Instructions to use Sky1701/BeejX_CoT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Sky1701/BeejX_CoT with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Sky1701/BeejX_CoT",
	filename="cot_model_q5_0.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use Sky1701/BeejX_CoT with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Sky1701/BeejX_CoT:Q5_0
# Run inference directly in the terminal:
llama-cli -hf Sky1701/BeejX_CoT:Q5_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Sky1701/BeejX_CoT:Q5_0
# Run inference directly in the terminal:
llama-cli -hf Sky1701/BeejX_CoT:Q5_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 Sky1701/BeejX_CoT:Q5_0
# Run inference directly in the terminal:
./llama-cli -hf Sky1701/BeejX_CoT:Q5_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 Sky1701/BeejX_CoT:Q5_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Sky1701/BeejX_CoT:Q5_0

Use Docker

docker model run hf.co/Sky1701/BeejX_CoT:Q5_0

LM Studio
Jan
Ollama
How to use Sky1701/BeejX_CoT with Ollama:
```
ollama run hf.co/Sky1701/BeejX_CoT:Q5_0
```

Unsloth Studio

How to use Sky1701/BeejX_CoT 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 Sky1701/BeejX_CoT 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 Sky1701/BeejX_CoT to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Sky1701/BeejX_CoT to start chatting

How to use Sky1701/BeejX_CoT with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf Sky1701/BeejX_CoT:Q5_0

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "Sky1701/BeejX_CoT:Q5_0"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use Sky1701/BeejX_CoT with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf Sky1701/BeejX_CoT:Q5_0

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default Sky1701/BeejX_CoT:Q5_0

Run Hermes

hermes

Docker Model Runner
How to use Sky1701/BeejX_CoT with Docker Model Runner:
```
docker model run hf.co/Sky1701/BeejX_CoT:Q5_0
```

Lemonade

How to use Sky1701/BeejX_CoT with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Sky1701/BeejX_CoT:Q5_0

Run and chat with the model

lemonade run user.BeejX_CoT-Q5_0

List all available models

lemonade list

Model Card: BeejX CoT Agriculture Model

Model Information

Model Name: BeejX CoT Agriculture Model (SmolLM3-3B)
Base Architecture: SmolLM3 (3.08B parameters)
Specialty: Agricultural knowledge with Chain-of-Thought reasoning
Language: English
License: Apache 2.0

Files in this Repository

BeejX_cot_model.gguf          (5.74 GB) - Original model, F16 precision
BeejX_cot_model_q8_0.gguf     (3.05 GB) - Q8_0 quantization, ~99% quality
BeejX_cot_model_q5_0.gguf     (2.01 GB) - Q5_0 quantization, recommended
chat_with_BeejX_cot.bat                - Windows batch script for easy use

Quick Usage Examples

Example 1: Disease Identification

Q: What are common diseases in maize and how can they be managed?

A: Understanding that maize (corn) is susceptible to various diseases...
[Provides detailed management practices including irrigation, fertilization, 
pest control, and disease prevention strategies]

Example 2: Late Blight in Potato

Q: What is late blight in potato and how can it be managed?

A: Potatoes are susceptible to fungal diseases such as late blight, 
powdery mildew, and stem canker that can cause significant yield loss...
[Provides information about disease characteristics and management]

Example 3: General Agriculture

Q: What are common diseases in guava and how can they be prevented?

A: Guava can suffer from various fungal, bacterial, and viral diseases, 
as well as nematode and root rot problems, especially in poorly drained soils...
[Continues with prevention and treatment methods]

Technical Specifications

Model Architecture

Parameters: 3.08 billion
Layers: 36 transformer layers
Embedding Dimension: 2048
Attention Heads: 16 (4 KV heads)
Feed-forward Dimension: 11,008
Context Window: 65,536 tokens (training), 4,096 typical usage
Vocabulary Size: 128,256 tokens
RoPE Frequency Base: 5,000,000

Quantization Details

Q8_0: 8-bit quantization, minimal quality loss
Q5_0: 5-bit quantization, balanced size/quality trade-off
Quantization Tool: llama.cpp quantize

Performance Benchmarks

Inference Speed (CPU - 12 threads)

Original (F16): ~15-20 tok/s
Q8_0: ~20-25 tok/s
Q5_0: ~25-30 tok/s

Memory Requirements

Original: ~6 GB RAM
Q8_0: ~3.5 GB RAM
Q5_0: ~2.5 GB RAM

Training Information

Dataset

Primary Focus: Agricultural knowledge
Sources: ICAR data, agricultural research papers, farming guides
Format: Chain-of-Thought reasoning format
Training Method: Fine-tuning on SmolLM3-3B base

Training Features

Step-by-step reasoning for complex questions
Agricultural domain specialization
Crop disease identification and management
Farming best practices
General knowledge retention

Ethical Considerations

Intended Use

Educational purposes
Agricultural research
Farming knowledge assistance
Decision support (with expert verification)

Limitations

Knowledge Cutoff: Based on training data available at time of fine-tuning
Regional Specificity: May not account for all regional agricultural practices
Verification Required: Always verify critical decisions with local agricultural experts
Language: Primarily English, may have limited multilingual capabilities

How to Use on HuggingFace

Download the model:
- Choose your preferred quantization (Q5_0 recommended)
- Download using HuggingFace Hub or direct download

Run with llama.cpp:

llama-cli -m BeejX_cot_model_q5_0.gguf -p "What are diseases in maize?" -n 512

Use in Python:

from llama_cpp import Llama
llm = Llama(model_path="BeejX_cot_model_q5_0.gguf")
response = llm("What are diseases in maize?", max_tokens=512)

Community & Support

Issues: Report bugs or issues
Discussions: Ask questions and share experiences
Contributions: Suggestions for improvements welcome

Version History

v1.0: Initial release with Q5_0 and Q8_0 quantizations
- Base: SmolLM3-3B
- Specialty: Agricultural CoT reasoning
- Context: 4096 tokens

Citation

If you use this model in your research or applications, please cite:

@misc{BeejX-cot-agriculture-model,
  title={BeejX CoT Agriculture Model: Chain-of-Thought Reasoning for Agricultural Knowledge},
  author={[Sky1701/BeejX]},
  year={2025},
  publisher={HuggingFace},
  base_model={HuggingFaceTB/SmolLM3-3B}
}

Disclaimer: This model is for educational and research purposes. Always consult with agricultural experts and local extension services for critical farming decisions.

Downloads last month: 3

GGUF

Model size

3B params

Architecture

smollm3

Hardware compatibility

5-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for Sky1701/BeejX_CoT

Base model

HuggingFaceTB/SmolLM3-3B-Base

Finetuned

HuggingFaceTB/SmolLM3-3B

Quantized

(89)

this model