Mercurial > repos > bgruening > whisper

comparison whisper.xml @ 0:2706ea308f94 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/whisper commit 65bc65452f0cb44220555d6295106ea525038c70
author bgruening
date Wed, 24 Apr 2024 22:10:03 +0000
parents
children 02fa7b2245e6
comparison
equal deleted inserted replaced
-1:000000000000 0:2706ea308f94
1 <tool id="whisper" name="Speach to Text" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="20.01" license="MIT">
2 <description>
3 Transcribe audio or video files to text using the OpenAI Whisper
4 </description>
5 <macros>
6 <token name="@TOOL_VERSION@">20231117</token>
7 <token name="@VERSION_SUFFIX@">0</token>
8 </macros>
9 <requirements>
10 <container type="docker">quay.io/galaxy/whisper:20231117</container>
11 </requirements>
12 <command detect_errors="exit_code"><![CDATA[
13 mkdir -p ./outs ./models &&
14 ln -s '$infile' ./input.${infile.ext} &&
15 whisper ./input.${infile.ext}
16 --model $model
17 --output_dir ./outs
18 --threads \${GALAXY_SLOTS:-2}
19 --task transcribe
20 --output_format all
21 --verbose False
22 --model_dir \${OPENAI_WHISPER_MODEL_DIR:-./models}
23 #if str($language).strip():
24 --language '$language'
25 #end if
26 #if $advanced.temperature:
27 --temperature '$temperature'
28 #end if
29 #if $advanced.best_of:
30 --best_of '$advanced.best_of'
31 #end if
32 #if $advanced.beam_size:
33 --beam_size '$advanced.beam_size'
34 #end if
35 #if $advanced.patience:
36 --patience '$advanced.patience'
37 #end if
38 #if $advanced.length_penalty:
39 --length_penalty '$advanced.length_penalty'
40 #end if
41 #if $advanced.suppress_tokens:
42 --suppress_tokens '$advanced.suppress_tokens'
43 #end if
44 #if str($advanced.initial_prompt).strip():
45 --initial_prompt '$advanced.initial_prompt'
46 #end if
47 #if $advanced.condition_on_previous_text:
48 --condition_on_previous_text '$advanced.condition_on_previous_text'
49 #end if
50 #if $advanced.temperature_increment_on_fallback:
51 --temperature_increment_on_fallback '$advanced.temperature_increment_on_fallback'
52 #end if
53 #if $advanced.compression_ratio_threshold:
54 --compression_ratio_threshold '$advanced.compression_ratio_threshold'
55 #end if
56 #if $advanced.logprob_threshold:
57 --logprob_threshold '$advanced.logprob_threshold'
58 #end if
59 #if $advanced.no_speech_threshold:
60 --no_speech_threshold '$advanced.no_speech_threshold'
61 #end if
62 #if $advanced.word_timestamps.word_timestamps == "True":
63 --word_timestamps '$advanced.word_timestamps.word_timestamps'
64 #if $advanced.word_timestamps.highlight_words:
65 --highlight_words '$advanced.word_timestamps.highlight_words'
66 #end if
67 #if $advanced.word_timestamps.max_line_width:
68 --max_line_width '$advanced.word_timestamps.max_line_width'
69 #end if
70 #if $advanced.word_timestamps.max_line_count:
71 --max_line_count '$advanced.word_timestamps.max_line_count'
72 #end if
73 #if $advanced.word_timestamps.max_words_per_line:
74 --max_words_per_line '$advanced.word_timestamps.max_words_per_line'
75 #end if
76 #end if
77 ]]>
78 </command>
79 <environment_variables>
80 <!-- we will disable the progress bar which is printed to stderr -->
81 <environment_variable name="TQDM_DISABLE">1</environment_variable>
82 </environment_variables>
83 <inputs>
84 <param name="infile" type="data" format="wav,mp3,mkv,flv,mpg,ogg,wma,mp4" label="Select audio or video file" />
85 <param argument="--model" type="select" label="Speech to Text Model">
86 <option value="tiny">Tiny (~32x faster than the large model)</option>
87 <option value="base">Base (~16x faster than the large model)</option>
88 <option value="small" selected="true">Small (~6x faster than the large model)</option>
89 <option value="medium">Medium (~2x faster than the large model)</option>
90 <option value="large">Large</option>
91 </param>
92 <param argument="--language" type="select" label="Language">
93 <option value="">Auto (detect language)</option>
94 <option value="Afrikaans">Afrikaans</option>
95 <option value="Albanian">Albanian</option>
96 <option value="Amharic">Amharic</option>
97 <option value="Arabic">Arabic</option>
98 <option value="Armenian">Armenian</option>
99 <option value="Assamese">Assamese</option>
100 <option value="Azerbaijani">Azerbaijani</option>
101 <option value="Bashkir">Bashkir</option>
102 <option value="Basque">Basque</option>
103 <option value="Belarusian">Belarusian</option>
