Merge branch 'espnet:master' into st_bugfix

chintu619 · May 6, 2022 · 6d1bd3a · 6d1bd3a
2 parents eb6dc2d + 793b999
commit 6d1bd3a
Show file tree

Hide file tree

Showing 40 changed files with 429 additions and 17 deletions.
diff --git a/README.md b/README.md
@@ -97,7 +97,7 @@ Demonstration
     - Conformer FastSpeech & FastSpeech2
     - VITS
 - Multi-speaker & multi-language extention
-    - Pretrined speaker embedding (e.g., X-vector)
+    - Pretrained speaker embedding (e.g., X-vector)
     - Speaker ID embedding
     - Language ID embedding
     - Global style token (GST) embedding

diff --git a/egs2/README.md b/egs2/README.md
@@ -19,6 +19,7 @@ See: https://espnet.github.io/espnet/espnet2_tutorial.html#recipes-using-espnet2
 | bur_openslr80           | Burmese ASR training dataset                                                                                                     | ASR                     | BUR                   | https://openslr.org/80/                                                                                      |              |
 | catslu               	  | CATSLU-MAPS                                                                                                                      | SLU                     | CMN           	      | https://sites.google.com/view/catslu/home                                                                     |              |
 | chime4                  | The 4th CHiME Speech Separation and Recognition Challenge                                                                        | ASR/Multichannel ASR    | ENG                   | http://spandh.dcs.shef.ac.uk/chime_challenge/chime2016/                                                      |              |
+| chime6                  | The 6th CHiME Speech Separation and Recognition Challenge    | ASR                                        | ENG             | https://chimechallenge.github.io/chime6/                |                               |
 | clarity21               | The First Clarity Enhancement Challenge CEC1                                                                                     | SE                      | ENG                   | https://claritychallenge.github.io/clarity_CEC1_doc/                                                         |              |
 | cmu_indic               | CMU INDIC                                                                                                                        | TTS                     | 7 languages           | http://festvox.org/cmu_indic/                                                                                |              |
 | commonvoice             | The Mozilla Common Voice                                                                                                         | ASR                     | 13 languages          | https://voice.mozilla.org/datasets                                                                           |              |

diff --git a/egs2/TEMPLATE/asr1/db.sh b/egs2/TEMPLATE/asr1/db.sh
@@ -23,6 +23,7 @@ REVERB=
 REVERB_OUT="${PWD}/REVERB"  # Output file path
 CHIME3=
 CHIME4=
+CHIME5=
 CSJDATATOP=
 CSJVER=dvd  ## Set your CSJ format (dvd or usb).
             ## Usage    :

diff --git a/egs2/TEMPLATE/asr1/scripts/utils/show_translation_result.sh b/egs2/TEMPLATE/asr1/scripts/utils/show_translation_result.sh
@@ -1,6 +1,6 @@
 #!/usr/bin/env bash
 mindepth=0
-maxdepth=3
+maxdepth=1
 case=tc
 
 . utils/parse_options.sh
@@ -44,24 +44,27 @@ cat << EOF
 
 EOF
 
+# only show BLEU score for now
 metrics="bleu"
-
 while IFS= read -r expdir; do
     if ls "${expdir}"/*/*/score_*/result.${case}.txt &> /dev/null; then
         echo "## $(basename ${expdir})"
-        for type in $metrics; do
-                	cat << EOF
+        for type in ${metrics}; do
+            cat << EOF
+
 ### ${type^^}
 
-|dataset|bleu_score|verbose_score|
+|dataset|score|verbose_score|
 |---|---|---|
 EOF
-    data=$(echo "${expdir}"/*/*/score_*/result.${case}.txt | cut -d '/' -f4)
-    bleu=$(sed -n '5p' "${expdir}"/*/*/score_*/result.${case}.txt | cut -d ' ' -f 3 | tr -d ',')
-    verbose=$(sed -n '7p' "${expdir}"/*/*/score_*/result.${case}.txt | cut -d ' ' -f 3- | tr -d '",')
-    echo "${data}|${bleu}|${verbose}"
 
