shikhar7ssu commited on
Commit
7d74274
·
verified ·
1 Parent(s): 763a64f

Upload 19 files

Browse files
README.md ADDED
@@ -0,0 +1,541 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - espnet
4
+ - audio
5
+ - classification
6
+ language: en
7
+ datasets:
8
+ - bean
9
+ license: cc-by-4.0
10
+ ---
11
+
12
+ ## ESPnet2 CLS model
13
+
14
+ ### `espnet/BEATs-BEAN.CornellBirdIdentification`
15
+
16
+ This model was trained by Shikhar Bharadwaj using bean recipe in [espnet](https://github.com/espnet/espnet/).
17
+
18
+ ### Demo: How to use in ESPnet2
19
+
20
+ Follow the [ESPnet installation instructions](https://espnet.github.io/espnet/installation.html)
21
+ if you haven't done that already.
22
+
23
+ ```bash
24
+ cd espnet
25
+ git checkout 9191aa59acc7d3ceaca1f48dcc8fbdad2e03484b
26
+ pip install -e .
27
+ cd egs2/bean/cls1
28
+ ./run.sh --skip_data_prep false --skip_train true --download_model espnet/BEATs-BEAN.CornellBirdIdentification
29
+ ```
30
+
31
+ <!-- Generated by scripts/utils/show_cls_result.sh -->
32
+ # RESULTS
33
+ ## Environments
34
+ - date: `Wed Jan 8 05:51:08 EST 2025`
35
+ - python version: `3.9.20 (main, Oct 3 2024, 07:27:41) [GCC 11.2.0]`
36
+ - espnet version: `espnet 202412`
37
+ - pytorch version: `pytorch 2.4.0`
38
+ - Git hash: `9191aa59acc7d3ceaca1f48dcc8fbdad2e03484b`
39
+ - Commit date: `Tue Jan 7 04:34:03 2025 -0500`
40
+
41
+ ## cls_cbi.20250107.141123
42
+ |Dataset|Metric|Value|
43
+ |---|---|---|
44
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.dev/score|mean_acc|68.12
45
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.dev/score|mAP|67.80
46
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.dev/score|mean_auc|93.70
47
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.dev/score|n_labels|264.00
48
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.dev/score|n_instances|3548.00
49
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.test/score|mean_acc|63.95
50
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.test/score|mAP|66.58
51
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.test/score|mean_auc|95.61
52
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.test/score|n_labels|264.00
53
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.test/score|n_instances|3620.00
54
+
55
+ ## CLS config
56
+
57
+ <details><summary>expand</summary>
58
+
59
+ ```
60
+ config: conf/beats_cbi.yaml
61
+ print_config: false
62
+ log_level: INFO
63
+ drop_last_iter: false
64
+ dry_run: false
65
+ iterator_type: sequence
66
+ valid_iterator_type: null
67
+ output_dir: /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123
68
+ ngpu: 1
69
+ seed: 0
70
+ num_workers: 2
71
+ num_att_plot: 0
72
+ dist_backend: nccl
73
+ dist_init_method: env://
74
+ dist_world_size: null
75
+ dist_rank: null
76
+ local_rank: 0
77
+ dist_master_addr: null
78
+ dist_master_port: null
79
+ dist_launcher: null
80
+ multiprocessing_distributed: false
81
+ unused_parameters: true
82
+ sharded_ddp: false
83
+ use_deepspeed: false
84
+ deepspeed_config: null
85
+ cudnn_enabled: true
86
+ cudnn_benchmark: false
87
+ cudnn_deterministic: true
88
+ use_tf32: false
89
+ collect_stats: false
90
+ write_collected_feats: false
91
+ max_epoch: 250
92
+ patience: null
93
+ val_scheduler_criterion:
94
+ - valid
95
+ - loss
96
+ early_stopping_criterion:
97
+ - valid
98
+ - loss
99
+ - min
100
+ best_model_criterion:
101
+ - - valid
102
+ - acc
103
+ - max
104
