Merge pull request espnet#4243 from D-Keqi/master

Add streaming ST/SLU

ci/test_integration_espnet2.sh

@@ -110,6 +110,40 @@ if python -c 'import torch as t; from distutils.version import LooseVersion as L
     cd "${cwd}"
 fi
 
+# [ESPnet2] test st recipe
+cd ./egs2/mini_an4/st1
+echo "==== [ESPnet2] ST ==="
+./run.sh --stage 1 --stop-stage 1
+feats_types="raw fbank_pitch"
+token_types="bpe char"
+for t in ${feats_types}; do
+    ./run.sh --stage 2 --stop-stage 4 --feats-type "${t}" --python "${python}"
+done
+for t in ${token_types}; do
+    ./run.sh --stage 5 --stop-stage 5 --tgt_token_type "${t}" --src_token_type "${t}" --python "${python}"
+done
+for t in ${feats_types}; do 
+    for t2 in ${token_types}; do
+        echo "==== feats_type=${t}, token_types=${t2} ==="
+        ./run.sh --ngpu 0 --stage 6 --stop-stage 13 --skip-upload false --feats-type "${t}" --tgt_token_type "${t2}" --src_token_type "${t2}" \
+            --st-args "--max_epoch=1" --lm-args "--max_epoch=1" --inference_args "--beam_size 5" --python "${python}"
+    done
+done
+echo "==== feats_type=raw, token_types=bpe, model_conf.extract_feats_in_collect_stats=False, normalize=utt_mvn ==="
+./run.sh --ngpu 0 --stage 10 --stop-stage 13 --skip-upload false --feats-type "raw" --tgt_token_type "bpe" --src_token_type "bpe" \
+    --feats_normalize "utterance_mvn" --lm-args "--max_epoch=1" --inference_args "--beam_size 5" --python "${python}" \
+    --st-args "--model_conf extract_feats_in_collect_stats=false --max_epoch=1"
+
+echo "==== use_streaming, feats_type=raw, token_types=bpe, model_conf.extract_feats_in_collect_stats=False, normalize=utt_mvn ==="
+./run.sh --use_streaming true --ngpu 0 --stage 6 --stop-stage 13 --skip-upload false --feats-type "raw" --tgt_token_type "bpe" --src_token_type "bpe" \
+    --feats_normalize "utterance_mvn" --lm-args "--max_epoch=1" --inference_args "--beam_size 5" --python "${python}" \
+    --st-args "--model_conf extract_feats_in_collect_stats=false --max_epoch=1 --encoder=contextual_block_transformer --decoder=transformer
+                --encoder_conf block_size=40 --encoder_conf hop_size=16 --encoder_conf look_ahead=16"
+
+# Remove generated files in order to reduce the disk usage
+rm -rf exp dump data
+cd "${cwd}"
+
 # [ESPnet2] Validate configuration files
 echo "<blank>" > dummy_token_list
 echo "==== [ESPnet2] Validation configuration files ==="

egs2/TEMPLATE/st1/st.sh

@@ -111,6 +111,7 @@ hf_repo=
 
 # Decoding related
 use_k2=false      # Whether to use k2 based decoder
+use_streaming=false # Whether to use streaming decoding
 batch_size=1
 inference_tag=    # Suffix to the result dir for decoding.
 inference_config= # Config for decoding.
@@ -1412,7 +1413,11 @@ if ! "${skip_eval}"; then
             key_file=${_data}/${_scp}
             split_scps=""
             _nj=$(min "${inference_nj}" "$(<${key_file} wc -l)")
-            st_inference_tool="espnet2.bin.st_inference"
+            if "${use_streaming}"; then
+                st_inference_tool="espnet2.bin.st_inference_streaming"
+            else
+                st_inference_tool="espnet2.bin.st_inference"
+            fi
 
             for n in $(seq "${_nj}"); do
                 split_scps+=" ${_logdir}/keys.${n}.scp"
@@ -1552,7 +1557,7 @@ if ! "${skip_eval}"; then
 
