HENet-tinyE
HENet uses depthwise conv to build lightweight blocks with LayerScale for stable training; the tiny-E configuration has widths 48/96/192/384 and depths 3/3/8/6, with parameter count in the tiny range.
Deployment Metrics
Model Parameters
| Model | Model Input | Backbone | Neck | Model Output |
|---|---|---|---|---|
| HENet-tinye | 1x3x224x224 |
HENet | — | classification logits (B,1000) |
Accuracy Metrics
| March | Metric | float | calibration | qat | hbm |
|---|---|---|---|---|---|
| J6M | Accuracy | 0.7762 | 0.7693 | — | 0.7719 |
| TopKAccuracy(5) | 0.9372 | — | — | — |
Data tested with
march = March.NASH_M(J6M); this task has no QAT stage (—in the qat column).HEAL versions: heal 0.0.2 / hbdk4-compiler 4.11.11 / horizon_plugin_pytorch 3.3.10.
Performance Metrics
Performance test methodology: FPS is measured with 8 threads on a single core; Latency is measured with single core, single thread; Memory is peak DDR usage.
| March | latency (ms) | fps | Memory Usage |
|---|---|---|---|
| J6M | 0.50 | 3667.03 | 12.70 |
| J6P | 0.40 | 11466.49 | 12.90 |
| J6B | 1.13 | 1616.56 | 10.00 |
Model Overview
Core Design
HENet uses depthwise conv to build lightweight blocks with LayerScale for stable training; the tiny-E configuration has widths 48/96/192/384 and depths 3/3/8/6, with parameter count in the tiny range.
- Task type: Image Classification.
- backbone: HENet (tiny-E configuration,
depth/block_nums=[3,3,8,6],width/embed_dims=[48,96,192,384],block_cls=["DWCB","GroupDWCB","AltDWCB","DWCB"], GELU activation, LayerScale, S2D downsampling,num_classes=1000), uses depthwise conv to build lightweight blocks with LayerScale for stable training. - neck: — (HENet has a built-in fully connected classification head; no separate neck).
- classification head: HENet built-in fully connected classification head (
include_top=True,feature_mix_channel=1024). - Loss function:
SoftTargetCrossEntropy(soft-label cross-entropy with mixup). - Model input: single RGB image at resolution
224 × 224(1x3x224x224). - Model output: 1000-class prediction logits; argmax gives the predicted class.
Official Repo and Paper
HybridEfficient Network is an efficient backbone designed by Horizon Robotics for the Journey series chips.
Reference
For more J6 chip deployment details, see https://developer.horizon.auto/blog/14094