All variants take and return fp32 tensors — swap the
.pte
file, keep your app code.
precision
file
size (MB)
parity vs fp32 eager (worst corr)
Mac median (ms)*
fp32
yolox_s_xnnpack_fp32.pte
35.9
1.000000
25.0
int8
yolox_s_xnnpack_int8.pte
9.2
0.999800
77.3
*Mac arm64, single process, median of 10 — a reference point for relative cost
only, not a device number (torch eager fp32 on the same machine: 38.6 ms).
Checked in the task's own units
Correlation is a first filter. These are the numbers that decide:
int8
— measured in the units that matter for this model — fraction of post-NMS detections matched at IoU 0.5 and same class: 0.938 of the fp32 build's detections are matched (90 of 96 across 10 images), worst single image 0.571.
Precisions that did not earn a slot
fp16 is not shipped
: it comes out at 101% of the fp32 file (36.1 MB vs 35.9 MB), so it buys nothing. XNNPACK serializes convolution weights as fp32 no matter what dtype the graph carries, so on a conv-heavy model fp16 saves no disk and only adds cast operations. Reach for int8 here, not fp16.
Verification (executorch 1.4.0, torch 2.13.0)
Parity is measured against the fp32 eager model on real image input;
corr
is
the correlation over all elements of each output tensor.
YOLOX-s-ExecuTorch huggingface.co is an AI model on huggingface.co that provides YOLOX-s-ExecuTorch's model effect (), which can be used instantly with this mlboydaisuke YOLOX-s-ExecuTorch model. huggingface.co supports a free trial of the YOLOX-s-ExecuTorch model, and also provides paid use of the YOLOX-s-ExecuTorch. Support call YOLOX-s-ExecuTorch model through api, including Node.js, Python, http.
YOLOX-s-ExecuTorch huggingface.co is an online trial and call api platform, which integrates YOLOX-s-ExecuTorch's modeling effects, including api services, and provides a free online trial of YOLOX-s-ExecuTorch, you can try YOLOX-s-ExecuTorch online for free by clicking the link below.
mlboydaisuke YOLOX-s-ExecuTorch online free url in huggingface.co:
YOLOX-s-ExecuTorch is an open source model from GitHub that offers a free installation service, and any user can find YOLOX-s-ExecuTorch on GitHub to install. At the same time, huggingface.co provides the effect of YOLOX-s-ExecuTorch install, users can directly use YOLOX-s-ExecuTorch installed effect in huggingface.co for debugging and trial. It also supports api for free installation.