Mercurial > repos > goeckslab > image_learner

comparison image_learner.xml @ 0:54b871dfc51e draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/goeckslab/gleam.git commit b7411ff35b6228ccdfd36cd4ebd946c03ac7f7e9
author goeckslab
date Tue, 03 Jun 2025 21:22:11 +0000
parents
children 39202fe5cf97
comparison
equal deleted inserted replaced
-1:000000000000 0:54b871dfc51e
1 <tool id="image_learner" name="Image Learner for Classification" version="0.1.0" profile="22.05">
2 <description>trains and evaluates a image classification model</description>
3 <requirements>
4 <container type="docker">quay.io/goeckslab/galaxy-ludwig-gpu:0.10.1</container>
5 </requirements>
6 <required_files>
7 <include path="utils.py" />
8 <include path="image_learner_cli.py" />
9 </required_files>
10 <stdio>
11 <exit_code range="137" level="fatal_oom" description="Out of Memory" />
12 <exit_code range="1:" level="fatal" description="Error occurred. Please check Tool Standard Error" />
13 </stdio>
14 <command>
15 <![CDATA[
16 #import re
17
18 #if $input_csv
19
20 #set $sanitized_input_csv = re.sub('[^\w\-_\.]', '_', $input_csv.element_identifier.strip())
21 ln -sf '$input_csv' "./${sanitized_input_csv}";
22 #end if
23
24 python '$__tool_directory__/image_learner_cli.py'
25 --csv-file "./${sanitized_input_csv}"
26 --image-zip "$image_zip"
27 --model-name "$model_name"
28 #if $use_pretrained == "true"
29 --use-pretrained
30 #if $fine_tune == "true"
31 --fine-tune
32 #end if
33 #end if
34 #if $customize_defaults == "true"
35 #if $epochs
36 --epochs "$epochs"
37 #end if
38 #if $early_stop
39 --early-stop "$early_stop"
40 #end if
41 #if $learning_rate_define == "true"
42 --learning-rate "$learning_rate"
43 #end if
44 #if $batch_size_define == "true"
45 --batch-size "$batch_size"
46 #end if
47 --split-probabilities "$train_split" "$val_split" "$test_split"
48 #end if
49 --random-seed "$random_seed"
50 --output-dir "." &&
51
52 mkdir -p '$output_model.extra_files_path' &&
53 cp -r experiment_run/model/*.json experiment_run/model/model_weights '$output_model.extra_files_path' &&
54
55 echo "Image Learner Classification Experiment is Done!"
56 ]]>
57 </command>
58
59 <inputs>
60 <param argument="input_csv" type="data" format="csv" optional="false" label="the metadata csv containing image_path column, label column and optional split column" />
61 <param name="image_zip" type="data" format="zip" optional="false" label="Image zip" help="Image zip file containing your image data"/>
62 <param name="model_name" type="select" optional="false" label="Select a model for your experiment" >
63
64 <option value="resnet18">Resnet18</option>
65 <option value="resnet34">Resnet34</option>
66 <option value="resnet50">Resnet50</option>
67 <option value="resnet101">Resnet101</option>
68 <option value="resnet152">Resnet152</option>
69 <option value="resnext50_32x4d">Resnext50_32x4d</option>
70 <option value="resnext101_32x8d">Resnext101_32x8d</option>
71 <option value="resnext101_64x4d">Resnext101_64x4d</option>
72 <option value="resnext152_32x8d">Resnext152_32x8d</option>
73 <option value="wide_resnet50_2">Wide_resnet50_2</option>
74 <option value="wide_resnet101_2">Wide_resnet101_2</option>
75 <option value="wide_resnet103_2">Wide_resnet103_2</option>
76 <option value="efficientnet_b0">Efficientnet_b0</option>
77 <option value="efficientnet_b1">Efficientnet_b1</option>
78 <option value="efficientnet_b2">Efficientnet_b2</option>
79 <option value="efficientnet_b3">Efficientnet_b3</option>
80 <option value="efficientnet_b4">Efficientnet_b4</option>
81 <option value="efficientnet_b5">Efficientnet_b5</option>
