--- tags: - espnet - audio - automatic-speech-recognition language: noinfo datasets: - speechcommands license: cc-by-4.0 --- ## ESPnet2 ASR model ### `pyf98/speechcommands_35commands_conformer` This model was trained by Yifan Peng using speechcommands recipe in [espnet](https://github.com/espnet/espnet/). ### Demo: How to use in ESPnet2 ```bash cd espnet git checkout bf523b70cae8300da004b41ec6a0d1b57c7ae8bb pip install -e . cd egs2/speechcommands/asr1 ./run.sh --skip_data_prep false --skip_train true --download_model pyf98/speechcommands_35commands_conformer ``` # RESULTS ## Environments - date: `Tue Dec 28 20:39:29 EST 2021` - python version: `3.9.7 (default, Sep 16 2021, 13:09:58) [GCC 7.5.0]` - espnet version: `espnet 0.10.5a1` - pytorch version: `pytorch 1.9.0` - Git hash: `bf523b70cae8300da004b41ec6a0d1b57c7ae8bb` - Commit date: `Tue Dec 28 14:53:22 2021 -0500` ## asr_35commands_conformer_noBatchNorm_warmup5k_lr2e-4_accum3_conv15_5speeds ### WER |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err| |---|---|---|---|---|---|---|---|---| |infer/dev|9981|9981|97.4|2.6|0.0|0.0|2.6|2.6| |infer/test|11005|11005|97.5|2.5|0.0|0.0|2.5|2.5| ### CER |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err| |---|---|---|---|---|---|---|---|---| |infer/dev|9981|38699|98.2|1.3|0.5|0.7|2.5|2.6| |infer/test|11005|42682|98.3|1.2|0.5|0.6|2.3|2.5| ### TER |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err| |---|---|---|---|---|---|---|---|---| ## ASR config
expand ``` config: conf/train_asr_conformer_noBatchNorm.yaml print_config: false log_level: INFO dry_run: false iterator_type: sequence output_dir: exp/asr_35commands_conformer_noBatchNorm_warmup5k_lr2e-4_accum3_conv15_5speeds ngpu: 1 seed: 0 num_workers: 1 num_att_plot: 3 dist_backend: nccl dist_init_method: env:// dist_world_size: null dist_rank: null local_rank: 0 dist_master_addr: null dist_master_port: null dist_launcher: null multiprocessing_distributed: false unused_parameters: false sharded_ddp: false cudnn_enabled: true cudnn_benchmark: false cudnn_deterministic: true collect_stats: false write_collected_feats: false max_epoch: 150 patience: null val_scheduler_criterion: - valid - loss early_stopping_criterion: - valid - loss - min best_model_criterion: - - valid - loss - min - - valid - acc - max keep_nbest_models: 10 grad_clip: 5.0 grad_clip_type: 2.0 grad_noise: false accum_grad: 3 no_forward_run: false resume: true train_dtype: float32 use_amp: false log_interval: null use_tensorboard: true use_wandb: false wandb_project: null wandb_id: null wandb_entity: null wandb_name: null wandb_model_log_interval: -1 detect_anomaly: false pretrain_path: null init_param: [] ignore_init_mismatch: false freeze_param: [] num_iters_per_epoch: null batch_size: 20 valid_batch_size: null batch_bins: 4000000 valid_batch_bins: null train_shape_file: - exp/asr_stats_fbank_pitch_word_sp/train/speech_shape - exp/asr_stats_fbank_pitch_word_sp/train/text_shape.word valid_shape_file: - exp/asr_stats_fbank_pitch_word_sp/valid/speech_shape - exp/asr_stats_fbank_pitch_word_sp/valid/text_shape.word batch_type: numel valid_batch_type: null fold_length: - 800 - 150 sort_in_batch: descending sort_batch: descending multiple_iterator: false chunk_length: 500 