From 100ea0304b956e55a9c2fe284b1ee1a26bdf2b7c Mon Sep 17 00:00:00 2001
From: 游雁 <zhifu.gzf@alibaba-inc.com>
Date: 星期四, 20 四月 2023 23:49:15 +0800
Subject: [PATCH] docs
---
egs_modelscope/asr/TEMPLATE/README.md | 75 +++++++++++++++---------
egs_modelscope/asr/TEMPLATE/infer.sh | 7 ++
egs/aishell/transformer/utils/prepare_checkpoint.py | 21 +++---
docs/docker.md | 13 +++-
docs/modescope_pipeline/quick_start.md | 16 ++--
5 files changed, 82 insertions(+), 50 deletions(-)
diff --git a/docs/docker.md b/docs/docker.md
index 90616fd..95a75f2 100644
--- a/docs/docker.md
+++ b/docs/docker.md
@@ -1,6 +1,6 @@
# Docker
-## Install
+## Install Docker
### Ubuntu
```shell
@@ -16,6 +16,11 @@
### CentOS
```shell
curl -fsSL https://get.docker.com | bash -s docker --mirror Aliyun
+```
+
+### MacOS
+```shell
+brew install --cask --appdir=/Applications docker
```
### Windows
@@ -40,18 +45,20 @@
sudo docker pull <image-name>:<tag>
```
-### Check Downloaded Image
+### Check Image
```shell
sudo docker images
```
## Run Docker
```shell
-sudo docker run -it <image-name>:<tag> bash
+sudo docker run -itd --name funasr <image-name>:<tag> bash
+sudo docker exec -it funasr bash
```
## Stop Docker
```shell
+exit
sudo docker ps
sudo docker stop <container-id>
```
diff --git a/docs/modescope_pipeline/quick_start.md b/docs/modescope_pipeline/quick_start.md
index b1614f5..50dac36 100644
--- a/docs/modescope_pipeline/quick_start.md
+++ b/docs/modescope_pipeline/quick_start.md
@@ -3,7 +3,7 @@
## Inference with pipeline
### Speech Recognition
-#### Paraformer model
+#### Paraformer Model
```python
from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks
@@ -18,7 +18,7 @@
```
### Voice Activity Detection
-#### FSMN-VAD
+#### FSMN-VAD Model
```python
from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks
@@ -37,7 +37,7 @@
```
### Punctuation Restoration
-#### CT_Transformer
+#### CT_Transformer Model
```python
from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks
@@ -52,7 +52,7 @@
```
### Timestamp Prediction
-#### TP-Aligner
+#### TP-Aligner Model
```python
from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks
@@ -68,7 +68,7 @@
```
### Speaker Verification
-#### X-vector
+#### X-vector Model
```python
from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks
@@ -87,8 +87,8 @@
print(rec_result["scores"][0])
```
-### Speaker diarization
-#### SOND
+### Speaker Diarization
+#### SOND Model
```python
from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks
@@ -154,7 +154,7 @@
## Finetune with pipeline
### Speech Recognition
-#### Paraformer model
+#### Paraformer Model
finetune.py
```python
diff --git a/egs_modelscope/asr/TEMPLATE/infer_after_finetune.py b/egs/aishell/transformer/utils/prepare_checkpoint.py
similarity index 75%
rename from egs_modelscope/asr/TEMPLATE/infer_after_finetune.py
rename to egs/aishell/transformer/utils/prepare_checkpoint.py
index 2d311dd..01763d4 100644
--- a/egs_modelscope/asr/TEMPLATE/infer_after_finetune.py
+++ b/egs/aishell/transformer/utils/prepare_checkpoint.py
@@ -1,12 +1,9 @@
-import json
import os
import shutil
from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks
from modelscope.hub.snapshot_download import snapshot_download
-
-from funasr.utils.compute_wer import compute_wer
def modelscope_infer_after_finetune(params):
# prepare for decoding
@@ -39,10 +36,14 @@
if __name__ == '__main__':
- params = {}
- params["modelscope_model_name"] = "damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
- params["output_dir"] = "./checkpoint"
- params["data_dir"] = "./data/test"
- params["decoding_model_name"] = "valid.acc.ave_10best.pb"
- params["batch_size"] = 64
- modelscope_infer_after_finetune(params)
\ No newline at end of file
+ import sys
+
+ model = sys.argv[1]
+ checkpoint_dir = sys.argv[2]
+ checkpoint_name = sys.argv[3]
+
+ try:
+ pretrained_model_path = snapshot_download(model, cache_dir=checkpoint_dir)
+ except BaseException:
+ raise BaseException(f"Please download pretrain model from ModelScope firstly.")
