add icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/README.md

@@ -0,0 +1,18 @@
+---
+language:
+- en
+library_name: k2
+tags:
+- automatic-speech-recognition
+- k2
+widget:
+- example_title: Librispeech sample 1
+  src: https://huggingface.co/csukuangfj/icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/resolve/main/test_wavs/1089-134686-0001.wav
+- example_title: Librispeech sample 2
+  src: https://huggingface.co/csukuangfj/icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/resolve/main/test_wavs/1221-135766-0001.wav
+- example_title: Librispeech sample 3
+  src: https://huggingface.co/csukuangfj/icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/resolve/main/test_wavs/1221-135766-0002.wav
+---
+
+# Introduction
+See https://github.com/k2-fsa/icefall/pull/363

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/config.json

@@ -0,0 +1,8 @@
+{
+  "hf_demo": {
+    "nn_model_filename": "exp/cpu_jit-torch-1.6.0.pt",
+    "bpe_model_filename": "data/lang_bpe_500/bpe.model",
+    "num_active_paths": 4,
+    "sample_rate": 16000
+  }
+}

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/data/lang_bpe_500/L.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1b88996f918737fba67fbd29152018b51a537c16ce0718a2b43d5140583224e
+size 19025703

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/data/lang_bpe_500/LG.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3bb9f021c7aad79d45dc275ba8154a430c4f660a319dcb872cd52500f25553d6
+size 249852195

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/data/lang_bpe_500/Linv.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbc8b3687a1b8f0811a84106b3b310642566c7b1bc282a929878f9269507a2c6
+size 19025703

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/data/lang_bpe_500/bpe.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c53433de083c4a6ad12d034550ef22de68cec62c4f58932a7b6b8b2f1e743fa5
+size 244865

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/data/lang_bpe_500/tokens.txt

