TR-HASH-Vision-v8-2M-COCO

TR-HASH Vision v8 Nano — COCO

Final checkpoint from the first 50-epoch COCO pretraining run of the corrected 2.53M-parameter TR-HASH Vision v8 architecture.

TR-HASH Vision is a compact anchor-free detector with a hierarchical vision tower. Spatial-token identities route activations through two narrow experts among eight, while a wider shared SwiGLU path remains active for every token. The model was trained from random initialization without an external detector or classification backbone.

Architecture

Evaluation

Official COCO evaluation on the 5,000-image val2017 split at 640 px. The checkpoint was selected at the final epoch by official COCO mAP50-95.

<p align="center"> <img alt="COCO AP50-95 16.59" src="https://img.shields.io/badge/COCO_AP50--95-16.59-06b6d4?style=for-the-badge"> <img alt="COCO AP50 27.81" src="https://img.shields.io/badge/COCO_AP50-27.81-6366f1?style=for-the-badge"> <img alt="COCO AR100 34.35" src="https://img.shields.io/badge/COCO_AR100-34.35-8b5cf6?style=for-the-badge"> <img alt="2.53 million parameters" src="https://img.shields.io/badge/Parameters-2.53M-111827?style=for-the-badge"> </p>

Example detection

<p align="center"> <img alt="TR-HASH Vision v8 detecting people and a bus in a street scene" src="./assets/example-detection.png" width="640"> </p> <p align="center"><em>Qualitative O2M inference example from the published checkpoint. This image is not part of the COCO evaluation set.</em></p>

Plain-language context: the final O2M checkpoint reaches 16.59 COCO AP, or approximately 52% of the 32.2 AP YOLO26 reference used for this comparison. TR-HASH reached this point in one 50-epoch run from random initialization. This is a progress indicator rather than a strict apples-to-apples efficiency claim: training recipes, initialization and compute budgets differ.

<details open> <summary><strong>O2M + NMS — recommended branch</strong></summary> </details> <details> <summary><strong>One-to-one NMS-free — experimental branch</strong></summary> </details>

The O2M branch is the recommended inference path. NMS-free reaches similar recall but still trails in score calibration and average precision. Expand the panels above for branch-level results; the raw evaluator outputs remain in validation.json and validation_nms_free.json.

Inference

from PIL import Image
import torch

from complexity.generative.detection import (
    load_detector_from_hub,
    preprocess_detector_image,
    restore_detector_boxes,
)

device = "cuda" if torch.cuda.is_available() else "cpu"
model = load_detector_from_hub(
    "AETHORIA-AI/TR-HASH-Vision-v8-2M-COCO",
    device=device,
)
pixels, metadata = preprocess_detector_image(
    Image.open("image.jpg"), model.config.image_size
)

with torch.inference_mode():
    prediction = model.predict(
        pixels[None].to(device),
        nms_free=False,
        confidence_threshold=0.19,
    )[0]
prediction["boxes"] = restore_detector_boxes(
    prediction["boxes"].cpu(), metadata
)

Class IDs are listed in class_names.json. The reported O2M metrics use the official COCO evaluator; validation.json contains the complete result.

Files

• ema.safetensors: recommended EMA detector weights;
• model.safetensors: non-EMA detector weights;
• training_state.pt: optimizer, scheduler and exact training state;
• tower.safetensors and ema_tower.safetensors: standalone vision tower;
• config.json: complete v8 architecture configuration;
• validation.json: official O2M + NMS COCO metrics;
• validation_nms_free.json: official NMS-free COCO metrics;
• metrics.csv: complete pretraining and validation history;
• class_names.json: the 80 COCO classes.

Both inference branches are contained in the same weights. The two metric files record their independent evaluations.

Training

• Dataset: COCO 2017
• Initialization: random
• Epochs: 50
• Optimizer: MuSGD
• Precision: BF16
• EMA: enabled
• External pretrained weights: none

Supervised clean-image refinement is planned as a separate follow-up phase and is not included in the metrics above.

Limitations

This is a research checkpoint, not a production release. Accuracy remains below mature production detectors, and the NMS-free branch requires further calibration. Validate accuracy, latency and failure modes on your own target domain before deployment.