Add model files

RESULTS.md

@@ -0,0 +1,48 @@
+<!-- Generated by scripts/utils/show_asr_result.sh -->
+# RESULTS
+## Environments
+- date: `Tue Apr 11 01:15:36 EDT 2023`
+- python version: `3.8.16 (default, Mar  2 2023, 03:21:46)  [GCC 11.2.0]`
+- espnet version: `espnet 202301`
+- pytorch version: `pytorch 1.8.1`
+- Git hash: `b0cceeac2ecd330e8270789cef945e49058858fa`
+  - Commit date: `Thu Mar 30 08:26:54 2023 -0400`
+
+## exp/asr_train_asr_e_branchformer_size256_mlp1024_e12_mactrue_raw_en_bpe500_sp
+### WER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/test|1155|27500|94.2|2.5|3.3|0.6|6.4|59.2|
+
+### CER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/test|1155|145066|96.8|0.5|2.7|0.6|3.8|59.2|
+
+### TER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/test|1155|54206|95.8|1.6|2.6|0.5|4.7|59.2|
+
+## exp/asr_train_asr_e_branchformer_size256_mlp1024_e12_mactrue_raw_en_bpe500_sp/decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave
+### WER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|org/dev|507|17783|93.6|3.1|3.3|0.9|7.3|69.0|
+
+### CER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|org/dev|507|95429|96.5|0.7|2.8|0.8|4.4|69.0|
+
+### TER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|org/dev|507|36002|95.4|2.0|2.6|0.8|5.5|69.0|
+

asr_train_asr_e_branchformer_size256_mlp1024_e12_mactrue_raw_en_bpe500_sp/RESULTS.md

@@ -0,0 +1,48 @@
+<!-- Generated by scripts/utils/show_asr_result.sh -->
+# RESULTS
+## Environments
+- date: `Tue Apr 11 01:15:36 EDT 2023`
+- python version: `3.8.16 (default, Mar  2 2023, 03:21:46)  [GCC 11.2.0]`
+- espnet version: `espnet 202301`
+- pytorch version: `pytorch 1.8.1`
+- Git hash: `b0cceeac2ecd330e8270789cef945e49058858fa`
+  - Commit date: `Thu Mar 30 08:26:54 2023 -0400`
+
+## exp/asr_train_asr_e_branchformer_size256_mlp1024_e12_mactrue_raw_en_bpe500_sp
+### WER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/test|1155|27500|94.2|2.5|3.3|0.6|6.4|59.2|
+
+### CER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/test|1155|145066|96.8|0.5|2.7|0.6|3.8|59.2|
+
+### TER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/test|1155|54206|95.8|1.6|2.6|0.5|4.7|59.2|
+
+## exp/asr_train_asr_e_branchformer_size256_mlp1024_e12_mactrue_raw_en_bpe500_sp/decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave
+### WER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|org/dev|507|17783|93.6|3.1|3.3|0.9|7.3|69.0|
+
+### CER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|org/dev|507|95429|96.5|0.7|2.8|0.8|4.4|69.0|
+
+### TER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|org/dev|507|36002|95.4|2.0|2.6|0.8|5.5|69.0|
+

README.md → asr_train_asr_e_branchformer_size256_mlp1024_e12_mactrue_raw_en_bpe500_sp/config.yaml

