# How to use an unsupported Beam Search decoder in ASR Pipeline?

**URL:** <https://discuss.huggingface.co/t/how-to-use-an-unsupported-beam-search-decoder-in-asr-pipeline/49506>\
**Category:** 🤗Transformers\
**Created:** [August 4, 2023, 5:27pm UTC](https://discuss.huggingface.co/t/how-to-use-an-unsupported-beam-search-decoder-in-asr-pipeline/49506 "2023-08-04T17:27:02Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![jlonsako](https://avatars.discourse-cdn.com/v4/letter/j/7c8e57/32.png) [@jlonsako](https://discuss.huggingface.co/u/jlonsako)\
**Post date:** [August 4, 2023, 5:27pm UTC](https://discuss.huggingface.co/t/how-to-use-an-unsupported-beam-search-decoder-in-asr-pipeline/49506/1 "2023-08-04T17:27:02Z")

</div>

I am using Meta’s new MMS model alongside a language model developed by Meta to transcribe some long form Amharic audio. As you can see in the code from [this space](https://huggingface.co/spaces/mms-meta/MMS/blob/main/asr.py), a beam-search-decoder is built from ‘torchaudio.models.ctc\_decoder’. Because I want to utilize chunking and striding found in the [ASR pipeline from the Transformers library](https://github.com/huggingface/transformers/blob/main/src/transformers/pipelines/automatic_speech_recognition.py), I have been trying to use this same decoder via the ASR pipeline by passing it in as an attribute. For reference, this is the code used to build the decoder:

```auto
from torchaudio.models.decoder import ctc_decoder
import json
from huggingface_hub import hf_hub_download

lm_decoding_config = {}
lm_decoding_configfile = hf_hub_download(
    repo_id="facebook/mms-cclms",
    filename="decoding_config.json",
    subfolder="mms-1b-all",
)

with open(lm_decoding_configfile) as f:
    lm_decoding_config = json.loads(f.read())

# allow language model decoding for "eng"

decoding_config = lm_decoding_config["eng"]

lm_file = hf_hub_download(
    repo_id="facebook/mms-cclms",
    filename=decoding_config["lmfile"].rsplit("/", 1)[1],
    subfolder=decoding_config["lmfile"].rsplit("/", 1)[0],
)
token_file = hf_hub_download(
    repo_id="facebook/mms-cclms",
    filename=decoding_config["tokensfile"].rsplit("/", 1)[1],
    subfolder=decoding_config["tokensfile"].rsplit("/", 1)[0],
)
lexicon_file = None
if decoding_config["lexiconfile"] is not None:
    lexicon_file = hf_hub_download(
        repo_id="facebook/mms-cclms",
        filename=decoding_config["lexiconfile"].rsplit("/", 1)[1],
        subfolder=decoding_config["lexiconfile"].rsplit("/", 1)[0],
    )

beam_search_decoder = ctc_decoder(
    lexicon=lexicon_file,
    tokens=token_file,
    lm=lm_file,
    nbest=1,
    beam_size=500,
    beam_size_token=50,
    lm_weight=float(decoding_config["lmweight"]),
    word_score=float(decoding_config["wordscore"]),
    sil_score=float(decoding_config["silweight"]),
    blank_token="<s>",
)

```

Because this decoder is not of type ‘BeamSearchDecoderCTC’ from ‘pyctcdecode’, it is not supported by the ASR pipeline. My question is this: what is the wisest way to make use of the language models developed by Meta in this pipeline? Should I try to build a Wav2Vec2ProcessorWithLM using the 5gram.bin file provided for the CC LM, or would it be prudent to add support for torchaudio.models.ctc\_decoder? Or is there another wise option to go after? For now I have jerryrigged the ASR pipeline to use ctc\_decoder instead of BeamSearchDecoderCTC, but that doesn’t seem like a wise long-term solution.