mirror of
https://github.com/hanzoai/anomalib.git
synced 2026-10-30 18:48:11 +00:00
zeekay
dda5b2a367
The previous commit was meant to remove .github/workflows and removed every tracked path instead — 892 files, the whole repo. This puts back the 874 that were never in scope; the eighteen workflow files stay gone, for the reasons that commit gives. Content is byte-identical to 9d2aec9 outside .github/workflows.
49 lines
1.5 KiB
Bash
49 lines
1.5 KiB
Bash
#!/bin/bash
|
|
|
|
# Copyright (C) 2024 Intel Corporation
|
|
# SPDX-License-Identifier: Apache-2.0
|
|
|
|
# Using Custom Folder Dataset with Anomalib CLI
|
|
# ------------------------------------------
|
|
# This example shows how to use your own dataset organized in folders.
|
|
|
|
# 1. Basic Usage with Default Structure
|
|
# Train using the default folder structure
|
|
echo "Training on custom dataset with default structure..."
|
|
anomalib train \
|
|
--model efficient_ad \
|
|
--data Folder \
|
|
--data.name my_dataset \
|
|
--data.root ./datasets/my_dataset \
|
|
--data.normal_dir good \
|
|
--data.abnormal_dir defect
|
|
|
|
# 2. Custom Configuration
|
|
# Train with custom data settings
|
|
echo -e "\nTraining with custom data settings..."
|
|
anomalib train \
|
|
--model efficient_ad \
|
|
--data Folder \
|
|
--data.name my_dataset \
|
|
--data.root ./datasets/my_dataset \
|
|
--data.normal_dir normal \
|
|
--data.abnormal_dir anomaly \
|
|
--data.train_batch_size 32 \
|
|
--data.eval_batch_size 32 \
|
|
--data.num_workers 8
|
|
|
|
# 3. Training with Multiple Dataset Variations
|
|
# Train on different subsets or configurations
|
|
echo -e "\nTraining on multiple dataset variations..."
|
|
for defect_type in "scratch" "crack" "stain"; do
|
|
echo "Training on defect type: $defect_type"
|
|
anomalib train \
|
|
--model efficient_ad \
|
|
--data Folder \
|
|
--data.name my_dataset \
|
|
--data.root "./datasets/my_dataset/$defect_type" \
|
|
--data.normal_dir good \
|
|
--data.abnormal_dir "$defect_type" \
|
|
--trainer.default_root_dir "results/$defect_type"
|
|
done
|