+            for result in "${expdir}"/*/*/score_"${type}"/result."${case}".txt; do
+                inference_tag=$(echo "${result}" | rev | cut -d/ -f4 | rev)
+                test_set=$(echo "${result}" | rev | cut -d/ -f3 | rev)
+                score=$(sed -n '5p' "${result}" | cut -d ' ' -f 3 | tr -d ',')
+                verbose=$(sed -n '7p' "${result}" | cut -d ' ' -f 3- | tr -d '",')
+                echo "|${inference_tag}/${test_set}|${score}|${verbose}|"
+            done
         done
     fi
-
 done < <(find ${exp} -mindepth ${mindepth} -maxdepth ${maxdepth} -type d)
diff --git a/egs2/TEMPLATE/mt1/mt.sh b/egs2/TEMPLATE/mt1/mt.sh
@@ -1215,7 +1215,7 @@ if ! "${skip_eval}"; then
             detokenizer.perl -l ${tgt_lang} -q < "${_scoredir}/hyp.trn" > "${_scoredir}/hyp.trn.detok"
 
             if [ ${tgt_case} = "tc" ]; then
-                echo "Case sensitive BLEU result (single-reference)" >> ${_scoredir}/result.tc.txt
+                echo "Case sensitive BLEU result (single-reference)" > ${_scoredir}/result.tc.txt
                 sacrebleu "${_scoredir}/ref.trn.detok" \
                           -i "${_scoredir}/hyp.trn.detok" \
                           -m bleu chrf ter \
@@ -1227,7 +1227,7 @@ if ! "${skip_eval}"; then
             # detokenize & remove punctuation except apostrophe
             remove_punctuation.pl < "${_scoredir}/ref.trn.detok" > "${_scoredir}/ref.trn.detok.lc.rm"
             remove_punctuation.pl < "${_scoredir}/hyp.trn.detok" > "${_scoredir}/hyp.trn.detok.lc.rm"
-            echo "Case insensitive BLEU result (single-reference)" >> ${_scoredir}/result.lc.txt
+            echo "Case insensitive BLEU result (single-reference)" > ${_scoredir}/result.lc.txt
             sacrebleu -lc "${_scoredir}/ref.trn.detok.lc.rm" \
                       -i "${_scoredir}/hyp.trn.detok.lc.rm" \
                       -m bleu chrf ter \
@@ -1279,7 +1279,7 @@ if ! "${skip_eval}"; then
 
         # Show results in Markdown syntax
         scripts/utils/show_translation_result.sh --case $tgt_case "${mt_exp}" > "${mt_exp}"/RESULTS.md
-        cat "${cat_exp}"/RESULTS.md
+        cat "${mt_exp}"/RESULTS.md
     fi
 else
     log "Skip the evaluation stages"