@@ -0,0 +1,502 @@
+<blk> 0
+<sos/eos> 1
+<unk> 2
+S 3
+▁THE 4
+▁A 5
+T 6
+▁AND 7
+ED 8
+▁OF 9
+▁TO 10
+E 11
+D 12
+N 13
+ING 14
+▁IN 15
+Y 16
+M 17
+C 18
+▁I 19
+A 20
+P 21
+▁HE 22
+R 23
+O 24
+L 25
+RE 26
+I 27
+U 28
+ER 29
+▁IT 30
+LY 31
+▁THAT 32
+▁WAS 33
+▁ 34
+▁S 35
+AR 36
+▁BE 37
+F 38
+▁C 39
+IN 40
+B 41
+▁FOR 42
+OR 43
+LE 44
+' 45
+▁HIS 46
+▁YOU 47
+AL 48
+▁RE 49
+V 50
+▁B 51
+G 52
+RI 53
+▁E 54
+▁WITH 55
+▁T 56
+▁AS 57
+LL 58
+▁P 59
+▁HER 60
+ST 61
+▁HAD 62
+▁SO 63
+▁F 64
+W 65
+CE 66
+▁IS 67
+ND 68
+▁NOT 69
+TH 70
+▁BUT 71
+EN 72
+▁SHE 73
+▁ON 74
+VE 75
+ON 76
+SE 77
+▁DE 78
+UR 79
+▁G 80
+CH 81
+K 82
+TER 83
+▁AT 84
+IT 85
+▁ME 86
+RO 87
+NE 88
+RA 89
+ES 90
+IL 91
+NG 92
+IC 93
+▁NO 94
+▁HIM 95
+ENT 96
+IR 97
+▁WE 98
+H 99
+▁DO 100
+▁ALL 101
+▁HAVE 102
+LO 103
+▁BY 104
+▁MY 105
+▁MO 106
+▁THIS 107
+LA 108
+▁ST 109
+▁WHICH 110
+▁CON 111
+▁THEY 112
+CK 113
+TE 114
+▁SAID 115
+▁FROM 116
+▁GO 117
+▁WHO 118
+▁TH 119
+▁OR 120
+▁D 121
+▁W 122
+VER 123
+LI 124
+▁SE 125
+▁ONE 126
+▁CA 127
+▁AN 128
+▁LA 129
+▁WERE 130
+EL 131
+▁HA 132
+▁MAN 133
+▁FA 134
+▁EX 135
+AD 136
+▁SU 137
+RY 138
+▁MI 139
+AT 140
+▁BO 141
+▁WHEN 142
+AN 143
+THER 144
+PP 145
+ATION 146
+▁FI 147
+▁WOULD 148
+▁PRO 149
+OW 150
+ET 151
+▁O 152
+▁THERE 153
+▁HO 154
+ION 155
+▁WHAT 156
+▁FE 157
+▁PA 158
+US 159
+MENT 160
+▁MA 161
+UT 162
+▁OUT 163
+▁THEIR 164
+▁IF 165
+▁LI 166
+▁K 167
+▁WILL 168
+▁ARE 169
+ID 170
+▁RO 171
+DE 172
+TION 173
+▁WA 174
+PE 175
+▁UP 176
+▁SP 177
+▁PO 178
+IGHT 179
+▁UN 180
+RU 181
+▁LO 182
+AS 183
+OL 184
+▁LE 185
+▁BEEN 186
+▁SH 187
+▁RA 188
+▁SEE 189
+KE 190
+UL 191
+TED 192
+▁SA 193
+UN 194
+UND 195
+ANT 196
+▁NE 197
+IS 198
+▁THEM 199
+CI 200
+GE 201
+▁COULD 202
+▁DIS 203
+OM 204
+ISH 205
+HE 206
+EST 207
+▁SOME 208
+ENCE 209
+ITY 210
+IVE 211
+▁US 212
+▁MORE 213
+▁EN 214
+ARD 215
+ATE 216
+▁YOUR 217
+▁INTO 218
+▁KNOW 219
+▁CO 220
+ANCE 221
+▁TIME 222
+▁WI 223
+▁YE 224
+AGE 225
+▁NOW 226
+TI 227
+FF 228
+ABLE 229
+▁VERY 230
+▁LIKE 231
+AM 232
+HI 233
+Z 234
+▁OTHER 235
+▁THAN 236
+▁LITTLE 237
+▁DID 238
+▁LOOK 239
+TY 240
+ERS 241
+▁CAN 242
+▁CHA 243
+▁AR 244
+X 245
+FUL 246
+UGH 247
+▁BA 248
+▁DAY 249
+▁ABOUT 250
+TEN 251
+IM 252
+▁ANY 253
+▁PRE 254
+▁OVER 255
+IES 256
+NESS 257
+ME 258
+BLE 259
+▁M 260
+ROW 261
+▁HAS 262
+▁GREAT 263
+▁VI 264
+TA 265
+▁AFTER 266
+PER 267
+▁AGAIN 268
+HO 269
+SH 270
+▁UPON 271
+▁DI 272
+▁HAND 273
+▁COM 274
+IST 275
+TURE 276
+▁STA 277
+▁THEN 278
+▁SHOULD 279
+▁GA 280
