Refactor CI scripts.

This commit is contained in:
Fangjun Kuang 2022-04-19 16:57:17 +08:00
parent 551f4631a3
commit e9a3e8376c
18 changed files with 519 additions and 657 deletions

View File

@ -0,0 +1,47 @@
#!/usr/bin/env bash
log() {
# This function is from espnet
local fname=${BASH_SOURCE[1]##*/}
echo -e "$(date '+%Y-%m-%d %H:%M:%S') (${fname}:${BASH_LINENO[0]}:${FUNCNAME[1]}) $*"
}
cd egs/librispeech/ASR
repo_url=https://huggingface.co/csukuangfj/icefall-asr-librispeech-pruned-transducer-stateless-2022-03-12
log "Downloading pre-trained model from $repo_url"
git lfs install
git clone $repo_url
repo=$(basename $repo_url)
log "Display test files"
tree $repo/
soxi $repo/test_wavs/*.wav
ls -lh $repo/test_wavs/*.wav
for sym in 1 2 3; do
log "Greedy search with --max-sym-per-frame $sym"
./pruned_transducer_stateless/pretrained.py \
--method greedy_search \
--max-sym-per-frame $sym \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
$repo/test_wavs/1089-134686-0001.wav \
$repo/test_wavs/1221-135766-0001.wav \
$repo/test_wavs/1221-135766-0002.wav
done
for method in modified_beam_search beam_search; do
log "$method"
./pruned_transducer_stateless/pretrained.py \
--method $method \
--beam-size 4 \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
$repo/test_wavs/1089-134686-0001.wav \
$repo/test_wavs/1221-135766-0001.wav \
$repo/test_wavs/1221-135766-0002.wav
done

View File

@ -0,0 +1,47 @@
#!/usr/bin/env bash
log() {
# This function is from espnet
local fname=${BASH_SOURCE[1]##*/}
echo -e "$(date '+%Y-%m-%d %H:%M:%S') (${fname}:${BASH_LINENO[0]}:${FUNCNAME[1]}) $*"
}
cd egs/librispeech/ASR
repo_url=https://huggingface.co/csukuangfj/icefall-asr-librispeech-transducer-stateless2-torchaudio-2022-04-19
log "Downloading pre-trained model from $repo_url"
git lfs install
git clone $repo_url
repo=$(basename $repo_url)
log "Display test files"
tree $repo/
soxi $repo/test_wavs/*.wav
ls -lh $repo/test_wavs/*.wav
for sym in 1 2 3; do
log "Greedy search with --max-sym-per-frame $sym"
./transducer_stateless2/pretrained.py \
--method greedy_search \
--max-sym-per-frame $sym \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
$repo/test_wavs/1089-134686-0001.wav \
$repo/test_wavs/1221-135766-0001.wav \
$repo/test_wavs/1221-135766-0002.wav
done
for method in modified_beam_search beam_search; do
log "$method"
./transducer_stateless2/pretrained.py \
--method $method \
--beam-size 4 \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
$repo/test_wavs/1089-134686-0001.wav \
$repo/test_wavs/1221-135766-0001.wav \
$repo/test_wavs/1221-135766-0002.wav
done

View File

@ -0,0 +1,46 @@
#!/usr/bin/env bash
log() {
# This function is from espnet
local fname=${BASH_SOURCE[1]##*/}
echo -e "$(date '+%Y-%m-%d %H:%M:%S') (${fname}:${BASH_LINENO[0]}:${FUNCNAME[1]}) $*"
}
cd egs/librispeech/ASR
repo_url=https://github.com/csukuangfj/icefall-asr-conformer-ctc-bpe-500
git lfs install
git clone $repo
log "Downloading pre-trained model from $repo_url"
git clone $repo_url
repo=$(basename $repo_url)
log "Display test files"
tree $repo/
soxi $repo/test_wavs/*.flac
ls -lh $repo/test_wavs/*.flac
log "CTC decoding"
./conformer_ctc/pretrained.py \
--method ctc-decoding \
--num-classes 500 \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
$repo/test_wavs/1089-134686-0001.flac \
$repo/test_wavs/1221-135766-0001.flac \
$repo/test_wavs/1221-135766-0002.flac
log "HLG decoding"
./conformer_ctc/pretrained.py \
--method 1best \
--num-classes 500 \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
--words-file $repo/data/lang_bpe_500/words.txt \
--HLG $repo/data/lang_bpe_500/HLG.pt \
$repo/test_wavs/1089-134686-0001.flac \
$repo/test_wavs/1221-135766-0001.flac \
$repo/test_wavs/1221-135766-0002.flac

