open-mmlab/mmsegmentation

[KeyError: 'PALETTE'] while loading model after retraining

Open

#2,039 opened on Sep 7, 2022

View on GitHub
 (5 comments) (0 reactions) (1 assignee)Python (2,386 forks)batch import
help wanted

Repository metrics

Stars
 (6,867 stars)
PR merge metrics
 (No merged PRs in 30d)

Description

Hi, I am getting following error while loading the model after retraining on my custom dataet.

mmseg/apis/inference.py:36, in init_segmentor(config, checkpoint, device)
     34     checkpoint = load_checkpoint(model, checkpoint, map_location='cpu')
     35     model.CLASSES = checkpoint['meta']['CLASSES']
---> 36     model.PALETTE = checkpoint['meta']['PALETTE']
     37 model.cfg = config  # save the config in the model for convenience
     38 model.to(device)

KeyError: 'PALETTE'

I checked the keys in my saved model which are these

dict_keys(['epoch', 'iter', 'mmcv_version', 'time', 'CLASSES'])

Did training with following code

# Build the detector
model = build_segmentor(cfg.model)
# Add an attribute for visualization convenience
model.CLASSES = datasets[0].CLASSES
model.PALETTE = datasets[0].PALETTE

# Create work_dir
mmcv.mkdir_or_exist(osp.abspath(cfg.work_dir))
train_segmentor(model, datasets, cfg, distributed=False, validate=True, 
                meta=dict())

Contributor guide

Research direction
Trace the training code to confirm that PALETTE is not being saved in the checkpoint meta. Modify either the training or inference code to ensure PALETTE is present, or add a fallback for missing PALETTE.
Tech stack
pythonpytorch
Domain
machine learningai
Issue type
Bug
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
2
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-3 hours
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
Pythonpytorch
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
50