+
j ^ RI Ht ^ RIHt ^ RIt^ RIt^RIHtH t ^RI
H t ^RI H
t
Ht ] P ! ]4 t]! ]! RRR7 4 ! R R
]
4 4 tR# ) )UserDict)AnyN)add_end_docstringslogging) ffmpeg_read)Pipelinebuild_pipeline_init_argsT)has_feature_extractor
has_tokenizerc z a a ] tR t^ t oRtRtRtRtRtV 3R lt V3R lV 3R llt
R t R R lt R t
R
tR tVtV ;t# )
#ZeroShotAudioClassificationPipelinea
Zero shot audio classification pipeline using `ClapModel`. This pipeline predicts the class of an audio when you
provide an audio and a set of `candidate_labels`.
The default `hypothesis_template` is : `"This is a sound of {}."`. Make sure you update it for your usage.
Example:
```python
>>> from transformers import pipeline
>>> from datasets import load_dataset
>>> dataset = load_dataset("ashraq/esc50")
>>> audio = next(iter(dataset["train"]["audio"]))["array"]
>>> classifier = pipeline(task="zero-shot-audio-classification", model="laion/clap-htsat-unfused")
>>> classifier(audio, candidate_labels=["Sound of a dog", "Sound of vacuum cleaner"])
[{'score': 0.9996, 'label': 'Sound of a dog'}, {'score': 0.0004, 'label': 'Sound of vacuum cleaner'}]
```
Learn more about the basics of using a pipeline in the [pipeline tutorial](../pipeline_tutorial). This audio
classification pipeline can currently be loaded from [`pipeline`] using the following task identifier:
`"zero-shot-audio-classification"`. See the list of available models on
[huggingface.co/models](https://huggingface.co/models?filter=zero-shot-audio-classification).
FTc ( < \ SV ` ! R/ VB R # )N )super__init__)selfkwargs __class__s &,v/home/luca/screen-checking/.venv/lib/python3.14/site-packages/transformers/pipelines/zero_shot_audio_classification.pyr ,ZeroShotAudioClassificationPipeline.__init__C s
"6" c < V ^8 d QhRS[ P S[, S[, S[, RS[RS[S[S[S[3, , /# ) audiosr return)npndarraybytesstrdictr list)format
__classdict__s "r __annotate__0ZeroShotAudioClassificationPipeline.__annotate__F sH 2 2rzzE 1C 7$ > 2# 2RVW[\_ad\dWeRf 2r c & < \ SV ` ! V3/ VB # )ah
Assign labels to the audio(s) passed as inputs.
Args:
audios (`str`, `list[str]`, `np.array` or `list[np.array]`):
The pipeline handles three types of inputs:
- A string containing a http link pointing to an audio
- A string containing a local path to an audio
- An audio loaded in numpy
candidate_labels (`list[str]`):
The candidate labels for this audio. They will be formatted using *hypothesis_template*.
hypothesis_template (`str`, *optional*, defaults to `"This is a sound of {}"`):
The format used in conjunction with *candidate_labels* to attempt the audio classification by
replacing the placeholder with the candidate_labels. Pass "{}" if *candidate_labels* are
already formatted.
Return:
A list of dictionaries containing one entry per proposed label. Each dictionary contains the
following keys:
- **label** (`str`) -- One of the suggested *candidate_labels*.
- **score** (`float`) -- The score attributed by the model to that label. It is a value between
0 and 1, computed as the `softmax` of `logits_per_audio`.
)r __call__)r r r r s &&,r r' ,ZeroShotAudioClassificationPipeline.__call__F s . w 1&11r c \ / pR V9 d
VR , VR & RV9 d
VR, VR& V/ / 3# )candidate_labelshypothesis_templater )r r preprocess_paramss &, r _sanitize_parameters8ZeroShotAudioClassificationPipeline._sanitize_parameters_ sI '4:;M4N 0 1 F *7=>S7T 3 4 "b((r c l \ V\ 4 ' d VP R 4 ' g VP R4 ' d$ \ P ! VRR7 P
pM-\
VR4 ;_uu_ 4 pVP 4 pRRR4 \ V\ 4 ' d \ WP P 4 p\ V\ P 4 ' g \ R4 h\ VP 4 ^8w d \# R4 hV P V.V P P RR 7 pVP% V P& 4 pW%R
& V Uu. uF qcP) V4 NK ppV P+ VRRR 7 pV.VR & V# + ' g i EL; iu upi )
zhttp://zhttps://T)follow_redirectsrbNz"We expect a numpy ndarray as inputzNWe expect a single channel audio input for ZeroShotAudioClassificationPipelinept)
sampling_ratereturn_tensorsr* )r4 padding text_inputs)
isinstancer
startswithhttpxgetcontentopenreadr r feature_extractorr3 r r TypeErrorlenshape
ValueErrortodtyper" tokenizer) r audior* r+ finputsx sequencesr6 s &&&& r
preprocess.ZeroShotAudioClassificationPipeline.preprocessh sc eS ! ! **e.>.>z.J.J %$?GG%&&! FFHE ' eU # # '='='K'KLE%,, @A A u{{ q mn n''
G4#9#9#G#GX\ (
4::&%5!"?"&&}5 k!nh / /%a.K &a.+K**;{;l;
0 g..
r c VP R 4 pVR, ^ , pVP ^ R7 pVP 4 p\ \ WR4 R R7 UUu. uF w rgRVRV/NK
pppV# u uppi )r* rN )dimc V ^ , ) # )r r )rI s &r AZeroShotAudioClassificationPipeline.postprocess.. s _`ab_c^cr )keyscorelabel)rO softmaxtolistsortedzip) r rT r* rN probsscoresr] candidate_labelresults && r postprocess/ZeroShotAudioClassificationPipeline.postprocess s (,,-?@ x(+1% +1V1NTc*d
*d& eWo 6*d
s A4r )NzThis is a sound of {}.)__name__
__module__ __qualname____firstlineno____doc___load_processor_load_image_processor_load_feature_extractor_load_tokenizerr r' r- rK rU rg __static_attributes____classdictcell__
__classcell__)r r# s @@r r
r
sJ : O ! "O#2 22)8 " r r
) collectionsr typingr r9 numpyr utilsr r audio_classificationr baser r
get_loggerri loggerr
r r r r} s_ ! . 4
H % ,4W[\]@( @ ^@r