View File

@ -0,0 +1,47 @@
#!/usr/bin/env bash
log() {
# This function is from espnet
local fname=${BASH_SOURCE[1]##*/}
echo -e "$(date '+%Y-%m-%d %H:%M:%S') (${fname}:${BASH_LINENO[0]}:${FUNCNAME[1]}) $*"
}
cd egs/librispeech/ASR
repo_url=https://huggingface.co/csukuangfj/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21
log "Downloading pre-trained model from $repo_url"
git lfs install
git clone $repo_url
repo=$(basename $repo_url)
log "Display test files"
tree $repo/
soxi $repo/test_wavs/*.wav
ls -lh $repo/test_wavs/*.wav
for sym in 1 2 3; do
log "Greedy search with --max-sym-per-frame $sym"
./transducer_stateless_multi_datasets/pretrained.py \
--method greedy_search \
--max-sym-per-frame $sym \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
$repo/test_wavs/1089-134686-0001.wav \
$repo/test_wavs/1221-135766-0001.wav \
$repo/test_wavs/1221-135766-0002.wav
done
for method in modified_beam_search beam_search; do
log "$method"
./transducer_stateless_multi_datasets/pretrained.py \
--method $method \
--beam-size 4 \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
$repo/test_wavs/1089-134686-0001.wav \
$repo/test_wavs/1221-135766-0001.wav \
$repo/test_wavs/1221-135766-0002.wav
done

View File

@ -0,0 +1,47 @@
#!/usr/bin/env bash
log() {
# This function is from espnet
local fname=${BASH_SOURCE[1]##*/}
echo -e "$(date '+%Y-%m-%d %H:%M:%S') (${fname}:${BASH_LINENO[0]}:${FUNCNAME[1]}) $*"
}
cd egs/librispeech/ASR
repo_url=https://huggingface.co/csukuangfj/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01
log "Downloading pre-trained model from $repo_url"
git lfs install
git clone $repo_url
repo=$(basename $repo_url)
log "Display test files"
tree $repo/
soxi $repo/test_wavs/*.wav
ls -lh $repo/test_wavs/*.wav
for sym in 1 2 3; do
log "Greedy search with --max-sym-per-frame $sym"
./transducer_stateless_multi_datasets/pretrained.py \
--method greedy_search \
--max-sym-per-frame $sym \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
$repo/test_wavs/1089-134686-0001.wav \
$repo/test_wavs/1221-135766-0001.wav \
$repo/test_wavs/1221-135766-0002.wav
done
for method in modified_beam_search beam_search; do
log "$method"
./transducer_stateless_multi_datasets/pretrained.py \
--method $method \
--beam-size 4 \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
$repo/test_wavs/1089-134686-0001.wav \
$repo/test_wavs/1221-135766-0001.wav \
$repo/test_wavs/1221-135766-0002.wav
done

View File

@ -0,0 +1,47 @@
#!/usr/bin/env bash
log() {
# This function is from espnet
local fname=${BASH_SOURCE[1]##*/}
echo -e "$(date '+%Y-%m-%d %H:%M:%S') (${fname}:${BASH_LINENO[0]}:${FUNCNAME[1]}) $*"
}
cd egs/aishell/ASR
repo_url=https://huggingface.co/csukuangfj/icefall-aishell-transducer-stateless-modified-2-2022-03-01
log "Downloading pre-trained model from $repo_url"
git lfs install
git clone $repo_url
repo=$(basename $repo_url)
log "Display test files"
tree $repo/
soxi $repo/test_wavs/*.wav
ls -lh $repo/test_wavs/*.wav
for sym in 1 2 3; do
log "Greedy search with --max-sym-per-frame $sym"
./transducer_stateless_modified-2/pretrained.py \
--method greedy_search \
--max-sym-per-frame $sym \
--checkpoint $repo/exp/pretrained.pt \
--lang-dir $repo/data/lang_char \
$repo/test_wavs/BAC009S0764W0121.wav \
$repo/test_wavs/BAC009S0764W0122.wav \
$repo/test_wavs/BAC009S0764W0123.wav
done
for method in modified_beam_search beam_search; do
log "$method"
./transducer_stateless_modified-2/pretrained.py \
--method $method \
--beam-size 4 \
--checkpoint $repo/exp/pretrained.pt \
--lang-dir $repo/data/lang_char \
$repo/test_wavs/BAC009S0764W0121.wav \
$repo/test_wavs/BAC009S0764W0122.wav \
$repo/test_wavs/BAC009S0764W0123.wav
done