+ keep_nbest_models: 1
105
+ nbest_averaging_interval: 0
106
+ grad_clip: 1
107
+ grad_clip_type: 2.0
108
+ grad_noise: false
109
+ accum_grad: 1
110
+ no_forward_run: false
111
+ resume: true
112
+ train_dtype: float32
113
+ use_amp: false
114
+ log_interval: null
115
+ use_matplotlib: true
116
+ use_tensorboard: true
117
+ create_graph_in_tensorboard: false
118
+ use_wandb: false
119
+ wandb_project: null
120
+ wandb_id: null
121
+ wandb_entity: null
122
+ wandb_name: null
123
+ wandb_model_log_interval: -1
124
+ detect_anomaly: false
125
+ use_adapter: false
126
+ adapter: lora
127
+ save_strategy: all
128
+ adapter_conf: {}
129
+ pretrain_path: null
130
+ init_param: []
131
+ ignore_init_mismatch: false
132
+ freeze_param: []
133
+ num_iters_per_epoch: null
134
+ batch_size: 32
135
+ valid_batch_size: 32
136
+ batch_bins: 1000000
137
+ valid_batch_bins: null
138
+ category_sample_size: 10
139
+ train_shape_file:
140
+ - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_stats_16k/train/speech_shape
141
+ - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_stats_16k/train/label_shape
142
+ valid_shape_file:
143
+ - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_stats_16k/valid/speech_shape
144
+ - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_stats_16k/valid/label_shape
145
+ batch_type: folded
146
+ valid_batch_type: null
147
+ fold_length:
148
+ - 160000
149
+ - 5
150
+ sort_in_batch: descending
151
+ shuffle_within_batch: false
152
+ sort_batch: descending
153
+ multiple_iterator: false
154
+ chunk_length: 500
155
+ chunk_shift_ratio: 0.5
156
+ num_cache_chunks: 1024
157
+ chunk_excluded_key_prefixes: []
158
+ chunk_default_fs: null
159
+ chunk_max_abs_length: null
160
+ chunk_discard_short_samples: true
161
+ train_data_path_and_name_and_type:
162
+ - - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/dump/cbi.train/wav.scp
163
+ - speech
164
+ - sound
165
+ - - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/dump/cbi.train/text
166
+ - label
167
+ - text
168
+ valid_data_path_and_name_and_type:
169
+ - - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/dump/cbi.dev/wav.scp
170
+ - speech
171
+ - sound
172
+ - - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/dump/cbi.dev/text
173
+ - label
174
+ - text
175
+ multi_task_dataset: false
176
+ allow_variable_data_keys: false
177
+ max_cache_size: 0.0
178
+ max_cache_fd: 32
179
+ allow_multi_rates: false
180
+ valid_max_cache_size: null
181
+ exclude_weight_decay: false
182
+ exclude_weight_decay_conf: {}
183
+ optim: adamw
184
+ optim_conf:
185
+ lr: 3.0e-05
186
+ weight_decay: 0.01
187
+ betas:
188
+ - 0.9
189
+ - 0.98
190
+ scheduler: cosineannealingwarmuprestarts
191
+ scheduler_conf:
192
+ first_cycle_steps: 95000
193
+ warmup_steps: 8000
194
+ max_lr: 3.0e-05
195
+ min_lr: 5.0e-06
196
+ token_list:
197
+ - scoori
198
+ - bulori
199
+ - bushti
200
+ - blkpho
201
+ - brthum
202
+ - cacwre
203
+ - pasfly
204
+ - lesgol
205
+ - logshr
206
+ - macwar
207
+ - pinsis
208
+ - whbnut
209
+ - hamfly
210
+ - normoc
211
+ - grtgra
212
+ - houwre
213
+ - comyel
214
+ - grhowl
215
+ - houfin
216
+ - rocpig
217
+ - annhum
218
+ - astfly
219
+ - magwar
220
+ - wesmea
221
+ - wewpew
222
+ - spotow
223
+ - amerob
224
+ - daejun
225
+ - easmea
226
+ - greroa
227
+ - mouchi
228
+ - pilwoo
229
+ - comrav
230
+ - hoowar
231
+ - savspa
232
+ - warvir
233
+ - easblu
234
+ - gnttow
235
+ - ovenbi1
236
+ - rewbla