+OUS 281
+OUR 282
+▁WELL 283
+▁ONLY 284
+MAN 285
+▁GOOD 286
+▁TWO 287
+▁MAR 288
+▁SAY 289
+▁HU 290
+TING 291
+▁OUR 292
+RESS 293
+▁DOWN 294
+IOUS 295
+▁BEFORE 296
+▁DA 297
+▁NA 298
+QUI 299
+▁MADE 300
+▁EVERY 301
+▁OLD 302
+▁EVEN 303
+IG 304
+▁COME 305
+▁GRA 306
+▁RI 307
+▁LONG 308
+OT 309
+SIDE 310
+WARD 311
+▁FO 312
+▁WHERE 313
+MO 314
+LESS 315
+▁SC 316
+▁MUST 317
+▁NEVER 318
+▁HOW 319
+▁CAME 320
+▁SUCH 321
+▁RU 322
+▁TAKE 323
+▁WO 324
+▁CAR 325
+UM 326
+AK 327
+▁THINK 328
+▁MUCH 329
+▁MISTER 330
+▁MAY 331
+▁JO 332
+▁WAY 333
+▁COMP 334
+▁THOUGHT 335
+▁STO 336
+▁MEN 337
+▁BACK 338
+▁DON 339
+J 340
+▁LET 341
+▁TRA 342
+▁FIRST 343
+▁JUST 344
+▁VA 345
+▁OWN 346
+▁PLA 347
+▁MAKE 348
+ATED 349
+▁HIMSELF 350
+▁WENT 351
+▁PI 352
+GG 353
+RING 354
+▁DU 355
+▁MIGHT 356
+▁PART 357
+▁GIVE 358
+▁IMP 359
+▁BU 360
+▁PER 361
+▁PLACE 362
+▁HOUSE 363
+▁THROUGH 364
+IAN 365
+▁SW 366
+▁UNDER 367
+QUE 368
+▁AWAY 369
+▁LOVE 370
+QUA 371
+▁LIFE 372
+▁GET 373
+▁WITHOUT 374
+▁PASS 375
+▁TURN 376
+IGN 377
+▁HEAD 378
+▁MOST 379
+▁THOSE 380
+▁SHALL 381
+▁EYES 382
+▁COL 383
+▁STILL 384
+▁NIGHT 385
+▁NOTHING 386
+ITION 387
+HA 388
+▁TELL 389
+▁WORK 390
+▁LAST 391
+▁NEW 392
+▁FACE 393
+▁HI 394
+▁WORD 395
+▁FOUND 396
+▁COUNT 397
+▁OB 398
+▁WHILE 399
+▁SHA 400
+▁MEAN 401
+▁SAW 402
+▁PEOPLE 403
+▁FRIEND 404
+▁THREE 405
+▁ROOM 406
+▁SAME 407
+▁THOUGH 408
+▁RIGHT 409
+▁CHILD 410
+▁FATHER 411
+▁ANOTHER 412
+▁HEART 413
+▁WANT 414
+▁TOOK 415
+OOK 416
+▁LIGHT 417
+▁MISSUS 418
+▁OPEN 419
+▁JU 420
+▁ASKED 421
+PORT 422
+▁LEFT 423
+▁JA 424
+▁WORLD 425
+▁HOME 426
+▁WHY 427
+▁ALWAYS 428
+▁ANSWER 429
+▁SEEMED 430
+▁SOMETHING 431
+▁GIRL 432
+▁BECAUSE 433
+▁NAME 434
+▁TOLD 435
+▁NI 436
+▁HIGH 437
+IZE 438
+▁WOMAN 439
+▁FOLLOW 440
+▁RETURN 441
+▁KNEW 442
+▁EACH 443
+▁KIND 444
+▁JE 445
+▁ACT 446
+▁LU 447
+▁CERTAIN 448
+▁YEARS 449
+▁QUITE 450
+▁APPEAR 451
+▁BETTER 452
+▁HALF 453
+▁PRESENT 454
+▁PRINCE 455
+SHIP 456
+▁ALSO 457
+▁BEGAN 458
+▁HAVING 459
+▁ENOUGH 460
+▁PERSON 461
+▁LADY 462
+▁WHITE 463
+▁COURSE 464
+▁VOICE 465
+▁SPEAK 466
+▁POWER 467
+▁MORNING 468
+▁BETWEEN 469
+▁AMONG 470
+▁KEEP 471
+▁WALK 472
+▁MATTER 473
+▁TEA 474
+▁BELIEVE 475
+▁SMALL 476
+▁TALK 477
+▁FELT 478
+▁HORSE 479
+▁MYSELF 480
+▁SIX 481
+▁HOWEVER 482
+▁FULL 483
+▁HERSELF 484
+▁POINT 485
+▁STOOD 486
+▁HUNDRED 487
+▁ALMOST 488
+▁SINCE 489
+▁LARGE 490
+▁LEAVE 491
+▁PERHAPS 492
+▁DARK 493
+▁SUDDEN 494
+▁REPLIED 495
+▁ANYTHING 496
+▁WONDER 497
+▁UNTIL 498
+Q 499
+#0 500
+#1 501

