# OLMo-3-7B-RLZero-Math
A 7B parameter model fine-tuned for mathematical reasoning using the RL-Zero approach from Allen Institute for AI.
## Key Features
- **65,536 token context** - Extended context with YaRN scaling for complex multi-step problems
- **RL-Zero training** - Reinforcement learning from scratch without human demonstrations
- **Step-by-step reasoning** - Trained to show work and explain mathematical logic
- **Fully open source** - Apache 2.0 license, weights and training code available
## Quick Start
```bash
ollama run richardyoung/olmo-3-7b-rlzero-math
```
## Example Prompts
**Algebra:**
```
Solve for x: 3x^2 - 12x + 9 = 0
```
**Calculus:**
```
Find the derivative of f(x) = x^3 * ln(x)
```
**Word Problems:**
```
A train leaves Station A at 60 mph. Another train leaves Station B (300 miles away) at 40 mph heading toward A. When do they meet?
```
**Proofs:**
```
Prove that the sum of the first n odd numbers equals n^2
```
## Available Formats
| Format | Size | Description |
|--------|------|-------------|
| Q4_K_M (default) | 4.2 GB | Best balance of quality and speed |
| GGUF/MLX/More | Various | See HuggingFace repo |
## Technical Requirements
- **Minimum:** 8GB RAM, any modern CPU
- **Recommended:** 16GB RAM or Apple Silicon Mac
## Links
- **Original Model:** [allenai/OLMo-3-7B-RLZero-Math](https://huggingface.co/allenai/OLMo-3-7B-RLZero-Math)
- **GGUF/MLX Quants:** [richardyoung/OLMo-3-7B-RLZero-Math-GGUF](https://huggingface.co/richardyoung/OLMo-3-7B-RLZero-Math-GGUF)
## Credits
- **Original Model:** Allen Institute for AI (Ai2)
- **Quantization:** Richard Young ([deepneuro.ai](https://deepneuro.ai/richard))
## License
Apache 2.0