Uni-NaVid for SatNav — Continue

This checkpoint is designed for use with SatNav, the continuous-state vision-and-language navigation platform for satellite maps. The SatNav Uni-NaVid integration uses it to predict short action sequences from a language instruction and a stream of RGB observations.

Model

Starting checkpoint: Jzzhang/Uni-NaVid
Vision encoder: EVA-CLIP
Training data: SatNav-v0.1 offline trajectories
Training: 1 epoch, learning rate 1e-5
Action window: up to 4 actions

Repository name

uninavid-satnav: Uni-NaVid adapted to SatNav
continue: starts from the Uni-NaVid checkpoint
1ep: trained for one epoch
lr1e-5: uses a learning rate of 1e-5

Reference Results

Evaluated in SatSim with the full split and a 500-step cap.

Split	Episodes	NE	OS	SR	SPL
`val_seen`	4,574	87.11	68.17	49.69	49.15
`val_unseen`	8,756	149.85	55.85	36.72	36.29

Usage

Use this checkpoint with the SatNav Uni-NaVid integration. It is also listed in the SatNav Baseline Model Zoo.

License

The checkpoint follows the applicable Llama 2 terms. Uni-NaVid, EVA-CLIP, SatNav, and other components retain their own licenses.

Downloads last month: 13

Inference Providers NEW

Video-Text-to-Text

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for Eku127/uninavid-satnav-continue-1ep-lr1e-5

Base model

Jzzhang/Uni-NaVid

Finetuned

(1)

this model

Collection including Eku127/uninavid-satnav-continue-1ep-lr1e-5

SatNav Baseline Model Zoo

Collection

Baseline checkpoints for SatNav vision-language navigation. This collection includes StreamVLN, OpenFly, UniNaVid, and NaVILA • 8 items • Updated Jul 4