Download README.md from EdyVision/foliage-hunter-experiments: direct link, hf CLI and curl.
- Browser
- Download file 3.85 kB
-
https://huggingface.co/datasets/EdyVision/foliage-hunter-experiments/resolve/main/README.md
- Command line
-
hf download hf://datasets/EdyVision/foliage-hunter-experiments/README.md
-
curl -L -o README.md https://huggingface.co/datasets/EdyVision/foliage-hunter-experiments/resolve/main/README.md
license: cc-by-4.0
task_categories:
- image-classification
pretty_name: Foliage Hunter Experiments
tags:
- inaturalist
- plant-identification
- foliage
- colorado
- evaluation
- gemma
Foliage Hunter Experiments
Held-out methodology, split manifests, and metrics for the
EdyVision/foliage-hunter-gemma3-12b
LoRA adapter (adapters/colorado_v1), the model behind
Foliage Hunter.
- Project repo, app, training and serving code: https://github.com/EdyVision/foliage-hunter
- Adapter: https://huggingface.co/EdyVision/foliage-hunter-gemma3-12b
- Base model: https://huggingface.co/google/gemma-3-12b-it
The CC BY 4.0 license on this dataset covers the metadata, manifests, and results published here. It does not relicense the underlying iNaturalist photos, which keep their own per-photo licenses (see below).
Important: no image bytes
This dataset publishes metadata only (observation IDs, licenses, species,
split membership, eval predictions). The photos include NC and ND material and
are not redistributed. Rebuild them with
training/scripts/build_inat_corpus.py
from the repo against colorado_v1/*_manifest.jsonl.
Layout
colorado_v1/
METHODOLOGY.json # train/eval protocol
train_manifest.jsonl # 1152 rows (observation_id, species, license, ...)
eval_manifest.jsonl # 288 rows
eval_results.json # per-image generate-and-match outcomes
colorado_priority_species.json
manifest.json
Manifest row fields: observation_id, species, scientific, tier,
inat_taxon_id, license, image_relpath.
Held-out result
| Metric | Value |
|---|---|
| n | 288 |
| hits | 245 |
| accuracy | 85.1% |
| matching | case-insensitive exact / substring on common name |
Stratification: every 5th image per species to eval (~20%).
Source data and attribution
Every photo referenced here is a research-grade observation photo contributed
by an iNaturalist user and retrieved through the public
iNaturalist API. 1,440 photos from 729
observations across 18 species (80 photos per species; Colorado place_id=34
filter for the local tiers). Species list in
colorado_v1/colorado_priority_species.json.
Per-photo licenses in the corpus:
| License | Photos |
|---|---|
| CC BY-NC | 1016 |
| CC BY | 301 |
| CC0 | 47 |
| CC BY-NC-ND | 38 |
| CC BY-NC-SA | 28 |
| CC BY-ND | 6 |
| CC BY-SA | 4 |
The photos were used for private model training and evaluation only. To credit
an individual photographer, open
https://www.inaturalist.org/observations/<observation_id> for any row in the
manifests; the observer's name and the photo license are shown there.
Thank you to the iNaturalist observers whose photos taught the model what a Colorado leaf looks like.
Citation
@misc{inaturalist2026,
title = {iNaturalist research-grade observations},
author = {{iNaturalist contributors}},
year = {2026},
howpublished = {\url{https://www.inaturalist.org}},
note = {Retrieved via \url{https://api.inaturalist.org/v1/observations}, October 2026. Observation IDs listed in EdyVision/foliage-hunter-experiments}
}
@misc{rosado2026foliagehunter,
title = {Foliage Hunter: a Colorado-tuned Gemma 3 LoRA leaf guide},
author = {Rosado, Eidan},
year = {2026},
howpublished = {\url{https://github.com/EdyVision/foliage-hunter}},
note = {Adapter: \url{https://huggingface.co/EdyVision/foliage-hunter-gemma3-12b}; experiments: \url{https://huggingface.co/datasets/EdyVision/foliage-hunter-experiments}}
}