View File

@ -0,0 +1,47 @@
#!/usr/bin/env bash
log() {
# This function is from espnet
local fname=${BASH_SOURCE[1]##*/}
echo -e "$(date '+%Y-%m-%d %H:%M:%S') (${fname}:${BASH_LINENO[0]}:${FUNCNAME[1]}) $*"
}
cd egs/aishell/ASR
repo_url=https://huggingface.co/csukuangfj/icefall-aishell-transducer-stateless-modified-2022-03-01
log "Downloading pre-trained model from $repo_url"
git lfs install
git clone $repo_url
repo=$(basename $repo_url)
log "Display test files"
tree $repo/
soxi $repo/test_wavs/*.wav
ls -lh $repo/test_wavs/*.wav
for sym in 1 2 3; do
log "Greedy search with --max-sym-per-frame $sym"
./transducer_stateless_modified/pretrained.py \
--method greedy_search \
--max-sym-per-frame $sym \
--checkpoint $repo/exp/pretrained.pt \
--lang-dir $repo/data/lang_char \
$repo/test_wavs/BAC009S0764W0121.wav \
$repo/test_wavs/BAC009S0764W0122.wav \
$repo/test_wavs/BAC009S0764W0123.wav
done
for method in modified_beam_search beam_search; do
log "$method"
./transducer_stateless_modified/pretrained.py \
--method $method \
--beam-size 4 \
--checkpoint $repo/exp/pretrained.pt \
--lang-dir $repo/data/lang_char \
$repo/test_wavs/BAC009S0764W0121.wav \
$repo/test_wavs/BAC009S0764W0122.wav \
$repo/test_wavs/BAC009S0764W0123.wav
done

View File

@ -0,0 +1,60 @@
#!/usr/bin/env bash
log() {
# This function is from espnet
local fname=${BASH_SOURCE[1]##*/}
echo -e "$(date '+%Y-%m-%d %H:%M:%S') (${fname}:${BASH_LINENO[0]}:${FUNCNAME[1]}) $*"
}
cd egs/librispeech/ASR
repo_url=https://huggingface.co/csukuangfj/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07
log "Downloading pre-trained model from $repo_url"
git lfs install
git clone $repo_url
repo=$(basename $repo_url)
log "Display test files"
tree $repo/
soxi $repo/test_wavs/*.wav
ls -lh $repo/test_wavs/*.wav
for sym in 1 2 3; do
log "Greedy search with --max-sym-per-frame $sym"
./transducer_stateless/pretrained.py \
--method greedy_search \
--max-sym-per-frame $sym \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
$repo/test_wavs/1089-134686-0001.wav \
$repo/test_wavs/1221-135766-0001.wav \
$repo/test_wavs/1221-135766-0002.wav
done
for method in modified_beam_search beam_search; do
log "$method"
./transducer_stateless/pretrained.py \
--method $method \
--beam-size 4 \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
$repo/test_wavs/1089-134686-0001.wav \
$repo/test_wavs/1221-135766-0001.wav \
$repo/test_wavs/1221-135766-0002.wav
done
for method in modified_beam_search beam_search; do
log "$method"
./transducer_stateless_multi_datasets/pretrained.py \
--method $method \
--beam-size 4 \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
$repo/test_wavs/1089-134686-0001.wav \
$repo/test_wavs/1221-135766-0001.wav \
$repo/test_wavs/1221-135766-0002.wav
done

32
.github/scripts/run-pre-trained-transducer.sh vendored Executable file
View File

@ -0,0 +1,32 @@
#!/usr/bin/env bash
log() {
# This function is from espnet
local fname=${BASH_SOURCE[1]##*/}
echo -e "$(date '+%Y-%m-%d %H:%M:%S') (${fname}:${BASH_LINENO[0]}:${FUNCNAME[1]}) $*"
}
cd egs/librispeech/ASR
repo_url=https://huggingface.co/csukuangfj/icefall-asr-librispeech-transducer-bpe-500-2021-12-23
log "Downloading pre-trained model from $repo_url"
git lfs install
git clone $repo_url
repo=$(basename $repo_url)
log "Display test files"
tree $repo/
soxi $repo/test_wavs/*.wav
ls -lh $repo/test_wavs/*.wav
log "Beam search decoding"
./transducer/pretrained.py \
--method beam_search \
--beam-size 4 \
--checkpoint $repo/exp/pretrained.pt \
--bpe-model $repo/data/lang_bpe_500/bpe.model \
$repo/test_wavs/1089-134686-0001.wav \
$repo/test_wavs/1221-135766-0001.wav \
$repo/test_wavs/1221-135766-0002.wav

