Tieto Code (mini) is a fine-tuned variant of Qwen/Qwen3-4B-Instruct-2507 model.

The model focuses on producing clear, structured, and step-by-step answers across multiple topics.

Trained on a JSONL chat dataset with system/user/assistant message roles.

Base model: Qwen/Qwen3-4B-Instruct-2507

Type: Full fine-tune (chat/instruct)

Primary use: Software engineering (Q&A, troubleshooting, best practices)

Safetensors

Model size

4B params

Tensor type

BF16

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for vpakarinen/tieto-code-mini-4b-instruct

Base model

Finetuned

this model