104 <option value="Bengali">Bengali</option>
105 <option value="Bosnian">Bosnian</option>
106 <option value="Breton">Breton</option>
107 <option value="Bulgarian">Bulgarian</option>
108 <option value="Burmese">Burmese</option>
109 <option value="Cantonese">Cantonese</option>
110 <option value="Castilian">Castilian</option>
111 <option value="Catalan">Catalan</option>
112 <option value="Chinese">Chinese</option>
113 <option value="Croatian">Croatian</option>
114 <option value="Czech">Czech</option>
115 <option value="Danish">Danish</option>
116 <option value="Dutch">Dutch</option>
117 <option value="English">English</option>
118 <option value="Estonian">Estonian</option>
119 <option value="Faroese">Faroese</option>
120 <option value="Finnish">Finnish</option>
121 <option value="Flemish">Flemish</option>
122 <option value="French">French</option>
123 <option value="Galician">Galician</option>
124 <option value="Georgian">Georgian</option>
125 <option value="German">German</option>
126 <option value="Greek">Greek</option>
127 <option value="Gujarati">Gujarati</option>
128 <option value="Haitian">Haitian</option>
129 <option value="Haitian Creole">Haitian Creole</option>
130 <option value="Hausa">Hausa</option>
131 <option value="Hawaiian">Hawaiian</option>
132 <option value="Hebrew">Hebrew</option>
133 <option value="Hindi">Hindi</option>
134 <option value="Hungarian">Hungarian</option>
135 <option value="Icelandic">Icelandic</option>
136 <option value="Indonesian">Indonesian</option>
137 <option value="Italian">Italian</option>
138 <option value="Japanese">Japanese</option>
139 <option value="Javanese">Javanese</option>
140 <option value="Kannada">Kannada</option>
141 <option value="Kazakh">Kazakh</option>
142 <option value="Khmer">Khmer</option>
143 <option value="Korean">Korean</option>
144 <option value="Lao">Lao</option>
145 <option value="Latin">Latin</option>
146 <option value="Latvian">Latvian</option>
147 <option value="Letzeburgesch">Letzeburgesch</option>
148 <option value="Lingala">Lingala</option>
149 <option value="Lithuanian">Lithuanian</option>
150 <option value="Luxembourgish">Luxembourgish</option>
151 <option value="Macedonian">Macedonian</option>
152 <option value="Malagasy">Malagasy</option>
153 <option value="Malay">Malay</option>
154 <option value="Malayalam">Malayalam</option>
155 <option value="Maltese">Maltese</option>
156 <option value="Mandarin">Mandarin</option>
157 <option value="Maori">Maori</option>
158 <option value="Marathi">Marathi</option>
159 <option value="Moldavian">Moldavian</option>
160 <option value="Moldovan">Moldovan</option>
161 <option value="Mongolian">Mongolian</option>
162 <option value="Myanmar">Myanmar</option>
163 <option value="Nepali">Nepali</option>
164 <option value="Norwegian">Norwegian</option>
165 <option value="Nynorsk">Nynorsk</option>
166 <option value="Occitan">Occitan</option>
167 <option value="Panjabi">Panjabi</option>
168 <option value="Pashto">Pashto</option>
169 <option value="Persian">Persian</option>
170 <option value="Polish">Polish</option>
171 <option value="Portuguese">Portuguese</option>
172 <option value="Punjabi">Punjabi</option>
173 <option value="Pushto">Pushto</option>
174 <option value="Romanian">Romanian</option>
175 <option value="Russian">Russian</option>
176 <option value="Sanskrit">Sanskrit</option>
177 <option value="Serbian">Serbian</option>
178 <option value="Shona">Shona</option>
179 <option value="Sindhi">Sindhi</option>
180 <option value="Sinhala">Sinhala</option>
181 <option value="Sinhalese">Sinhalese</option>
182 <option value="Slovak">Slovak</option>
183 <option value="Slovenian">Slovenian</option>
184 <option value="Somali">Somali</option>
185 <option value="Spanish">Spanish</option>
186 <option value="Sundanese">Sundanese</option>
187 <option value="Swahili">Swahili</option>
188 <option value="Swedish">Swedish</option>
189 <option value="Tagalog">Tagalog</option>
190 <option value="Tajik">Tajik</option>
191 <option value="Tamil">Tamil</option>
192 <option value="Tatar">Tatar</option>
193 <option value="Telugu">Telugu</option>
194 <option value="Thai">Thai</option>
195 <option value="Tibetan">Tibetan</option>
196 <option value="Turkish">Turkish</option>
197 <option value="Turkmen">Turkmen</option>
198 <option value="Ukrainian">Ukrainian</option>
199 <option value="Urdu">Urdu</option>
