Update model

README.md

@@ -35,12 +35,12 @@ cd egs2/librimix/enh_diar1_2
 <details><summary>expand</summary>
 
 ```
-config: conf/tuning/train_diar_enh_convtasnet_concat_feats.yaml
+config: conf/tuning/train_diar_enh_convtasnet_2.yaml
 print_config: false
 log_level: INFO
 dry_run: false
 iterator_type: chunk
-output_dir: exp/diar_enh_train_diar_enh_convtasnet_concat_feats_raw
+output_dir: exp/diar_enh_train_diar_enh_convtasnet_2_raw
 ngpu: 1
 seed: 0
 num_workers: 4
@@ -262,7 +262,7 @@ enh_mask_module_conf:
     max_num_spk: 3
     mask_nonlinear: relu
     bottleneck_dim: 128
-frontend: default
+frontend: null
 frontend_conf: {}
 specaug: null
 specaug_conf: {}
@@ -288,10 +288,8 @@ st_extra_asr_decoder: rnn
 st_extra_asr_decoder_conf: {}
 st_extra_mt_decoder: rnn
 st_extra_mt_decoder_conf: {}
-diar_frontend: default
-diar_frontend_conf:
-    hop_length: 64
-    fs: 8000
+diar_frontend: null
+diar_frontend_conf: {}
 diar_specaug: null
 diar_specaug_conf: {}
 diar_normalize: utterance_mvn

exp/diar_enh_train_diar_enh_convtasnet_2_raw/DIAR_RESULTS.md

@@ -0,0 +1,55 @@
+<!-- Generated by scripts/utils/show_diar_result.sh -->
+# RESULTS
+## Environments
+- date: `Mon Mar 13 03:12:32 EDT 2023`
+- python version: `3.9.12 (main, Apr  5 2022, 06:56:58)  [GCC 7.5.0]`
+- espnet version: `espnet 202205`
+- pytorch version: `pytorch 1.8.1+cu111`
+- Git hash: `f2778f798b76e102602078213a21b63aa592be70`
+  - Commit date: `Sun Jul 10 23:45:31 2022 +0000`
+
+## diar_enh_train_diar_enh_convtasnet_2_raw
+### DER
+diarized_enhanced_test_decode_diar_enh_wo_diar
+|threshold_median_collar|DER|
+|---|---|
+|result_th0.3_med11_collar0.0|6.19|
+|result_th0.3_med1_collar0.0|6.23|
+|result_th0.4_med11_collar0.0|6.05|
+|result_th0.4_med1_collar0.0|6.13|
+|result_th0.5_med11_collar0.0|6.20|
+|result_th0.5_med1_collar0.0|6.30|
+|result_th0.6_med11_collar0.0|6.66|
+|result_th0.6_med1_collar0.0|6.80|
+|result_th0.7_med11_collar0.0|7.62|
+|result_th0.7_med1_collar0.0|7.77|
+## diar_enh_train_diar_enh_convtasnet_2_raw
+### DER
+diarized_enhanced_test_decode_diar_enh
+|threshold_median_collar|DER|
+|---|---|
+|result_th0.3_med11_collar0.0|6.20|
+|result_th0.3_med1_collar0.0|6.24|
+|result_th0.4_med11_collar0.0|6.04|
+|result_th0.4_med1_collar0.0|6.11|
+|result_th0.5_med11_collar0.0|6.16|
+|result_th0.5_med1_collar0.0|6.26|
+|result_th0.6_med11_collar0.0|6.59|
+|result_th0.6_med1_collar0.0|6.72|
+|result_th0.7_med11_collar0.0|7.46|
+|result_th0.7_med1_collar0.0|7.60|
+## diar_enh_train_diar_enh_convtasnet_2_raw
+### DER
+diarized_enhanced_test
+|threshold_median_collar|DER|
+|---|---|
+|result_th0.3_med11_collar0.0|6.19|
+|result_th0.3_med1_collar0.0|6.23|
+|result_th0.4_med11_collar0.0|6.05|
+|result_th0.4_med1_collar0.0|6.13|
+|result_th0.5_med11_collar0.0|6.20|
+|result_th0.5_med1_collar0.0|6.30|
+|result_th0.6_med11_collar0.0|6.66|
+|result_th0.6_med1_collar0.0|6.80|
+|result_th0.7_med11_collar0.0|7.62|
+|result_th0.7_med1_collar0.0|7.77|

exp/diar_enh_train_diar_enh_convtasnet_2_raw/ENH_RESULTS.md

@@ -0,0 +1,21 @@
+<!-- Generated by scripts/utils/show_enh_score.sh -->
+# RESULTS
+## Environments
+- date: `Mon Mar 13 03:15:55 EDT 2023`
+- python version: `3.9.12 (main, Apr  5 2022, 06:56:58)  [GCC 7.5.0]`
+- espnet version: `espnet 202205`
+- pytorch version: `pytorch 1.8.1+cu111`
+- Git hash: `f2778f798b76e102602078213a21b63aa592be70`
+  - Commit date: `Sun Jul 10 23:45:31 2022 +0000`
+
+