237
+ - robgro
238
+ - swathr
239
+ - tuftit
240
+ - westan
241
+ - winwre3
242
+ - btywar
243
+ - carwre
244
+ - herthr
245
+ - bewwre
246
+ - sora
247
+ - brdowl
248
+ - buggna
249
+ - casvir
250
+ - chispa
251
+ - fiespa
252
+ - aldfly
253
+ - killde
254
+ - moudov
255
+ - rebwoo
256
+ - bkpwar
257
+ - dowwoo
258
+ - greegr
259
+ - banswa
260
+ - orcwar
261
+ - plsvir
262
+ - y00475
263
+ - blugrb1
264
+ - gockin
265
+ - greyel
266
+ - larspa
267
+ - osprey
268
+ - sonspa
269
+ - yebfly
270
+ - blujay
271
+ - brnthr
272
+ - canwre
273
+ - clanut
274
+ - comred
275
+ - eastow
276
+ - haiwoo
277
+ - lesyel
278
+ - amepip
279
+ - easpho
280
+ - fiscro
281
+ - sposan
282
+ - wooscj2
283
+ - bkhgro
284
+ - labwoo
285
+ - lazbun
286
+ - marwre
287
+ - stejay
288
+ - weskin
289
+ - bkbwar
290
+ - buhvir
291
+ - cangoo
292
+ - canwar
293
+ - dusfly
294
+ - grcfly
295
+ - norcar
296
+ - wilsni1
297
+ - yerwar
298
+ - yetvir
299
+ - eucdov
300
+ - linspa
301
+ - norpar
302
+ - olsfly
303
+ - rebnut
304
+ - scatan
305
+ - bnhcow
306
+ - louwat
307
+ - norfli
308
+ - veery
309
+ - woothr
310
+ - btnwar
311
+ - cedwax
312
+ - chswar
313
+ - comgra
314
+ - indbun
315
+ - leabit
316
+ - leafly
317
+ - pinwar
318
+ - reevir1
319
+ - solsan
320
+ - bktspa
321
+ - foxspa
322
+ - houspa
323
+ - snobun
324
+ - vesspa
325
+ - yelwar
326
+ - brespa
327
+ - comgol
328
+ - coohaw
329
+ - gnwtea
330
+ - grbher3
331
+ - hergul
332
+ - mallar3
333
+ - swaspa
334
+ - brncre
335
+ - btbwar
336
+ - caster1
337
+ - eawpew
338
+ - rethaw
339
+ - rocwre
340
+ - ruckin
341
+ - semsan
342
+ - whtspa
343
+ - wlswar
344
+ - bkcchi
345
+ - bkchum
346
+ - amered
347
+ - norwat
348
+ - whcspa
349
+ - grycat
350
+ - balori
351
+ - purfin
352
+ - treswa
353
+ - wilfly
354
+ - comter
355
+ - belspa2
356
+ - juntit1
357
+ - comnig
358
+ - reshaw
359
+ - snogoo
360
+ - gadwal
361
+ - perfal
362
+ - grnher
363
+ - horlar
364
+ - lobdow
365
+ - bawwar
366
+ - amegfi
367
+ - commer
368
+ - ribgul
369
+ - casfin
370
+ - pibgre
371
+ - evegro
372
+ - pygnut
373
+ - brwhaw
374
+ - gryfly
375
+ - leasan
376
+ - barswa
377
+ - phaino
378
+ - amecro
379
+ - calqua
380
+ - amewoo
381
+ - pingro
382
+ - saypho
383
+ - semplo
384
+ - buwwar
385
+ - boboli
386
+ - amekes
387
+ - cowscj1
388
+ - amtspa
389
+ - lobcur
390
+ - belkin1
391
+ - pecsan
392
+ - prawar
393
+ - vigswa
394
+ - camwar
395
+ - easkin
396
+ - yebsap
397
+ - norsho
398
+ - gocspa
399
+ - rufhum
400
+ - baisan
401
+ - cliswa
402
+ - pinjay
403
+ - comloo
404
+ - baleag
405
+ - merlin
406
+ - yehbla
407
+ - calgul
408
+ - goleag
409
+ - nutwoo
410
+ - rusbla
411
+ - eursta
412
+ - ameavo
413
+ - lesnig
414
+ - palwar
415
+ - bkbmag1
416
+ - brebla
417
+ - sagthr
418
+ - bkbcuc
419
+ - wesgre
420
+ - redcro
421
+ - wiltur
422
+ - amebit
423
+ - sagspa1
424
+ - tunswa
425
+ - wooduc
426
+ - renpha
427
+ - whtswi
428
+ - bongul
429
+ - norhar2
430
+ - doccor
431
+ - lotduc
432
+ - chukar
433
+ - horgre
434
+ - nrwswa
435
+ - sheowl
436
+ - wesblu
437
+ - whfibi
438
+ - buwtea
439
+ - norpin
440
+ - eargre
441
+ - rebsap
442
+ - lewwoo
443
+ - rebmer
444
+ - wessan
445
+ - chiswi
446
+ - lecthr
447
+ - rthhum
448
+ - moublu
449
+ - amewig
450
+ - rinduc
451
+ - shshaw
452
+ - rufgro
453
+ - swahaw
454
+ - coshum
455
+ - truswa
456
+ - rudduc
457
+ - buffle
458
+ - hoomer
459
+ - gcrfin
460
+ - redhea
461
+ - <unk>
462
+ token_type: word
463
+ init: xavier_normal
464