200 <option value="Uzbek">Uzbek</option>
201 <option value="Valencian">Valencian</option>
202 <option value="Vietnamese">Vietnamese</option>
203 <option value="Welsh">Welsh</option>
204 <option value="Yiddish">Yiddish</option>
205 <option value="Yoruba">Yoruba</option>
206 </param>
207 <param argument="--output_format" type="select" label="Output Format" multiple="true">
208 <option value="txt" selected="true">Text</option>
209 <option value="json">JSON</option>
210 <option value="srt">SubRip</option>
211 <option value="vtt">WebVTT</option>
212 <option value="tsv">Tab-separated values</option>
213 </param>
214 <section name="advanced" title="Advanced Options">
215 <param argument="--temperature" type="integer" value="0" optional="true" label="Temperature" help="Temperature to use for sampling" />
216 <param argument="--best_of" type="integer" value="5" optional="true" label="Best of" help="Number of candidates when sampling with non-zero temperature" />
217 <param argument="--beam_size" type="integer" value="5" optional="true" label="Beam size" help="Number of beams in beam search, only applicable when temperature is zero" />
218 <param argument="--patience" type="float" value="" optional="true" label="Optional patience value to use in beam decoding" help="As in https://arxiv.org/abs/2204.05424, the default (1.0) is equivalent to conventional beam search" />
219 <param argument="--length_penalty" type="float" value="" optional="true" label="Optional token length penalty coefficient (alpha)" help="As in https://arxiv.org/abs/1609.08144, uses simple length normalization by default" />
220 <param argument="--suppress_tokens" type="integer" value="-1" optional="true" label="Suppress tokens" help="Comma-separated list of token ids to suppress during sampling; -1 will suppress most special characters except common punctuations" />
221 <param argument="--initial_prompt" type="text" value="" optional="true" label="Initial prompt" help="Optional text to provide as a prompt for the first window" />
222 <param argument="--condition_on_previous_text" type="boolean" truevalue="True" falsevalue="False" checked="true" optional="true" label="Condition on previous text" help="If True, provide the previous output of the model as a prompt for the next window; disabling may make the text inconsistent across windows, but the model becomes less prone to getting stuck in a failure loop" />
223 <param argument="--temperature_increment_on_fallback" type="float" value="0.2" optional="true" label="Temperature increment on fallback" help="Temperature to increase when falling back when the decoding fails to meet either of the thresholds below" />
224 <param argument="--compression_ratio_threshold" type="float" value="2.4" optional="true" label="Compression ratio threshold" help="If the gzip compression ratio is higher than this value, treat the decoding as failed" />
225 <param argument="--logprob_threshold" type="float" value="-1.0" optional="true" label="Logprob threshold" help="If the average log probability is lower than this value, treat the decoding as failed" />
226 <param argument="--no_speech_threshold" type="float" value="0.6" optional="true" label="No speech threshold" help="If the probability of the |nospeech| token is higher than this value AND the decoding has failed due to `logprob_threshold`, consider the segment as silence" />
227 <conditional name="word_timestamps">
228 <param argument="--word_timestamps" type="select" label="Extract word timestamps?" help="(experimental) Extract word-level timestamps and refine the results based on them">
229 <option value="False">False</option>
230 <option value="True">True</option>
231 </param>
232 <when value="True">
233 <param argument="--highlight_words" type="boolean" truevalue="True" falsevalue="False" value="False" optional="true" label="Highlight words" help="Underline each word as it is spoken in srt and vtt" />
234 <param argument="--max_line_width" type="integer" value="" optional="true" label="Max line width" help="The maximum number of characters in a line before breaking the line" />
235 <param argument="--max_line_count" type="integer" value="" optional="true" label="Max line count" help="The maximum number of lines in a segment" />
236 <param argument="--max_words_per_line" type="integer" value="" optional="true" label="Max words per line" help="No effect with --max_line_width. the maximum number of words in a segment" />
