Datasets:
Publish final MapPool predictions
Browse files- README.md +4 -5
- build-report.json +12 -1
- figures.jsonl +0 -0
- mappool_predictions.csv +0 -0
README.md
CHANGED
|
@@ -18,8 +18,7 @@ the application repository and submitted through pull requests.
|
|
| 18 |
- `media/figures/brgm-v2/`: 6,046 figures from the newer BRGM parsing batch.
|
| 19 |
- `figures.jsonl`: portable figure manifest with dimensions, content SHA-256, source
|
| 20 |
document, relative media path, and MapPool prediction metadata.
|
| 21 |
-
- `annotations.jsonl`:
|
| 22 |
-
`brgm-worklist-v1` and 156 for `dataset2-v1`.
|
| 23 |
- `mappool_predictions.csv`: portable map/not-map labels and decision margins.
|
| 24 |
- `build-report.json`: input/output counts, model provenance, and manifest checksum.
|
| 25 |
|
|
@@ -27,9 +26,9 @@ MapPool predictions are queue-prioritization hints, not ground truth. The classi
|
|
| 27 |
uses the `sraimund/MapPool` head with the OpenCLIP
|
| 28 |
`ViT-L-14-quickgelu/openai` backbone. Neither model weight is included here.
|
| 29 |
|
| 30 |
-
|
| 31 |
-
GCPs and geometry, contributor provenance, and timestamps. Annotation records
|
| 32 |
-
CC BY 4.0; that does not relicense the source figure media.
|
| 33 |
|
| 34 |
## Licensing and release status
|
| 35 |
|
|
|
|
| 18 |
- `media/figures/brgm-v2/`: 6,046 figures from the newer BRGM parsing batch.
|
| 19 |
- `figures.jsonl`: portable figure manifest with dimensions, content SHA-256, source
|
| 20 |
document, relative media path, and MapPool prediction metadata.
|
| 21 |
+
- `annotations.jsonl`: migrated human ground-truth annotations when published.
|
|
|
|
| 22 |
- `mappool_predictions.csv`: portable map/not-map labels and decision margins.
|
| 23 |
- `build-report.json`: input/output counts, model provenance, and manifest checksum.
|
| 24 |
|
|
|
|
| 26 |
uses the `sraimund/MapPool` head with the OpenCLIP
|
| 27 |
`ViT-L-14-quickgelu/openai` backbone. Neither model weight is included here.
|
| 28 |
|
| 29 |
+
Published human annotations include classification, map type, georeferencing status,
|
| 30 |
+
optional GCPs and geometry, contributor provenance, and timestamps. Annotation records
|
| 31 |
+
default to CC BY 4.0; that does not relicense the source figure media.
|
| 32 |
|
| 33 |
## Licensing and release status
|
| 34 |
|
build-report.json
CHANGED
|
@@ -6,5 +6,16 @@
|
|
| 6 |
},
|
| 7 |
"media_staged": true,
|
| 8 |
"brgm_source": "/scratch/msoric/OUTPUT_GEOREF/docling_brgm",
|
| 9 |
-
"dataset2_source": "/scratch/msoric/OUTPUT_GEOREF/parsing"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
}
|
|
|
|
| 6 |
},
|
| 7 |
"media_staged": true,
|
| 8 |
"brgm_source": "/scratch/msoric/OUTPUT_GEOREF/docling_brgm",
|
| 9 |
+
"dataset2_source": "/scratch/msoric/OUTPUT_GEOREF/parsing",
|
| 10 |
+
"detector": {
|
| 11 |
+
"counts": {
|
| 12 |
+
"map": 2541,
|
| 13 |
+
"not_map": 5017
|
| 14 |
+
},
|
| 15 |
+
"model_id": "sraimund/MapPool:model.pkl",
|
| 16 |
+
"backbone": "ViT-L-14-quickgelu/openai",
|
| 17 |
+
"model_sha256": "b65d6d468be49c9bad63ca9f2899fd0946998d68ec72bbd7c9592b98150b7cd3",
|
| 18 |
+
"scores_file": "mappool_scores.csv"
|
| 19 |
+
},
|
| 20 |
+
"manifest_sha256": "1c6ae66d17f19971b3d31461d25f983ffac780eaed488a5867a71a52abc67d05"
|
| 21 |
}
|
figures.jsonl
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
mappool_predictions.csv
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|