Datasets:

nvidia
/

Nemotron-SFT-Multilingual-v1

domain string	language string	license string	messages list	used_in list	uuid string
multilingual_code	de	cc-by-4.0	[ { "role": "user", "content": "Sie sind ein Geheimagent des Friedenskomitees Intelligenzcentrums, der sich einer geheimen Laboranlage auf dem Mars eingeschlichen hat, die von den Automatischen Kriminellen Maschinen betrieben wird. Ihre Mission besteht darin, ein vertrauliches Dokument zu sichern, das hinter ...	[ "super_v3" ]	53e28a72-625b-4d5c-91b1-c43b7c7fa668
multilingual_code	de	cc-by-4.0	[ { "role": "user", "content": "Betrachte die modifizierte Collatz-Folge, in der gerene Begriffe durch 2 geteilt (D-Schritt) und ungerene Begriffe durch (3x + 1)/2 transformiert werden (U-Schritt). Gegeben eine positive Ganzzahl n und eine Zeichenkette aus U/D-Schritten, finde die kleinste Ganzzahl x ≥ n, sod...	[ "super_v3" ]	d0469245-0996-43dd-854c-f648fb577545
multilingual_code	de	cc-by-4.0	[ { "role": "user", "content": "Sie erhalten zwei Strings A und B der Länge N und eine Konstante β > 0. Die Strings werden entlang eines linearen optischen Mediums ausgerichtet, wobei jeder Position i (0 ≤ i < N) ein radialer Abstand r_i = \|i - (N-1)/2\| vom Zentrum zugeordnet ist. Der Kostenfaktor für die Dur...	[ "super_v3" ]	dc9c7edc-eedc-427b-b86b-c0987e30798f
multilingual_code	de	cc-by-4.0	[{"role":"user","content":"Ashish besitzt eine Zeichenkette S der Länge N, die einem mehrdimensiona(...TRUNCATED)	[ "super_v3" ]	2cb51356-cbcf-4b51-98be-d5217cab2c03
multilingual_code	de	cc-by-4.0	[{"role":"user","content":"In einer hierarchischen federierten Lernarchitektur, die als binärer Bau(...TRUNCATED)	[ "super_v3" ]	bed38bc2-9cf5-4fb3-8ca0-e7c590700beb
multilingual_code	de	cc-by-4.0	[{"role":"user","content":"Chef hat das Quick Battle Royale-Spiel mit einem magnetischen Feld erweit(...TRUNCATED)	[ "super_v3" ]	215db1a3-7078-4f00-8f76-382eec525fdd
multilingual_code	de	cc-by-4.0	[{"role":"user","content":"Rohan justiert ein optisches Spektrometer mit diskreten spektralen Kanäl(...TRUNCATED)	[ "super_v3" ]	b4e4881d-dfde-45c0-b2c2-20cf3537dbb5
multilingual_code	de	cc-by-4.0	[{"role":"user","content":"Für genauen genomischen Editing über site-spezifische Rekombination mü(...TRUNCATED)	[ "super_v3" ]	45bc6bc9-da4a-4449-9e66-132119eba937
multilingual_code	de	cc-by-4.0	[{"role":"user","content":"In einem fortschrittlichen Apothekenbetrieb, der thermodynamische Prinzip(...TRUNCATED)	[ "super_v3" ]	c28f5971-a438-46f6-8b67-3e293e61069e
multilingual_code	de	cc-by-4.0	[{"role":"user","content":"Doremy baut eine Stadt auf einer Insel, die auch 100 Gefangene beherbergt(...TRUNCATED)	[ "super_v3" ]	e2e58500-d3a4-451a-80b7-c5e1e38e900f

End of preview.

Dataset Description:

Nemotron-Multilingual-v1 is a multilingual reasoning dataset made by translating a subsample of SFT data from Nemotron-Math-v2, Nemotron-Competitive-Programming-v1, and Nemotron-Science-v1 into to 6 languages (German, French, Japanese, German, Italian, Japanese, Chinese).
The original datasets were translated with Qwen2.5-14B-Instruct, then filtered with heuristics to remove translation failures and hallucinations. The STEM subsets are further post-edited with an LLM Qwen3-4B-Thinking-2507 to fix format mismatching problems.
We provide the prompt and the final answer in the target language, while the reasoning trace is kept in English. This is due to architectural decisions for Nemotron 3 series.

For details of dataset in each domain, please refer to the original data cards mentioned before.

This dataset is ready for commercial use.

Dataset Owner(s):

NVIDIA Corporation

Dataset Creation Date:

Created on: Jan 28, 2026 Last Modified on: Jan 28, 2026

License/Terms of Use:

This dataset is governed by the Creative Commons Attribution 4.0 International License (CC BY 4.0), except for the StackOverflow and MathGenSelect data, which is governed by the Creative Commons Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0)

Intended Usage:

This dataset is intended for post-training large language models with multilingual capabilities, with a special focus on improving model's capability of handling STEM, math and coding applications under multilingual setup.

Because all the examples are translated from English-sourced datasets, it is not intended to instill local or regional knowledge specific to where a specific language is spoken.

Dataset Characterization

Data Collection Method

[Hybrid]

Labeling Method

[Synthetic]

Dataset Format

Modality: Text
Format: JSONL
Structure: Text + Metadata

Dataset Quantification

The samples below represent the number of Q&A prompts, along with the reasoning trace in English.

Subset	Samples
code_de	133322
code_es	131578
code_fr	136045
code_it	143122
code_ja	126393
code_zh	154653
math_de	128846
math_es	102866
math_fr	115916
math_it	130388
math_ja	101820
math_zh	88594
stem_de	261205
stem_es	262353
stem_fr	264221
stem_it	269240
stem_ja	256624
stem_zh	258069
Total	3065255

Measurement of Total Data Storage: ~90GB

Reference(s):

You can find our recipe for translation here.

Ethical Considerations:

NVIDIA believes Trustworthy AI is a shared responsibility and we have NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal developer teams to ensure this dataset meets requirements for the relevant industry and use case and addresses unforeseen product misuse. Please report quality, risk, security vulnerabilities or NVIDIA AI Concerns here

Downloads last month: -

Collection including nvidia/Nemotron-SFT-Multilingual-v1

Nemotron-Post-Training-v3

Collection

Collection of datasets used in the post-training phase of Nemotron Nano and Super v3. • 26 items • Updated about 22 hours ago • 83