Comments (3)
Hey @Killshot667! Thanks for raising this interesting point. Indeed, distillation has, for the moment, been targeted at single languages.
For distillation, the approach was initially to shrink the model as much as possible while maximizing its performance by training a smaller decoder on a targeted language. The idea is to trade the multilingual capacities of the 32 layers of the decoder for size and speed improvement brought by a smaller decoder (therefore with smaller learning capacities). In this context, two layers appeared to be Pareto optimal. Were we to train on a multilingual dataset, more decoder layers might be needed to enhance learning capacities. Such an adaptation of the student model’s decoder layers can be easily done by changing --decoder_layers
when initializing.
Secondly, note there is nothing restraining a distilled model from having multilingual transcription capacities. First, the encoder is identical to Whisper’s, so robustness in creating a representation of speech for different languages remains unchanged. Secondly, when initializing the student model, we keep Whisper’s vocabulary and start from Whisper input embeddings, coming with inherent multilingual tokens. To this extent, the only thing restraining distil-large-v3 from being multilingual is the dataset it has been distilled on. You could perfectly train, for example, a 4-decoder-layer distilled model on European languages (easily done by pseudo-labeling each set with the correct --language
flag as explained in language-mixing). Actually, language-mixing experiments showed that mixing close languages could improve the model’s performance.
from distil-whisper.
I too have the same question.
@sanchit-gandhi Please try distilling whisper-small on kathbath dataset and share the results.
from distil-whisper.
Hey @Killshot667 - that's a great question, and super sorry for the late reply here! I'll defer to @eustlb, who has been running some preliminary experiments on distilling Whisper jointly for French and Spanish. You can read about the initial results and how to reproduce them on the README here: https://github.com/huggingface/distil-whisper/tree/main/training#3-language-mixing
from distil-whisper.
Related Issues (20)
- large-v2 for english lost voice to text HOT 1
- Finetuning on which model? HOT 1
- Resuming training fails HOT 3
- [Issue] latest run_pseudo_labelling.py
- Quantize distil-whisper?
- perceptually faster inference through pre-completion inference of audio
- RuntimeError: User specified an unsupported autocast device_type 'mps'
- question about when to apply WER threshold filtering strategy with concatenated audio
- Problems in concatenate_dataset
- How to set the target language for examples in README? HOT 7
- Unable to reproduce results from the paper HOT 6
- Unable to set concatenate_audio parameter to False in run_pseudo_labelling.py
- Discrepancy on WER benchmark result in Tedlium dataset HOT 1
- Repository Not Found for url
- ZeroDivisionError: division by zero
- Pseudo-labelling librispeech_asr (train.360): KeyError `train-360` when not streaming. HOT 1
- Training README datasets table: text column and id column HOT 4
- Voxpopuli text column "raw_text" HF dataset card shows empty string. HOT 1
- any executable script for running on custom data/given dataset HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from distil-whisper.