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/decoding-results/fast_beam_search/log-decode-iter-1224000-avg-14-beam-4.0-max-contexts-32-max-states-8-2022-05-13-10-11-35

@@ -0,0 +1,31 @@
+2022-05-13 10:11:35,141 INFO [decode.py:531] Decoding started
+2022-05-13 10:11:35,142 INFO [decode.py:537] Device: cuda:0
+2022-05-13 10:11:35,144 INFO [decode.py:547] {'best_train_loss': inf, 'best_valid_loss': inf, 'best_train_epoch': -1, 'best_valid_epoch': -1, 'batch_idx_train': 0, 'log_interval': 50, 'reset_interval': 200, 'valid_interval': 3000, 'feature_dim': 80, 'subsampling_factor': 4, 'encoder_dim': 512, 'nhead': 8, 'dim_feedforward': 2048, 'num_encoder_layers': 12, 'decoder_dim': 512, 'joiner_dim': 512, 'model_warm_step': 3000, 'env_info': {'k2-version': '1.15.1', 'k2-build-type': 'Release', 'k2-with-cuda': True, 'k2-git-sha1': 'f8d2dba06c000ffee36aab5b66f24e7c9809f116', 'k2-git-date': 'Thu Apr 21 12:20:34 2022', 'lhotse-version': '1.1.0.dev+missing.version.file', 'torch-version': '1.10.0+cu102', 'torch-cuda-available': True, 'torch-cuda-version': '10.2', 'python-version': '3.8', 'icefall-git-branch': 'modified-conformer-with-multi-datasets', 'icefall-git-sha1': '00fd664-dirty', 'icefall-git-date': 'Fri Apr 29 15:28:42 2022', 'icefall-path': '/ceph-fj/fangjun/open-source-2/icefall-multi-4', 'k2-path': '/ceph-fj/fangjun/open-source-2/k2-multi-22/k2/python/k2/__init__.py', 'lhotse-path': '/ceph-fj/fangjun/open-source-2/lhotse-multi-3/lhotse/__init__.py', 'hostname': 'de-74279-k2-train-2-0307200233-b554c565c-lf9qd', 'IP address': '10.177.74.201'}, 'epoch': 28, 'iter': 1224000, 'avg': 14, 'exp_dir': PosixPath('pruned_transducer_stateless3/exp-0.9'), 'bpe_model': 'data/lang_bpe_500/bpe.model', 'decoding_method': 'fast_beam_search', 'beam_size': 4, 'beam': 4.0, 'max_contexts': 32, 'max_states': 8, 'context_size': 2, 'max_sym_per_frame': 1, 'num_paths': 100, 'nbest_scale': 0.5, 'max_duration': 600, 'bucketing_sampler': True, 'num_buckets': 30, 'shuffle': True, 'return_cuts': True, 'num_workers': 2, 'on_the_fly_num_workers': 0, 'enable_spec_aug': True, 'spec_aug_time_warp_factor': 80, 'enable_musan': True, 'manifest_dir': PosixPath('data/fbank'), 'on_the_fly_feats': False, 'res_dir': PosixPath('pruned_transducer_stateless3/exp-0.9/fast_beam_search'), 'suffix': 'iter-1224000-avg-14-beam-4.0-max-contexts-32-max-states-8', 'blank_id': 0, 'unk_id': 2, 'vocab_size': 500}
+2022-05-13 10:11:35,144 INFO [decode.py:549] About to create model
+2022-05-13 10:11:35,622 INFO [decode.py:566] averaging ['pruned_transducer_stateless3/exp-0.9/checkpoint-1224000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1216000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1208000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1200000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1192000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1184000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1176000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1168000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1160000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1152000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1144000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1136000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1128000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1120000.pt']
+2022-05-13 10:11:56,508 INFO [decode.py:595] Number of model parameters: 80199888
+2022-05-13 10:11:56,508 INFO [librispeech.py:58] About to get test-clean cuts from data/fbank/cuts_test-clean.json.gz
+2022-05-13 10:11:56,621 INFO [librispeech.py:63] About to get test-other cuts from data/fbank/cuts_test-other.json.gz
+2022-05-13 10:11:59,087 INFO [decode.py:438] batch 0/?, cuts processed until now is 123
+2022-05-13 10:12:15,259 INFO [decode.py:438] batch 10/?, cuts processed until now is 945
+2022-05-13 10:12:37,390 INFO [decode.py:438] batch 20/?, cuts processed until now is 1558
+2022-05-13 10:12:52,210 INFO [decode.py:438] batch 30/?, cuts processed until now is 2383
+2022-05-13 10:13:08,683 INFO [decode.py:455] The transcripts are stored in pruned_transducer_stateless3/exp-0.9/fast_beam_search/recogs-test-clean-beam_4.0_max_contexts_32_max_states_8-iter-1224000-avg-14-beam-4.0-max-contexts-32-max-states-8.txt
+2022-05-13 10:13:08,776 INFO [utils.py:405] [test-clean-beam_4.0_max_contexts_32_max_states_8] %WER 2.10% [1102 / 52576, 115 ins, 115 del, 872 sub ]
+2022-05-13 10:13:09,068 INFO [decode.py:468] Wrote detailed error stats to pruned_transducer_stateless3/exp-0.9/fast_beam_search/errs-test-clean-beam_4.0_max_contexts_32_max_states_8-iter-1224000-avg-14-beam-4.0-max-contexts-32-max-states-8.txt
+2022-05-13 10:13:09,069 INFO [decode.py:485] 
+For test-clean, WER of different settings are:
+beam_4.0_max_contexts_32_max_states_8	2.1	best for test-clean
+
+2022-05-13 10:13:10,920 INFO [decode.py:438] batch 0/?, cuts processed until now is 138
+2022-05-13 10:13:25,921 INFO [decode.py:438] batch 10/?, cuts processed until now is 1070
+2022-05-13 10:13:46,529 INFO [decode.py:438] batch 20/?, cuts processed until now is 1765
+2022-05-13 10:13:59,688 INFO [decode.py:438] batch 30/?, cuts processed until now is 2653
+2022-05-13 10:14:14,049 INFO [decode.py:455] The transcripts are stored in pruned_transducer_stateless3/exp-0.9/fast_beam_search/recogs-test-other-beam_4.0_max_contexts_32_max_states_8-iter-1224000-avg-14-beam-4.0-max-contexts-32-max-states-8.txt
+2022-05-13 10:14:14,142 INFO [utils.py:405] [test-other-beam_4.0_max_contexts_32_max_states_8] %WER 4.68% [2449 / 52343, 245 ins, 223 del, 1981 sub ]
+2022-05-13 10:14:14,353 INFO [decode.py:468] Wrote detailed error stats to pruned_transducer_stateless3/exp-0.9/fast_beam_search/errs-test-other-beam_4.0_max_contexts_32_max_states_8-iter-1224000-avg-14-beam-4.0-max-contexts-32-max-states-8.txt
+2022-05-13 10:14:14,354 INFO [decode.py:485] 
+For test-other, WER of different settings are:
+beam_4.0_max_contexts_32_max_states_8	4.68	best for test-other
+
+2022-05-13 10:14:14,354 INFO [decode.py:624] Done!

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/decoding-results/greedy_search/log-decode-iter-1224000-avg-14-context-2-max-sym-per-frame-1-2022-05-13-09-30-57