+ input_size: 1
465
+ use_preprocessor: true
466
+ frontend: null
467
+ frontend_conf: {}
468
+ specaug: null
469
+ specaug_conf: {}
470
+ normalize: null
471
+ normalize_conf: {}
472
+ preencoder: null
473
+ preencoder_conf: {}
474
+ encoder: beats
475
+ encoder_conf:
476
+ beats_ckpt_path: /compute/babel-13-33/sbharad2/models/BEATs/BEATs_iter3.pt
477
+ beats_config:
478
+ layer_wise_gradient_decay_ratio: 0.3
479
+ encoder_layerdrop: 0.1
480
+ dropout: 0.0
481
+ use_weighted_representation: false
482
+ specaug_config:
483
+ apply_time_warp: true
484
+ apply_freq_mask: false
485
+ apply_time_mask: true
486
+ time_mask_width_ratio_range:
487
+ - 0
488
+ - 0.06
489
+ num_time_mask: 1
490
+ roll_augment: true
491
+ roll_interval: 1
492
+ decoder: linear
493
+ decoder_conf: {}
494
+ model: espnet
495
+ model_conf:
496
+ classification_type: multi-class
497
+ mixup_augmentation: false
498
+ lsm_weight: 0.1
499
+ required:
500
+ - output_dir
501
+ - token_list
502
+ version: '202412'
503
+ distributed: false
504
+ ```
505
+
506
+ </details>
507
+
508
+
509
+
510
+ ### Citing ESPnet
511
+
512
+ ```BibTex
513
+ @inproceedings{watanabe2018espnet,
514
+ 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},
515
+ title={{ESPnet}: End-to-End Speech Processing Toolkit},
516
+ year={2018},
517
+ booktitle={Proceedings of Interspeech},
518
+ pages={2207--2211},
519
+ doi={10.21437/Interspeech.2018-1456},
520
+ url={http://dx.doi.org/10.21437/Interspeech.2018-1456}
521
+ }
522
+
523
+
524
+
525
+
526
+
527
+
528
+ ```
529
+
530
+ or arXiv:
531
+
532
+ ```bibtex
533
+ @misc{watanabe2018espnet,
534
+ title={ESPnet: End-to-End Speech Processing Toolkit},
535
+ 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},
536
+ year={2018},
537
+ eprint={1804.00015},
538
+ archivePrefix={arXiv},
539
+ primaryClass={cs.CL}
540
+ }
541
+ ```
bean.cbi/data/token_list ADDED
@@ -0,0 +1,265 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ scoori
2
+ bulori
3
+ bushti
4
+ blkpho
5
+ brthum
6
+ cacwre
7
+ pasfly
8
+ lesgol
9
+ logshr
10
+ macwar
11
+ pinsis
12
+ whbnut
13
+ hamfly
14
+ normoc
15
+ grtgra
16
+ houwre
17
+ comyel
18
+ grhowl
19
+ houfin
20
+ rocpig
21
+ annhum
22
+ astfly
23
+ magwar
24
+ wesmea
25
+ wewpew
26
+ spotow
27
+ amerob
28
+ daejun
29
+ easmea
30
+ greroa
31
+ mouchi
32
+ pilwoo
33
+ comrav
34
+ hoowar
35
+ savspa
36
+ warvir
37
+ easblu
38
+ gnttow
39
+ ovenbi1
40
+ rewbla
41
+ robgro
42
+ swathr
43
+ tuftit
44
+ westan
45
+ winwre3
46
+ btywar
47
+ carwre
48
+ herthr
49
+ bewwre
50
+ sora
51
+ brdowl
52
+ buggna
53
+ casvir
54
+ chispa
55
+ fiespa
56
+ aldfly
57
+ killde
58
+ moudov
59
+ rebwoo
60
+ bkpwar
61
+ dowwoo
62
+ greegr
63
+ banswa
64
+ orcwar
65
+ plsvir
66
+ y00475
67
+ blugrb1
68
+ gockin
69
+ greyel
70
+ larspa
71
+ osprey
72
+ sonspa
73
+ yebfly
74
+ blujay
75
+ brnthr
76
+ canwre
77
+ clanut
78
+ comred
79
+ eastow
80
+ haiwoo
81
+ lesyel
82
+ amepip
83
+ easpho
84
+ fiscro
85
+ sposan
86
+ wooscj2
87
+ bkhgro
88
+ labwoo
89
+ lazbun
90
+ marwre
91
+ stejay
92
+ weskin
93
+ bkbwar
94
+ buhvir
95
+ cangoo
96
+ canwar
97
+ dusfly
98
+ grcfly
99
+ norcar
100
+ wilsni1
101
+ yerwar
102
+ yetvir
103
+ eucdov
104
+ linspa
105
+ norpar
106
+ olsfly
107
+ rebnut
108
+ scatan
109
+ bnhcow
110
+ louwat
111
+ norfli
112
+ veery
113
+ woothr
114
+ btnwar
115
+ cedwax
116
+ chswar
117
+ comgra
118
+ indbun
119
+ leabit
120
+ leafly
121
+ pinwar
122
+ reevir1
123
+ solsan
124
