Model components / Training & synthetic datasets

SmolTalk

SmolTalk (and SmolTalk2) is Hugging Face's open SFT mixture behind the SmolLM instruct models, combining new Apache-2.0 synthetic sets (Smol-Magpie-Ultra and others) with established public datasets. Marin 8B Instruct also trains on it; component licenses follow their sources.

Family bucket: SmolTalk and SmolTalk2.

Openness

4 high confidence

4.0

access: ungated
license: apache-2.0(new-subsets)+per-component
dataset_card: present

Ungated; newly created subsets are Apache-2.0, incorporated public datasets keep their own licenses.

https://huggingface.co/datasets/HuggingFaceTB/smoltalk recorded 2026-07-04
ungated, mixed licensing documented on card

Adoption

3 high confidence

3.0

17,489 monthly downloads on Hugging Face, graded on the training-corpus bands.

https://huggingface.co/api/datasets/HuggingFaceTB/smoltalk recorded 2026-07-04
downloads field = 17,489 (30-day)

Capability

5 high confidence

5.0

Leads head-to-head SFT-mixture ablations (beats OpenHermes-2.5, UltraChat, MagPie-Pro) and is the current default across SmolLM2/3 and Marin 8B Instruct.

https://arxiv.org/html/2502.02737v1 recorded 2026-07-04
SmolLM2 paper: SmolTalk surpasses OpenHermes-2.5, UltraChat, and MagPie-Pro
https://huggingface.co/marin-community/marin-8b-instruct recorded 2026-07-04
Marin 8B Instruct lists SmolTalk in its SFT mix

Unchanged since 2026-07-04 (last edited, not re-checked)