@@ -0,0 +1,25 @@
+2022-05-13 09:30:57,234 INFO [decode.py:531] Decoding started
+2022-05-13 09:30:57,234 INFO [decode.py:537] Device: cuda:0
+2022-05-13 09:30:57,237 INFO [decode.py:547] {'best_train_loss': inf, 'best_valid_loss': inf, 'best_train_epoch': -1, 'best_valid_epoch': -1, 'batch_idx_train': 0, 'log_interval': 50, 'reset_interval': 200, 'valid_interval': 3000, 'feature_dim': 80, 'subsampling_factor': 4, 'encoder_dim': 512, 'nhead': 8, 'dim_feedforward': 2048, 'num_encoder_layers': 12, 'decoder_dim': 512, 'joiner_dim': 512, 'model_warm_step': 3000, 'env_info': {'k2-version': '1.15.1', 'k2-build-type': 'Release', 'k2-with-cuda': True, 'k2-git-sha1': 'f8d2dba06c000ffee36aab5b66f24e7c9809f116', 'k2-git-date': 'Thu Apr 21 12:20:34 2022', 'lhotse-version': '1.1.0.dev+missing.version.file', 'torch-version': '1.10.0+cu102', 'torch-cuda-available': True, 'torch-cuda-version': '10.2', 'python-version': '3.8', 'icefall-git-branch': 'modified-conformer-with-multi-datasets', 'icefall-git-sha1': '00fd664-dirty', 'icefall-git-date': 'Fri Apr 29 15:28:42 2022', 'icefall-path': '/ceph-fj/fangjun/open-source-2/icefall-multi-4', 'k2-path': '/ceph-fj/fangjun/open-source-2/k2-multi-22/k2/python/k2/__init__.py', 'lhotse-path': '/ceph-fj/fangjun/open-source-2/lhotse-multi-3/lhotse/__init__.py', 'hostname': 'de-74279-k2-train-2-0307200233-b554c565c-lf9qd', 'IP address': '10.177.74.201'}, 'epoch': 28, 'iter': 1224000, 'avg': 14, 'exp_dir': PosixPath('pruned_transducer_stateless3/exp-0.9'), 'bpe_model': 'data/lang_bpe_500/bpe.model', 'decoding_method': 'greedy_search', 'beam_size': 4, 'beam': 4.0, 'max_contexts': 32, 'max_states': 8, 'context_size': 2, 'max_sym_per_frame': 1, 'num_paths': 100, 'nbest_scale': 0.5, 'max_duration': 600, 'bucketing_sampler': True, 'num_buckets': 30, 'shuffle': True, 'return_cuts': True, 'num_workers': 2, 'on_the_fly_num_workers': 0, 'enable_spec_aug': True, 'spec_aug_time_warp_factor': 80, 'enable_musan': True, 'manifest_dir': PosixPath('data/fbank'), 'on_the_fly_feats': False, 'res_dir': PosixPath('pruned_transducer_stateless3/exp-0.9/greedy_search'), 'suffix': 'iter-1224000-avg-14-context-2-max-sym-per-frame-1', 'blank_id': 0, 'unk_id': 2, 'vocab_size': 500}
+2022-05-13 09:30:57,237 INFO [decode.py:549] About to create model
+2022-05-13 09:30:57,725 INFO [decode.py:566] averaging ['pruned_transducer_stateless3/exp-0.9/checkpoint-1224000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1216000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1208000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1200000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1192000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1184000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1176000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1168000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1160000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1152000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1144000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1136000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1128000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1120000.pt']
+2022-05-13 09:31:34,244 INFO [decode.py:595] Number of model parameters: 80199888
+2022-05-13 09:31:34,245 INFO [librispeech.py:58] About to get test-clean cuts from data/fbank/cuts_test-clean.json.gz
+2022-05-13 09:31:34,385 INFO [librispeech.py:63] About to get test-other cuts from data/fbank/cuts_test-other.json.gz
+2022-05-13 09:31:36,351 INFO [decode.py:438] batch 0/?, cuts processed until now is 123
+2022-05-13 09:32:07,927 INFO [decode.py:455] The transcripts are stored in pruned_transducer_stateless3/exp-0.9/greedy_search/recogs-test-clean-greedy_search-iter-1224000-avg-14-context-2-max-sym-per-frame-1.txt
+2022-05-13 09:32:08,021 INFO [utils.py:405] [test-clean-greedy_search] %WER 2.03% [1069 / 52576, 108 ins, 91 del, 870 sub ]
+2022-05-13 09:32:08,353 INFO [decode.py:468] Wrote detailed error stats to pruned_transducer_stateless3/exp-0.9/greedy_search/errs-test-clean-greedy_search-iter-1224000-avg-14-context-2-max-sym-per-frame-1.txt
+2022-05-13 09:32:08,354 INFO [decode.py:485] 
+For test-clean, WER of different settings are:
+greedy_search	2.03	best for test-clean
+
+2022-05-13 09:32:09,808 INFO [decode.py:438] batch 0/?, cuts processed until now is 138
+2022-05-13 09:32:39,850 INFO [decode.py:455] The transcripts are stored in pruned_transducer_stateless3/exp-0.9/greedy_search/recogs-test-other-greedy_search-iter-1224000-avg-14-context-2-max-sym-per-frame-1.txt
+2022-05-13 09:32:39,924 INFO [utils.py:405] [test-other-greedy_search] %WER 4.70% [2461 / 52343, 227 ins, 234 del, 2000 sub ]
+2022-05-13 09:32:40,108 INFO [decode.py:468] Wrote detailed error stats to pruned_transducer_stateless3/exp-0.9/greedy_search/errs-test-other-greedy_search-iter-1224000-avg-14-context-2-max-sym-per-frame-1.txt
+2022-05-13 09:32:40,109 INFO [decode.py:485] 
+For test-other, WER of different settings are:
+greedy_search	4.7	best for test-other
+
+2022-05-13 09:32:40,109 INFO [decode.py:624] Done!

