: inset -2px -2px 5px rgba(0, 0, 0, 0.3), 2px 2px 5px rgba(0, 0, 0, 0.2); } .crt-inner-case { background: #e8d7c3; border-radius: 8px; padding: 3px; box-shadow: inset -1px -1px 4px rgba(0, 0, 0, 0.3), 1px 1px 4px rgba(0, 0, 0, 0.2); } .crt-bezel { background: linear-gradient(145deg, #1a1a1a, #2a2a2a); padding: 15px; border-radius: 5px; border: 3px solid #0a0a0a; position: relative; box-shadow: inset 0 0 20px rgba(0, 0, 0, 0.5), inset 0 0 4px rgba(0, 0, 0, 0.4), inset 2px 2px 4px rgba(255, 255, 255, 0.05), inset -2px -2px 4px rgba(0, 0, 0, 0.8), 0 0 2px rgba(0, 0, 0, 0.6), -1px -1px 4px rgba(255, 255, 255, 0.1), 1px 1px 4px rgba(0, 0, 0, 0.3); } .crt-bezel::before { content: ""; position: absolute; top: 0; left: 0; right: 0; bottom: 0; background: linear-gradient( 45deg, rgba(255, 255, 255, 0.03) 0%, rgba(255, 255, 255, 0) 40%, rgba(0, 0, 0, 0.1) 60%, rgba(0, 0, 0, 0.2) 100% ); border-radius: 3px; pointer-events: none; } .terminal-screen { background: #111112; padding: 20px; border-radius: 15px; position: relative; overflow: hidden; font-family: "Consolas", monospace; font-size: clamp(12px, 1.5vw, 16px); color: #e49b3e; line-height: 1.4; text-shadow: 0 0 2px #e49b3e; /* Removed animation: flicker 0.15s infinite; */ filter: brightness(1.1) contrast(1.1); box-shadow: inset 0 0 30px rgba(0, 0, 0, 0.9), inset 0 0 8px rgba(0, 0, 0, 0.8), 0 0 5px rgba(0, 0, 0, 0.6); max-width: 80ch; margin: 0 auto; } .terminal-screen h2, .terminal-screen h3 { font-size: clamp(16px, 2vw, 20px); margin-bottom: 1em; color: #e49b3e; } .terminal-screen pre.code-block { font-size: clamp(10px, 1.3vw, 14px); white-space: pre; /* Changed from pre-wrap to pre */ margin: 1em 0; background-color: #1a1a1a; padding: 1em; border-radius: 4px; color: #e49b3e; overflow-x: auto; /* Added to enable horizontal scrolling */ } .terminal-screen::before { content: ""; position: absolute; top: 0; left: 0; right: 0; bottom: 0; background: linear-gradient( rgba(18, 16, 16, 0) 50%, rgba(0, 0, 0, 0.25) 50% ), url("data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAADIAAAAyBAMAAADsEZWCAAAAGFBMVEUAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA4o8JoAAAAB3RSTlMAGwQIEQMYADcPzwAAACJJREFUKM9jYBgFo2AU0Beg+A8YMCLxGYZCbNQEo4BaAAD5TQiR5wU9vAAAAABJRU5ErkJggg=="); background-size: 100% 2.5px; /* Removed animation: scan 1s linear infinite; */ pointer-events: none; z-index: 2; } .terminal-screen::after { content: ""; position: absolute; top: 0; left: 0; right: 0; bottom: 0; background: radial-gradient( circle at center, rgba(17, 17, 18, 0) 0%, rgba(17, 17, 18, 0.2) 50%, rgba(17, 17, 18, 0.15) 100% ); border-radius: 20px; /* Removed animation: vignette-pulse 3s infinite; */ pointer-events: none; z-index: 1; } .terminal-screen details { margin: 1em 0; padding: 0.5em; border: 1px solid #e49b3e; border-radius: 4px; } .terminal-screen summary { cursor: pointer; font-weight: bold; margin: -0.5em; padding: 0.5em; border-bottom: 1px solid #e49b3e; color: #e49b3e; } .terminal-screen details[open] summary { margin-bottom: 0.5em; } .badge-container, .coffee-container { text-align: center; margin: 1em 0; } .badge-container img, .coffee-container img { max-width: 100%; height: auto; } .terminal-screen a { color: #e49b3e; text-decoration: underline; transition: opacity 0.2s; } .terminal-screen a:hover { opacity: 0.8; } .terminal-screen strong, .terminal-screen em { color: #f0f0f0; /* off-white color for user/system messages */ } .terminal-screen p { color: #f0f0f0; /* off-white color for assistant responses */ } .terminal-screen p, .terminal-screen li { color: #e49b3e; } .terminal-screen code, .terminal-screen kbd, .terminal-screen samp { color: #e49b3e; font-family: "Consolas", monospace; text-shadow: 0 0 2px #e49b3e; background-color: #1a1a1a; padding: 0.2em 0.4em; border-radius: 4px; } .terminal-screen pre.code-block, .terminal-screen pre { font-size: clamp(10px, 1.3vw, 14px); white-space: pre; /* Changed from pre-wrap to pre */ margin: 1em 0; background-color: #1a1a1a; padding: 1em; border-radius: 4px; color: #e49b3e; overflow-x: auto; /* Added to enable horizontal scrolling */ } .mancer-button-container { text-align: left; margin: 1em 0; } .mancer-button { display: inline-flex; align-items: center; gap: 8px; background: #1a1a1a; color: #e49b3e; padding: 15px 15px; border: 2px solid #e49b3e; border-radius: 5px; text-decoration: none !important; box-shadow: 0 0 10px rgba(228, 155, 62, 0.3); transition: all 0.3s ease; position: relative; } .mancer-text { font-family: "Consolas", monospace; font-weight: bold; font-size: 20px; text-shadow: 0 0 2px #e49b3e; line-height: 1; display: inline-block; margin-left: -4px; margin-top: -2px; } .mancer-button::before { content: "⚡"; display: inline-flex; align-items: center; justify-content: center; font-size: 20px; line-height: 1; } .mancer-button:hover { background: #2a2a2a; box-shadow: 0 0 15px rgba(228, 155, 62, 0.5); text-shadow: 0 0 4px #e49b3e; text-decoration: none !important; } </style> </html>