Merge pull request #4771 from realzza/master

[Recipe PR] MELD: Multimodal EmotionLines Dataset

egs2/README.md

@@ -84,6 +84,7 @@ See: https://espnet.github.io/espnet/espnet2_tutorial.html#recipes-using-espnet2
 | lt_slurp_spatialized    | Spatialized Libri-Trans and Spatialized SLURP (LT-S and SLURP-S), Enhancement for Translation and Understanding Dataset          | SE/ST/SLU               | ENG                  |                                                                                                              |              |
 | magicdata               | MAGICDATA Mandarin Chinese Read Speech Corpus                                                                                    | ASR                     | ENG                  | https://www.openslr.org/68/                                                                                  |              |
 | mediaspeech             | MediaSpeech: Multilanguage ASR Benchmark and Dataset                                                                             | ASR                     | FRA                  | https://www.openslr.org/108/                                                                                 |              |
+| meld                    | MELD: Multimodal EmotionLines Dataset                                                                                            | SLU                     | ENG                  | https://affective-meld.github.io/									     |              |
 | microsoft_speech        | Microsoft Speech Corpus (Indian languages)                                                                                       | ASR                     | 3 languages          | https://msropendata.com/datasets/7230b4b1-912d-400e-be58-f84e0512985e                                        |              |
 | mini_an4                | Mini version of CMU AN4 database for the integration test                                                                        | ASR/TTS/SE              | ENG                  | http://www.speech.cs.cmu.edu/databases/an4/                                                                  |              |
 | mini_librispeech        | Mini version of Librispeech corpus                                                                                               | DIAR                    | ENG                  | https://openslr.org/31/                                                                                      |              |

egs2/TEMPLATE/asr1/db.sh

@@ -48,6 +48,7 @@ TEDXJP=
 LIBRISPEECH=downloads
 LIBRILIGHT_LIMITED=
 FSC=
+MELD=downloads
 SLURP=
 SLURP_S= # Output file path
 LIBRITRANS_S= # Output file path

egs2/meld/asr1/README.md

@@ -0,0 +1,45 @@
+# MELD RESULTS
+## Environments
+- date: `Thu Nov 10 09:07:40 EST 2022`
+- python version: `3.8.6 (default, Dec 17 2020, 16:57:01)  [GCC 10.2.0]`
+- espnet version: `espnet 202207`
+- pytorch version: `pytorch 1.8.1+cu102`
+- Git hash: `a7bd6522b32ec6472c13f6a2289dcdff4a846c12`
+  - Commit date: `Wed Sep 14 08:34:27 2022 -0400`
+- Pretrained Model
+  - Huggingface Hub: [ [model card](https://huggingface.co/espnet/realzza-meld-asr-hubert-transformer) | [model](https://huggingface.co/espnet/realzza-meld-asr-hubert-transformer/blob/main/exp/asr_train_asr_hubert_transformer_adam_specaug_meld_raw_en_bpe850/valid.acc.ave_5best.pth) ]
+
+## asr_train_asr_hubert_transformer_adam_specaug_meld_raw_en_bpe850
+- ASR config: conf/tuning/train_asr_hubert_transformer_adam_specaug_meld.yaml
+- token_type: bpe
+- keep_nbest_models: 5
+
+|dataset|Snt|Emotion Classification (%)|
+|---|---|---|
+|decoder_asr_asr_model_valid.acc.ave_5best/test|2608|39.22|
+|decoder_asr_asr_model_valid.acc.ave_5best/valid|1104|42.64|
+
+### ASR results
+
+#### WER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|decoder_asr_asr_model_valid.acc.ave_5best/test|2608|24809|55.5|28.0|16.5|8.4|52.9|96.5|
+|decoder_asr_asr_model_valid.acc.ave_5best/valid|1104|10171|55.3|29.4|15.3|7.0|51.7|96.2|
+
+#### CER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|decoder_asr_asr_model_valid.acc.ave_5best/test|2608|120780|71.1|10.7|18.2|10.6|39.5|96.5|
+|decoder_asr_asr_model_valid.acc.ave_5best/valid|1104|49323|71.3|11.1|17.6|9.4|38.1|96.2|
+
+#### TER
+
+|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
+|---|---|---|---|---|---|---|---|---|
+|decoder_asr_asr_model_valid.acc.ave_5best/test|2608|35287|57.6|21.8|20.5|7.8|50.2|96.5|
+|decoder_asr_asr_model_valid.acc.ave_5best/valid|1104|14430|57.4|23.2|19.4|6.1|48.6|96.2|
+
+

egs2/meld/asr1/asr.sh

@@ -0,0 +1 @@
+../../TEMPLATE/asr1/asr.sh

egs2/meld/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

egs2/meld/asr1/conf/decoder_asr.yaml

@@ -0,0 +1,6 @@
+beam_size: 10
+ctc_weight: 0.3
+lm_weight: 0.1
+penalty: 0.0
+maxlenratio: 0.0
+minlenratio: 0.0

egs2/meld/asr1/conf/fbank.conf

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

egs2/meld/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

egs2/meld/asr1/conf/pitch.conf

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

egs2/meld/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

egs2/meld/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.