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/decoding-results/modified_beam_search/log-decode-iter-1224000-avg-14-modified_beam_search-beam-size-4-2022-05-13-11-51-15

@@ -0,0 +1,31 @@
+2022-05-13 11:51:15,903 INFO [decode.py:531] Decoding started
+2022-05-13 11:51:15,904 INFO [decode.py:537] Device: cuda:0
+2022-05-13 11:51:15,906 INFO [decode.py:547] {'best_train_loss': inf, 'best_valid_loss': inf, 'best_train_epoch': -1, 'best_valid_epoch': -1, 'batch_idx_train': 0, 'log_interval': 50, 'reset_interval': 200, 'valid_interval': 3000, 'feature_dim': 80, 'subsampling_factor': 4, 'encoder_dim': 512, 'nhead': 8, 'dim_feedforward': 2048, 'num_encoder_layers': 12, 'decoder_dim': 512, 'joiner_dim': 512, 'model_warm_step': 3000, 'env_info': {'k2-version': '1.15.1', 'k2-build-type': 'Release', 'k2-with-cuda': True, 'k2-git-sha1': 'f8d2dba06c000ffee36aab5b66f24e7c9809f116', 'k2-git-date': 'Thu Apr 21 12:20:34 2022', 'lhotse-version': '1.1.0.dev+missing.version.file', 'torch-version': '1.10.0+cu102', 'torch-cuda-available': True, 'torch-cuda-version': '10.2', 'python-version': '3.8', 'icefall-git-branch': 'modified-conformer-with-multi-datasets', 'icefall-git-sha1': '00fd664-dirty', 'icefall-git-date': 'Fri Apr 29 15:28:42 2022', 'icefall-path': '/ceph-fj/fangjun/open-source-2/icefall-multi-4', 'k2-path': '/ceph-fj/fangjun/open-source-2/k2-multi-22/k2/python/k2/__init__.py', 'lhotse-path': '/ceph-fj/fangjun/open-source-2/lhotse-multi-3/lhotse/__init__.py', 'hostname': 'de-74279-k2-train-2-0307200233-b554c565c-lf9qd', 'IP address': '10.177.74.201'}, 'epoch': 28, 'iter': 1224000, 'avg': 14, 'exp_dir': PosixPath('pruned_transducer_stateless3/exp-0.9'), 'bpe_model': 'data/lang_bpe_500/bpe.model', 'decoding_method': 'modified_beam_search', 'beam_size': 4, 'beam': 4.0, 'max_contexts': 32, 'max_states': 8, 'context_size': 2, 'max_sym_per_frame': 1, 'num_paths': 100, 'nbest_scale': 0.5, 'max_duration': 600, 'bucketing_sampler': True, 'num_buckets': 30, 'shuffle': True, 'return_cuts': True, 'num_workers': 2, 'on_the_fly_num_workers': 0, 'enable_spec_aug': True, 'spec_aug_time_warp_factor': 80, 'enable_musan': True, 'manifest_dir': PosixPath('data/fbank'), 'on_the_fly_feats': False, 'res_dir': PosixPath('pruned_transducer_stateless3/exp-0.9/modified_beam_search'), 'suffix': 'iter-1224000-avg-14-modified_beam_search-beam-size-4', 'blank_id': 0, 'unk_id': 2, 'vocab_size': 500}
+2022-05-13 11:51:15,907 INFO [decode.py:549] About to create model
+2022-05-13 11:51:16,527 INFO [decode.py:566] averaging ['pruned_transducer_stateless3/exp-0.9/checkpoint-1224000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1216000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1208000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1200000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1192000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1184000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1176000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1168000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1160000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1152000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1144000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1136000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1128000.pt', 'pruned_transducer_stateless3/exp-0.9/checkpoint-1120000.pt']
+2022-05-13 11:53:04,387 INFO [decode.py:595] Number of model parameters: 80199888
+2022-05-13 11:53:04,388 INFO [librispeech.py:58] About to get test-clean cuts from data/fbank/cuts_test-clean.json.gz
+2022-05-13 11:53:04,507 INFO [librispeech.py:63] About to get test-other cuts from data/fbank/cuts_test-other.json.gz
+2022-05-13 11:53:37,741 INFO [decode.py:438] batch 0/?, cuts processed until now is 123
+2022-05-13 11:58:10,766 INFO [decode.py:438] batch 10/?, cuts processed until now is 945
+2022-05-13 12:01:44,946 INFO [decode.py:438] batch 20/?, cuts processed until now is 1558
+2022-05-13 12:04:34,252 INFO [decode.py:438] batch 30/?, cuts processed until now is 2383
+2022-05-13 12:07:23,218 INFO [decode.py:455] The transcripts are stored in pruned_transducer_stateless3/exp-0.9/modified_beam_search/recogs-test-clean-beam_size_4-iter-1224000-avg-14-modified_beam_search-beam-size-4.txt
+2022-05-13 12:07:23,293 INFO [utils.py:405] [test-clean-beam_size_4] %WER 2.00% [1049 / 52576, 111 ins, 85 del, 853 sub ]
+2022-05-13 12:07:23,471 INFO [decode.py:468] Wrote detailed error stats to pruned_transducer_stateless3/exp-0.9/modified_beam_search/errs-test-clean-beam_size_4-iter-1224000-avg-14-modified_beam_search-beam-size-4.txt
+2022-05-13 12:07:23,472 INFO [decode.py:485] 
+For test-clean, WER of different settings are:
+beam_size_4	2.0	best for test-clean
+
+2022-05-13 12:07:37,469 INFO [decode.py:438] batch 0/?, cuts processed until now is 138
+2022-05-13 12:10:52,601 INFO [decode.py:438] batch 10/?, cuts processed until now is 1070
+2022-05-13 12:13:58,872 INFO [decode.py:438] batch 20/?, cuts processed until now is 1765
+2022-05-13 12:19:26,416 INFO [decode.py:438] batch 30/?, cuts processed until now is 2653
+2022-05-13 12:22:25,967 INFO [decode.py:455] The transcripts are stored in pruned_transducer_stateless3/exp-0.9/modified_beam_search/recogs-test-other-beam_size_4-iter-1224000-avg-14-modified_beam_search-beam-size-4.txt
+2022-05-13 12:22:26,158 INFO [utils.py:405] [test-other-beam_size_4] %WER 4.63% [2422 / 52343, 247 ins, 197 del, 1978 sub ]
+2022-05-13 12:22:26,334 INFO [decode.py:468] Wrote detailed error stats to pruned_transducer_stateless3/exp-0.9/modified_beam_search/errs-test-other-beam_size_4-iter-1224000-avg-14-modified_beam_search-beam-size-4.txt
+2022-05-13 12:22:26,335 INFO [decode.py:485] 
+For test-other, WER of different settings are:
+beam_size_4	4.63	best for test-other
+
+2022-05-13 12:22:26,335 INFO [decode.py:624] Done!

