⚠️ TEST MODEL - NOT FOR PRODUCTION USE ⚠️

This is my first model created with GRPO Reinforcement Learning! It was trained only for 1 epoch to produce 50 token answers. The model was created based entirely off the associated Hugging Face 🤗 course.

Downloads last month
7
Safetensors
Model size
135M params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for dzur658/SmolGRPO-135M

Finetuned
(173)
this model

Dataset used to train dzur658/SmolGRPO-135M