diff --git a/egs2/chime6/asr1/README.md b/egs2/chime6/asr1/README.md
@@ -0,0 +1,30 @@
+<!-- Generated by scripts/utils/show_asr_result.sh -->
+# RESULTS
+## Environments
+- date: `Tue May  3 16:47:10 EDT 2022`
+- python version: `3.9.12 (main, Apr  5 2022, 06:56:58)  [GCC 7.5.0]`
+- espnet version: `espnet 202204`
+- pytorch version: `pytorch 1.10.1`
+- Git hash: `b757b89d45d5574cebf44e225cbe32e3e9e4f522`
+  - Commit date: `Mon May 2 09:21:08 2022 -0400`
+
+## asr_train_asr_transformer_wavlm_lr1e-3_specaug_accum1_preenc128_warmup20k_raw_en_bpe1000_sp
+- Pretrained model: https://huggingface.co/espnet/simpleoier_chime6_asr_transformer_wavlm_lr1e-3
+### WER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|decode_asr_transformer_asr_model_valid.acc.ave_5best/dev_gss_multiarray|7437|58881|69.4|20.2|10.4|8.6|39.1|75.8|
+
+### CER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|decode_asr_transformer_asr_model_valid.acc.ave_5best/dev_gss_multiarray|7437|280767|80.6|7.4|12.0|8.9|28.3|76.6|
+
+### TER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|decode_asr_transformer_asr_model_valid.acc.ave_5best/dev_gss_multiarray|7437|92680|68.9|17.7|13.4|8.2|39.3|76.6|
+
diff --git a/egs2/chime6/asr1/asr.sh b/egs2/chime6/asr1/asr.sh
@@ -0,0 +1 @@
+../../TEMPLATE/asr1/asr.sh
diff --git a/egs2/chime6/asr1/cmd.sh b/egs2/chime6/asr1/cmd.sh
@@ -0,0 +1,110 @@
+# ====== About run.pl, queue.pl, slurm.pl, and ssh.pl ======
+# Usage: <cmd>.pl [options] JOB=1:<nj> <log> <command...>
+# e.g.
+#   run.pl --mem 4G JOB=1:10 echo.JOB.log echo JOB
+#
+# Options:
+#   --time <time>: Limit the maximum time to execute.
+#   --mem <mem>: Limit the maximum memory usage.
+#   -–max-jobs-run <njob>: Limit the number parallel jobs. This is ignored for non-array jobs.
+#   --num-threads <ngpu>: Specify the number of CPU core.
+#   --gpu <ngpu>: Specify the number of GPU devices.
+#   --config: Change the configuration file from default.
+#
+# "JOB=1:10" is used for "array jobs" and it can control the number of parallel jobs.
+# The left string of "=", i.e. "JOB", is replaced by <N>(Nth job) in the command and the log file name,
+# e.g. "echo JOB" is changed to "echo 3" for the 3rd job and "echo 8" for 8th job respectively.
+# Note that the number must start with a positive number, so you can't use "JOB=0:10" for example.
+#
+# run.pl, queue.pl, slurm.pl, and ssh.pl have unified interface, not depending on its backend.
+# These options are mapping to specific options for each backend and
+# it is configured by "conf/queue.conf" and "conf/slurm.conf" by default.
+# If jobs failed, your configuration might be wrong for your environment.
+#
+#
+# The official documentation for run.pl, queue.pl, slurm.pl, and ssh.pl:
+#   "Parallelization in Kaldi": http://kaldi-asr.org/doc/queue.html
+# =========================================================~
+
+
+# Select the backend used by run.sh from "local", "stdout", "sge", "slurm", or "ssh"
+cmd_backend='local'
+
+# Local machine, without any Job scheduling system
+if [ "${cmd_backend}" = local ]; then
+
+    # The other usage
+    export train_cmd="run.pl"
+    # Used for "*_train.py": "--gpu" is appended optionally by run.sh
+    export cuda_cmd="run.pl"
+    # Used for "*_recog.py"
+    export decode_cmd="run.pl"
+
+# Local machine logging to stdout and log file, without any Job scheduling system
+elif [ "${cmd_backend}" = stdout ]; then