82 <option value="efficientnet_b6">Efficientnet_b6</option>
83 <option value="efficientnet_b7">Efficientnet_b7</option>
84 <option value="efficientnet_v2_s">Efficientnet_v2_s</option>
85 <option value="efficientnet_v2_m">Efficientnet_v2_m</option>
86 <option value="efficientnet_v2_l">Efficientnet_v2_l</option>
87 <option value="regnet_y_400mf">Regnet_y_400mf</option>
88 <option value="regnet_y_800mf">Regnet_y_800mf</option>
89 <option value="regnet_y_1_6gf">Regnet_y_1_6gf</option>
90 <option value="regnet_y_3_2gf">Regnet_y_3_2gf</option>
91 <option value="regnet_y_8gf">Regnet_y_8gf</option>
92 <option value="regnet_y_16gf">Regnet_y_16gf</option>
93 <option value="regnet_y_32gf">Regnet_y_32gf</option>
94 <option value="regnet_y_128gf">Regnet_y_128gf</option>
95 <option value="regnet_x_400mf">Regnet_x_400mf</option>
96 <option value="regnet_x_800mf">Regnet_x_800mf</option>
97 <option value="regnet_x_1_6gf">Regnet_x_1_6gf</option>
98 <option value="regnet_x_3_2gf">Regnet_x_3_2gf</option>
99 <option value="regnet_x_8gf">Regnet_x_8gf</option>
100 <option value="regnet_x_16gf">Regnet_x_16gf</option>
101 <option value="regnet_x_32gf">Regnet_x_32gf</option>
102 <option value="vgg11">Vgg11</option>
103 <option value="vgg11_bn">Vgg11_bn</option>
104 <option value="vgg13">Vgg13</option>
105 <option value="vgg13_bn">Vgg13_bn</option>
106 <option value="vgg16">Vgg16</option>
107 <option value="vgg16_bn">Vgg16_bn</option>
108 <option value="vgg19">Vgg19</option>
109 <option value="vgg19_bn">Vgg19_bn</option>
110 <option value="shufflenet_v2_x0_5">Shufflenet_v2_x0_5</option>
111 <option value="shufflenet_v2_x1_0">Shufflenet_v2_x1_0</option>
112 <option value="shufflenet_v2_x1_5">Shufflenet_v2_x1_5</option>
113 <option value="shufflenet_v2_x2_0">Shufflenet_v2_x2_0</option>
114 <option value="squeezenet1_0">Squeezenet1_0</option>
115 <option value="squeezenet1_1">Squeezenet1_1</option>
116 <option value="swin_t">Swin_t</option>
117 <option value="swin_s">Swin_s</option>
118 <option value="swin_b">Swin_b</option>
119 <option value="swin_v2_t">Swin_v2_t</option>
120 <option value="swin_v2_s">Swin_v2_s</option>
121 <option value="swin_v2_b">Swin_v2_b</option>
122 <option value="vit_b_16">Vit_b_16</option>
123 <option value="vit_b_32">Vit_b_32</option>
124 <option value="vit_l_16">Vit_l_16</option>
125 <option value="vit_l_32">Vit_l_32</option>
126 <option value="vit_h_14">Vit_h_14</option>
127 <option value="convnext_tiny">Convnext_tiny</option>
128 <option value="convnext_small">Convnext_small</option>
129 <option value="convnext_base">Convnext_base</option>
130 <option value="convnext_large">Convnext_large</option>
131 <option value="maxvit_t">Maxvit_t</option>
132 <option value="alexnet">Alexnet</option>
133 <option value="googlenet">Googlenet</option>
134 <option value="inception_v3">Inception_v3</option>
135 <option value="mobilenet_v2">Mobilenet_v2</option>
136 <option value="mobilenet_v3_large">Mobilenet_v3_large</option>
137 <option value="mobilenet_v3_small">Mobilenet_v3_small</option>
138 </param>
139
140 <conditional name="scratch_fine_tune">
141 <param name="use_pretrained" type="select"
142 label="Use pretrained weights?"
143 help="If select no, the encoder, combiner, and decoder will all be initialized and trained from scratch.
144 (e.g. when your images are very different from ImageNet or no suitable pretrained model exists.)">
145 <option value="false">No</option>
146 <option value="true" selected="true">Yes</option>
147 </param>
148 <when value="true">
149 <param name="fine_tune" type="select" label="Fine tune the encoder?"
