FrancescoArno94

Qwen3-1.7B-grpo-math-code

README

License: apache-2.0

Uploaded finetuned model

Developed by: FrancescoArno94
License: apache-2.0
Finetuned from model : FrancescoArno94/Qwen3-1.7B-grpo-math-code

This qwen3 model was trained 2x faster with Unsloth. This Qwen3 model was obtained with the following post-training-pipeline: Qwen3-1.7B --> sft on tulu dataset -> Qwen3-1.7B instruct ---> dpo on ultra-feedback dataset --- -> Qwen3.1 7B dpo --> grpo on a mix of 'open-r1/DAPO-Math-17k-Processed' and 'open-r1/verifiable-coding-problems-python' -- > Qwen3-1.7B grpo

Available on FriendliAI

Dedicated Endpoints

Run this model inference on single tenant GPU with unmatched speed and reliability at scale.

Learn more

Model Details

Model Provider

FrancescoArno94

Model Tree

Base

FrancescoArno94/Qwen3-1.7B-grpo-math-code

Fine-tuned

this model