Open Models Almanac
Methodspost-trainingsupervised fine-tuning

specific method · filed under post-training

Two-stage supervised fine-tuning

Uses two supervised fine-tuning stages, with the evidence describing multimodal capability development across stages.

The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See supervised fine-tuning for the methods filed beside it.

Also called two-stage SFT process, two-stage SFT pipeline, Two-stage supervised fine-tuning pipeline.

sources
4
models
2
lab adopt it
1
strongest
core

Documented in

Evidence

5 quotes from 4 sources, one card per model, each opening with its strongest.

How sources treat it
used 4core 1

A two-stage SFT pipeline trains the language and vision components

corepost trainingin Nemotron 3.5 Super VLNVIDIA
4 more from this model

Our two-stage SFT pipeline transforms the text-only base model into a vision-language model while jointly developing reasoning, chat, safety, long-context, and agentic capabilities.

usedpost trainingin Nemotron 3.5 SuperNVIDIA

we use a two-stage SFT process because learning vision and text capabilities simultaneously is challenging.

usedpost trainingin Nemotron 3.5 SuperNVIDIA

A two-stage supervised fine-tuning pipeline jointly trains the language and vision components

usedpost trainingin Nemotron 3.5 Super VLNVIDIA

A two-stage supervised fine-tuning pipeline jointly trains the language and vision components

usedpost trainingin Nemotron 3.5 Super VLNVIDIA

Our two-stage SFT pipeline transforms the text-only base model into a vision-language model while jointly developing reasoning, chat, safety, long-context, and agentic capabilities.

usedpost trainingin Nemotron 3.5 SuperNVIDIA
1 more from this model

we use a two-stage SFT process because learning vision and text capabilities simultaneously is challenging.

usedpost trainingin Nemotron 3.5 SuperNVIDIA
↑