mlboydaisuke/Depth-Anything-V2-Small-ExecuTorch

depth_anything_v2_small — ExecuTorch

• Source: depth-anything/Depth-Anything-V2-Small-hf
• License: Apache-2.0
• Input: [[1, 3, 518, 518]] — RGB, ImageNet norm, 518x518
• Output: relative inverse depth [1,518,518]

Variants

All variants take and return fp32 tensors — swap the .pte file, keep your app code.

build	file	size (MB)	parity vs fp32 eager (worst corr)	Mac median (ms)*
fp32	depth_anything_v2_small_xnnpack_fp32.pte	99.0	1.000000	167.3
fp16	depth_anything_v2_small_xnnpack_fp16.pte	55.5	0.999992	289.7
Core ML (fp16, iOS)	depth_anything_v2_small_coreml_all.pte	50.2	0.999992	48.1

The Core ML build is the same graph lowered to Apple's Neural Engine instead of XNNPACK, which is CPU-only. Measured on an iPhone 17 Pro across seven models, it runs 3.5x to 13.9x faster (median 12x) at roughly half the file size — for example Depth-Anything-V2-Small at 500.8 ms against 42.7 ms, and MODNet at 81.7 ms against 5.9 ms. It computes in fp16 and is iOS-only; the XNNPACK files stay the portable option and are what runs on Android.

*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: 85.0 ms).

Builds that did not earn a slot

• int8 (dynamic) is not shipped: measured in the units that matter for this model — fraction of pixels within 1.25x of the fp32 depth: median 0.9941 over 10 real images, worst 0.9748.

Verification (executorch 1.4.0, torch 2.13.0)

Parity is measured against the fp32 eager model on random input; corr is the correlation over all elements of each output tensor.

output	shape	max_abs_diff	corr
0	[1, 518, 518]	4.768e-06	1.000000

XNNPACK delegate coverage (fp32): 73.4% (482/657 ops); ops left on the portable kernels: aten.expand_copy.default x49, aten.native_layer_norm.default x28, aten.mul.Scalar x24, aten.logical_not.default x24, aten.eq.Scalar x12, aten.full_like.default x12, aten.any.dim x12, aten.where.self x12, dim_order_ops._to_dim_order_copy.default x1, aten.squeeze_copy.dims x1

Conversion

torch.export -> to_edge_transform_and_lower(partitioner) -> .pte (conversion scripts: executorch-models)