View File

@ -40,11 +40,6 @@ jobs:
with:
fetch-depth: 0
- name: Install graphviz
shell: bash
run: |
sudo apt-get -qq install graphviz
- name: Setup Python ${{ matrix.python-version }}
uses: actions/setup-python@v2
with:
@ -77,104 +72,11 @@ jobs:
cmake -DCMAKE_BUILD_TYPE=Release ..
make -j2 _kaldifeat
- name: Download pre-trained model
shell: bash
run: |
sudo apt-get -qq install git-lfs
mkdir -p ~/tmp
cd ~/tmp
git lfs install
git clone https://huggingface.co/csukuangfj/icefall-asr-librispeech-pruned-transducer-stateless-2022-03-12
- name: Display test files
shell: bash
run: |
sudo apt-get -qq install tree sox
tree ~/tmp/icefall-asr-librispeech-pruned-transducer-stateless-2022-03-12
soxi ~/tmp/icefall-asr-librispeech-pruned-transducer-stateless-2022-03-12/test_wavs/*.wav
ls -lh ~/tmp/icefall-asr-librispeech-pruned-transducer-stateless-2022-03-12/test_wavs/*.wav
- name: Run greedy search decoding (max-sym-per-frame 1)
- name: Inference with pre-trained model
shell: bash
run: |
sudo apt-get -qq install git-lfs tree sox
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
dir=~/tmp/icefall-asr-librispeech-pruned-transducer-stateless-2022-03-12
cd egs/librispeech/ASR
./pruned_transducer_stateless/pretrained.py \
--method greedy_search \
--max-sym-per-frame 1 \
--checkpoint $dir/exp/pretrained.pt \
--bpe-model $dir/data/lang_bpe_500/bpe.model \
$dir/test_wavs/1089-134686-0001.wav \
$dir/test_wavs/1221-135766-0001.wav \
$dir/test_wavs/1221-135766-0002.wav
- name: Run greedy search decoding (max-sym-per-frame 2)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
dir=~/tmp/icefall-asr-librispeech-pruned-transducer-stateless-2022-03-12
cd egs/librispeech/ASR
./pruned_transducer_stateless/pretrained.py \
--method greedy_search \
--max-sym-per-frame 2 \
--checkpoint $dir/exp/pretrained.pt \
--bpe-model $dir/data/lang_bpe_500/bpe.model \
$dir/test_wavs/1089-134686-0001.wav \
$dir/test_wavs/1221-135766-0001.wav \
$dir/test_wavs/1221-135766-0002.wav
- name: Run greedy search decoding (max-sym-per-frame 3)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
dir=~/tmp/icefall-asr-librispeech-pruned-transducer-stateless-2022-03-12
cd egs/librispeech/ASR
./pruned_transducer_stateless/pretrained.py \
--method greedy_search \
--max-sym-per-frame 3 \
--checkpoint $dir/exp/pretrained.pt \
--bpe-model $dir/data/lang_bpe_500/bpe.model \
$dir/test_wavs/1089-134686-0001.wav \
$dir/test_wavs/1221-135766-0001.wav \
$dir/test_wavs/1221-135766-0002.wav
- name: Run beam search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
dir=~/tmp/icefall-asr-librispeech-pruned-transducer-stateless-2022-03-12
cd egs/librispeech/ASR
./pruned_transducer_stateless/pretrained.py \
--method beam_search \
--beam-size 4 \
--checkpoint $dir/exp/pretrained.pt \
--bpe-model $dir/data/lang_bpe_500/bpe.model \
$dir/test_wavs/1089-134686-0001.wav \
$dir/test_wavs/1221-135766-0001.wav \
$dir/test_wavs/1221-135766-0002.wav
- name: Run modified beam search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
dir=~/tmp/icefall-asr-librispeech-pruned-transducer-stateless-2022-03-12
cd egs/librispeech/ASR
./pruned_transducer_stateless/pretrained.py \
--method modified_beam_search \
--beam-size 4 \
--checkpoint $dir/exp/pretrained.pt \
--bpe-model $dir/data/lang_bpe_500/bpe.model \
$dir/test_wavs/1089-134686-0001.wav \
$dir/test_wavs/1221-135766-0001.wav \
$dir/test_wavs/1221-135766-0002.wav
.github/scripts/run-librispeech-pruned-transducer-stateless-2022-03-12.sh

