NEDO Turkish SFT Mixtures: Supervised Fine-Tuning Datasets

Turkish supervised fine-tuning datasets prepared for the NEDO Turkish SLM project. The data was built to fine-tune NEDOQwen-style Turkish decoder-only language models after base pretraining. It was authored by Ethosoft and last updated on 2026-05-26.

Use Cases

• Fine-tuning Turkish decoder-only language models based on the described SFT mixtures.
• Training models for Turkish language tasks using supervised fine-tuning data.
• Benchmarking or extending models pre-trained on the related NEDO Turkish 65K tokenized web corpus.

Strengths

• Designed for a specific, documented project (NEDO Turkish SLM project).
• Intended for use with a related, named pretraining dataset (Ethosoft/nedo-turkish-65k-tokenized-60b).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Last updated 2026-05-26 11:14:28; freshness should be verified.

Provenance

Source

Ethosoft

Freshness

2026-05-26