mmseg CNN classics

• Original code: https://github.com/open-mmlab/mmsegmentation at c685fe6767c4
• Code licence: Apache-2.0 (permissive); files: LICENSE
• Weights licence: Apache-2.0
• Licence evidence: https://github.com/open-mmlab/mmsegmentation/blob/c685fe6767c4cadf6b051983ca6208f1b9d1ccb8/README.md#license
• Training data: ImageNet-1k (ILSVRC 2012 classification), ADE20K (SceneParsing, 150 classes)
• Kernel libraries: igev, cutlass_conv, cudnn_hacc, seg_conv, seg_up
• Variants: fcn-r101-d8-ade20k, pspnet-r101-d8-ade20k, deeplabv3-r101-d8-ade20k, deeplabv3plus-r101-d8-ade20k, hrnet-w48-fcn-ade20k
• Outputs: logits; label_set: ade20k-150
• Input size: any H x W >= 1, resized or padded inside the model
• Tested up to (portable, fp16, RTX 5060 Ti 16 GB): fcn-r101-d8-ade20k 3000x4000 (7.2 GiB peak allocated, 7.5 reserved); pspnet-r101-d8-ade20k 3000x4000 (7.2 GiB peak allocated, 7.5 reserved); deeplabv3-r101-d8-ade20k 3000x4000 (7.3 GiB peak allocated, 7.5 reserved); deeplabv3plus-r101-d8-ade20k 3000x4000 (7.2 GiB peak allocated, 7.5 reserved); hrnet-w48-fcn-ade20k 3000x4000 (7.2 GiB peak allocated, 7.5 reserved)
• hrnet-w48-fcn-ade20k 1024x2048, first call: portable + torch.compile 36 s (cold Inductor cache), TensorRT fp16 30 s (ONNX + engine).
• fcn-r101-d8-ade20k original 1024x2048: torch.autocast(fp16) around the forward
• pspnet-r101-d8-ade20k original 1024x2048: torch.autocast(fp16) around the forward
• deeplabv3-r101-d8-ade20k original 1024x2048: torch.autocast(fp16) around the forward
• deeplabv3plus-r101-d8-ade20k original 1024x2048: torch.autocast(fp16) around the forward
• hrnet-w48-fcn-ade20k original 1024x2048: torch.autocast(fp16) around the forward
• fcn-r101-d8-ade20k load / first call: original 4.72 s / 0.393 s (512x512), portable 0.53 s / 0.073 s, kernels 0.91† s / 0.018† s
• pspnet-r101-d8-ade20k load / first call: original 1.35 s / 0.092 s (512x512), portable 0.47 s / 0.082 s, kernels 0.77† s / 0.026† s
• deeplabv3-r101-d8-ade20k load / first call: original 1.64 s / 0.031 s (512x512), portable 0.73 s / 0.081 s, kernels 0.91† s / 0.021† s
• deeplabv3plus-r101-d8-ade20k load / first call: original 4.72 s / 0.662 s (512x512), portable 0.62 s / 0.095 s, kernels 0.75† s / 0.019† s
• hrnet-w48-fcn-ade20k load / first call: original 1.48 s / 0.079 s (512x512), portable 0.62 s / 0.071 s, kernels 0.85† s / 0.014† s
• † measured in the speed sweep after the portable load (or an earlier size) in the same process: file cache and kernels are warm, so the time is lower than a fresh start.
• Protocol, first call, 3 warm-up, peak reset, median of 3 rounds x median of 5 (tools/speed_check.py), synthetic CUDA input (bench_inputs); peak = max over the first call and the timed calls

Measured

Each image at its own size, argmax of the logits. Original = the original code named under Details, same checkpoint and protocol.

Measured on NVIDIA GeForce RTX 5060 Ti, torch 2.14.1+cu130, fastforks 939adfe8, 2026-11-06.

Parity

CPU, float32: max |diff| <= 0.002 vs the original. GPU, float16 (the default), portable and kernels: mean |diff| <= 0.04 and p99 <= 0.15 (absolute error of the output logits).