View File

@ -0,0 +1,82 @@
# Copyright 2021 Fangjun Kuang (csukuangfj@gmail.com)
# See ../../LICENSE for clarification regarding multiple authors
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
name: run-librispeech-2022-04-19
# stateless transducer + torchaudio rnn-t loss
on:
push:
branches:
- master
pull_request:
types: [labeled]
jobs:
run_librispeech_2022_04_19:
if: github.event.label.name == 'ready' || github.event_name == 'push'
runs-on: ${{ matrix.os }}
strategy:
matrix:
os: [ubuntu-18.04]
python-version: [3.7, 3.8, 3.9]
fail-fast: false
steps:
- uses: actions/checkout@v2
with:
fetch-depth: 0
- name: Setup Python ${{ matrix.python-version }}
uses: actions/setup-python@v2
with:
python-version: ${{ matrix.python-version }}
cache: 'pip'
cache-dependency-path: '**/requirements-ci.txt'
- name: Install Python dependencies
run: |
grep -v '^#' ./requirements-ci.txt | xargs -n 1 -L 1 pip install
- name: Cache kaldifeat
id: my-cache
uses: actions/cache@v2
with:
path: |
~/tmp/kaldifeat
key: cache-tmp-${{ matrix.python-version }}
- name: Install kaldifeat
if: steps.my-cache.outputs.cache-hit != 'true'
shell: bash
run: |
mkdir -p ~/tmp
cd ~/tmp
git clone https://github.com/csukuangfj/kaldifeat
cd kaldifeat
mkdir build
cd build
cmake -DCMAKE_BUILD_TYPE=Release ..
make -j2 _kaldifeat
- name: Inference with pre-trained model
shell: bash
run: |
sudo apt-get -qq install git-lfs tree sox
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
.github/scripts/run-librispeech-transducer-stateless2-2022-04-19.sh

View File

@ -39,11 +39,6 @@ jobs:
with:
fetch-depth: 0
- name: Install graphviz
shell: bash
run: |
sudo apt-get -qq install graphviz
- name: Setup Python ${{ matrix.python-version }}
uses: actions/setup-python@v2
with:
@ -76,48 +71,11 @@ jobs:
cmake -DCMAKE_BUILD_TYPE=Release ..
make -j2 _kaldifeat
- name: Download pre-trained model
- name: Inference with pre-trained model
shell: bash
run: |
sudo apt-get -qq install git-lfs tree sox
cd egs/librispeech/ASR
mkdir tmp
cd tmp
git lfs install
git clone https://github.com/csukuangfj/icefall-asr-conformer-ctc-bpe-500
cd ..
tree tmp
soxi tmp/icefall-asr-conformer-ctc-bpe-500/test_wavs/*.flac
ls -lh tmp/icefall-asr-conformer-ctc-bpe-500/test_wavs/*.flac
- name: Run CTC decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./conformer_ctc/pretrained.py \
--num-classes 500 \
--checkpoint ./tmp/icefall-asr-conformer-ctc-bpe-500/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-conformer-ctc-bpe-500/data/lang_bpe_500/bpe.model \
--method ctc-decoding \
./tmp/icefall-asr-conformer-ctc-bpe-500/test_wavs/1089-134686-0001.flac \
./tmp/icefall-asr-conformer-ctc-bpe-500/test_wavs/1221-135766-0001.flac \
./tmp/icefall-asr-conformer-ctc-bpe-500/test_wavs/1221-135766-0002.flac
- name: Run HLG decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./conformer_ctc/pretrained.py \
--num-classes 500 \
--checkpoint ./tmp/icefall-asr-conformer-ctc-bpe-500/exp/pretrained.pt \
--words-file ./tmp/icefall-asr-conformer-ctc-bpe-500/data/lang_bpe_500/words.txt \
--HLG ./tmp/icefall-asr-conformer-ctc-bpe-500/data/lang_bpe_500/HLG.pt \
./tmp/icefall-asr-conformer-ctc-bpe-500/test_wavs/1089-134686-0001.flac \
./tmp/icefall-asr-conformer-ctc-bpe-500/test_wavs/1221-135766-0001.flac \
./tmp/icefall-asr-conformer-ctc-bpe-500/test_wavs/1221-135766-0002.flac
.github/scripts/run-pre-trained-conformer-ctc.sh