+
+    # The other usage
+    export train_cmd="stdout.pl"
+    # Used for "*_train.py": "--gpu" is appended optionally by run.sh
+    export cuda_cmd="stdout.pl"
+    # Used for "*_recog.py"
+    export decode_cmd="stdout.pl"
+
+
+# "qsub" (Sun Grid Engine, or derivation of it)
+elif [ "${cmd_backend}" = sge ]; then
+    # The default setting is written in conf/queue.conf.
+    # You must change "-q g.q" for the "queue" for your environment.
+    # To know the "queue" names, type "qhost -q"
+    # Note that to use "--gpu *", you have to setup "complex_value" for the system scheduler.
+
+    export train_cmd="queue.pl"
+    export cuda_cmd="queue.pl"
+    export decode_cmd="queue.pl"
+
+
+# "qsub" (Torque/PBS.)
+elif [ "${cmd_backend}" = pbs ]; then
+    # The default setting is written in conf/pbs.conf.
+
+    export train_cmd="pbs.pl"
+    export cuda_cmd="pbs.pl"
+    export decode_cmd="pbs.pl"
+
+
+# "sbatch" (Slurm)
+elif [ "${cmd_backend}" = slurm ]; then
+    # The default setting is written in conf/slurm.conf.
+    # You must change "-p cpu" and "-p gpu" for the "partition" for your environment.
+    # To know the "partion" names, type "sinfo".
+    # You can use "--gpu * " by default for slurm and it is interpreted as "--gres gpu:*"
+    # The devices are allocated exclusively using "${CUDA_VISIBLE_DEVICES}".
+
+    export train_cmd="slurm.pl"
+    export cuda_cmd="slurm.pl"
+    export decode_cmd="slurm.pl"
+
+elif [ "${cmd_backend}" = ssh ]; then
+    # You have to create ".queue/machines" to specify the host to execute jobs.
+    # e.g. .queue/machines
+    #   host1
+    #   host2
+    #   host3
+    # Assuming you can login them without any password, i.e. You have to set ssh keys.
+
+    export train_cmd="ssh.pl"
+    export cuda_cmd="ssh.pl"
+    export decode_cmd="ssh.pl"
+
+# This is an example of specifying several unique options in the JHU CLSP cluster setup.
+# Users can modify/add their own command options according to their cluster environments.
+elif [ "${cmd_backend}" = jhu ]; then
+
+    export train_cmd="queue.pl --mem 2G"
+    export cuda_cmd="queue-freegpu.pl --mem 2G --gpu 1 --config conf/queue.conf"
+    export decode_cmd="queue.pl --mem 4G"
+
+else
+    echo "$0: Error: Unknown cmd_backend=${cmd_backend}" 1>&2
+    return 1
+fi
diff --git a/egs2/chime6/asr1/conf/decode_asr_transformer.yaml b/egs2/chime6/asr1/conf/decode_asr_transformer.yaml
@@ -0,0 +1,7 @@
+batch_size: 0
+beam_size: 10
+penalty: 0.0
+maxlenratio: 0.0
+minlenratio: 0.0
+ctc_weight: 0.3
+lm-weight: 0.0
diff --git a/egs2/chime6/asr1/conf/fbank.conf b/egs2/chime6/asr1/conf/fbank.conf
@@ -0,0 +1,2 @@
+--sample-frequency=16000 
+--num-mel-bins=80
diff --git a/egs2/chime6/asr1/conf/pbs.conf b/egs2/chime6/asr1/conf/pbs.conf
@@ -0,0 +1,11 @@
+# Default configuration
+command qsub -V -v PATH -S /bin/bash
+option name=* -N $0
+option mem=* -l mem=$0
+option mem=0          # Do not add anything to qsub_opts
+option num_threads=* -l ncpus=$0
+option num_threads=1  # Do not add anything to qsub_opts
+option num_nodes=* -l nodes=$0:ppn=1
+default gpu=0
+option gpu=0
+option gpu=* -l ngpus=$0
diff --git a/egs2/chime6/asr1/conf/pitch.conf b/egs2/chime6/asr1/conf/pitch.conf
@@ -0,0 +1 @@
+--sample-frequency=16000
diff --git a/egs2/chime6/asr1/conf/queue.conf b/egs2/chime6/asr1/conf/queue.conf
@@ -0,0 +1,12 @@
+# Default configuration
+command qsub -v PATH -cwd -S /bin/bash -j y -l arch=*64*