@@ -1,80 +1,3 @@
----
-datasets:
-- LIUM/tedlium
-language:
-- en
-metrics:
-- wer
-library_name: espnet
-pipeline_tag: automatic-speech-recognition
----
-
-## ESPnet2 ASR model 
-
-### `espnet/tedlium3`
-
-This model was trained by Dongwei Jiang using tedlium3 recipe in [espnet](https://github.com/espnet/espnet/).
-
-### Demo: How to use in ESPnet2
-
-```bash
-cd espnet
-git checkout ff841366229d539eb74d23ac999cae7c0cc62cad
-pip install -e .
-cd egs2/tedlium3/asr1
-./run.sh --skip_data_prep false --skip_train true --download_model espnet/dongwei_tedlium3_asr_e-branchformer_external_lm
-```
-
-<!-- Generated by scripts/utils/show_asr_result.sh -->
-# RESULTS
-## Environments
-- date: `Tue Apr 11 01:15:36 EDT 2023`
-- python version: `3.8.16 (default, Mar  2 2023, 03:21:46)  [GCC 11.2.0]`
-- espnet version: `espnet 202301`
-- pytorch version: `pytorch 1.8.1`
-- Git hash: `b0cceeac2ecd330e8270789cef945e49058858fa`
-  - Commit date: `Thu Mar 30 08:26:54 2023 -0400`
-
-
-## exp/asr_train_asr_e_branchformer_size256_mlp1024_e12_mactrue_raw_en_bpe500_sp
-### WER
-|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
-|---|---|---|---|---|---|---|---|---|
-|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/test|1155|27500|94.2|2.5|3.3|0.6|6.4|59.2|
-
-### CER
-|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
-|---|---|---|---|---|---|---|---|---|
-|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/test|1155|145066|96.8|0.5|2.7|0.6|3.8|59.2|
-
-### TER
-|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
-|---|---|---|---|---|---|---|---|---|
-|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/test|1155|54206|95.8|1.6|2.6|0.5|4.7|59.2|
-
-## exp/asr_train_asr_e_branchformer_size256_mlp1024_e12_mactrue_raw_en_bpe500_sp/decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave
-### WER
-|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
-|---|---|---|---|---|---|---|---|---|
-|org/dev|507|17783|93.6|3.1|3.3|0.9|7.3|69.0|
-
-### CER
-|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
-|---|---|---|---|---|---|---|---|---|
-|org/dev|507|95429|96.5|0.7|2.8|0.8|4.4|69.0|
-
-### TER
-|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
-|---|---|---|---|---|---|---|---|---|
-|org/dev|507|36002|95.4|2.0|2.6|0.8|5.5|69.0|
-
-
-
-## ASR config
-
-<details><summary>expand</summary>
-
-```
 config: conf/tuning/train_asr_e_branchformer_size256_mlp1024_e12_mactrue.yaml
 print_config: false
 log_level: INFO
@@ -784,39 +707,3 @@ required:
 - token_list
 version: '202301'
 distributed: true
-```
-
-</details>
-
-
-
-### 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}
-}
-```

asr_train_asr_e_branchformer_size256_mlp1024_e12_mactrue_raw_en_bpe500_sp/valid.acc.ave_10best.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82fbcf415b6d04085f3a40cfc2cd5ed2434f0a77aea8b782be77287424f9008c
+size 140364237