View File

@ -39,11 +39,6 @@ jobs:
with:
fetch-depth: 0
- name: Install graphviz
shell: bash
run: |
sudo apt-get -qq install graphviz
- name: Setup Python ${{ matrix.python-version }}
uses: actions/setup-python@v2
with:
@ -76,97 +71,11 @@ jobs:
cmake -DCMAKE_BUILD_TYPE=Release ..
make -j2 _kaldifeat
- name: Download pre-trained model
- name: Inference with pre-trained model
shell: bash
run: |
sudo apt-get -qq install git-lfs tree sox
cd egs/librispeech/ASR
mkdir tmp
cd tmp
git lfs install
git clone https://huggingface.co/csukuangfj/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21
cd ..
tree tmp
soxi tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/*.wav
ls -lh tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/*.wav
- name: Run greedy search decoding (max-sym-per-frame 1)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless_multi_datasets/pretrained.py \
--method greedy_search \
--max-sym-per-frame 1 \
--checkpoint ./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1221-135766-0002.wav
- name: Run greedy search decoding (max-sym-per-frame 2)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless_multi_datasets/pretrained.py \
--method greedy_search \
--max-sym-per-frame 2 \
--checkpoint ./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1221-135766-0002.wav
- name: Run greedy search decoding (max-sym-per-frame 3)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless_multi_datasets/pretrained.py \
--method greedy_search \
--max-sym-per-frame 3 \
--checkpoint ./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1221-135766-0002.wav
- name: Run beam search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless_multi_datasets/pretrained.py \
--method beam_search \
--beam-size 4 \
--checkpoint ./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1221-135766-0002.wav
- name: Run modified beam search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless_multi_datasets/pretrained.py \
--method modified_beam_search \
--beam-size 4 \
--checkpoint ./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-100h-transducer-stateless-multi-datasets-bpe-500-2022-02-21/test_wavs/1221-135766-0002.wav
.github/scripts/run-pre-trained-transducer-stateless-librispeech-100h.sh

View File

@ -39,11 +39,6 @@ jobs:
with:
fetch-depth: 0
- name: Install graphviz
shell: bash
run: |
sudo apt-get -qq install graphviz
- name: Setup Python ${{ matrix.python-version }}
uses: actions/setup-python@v2
with:
@ -76,99 +71,11 @@ jobs:
cmake -DCMAKE_BUILD_TYPE=Release ..
make -j2 _kaldifeat
- name: Download pre-trained model
- name: Inference with pre-trained model
shell: bash
run: |
sudo apt-get -qq install git-lfs tree sox
cd egs/librispeech/ASR
mkdir tmp
cd tmp
git lfs install
git clone https://huggingface.co/csukuangfj/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01
cd ..
tree tmp
soxi tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/*.wav
ls -lh tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/*.wav
- name: Run greedy search decoding (max-sym-per-frame 1)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless_multi_datasets/pretrained.py \
--method greedy_search \
--max-sym-per-frame 1 \
--checkpoint ./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1221-135766-0002.wav
- name: Run greedy search decoding (max-sym-per-frame 2)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless_multi_datasets/pretrained.py \
--method greedy_search \
--max-sym-per-frame 2 \
--checkpoint ./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1221-135766-0002.wav
- name: Run greedy search decoding (max-sym-per-frame 3)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless_multi_datasets/pretrained.py \
--method greedy_search \
--max-sym-per-frame 3 \
--checkpoint ./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1221-135766-0002.wav
- name: Run beam search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless_multi_datasets/pretrained.py \
--method beam_search \
--beam-size 4 \
--checkpoint ./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1221-135766-0002.wav
- name: Run modified beam search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless_multi_datasets/pretrained.py \
--method modified_beam_search \
--beam-size 4 \
--checkpoint ./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-multi-datasets-bpe-500-2022-03-01/test_wavs/1221-135766-0002.wav
.github/scripts/run-pre-trained-transducer-stateless-librispeech-960h.sh

