SwiftVLN for SatNav — Short-Term Only

Related Repositories

• SwiftVLN: training and evaluation code for these checkpoints.
• SatNav: satellite-image navigation environments, datasets, and evaluation tools.

This checkpoint is designed for SwiftVLN on SatNav, the continuous-state vision-and-language navigation benchmark over satellite imagery. It predicts short navigation-action sequences from an instruction and the current RGB window, without observations from earlier windows.

Model

• Starting checkpoint: Qwen/Qwen2.5-VL-3B-Instruct
• Training data: SatNav-v0.1 offline expert trajectories
• Training: 1 epoch, full-parameter fine-tuning, learning rate 2e-5
• Context: 32 current RGB frames and no long-term visual memory
• Prediction horizon: 4 actions

Repository name

• swiftvln-satnav: SwiftVLN trained and evaluated on SatNav
• 3b: Qwen2.5-VL 3B backbone
• 1ep: trained for one epoch
• f32s4: uses a 32-frame window and predicts four actions
• overlap0: uses non-overlapping training windows
• pf-h0-nomem: sets the number of history frames to zero
• pool-s2: retains the per-frame processor configuration used by the reference model
• noembed: uses no embedding enhancement

Reference Results

Results reported in the SatNav paper, memory-design ablation, row Short-term only.

SR, SPL, and OS are percentages. Changes in SR and OS are percentage points relative to the SwiftVLN reference model.

Usage

Use this checkpoint with the SwiftVLN evaluation guide. Keep the full repository name unchanged because SwiftVLN derives the evaluation configuration from it.

This checkpoint is the short-term-only ablation in the SwiftVLN SatNav Ablation Model Zoo.