+ bktspa
125
+ foxspa
126
+ houspa
127
+ snobun
128
+ vesspa
129
+ yelwar
130
+ brespa
131
+ comgol
132
+ coohaw
133
+ gnwtea
134
+ grbher3
135
+ hergul
136
+ mallar3
137
+ swaspa
138
+ brncre
139
+ btbwar
140
+ caster1
141
+ eawpew
142
+ rethaw
143
+ rocwre
144
+ ruckin
145
+ semsan
146
+ whtspa
147
+ wlswar
148
+ bkcchi
149
+ bkchum
150
+ amered
151
+ norwat
152
+ whcspa
153
+ grycat
154
+ balori
155
+ purfin
156
+ treswa
157
+ wilfly
158
+ comter
159
+ belspa2
160
+ juntit1
161
+ comnig
162
+ reshaw
163
+ snogoo
164
+ gadwal
165
+ perfal
166
+ grnher
167
+ horlar
168
+ lobdow
169
+ bawwar
170
+ amegfi
171
+ commer
172
+ ribgul
173
+ casfin
174
+ pibgre
175
+ evegro
176
+ pygnut
177
+ brwhaw
178
+ gryfly
179
+ leasan
180
+ barswa
181
+ phaino
182
+ amecro
183
+ calqua
184
+ amewoo
185
+ pingro
186
+ saypho
187
+ semplo
188
+ buwwar
189
+ boboli
190
+ amekes
191
+ cowscj1
192
+ amtspa
193
+ lobcur
194
+ belkin1
195
+ pecsan
196
+ prawar
197
+ vigswa
198
+ camwar
199
+ easkin
200
+ yebsap
201
+ norsho
202
+ gocspa
203
+ rufhum
204
+ baisan
205
+ cliswa
206
+ pinjay
207
+ comloo
208
+ baleag
209
+ merlin
210
+ yehbla
211
+ calgul
212
+ goleag
213
+ nutwoo
214
+ rusbla
215
+ eursta
216
+ ameavo
217
+ lesnig
218
+ palwar
219
+ bkbmag1
220
+ brebla
221
+ sagthr
222
+ bkbcuc
223
+ wesgre
224
+ redcro
225
+ wiltur
226
+ amebit
227
+ sagspa1
228
+ tunswa
229
+ wooduc
230
+ renpha
231
+ whtswi
232
+ bongul
233
+ norhar2
234
+ doccor
235
+ lotduc
236
+ chukar
237
+ horgre
238
+ nrwswa
239
+ sheowl
240
+ wesblu
241
+ whfibi
242
+ buwtea
243
+ norpin
244
+ eargre
245
+ rebsap
246
+ lewwoo
247
+ rebmer
248
+ wessan
249
+ chiswi
250
+ lecthr
251
+ rthhum
252
+ moublu
253
+ amewig
254
+ rinduc
255
+ shshaw
256
+ rufgro
257
+ swahaw
258
+ coshum
259
+ truswa
260
+ rudduc
261
+ buffle
262
+ hoomer
263
+ gcrfin
264
+ redhea
265
+ <unk>
bean.cbi/exp/cls_cbi.20250107.141123/201epoch.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:998d35e595dcad4f829e5914872d1fee872e4ea53ff5a1cbb3f6fe02ec54aad6
3
+ size 362154053
bean.cbi/exp/cls_cbi.20250107.141123/RESULTS.md ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ <!-- Generated by scripts/utils/show_cls_result.sh -->
2
+ # RESULTS
3
+ ## Environments
4
+ - date: `Wed Jan 8 05:51:08 EST 2025`
5
+ - python version: `3.9.20 (main, Oct 3 2024, 07:27:41) [GCC 11.2.0]`
6
+ - espnet version: `espnet 202412`
7
+ - pytorch version: `pytorch 2.4.0`
8
+ - Git hash: `9191aa59acc7d3ceaca1f48dcc8fbdad2e03484b`
9
+ - Commit date: `Tue Jan 7 04:34:03 2025 -0500`
10
+
11
+ ## cls_cbi.20250107.141123
12
+ |Dataset|Metric|Value|
13
+ |---|---|---|
14
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.dev/score|mean_acc|68.12
15
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.dev/score|mAP|67.80
16
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.dev/score|mean_auc|93.70
17
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.dev/score|n_labels|264.00
18
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.dev/score|n_instances|3548.00
19
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.test/score|mean_acc|63.95
20
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.test/score|mAP|66.58
21
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.test/score|mean_auc|95.61
22
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.test/score|n_labels|264.00
23
+ /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/cls_cbi.test/score|n_instances|3620.00
