samadon1/GalamseyWatch

Two-layer agentic Earth Observation on satellite-class compute. Fine-tuned LFM2.5-VL perception + tool-calling agent over Sentinel-2 imagery. Galamsey detection in Ghana as the worked example. # GalamseyWatch Two artifacts in one repo, both built around a fine-tuned vision-language model that detects illegal small-scale gold mining (galamsey) in Sentinel-2 imagery: - **`app/`**: a browser-native dashboard that runs the VLM via WebGPU on an enforcement officer's laptop. Live at galamseywatch.vercel.app. Click the map; nothing leaves the device. - **`orchestrator/`**: a FastAPI service that runs a two-layer agentic-EO pipeline (VLM perception + tool-calling LLM policy) over a simulated satellite pass and decides, per tile, what's worth downlinking. The contribution of the orchestrator isn't the galamsey detector. That's the worked example. The contribution is the **architecture**: a four-interface contract (`VLMProvider`, `AgentPolicy`, `ImagerySource`, `Task`) that any fork can reskin in a single day for wildfire detection, illegal fishing, oil spills, etc. ## Latest: unified VLM replaces the two-layer pipeline A follow-up result from May 2026: a single fine-tuned **450M LFM2.5-VL** picks the action directly, removing the description-string bottleneck between the perception VLM and the LFM2 policy. A multitask SFT mixture lets one weight set serve all three jobs (action policy, grounding, scene description). | System | Total params | 99-tile action-match accuracy | |---|---:|---:| | Two-layer (bare) | 3.05 B | 65.7 % | | Two-layer (rich-context) | 3.05 B | 63.6 % | | Unified v3 (action-only LoRA) (`samwell/galamsey-unified-v3`) | 450 M | 76.8 % | | **Unified v4.1 (multitask LoRA)** (`samwell/galamsey-unified-v4-1`) | **450 M** | **77.8 %** | **+12.1 pp over the strongest baseline at 6.8× fewer parameters**, with grounding mIoU and description BLEU both within noise of the specialist perception model. The win comes from three design choices: (a) the action LoRA stacks on top of the `samwell/galamsey-v9-e3` perception fine-tune (frees capacity for action selection), (b) the assistant target for action examples is action-only (concentrates LM loss on …