+ shutil.copy(os.path.join(checkpoint_dir, checkpoint_name), os.path.join(pretrained_model_path, "model.pb"))
diff --git a/egs_modelscope/asr/TEMPLATE/README.md b/egs_modelscope/asr/TEMPLATE/README.md
index 8b6b24d..a5d7d6e 100644
--- a/egs_modelscope/asr/TEMPLATE/README.md
+++ b/egs_modelscope/asr/TEMPLATE/README.md
@@ -1,12 +1,12 @@
# Speech Recognition
> **Note**:
-> The modelscope pipeline supports all the models in [model zoo](https://alibaba-damo-academy.github.io/FunASR/en/modelscope_models.html#pretrained-models-on-modelscope) to inference and finetine. Here we take model of Paraformer and Paraformer-online as example to demonstrate the usage.
+> The modelscope pipeline supports all the models in [model zoo](https://alibaba-damo-academy.github.io/FunASR/en/modelscope_models.html#pretrained-models-on-modelscope) to inference and finetine. Here we take typic model as example to demonstrate the usage.
## Inference
### Quick start
-#### [Paraformer model](https://www.modelscope.cn/models/damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch/summary)
+#### [Paraformer Model](https://www.modelscope.cn/models/damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch/summary)
```python
from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks
@@ -19,7 +19,7 @@
rec_result = inference_pipeline(audio_in='https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_zh.wav')
print(rec_result)
```
-#### [Paraformer-online model](https://www.modelscope.cn/models/damo/speech_paraformer_asr_nat-zh-cn-16k-common-vocab8404-online/summary)
+#### [Paraformer-online Model](https://www.modelscope.cn/models/damo/speech_paraformer_asr_nat-zh-cn-16k-common-vocab8404-online/summary)
```python
inference_pipeline = pipeline(
task=Tasks.auto_speech_recognition,
@@ -41,7 +41,7 @@
```
Full code of demo, please ref to [demo](https://github.com/alibaba-damo-academy/FunASR/discussions/241)
-#### [UniASR model](https://www.modelscope.cn/models/damo/speech_UniASR_asr_2pass-zh-cn-8k-common-vocab3445-pytorch-online/summary)
+#### [UniASR Model](https://www.modelscope.cn/models/damo/speech_UniASR_asr_2pass-zh-cn-8k-common-vocab3445-pytorch-online/summary)
There are three decoding mode for UniASR model(`fast`銆乣normal`銆乣offline`), for more model detailes, please refer to [docs](https://www.modelscope.cn/models/damo/speech_UniASR_asr_2pass-zh-cn-8k-common-vocab3445-pytorch-online/summary)
```python
decoding_model = "fast" # "fast"銆�"normal"銆�"offline"
@@ -59,21 +59,21 @@
Undo
#### API-reference
-##### define pipeline
+##### Define pipeline
- `task`: `Tasks.auto_speech_recognition`
- `model`: model name in [model zoo](https://alibaba-damo-academy.github.io/FunASR/en/modelscope_models.html#pretrained-models-on-modelscope), or model path in local disk
- `ngpu`: 1 (Defalut), decoding on GPU. If ngpu=0, decoding on CPU
- `ncpu`: 1 (Defalut), sets the number of threads used for intraop parallelism on CPU
- `output_dir`: None (Defalut), the output path of results if set
- `batch_size`: 1 (Defalut), batch size when decoding
-##### infer pipeline
+##### Infer pipeline
- `audio_in`: the input to decode, which could be:
- wav_path, `e.g.`: asr_example.wav,
- pcm_path, `e.g.`: asr_example.pcm,
- audio bytes stream, `e.g.`: bytes data from a microphone
- audio sample point锛宍e.g.`: `audio, rate = soundfile.read("asr_example_zh.wav")`, the dtype is numpy.ndarray or torch.Tensor
- wav.scp, kaldi style wav list (`wav_id \t wav_path``), `e.g.`:
- ```cat wav.scp
+ ```text
asr_example1 ./audios/asr_example1.wav
asr_example2 ./audios/asr_example2.wav
```
@@ -85,13 +85,15 @@
FunASR also offer recipes [infer.sh](https://github.com/alibaba-damo-academy/FunASR/blob/main/egs_modelscope/asr/TEMPLATE/infer.sh) to decode with multi-thread CPUs, or multi GPUs.
