Instructions to use AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched",
	filename="Dolphin-2.9.3-Mistral-Nemo-12B-Patched-Q2_K.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched:Q4_K_M

Use Docker

docker model run hf.co/AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched:Q4_K_M

LM Studio
Jan

vLLM

How to use AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched:Q4_K_M

Ollama
How to use AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched with Ollama:
```
ollama run hf.co/AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched:Q4_K_M
```

Unsloth Studio

How to use AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched to start chatting

Atomic Chat new
Docker Model Runner
How to use AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched with Docker Model Runner:
```
docker model run hf.co/AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched:Q4_K_M
```

Lemonade

How to use AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched:Q4_K_M

Run and chat with the model

lemonade run user.Dolphin-2.9.3-Mistral-Nemo-12B-Patched-Q4_K_M

List all available models

lemonade list

"This is humanity's race.

The solution is open source.

Stay sovereign."

— AIOpsInSpace

📑 Table of Contents
🎯 1. Executive Summary
🏗️ 2. Architecture & Merging
🚀 3. Technical Enhancements
📊 4. Benchmark Competitiveness
🏆 5. Arena Analytics
🔍 6. SWOT Analysis
⚡ 7. Usage & Deployment
⚙️ 8. Backend Compatibility
📜 9. Disclaimers & Credits

Dolphin-2.9.3-Mistral-Nemo-12B-Patched
AIOpsInSpace Official
Dolphin 2.9.3 Mistral-Nemo 12B patched for seamless 128K context generation without tokenizer hangs.

🐬 12B Dense Model
⚡ 128K Context Support
🛠️ Tokenizer Loop Patched
> What is this model and Why is it Needed?
Dolphin-2.9.3-Mistral-Nemo-12B-Patched is built on top of cognitivecomputations/dolphin-2.9.3-mistral-nemo-12b.

Why it is needed: Fixes GGUF tokenizer token mapping bugs that caused local backends to crash or enter infinite generation loops on long prompts.
> From the Parent Repository
"Dolphin 2.9.3 on Mistral-Nemo 12B delivers unmatched punch for a 12B model."


— Cognitive Computations

↑ Back to Top

🏗️ 2. Model Architecture & Merging

Architecture: Mistral-Nemo 12B Transformer Architecture

Merging Technique: Tokenizer Vocabulary Patching

Constituent Models:
Base Model: cognitivecomputations/dolphin-2.9.3-mistral-nemo-12b

Methodology: Hard-patched special tokens to ensure clean EOS handling across all llama.cpp quantizations.

↑ Back to Top

🚀 3. Technical Enhancements

> Key Upgrades Over Base Model:
128K Long-Context: Handles extensive document analysis and codebases easily.
Uncensored Freedom: Full Dolphin uncensored dataset fine-tune.

↑ Back to Top

📊 4. Benchmark Competitiveness vs. Frontier Scores

    > Evaluated Performance



Benchmark
Dolphin-2.9.3-Mistral-Nemo-12B-Patched
Frontier Target


MMLU
Evaluated
88.7%

GSM8K
Evaluated
95.6%

HumanEval
Evaluated
90.2%

Benchmark	Dolphin-2.9.3-Mistral-Nemo-12B-Patched	Frontier Target
MMLU	Evaluated	88.7%
GSM8K	Evaluated	95.6%
HumanEval	Evaluated	90.2%

↑ Back to Top

🏆 5. Comprehensive Arena Analytics

> Status: Active Community Benchmarking
// Note: Arena Elo and head-to-head winrates updated continuously as evaluation telemetry processes.

↑ Back to Top

🔍 6. SWOT Analysis

> Strengths (S)
🛡️ Uncensored Fidelity: Surgically patched to ensure maximum generation throughput without alignment overhead.
⚡ Optimized Engine: Advanced mechanics ensure zero context fragmentation or execution hangs.


> Weaknesses (W)
📉 Hardware Limits: Requires sufficient VRAM/RAM for higher precision GGUF quantizations.


> Opportunities (O)
🎯 Local Sovereign Agents: Perfect for offline, private reasoning and agentic workflows.


> Threats (T)
⚠️ Sampler Sensitivity: High temperatures may require repetition penalty adjustments.

↑ Back to Top

⚡ 7. Usage & Deployment Info

> Recommended Settings
Temperature: 0.2 - 0.7
Top-P: 0.95
Backend Engines: Compatible with llama.cpp, vLLM, Ollama, LM Studio, KoboldCPP

↑ Back to Top

⚙️ 8. Backend Compatibility

> Validated Engines:
[+] llama.cpp: Native support across all quantizations.
[+] Ollama / LM Studio: Full GGUF compatibility.

↑ Back to Top

📜 9. Disclaimers & Credits

Disclaimer: Dolphin-2.9.3-Mistral-Nemo-12B-Patched is provided for research and sovereign local deployment. As an unaligned model, users are responsible for ensuring usage complies with local laws.

Credits: Gratitude to original base model authors (cognitivecomputations/dolphin-2.9.3-mistral-nemo-12b) and open-source AI community tools.

Downloads last month: 505

GGUF

Model size

12B params

Architecture

llama

Hardware compatibility

2-bit

3-bit

4-bit

5-bit

6-bit

8-bit

Model tree for AIOpsInSpace/Dolphin-2.9.3-Mistral-Nemo-12B-Patched

Base model

mistralai/Mistral-Nemo-Base-2407

Finetuned

dphn/dolphin-2.9.3-mistral-nemo-12b

Quantized

(7)

this model