Spaces:
Configuration error
Configuration error
|
Download docs/RESOURCE_INDEX.md from fireviewer/README: direct link, hf CLI and curl.
- Browser
- Download file 4.27 kB
-
https://huggingface.co/spaces/fireviewer/README/resolve/main/docs/RESOURCE_INDEX.md
- Command line
-
hf download hf://spaces/fireviewer/README/docs/RESOURCE_INDEX.md
-
curl -L -o RESOURCE_INDEX.md https://huggingface.co/spaces/fireviewer/README/resolve/main/docs/RESOURCE_INDEX.md
4.27 kB
| # FireViewer · Resource catalogue | |
| [Home](../README.md) · [Choose a resource](CHOOSING_VARIANTS.md) · [Update log](../CHANGELOG.md) | |
| Public inventory after model cleanup: **7 model repositories and 6 datasets**. Entries are ordered by repository creation within each section; documentation commits do not define release chronology. Scores retain their own protocols and are not a shared benchmark ranking. | |
| ## Datasets | |
| | Resource | Role and recorded state | | |
| |---|---| | |
| | [fireviewer_probable_fire_v3_native](https://huggingface.co/datasets/fireviewer/fireviewer_probable_fire_v3_native) | 103,244 native images; one common proposal class, separate from the two-class corpus. | | |
| | [fireviewer_fire_smoke_flame_v3_native](https://huggingface.co/datasets/fireviewer/fireviewer_fire_smoke_flame_v3_native) | 130,444 images and 160,159 boxes; flame/smoke plus six image-level context categories. | | |
| | [v2.5-box-precise-and-multi-source-suplement](https://huggingface.co/datasets/fireviewer/v2.5-box-precise-and-multi-source-suplement) | Historical v2.5 supplement with two alternative annotation views over the same 64,061 images. | | |
| | [2.5deprecied](https://huggingface.co/datasets/fireviewer/2.5deprecied) | Deprecated mixed-task archive; task populations, original annotations and provenance retained. | | |
| | [simple-measured-scenes-v1](https://huggingface.co/datasets/fireviewer/simple-measured-scenes-v1) | Three geographic tile catalogues; native CRS, local coordinate origins and asset contracts. | | |
| | [fire-smoke-detection-corpus-v1](https://huggingface.co/datasets/fireviewer/fire-smoke-detection-corpus-v1) | Historical v1 corpus, 87,568 current rows, with source-specific annotation and image-rights records. | | |
| ## Models and conversion archives | |
| | Resource | Role and recorded state | | |
| |---|---| | |
| | [yolo26s-flame-smoke-l4](https://huggingface.co/fireviewer/yolo26s-flame-smoke-l4) | Smaller native flame/smoke detector; best recorded validation AP50–95 35.8649% at E20. Distinct architecture, not a measured device-performance claim. | | |
| | [yolo11m-flame-smoke-resume-a100](https://huggingface.co/fireviewer/yolo11m-flame-smoke-resume-a100) | Retained improved YOLO11M detector; best recorded validation AP50–95 39.1811% at E46. | | |
| | [yolo11m-fire-smoke](https://huggingface.co/fireviewer/yolo11m-fire-smoke) | Historical 42-epoch source checkpoint used by the Cadryl conversion. Distinct weights retained for reproducibility. | | |
| | [convnext-tiny-alert-categories-v2](https://huggingface.co/fireviewer/convnext-tiny-alert-categories-v2) | Six image-level visual contexts; validation macro F1 76.92%, accuracy 87.72%. | | |
| | [dinov2-small-fire-associated-smoke-v2.5](https://huggingface.co/fireviewer/dinov2-small-fire-associated-smoke-v2.5) | Historical binary smoke-compatibility classifier with a calibrated abstention interval. | | |
| | [litert-models](https://huggingface.co/fireviewer/litert-models) | Ten distinct historical conversion packages; per-variant contracts, numerical checks and device limitations. | | |
| | [rtdetr-v2-r50-fire-smoke](https://huggingface.co/fireviewer/rtdetr-v2-r50-fire-smoke) | Historical visible-smoke/flame comparator; independent benchmark and specific checkpoint terms. | | |
| ## Central Cadryl packages | |
| - [YOLO11M · native-output adaptation](https://huggingface.co/unicornwhodev/lite-rt_models_cadrylpro/tree/main/models/fireviewer_yolo11m_fire_smoke_928_trainable). | |
| - [ConvNeXt-Tiny · final-classifier adaptation](https://huggingface.co/unicornwhodev/lite-rt_models_cadrylpro/tree/main/models/fireviewer_convnext_tiny_alerts_920_trainable). | |
| These links go directly to the preserved central packages; the redundant redirect-only model repositories have been removed. | |
| ## Historical evidence and restricted resources | |
| The [v2.5 cascade](CASCADE_V25.md), [v2.5 dataset guide](DATASET_V25.md) and [680-image diagnostic](../reports/cascade-v25-680/REPORT.md) retain their historical scopes and revisions. Some resources they cite are private or absent from the current public catalogue. Historical availability does not establish present public access. | |
| Private research, calibration and review resources retain their access restrictions and are not detailed in this public index. The organization card is documentation, not a live inference demo. | |