+
j: ^ RI t ^ RIHtHtHt ^RIHt ^RIHt ^RI H
t
^RI H t H
t
^RIHt ]
! 4 ' d ^ RIt^RIHt ^R IHt R
t ! R R ]R
R7 t ! R R]4 tR# ) N)Any TypedDictoverload)
AudioInput)GenerationConfig)is_torch_available)ChatChatType)Pipeline)%MODEL_FOR_TEXT_TO_SPECTROGRAM_MAPPING)SpeechT5HifiGanzmicrosoft/speecht5_hifiganc 0 a ] tR t^"t o RtV 3R ltRtV tR# ) AudioOutputz
audio (`AudioInput`):
The generated audio waveform.
sampling_rate (`int`):
The sampling rate of the generated audio waveform.
c 2 < V ^8 d Qh/ S[ ;R& S[;R& # ) audio
sampling_rate)r int)format
__classdict__s "e/home/luca/screen-checking/.venv/lib/python3.14/site-packages/transformers/pipelines/text_to_audio.py __annotate__AudioOutput.__annotate__" s N)__name__
__module__ __qualname____firstlineno____doc____annotate_func____static_attributes____classdictcell__)r s @r r r " s r r F)totalc a a ] tR t^.t oRtRtRtRtRtRt ]
! RR7 t RRRRR R/V 3R
llt R t
R t]V3R
lR l4 t]V3R lR l4 t]V3R lR l4 t]V3R lR l4 tV 3R ltRR ltR tRtVtV ;t# )TextToAudioPipelinea
Text-to-audio generation pipeline using any `AutoModelForTextToWaveform` or `AutoModelForTextToSpectrogram`. This
pipeline generates an audio file from an input text and optional other conditional inputs.
Unless the model you're using explicitly sets generation parameters in `generation_config.json`, the default values
from `generation.configuration_utils.py._get_default_generation_params()` will be used.
- max_new_tokens: 256
Example:
```python
>>> from transformers import pipeline
>>> pipe = pipeline(model="suno/bark-small")
>>> output = pipe("Hey it's HuggingFace on the phone!")
>>> audio = output["audio"]
>>> sampling_rate = output["sampling_rate"]
```
Learn more about the basics of using a pipeline in the [pipeline tutorial](../pipeline_tutorial)
You can specify parameters passed to the model by using [`TextToAudioPipeline.__call__.forward_params`] or
[`TextToAudioPipeline.__call__.generate_kwargs`].
Example:
```python
>>> from transformers import pipeline
>>> music_generator = pipeline(task="text-to-audio", model="facebook/musicgen-small")
>>> # diversify the music generation by adding randomness with a high temperature and set a maximum music length
>>> generate_kwargs = {
... "do_sample": True,
... "temperature": 0.7,
... "max_new_tokens": 35,
... }
>>> outputs = music_generator("Techno music with high melodic riffs", generate_kwargs=generate_kwargs)
```
This pipeline can currently be loaded from [`pipeline`] using the following task identifiers: `"text-to-speech"` or
`"text-to-audio"`.
See the list of available models on [huggingface.co/models](https://huggingface.co/models?filter=text-to-speech).
TNF )max_new_tokensvocoderr noise_schedulerc < Ve W5R& W0n \ S V ` ! V/ VB R V n V P P
\ P ! 4 9 dI Vf> \ P ! \ 4 P V P P 4 MTV n V P P P R9 d R V n W n V P e& V P P P V n V P f V P P pV P P" P% RR 4 pVeH TP' VP) 4 P+ 4 UU u/ uF
w rV f K WbK up p4 R FO p
\- WjR 4 pVe W n K \- VRR 4 f K, \- VP. V
R 4 pVf KI W n KQ V P fV V P eF \1 V P R4 ' d( V P P2 P V n R # R # R # R # u up pi )Nr* generation_config codec_configfeature_extractor)musicgenspeecht5) sample_rater )r* super__init__r) model __class__r valuesr
from_pretrainedDEFAULT_VOCODER_IDtodeviceconfig
model_type processorr __dict__getupdateto_dictitemsgetattrr- hasattrr. ) selfr) r r* argskwargsr;
gen_configkvsampling_rate_namer5 s &$$$*, r r3 TextToAudioPipeline.__init__l s &(7$ % .