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/.gitattributes

@@ -0,0 +1,10 @@
+tensorboard/events.out.tfevents.1651022882.de-74279-k2-train-6-0415002726-7dc5bf9fdc-w24k9.3431310.0 filter=lfs diff=lfs merge=lfs -text
+log filter=lfs diff=lfs merge=lfs -text
+log/log-train-2022-04-27-09-28-02-6 filter=lfs diff=lfs merge=lfs -text
+log/log-train-2022-04-27-09-28-02-7 filter=lfs diff=lfs merge=lfs -text
+log/log-train-2022-04-27-09-28-02-0 filter=lfs diff=lfs merge=lfs -text
+log/log-train-2022-04-27-09-28-02-1 filter=lfs diff=lfs merge=lfs -text
+log/log-train-2022-04-27-09-28-02-2 filter=lfs diff=lfs merge=lfs -text
+log/log-train-2022-04-27-09-28-02-3 filter=lfs diff=lfs merge=lfs -text
+log/log-train-2022-04-27-09-28-02-4 filter=lfs diff=lfs merge=lfs -text
+log/log-train-2022-04-27-09-28-02-5 filter=lfs diff=lfs merge=lfs -text

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/cpu_jit-torch-1.10.0.pt

@@ -0,0 +1 @@
+cpu_jit.pt

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/cpu_jit-torch-1.6.0.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad1617a2d7e4a71838c396cb57560886ab4d4b71d51af71d8678834d35226716
+size 341490258

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/cpu_jit.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd9f203088b2ec97013e4bf27d2b531d2f27bb472cdda560cbd4c9670e8b0956
+size 341499218

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/decode-pruned-rnnt-3-giga.sh

@@ -0,0 +1,17 @@
+#!/usr/bin/env bash
+
+export CUDA_VISIBLE_DEVICES="7"
+
+set -ex
+
+for iter in 1224000; do
+  for avg in 14; do
+    ./pruned_transducer_stateless3/decode-giga.py \
+      --iter $iter \
+      --avg $avg \
+      --exp-dir ./pruned_transducer_stateless3/exp-0.9 \
+      --max-duration 600 \
+      --decoding-method fast_beam_search \
+      --max-sym-per-frame 1
+  done
+done

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/decode-pruned-rnnt-3.sh

@@ -0,0 +1,21 @@
+#!/usr/bin/env bash
+
+export CUDA_VISIBLE_DEVICES="7"
+
+set -ex
+
+for iter in 1224000; do
+  for avg in 14; do
+    for method in greedy_search modified_beam_search fast_beam_search ; do
+      ./pruned_transducer_stateless3/decode.py \
+        --iter $iter \
+        --avg $avg \
+        --exp-dir ./pruned_transducer_stateless3/exp-0.9/ \
+        --max-duration 600 \
+        --decoding-method $method \
+        --max-sym-per-frame 1 \
+        --beam 4 \
+        --max-contexts 32
+    done
+  done
+done

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/export.sh

@@ -0,0 +1,17 @@
+#!/usr/bin/env bash
+
+set -e
+
+export CUDA_VISIBLE_DEVICES=7
+
+iter=1224000
+avg=14
+
+
+./pruned_transducer_stateless3/export.py \
+  --exp-dir ./pruned_transducer_stateless3/exp-0.9 \
+  --bpe-model data/lang_bpe_500/bpe.model \
+  --iter $iter \
+  --avg $avg
+
+mv -v ./pruned_transducer_stateless3/exp-0.9/pretrained.pt ./pruned_transducer_stateless3/exp-0.9/pretrained-iter-${iter}-avg-${avg}.pt

