YAML Metadata Warning: empty or missing yaml metadata in repo card (https://huggingface.co/docs/hub/model-cards#model-card-metadata)

Indo Math Teacher (GGUF)

A fine-tuned math teaching model using the Gasing method for teaching mathematics in Bahasa Indonesia.

Model Description

This is the GGUF version of the Indo Math Teacher model, fine-tuned from Qwen2.5-Math-1.5B to provide step-by-step mathematical explanations using the Gasing method in Bahasa Indonesia.

Gasing Method

The Gasing method (Gampang, ASyik, dan menyenaNGkan - Easy, Fun, and Enjoyable) is an approach to teaching mathematics that focuses on:

  • Building strong foundational understanding through visual and practical examples
  • Using step-by-step explanations
  • Making math enjoyable and accessible

Usage

You can use this model with llama.cpp, text-generation-webui, or any other tool that supports GGUF models.

Using with llama.cpp

  1. Clone llama.cpp repository:
git clone https://github.com/ggerganov/llama.cpp
cd llama.cpp
make
  1. Download the model:
# Replace Lckoo1230 with the actual repository username
wget https://huggingface.co/Lckoo1230/indo-math-teacher-gguf/resolve/main/indo_math_teacher_q8_0.gguf
  1. Run the model:
./main -m indo_math_teacher_q8_0.gguf -n 1024 -p "Kamu adalah guru matematika yang menggunakan metode Gasing. Tolong jelaskan cara menghitung 7 x 8:"

Using with Ollama

Create a Modelfile:

FROM gguf:indo_math_teacher_q8_0.gguf
PARAMETER temperature 0.7
PARAMETER num_ctx 4096
SYSTEM "You are a math teacher using the Gasing method."

Create the model and chat with it:

ollama create indo-math-teacher -f Modelfile
ollama run indo-math-teacher

Examples

Basic Multiplication

Input: "Bagaimana cara mengalikan 7 x 8 dengan metode Gasing?"

Output:

Untuk mengalikan 7 x 8 dengan metode Gasing, kita bisa menggunakan pendekatan berikut:

1. Pertama, kita ingat bahwa 5 x 8 = 40
2. 7 berarti 5 + 2
3. Jadi 7 x 8 = (5 x 8) + (2 x 8) = 40 + 16 = 56

Atau cara lain:
1. 7 x 8 = 7 x (10 - 2) = 70 - 14 = 56

Dengan metode Gasing, kita memecah perhitungan menjadi langkah-langkah yang lebih mudah sehingga siswa dapat memahami konsep perkalian dengan lebih baik.

Model Details

  • Base Model: Qwen2.5-Math-1.5B
  • Training Data: Collection of mathematical explanations using the Gasing method in Bahasa Indonesia
  • Quantization: Q8_0
  • Context Length: 4096 tokens
  • License: The same license as the base model (Qwen2.5)
Downloads last month
5
GGUF
Model size
1.54B params
Architecture
qwen2
Hardware compatibility
Log In to view the estimation

8-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support