fireviewer/litert-models

FireViewer · LiteRT conversion archive

Quick card

Field	Reference
Use	Inspect an archived LiteRT conversion and its precise output contract.
Expected input	Matching image/ROI tensors, labels, coordinate transforms and runtime contract.
Example	Invoke the documented inference signature; adapt reviewed targets only for historical graphs exposing genuine training signatures.
Limits	Export and SDK receipts concern these converted graphs. Native research-model evaluation remains separate.
Contract	Conversion guide · Per-variant cards

Ten retained conversion packages: six historical variants, including five trainable heads/output adapters, and four FP32 inference-only exports. The original artifacts, technical contracts and SDK receipts are preserved.

Conversion guide · Usage guide · Conversion evidence · Update log

Retained FP32 exports

Package	Input	Conversion contract
fireviewer_dinov2_small_fire_associated_smoke_v25_fp32	Exact tensor signature in contract	runtime_contract.json
fireviewer_dinov2_small_visible_flame_monopoint_v25_fp32	Exact tensor signature in contract	runtime_contract.json
fireviewer_rfdetr_medium_probable_fire_v25_fp32	Exact tensor signature in contract	runtime_contract.json
fireviewer_rfdetr_medium_visible_flame_v25_fp32	Exact tensor signature in contract	runtime_contract.json

These four exports expose inference only. Their tensor-parity and frozen operating-point checks are conversion evidence; the native cascade diagnostic is documented in its own research presentation.

Catalogue and recorded qualification

Variant	Synthetic host learning	Standalone Android SDK	Vision Dataset Studio
D-FINE M strict v1, learning	PASS	Not qualified	Not qualified
RT-DETRv2 R50, learning	PASS	Not qualified	Not qualified
YOLO11 M strict v1, inference	N/A	Not qualified	Not qualified
YOLO11 M strict v1, learning	PASS	PASS on emulator	Not qualified
RF-DETR Medium 1.10, learning	PASS	Not qualified	Not qualified
DINOv3 pointing pilot v1, learning	PASS	Not qualified	Not qualified

The retained delivery receipt covers five learning conversions. The YOLO inference-only graph is a sixth delivered variant. No SegFormer graph is delivered because the optional source was unavailable. Detection adapters cannot create proposals absent from the frozen detector.

The six historical variants remain unqualified inside Vision Dataset Studio. The YOLO learning SDK pass does not qualify other models, application integration or a physical ARM phone. For those six historical variants, task accuracy, generalization and forgetting are unmeasured.

Download and integrate

Choose a variant and read its README, runtime contract, artifact manifest and source notices. Retrieve all matching files at the same immutable revision:

from huggingface_hub import hf_hub_download

contract = hf_hub_download(
    repo_id="fireviewer/litert-models",
    revision="bdcd483d5c5aadeff1b531a35b536f3530906e18",
    filename="models/fireviewer_yolo11m_strict_v1_learning/runtime_contract.json",
)

Apply the exact normalization, tensor layout, labels and coordinate transform. Check inference before learning, train only with reviewed targets, keep the original weights, and store checkpoints separately. Evaluate adapted candidates before activation. See the complete usage guide.

SDK and runtime contract

The Kotlin SDK exposes infer, train, save and restore. Checkpoints preserve parameters, SGD momentum and step counters; receipts check model/label identity and file integrity. The integrating application selects the active checkpoint.

YOLO supports image dimensions that are multiples of 32. Other graphs resize to their internal grid. The historical SDK uses TensorFlow Lite and Select TF Ops 2.16.1. Vision Dataset Studio's separate 16 KB Flex fix does not retroactively qualify this SDK. Conversion sources.

Rights and responsible use

Source weights, datasets and software retain their own licences and restrictions. Read each variant's notices; download access alone grants no commercial-use or redistribution permission. In particular, YOLO upstream obligations and the DINOv3 pilot's retained release policy remain applicable.

Synthetic checks establish parameter updates and persistence, not field accuracy or reliable alerts. These models must not be used alone to confirm a fire, trigger an emergency alert, order an evacuation or direct emergency response.

Documentation and revision

Historical catalogue documentation updated in English on 5 October 2026, from the public repository at bdcd483d5c5aadeff1b531a35b536f3530906e18 and its retained reports. This update checks documentation, repository metadata and small evidence files; it does not rerun training, inference, dataset payload verification or device qualification. Historical receipts keep their original dates, revisions and scope. Earlier README hashes in artifact manifests refer to those earlier releases.

The association FIRE-VIEWER provides administrative and financial stewardship for resources under its control. Technical governance is maintainer-led. Upstream, source-specific and pre-association rights remain separate; repository placement does not transfer ownership or grant additional rights.

Project documentation · Institutional website. Contact: contact@fire-viewer.fr.

Likes are always welcome! ❤️ If these resources help you, a little heart makes our day and helps others discover the work. Thank you!