$)&) :: #H#O#O#Q Q ? //0BCFFtzzGXGXY
L :: ' '+C C !DN * << #!% !4!4!B!BD % ZZ&&F ,,001DdKJ%
0B0B0D0J0J0L ^0LPQtqt0L ^_&F" 'D I
,)6& V^T:F$+F,?,?ASUY$ZM$0-:* 'G %$..*DQUQ_Q_atIuIu!% !A!A!O!OD Jv*D % _s <I
I
c \ V\ 4 ' d V.pV P P P R 8X d] Rp\ V P R4 ' d" \ V P P RR4 pRVRRRRR R/pVP V4 TpV P e
V P M V P p\ V\ 4 ' dG VP R R
4 ;' g / pR RR
RRRR V/VCpVP ! VP 3/ VB pMV P P P R8X d? V U u. uF qP! R4 ' g RV 2MT NK" pp VP# RR4 V P P P R8X d- V U u. uF qP! R4 ' g RV 2MT NK" pp V! V3/ VBRR/B pVP% V P P& R7 pV# u up i u up i )barkr' semantic_configmax_input_semantic_length
max_lengthadd_special_tokensFreturn_attention_maskTreturn_token_type_idsNprocessor_kwargstokenize return_dictadd_generation_promptcsm[z[0]diaz[S1] return_tensorspt)dtype)
isinstancestrr4 r; r< rD r, rC rO r@ r= tokenizerr popapply_chat_templatemessages
startswith
setdefaultr9 r^ )
rE textrG rQ
new_kwargs preprocessorrU chat_template_kwargsoutputts
&&, r
preprocess TextToAudioPipeline.preprocess s dC 6D :: ' '6 1 Jt--/@AA $T%;%;%K%KMhjm n
j$e'' J
f % F)- )Ct~~ dD ! ! %zz*2##/4E^4[484H4H 01ZZ((J<J>JF?## KKZK_K_KaJbd
ZZA,A.A!DF << #\\&)F
r c , < V ^8 d QhRS[ RS[RS[/# r text_inputsrp return)r` r r )r r s "r r TextToAudioPipeline.__annotate__ s SSCS3S;Sr c R # Nr rE r rp s &&,r __call__ TextToAudioPipeline.__call__ s PSr c L < V ^8 d QhRS[ S[, RS[RS[ S[, /# r )listr` r r )r r s "r r r s! __DI__kIZ_r c R # r r r s &&,r r r s \_r c , < V ^8 d QhRS[ RS[RS[/# r )r
r r )r r s "r r r s XXHXX Xr c R # r r r s &&,r r r
s UXr c L < V ^8 d QhRS[ S[, RS[RS[ S[, /# r )r r
r r )r r s "r r r s" ddDNdcddS^N_dr c R # r r r s &&,r r r
s adr c & < \ SV ` ! V3/ VB # )a
Generates speech/audio from the inputs. See the [`TextToAudioPipeline`] documentation for more information.
Args:
text_inputs (`str`, `list[str]`, `ChatType`, or `list[ChatType]`):
One or several texts to generate. If strings or a list of string are passed, this pipeline will
generate the corresponding text. Alternatively, a "chat", in the form of a list of dicts with "role"
and "content" keys, can be passed, or a list of such chats. When chats are passed, the model's chat
template will be used to format them before passing them to the model.
forward_params (`dict`, *optional*):
Parameters passed to the model generation/forward method. `forward_params` are always passed to the
underlying model.
generate_kwargs (`dict`, *optional*):
The dictionary of ad-hoc parametrization of `generate_config` to be used for the generation call. For a
complete overview of generate, check the [following
guide](https://huggingface.co/docs/transformers/en/main_classes/text_generation). `generate_kwargs` are
only passed to the underlying model if the latter is a generative model.
Return:
`AudioOutput` or a list of `AudioOutput`, which is a `TypedDict` with two keys:
- **audio** (`np.ndarray` of shape `(nb_channels, audio_length)`) -- The generated audio waveform.
- **sampling_rate** (`int`) -- The sampling rate of the generated audio waveform.
)r2 r )rE r rp r5 s &&,r r r s 2 w >~>>r c \ V R R4 e V P VR & \ V RR4 e V P VR& V P VR& RV' d TM/ RV' d TM/ /pVf / pVe WAR& / pWV3# )assistant_modelNassistant_tokenizerra rp rq rU )rC r ra r )rE preprocess_paramsrp rq rU paramspostprocess_paramss &&&&& r _sanitize_parameters(TextToAudioPipeline._sanitize_parameters+ s 4*D 1 =151E1EO - . 4. 5 A+/>>OK (595M5MO 1 2
nB /r
$ " '4D 0 1 *<<r sU + + $ ) & 6 QC1 )5 n
( n
r