lgris commited on
Commit
61d69f4
1 Parent(s): 761b97f

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +25 -0
README.md ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language: pt
3
+ tags:
4
+ - speech
5
+ license: apache-2.0
6
+ ---
7
+
8
+ # DistilXLSR-53 for BP
9
+ [DistilXLSR-53 for BP: DistilHuBERT applied to Wav2vec XLSR-53 for Brazilian Portuguese](https://github.com/s3prl/s3prl/tree/master/s3prl/upstream/distiller)
10
+
11
+ The base model pretrained on 16kHz sampled speech audio. When using the model make sure that your speech input is also sampled at 16Khz.
12
+
13
+ **Note**: This model does not have a tokenizer as it was pretrained on audio alone. In order to use this model **speech recognition**, a tokenizer should be created and the model should be fine-tuned on labeled text data. Check out [this blog](https://huggingface.co/blog/fine-tune-wav2vec2-english) for more in-detail explanation of how to fine-tune the model.
14
+ Paper: [DistilHuBERT: Speech Representation Learning by Layer-wise Distillation of Hidden-unit BERT](https://arxiv.org/abs/2110.01900)
15
+ Authors: Heng-Jui Chang, Shu-wen Yang, Hung-yi Lee
16
+
17
+ **Note 2**: The XLSR-53 model was distilled using [Brazilian Portuguese Datasets](https://huggingface.co/lgris/bp400-xlsr) for test purposes. The dataset is quite small to perform such task (the performance might not be so good as the [original work](https://arxiv.org/abs/2110.01900)).
18
+
19
+
20
+ **Abstract**
21
+
22
+ Self-supervised speech representation learning methods like wav2vec 2.0 and Hidden-unit BERT (HuBERT) leverage unlabeled speech data for pre-training and offer good representations for numerous speech processing tasks. Despite the success of these methods, they require large memory and high pre-training costs, making them inaccessible for researchers in academia and small companies. Therefore, this paper introduces DistilHuBERT, a novel multi-task learning framework to distill hidden representations from a HuBERT model directly. This method reduces HuBERT's size by 75% and 73% faster while retaining most performance in ten different tasks. Moreover, DistilHuBERT required little training time and data, opening the possibilities of pre-training personal and on-device SSL models for speech.
23
+
24
+ # Usage
25
+ See [this blog](https://huggingface.co/blog/fine-tune-wav2vec2-english) for more information on how to fine-tune the model.