+## diar_enh_train_diar_enh_convtasnet_2_raw
+
+config: conf/tuning/train_diar_enh_convtasnet_2.yaml
+
+|dataset|STOI|SAR|SDR|SIR|SI_SNR|
+|---|---|---|---|---|---|
+|diarized_enhanced_test|70.1006|5.7082|3.8434|11.2483|2.7215|
+|diarized_enhanced_test_decode_diar_enh|71.4425|6.3658|4.5344|11.8239|3.4320|
+|diarized_enhanced_test_decode_diar_enh_wo_diar|71.5192|6.3575|4.4915|11.7014|3.4049|
+

exp/diar_enh_train_diar_enh_convtasnet_2_raw/config.yaml

@@ -0,0 +1,284 @@
+config: conf/tuning/train_diar_enh_convtasnet_2.yaml
+print_config: false
+log_level: INFO
+dry_run: false
+iterator_type: chunk
+output_dir: exp/diar_enh_train_diar_enh_convtasnet_2_raw
+ngpu: 1
+seed: 0
+num_workers: 4
+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: 100
+patience: 50
+val_scheduler_criterion:
+- valid
+- loss
+early_stopping_criterion:
+- valid
+- loss
+- min
+best_model_criterion:
+-   - valid
+    - loss_enh
+    - min
+keep_nbest_models: 1
+nbest_averaging_interval: 0
+grad_clip: 5.0
+grad_clip_type: 2.0
+grad_noise: false
+accum_grad: 4
+no_forward_run: false
+resume: true
+train_dtype: float32
+use_amp: false
+log_interval: null
+use_matplotlib: true
+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: 4
+valid_batch_size: null
+batch_bins: 1000000
+valid_batch_bins: null
+train_shape_file:
+- exp/diar_enh_stats_8k/train/speech_shape
+- exp/diar_enh_stats_8k/train/text_shape
+- exp/diar_enh_stats_8k/train/speech_ref1_shape
+- exp/diar_enh_stats_8k/train/speech_ref2_shape
+- exp/diar_enh_stats_8k/train/speech_ref3_shape
+- exp/diar_enh_stats_8k/train/noise_ref1_shape
+valid_shape_file:
+- exp/diar_enh_stats_8k/valid/speech_shape
+- exp/diar_enh_stats_8k/valid/text_shape
+- exp/diar_enh_stats_8k/valid/speech_ref1_shape
+- exp/diar_enh_stats_8k/valid/speech_ref2_shape
+- exp/diar_enh_stats_8k/valid/speech_ref3_shape
+- exp/diar_enh_stats_8k/valid/noise_ref1_shape
+batch_type: folded
+valid_batch_type: null
+fold_length:
+- 800
+- 80000
+- 80000
+- 80000
+- 80000
+- 80000
+sort_in_batch: descending
+sort_batch: descending
+multiple_iterator: false
+chunk_length: 24000
+chunk_shift_ratio: 0.5
+num_cache_chunks: 1024
+train_data_path_and_name_and_type:
+-   - dump/raw/train/wav.scp
+    - speech
+    - sound
+-   - dump/raw/train/espnet_rttm
+    - text
+    - rttm
+-   - dump/raw/train/spk1.scp
+    - speech_ref1
+    - sound
+-   - dump/raw/train/spk2.scp
+    - speech_ref2
+    - sound
+-   - dump/raw/train/spk3.scp
+    - speech_ref3
+    - sound
+-   - dump/raw/train/noise1.scp
+    - noise_ref1
+    - sound
+valid_data_path_and_name_and_type:
+-   - dump/raw/dev/wav.scp
+    - speech
+    - sound
+-   - dump/raw/dev/espnet_rttm
+    - text
+    - rttm
+-   - dump/raw/dev/spk1.scp
+    - speech_ref1
+    - sound
+-   - dump/raw/dev/spk2.scp
+    - speech_ref2
+    - sound
+-   - dump/raw/dev/spk3.scp
+    - speech_ref3
+    - sound
+-   - dump/raw/dev/noise1.scp
+    - noise_ref1
+    - sound
+allow_variable_data_keys: false
+max_cache_size: 0.0
+max_cache_fd: 32
+valid_max_cache_size: null
+optim: adam
+optim_conf:
+    lr: 0.001
+    eps: 1.0e-07
+    weight_decay: 0
+scheduler: reducelronplateau
+scheduler_conf:
+    mode: min
+    factor: 0.5
+    patience: 1
+token_list: null
+src_token_list: null
+init: xavier_uniform
+input_size: null
+ctc_conf:
+    dropout_rate: 0.0
+    ctc_type: builtin
+    reduce: true
+    ignore_nan_grad: null
+    zero_infinity: true