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/log/log-train-2022-04-27-09-28-02-0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:438e22f2ed49ba049e4c8c353e523868f1baa470b03504759ee133c974e9a340
+size 11931394

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/log/log-train-2022-04-27-09-28-02-1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e374c828c348adfa38263f7819f41048cc71164a0a836e0eae496c6d5f8db7c3
+size 11907390

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/log/log-train-2022-04-27-09-28-02-2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:686061f49a6ce37158717d089aca12733ac14fc9d4a708a0fba4bc08b54e2f53
+size 11907327

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/log/log-train-2022-04-27-09-28-02-3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:abea382eb444e2e92aed1bd68520ee421e87309194beae3ccf14bc302b808d18
+size 11906935

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/log/log-train-2022-04-27-09-28-02-4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8967e0dd070f35bfce49b819408e0ca507e3397f649dddc3e15bcf949f487654
+size 11906895

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/log/log-train-2022-04-27-09-28-02-5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0fdbf7210b862ea50c22309675b10efd263c3cae12ee4373ad0180e4f115add6
+size 11907394

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/log/log-train-2022-04-27-09-28-02-6

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eaac6c135f9fe36c779568cda4e240ada6cefb1aa9795ee34bf0f218a4d8204c
+size 11907157

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/log/log-train-2022-04-27-09-28-02-7

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13d4a44dd6fa37f5a0df06865ec1886d214a05f6cea2779856fc39c5cff243a3
+size 11907279

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/onnx/README.md

@@ -0,0 +1,5 @@
+# Introduction
+
+Please refer to
+<https://github.com/k2-fsa/sherpa-onnx>
+for usage

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/onnx/decoder.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7b09a0bfbe3896a34117c048075d6cc565991eb5b92f3953d732b32cb5954ca
+size 1028645

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/onnx/encoder.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7bc2d5fe5b5e39dc05464826f55515f2cd27fc287cbc742be549d763df74d050
+size 329041122

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/onnx/export-3.sh

@@ -0,0 +1,24 @@
+#!/usr/bin/env bash
+
+set -e
+
+export CUDA_VISIBLE_DEVICES=
+
+epoch=99
+avg=1
+
+if [ ! -f pruned_transducer_stateless3/exp-0.9/epoch-99.pt ]; then
+  mkdir -p pruned_transducer_stateless3/exp-0.9
+  pushd pruned_transducer_stateless3/exp-0.9
+  ln -s /ceph-fj/fangjun/open-source-2/icefall-models/icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/pretrained-iter-1224000-avg-14.pt epoch-99.pt
+  popd
+fi
+
+./pruned_transducer_stateless3/export.py \
+  --exp-dir ./pruned_transducer_stateless3/exp-0.9 \
+  --bpe-model data/lang_bpe_500/bpe.model \
+  --epoch $epoch \
+  --avg $avg \
+  --jit 0 \
+  --jit-trace 0 \
+  --onnx 1

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/onnx/joiner.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:143b4ea96d4e2f230b7dcfafbceedab905b8222e05a4fcccf55bf41f5c1ef727
+size 1026428

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/onnx/joiner_decoder_proj.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:826a5570d1e18caa3d733211e7e52587be6841c7188f0bd6b0e2192b3a6564dc
+size 1050898

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/onnx/joiner_encoder_proj.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2ad02e058ad80a52d8106bcc85ef78a693357a3d9a271fe5374f123020f5c1d
+size 1050898

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/pretrained-iter-1224000-avg-14-torch-1.6.0.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37f9c4b1ff7611893690c0a0dd562444719cd4c0e3d8cfffeadf98e7e0053d0e
+size 321012606

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/pretrained-iter-1224000-avg-14.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb2152556d6029c629c8a23255bc00c1666ae7ee3066d734ddb9262aadfebfc4
+size 321006327

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/pretrained-iter-808000-avg-17.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aaea20eb609384896c391be96160fb946e7ce798e554573237342d2074702d5d
+size 321006327

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/pretrained-iter-824000-avg-18.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50ee614bf40e6b162fb11746d97131de4cb40ef6b10f3e765a6b97de1621a6ee
+size 321006327

icefall-asr-librispeech-pruned-transducer-stateless3-2022-05-13/exp/run-pruned-rnnt-3-2.sh

@@ -0,0 +1,15 @@
+#!/usr/bin/env bash
+
+export CUDA_VISIBLE_DEVICES="0,1,2,3,4,5,6,7"
+
+./pruned_transducer_stateless3/train.py \
+  --world-size 8 \
+  --num-epochs 30 \
+  --start-epoch 0 \
+  --full-libri 1 \
+  --exp-dir pruned_transducer_stateless3/exp-0.9 \
+  --max-duration 300 \
+  --use-fp16 1 \
+  --lr-epochs 4 \
+  --num-workers 2 \
+  --giga-prob 0.9