View File

@ -39,11 +39,6 @@ jobs:
with:
fetch-depth: 0
- name: Install graphviz
shell: bash
run: |
sudo apt-get -qq install graphviz
- name: Setup Python ${{ matrix.python-version }}
uses: actions/setup-python@v2
with:
@ -76,98 +71,11 @@ jobs:
cmake -DCMAKE_BUILD_TYPE=Release ..
make -j2 _kaldifeat
- name: Download pre-trained model
- name: Inference with pre-trained model
shell: bash
run: |
sudo apt-get -qq install git-lfs tree sox
cd egs/aishell/ASR
mkdir tmp
cd tmp
git lfs install
git clone https://huggingface.co/csukuangfj/icefall-aishell-transducer-stateless-modified-2-2022-03-01
cd ..
tree tmp
soxi tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/*.wav
ls -lh tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/*.wav
- name: Run greedy search decoding (max-sym-per-frame 1)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/aishell/ASR
./transducer_stateless_modified-2/pretrained.py \
--method greedy_search \
--max-sym-per-frame 1 \
--checkpoint ./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/exp/pretrained.pt \
--lang-dir ./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/data/lang_char \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0121.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0122.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0123.wav
- name: Run greedy search decoding (max-sym-per-frame 2)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/aishell/ASR
./transducer_stateless_modified-2/pretrained.py \
--method greedy_search \
--max-sym-per-frame 2 \
--checkpoint ./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/exp/pretrained.pt \
--lang-dir ./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/data/lang_char \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0121.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0122.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0123.wav
- name: Run greedy search decoding (max-sym-per-frame 3)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/aishell/ASR
./transducer_stateless_modified-2/pretrained.py \
--method greedy_search \
--max-sym-per-frame 3 \
--checkpoint ./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/exp/pretrained.pt \
--lang-dir ./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/data/lang_char \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0121.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0122.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0123.wav
- name: Run beam search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/aishell/ASR
./transducer_stateless_modified-2/pretrained.py \
--method beam_search \
--beam-size 4 \
--checkpoint ./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/exp/pretrained.pt \
--lang-dir ./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/data/lang_char \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0121.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0122.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0123.wav
- name: Run modified beam search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/aishell/ASR
./transducer_stateless_modified-2/pretrained.py \
--method modified_beam_search \
--beam-size 4 \
--checkpoint ./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/exp/pretrained.pt \
--lang-dir ./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/data/lang_char \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0121.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0122.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2-2022-03-01/test_wavs/BAC009S0764W0123.wav
.github/scripts/run-pre-trained-transducer-stateless-modified-2-aishell.sh

View File

@ -39,11 +39,6 @@ jobs:
with:
fetch-depth: 0
- name: Install graphviz
shell: bash
run: |
sudo apt-get -qq install graphviz
- name: Setup Python ${{ matrix.python-version }}
uses: actions/setup-python@v2
with:
@ -76,98 +71,11 @@ jobs:
cmake -DCMAKE_BUILD_TYPE=Release ..
make -j2 _kaldifeat
- name: Download pre-trained model
- name: Inference with pre-trained model
shell: bash
run: |
sudo apt-get -qq install git-lfs tree sox
cd egs/aishell/ASR
mkdir tmp
cd tmp
git lfs install
git clone https://huggingface.co/csukuangfj/icefall-aishell-transducer-stateless-modified-2022-03-01
cd ..
tree tmp
soxi tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/*.wav
ls -lh tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/*.wav
- name: Run greedy search decoding (max-sym-per-frame 1)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/aishell/ASR
./transducer_stateless_modified/pretrained.py \
--method greedy_search \
--max-sym-per-frame 1 \
--checkpoint ./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/exp/pretrained.pt \
--lang-dir ./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/data/lang_char \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0121.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0122.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0123.wav
- name: Run greedy search decoding (max-sym-per-frame 2)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/aishell/ASR
./transducer_stateless_modified/pretrained.py \
--method greedy_search \
--max-sym-per-frame 2 \
--checkpoint ./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/exp/pretrained.pt \
--lang-dir ./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/data/lang_char \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0121.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0122.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0123.wav
- name: Run greedy search decoding (max-sym-per-frame 3)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/aishell/ASR
./transducer_stateless_modified/pretrained.py \
--method greedy_search \
--max-sym-per-frame 3 \
--checkpoint ./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/exp/pretrained.pt \
--lang-dir ./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/data/lang_char \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0121.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0122.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0123.wav
- name: Run beam search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/aishell/ASR
./transducer_stateless_modified/pretrained.py \
--method beam_search \
--beam-size 4 \
--checkpoint ./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/exp/pretrained.pt \
--lang-dir ./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/data/lang_char \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0121.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0122.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0123.wav
- name: Run modified beam search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/aishell/ASR
./transducer_stateless_modified/pretrained.py \
--method modified_beam_search \
--beam-size 4 \
--checkpoint ./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/exp/pretrained.pt \
--lang-dir ./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/data/lang_char \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0121.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0122.wav \
./tmp/icefall-aishell-transducer-stateless-modified-2022-03-01/test_wavs/BAC009S0764W0123.wav
.github/scripts/run-pre-trained-transducer-stateless-modified-aishell.sh