24
+
bean.cbi/exp/cls_cbi.20250107.141123/config.yaml ADDED
@@ -0,0 +1,444 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ config: conf/beats_cbi.yaml
2
+ print_config: false
3
+ log_level: INFO
4
+ drop_last_iter: false
5
+ dry_run: false
6
+ iterator_type: sequence
7
+ valid_iterator_type: null
8
+ output_dir: /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123
9
+ ngpu: 1
10
+ seed: 0
11
+ num_workers: 2
12
+ num_att_plot: 0
13
+ dist_backend: nccl
14
+ dist_init_method: env://
15
+ dist_world_size: null
16
+ dist_rank: null
17
+ local_rank: 0
18
+ dist_master_addr: null
19
+ dist_master_port: null
20
+ dist_launcher: null
21
+ multiprocessing_distributed: false
22
+ unused_parameters: true
23
+ sharded_ddp: false
24
+ use_deepspeed: false
25
+ deepspeed_config: null
26
+ cudnn_enabled: true
27
+ cudnn_benchmark: false
28
+ cudnn_deterministic: true
29
+ use_tf32: false
30
+ collect_stats: false
31
+ write_collected_feats: false
32
+ max_epoch: 250
33
+ patience: null
34
+ val_scheduler_criterion:
35
+ - valid
36
+ - loss
37
+ early_stopping_criterion:
38
+ - valid
39
+ - loss
40
+ - min
41
+ best_model_criterion:
42
+ - - valid
43
+ - acc
44
+ - max
45
+ keep_nbest_models: 1
46
+ nbest_averaging_interval: 0
47
+ grad_clip: 1
48
+ grad_clip_type: 2.0
49
+ grad_noise: false
50
+ accum_grad: 1
51
+ no_forward_run: false
52
+ resume: true
53
+ train_dtype: float32
54
+ use_amp: false
55
+ log_interval: null
56
+ use_matplotlib: true
57
+ use_tensorboard: true
58
+ create_graph_in_tensorboard: false
59
+ use_wandb: false
60
+ wandb_project: null
61
+ wandb_id: null
62
+ wandb_entity: null
63
+ wandb_name: null
64
+ wandb_model_log_interval: -1
65
+ detect_anomaly: false
66
+ use_adapter: false
67
+ adapter: lora
68
+ save_strategy: all
69
+ adapter_conf: {}
70
+ pretrain_path: null
71
+ init_param: []
72
+ ignore_init_mismatch: false
73
+ freeze_param: []
74
+ num_iters_per_epoch: null
75
+ batch_size: 32
76
+ valid_batch_size: 32
77
+ batch_bins: 1000000
78
+ valid_batch_bins: null
79
+ category_sample_size: 10
80
+ train_shape_file:
81
+ - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_stats_16k/train/speech_shape
82
+ - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_stats_16k/train/label_shape
83
+ valid_shape_file:
84
+ - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_stats_16k/valid/speech_shape
85
+ - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_stats_16k/valid/label_shape
86
+ batch_type: folded
87
+ valid_batch_type: null
88
+ fold_length:
89
+ - 160000
90
+ - 5
91
+ sort_in_batch: descending
92
+ shuffle_within_batch: false
93
+ sort_batch: descending
94
+ multiple_iterator: false
95
+ chunk_length: 500
96
+ chunk_shift_ratio: 0.5
97
+ num_cache_chunks: 1024
98
+ chunk_excluded_key_prefixes: []
99
+ chunk_default_fs: null
100
+ chunk_max_abs_length: null
101
+ chunk_discard_short_samples: true
102
+ train_data_path_and_name_and_type:
103
+ - - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/dump/cbi.train/wav.scp
104
+ - speech
105
+ - sound
106
+ - - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/dump/cbi.train/text
107
+ - label
108
+ - text
109
+ valid_data_path_and_name_and_type:
110
+ - - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/dump/cbi.dev/wav.scp
111
+ - speech
112
+ - sound
113
+ - - /compute/babel-11-13/sbharad2/beats_run/bean.cbi/dump/cbi.dev/text
114
+ - label
115
+ - text
116
+ multi_task_dataset: false