+enh_criterions:
+-   name: si_snr
+    conf:
+        eps: 1.0e-07
+    wrapper: pit
+    wrapper_conf:
+        weight: 1.0
+        independent_perm: true
+diar_num_spk: 3
+diar_input_size: 128
+enh_model_conf:
+    loss_type: si_snr
+asr_model_conf:
+    ctc_weight: 0.5
+    interctc_weight: 0.0
+    ignore_id: -1
+    lsm_weight: 0.0
+    length_normalized_loss: false
+    report_cer: true
+    report_wer: true
+    sym_space: <space>
+    sym_blank: <blank>
+    extract_feats_in_collect_stats: true
+st_model_conf:
+    stft_consistency: false
+    loss_type: mask_mse
+    mask_type: null
+diar_model_conf:
+    diar_weight: 0.2
+    attractor_weight: 0.2
+subtask_series:
+- enh
+- diar
+model_conf:
+    calc_enh_loss: true
+    bypass_enh_prob: 0
+use_preprocessor: true
+token_type: bpe
+bpemodel: null
+src_token_type: bpe
+src_bpemodel: null
+non_linguistic_symbols: null
+cleaner: null
+g2p: null
+enh_encoder: conv
+enh_encoder_conf:
+    channel: 512
+    kernel_size: 16
+    stride: 8
+enh_separator: tcn_nomask
+enh_separator_conf:
+    layer: 8
+    stack: 3
+    bottleneck_dim: 128
+    hidden_dim: 512
+    kernel: 3
+    causal: false
+    norm_type: gLN
+enh_decoder: conv
+enh_decoder_conf:
+    channel: 512
+    kernel_size: 16
+    stride: 8
+enh_mask_module: multi_mask
+enh_mask_module_conf:
+    max_num_spk: 3
+    mask_nonlinear: relu
+    bottleneck_dim: 128
+frontend: null
+frontend_conf: {}
+specaug: null
+specaug_conf: {}
+normalize: utterance_mvn
+normalize_conf: {}
+asr_preencoder: null
+asr_preencoder_conf: {}
+asr_encoder: rnn
+asr_encoder_conf: {}
+asr_postencoder: null
+asr_postencoder_conf: {}
+asr_decoder: rnn
+asr_decoder_conf: {}
+st_preencoder: null
+st_preencoder_conf: {}
+st_encoder: rnn
+st_encoder_conf: {}
+st_postencoder: null
+st_postencoder_conf: {}
+st_decoder: rnn
+st_decoder_conf: {}
+st_extra_asr_decoder: rnn
+st_extra_asr_decoder_conf: {}
+st_extra_mt_decoder: rnn
+st_extra_mt_decoder_conf: {}
+diar_frontend: null
+diar_frontend_conf: {}
+diar_specaug: null
+diar_specaug_conf: {}
+diar_normalize: utterance_mvn
+diar_normalize_conf: {}
+diar_encoder: transformer
+diar_encoder_conf:
+    input_layer: conv2d8
+    num_blocks: 4
+    linear_units: 512
+    dropout_rate: 0.1
+    output_size: 256
+    attention_heads: 4
+    attention_dropout_rate: 0.1
+diar_decoder: linear
+diar_decoder_conf: {}
+label_aggregator: label_aggregator
+label_aggregator_conf:
+    win_length: 256
+    hop_length: 64
+diar_attractor: rnn
+diar_attractor_conf:
+    unit: 256
+    layer: 1
+    dropout: 0.0
+    attractor_grad: true
+required:
+- output_dir
+version: '202205'
+distributed: false

exp/diar_enh_train_diar_enh_convtasnet_2_raw/valid.loss_enh.best.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64fe8f116078f977e694d571382cdcd28a7352f5dd3a829e06667190a3d227b1
+size 36279117

meta.yaml

@@ -1,8 +1,8 @@
 espnet: '202205'
 files:
-  model_file: exp/diar_enh_train_diar_enh_convtasnet_concat_feats_raw/valid.loss_enh.best.pth
+  model_file: exp/diar_enh_train_diar_enh_convtasnet_2_raw/valid.loss_enh.best.pth
 python: "3.9.16 | packaged by conda-forge | (main, Feb  1 2023, 21:39:03) \n[GCC 11.3.0]"
-timestamp: 1683210501.537852
+timestamp: 1683210536.396694
 torch: 1.8.1+cu102
 yaml_files:
-  train_config: exp/diar_enh_train_diar_enh_convtasnet_concat_feats_raw/config.yaml
+  train_config: exp/diar_enh_train_diar_enh_convtasnet_2_raw/config.yaml