Instructions to use Johnohhh1/gemma-3N-finetune-gguf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use Johnohhh1/gemma-3N-finetune-gguf 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 Johnohhh1/gemma-3N-finetune-gguf:Q8_0 # Run inference directly in the terminal: llama cli -hf Johnohhh1/gemma-3N-finetune-gguf:Q8_0
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf Johnohhh1/gemma-3N-finetune-gguf:Q8_0 # Run inference directly in the terminal: llama cli -hf Johnohhh1/gemma-3N-finetune-gguf:Q8_0
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 Johnohhh1/gemma-3N-finetune-gguf:Q8_0 # Run inference directly in the terminal: ./llama-cli -hf Johnohhh1/gemma-3N-finetune-gguf:Q8_0
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 Johnohhh1/gemma-3N-finetune-gguf:Q8_0 # Run inference directly in the terminal: ./build/bin/llama-cli -hf Johnohhh1/gemma-3N-finetune-gguf:Q8_0
Use Docker
docker model run hf.co/Johnohhh1/gemma-3N-finetune-gguf:Q8_0
- LM Studio
- Jan
- Ollama
How to use Johnohhh1/gemma-3N-finetune-gguf with Ollama:
ollama run hf.co/Johnohhh1/gemma-3N-finetune-gguf:Q8_0
- Unsloth Studio
How to use Johnohhh1/gemma-3N-finetune-gguf 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 Johnohhh1/gemma-3N-finetune-gguf 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 Johnohhh1/gemma-3N-finetune-gguf to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for Johnohhh1/gemma-3N-finetune-gguf to start chatting
- Atomic Chat new
- Docker Model Runner
How to use Johnohhh1/gemma-3N-finetune-gguf with Docker Model Runner:
docker model run hf.co/Johnohhh1/gemma-3N-finetune-gguf:Q8_0
- Lemonade
How to use Johnohhh1/gemma-3N-finetune-gguf with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull Johnohhh1/gemma-3N-finetune-gguf:Q8_0
Run and chat with the model
lemonade run user.gemma-3N-finetune-gguf-Q8_0
List all available models
lemonade list
Gemma-3N-ERG-Hazmat-Assistant
Model Description
A specialized fine-tune of Gemma-3N focused on hazardous materials emergency response. This model combines ERG (Emergency Response Guidebook) protocols with atmospheric dispersion modeling capabilities similar to MARPLOT/ALOHA software.
Intended Use
Primary Applications:
- Emergency response planning and training
- Hazmat incident initial assessment
- Chemical safety education
- Plume dispersion estimation
- UN dangerous goods classification assistance
NOT intended for:
- Replacement of professional emergency response
- Final decision-making in actual emergencies
- Legal or regulatory compliance determinations
Training Data
Fine-tuned on:
- ERG 2024 emergency response protocols
- MARPLOT atmospheric dispersion models
- UN dangerous goods classifications (UN1000-UN3500)
- Chemical safety data sheets
- Plume modeling scenarios
Dataset: Johnohhh1/erg_chemical_dataset (36,421 examples)
Capabilities
The model demonstrates knowledge in:
- Chemical Identification: UN numbers, chemical properties, hazard classes
- Emergency Response: Isolation distances, evacuation zones, PPE requirements
- Atmospheric Modeling: Plume dispersion, dense gas behavior, stability classes
- Complex Scenarios: BLEVEs, water-reactive chemicals, multi-chemical incidents
- Decontamination: Procedures for various chemical exposures
Technical Specifications
- Base Model: unsloth/gemma-3n-E4B-it (7.87B parameters)
- Fine-tuned Parameters: 19.2M (0.24% of total)
- Quantization: Q8_0 GGUF format (7.3GB)
- Training: 1 epoch on 36,421 examples
- Hardware: Trained on A100 GPU via Google Colab
Example Usage
With Ollama:
ollama run gemma-erg
Sample Prompts:
"What are the initial isolation distances for UN1005 (anhydrous ammonia)?"
"Calculate plume dispersion for 1000 gallons of chlorine in D stability conditions"
"What PPE is required for sulfuric acid decontamination?"
Performance
The model shows strong performance in:
- Providing structured emergency response guidance
- Calculating atmospheric dispersion estimates
- Identifying chemical-specific hazards
- Recommending appropriate PPE and decontamination
Limitations
- Not a replacement for professional hazmat response
- Atmospheric calculations are estimates requiring field validation
- May not reflect latest regulatory changes post-training
- Regional response protocols may vary from model recommendations
Safety & Ethics
⚠️ Critical Safety Notice: This model provides educational estimates only. Always defer to:
- Local emergency response authorities
- Professional hazmat teams
- Official ERG documentation
- Site-specific assessments
Citation
If you use this model, please cite:
@misc{gemma-erg-hazmat,
author = {Johnohhh1},
title = {Gemma-3N-ERG-Hazmat-Assistant},
year = {2024},
publisher = {HuggingFace},
url = {https://huggingface.co/Johnohhh1/gemma-3N-finetune-gguf}
}
Acknowledgments
- Trained using Unsloth for efficient fine-tuning
- Based on Google's Gemma architecture
- ERG data from public emergency response resources
- Downloads last month
- 3
8-bit