Juvenile se[y]mour[...]iamorphs, temnospondyls, and others have external gills and were fully aquatic: "tadpoles" in the broad sense.
Of course there's a story behind this.The plesiomorphic state is to have a gradual ontogeny from a salamander-larva-like larva to a more or less terrestrial adult. In lissamphibians, many of the events of ontogeny -- loss of gills, closure of gill slits, appearance of many skull bones, dramatic restructuring of the palate -- happen at the same time (that's called metamorphosis), and the frogs took this to the next level after greatly extending and elaborating the earliest larval stage* at the expense of all others, concentrating even more events (appearance of more skull bones, sexual organs**, lungs... and the loss of the tail) in metamorphosis.
Amphibamid temnospondyls concentrated some of these events. That's the only known case of even partial metamorphosis outside Lissamphibia, though I must caution that the ontogeny of most limbed vertebrates other than amphibamids, a handful of other temnospondyls like *Sclerocephalus*, the seymouriamorph *Discosauriscus*, lissamphibians and amniotes is very poorly known. This holds all the way down to *Eusthenopteron*, of which there are long and detailed growth series, but the smallest individuals already have a completely ossified skull...
People do talk about "metamorphosis" in *Discosauriscus*, but this is arbitrarily defined as the disappearance of the external gills; no other event happens at the same time.
* Tadpoles have for instance taken the embryonic cartilage skeleton of the head and run with it. They're also mostly herbivorous, while postmetamorphic frogs are all carnivorous (in the broad sense); this means the gut changes drastically in metamorphosis. ** That must be why there are so many salamanders without metamorphosis or with incomplete metamorphosis, but no frogs without complete metamorphosis.