Mercurial > repos > goeckslab > ludwig_predict

comparison ludwig_predict.xml @ 0:0a7b83ddda17 draft default tip

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/goeckslab/Galaxy-Ludwig.git commit bdea9430787658783a51cc6c2ae951a01e455bb4
author goeckslab
date Tue, 07 Jan 2025 22:45:18 +0000
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:0a7b83ddda17
1 <tool id="ludwig_predict" name="Ludwig Predict" version="@VERSION@" profile="@PROFILE@">
2 <description>loads a pretrained model to do prediction</description>
3 <macros>
4 <import>ludwig_macros.xml</import>
5 </macros>
6 <expand macro="python_requirements_gpu" />
7 <expand macro="macro_stdio" />
8 <version_command>echo "@VERSION@"</version_command>
9 <command>
10 <![CDATA[
11 #import re
12 mkdir -p outputs &&
13 #if $dataset
14 #set $sanitized_dataset = re.sub('[^\w\-_\.]', '_', $dataset.element_identifier.strip())
15 ln -sf '$dataset' "./${sanitized_dataset}";
16 #end if
17
18 #if $raw_data
19 unzip -o -q '$raw_data' -d ./;
20 #end if
21
22 python '$__tool_directory__/ludwig_predict.py'
23 #if $model_path
24 --model_path '$model_path.extra_files_path'
25 #end if
26 #if $dataset
27 --dataset "./${sanitized_dataset}"
28 #end if
29 #if $disable_parallel_threads
30 --disable_parallel_threads
31 #end if
32 --output_directory "./outputs"
33 --data_format '$data_format'
34 --split '$split'
35 --backend local
36 --skip_save_unprocessed_output &&
37 echo "Prediction is done!"
38
39 ]]>
40 </command>
41 <configfiles>
42 <inputs name="inputs" />
43 </configfiles>
44 <inputs>
45 <param name="model_path" type="data" format="ludwig_model" label="Load the pretrained model" />
46 <param name="dataset" type="data" format="tabular,csv,h5,json,txt" label="Input dataset" />
47 <param name="data_format" type="select" label="Data format">
48 <option value="auto" selected="true">auto</option>
49 <option value="tsv">tsv</option>
50 <option value="csv">csv</option>
51 <option value="h5">h5</option>
52 <option value="json">json</option>
53 </param>
54 <param name="split" type="select" label="Select the split portion to test the model on">
55 <option value="training">training</option>
56 <option value="validation">validation</option>
57 <option value="test">test</option>
58 <option value="full" selected="true">full</option>
59 </param>
60 <param name="batch_size" type="integer" value="128" optional="true" label="Batch size" min="1" max="4096" />
61 <param name="disable_parallel_threads" type="boolean" checked="false" label="Whether to disable parallel threads for reproducibility?" />
62 <param name="raw_data" type="data" format="zip" optional="true" label="Raw data" help="Optional. Needed for images."/>
63 </inputs>
64 <outputs>
65 <collection type="list" name="output_csv" label="${tool.name} CSV on ${on_string}" >
66 <discover_datasets pattern="(?P&lt;designation&gt;predictions_parquet\.csv)" format="csv" directory="outputs" />
67 <discover_datasets pattern="(?P&lt;designation&gt;.+)\.json" format="json" directory="outputs" />
68 </collection>
69 <!-- <data format="html" name="output_report" from_work_dir="outputs/smart_report.html" label="${tool.name} report on ${on_string}" /> -->
70 <!-- <data format="csv" name="output_top_k" label="${tool.name} top K predictions on ${on_string}" /> -->
71 </outputs>
72 <tests>
73 <test>
74 <param name="model_path" value="" ftype="ludwig_model">
75 <composite_data value="temp_model01/model_hyperparameters.json" />
76 <composite_data value="temp_model01/model_weights" />
77 <composite_data value="temp_model01/training_set_metadata.json" />
78 <composite_data value="temp_model01/training_progress.json" />
79 </param>
80 <param name="dataset" value="temperature_la.csv" ftype="csv" />
81 <param name="split" value="test" />
82 <output_collection name="output_csv">
83 <element name="predictions_parquet.csv">
84 <assert_contents>
85 <has_n_columns n="1" />
86 </assert_contents>
87 </element>
88 </output_collection>
89 </test>
90 </tests>
91 <help>
92 <![CDATA[
93 **What it does**
94 This tool conducts `ludwig predict`.
95
96
97 **Input**
98 - a trained ludwig model.
99 - dataset to be evaluate.
100
101
102 **Output**
103 - report in html.
104 - a collection of prediction results.
105
106
107
108 ]]>
109 </help>
110 <expand macro="macro_citations" />
111 </tool>