lm_train_lm_en_bpe500/config.yaml

@@ -0,0 +1,627 @@
+config: conf/train_lm.yaml
+print_config: false
+log_level: INFO
+dry_run: false
+iterator_type: sequence
+output_dir: exp/lm_train_lm_en_bpe500
+ngpu: 1
+seed: 0
+num_workers: 1
+num_att_plot: 3
+dist_backend: nccl
+dist_init_method: env://
+dist_world_size: 4
+dist_rank: 0
+local_rank: 0
+dist_master_addr: localhost
+dist_master_port: 60077
+dist_launcher: null
+multiprocessing_distributed: true
+unused_parameters: false
+sharded_ddp: false
+cudnn_enabled: true
+cudnn_benchmark: false
+cudnn_deterministic: true
+collect_stats: false
+write_collected_feats: false
+max_epoch: 50
+patience: null
+val_scheduler_criterion:
+- valid
+- loss
+early_stopping_criterion:
+- valid
+- loss
+- min
+best_model_criterion:
+-   - valid
+    - loss
+    - min
+keep_nbest_models: 10
+nbest_averaging_interval: 0
+grad_clip: 5.0
+grad_clip_type: 2.0
+grad_noise: false
+accum_grad: 1
+no_forward_run: false
+resume: true
+train_dtype: float32
+use_amp: false
+log_interval: null
+use_matplotlib: true
+use_tensorboard: true
+create_graph_in_tensorboard: false
+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: 20000000
+valid_batch_bins: null
+train_shape_file:
+- exp/lm_stats_en_bpe500/train/text_shape.bpe
+valid_shape_file:
+- exp/lm_stats_en_bpe500/valid/text_shape.bpe
+batch_type: numel
+valid_batch_type: null
+fold_length:
+- 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/raw/lm_train.txt
+    - text
+    - text
+valid_data_path_and_name_and_type:
+-   - dump/raw/dev/text
+    - text
+    - text
+allow_variable_data_keys: false
+max_cache_size: 0.0
+max_cache_fd: 32
+valid_max_cache_size: null
+exclude_weight_decay: false
+exclude_weight_decay_conf: {}
+optim: adam
+optim_conf:
+    lr: 0.001
+scheduler: warmuplr
+scheduler_conf:
+    warmup_steps: 25000
+token_list:
+- <blank>
+- <unk>
+- '[unk]'
+- ▁
+- s
+- ▁the
+- t
+- ▁and
+- e
+- ▁a
+- ▁to
+- d
+- ▁of
+- ''''
+- n
+- ing
+- ▁in
+- ▁that
+- re
+- ▁i
+- c
+- o
+- u
+- ▁we
+- y
+- a
+- ed
+- ▁it
+- ▁you
+- i
+- m
+- ▁is
+- er
+- p
+- g
+- w
+- al
+- ▁this
+- ▁so
+- f
+- le
+- b
+- ar
+- ▁f
+- k
+- ▁c
+- r
+- in
+- or
+- ▁for
+- ▁be
+- ve
+- ▁was
+- te
+- th
+- ▁do
+- es
+- ly
+- ▁they
+- ro
+- ▁are
+- ▁with
+- ▁have
+- an
+- v
+- ch
+- ▁on
+- se
+- lo
+- ▁but
+- en
+- ri
+- li
+- ▁what
+- it
+- ic
+- ▁can
+- l
+- ur
+- ce
+- ent
+- ▁me
+- ▁b
+- ▁ma
+- ▁he
+- ra
+- ▁de
+- ll
+- at
+- ▁about
+- ▁one
+- ▁not
+- ne
+- ▁all
+- ▁my
+- ter
+- el
+- il
+- ▁there
+- 'on'
+- ad
+- ▁mo
+- ol
+- ation
+- nd
+- ▁like
+- ▁people
+- po
+- ▁at
+- ▁us
+- us
+- ▁g
+- ci
+- ▁our
+- h
+- pe
+- ▁as
+- ▁from
+- vi
+- ▁if
+- as
+- ▁ex
+- ▁con
+- ▁an
+- ver
+- ▁out
+- ▁just
+- un
+- ▁see
+- la
+- ▁di
+- ▁when
+- ▁now
+- ▁p
+- ha
+- ▁who
+- ck
+- ▁these
+- ▁because
+- ▁or
+- ▁know
+- ion
+- ir
+- ▁co
+- ▁up
+- ▁pa
+- ment
+- ▁think
+- ge
+- ▁how
+- ide
+- ▁by
+- ul
+- ity
+- ▁go
+- ▁get
+- ▁ho
+- ive
+- ▁very
+- ate
+- ng
+- ▁no
+- ▁had
+- ac
+- ▁bo
+- ry
+- ▁more
+- ▁them
+- ▁some
+- mi
+- ▁time
+- ▁your
+- me
+- ▁going
+- op
+- am
+- per
+- et
+- ▁would
+- ru
+- ure
+- ti
+- ist
+- ▁their
+- x
+- ▁were
+- ▁look
+- ▁pro
+- ▁which
+- ▁work
+- tion
+- est
+- ty
+- im
+- z
+- ta
+- ▁want
+- ▁two
+- age
+- ▁really
+- om
+- ma
+- ers
+- ting
+- ▁world