117
+ allow_variable_data_keys: false
118
+ max_cache_size: 0.0
119
+ max_cache_fd: 32
120
+ allow_multi_rates: false
121
+ valid_max_cache_size: null
122
+ exclude_weight_decay: false
123
+ exclude_weight_decay_conf: {}
124
+ optim: adamw
125
+ optim_conf:
126
+ lr: 3.0e-05
127
+ weight_decay: 0.01
128
+ betas:
129
+ - 0.9
130
+ - 0.98
131
+ scheduler: cosineannealingwarmuprestarts
132
+ scheduler_conf:
133
+ first_cycle_steps: 95000
134
+ warmup_steps: 8000
135
+ max_lr: 3.0e-05
136
+ min_lr: 5.0e-06
137
+ token_list:
138
+ - scoori
139
+ - bulori
140
+ - bushti
141
+ - blkpho
142
+ - brthum
143
+ - cacwre
144
+ - pasfly
145
+ - lesgol
146
+ - logshr
147
+ - macwar
148
+ - pinsis
149
+ - whbnut
150
+ - hamfly
151
+ - normoc
152
+ - grtgra
153
+ - houwre
154
+ - comyel
155
+ - grhowl
156
+ - houfin
157
+ - rocpig
158
+ - annhum
159
+ - astfly
160
+ - magwar
161
+ - wesmea
162
+ - wewpew
163
+ - spotow
164
+ - amerob
165
+ - daejun
166
+ - easmea
167
+ - greroa
168
+ - mouchi
169
+ - pilwoo
170
+ - comrav
171
+ - hoowar
172
+ - savspa
173
+ - warvir
174
+ - easblu
175
+ - gnttow
176
+ - ovenbi1
177
+ - rewbla
178
+ - robgro
179
+ - swathr
180
+ - tuftit
181
+ - westan
182
+ - winwre3
183
+ - btywar
184
+ - carwre
185
+ - herthr
186
+ - bewwre
187
+ - sora
188
+ - brdowl
189
+ - buggna
190
+ - casvir
191
+ - chispa
192
+ - fiespa
193
+ - aldfly
194
+ - killde
195
+ - moudov
196
+ - rebwoo
197
+ - bkpwar
198
+ - dowwoo
199
+ - greegr
200
+ - banswa
201
+ - orcwar
202
+ - plsvir
203
+ - y00475
204
+ - blugrb1
205
+ - gockin
206
+ - greyel
207
+ - larspa
208
+ - osprey
209
+ - sonspa
210
+ - yebfly
211
+ - blujay
212
+ - brnthr
213
+ - canwre
214
+ - clanut
215
+ - comred
216
+ - eastow
217
+ - haiwoo
218
+ - lesyel
219
+ - amepip
220
+ - easpho
221
+ - fiscro
222
+ - sposan
223
+ - wooscj2
224
+ - bkhgro
225
+ - labwoo
226
+ - lazbun
227
+ - marwre
228
+ - stejay
229
+ - weskin
230
+ - bkbwar
231
+ - buhvir
232
+ - cangoo
233
+ - canwar
234
+ - dusfly
235
+ - grcfly
236
+ - norcar
237
+ - wilsni1
238
+ - yerwar
239
+ - yetvir
240
+ - eucdov
241
+ - linspa
242
+ - norpar
243
+ - olsfly
244
+ - rebnut
245
+ - scatan
246
+ - bnhcow
247
+ - louwat
248
+ - norfli
249
+ - veery
250
+ - woothr
251
+ - btnwar
252
+ - cedwax
253
+ - chswar
254
+ - comgra
255
+ - indbun
256
+ - leabit
257
+ - leafly
258
+ - pinwar
259
+ - reevir1
260
+ - solsan
261
+ - bktspa
262
+ - foxspa
263
+ - houspa
264
+ - snobun
265
+ - vesspa
266
+ - yelwar
267
+ - brespa
268
+ - comgol
269
+ - coohaw
270
+ - gnwtea
271
+ - grbher3
272
+ - hergul
273
+ - mallar3
274
+ - swaspa
275
+ - brncre
276
+ - btbwar
277
+ - caster1
278
+ - eawpew
279
+ - rethaw
280
+ - rocwre
281
+ - ruckin
282
+ - semsan
283
+ - whtspa
284
+ - wlswar
285
+ - bkcchi
286
+ - bkchum
287
+ - amered
288
+ - norwat
289
+ - whcspa
290
+ - grycat
291
+ - balori
292
+ - purfin
293
+ - treswa
294
+ - wilfly
295
+ - comter
296
+ - belspa2
297
+ - juntit1
298
+ - comnig
299
+ - reshaw
300
+ - snogoo
301
+ - gadwal
302
+ - perfal
303
+ - grnher
304
+ - horlar
305
+ - lobdow
306
+ - bawwar
307
+ - amegfi
308
+ - commer
309
+ - ribgul
310
+ - casfin
311
+ - pibgre
312
+ - evegro
313
+ - pygnut
314
+ - brwhaw
315
+ - gryfly
316
+ - leasan
317
+ - barswa
318
+ - phaino
319
+ - amecro
320
+ - calqua
321
+ - amewoo
322
+ - pingro
323
+ - saypho
324
+ - semplo
325
+ - buwwar
326
+ - boboli
327
+ - amekes
328
+ - cowscj1
329
+ - amtspa
330
+ - lobcur
331
