metadata
license: apache-2.0
tags:
- unsloth
- trl
- sft
datasets:
- nvidia/OpenCodeReasoning
- nvidia/OpenMathReasoning
base_model:
- Qwen/Qwen3-14B
Qwen 3 14b a great medium sized model. This qwen3 model is trained on 20.000 samples taken from Nvidia/OpenCodeReasoning and Nvidia/OpenMathReasoning.
Model use cases:
- Coding (python and c++)
- Solving math questions
- General reasoning
Use thinking mode using "/think" for the best results