+- co
+- ▁way
+- ▁don
+- wa
+- hi
+- tra
+- ▁la
+- ▁here
+- able
+- lu
+- ▁other
+- mo
+- ies
+- ▁has
+- ▁could
+- j
+- ▁make
+- ally
+- ▁sta
+- ten
+- ▁will
+- ▁un
+- ig
+- ▁where
+- ▁into
+- ke
+- ▁than
+- ▁comp
+- ▁actually
+- tic
+- sh
+- ▁did
+- tor
+- fa
+- ical
+- ▁she
+- ▁years
+- ▁say
+- one
+- ted
+- ▁things
+- ph
+- ▁new
+- ▁pre
+- ▁any
+- ▁thousand
+- ▁been
+- ▁inter
+- ▁his
+- ▁com
+- ▁need
+- nce
+- ▁right
+- ▁take
+- ▁even
+- ▁over
+- ▁start
+- ▁hundred
+- min
+- ▁sp
+- ▁those
+- ▁car
+- ▁then
+- mp
+- ap
+- ▁first
+- les
+- ize
+- ▁every
+- ba
+- ▁something
+- ▁well
+- ard
+- ▁str
+- ▁back
+- und
+- ia
+- pl
+- ki
+- ho
+- ▁call
+- ▁most
+- ▁also
+- bi
+- ▁thing
+- ▁life
+- um
+- ▁said
+- ▁kind
+- ▁lot
+- ▁much
+- va
+- ▁ra
+- ▁little
+- ▁dr
+- ▁got
+- ▁come
+- ful
+- ▁talk
+- ▁part
+- ▁day
+- ant
+- ction
+- ▁happen
+- ▁only
+- ▁many
+- ▁wo
+- pri
+- ▁her
+- ▁br
+- qui
+- ▁mean
+- ▁three
+- iv
+- ▁different
+- ugh
+- ain
+- ▁human
+- ance
+- ▁change
+- ▁let
+- ▁real
+- ▁show
+- ▁good
+- ▁around
+- ▁through
+- ▁jo
+- bu
+- ▁down
+- ight
+- ga
+- ▁why
+- ▁live
+- ff
+- ▁tell
+- ▁put
+- ▁idea
+- port
+- ▁same
+- ▁give
+- ated
+- ish
+- ible
+- ▁though
+- ious
+- ▁problem
+- ▁five
+- par
+- ▁fact
+- ▁cha
+- ition
+- ▁year
+- ▁big
+- ▁plan
+- ▁great
+- ▁find
+- ▁four
+- ▁app
+- ▁after
+- ▁system
+- ▁place
+- ▁em
+- ▁build
+- ▁percent
+- ▁again
+- ▁point
+- ▁learn
+- ▁own
+- ▁long
+- ▁made
+- ▁today
+- ▁nine
+- ities
+- ▁gene
+- ▁six
+- ▁question
+- light
+- ▁should
+- ▁came
+- ▁feel
+- ▁turn
+- ▁person
+- ▁end
+- ▁hu
+- ▁design
+- ▁help
+- ▁brain
+- ▁last
+- ▁create
+- ▁important
+- ▁before
+- ▁high
+- ▁never
+- ▁trans
+- ▁another
+- ▁him
+- ▁eight
+- ▁might
+- ▁understand
+- ▁power
+- ▁better
+- q
+- ▁found
+- ▁play
+- ▁twenty
+- ▁still
+- ▁school
+- ▁each
+- ▁seven
+- ▁together
+- ▁few
+- ▁hand
+- ▁example
+- que
+- ▁next
+- ▁million
+- ▁story
+- ▁women
+- ▁under
+- ▁number
+- ▁course
+- ▁water
+- ▁ago
+- ▁grow
+- ▁between
+- ▁develop
+- ▁america
+- ▁sort
+- ▁technology
+- ▁believe
+- ▁second
+- ▁small
+- ▁maybe
+- ▁become
+- press
+- ▁health
+- ▁space
+- ▁word
+- ▁hard
+- ▁children
+- ▁organ
+- ▁always
+- ▁country
+- ▁reason
+- ▁experience
+- ▁large
+- ▁everything
+- ▁friend
+- ▁project
+- ▁computer
+- ▁fifty
+- ▁money
+- ▁information
+- graph
+- ▁walk
+- ization
+- ▁africa
+- ▁picture
+- ▁process
+- ▁teach
+- ▁enough
+- ▁elect
+- ▁thirty
+- '0'
+- '1'
+- '2'
+- '9'
+- '3'
+- '5'
+- '8'
+- '4'
+- '7'
+- '6'
+- '&'
+- +
+- '#'
+- '@'
+- '*'
+- \
+- ^
+- R
+- _
+- '-'
+- '%'
+- '='
+- $
+- M
+- ā
+- ']'
+- E
+- U
+- A
+- G
+- '['
+- <sos/eos>
+init: null
+model_conf:
+    ignore_id: 0
+use_preprocessor: true
+token_type: bpe
+bpemodel: data/en_token_list/bpe_unigram500/bpe.model
+non_linguistic_symbols: null
+cleaner: null
+g2p: null
+lm: transformer
+lm_conf:
+    pos_enc: null
+    embed_unit: 128
+    att_unit: 512
+    head: 8
+    unit: 2048
+    layer: 15
+    dropout_rate: 0.1
+required:
+- output_dir
+- token_list
+version: '202301'
+distributed: true

lm_train_lm_en_bpe500/perplexity_test/ppl

@@ -0,0 +1 @@
+10.155182315138273

lm_train_lm_en_bpe500/valid.loss.ave_10best.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec899e08f7764595b4bf1d45cd88d21156c92c7cb9885071e53d118232970df0
+size 190793638