237 </when>
238 <when value="False">
239 </when>
240 </conditional>
241 </section>
242 </inputs>
243 <outputs>
244 <data name="output_txt" format="txt" from_work_dir="./outs/input.txt" label="${tool.name} on ${on_string}.txt">
245 <filter>'txt' in output_format</filter>
246 </data>
247 <data name="output_json" format="json" from_work_dir="./outs/input.json" label="${tool.name} on ${on_string}.json">
248 <filter>'json' in output_format</filter>
249 </data>
250 <data name="output_srt" format="txt" from_work_dir="./outs/input.srt" label="${tool.name} on ${on_string}.srt">
251 <filter>'srt' in output_format</filter>
252 </data>
253 <data name="output_vtt" format="txt" from_work_dir="./outs/input.vtt" label="${tool.name} on ${on_string}.vtt">
254 <filter>'vtt' in output_format</filter>
255 </data>
256 <data name="output_tsv" format="tabular" from_work_dir="./outs/input.tsv" label="${tool.name} on ${on_string}.tsv">
257 <filter>'tsv' in output_format</filter>
258 </data>
259 </outputs>
260 <tests>
261 <test expect_num_outputs="1">
262 <param name="infile" value="english.wav" ftype="wav"/>
263 <param name="model" value="tiny"/>
264 <param name="language" value="English"/>
265 <param name="output_format" value="txt"/>
266 <output name="output_txt" file="transcribe.txt"/>
267 </test>
268 <test expect_num_outputs="3">
269 <param name="infile" value="english.wav" ftype="wav"/>
270 <param name="model" value="tiny"/>
271 <param name="language" value="English"/>
272 <param name="output_format" value="srt,tsv,json"/>
273 <output name="output_srt" file="transcribe_english.srt"/>
274 <output name="output_tsv" file="transcribe_english.tsv"/>
275 <output name="output_json">
276 <assert_contents>
277 <has_text text="21-year-old Jesus joined Manchester City last year in January 2017 from"/>
278 <has_text text="temperature"/>
279 <has_text text="no_speech_prob"/>
280 <has_text text="English"/>
281 <has_n_lines n="1"/>
282 </assert_contents>
283 </output>
284 </test>
285 <test expect_num_outputs="1">
286 <param name="infile" value="german.wav" ftype="wav"/>
287 <param name="model" value="small"/>
288 <param name="language" value="German"/>
289 <param name="output_format" value="txt"/>
290 <output name="output_txt" file="transcribe_german.txt"/>
291 </test>
292 <test expect_num_outputs="1">
293 <param name="infile" value="german_english.mp3" ftype="mp3"/>
294 <param name="model" value="medium"/>
295 <param name="output_format" value="txt"/>
296 <output name="output_txt" file="transcribe_german_english.txt"/>
297 </test>
298 <test expect_num_outputs="2">
299 <param name="infile" value="persian.wav" ftype="wav"/>
300 <param name="model" value="medium"/>
301 <param name="language" value="Persian"/>
302 <param name="output_format" value="srt,json"/>
303 <section name="advanced">
304 <param name="condition_on_previous_text" value="False"/>
305 <conditional name="word_timestamps">
306 <param name="word_timestamps" value="True"/>
307 <param name="max_words_per_line" value="5"/>
308 </conditional>
309 </section>
310 <output name="output_srt">
311 <assert_contents>
312 <has_n_lines n="32"/>
313 </assert_contents>
314 </output>
315 <output name="output_json">
316 <assert_contents>
317 <has_text text="\u0628\u0631\u062e\u06cc \u0627\u0632 \u0627\u06cc\u0646"/>
318 <has_text text="temperature"/>
319 <has_text text="no_speech_prob"/>
320 <has_text text="Persian"/>
321 <has_n_lines n="1"/>
322 </assert_contents>
323 </output>
324 </test>
325 </tests>
326 <help><![CDATA[
327
328 .. class:: infomark
329
330 **What it does**
331
332 Transcribe audio or video files to text using the `Whisper from OpenAI <https://github.com/openai/whisper>`_.
333
334 Usage
335 .....
336
337
338 **Input**
339 Audio or video file to transcribe in one of wav, mp3, mkv, flv, mpg, ogg, wma, or mp4.
340
341
342 **Output**
343 Transcribed text in the selected format. The output can be in text, JSON, SubRip, WebVTT, or tab-separated values (tabular) format.
344 ]]></help>
345 <creator>
346 <person givenName="Alireza" familyName="Heidari" url="http://github.com/itisalirh"/>
347 </creator>
348 <citations>
349 <citation type="bibtex">
350 @misc{openai2022whisper,
351 title={Whisper},
352 author={OpenAI},
353 year={2022},
354 url={https://github.com/openai/whisper}
355 }
356 </citation>
357 </citations>
358 </tool>