by facebook
Open source · 2M downloads · 213 likes
The W2v-BERT 2.0 model is an audio encoder based on a Conformer architecture, specifically designed to process and understand speech. Pre-trained on 4.5 million hours of unlabeled audio data spanning over 143 languages, it excels at extracting rich, multilingual representations from vocal signals. While it requires fine-tuning for specific tasks such as automatic speech recognition (ASR) or audio classification, it can already be used to generate high-quality audio embeddings from its top layer. Its integration into the Seamless Communication models highlights its versatility for multilingual communication applications. What sets it apart is its ability to capture diverse linguistic nuances while remaining effective for low-resource languages.
We are open-sourcing our Conformer-based W2v-BERT 2.0 speech encoder as described in Section 3.2.1 of the paper, which is at the core of our Seamless models.
This model was pre-trained on 4.5M hours of unlabeled audio data covering more than 143 languages. It requires finetuning to be used for downstream tasks such as Automatic Speech Recognition (ASR), or Audio Classification.
| Model Name | #params | checkpoint |
|---|---|---|
| W2v-BERT 2.0 | 600M | checkpoint |
This model and its training are supported by 🤗 Transformers, more on it in the docs.
This is a bare checkpoint without any modeling head, and thus requires finetuning to be used for downstream tasks such as ASR. You can however use it to extract audio embeddings from the top layer with this code snippet:
from transformers import AutoFeatureExtractor, Wav2Vec2BertModel
import torch
from datasets import load_dataset
dataset = load_dataset("hf-internal-testing/librispeech_asr_demo", "clean", split="validation")
dataset = dataset.sort("id")
sampling_rate = dataset.features["audio"].sampling_rate
processor = AutoProcessor.from_pretrained("facebook/w2v-bert-2.0")
model = Wav2Vec2BertModel.from_pretrained("facebook/w2v-bert-2.0")
# audio file is decoded on the fly
inputs = processor(dataset[0]["audio"]["array"], sampling_rate=sampling_rate, return_tensors="pt")
with torch.no_grad():
outputs = model(**inputs)
To learn more about the model use, refer to the following resources:
This model can be used in Seamless Communication, where it was released.
Here's how to make a forward pass through the voice encoder, after having completed the installation steps:
import torch
from fairseq2.data.audio import AudioDecoder, WaveformToFbankConverter
from fairseq2.memory import MemoryBlock
from fairseq2.nn.padding import get_seqs_and_padding_mask
from pathlib import Path
from seamless_communication.models.conformer_shaw import load_conformer_shaw_model
audio_wav_path, device, dtype = ...
audio_decoder = AudioDecoder(dtype=torch.float32, device=device)
fbank_converter = WaveformToFbankConverter(
num_mel_bins=80,
waveform_scale=2**15,
channel_last=True,
standardize=True,
device=device,
dtype=dtype,
)
collater = Collater(pad_value=1)
model = load_conformer_shaw_model("conformer_shaw", device=device, dtype=dtype)
model.eval()
with Path(audio_wav_path).open("rb") as fb:
block = MemoryBlock(fb.read())
decoded_audio = audio_decoder(block)
src = collater(fbank_converter(decoded_audio))["fbank"]
seqs, padding_mask = get_seqs_and_padding_mask(src)
with torch.inference_mode():
seqs, padding_mask = model.encoder_frontend(seqs, padding_mask)
seqs, padding_mask = model.encoder(seqs, padding_mask)