Model reference · open weights
DCCRNet_Libri1Mix_enhsingle_16k is an open-weight audio or speech model from JorisCos. AxForge deploys and operates it for you on dedicated EU-owned hardware — with the licence handled where one is required.
Available as managed deployment — configured and operated for you on dedicated EU hardware, quoted per deployment.
What it is
| Maker | JorisCos |
|---|---|
| Type | Audio & music |
| Task | Audio→audio |
| Runs with | asteroid |
| Released | 2022-03-02 |
| Popularity | 995 downloads / month |
| Licence | Open weights |
About
JorisCos/DCCRNet_Libri1Mix_enhsignle_16kDescription:
This model was trained by Joris Cosentino using the librimix recipe in Asteroid.
It was trained on the enh_single task of the Libri1Mix dataset.
Training config:
data:
n_src: 1
sample_rate: 16000
segment: 3
task: enh_single
train_dir: data/wav16k/min/train-360
valid_dir: data/wav16k/min/dev
filterbank:
stft_kernel_size: 400
stft_n_filters: 512
stft_stride: 100
masknet:
architecture: DCCRN-CL
n_src: 1
optim:
lr: 0.001
optimizer: adam
weight_decay: 1.0e-05
training:
batch_size: 12
early_stop: true
epochs: 200
gradient_clipping: 5
half_lr: true
num_workers: 4
Results:
On Libri1Mix min test set :
si_sdr: 13.329767398333798
si_sdr_imp: 9.879986092474098
sdr: 13.87279932997016
sdr_imp: 10.370136530757103
sir: Infinity
sir_imp: NaN
sar: 13.87279932997016
sar_imp: 10.370136530757103
stoi: 0.9140907015623948
stoi_imp: 0.11817087802185405
License notice:
This work "DCCRNet_Libri1Mix_enhsignle_16k" is a derivative of LibriSpeech ASR corpus by Vassil Panayotov, used under CC BY 4.0; of The WSJ0 Hipster Ambient Mixtures dataset by Whisper.ai, used under CC BY-NC 4.0 (Research only). "DCCRNet_Libri1Mix_enhsignle_16k" is licensed under Attribution-ShareAlike 3.0 Unported by Joris Cosentino
From the published model card. Full card on the HuggingFace links in the sidebar.
How it works
Using it via the API
Once AxForge deploys dccrnet-libri1mix-enhsingle-16k for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (dccrnet-libri1mix-enhsingle-16k below is illustrative; you get the exact model name on deployment.)
$ curl -sS https://api.axforge.ai/v1/audio/transcriptions \ -H "Authorization: Bearer $AXFORGE_API_KEY" \ -F model="dccrnet-libri1mix-enhsingle-16k" -F file=@audio.mp3
Create an account — your API key is available in the console. 5M tokens/month currently included with every new account at launch.