Upload 12 files

CKPT.yaml

@@ -0,0 +1,4 @@
+# yamllint disable
+end-of-epoch: true
+si-snr: -5.768610138794406
+unixtime: 1708444525.2560909

README.md

@@ -0,0 +1,133 @@
+---
+language: "en"
+thumbnail:
+tags:
+- Source Separation
+- Speech Separation
+- Audio Source Separation
+- Libri3Mix
+- SepFormer
+- Transformer 
+- audio-to-audio 
+- audio-source-separation
+- speechbrain
+license: "apache-2.0"
+datasets:
+- Libri3Mix
+metrics:
+- SI-SNRi
+- SDRi
+
+---
+
+<iframe src="https://ghbtns.com/github-btn.html?user=speechbrain&repo=speechbrain&type=star&count=true&size=large&v=2" frameborder="0" scrolling="0" width="170" height="30" title="GitHub"></iframe>
+<br/><br/>
+
+# SepFormer trained on Libri3Mix
+
+This repository provides all the necessary tools to perform audio source separation with a [SepFormer](https://arxiv.org/abs/2010.13154v2) 
+model, implemented with SpeechBrain, and pretrained on Libri3Mix dataset. For a better experience we encourage you to learn more about
+[SpeechBrain](https://speechbrain.github.io). The model performance is 19.8 dB SI-SNRi on the test set of Libri3Mix dataset.
+
+| Release | Train-Set SI-SNRi | Test-Set SI-SNRi |
+|:-------------:|:--------------:|:--------------:|
+| 16-09-22 | 9.51.0dB | 5.77.4dB |
+
+
+## Install SpeechBrain
+
+First of all, please install SpeechBrain with the following command:
+
+```
+pip install speechbrain
+```
+
+Please notice that we encourage you to read our tutorials and learn more about
+[SpeechBrain](https://speechbrain.github.io).
+
+### Perform source separation on your own audio file
+
+```python
+from speechbrain.pretrained import SepformerSeparation as separator
+import torchaudio
+
+model = separator.from_hparams(source="speechbrain/sepformer-libri3mix", savedir='pretrained_models/sepformer-libri3mix')
+
+est_sources = model.separate_file(path='speechbrain/sepformer-wsj03mix/test_mixture_3spks.wav') 
+
+torchaudio.save("source1hat.wav", est_sources[:, :, 0].detach().cpu(), 8000)
+torchaudio.save("source2hat.wav", est_sources[:, :, 1].detach().cpu(), 8000)
+torchaudio.save("source3hat.wav", est_sources[:, :, 2].detach().cpu(), 8000)
+
+```
+
+The system expects input recordings sampled at 8kHz (single channel).
+If your signal has a different sample rate, resample it (e.g, using torchaudio or sox) before using the interface.
+
+### Inference on GPU
+To perform inference on the GPU, add  `run_opts={"device":"cuda"}`  when calling the `from_hparams` method.
+
+### Training
+The model was trained with SpeechBrain (fc2eabb7).
+To train it from scratch follows these steps:
+1. Clone SpeechBrain:
+```bash
+git clone https://github.com/speechbrain/speechbrain/
+```
+2. Install it:
+```
+cd speechbrain
+pip install -r requirements.txt
+pip install -e .
+```
+
+3. Run Training:
+```
+cd  recipes/LibriMix/separation
+python train.py hparams/sepformer.yaml --data_folder=your_data_folder
+```
+Note: change num_spks to 3 in the yaml file.
+
+
+You can find our training results (models, logs, etc) [here](https://drive.google.com/drive/folders/1DN49LtAs6cq1X0jZ8tRMlh2Pj6AecClz).
+
+### Limitations
+The SpeechBrain team does not provide any warranty on the performance achieved by this model when used on other datasets.
+
+#### Referencing SpeechBrain
+
+```bibtex
+@misc{speechbrain,
+  title={{SpeechBrain}: A General-Purpose Speech Toolkit},
+  author={Mirco Ravanelli and Titouan Parcollet and Peter Plantinga and Aku Rouhe and Samuele Cornell and Loren Lugosch and Cem Subakan and Nauman Dawalatabad and Abdelwahab Heba and Jianyuan Zhong and Ju-Chieh Chou and Sung-Lin Yeh and Szu-Wei Fu and Chien-Feng Liao and Elena Rastorgueva and François Grondin and William Aris and Hwidong Na and Yan Gao and Renato De Mori and Yoshua Bengio},
+  year={2021},
+  eprint={2106.04624},
+  archivePrefix={arXiv},
+  primaryClass={eess.AS},
+  note={arXiv:2106.04624}
+}
+```
+
+
+#### Referencing SepFormer
+```bibtex
+@inproceedings{subakan2021attention,
+      title={Attention is All You Need in Speech Separation}, 
+      author={Cem Subakan and Mirco Ravanelli and Samuele Cornell and Mirko Bronzi and Jianyuan Zhong},
+      year={2021},
+      booktitle={ICASSP 2021}
+}
+
+@misc{subakan2022sepformer
+  author = {Subakan, Cem and Ravanelli, Mirco and Cornell, Samuele and Grondin, Francois and Bronzi, Mirko},
+  title = {On Using Transformers for Speech-Separation},
+  year = {2022},
+  copyright = {arXiv.org perpetual, non-exclusive license}
+}
+
+```
+
+# **About SpeechBrain**
+- Website: https://speechbrain.github.io/
+- Code: https://github.com/speechbrain/speechbrain/
+- HuggingFace: https://huggingface.co/speechbrain/