egs2/meld/asr1/conf/train_asr.yaml

@@ -0,0 +1 @@
+tuning/train_asr_hubert_transformer_adam_specaug_meld.yaml

egs2/meld/asr1/conf/tuning/train_asr_hubert_transformer_adam_specaug_meld.yaml

@@ -0,0 +1,74 @@
+# network architecture
+# encoder related
+encoder: conformer
+encoder_conf:
+    output_size: 512    # dimension of attention
+    attention_heads: 8
+    linear_units: 2048  # the number of units of position-wise feed forward
+    num_blocks: 12      # the number of encoder blocks
+    dropout_rate: 0.2
+    positional_dropout_rate: 0.1
+    attention_dropout_rate: 0.0
+    input_layer: conv2d2 # encoder architecture type
+    normalize_before: true
+
+# decoder related
+decoder: transformer
+decoder_conf:
+    attention_heads: 8
+    linear_units: 2048
+    num_blocks: 6
+    dropout_rate: 0.2
+    positional_dropout_rate: 0.1
+    self_attention_dropout_rate: 0.0
+    src_attention_dropout_rate: 0.0
+optim: adam
+optim_conf:
+    lr: 0.00015
+scheduler: warmuplr     # pytorch v1.1.0+ required
+scheduler_conf:
+    warmup_steps: 16000
+max_epoch: 2
+keep_nbest_models: 5
+
+freeze_param: [
+"frontend.upstream"
+]
+
+frontend_conf:
+  n_fft: 512
+  hop_length: 256
+
+frontend: s3prl
+frontend_conf:
+    frontend_conf:
+        upstream: hubert_large_ll60k  # 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: 80
+
+model_conf:
+    ctc_weight: 0.3
+    lsm_weight: 0.1
+    length_normalized_loss: false
+    extract_feats_in_collect_stats: false   # Note: "False" means during collect stats (stage 10), generating dummy stats files rather than extract_feats by forward frontend.
+
+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/meld/asr1/db.sh

@@ -0,0 +1 @@
+../../TEMPLATE/asr1/db.sh

egs2/meld/asr1/local/data.sh

@@ -0,0 +1,62 @@
+#!/usr/bin/env bash
+# Set bash to 'debug' mode, it will exit on :
+# -e 'error', -u 'undefined variable', -o ... 'error in pipeline', -x 'print commands',
+set -e
+set -u
+set -o pipefail
+
+log() {
+    local fname=${BASH_SOURCE[1]##*/}
+    echo -e "$(date '+%Y-%m-%dT%H:%M:%S') (${fname}:${BASH_LINENO[0]}:${FUNCNAME[1]}) $*"
+}
+SECONDS=0
+
+
+stage=1
+stop_stage=100
+
+datadir=./downloads
+data_url=https://web.eecs.umich.edu/~mihalcea/downloads/
+data_url2=https://huggingface.co/datasets/declare-lab/MELD/resolve/main/
+
+
+log "$0 $*"
+. utils/parse_options.sh
+
+. ./db.sh
+. ./path.sh
+. ./cmd.sh
+
+if [ $# -ne 0 ]; then
+    log "Error: No positional arguments are required."
+    exit 2
+fi
+
+if [ -z "${MELD}" ]; then
+    log "Fill the value of 'MELD' of db.sh"
+    exit 1
+fi
+
+if [ ${stage} -le 1 ] && [ ${stop_stage} -ge 1 ]; then
+    log "stage 1: Data Download"
+    mkdir -p ${datadir}
+    if ! local/download_and_untar.sh --remove-archive ${datadir} ${data_url}; then
+        log "Failed to download from the original site, try a backup site."
+        local/download_and_untar.sh --remove-archive ${datadir} ${data_url2}
+    fi
+fi
+
+if [ ${stage} -le 2 ] && [ ${stop_stage} -ge 2 ]; then
+    log "stage 2: Data Preparation"
+    mkdir -p data/{train,valid,test}
+    python3 local/data_prep.py ${MELD}
+    for x in test valid train; do
+        for f in text wav.scp utt2spk; do
+            sort data/${x}/${f} -o data/${x}/${f}
+        done
+        utils/utt2spk_to_spk2utt.pl data/${x}/utt2spk > "data/${x}/spk2utt"
+        utils/validate_data_dir.sh --no-feats data/${x} || exit 1
+    done
+fi
+
+log "Successfully finished. [elapsed=${SECONDS}s]"