150 help="Whether to fine tune the encoder(combiner and decoder will be fine-tued anyway)" >
151 <option value="false" >No</option>
152 <option value="true" selected="true">Yes</option>
153 </param>
154 </when>
155 <when value="false">
156 <!-- No additional parameters to show if the user selects 'No' -->
157 </when>
158 </conditional>
159 <param argument="random_seed" type="integer" value="42" optional="true" label="Random seed (set for reproducibility)" min="0" max="999999"/>
160 <conditional name="advanced_settings">
161 <param name="customize_defaults" type="select" label="Customize Default Settings?" help="Select yes if you want to customize the default settings of the experiment.">
162 <option value="false" selected="true">No</option>
163 <option value="true">Yes</option>
164 </param>
165 <when value="true">
166 <param name="epochs" type="integer" value="10" min="1" max="99999" label="Epochs" help="Total number of full passes through the training dataset. Higher values may improve accuracy but increase training time. Default: 10." />
167 <param name="early_stop" type="integer" value="5" min="1" max="99999" label="Early Stop" help="Number of epochs with no improvement after which training will be stopped. Default: 5." />
168 <conditional name="learning_rate_condition">
169 <param name="learning_rate_define" type="select" label="Define an initial learning rate?" help="Want to change the initial learning rate from default to a number? See ludwig.ai for more info. Default: No" >
170 <option value="false" selected="true" >No</option>
171 <option value="true">Yes</option>
172 </param>
173 <when value="true">
174 <param name="learning_rate" type="float" value="0.001" min="0.0001" max="1.0" label="Learning Rate" help="Initial learning rate for the optimizer. Default: 0.001." />
175 </when>
176 <when value="false">
177 <!-- No additional parameters to show if the user selects 'No' -->
178 </when>
179 </conditional>
180 <conditional name="batch_size_condition">
181 <param name="batch_size_define" type="select" label="Define your batch size?" help="Want to change the batch size from auto to a number? See ludwig.ai for more info. Default: No" >
182 <option value="false" selected="true" >No</option>
183 <option value="true">Yes</option>
184 </param>
185 <when value="true">
186 <param name="batch_size" type="integer" value="32" min="1" max="99999" label="Batch Size" help="Number of samples per gradient update. Default: 32." />
187 </when>
188 <when value="false">
189 <!-- No additional parameters to show if the user selects 'No' -->
190 </when>
191 </conditional>
192 <param name="train_split" type="float"
193 label="Training split proportion (only works if no split column in the metadata csv)"
194 value="0.7"
195 help="Fraction of data for training (e.g., 0.7). train split + val split + test split should = 1"/>
196 <param name="val_split" type="float"
197 label="Validation split proportion (only works if no split column in the metadata csv)"
198 value="0.1"
199 help="Fraction of data for validation (e.g., 0.1). train split + val split + test split should = 1"/>
200 <param name="test_split" type="float"
201 label="Test split proportion (only works if no split column in the metadata csv)"
202 value="0.2"
203 help="Fraction of data for testing (e.g., 0.2) train split + val split + test split should = 1."/>
204 </when>
205 <when value="false">
206 <!-- No additional parameters to show if the user selects 'No' -->
207 </when>
208 </conditional>
209 </inputs>
210 <outputs>
211 <data format="ludwig_model" name="output_model" label="${tool.name} trained model on ${on_string}" />
212 <data format="html" name="output_report" from_work_dir="image_classification_results_report.html" label="${tool.name} report on ${on_string}" />
213 <collection type="list" name="output_pred_csv" label="${tool.name} predictions CSVs/experiment stats/plots on ${on_string}" >
214 <discover_datasets pattern="(?P&lt;designation&gt;predictions\.csv)" format="csv" directory="experiment_run" />
215 <discover_datasets pattern="(?P&lt;designation&gt;.+)\.json" format="json" directory="experiment_run" />
216 <discover_datasets pattern="(?P&lt;designation&gt;.+)\.png" format="png" directory="experiment_run/visualizations/train" />
217 <discover_datasets pattern="(?P&lt;designation&gt;.+)\.png" format="png" directory="experiment_run/visualizations/test" />
218 </collection>
219 </outputs>
220 <tests>
221 <test expect_num_outputs="3">
222 <param name="input_csv" value="mnist_subset.csv" ftype="csv" />
223 <param name="image_zip" value="mnist_subset.zip" ftype="zip" />
224 <param name="model_name" value="resnet18" />
225 <output name="output_report" file="image_classification_results_report_mnist.html" compare="sim_size" delta="20000" >
226 <assert_contents>
227 <has_text text="Epochs" />
228 </assert_contents>
229 </output>
230
231 <output_collection name="output_pred_csv" type="list" >
232 <element name="predictions.csv" >
233 <assert_contents>
234 <has_n_columns n="1" />
235 </assert_contents>
236 </element>
237 </output_collection>
238 </test>
239 </tests>
240 <help>
241 <![CDATA[
242 **What it does**
243 Image Learner for Classification: trains and evaluates a image classification model.
244 It uses the metadata csv to find the image paths and labels.
245 The metadata csv should contain a column with the name 'image_path' and a column with the name 'label'.
246 Optionally, you can also add a column with the name 'split' to specify which split each row belongs to (train, val, test).
247 If you do not provide a split column, the tool will automatically split the data into train, val, and test sets based on the proportions you specify or [0.7, 0.1, 0.2] by default.
248
249
250 **Outputs**
251 The tool will output a trained model in the form of a ludwig_model file,
252 a report in the form of an HTML file, and a collection of CSV/json/png files containing the predictions, experiment stats and visualizations.
253 The html report will contain metrics&experiment setup parameters, train&val plots and test plots.
254
255 ]]>
256 </help>
257 <citations>
258 <citation type="bibtex">
259 @misc{https://doi.org/10.48550/arxiv.1909.07930,
260 doi = {10.48550/ARXIV.1909.07930},
261 url = {https://arxiv.org/abs/1909.07930},
262 author = {Molino, Piero and Dudin, Yaroslav and Miryala, Sai Sumanth},
263 title = {Ludwig: a type-based declarative deep learning toolbox},
264 publisher = {arXiv},
265 year = {2019},
266 copyright = {arXiv.org perpetual, non-exclusive license}
267 }
268 </citation>
269 </citations>
270 </tool>