A Partial Replication of MaskFormer in TensorFlow on TPUs for the TensorFlow Model Garden
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910427245445120 |
|---|---|
| author | Purohit, Vishal Jiang, Wenxin Ravikiran, Akshath R. Davis, James C. |
| author_facet | Purohit, Vishal Jiang, Wenxin Ravikiran, Akshath R. Davis, James C. |
| contents | This paper undertakes the task of replicating the MaskFormer model a universal image segmentation model originally developed using the PyTorch framework, within the TensorFlow ecosystem, specifically optimized for execution on Tensor Processing Units (TPUs). Our implementation exploits the modular constructs available within the TensorFlow Model Garden (TFMG), encompassing elements such as the data loader, training orchestrator, and various architectural components, tailored and adapted to meet the specifications of the MaskFormer model. We address key challenges encountered during the replication, non-convergence issues, slow training, adaptation of loss functions, and the integration of TPU-specific functionalities. We verify our reproduced implementation and present qualitative results on the COCO dataset. Although our implementation meets some of the objectives for end-to-end reproducibility, we encountered challenges in replicating the PyTorch version of MaskFormer in TensorFlow. This replication process is not straightforward and requires substantial engineering efforts. Specifically, it necessitates the customization of various components within the TFMG, alongside thorough verification and hyper-parameter tuning. The replication is available at: https://github.com/PurdueDualityLab/tf-maskformer/tree/main/official/projects/maskformer |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_18801 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | A Partial Replication of MaskFormer in TensorFlow on TPUs for the TensorFlow Model Garden Purohit, Vishal Jiang, Wenxin Ravikiran, Akshath R. Davis, James C. Computer Vision and Pattern Recognition Machine Learning Software Engineering This paper undertakes the task of replicating the MaskFormer model a universal image segmentation model originally developed using the PyTorch framework, within the TensorFlow ecosystem, specifically optimized for execution on Tensor Processing Units (TPUs). Our implementation exploits the modular constructs available within the TensorFlow Model Garden (TFMG), encompassing elements such as the data loader, training orchestrator, and various architectural components, tailored and adapted to meet the specifications of the MaskFormer model. We address key challenges encountered during the replication, non-convergence issues, slow training, adaptation of loss functions, and the integration of TPU-specific functionalities. We verify our reproduced implementation and present qualitative results on the COCO dataset. Although our implementation meets some of the objectives for end-to-end reproducibility, we encountered challenges in replicating the PyTorch version of MaskFormer in TensorFlow. This replication process is not straightforward and requires substantial engineering efforts. Specifically, it necessitates the customization of various components within the TFMG, alongside thorough verification and hyper-parameter tuning. The replication is available at: https://github.com/PurdueDualityLab/tf-maskformer/tree/main/official/projects/maskformer |
| title | A Partial Replication of MaskFormer in TensorFlow on TPUs for the TensorFlow Model Garden |
| topic | Computer Vision and Pattern Recognition Machine Learning Software Engineering |
| url | https://arxiv.org/abs/2404.18801 |