brain.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33809a026a2c1febce7b03c8aafaee4ddfc851b2c70f180f8c06bf1017f4df5c
+size 46

config.json

@@ -0,0 +1,3 @@
+{
+	  "speechbrain_interface": "SepformerSeparation"
+}

counter.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7688b6ef52555962d008fff894223582c484517cea7da49ee67800adc7fc8866
+size 2

dataloader-TRAIN.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc930d7c1a7c02a920e46b457a2d5b77a54d29739719b5da85a12a7eac2607c5
+size 5

decoder.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a8adecf5da46ffdf91d627cc9c3126057d32424abefb86886a0413224ff5abd
+size 17195

encoder.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f90be7401e7180f0d5b70583527afb50c837e1038675ce5bddf1368cb8cb7c0
+size 17259

hyperparams.yaml

@@ -0,0 +1,175 @@
+# Generated 2024-02-06 from:
+# /mnt3/speechbrain/recipes/LibriMix/separation/hparams/sepformer-libri3mix.yaml
+# yamllint disable
+# ################################
+# Model: SepFormer for source separation
+# https://arxiv.org/abs/2010.13154
+# Dataset : Libri3Mix
+# ################################
+#
+# Basic parameters
+# Seed needs to be set at top of yaml, before objects with parameters are made
+#
+seed: 1234
+__set_seed: !apply:torch.manual_seed [1234]
+
+# Data params
+
+# e.g. '/yourpath/Libri3Mix/train-clean-360/'
+# the data folder is needed even if dynamic mixing is applied
+data_folder: /mnt3/Libri3Mix_48k_own/Libri3Mix/
+
+# This is needed only if dynamic mixing is applied
+base_folder_dm: /yourpath/LibriSpeech/train-clean-360/
+
+experiment_name: sepformer-libri3mix-48k
+output_folder: results/sepformer-libri3mix-48k/1234
+train_log: results/sepformer-libri3mix-48k/1234/train_log.txt
+save_folder: results/sepformer-libri3mix-48k/1234/save
+train_data: results/sepformer-libri3mix-48k/1234/save/libri3mix_train-360.csv
+valid_data: results/sepformer-libri3mix-48k/1234/save/libri3mix_test.csv
+test_data: results/sepformer-libri3mix-48k/1234/save/libri3mix_test.csv
+skip_prep: false
+
+ckpt_interval_minutes: 60
+
+# Experiment params
+precision: fp16 # bf16, fp16 or fp32 # Set it to True for mixed precision
+num_spks: 3
+noprogressbar: false
+save_audio: false # Save estimated sources on disk
+sample_rate: 48000
+
+# Training parameters
+N_epochs: 50
+batch_size: 1
+lr: 0.0005
+clip_grad_norm: 5
+loss_upper_lim: 999999  # this is the upper limit for an acceptable loss
+# if True, the training sequences are cut to a specified length
+limit_training_signal_len: true
+# this is the length of sequences if we choose to limit
+# the signal length of training sequences
+training_signal_len: 100000
+
+# Set it to True to dynamically create mixtures at training time
+dynamic_mixing: false
+use_wham_noise: true
+
+# Parameters for data augmentation
+use_wavedrop: false
+use_speedperturb: true
+use_rand_shift: false
+min_shift: -8000
+max_shift: 8000
+
+speedperturb: !new:speechbrain.lobes.augment.TimeDomainSpecAugment