         # Show results in Markdown syntax
         scripts/utils/show_translation_result.sh --case $tgt_case "${st_exp}" > "${st_exp}"/RESULTS.md
-        cat "${cat_exp}"/RESULTS.md
+        cat "${st_exp}"/RESULTS.md
     fi
 else
     log "Skip the evaluation stages"

egs2/fisher_callhome_spanish/st1/conf/decode_streaming_st.yaml

@@ -0,0 +1,5 @@
+batch_size: 1
+beam_size: 10
+nbest: 1
+lm_weight: 0.0
+sim_chunk_length: 5120

egs2/fisher_callhome_spanish/st1/conf/train_st_streaming.yaml

@@ -0,0 +1,95 @@
+# network architecture
+
+# frontend related
+frontend: default
+frontend_conf:
+    n_fft: 512
+    win_length: 400
+    hop_length: 160
+
+# encoder related
+# encoder related
+encoder: contextual_block_transformer
+encoder_conf:
+    output_size: 256
+    attention_heads: 4
+    linear_units: 2048
+    num_blocks: 12
+    dropout_rate: 0.1
+    positional_dropout_rate: 0.1
+    attention_dropout_rate: 0.1
+    input_layer: conv2d
+    normalize_before: true
+    block_size: 40
+    hop_size: 16
+    look_ahead: 16
+    init_average: true
+    ctx_pos_enc: true
+
+# decoder related
+decoder: transformer
+decoder_conf:
+    input_layer: embed
+    num_blocks: 6
+    linear_units: 2048
+    dropout_rate: 0.1
+
+extra_asr_decoder: transformer
+extra_asr_decoder_conf:
+    input_layer: embed
+    num_blocks: 2
+    linear_units: 2048
+    dropout_rate: 0.1
+
+extra_mt_decoder: transformer
+extra_mt_decoder_conf:
+    input_layer: embed
+    num_blocks: 2
+    linear_units: 2048
+    dropout_rate: 0.1
+
+# loss related
+model_conf:
+    asr_weight: 0.3
+    mt_weight: 0.0
+    mtlalpha: 1.0
+    lsm_weight: 0.1
+    length_normalized_loss: false
+
+# optimization related
+optim: adam
+accum_grad: 1
+grad_clip: 3
+max_epoch: 50
+optim_conf:
+    lr: 2.5
+scheduler: noamlr
+scheduler_conf:
+    model_size: 256
+    warmup_steps: 25000
+
+# minibatch related
+batch_type: folded
+batch_size: 128
+
+best_model_criterion:
+-   - valid
+    - acc
+    - max
+keep_nbest_models: 10
+num_att_plot: 0
+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
+    - 30
+    num_freq_mask: 2
+    apply_time_mask: true
+    time_mask_width_range:
+    - 0
+    - 40
+    num_time_mask: 2

egs2/fisher_callhome_spanish/st1/run.sh

@@ -22,6 +22,7 @@ src_case=lc.rm
 tgt_case=lc.rm
 
 ./st.sh \
+    --use_streaming false \
     --local_data_opts "--stage 0" \
     --audio_format "flac.ark" \
     --use_lm false \

egs2/fsc/asr1/conf/train_asr_streaming_transformer.yaml

@@ -0,0 +1,58 @@
+# network architecture
+# encoder related
+encoder: contextual_block_transformer
+encoder_conf:
+    output_size: 256    # dimension of attention
+    attention_heads: 4
+    linear_units: 2048  # the number of units of position-wise feed forward
+    num_blocks: 12      # the number of encoder blocks
+    dropout_rate: 0.1
+    positional_dropout_rate: 0.1
+    attention_dropout_rate: 0.0
+    input_layer: conv2d # encoder architecture type
+    normalize_before: true
+    block_size: 40
+    hop_size: 16
+    look_ahead: 16
+    init_average: true
+    ctx_pos_enc: true
+
+# decoder related
+decoder: transformer
+decoder_conf:
+    attention_heads: 4
+    linear_units: 2048
+    num_blocks: 6
+    dropout_rate: 0.1
+    positional_dropout_rate: 0.1
+    self_attention_dropout_rate: 0.0
+    src_attention_dropout_rate: 0.0
+
+# minibatch related
+batch_type: folded
+batch_size: 64
+max_epoch: 200
+keep_nbest_models: 5
+
+optim: adam
+optim_conf:
+    lr: 0.0002
+scheduler: warmuplr     # pytorch v1.1.0+ required
+scheduler_conf:
+    warmup_steps: 25000
+num_att_plot: 0
+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
+    - 30
+    num_freq_mask: 2
+    apply_time_mask: true
+    time_mask_width_range:
+    - 0
+    - 40
+    num_time_mask: 2

egs2/mini_an4/st1/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

egs2/mini_an4/st1/conf/fbank.conf

@@ -0,0 +1,2 @@
+--sample-frequency=16000 
+--num-mel-bins=80

egs2/mini_an4/st1/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

egs2/mini_an4/st1/conf/pitch.conf

@@ -0,0 +1 @@
+--sample-frequency=16000

egs2/mini_an4/st1/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

egs2/mini_an4/st1/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.