View File

@ -39,11 +39,6 @@ jobs:
with:
fetch-depth: 0
- name: Install graphviz
shell: bash
run: |
sudo apt-get -qq install graphviz
- name: Setup Python ${{ matrix.python-version }}
uses: actions/setup-python@v2
with:
@ -76,96 +71,11 @@ jobs:
cmake -DCMAKE_BUILD_TYPE=Release ..
make -j2 _kaldifeat
- name: Download pre-trained model
- name: Inference with pre-trained model
shell: bash
run: |
sudo apt-get -qq install git-lfs tree sox
cd egs/librispeech/ASR
mkdir tmp
cd tmp
git lfs install
git clone https://huggingface.co/csukuangfj/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07
cd ..
tree tmp
soxi tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/*.wav
ls -lh tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/*.wav
- name: Run greedy search decoding (max-sym-per-frame 1)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless/pretrained.py \
--method greedy_search \
--max-sym-per-frame 1 \
--checkpoint ./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1221-135766-0002.wav
- name: Run greedy search decoding (max-sym-per-frame 2)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless/pretrained.py \
--method greedy_search \
--max-sym-per-frame 2 \
--checkpoint ./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1221-135766-0002.wav
- name: Run greedy search decoding (max-sym-per-frame 3)
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless/pretrained.py \
--method greedy_search \
--max-sym-per-frame 3 \
--checkpoint ./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1221-135766-0002.wav
- name: Run beam search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless/pretrained.py \
--method beam_search \
--beam-size 4 \
--checkpoint ./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1221-135766-0002.wav
- name: Run modified beam search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer_stateless/pretrained.py \
--method modified_beam_search \
--beam-size 4 \
--checkpoint ./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-transducer-stateless-bpe-500-2022-02-07/test_wavs/1221-135766-0002.wav
.github/scripts/run-pre-trained-transducer-stateless.sh

View File

@ -39,11 +39,6 @@ jobs:
with:
fetch-depth: 0
- name: Install graphviz
shell: bash
run: |
sudo apt-get -qq install graphviz
- name: Setup Python ${{ matrix.python-version }}
uses: actions/setup-python@v2
with:
@ -76,48 +71,11 @@ jobs:
cmake -DCMAKE_BUILD_TYPE=Release ..
make -j2 _kaldifeat
- name: Download pre-trained model
- name: Inference with pre-trained model
shell: bash
run: |
sudo apt-get -qq install git-lfs tree sox
cd egs/librispeech/ASR
mkdir tmp
cd tmp
git lfs install
git clone https://huggingface.co/csukuangfj/icefall-asr-librispeech-transducer-bpe-500-2021-12-23
cd ..
tree tmp
soxi tmp/icefall-asr-librispeech-transducer-bpe-500-2021-12-23/test_wavs/*.wav
ls -lh tmp/icefall-asr-librispeech-transducer-bpe-500-2021-12-23/test_wavs/*.wav
- name: Run greedy search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer/pretrained.py \
--method greedy_search \
--checkpoint ./tmp/icefall-asr-librispeech-transducer-bpe-500-2021-12-23/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-transducer-bpe-500-2021-12-23/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-transducer-bpe-500-2021-12-23/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-transducer-bpe-500-2021-12-23/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-transducer-bpe-500-2021-12-23/test_wavs/1221-135766-0002.wav
- name: Run beam search decoding
shell: bash
run: |
export PYTHONPATH=$PWD:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/kaldifeat/python:$PYTHONPATH
export PYTHONPATH=~/tmp/kaldifeat/build/lib:$PYTHONPATH
cd egs/librispeech/ASR
./transducer/pretrained.py \
--method beam_search \
--beam-size 4 \
--checkpoint ./tmp/icefall-asr-librispeech-transducer-bpe-500-2021-12-23/exp/pretrained.pt \
--bpe-model ./tmp/icefall-asr-librispeech-transducer-bpe-500-2021-12-23/data/lang_bpe_500/bpe.model \
./tmp/icefall-asr-librispeech-transducer-bpe-500-2021-12-23/test_wavs/1089-134686-0001.wav \
./tmp/icefall-asr-librispeech-transducer-bpe-500-2021-12-23/test_wavs/1221-135766-0001.wav \
./tmp/icefall-asr-librispeech-transducer-bpe-500-2021-12-23/test_wavs/1221-135766-0002.wav
.github/scripts/run-pre-trained-transducer.sh