:j5vddlZddlZddlmZmZddlmZddlmZddl m Z m Z m Z ddl Z ddlZddlmZddlmZmZmZgZdej,e j.e j0j2zZed d ezZd ZGd d e j<j>Z Gdde j<j>Z!GddeZ"GddeZ#Gdde j<j>e"Z$Gdde#Z%eGddZ&e&deeddddd d!d"d#d$d%d& Z'd'e'_(y)(N)ABCabstractmethod) dataclass)partial)CallableListTuple) module_utils)emformer_rnnt_baseRNNTRNNTBeamSearch( g?ctj||tjkD||tjkD<||tjktjz ||tjk<|SN)torchlogmathexs w/home/mcse/projects/srt_converter/srt-converter-venv/lib/python3.12/site-packages/torchaudio/pipelines/rnnt_pipeline.py_piecewise_linear_logrsSIIaDFF m,Aa$&&jMqDFF{^dff,Aa466kN Hc$eZdZfdZdZxZS)_FunctionalModulec0t|||_yr)super__init__ functional)selfr __class__s rrz_FunctionalModule.__init__s $rc$|j|Sr)r r!inputs rforwardz_FunctionalModule.forwardsu%%r__name__ __module__ __qualname__rr& __classcell__r"s@rrrs %&rrc$eZdZfdZdZxZS)_GlobalStatsNormalizationcHt|t|5}tj|j }ddd|j dtjd|j dtj|dy#1swYZxYw)Nmean invstddev) rropenjsonloadsreadregister_bufferrtensor)r!global_stats_pathfblobr"s rrz"_GlobalStatsNormalization.__init__$s  # $ (::affh'D ( VU\\$v,%?@ [%,,tK7H*IJ  ( (s $BB!c:||jz |jzSr)r0r1r$s rr&z!_GlobalStatsNormalization.forward-s !T^^33rr'r,s@rr.r.#sK4rr.cleZdZedej deej ej ffdZy)_FeatureExtractorr%returncy)XGenerates features and length output from the given input tensor. Args: input (torch.Tensor): input tensor. Returns: (torch.Tensor, torch.Tensor): torch.Tensor: Features, with shape `(length, *)`. torch.Tensor: Length, with shape `(1,)`. Nr$s r__call__z_FeatureExtractor.__call__2rN)r(r)r*rrTensorr rBrArrr=r=1s8  ell  uU\\5<<5O/P    rr=c,eZdZedeedefdZy)_TokenProcessortokensr>c y)zDecodes given list of tokens to text sequence. Args: tokens (List[int]): list of tokens to decode. Returns: str: Decoded text sequence. NrA)r!rGkwargss rrBz_TokenProcessor.__call__CrCrN)r(r)r*rrintstrrBrArrrFrFBs&  tCy  s    rrFceZdZdZdej j ddffd Zdejde ejejffdZ xZ S)_ModuleFeatureExtractorz``torch.nn.Module``-based feature extraction pipeline. Args: pipeline (torch.nn.Module): module that implements feature extraction logic. pipeliner>Nc0t|||_yr)rrrN)r!rNr"s rrz _ModuleFeatureExtractor.__init__Ws   rr%cr|j|}tj|jdg}||fS)r@r)rNrr7shape)r!r%featureslengths rr&z_ModuleFeatureExtractor.forward[s7=='x~~a012r) r(r)r*__doc__rnnModulerrDr r&r+r,s@rrMrMPsL !!T! U\\ eELL%,,4N.O rrMc<eZdZdZdeddfdZd deededefdZ y) _SentencePieceTokenProcessorztSentencePiece-model-based token processor. Args: sp_model_path (str): path to SentencePiece model. sp_model_pathr>Nctjds tdddl}|j ||_|j j |j j|j jh|_ y)N sentencepiecez2SentencePiece is not available. Please install it.r) model_file) r is_module_available RuntimeErrorr[SentencePieceProcessorsp_modelunk_ideos_idpad_idpost_process_remove_list)r!rYspms rrz%_SentencePieceTokenProcessor.__init__tsn//@ST T#22m2L MM " MM " MM ") %rrGlstripc|ddDcgc]}||jvs|}}dj|jj|j dd}|r|j S|Scc}w)aXDecodes given list of tokens to text sequence. Args: tokens (List[int]): list of tokens to decode. lstrip (bool, optional): if ``True``, returns text sequence with leading whitespace removed. (Default: ``True``). Returns: str: Decoded text sequence. Nu▁ )rdjoinr` id_to_piecereplacerf)r!rGrf token_indexfiltered_hypo_tokens output_strings rrBz%_SentencePieceTokenProcessor.__call__s~,2!": 'DLiLi9iK   9 9:N OPXXYacfg  '') )  s A0A0)T) r(r)r*rTrKrrrJboolrBrArrrXrXms8  c  d  !tCy!$!