+  perturb_prob: 1.0
+  drop_freq_prob: 0.0
+  drop_chunk_prob: 0.0
+  sample_rate: 48000
+  speeds: [95, 100, 105]
+
+wavedrop: !new:speechbrain.lobes.augment.TimeDomainSpecAugment
+  perturb_prob: 0.0
+  drop_freq_prob: 1.0
+  drop_chunk_prob: 1.0
+  sample_rate: 48000
+
+# loss thresholding -- this thresholds the training loss
+threshold_byloss: true
+threshold: -30
+
+# Encoder parameters
+N_encoder_out: 256
+out_channels: 256
+kernel_size: 16
+kernel_stride: 8
+d_ffn: 1024
+
+# Dataloader options
+dataloader_opts:
+  batch_size: 1
+  num_workers: 3
+
+
+# Specifying the network
+Encoder: &id004 !new:speechbrain.lobes.models.dual_path.Encoder
+  kernel_size: 16
+  out_channels: 256
+
+
+SBtfintra: &id002 !new:speechbrain.lobes.models.dual_path.SBTransformerBlock
+  num_layers: 8
+  d_model: 256
+  nhead: 8
+  d_ffn: 1024
+  dropout: 0
+  use_positional_encoding: true
+  norm_before: true
+
+SBtfinter: &id003 !new:speechbrain.lobes.models.dual_path.SBTransformerBlock
+  num_layers: 8
+  d_model: 256
+  nhead: 8
+  d_ffn: 1024
+  dropout: 0
+  use_positional_encoding: true
+  norm_before: true
+
+MaskNet: &id006 !new:speechbrain.lobes.models.dual_path.Dual_Path_Model
+
+  num_spks: 3
+  in_channels: 256
+  out_channels: 256
+  num_layers: 2
+  K: 250
+  intra_model: *id002
+  inter_model: *id003
+  norm: ln
+  linear_layer_after_inter_intra: false
+  skip_around_intra: true
+
+Decoder: &id005 !new:speechbrain.lobes.models.dual_path.Decoder
+  in_channels: 256
+  out_channels: 1
+  kernel_size: 16
+  stride: 8
+  bias: false
+
+optimizer: !name:torch.optim.Adam
+  lr: 0.0005
+  weight_decay: 0
+
+loss: !name:speechbrain.nnet.losses.get_si_snr_with_pitwrapper
+
+lr_scheduler: !new:speechbrain.nnet.schedulers.ReduceLROnPlateau
+  factor: 0.5
+  patience: 2
+  dont_halve_until_epoch: 5
+
+epoch_counter: &id007 !new:speechbrain.utils.epoch_loop.EpochCounter
+        # lr_scheduler: !ref <lr_scheduler>
+
+  limit: 50
+
+modules:
+  encoder: *id004
+  decoder: *id005
+  masknet: *id006
+checkpointer: !new:speechbrain.utils.checkpoints.Checkpointer
+  checkpoints_dir: results/sepformer-libri3mix-48k/1234/save
+  recoverables:
+    encoder: *id004
+    decoder: *id005
+    masknet: *id006
+    counter: *id007
+train_logger: !new:speechbrain.utils.train_logger.FileTrainLogger
+  save_file: results/sepformer-libri3mix-48k/1234/train_log.txt
+
+
+pretrainer: !new:speechbrain.utils.parameter_transfer.Pretrainer
+  loadables:
+     encoder: !ref <Encoder>
+     masknet: !ref <MaskNet>
+     decoder: !ref <Decoder>

masknet.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94b4ef6a585db8dbb4227ecaab99e573f1c0f1fe10fda2879f099e36d76de593
+size 113365891

optimizer.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83bc94f9f931c623c0c0aca4be210eab960eb4cb7029237b7ea77915f455c031
+size 206306100

scaler.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e6fcb92f2dfc84b1e44c4aef11d4ff249f17857427a73381f8289240ee90a9a
+size 557