+option name=* -N $0
+option mem=* -l mem_free=$0,ram_free=$0
+option mem=0          # Do not add anything to qsub_opts
+option num_threads=* -pe smp $0
+option num_threads=1  # Do not add anything to qsub_opts
+option max_jobs_run=* -tc $0
+option num_nodes=* -pe mpi $0  # You must set this PE as allocation_rule=1
+default gpu=0
+option gpu=0
+option gpu=* -l gpu=$0 -q g.q
diff --git a/egs2/chime6/asr1/conf/slurm.conf b/egs2/chime6/asr1/conf/slurm.conf
@@ -0,0 +1,14 @@
+# Default configuration
+command sbatch --export=PATH
+option name=* --job-name $0
+option time=* --time $0
+option mem=* --mem-per-cpu $0
+option mem=0
+option num_threads=* --cpus-per-task $0
+option num_threads=1 --cpus-per-task 1
+option num_nodes=* --nodes $0
+default gpu=0
+option gpu=0 -p cpu
+option gpu=* -p gpu --gres=gpu:$0 -c $0  # Recommend allocating more CPU than, or equal to the number of GPU
+# note: the --max-jobs-run option is supported as a special case
+# by slurm.pl and you don't have to handle it in the config file.
diff --git a/egs2/chime6/asr1/conf/train_lm.yaml b/egs2/chime6/asr1/conf/train_lm.yaml
@@ -0,0 +1,16 @@
+optim: sgd
+patience: 3
+max_epoch: 20
+batch_type: folded
+batch_size: 1024 # 300 for word LMs
+lm: seq_rnn
+lm_conf:
+    rnn_type: lstm
+    nlayers: 2  # 1 for word LMs
+    unit: 650   # 1000 for word LMs
+
+best_model_criterion:
+-   - valid
+    - loss
+    - min
+keep_nbest_models: 1
diff --git a/...r1/conf/tuning/train_asr_transformer_wavlm_lr1e-3_specaug_accum1_preenc128_warmup20k.yaml b/...r1/conf/tuning/train_asr_transformer_wavlm_lr1e-3_specaug_accum1_preenc128_warmup20k.yaml
@@ -0,0 +1,87 @@
+# minibatch related
+batch_type: folded
+batch_size: 48
+accum_grad: 1
+grad_clip: 5
+max_epoch: 8
+patience: 4
+# The initialization method for model parameters
+init: xavier_uniform
+val_scheduler_criterion:
+- valid
+- loss
+best_model_criterion:
+-   - valid
+    - acc
+    - max
+keep_nbest_models: 10
+unused_parameters: true
+freeze_param: [
+    "frontend.upstream"
+]
+
+# network architecture
+frontend: s3prl
+frontend_conf:
+    frontend_conf:
+        upstream: wavlm_large  # Note: If the upstream is changed, please change the input_size in the preencoder.
+    download_dir: ./hub
+    multilayer_feature: True
+
+preencoder: linear
+preencoder_conf:
+    input_size: 1024  # Note: If the upstream is changed, please change this value accordingly.
+    output_size: 128
+
+# encoder related
+encoder: transformer
+encoder_conf:
+    output_size: 256
+    attention_heads: 4
+    linear_units: 2048
+    num_blocks: 12
+    dropout_rate: 0.1
+    attention_dropout_rate: 0.0
+    input_layer: conv2d2
+    normalize_before: true
+
+# decoder related
+decoder: transformer
+decoder_conf:
+    input_layer: embed
+    attention_heads: 4
+    linear_units: 2048
+    num_blocks: 6
+    dropout_rate: 0.1
+    positional_dropout_rate: 0.0
+    self_attention_dropout_rate: 0.0
+    src_attention_dropout_rate: 0.0
+
+model_conf:
+    ctc_weight: 0.3
+    lsm_weight: 0.1
+    length_normalized_loss: false
+    extract_feats_in_collect_stats: false
+
+optim: adam
+optim_conf:
+    lr: 0.001
+scheduler: warmuplr
+scheduler_conf:
+    warmup_steps: 20000
+
+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
+    - 100
+    num_freq_mask: 4
+    apply_time_mask: true
+    time_mask_width_range:
+    - 0
+    - 40
+    num_time_mask: 2