#!rrXceZdZUdZGddeZGddeZee d<e ge fe d<ee d<ee d <e e d <e e d <e e d <e e d <e e d<e e d<e e d<e e d<de fdZ ede fdZede fdZede fdZede fdZede fdZede fdZdefdZdefdZdefdZdefdZy) RNNTBundleuDataclass that bundles components for performing automatic speech recognition (ASR, speech-to-text) inference with an RNN-T model. More specifically, the class provides methods that produce the featurization pipeline, decoder wrapping the specified RNN-T model, and output token post-processor that together constitute a complete end-to-end ASR inference pipeline that produces a text sequence given a raw waveform. It can support non-streaming (full-context) inference as well as streaming inference. Users should not directly instantiate objects of this class; rather, users should use the instances (representing pre-trained models) that exist within the module, e.g. :data:`torchaudio.pipelines.EMFORMER_RNNT_BASE_LIBRISPEECH`. Example >>> import torchaudio >>> from torchaudio.pipelines import EMFORMER_RNNT_BASE_LIBRISPEECH >>> import torch >>> >>> # Non-streaming inference. >>> # Build feature extractor, decoder with RNN-T model, and token processor. >>> feature_extractor = EMFORMER_RNNT_BASE_LIBRISPEECH.get_feature_extractor() 100%|███████████████████████████████| 3.81k/3.81k [00:00<00:00, 4.22MB/s] >>> decoder = EMFORMER_RNNT_BASE_LIBRISPEECH.get_decoder() Downloading: "https://download.pytorch.org/torchaudio/models/emformer_rnnt_base_librispeech.pt" 100%|███████████████████████████████| 293M/293M [00:07<00:00, 42.1MB/s] >>> token_processor = EMFORMER_RNNT_BASE_LIBRISPEECH.get_token_processor() 100%|███████████████████████████████| 295k/295k [00:00<00:00, 25.4MB/s] >>> >>> # Instantiate LibriSpeech dataset; retrieve waveform for first sample. >>> dataset = torchaudio.datasets.LIBRISPEECH("/home/librispeech", url="test-clean") >>> waveform = next(iter(dataset))[0].squeeze() >>> >>> with torch.no_grad(): >>> # Produce mel-scale spectrogram features. >>> features, length = feature_extractor(waveform) >>> >>> # Generate top-10 hypotheses. >>> hypotheses = decoder(features, length, 10) >>> >>> # For top hypothesis, convert predicted tokens to text. >>> text = token_processor(hypotheses[0][0]) >>> print(text) he hoped there would be stew for dinner turnips and carrots and bruised potatoes and fat mutton pieces to [...] >>> >>> >>> # Streaming inference. >>> hop_length = EMFORMER_RNNT_BASE_LIBRISPEECH.hop_length >>> num_samples_segment = EMFORMER_RNNT_BASE_LIBRISPEECH.segment_length * hop_length >>> num_samples_segment_right_context = ( >>> num_samples_segment + EMFORMER_RNNT_BASE_LIBRISPEECH.right_context_length * hop_length >>> ) >>> >>> # Build streaming inference feature extractor. >>> streaming_feature_extractor = EMFORMER_RNNT_BASE_LIBRISPEECH.get_streaming_feature_extractor() >>> >>> # Process same waveform as before, this time sequentially across overlapping segments >>> # to simulate streaming inference. Note the usage of ``streaming_feature_extractor`` and ``decoder.infer``. >>> state, hypothesis = None, None >>> for idx in range(0, len(waveform), num_samples_segment): >>> segment = waveform[idx: idx + num_samples_segment_right_context] >>> segment = torch.nn.functional.pad(segment, (0, num_samples_segment_right_context - len(segment))) >>> with torch.no_grad(): >>> features, length = streaming_feature_extractor(segment) >>> hypotheses, state = decoder.infer(features, length, 10, state=state, hypothesis=hypothesis) >>> hypothesis = hypotheses[0] >>> transcript = token_processor(hypothesis[0]) >>> if transcript: >>> print(transcript, end=" ", flush=True) he hoped there would be stew for dinner turn ips and car rots and bru 