---
license: creativeml-openrail-m
base_model:
- prithivMLmods/Tulu-MathLingo-8B
tags:
- ollama
- math
- trl
- llama
- text-generation-inference
- math_lingo
datasets:
- microsoft/orca-math-word-problems-200k
language:
- en
pipeline_tag: text-generation
library_name: transformers
---
# Tulu-MathLingo-8B-GGUF Model Files

The **Tulu-MathLingo-8B** model is a fine-tuned version of **meta-llama/Llama-3.1-8B**, optimized for solving mathematical word problems and reasoning tasks in English and the Tulu language. The model integrates advanced language understanding and reasoning capabilities with a focus on providing solutions to math-related queries.

| **File Name**                  | **Size**     | **Description**                                  | **Upload Status** |
|--------------------------------|--------------|-------------------------------------------------|-------------------|
| `.gitattributes`               | 1.78 kB     | Tracks large files using Git LFS.                | Uploaded          |
| `README.md`                    | 449 Bytes   | Basic information about the Tulu-MathLingo model. | Updated           |
| `Tulu-MathLingo-8B.F16.gguf`   | 16.1 GB     | Full precision (FP16) model weights.             | Uploaded (LFS)    |
| `Tulu-MathLingo-8B.Q4_K_M.gguf`| 4.92 GB     | Quantized Q4_K_M model weights.                  | Uploaded (LFS)    |
| `Tulu-MathLingo-8B.Q5_K_M.gguf`| 5.73 GB     | Quantized Q5_K_M model weights.                  | Uploaded (LFS)    |
| `Tulu-MathLingo-8B.Q8_0.gguf`  | 8.54 GB     | Quantized Q8_0 model weights.                    | Uploaded (LFS)    |
| `config.json`                  | 29 Bytes    | Minimal configuration file.                      | Uploaded          |

### Sample Solve

![xvxv.png](https://cdn-uploads.huggingface.co/production/uploads/65bb837dbfb878f46c77de4c/vX8m-ltsacAztTF9SqDxB.png)

### **Key Features**

1. **Multilingual Math Reasoning:**  
   - Designed for solving complex math problems in **English** and **Tulu**.  

2. **Text Generation:**  
   - Generates detailed and contextually accurate text responses.

3. **Fine-Tuned Specializations:**  
   - Trained on the **microsoft/orca-math-word-problems-200k** dataset for word problem-solving.  

4. **Special Token Mapping:**  
   - Configured to use tokens for specific functions such as `<PAD>` and `<EOS>` effectively.

5. **Secure and Efficient Storage:**  
   - Model weights are stored in the **Safetensors** format for secure and faster inference.

6. **Large Parameter Size:**  
   - 8.03 billion parameters enable handling complex queries and multi-turn conversations.

---

### **Training Details**

- **Base Model:** [meta-llama/Llama-3.1-8B](#)  
- **Fine-Tuned:**  
   - Through multiple stages: **SFT (Supervised Fine-Tuning)** and **DPO (Direct Preference Optimization)**.

- **Dataset:**  
   - Trained on **200k word problems** from the **Microsoft Orca Math Word Problems Dataset**.

- **Model Size:**  
   - 8.03B parameters, optimized for **FP16** tensor type.

---

### **Applications**

1. **Mathematical Word Problems:**  
   - Solve structured or unstructured math problems in natural language.

2. **Conversational AI for Math:**  
   - Engage users in interactive dialogues focused on math and logic reasoning.

3. **Multilingual Support:**  
   - Supports queries in **Tulu** and **English**, enhancing accessibility.

4. **Education Tools:**  
   - Useful in tutoring systems for math, helping students with problem-solving.

---

### **Usage**

#### **Loading the Model**
```python
from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "prithivMLmods/Tulu-MathLingo-8B"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="fp16")
```

---

##### **Math Word Problem**
```python
query = "If a train travels 60 miles in 2 hours, what is its average speed?"
inputs = tokenizer(query, return_tensors="pt")
outputs = model.generate(**inputs, max_length=100)

response = tokenizer.decode(outputs[0], skip_special_tokens=True)
print("Answer:", response)
```
### **Performance Requirements**

- **Hardware:**  
   - Requires a GPU with at least **24GB VRAM** for optimal performance due to model size and FP16 usage.

- **Optimization:**  
   - Use mixed precision (`fp16`) for reduced memory footprint.  
   - Split inference across multiple GPUs if necessary.

---
# Run with Ollama [ Ollama Run ]

## Overview

Ollama is a powerful tool that allows you to run machine learning models effortlessly. This guide will help you download, install, and run your own GGUF models in just a few minutes.

## Table of Contents

- [Download and Install Ollama](#download-and-install-ollama)
- [Steps to Run GGUF Models](#steps-to-run-gguf-models)
  - [1. Create the Model File](#1-create-the-model-file)
  - [2. Add the Template Command](#2-add-the-template-command)
  - [3. Create and Patch the Model](#3-create-and-patch-the-model)
- [Running the Model](#running-the-model)
- [Sample Usage](#sample-usage)

## Download and Install Ollama🦙

To get started, download Ollama from [https://ollama.com/download](https://ollama.com/download) and install it on your Windows or Mac system.

## Steps to Run GGUF Models

### 1. Create the Model File
First, create a model file and name it appropriately. For example, you can name your model file `metallama`.

### 2. Add the Template Command
In your model file, include a `FROM` line that specifies the base model file you want to use. For instance:

```bash
FROM Llama-3.2-1B.F16.gguf
```

Ensure that the model file is in the same directory as your script.

### 3. Create and Patch the Model
Open your terminal and run the following command to create and patch your model:

```bash
ollama create metallama -f ./metallama
```

Once the process is successful, you will see a confirmation message.

To verify that the model was created successfully, you can list all models with:

```bash
ollama list
```

Make sure that `metallama` appears in the list of models.

---

## Running the Model

To run your newly created model, use the following command in your terminal:

```bash
ollama run metallama
```

### Sample Usage / Test

In the command prompt, you can execute:

```bash
D:\>ollama run metallama
```

You can interact with the model like this:

```plaintext
>>> write a mini passage about space x
Space X, the private aerospace company founded by Elon Musk, is revolutionizing the field of space exploration.
With its ambitious goals to make humanity a multi-planetary species and establish a sustainable human presence in
the cosmos, Space X has become a leading player in the industry. The company's spacecraft, like the Falcon 9, have
demonstrated remarkable capabilities, allowing for the transport of crews and cargo into space with unprecedented
efficiency. As technology continues to advance, the possibility of establishing permanent colonies on Mars becomes
increasingly feasible, thanks in part to the success of reusable rockets that can launch multiple times without
sustaining significant damage. The journey towards becoming a multi-planetary species is underway, and Space X
plays a pivotal role in pushing the boundaries of human exploration and settlement.
```
---

## Conclusion

With these simple steps, you can easily download, install, and run your own models using Ollama. Whether you're exploring the capabilities of Llama or building your own custom models, Ollama makes it accessible and efficient.


- This README provides clear instructions and structured information to help users navigate the process of using Ollama effectively. Adjust any sections as needed based on your specific requirements or additional details you may want to include.

---