There is a newer version of the record available.

Published March 18, 2022 | Version v0.1-tpu-weights
Software Open

rwightman/pytorch-image-models: TPU VM Trained Weight release w/ PyTorch XLA

Creators

  • 1. Weights & Biases
  • 2. independent
  • 3. @huggingface
  • 4. MIT
  • 5. Kaggle Competition Master
  • 6. Neuro Event Labs Oy
  • 7. Indie developer
  • 8. University of Bern
  • 9. Mobility Technologies Co., Ltd.
  • 10. Xiamen University
  • 11. LG AI Research
  • 12. AI Lab, Naver Corp.
  • 13. @pytorch

Description

A wide range of mid-large sized models trained in PyTorch XLA on TPU VM instances. Demonstrating viability of the TPU + PyTorch combo for excellent image model results. All models trained w/ the bits_and_tpu branch of this codebase.

A big thanks to the TPU Research Cloud (https://sites.research.google/trc/about/) for the compute used in these experiments.

This set includes several novel weights, including EvoNorm-S RegNetZ (C/D timm variants) and ResNet-V2 model experiments, as well as custom pre-activation model variants of RegNet-Y (called RegNet-V) and Xception (Xception-P) models.

Many if not all of the included RegNet weights surpass original paper results by a wide margin and remain above other known results (e.g. recent torchvision updates) in ImageNet-1k validation and especially OOD test set / robustness performance and scaling to higher resolutions.

RegNets
  • regnety_040 - 82.3 @ 224, 82.96 @ 288
  • regnety_064 - 83.0 @ 224, 83.65 @ 288
  • regnety_080 - 83.17 @ 224, 83.86 @ 288
  • regnetv_040 - 82.44 @ 224, 83.18 @ 288 (timm pre-act)
  • regnetv_064 - 83.1 @ 224, 83.71 @ 288 (timm pre-act)
  • regnetz_040 - 83.67 @ 256, 84.25 @ 320
  • regnetz_040h - 83.77 @ 256, 84.5 @ 320 (w/ extra fc in head)
Alternative norm layers (no BN!)
  • resnetv2_50d_gn - 80.8 @ 224, 81.96 @ 288 (pre-act GroupNorm)
  • resnetv2_50d_evos 80.77 @ 224, 82.04 @ 288 (pre-act EvoNormS)
  • regnetz_c16_evos - 81.9 @ 256, 82.64 @ 320 (EvoNormS)
  • regnetz_d8_evos - 83.42 @ 256, 84.04 @ 320 (EvoNormS)
Xception redux
  • xception41p - 82 @ 299 (timm pre-act)
  • xception65 - 83.17 @ 299
  • xception65p - 83.14 @ 299 (timm pre-act)
ResNets (w/ SE and/or NeXT)
  • resnext101_64x4d - 82.46 @ 224, 83.16 @ 288
  • seresnext101_32x8d - 83.57 @ 224, 84.270 @ 288
  • resnetrs200 - 83.85 @ 256, 84.44 @ 320

Files

rwightman/pytorch-image-models-v0.1-tpu-weights.zip

Files (15.0 MB)

Additional details

Related works

Is supplement to
https://github.com/rwightman/pytorch-image-models/tree/v0.1-tpu-weights (URL)