Hugging Face's logo

adithya-balaji
/
rfdetr-cpp-base

Object Detection
GGUF
rfdetr.cpp
rfdetr
ggml
cpp-inference
Model card Files
xet
 Community
rfdetr-cpp-base / README.md
adithya-balaji's picture
adithya-balaji
Add/update model card for base
21859cd verified
preview code
|
Raw
History Blame Contribute Delete
5.33 kB
metadata 
license: apache-2.0
library_name: rfdetr.cpp
tags:
  - object-detection
  - rfdetr
  - gguf
  - ggml
  - cpp-inference
pipeline_tag: object-detection
base_model: roboflow/rfdetr

RF-DETR Base — GGUF for rfdetr.cpp

GGUF-format weights of Roboflow RF-DETR Base (detection variant) for use with rfdetr.cpp, a C++/ggml implementation that matches the upstream PyTorch model on CPU.

This repo contains all four standard quantizations of this variant. F16 is the recommended default — same accuracy as F32, 1.85× smaller, and typically the fastest on modern CPUs thanks to ggml's F32×F16 matmul fast path.

Available files

File Quant Size (MB) Recall @ IoU 0.5 Recall @ IoU 0.95 Mean |Δscore| Latency (median ms, T=8)
rfdetr-base-f32.gguf F32 119.2 1.0000 1.0000 0.0012 144.0
rfdetr-base-f16.gguf ← recommended F16 64.2 1.0000 1.0000 0.0014 136.1
rfdetr-base-q8_0.gguf Q8_0 38.5 0.9881 0.9881 0.0032 145.7
rfdetr-base-q4_K.gguf Q4_K 31.5 0.9542 0.8837 0.0196 166.3

All accuracy numbers above are computed against the upstream PyTorch reference (rfdetr 1.9.0) on 7 images (000000000139.jpg, 000000000632.jpg, 000000039769.jpg, 000000087038.jpg, 000000252219.jpg, 000000397133.jpg, bus.jpg) at threshold 0.5. Latency is measured separately with rfdetr-cli bench (8 iters + 3 warmup) at T=8 threads on a single Intel Core i7-12800HX, on tests/fixtures/ci/test_image.jpg.

Architecture

  • Backbone: DINOv2-small
  • Input resolution: 560×560
  • Patch size: 14
  • Decoder layers: 3
  • Object queries: 300
  • Task: object detection (boxes only)

Quantization notes

  • F32 — full-precision reference, ~120 MB. Bit-exact PyTorch parity.
  • F16 — matmul-multiplicand weights only; LayerNorms, conv kernels, embeddings, biases, and layer-scale gammas stay F32. Lossless on this model and consistently the fastest variant on CPU.
  • Q8_0 — best size/accuracy tradeoff under F16; ~3× smaller than F32 with effectively identical detections.
  • Q4_K — smallest practical quant. Rows with ne[0] % 256 != 0 (the decoder's 128-dim MLP halves, 60 tensors) silently fall back to Q8_0 per ggml's quantizer logic — net compression is still ~3.8× over F32. Use only when the size budget is tight; expect a measurable Recall@0.95 drop relative to F16/Q8_0 (see file table above).

Compatibility

These GGUFs stamp rfdetr.preprocess.resize_mode = "bilinear_no_antialias", matching RF-DETR 1.9's antialias-free float bilinear resize (align_corners=false, half-pixel coordinates, no intermediate uint8 rounding). rf-detr.cpp treats this key as optional: GGUFs that predate this metadata (no resize_mode key) keep using the legacy stb-based resize path, so older files continue to produce their original outputs unchanged. An unrecognized resize_mode value is rejected rather than guessed.

Keypoint-preview inference is not supported. rf-detr.cpp does not implement the keypoint output head; this repository only serves box detection outputs.

Usage 

# 1. Clone + build rfdetr.cpp
git clone https://github.com/adithyab94/rf-detr.cpp
cd rf-detr.cpp
cmake -B build -DRFDETR_BUILD_CLI=ON && cmake --build build -j

# 2. Download a quant (F16 recommended)
hf download adithya-balaji/rfdetr-cpp-base rfdetr-base-f16.gguf --local-dir models/

# 3. Run detection
build/bin/rfdetr-cli detect \
    --model models/rfdetr-base-f16.gguf \
    --input my_image.jpg \
    --threshold 0.5 --threads 8 \
    --output detections.json

Accuracy methodology

All accuracy metrics are computed against the upstream PyTorch reference (rfdetr 1.9.0) on 7 images (000000000139.jpg, 000000000632.jpg, 000000039769.jpg, 000000087038.jpg, 000000252219.jpg, 000000397133.jpg, bus.jpg) at threshold 0.5. Each detection match uses greedy Hungarian-style assignment by IoU (≥ 0.5 lenient, ≥ 0.95 strict) with class equality required.

See BENCHMARK.md and benchmarks/results/accuracy_sweep.json for the full sweep across the (variant × quant) cells.

Provenance

  • Source project: Roboflow RF-DETR
  • Upstream package: rfdetr==1.9.0
  • Converted with rfdetr.cpp at commit fbef9387bed3
  • Checkpoint: official pretrained rfdetr-base weights (downloaded by the rfdetr package on first use)

Checksums (SHA-256)

Also available as SHA256SUMS in this repo.

1d81763ef7ae83d2d0b1049a853b959c435aa8d3efe0944dcde4fd7151db0691  rfdetr-base-f32.gguf
3426bc8e6bdb714c4f42355695b469a483f2f79f6b8a569a04ac00b34555369a  rfdetr-base-f16.gguf
a6f36a8c251be2a57e6756cf3a0b79ce041007d0506e43d797d35f260a6ed746  rfdetr-base-q8_0.gguf
94e70960144b443ff05f4e76aa1bc2e7ca8cbf1c8dde50754e15712f2f88873f  rfdetr-base-q4_K.gguf

License

Apache-2.0 — matches the upstream rfdetr license.