'd oes and fat mut ton pieces to [...] ceZdZdZy)RNNTBundle.FeatureExtractorz:Interface of the feature extraction part of RNN-T pipelineNr(r)r*rTrArrFeatureExtractorrusHrrwceZdZdZy)RNNTBundle.TokenProcessorz7Interface of the token processor part of RNN-T pipelineNrvrArrTokenProcessorrysErrz _rnnt_path_rnnt_factory_func_global_stats_path_sp_model_path_right_padding_blank _sample_rate_n_fft_n_mels _hop_length_segment_length_right_context_lengthr>c|j}tjj|j}t j |}|j||j|Sr) r| torchaudioutilsdownload_assetr{rloadload_state_dicteval)r!modelpath state_dicts r _get_modelzRNNTBundle._get_modelsT'')..t?ZZ%  j)  rc|jS)zSSample rate (in cycles per second) of input waveforms. :type: int )rr!s r sample_ratezRNNTBundle.sample_rates    rc|jS)z7Size of FFT window to use. :type: int )rrs rn_fftzRNNTBundle.n_ffts {{rc|jS)z`Number of mel spectrogram features to extract from input waveforms. :type: int )rrs rn_melszRNNTBundle.n_mels s ||rc|jS)zdNumber of samples between successive frames in input expected by model. :type: int )rrs r hop_lengthzRNNTBundle.hop_lengths rc|jS)zTNumber of frames in segment in input expected by model. :type: int )rrs rsegment_lengthzRNNTBundle.segment_lengths ###rc|jS)zcNumber of frames in right contextual block in input expected by model. :type: int )rrs rright_context_lengthzRNNTBundle.right_context_length%s )))rcN|j}t||jS)zOConstructs RNN-T decoder. Returns: RNNTBeamSearch )rr r)r!rs r get_decoderzRNNTBundle.get_decoder-s! !eT[[11rc tjjj}t t j jtjjjjjjtdtdt|tfdS)zzConstructs feature extractor for non-streaming (full-context) ASR. Returns: FeatureExtractor rrrrc&|jddSNrhr transposers rz2RNNTBundle.get_feature_extractor..BAKK1,=rc&t|tzSrr_gainrs rrz2RNNTBundle.get_feature_extractor..C,A!e),Lrcttjjj|dddjfS)Nr)rrUr padr)rr!s rrz2RNNTBundle.get_feature_extractor..Es.EHH,?,?,C,CA1aQUQdQdGe,frrrrr}rMrrU Sequential transformsMelSpectrogramrrrrrr.r! local_paths` rget_feature_extractorz RNNTBundle.get_feature_extractor6s  %%44T5L5LM & HH  %%44 $ 0 0 4;;cgcrcr5""=>!"LM)*5!"fg   rc vtjj|j}t t j jtjj|j|j|j|jtdtdt|S)zvConstructs feature extractor for streaming (simultaneous) ASR. Returns: FeatureExtractor rc&|jddSrrrs rrz.Urrc&t|tzSrrrs rrz.Vrrrrs rget_streaming_feature_extractorz*RNNTBundle.get_streaming_feature_extractorIs  %%44T5L5LM & HH  %%44 $ 0 0 4;;cgcrcr5""=>!"LM)*5    rcjtjj|j}t |S)zQConstructs token processor. Returns: TokenProcessor )rrrr~rXrs rget_token_processorzRNNTBundle.get_token_processor[s+  %%44T5H5HI +J77rN)r(r)r*rTr=rwrFrzrK__annotations__rr rJrpropertyrrrrrrr rrrrrArrrsrssUFPI,IFFO T** K K LD!S!!s C  $$$*c**2^2 '7 & 1A $8^8rrsz(models/emformer_rnnt_base_librispeech.pti) num_symbolsz2pipeline-assets/global_stats_rnnt_librispeech.jsonz.pipeline-assets/spm_bpe_4096_librispeech.modelii>iP) r{r|r}r~rrrrrrrraASR pipeline based on Emformer-RNNT, pretrained on *LibriSpeech* dataset :cite:`7178964`, capable of performing both streaming and non-streaming inference. The underlying model is constructed by :py:func:`torchaudio.models.emformer_rnnt_base` and utilizes weights trained on LibriSpeech using training script ``train.py`` `here `__ with default arguments. Please refer to :py:class:`RNNTBundle` for usage instructions. ))r3rabcrr dataclassesr functoolsrtypingrrr rrtorchaudio._internalr torchaudio.modelsr r r __all__log10iinfoint16max_decibelpowrrrUrVrr.r=rFrMrXrsEMFORMER_RNNT_BASE_LIBRISPEECHrTrArrrs4 #!(( -FF  JDJJ{u{{5;;7;;< < Bx  && 4 4  "  c   ehhoo/@ :(!?(!V I8I8 I8X",91tDKC    " *&r