chunk_shift_ratio: 0.5 num_cache_chunks: 1024 train_data_path_and_name_and_type: - - dump/fbank_pitch/train_sp/feats.scp - speech - kaldi_ark - - dump/fbank_pitch/train_sp/text - text - text valid_data_path_and_name_and_type: - - dump/fbank_pitch/dev/feats.scp - speech - kaldi_ark - - dump/fbank_pitch/dev/text - text - text allow_variable_data_keys: false max_cache_size: 0.0 max_cache_fd: 32 valid_max_cache_size: null optim: adam optim_conf: lr: 0.0002 scheduler: warmuplr scheduler_conf: warmup_steps: 5000 token_list: - - - zero - five - 'yes' - seven - nine - one - down - 'no' - stop - two - go - six - 'on' - left - eight - right - 'off' - three - four - up - house - wow - dog - marvin - bird - cat - happy - sheila - bed - tree - backward - visual - learn - follow - forward - init: null input_size: 83 ctc_conf: dropout_rate: 0.0 ctc_type: builtin reduce: true ignore_nan_grad: true model_conf: ctc_weight: 0.0 lsm_weight: 0.1 length_normalized_loss: false use_preprocessor: true token_type: word bpemodel: null non_linguistic_symbols: null cleaner: null g2p: null speech_volume_normalize: null rir_scp: null rir_apply_prob: 1.0 noise_scp: null noise_apply_prob: 1.0 noise_db_range: '13_15' frontend: null frontend_conf: {} specaug: specaug specaug_conf: apply_time_warp: true time_warp_window: 5 time_warp_mode: bicubic apply_freq_mask: true freq_mask_width_range: - 0 - 30 num_freq_mask: 2 apply_time_mask: true time_mask_width_range: - 0 - 40 num_time_mask: 2 normalize: global_mvn normalize_conf: stats_file: exp/asr_stats_fbank_pitch_word_sp/train/feats_stats.npz preencoder: null preencoder_conf: {} encoder: conformer encoder_conf: output_size: 256 attention_heads: 4 linear_units: 2048 num_blocks: 12 dropout_rate: 0.1 positional_dropout_rate: 0.1 attention_dropout_rate: 0.1 input_layer: conv2d normalize_before: true macaron_style: true rel_pos_type: legacy pos_enc_layer_type: rel_pos selfattention_layer_type: rel_selfattn activation_type: swish use_cnn_module: true cnn_module_kernel: 15 postencoder: null postencoder_conf: {} decoder: transformer decoder_conf: attention_heads: 4 linear_units: 2048 num_blocks: 6 dropout_rate: 0.1 positional_dropout_rate: 0.1 self_attention_dropout_rate: 0.1 src_attention_dropout_rate: 0.1 required: - output_dir - token_list version: 0.10.3a3 distributed: false ```
### Citing ESPnet ```BibTex @inproceedings{watanabe2018espnet, author={Shinji Watanabe and Takaaki Hori and Shigeki Karita and Tomoki Hayashi and Jiro Nishitoba and Yuya Unno and Nelson Yalta and Jahn Heymann and Matthew Wiesner and Nanxin Chen and Adithya Renduchintala and Tsubasa Ochiai}, title={{ESPnet}: End-to-End Speech Processing Toolkit}, year={2018}, booktitle={Proceedings of Interspeech}, pages={2207--2211}, doi={10.21437/Interspeech.2018-1456}, url={http://dx.doi.org/10.21437/Interspeech.2018-1456} } ``` or arXiv: ```bibtex @misc{watanabe2018espnet, title={ESPnet: End-to-End Speech Processing Toolkit}, author={Shinji Watanabe and Takaaki Hori and Shigeki Karita and Tomoki Hayashi and Jiro Nishitoba and Yuya Unno and Nelson Yalta and Jahn Heymann and Matthew Wiesner and Nanxin Chen and Adithya Renduchintala and Tsubasa Ochiai}, year={2018}, eprint={1804.00015}, archivePrefix={arXiv}, primaryClass={cs.CL} } ```