+ - belkin1
332
+ - pecsan
333
+ - prawar
334
+ - vigswa
335
+ - camwar
336
+ - easkin
337
+ - yebsap
338
+ - norsho
339
+ - gocspa
340
+ - rufhum
341
+ - baisan
342
+ - cliswa
343
+ - pinjay
344
+ - comloo
345
+ - baleag
346
+ - merlin
347
+ - yehbla
348
+ - calgul
349
+ - goleag
350
+ - nutwoo
351
+ - rusbla
352
+ - eursta
353
+ - ameavo
354
+ - lesnig
355
+ - palwar
356
+ - bkbmag1
357
+ - brebla
358
+ - sagthr
359
+ - bkbcuc
360
+ - wesgre
361
+ - redcro
362
+ - wiltur
363
+ - amebit
364
+ - sagspa1
365
+ - tunswa
366
+ - wooduc
367
+ - renpha
368
+ - whtswi
369
+ - bongul
370
+ - norhar2
371
+ - doccor
372
+ - lotduc
373
+ - chukar
374
+ - horgre
375
+ - nrwswa
376
+ - sheowl
377
+ - wesblu
378
+ - whfibi
379
+ - buwtea
380
+ - norpin
381
+ - eargre
382
+ - rebsap
383
+ - lewwoo
384
+ - rebmer
385
+ - wessan
386
+ - chiswi
387
+ - lecthr
388
+ - rthhum
389
+ - moublu
390
+ - amewig
391
+ - rinduc
392
+ - shshaw
393
+ - rufgro
394
+ - swahaw
395
+ - coshum
396
+ - truswa
397
+ - rudduc
398
+ - buffle
399
+ - hoomer
400
+ - gcrfin
401
+ - redhea
402
+ - <unk>
403
+ token_type: word
404
+ init: xavier_normal
405
+ input_size: 1
406
+ use_preprocessor: true
407
+ frontend: null
408
+ frontend_conf: {}
409
+ specaug: null
410
+ specaug_conf: {}
411
+ normalize: null
412
+ normalize_conf: {}
413
+ preencoder: null
414
+ preencoder_conf: {}
415
+ encoder: beats
416
+ encoder_conf:
417
+ beats_ckpt_path: /compute/babel-13-33/sbharad2/models/BEATs/BEATs_iter3.pt
418
+ beats_config:
419
+ layer_wise_gradient_decay_ratio: 0.3
420
+ encoder_layerdrop: 0.1
421
+ dropout: 0.0
422
+ use_weighted_representation: false
423
+ specaug_config:
424
+ apply_time_warp: true
425
+ apply_freq_mask: false
426
+ apply_time_mask: true
427
+ time_mask_width_ratio_range:
428
+ - 0
429
+ - 0.06
430
+ num_time_mask: 1
431
+ roll_augment: true
432
+ roll_interval: 1
433
+ decoder: linear
434
+ decoder_conf: {}
435
+ model: espnet
436
+ model_conf:
437
+ classification_type: multi-class
438
+ mixup_augmentation: false
439
+ lsm_weight: 0.1
440
+ required:
441
+ - output_dir
442
+ - token_list
443
+ version: '202412'
444
+ distributed: false
bean.cbi/exp/cls_cbi.20250107.141123/images/acc.png ADDED
bean.cbi/exp/cls_cbi.20250107.141123/images/backward_time.png ADDED
bean.cbi/exp/cls_cbi.20250107.141123/images/clip.png ADDED
bean.cbi/exp/cls_cbi.20250107.141123/images/forward_time.png ADDED
bean.cbi/exp/cls_cbi.20250107.141123/images/gpu_max_cached_mem_GB.png ADDED
bean.cbi/exp/cls_cbi.20250107.141123/images/grad_norm.png ADDED
bean.cbi/exp/cls_cbi.20250107.141123/images/iter_time.png ADDED
bean.cbi/exp/cls_cbi.20250107.141123/images/loss.png ADDED
bean.cbi/exp/cls_cbi.20250107.141123/images/loss_scale.png ADDED
bean.cbi/exp/cls_cbi.20250107.141123/images/macro_precision.png ADDED
bean.cbi/exp/cls_cbi.20250107.141123/images/optim0_lr0.png ADDED
bean.cbi/exp/cls_cbi.20250107.141123/images/optim_step_time.png ADDED
bean.cbi/exp/cls_cbi.20250107.141123/images/train_time.png ADDED
meta.yaml ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ espnet: '202412'
2
+ files:
3
+ classification_model_file: /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/201epoch.pth
4
+ python: "3.9.20 (main, Oct 3 2024, 07:27:41) \n[GCC 11.2.0]"
5
+ timestamp: 1736334809.745881
6
+ torch: 2.4.0
7
+ yaml_files:
8
+ classification_train_config: /compute/babel-11-13/sbharad2/beats_run/bean.cbi/exp/cls_cbi.20250107.141123/config.yaml