- Setting parameters in `infer.sh`
- - <strong>model:</strong> # model name on ModelScope
- - <strong>data_dir:</strong> # the dataset dir needs to include `${data_dir}/wav.scp`. If `${data_dir}/text` is also exists, CER will be computed
- - <strong>output_dir:</strong> # result dir
- - <strong>batch_size:</strong> # batchsize of inference
- - <strong>gpu_inference:</strong> # whether to perform gpu decoding, set false for cpu decoding
- - <strong>gpuid_list:</strong> # set gpus, e.g., gpuid_list="0,1"
- - <strong>njob:</strong> # the number of jobs for CPU decoding, if `gpu_inference`=false, use CPU decoding, please set `njob`
+ - `model`: model name on ModelScope
+ - `data_dir`: the dataset dir needs to include `${data_dir}/wav.scp`. If `${data_dir}/text` is also exists, CER will be computed
+ - `output_dir`: result dir
+ - `batch_size`: batchsize of inference
+ - `gpu_inference`: whether to perform gpu decoding, set false for cpu decoding
+ - `gpuid_list`: set gpus, e.g., `gpuid_list`="0,1"
+ - `njob`: the number of jobs for CPU decoding, if `gpu_inference`=false, use CPU decoding, please set `njob`
+ - `checkpoint_dir`: only used for infer finetuned models, the path dir of finetuned models
+ - `checkpoint_name`: only used for infer finetuned models, `valid.cer_ctc.ave.pb` (Default), which checkpoint is used to infer
- Decode with multi GPUs:
```shell
@@ -167,12 +169,12 @@
### Finetune with your data
- Modify finetune training related parameters in [finetune.py](https://github.com/alibaba-damo-academy/FunASR/blob/main/egs_modelscope/asr/TEMPLATE/finetune.py)
- - <strong>output_dir:</strong> # result dir
- - <strong>data_dir:</strong> # the dataset dir needs to include files: `train/wav.scp`, `train/text`; `validation/wav.scp`, `validation/text`
- - <strong>dataset_type:</strong> # for dataset larger than 1000 hours, set as `large`, otherwise set as `small`
- - <strong>batch_bins:</strong> # batch size. For dataset_type is `small`, `batch_bins` indicates the feature frames. For dataset_type is `large`, `batch_bins` indicates the duration in ms
- - <strong>max_epoch:</strong> # number of training epoch
- - <strong>lr:</strong> # learning rate
+ - `output_dir`: result dir
+ - `data_dir`: the dataset dir needs to include files: `train/wav.scp`, `train/text`; `validation/wav.scp`, `validation/text`
+ - `dataset_type`: for dataset larger than 1000 hours, set as `large`, otherwise set as `small`
+ - `batch_bins`: batch size. For dataset_type is `small`, `batch_bins` indicates the feature frames. For dataset_type is `large`, `batch_bins` indicates the duration in ms
+ - `max_epoch`: number of training epoch
+ - `lr`: learning rate
- Then you can run the pipeline to finetune with:
```shell
@@ -183,14 +185,29 @@
CUDA_VISIBLE_DEVICES=1,2 python -m torch.distributed.launch --nproc_per_node 2 finetune.py > log.txt 2>&1
```
## Inference with your finetuned model
-- Modify inference related parameters in [infer_after_finetune.py](https://github.com/alibaba-damo-academy/FunASR/blob/main/egs_modelscope/asr/TEMPLATE/infer_after_finetune.py)
- - <strong>modelscope_model_name: </strong> # model name on ModelScope
- - <strong>output_dir:</strong> # result dir
- - <strong>data_dir:</strong> # the dataset dir needs to include `test/wav.scp`. If `test/text` is also exists, CER will be computed
- - <strong>decoding_model_name:</strong> # set the checkpoint name for decoding, e.g., `valid.cer_ctc.ave.pb`
- - <strong>batch_size:</strong> # batchsize of inference
-- Then you can run the pipeline to finetune with:
-```python
- python infer_after_finetune.py
+- Setting parameters in [infer.sh](https://github.com/alibaba-damo-academy/FunASR/blob/main/egs_modelscope/asr/TEMPLATE/infer.sh) is the same with [docs](https://github.com/alibaba-damo-academy/FunASR/tree/main/egs_modelscope/asr/TEMPLATE#inference-with-multi-thread-cpus-or-multi-gpus)
+
+- Decode with multi GPUs:
+```shell
+ bash infer.sh \
+ --model "damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch" \
+ --data_dir "./data/test" \
+ --output_dir "./results" \
+ --batch_size 64 \
+ --gpu_inference true \
+ --gpuid_list "0,1" \
+ --checkpoint_dir "./checkpoint" \
+ --checkpoint_name "valid.cer_ctc.ave.pb"
```
+- Decode with multi-thread CPUs:
+```shell
+ bash infer.sh \
+ --model "damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch" \
+ --data_dir "./data/test" \
+ --output_dir "./results" \
+ --gpu_inference false \
+ --njob 64 \
+ --checkpoint_dir "./checkpoint" \
+ --checkpoint_name "valid.cer_ctc.ave.pb"
+```
\ No newline at end of file
diff --git a/egs_modelscope/asr/TEMPLATE/infer.sh b/egs_modelscope/asr/TEMPLATE/infer.sh
index b8b011c..0d54abc 100644
--- a/egs_modelscope/asr/TEMPLATE/infer.sh
+++ b/egs_modelscope/asr/TEMPLATE/infer.sh
@@ -13,6 +13,8 @@
gpu_inference=true # whether to perform gpu decoding
gpuid_list="0,1" # set gpus, e.g., gpuid_list="0,1"
njob=4 # the number of jobs for CPU decoding, if gpu_inference=false, use CPU decoding, please set njob
+checkpoint_dir=
+checkpoint_name="valid.cer_ctc.ave.pb"
. utils/parse_options.sh || exit 1;
@@ -34,6 +36,11 @@
done
perl utils/split_scp.pl ${data_dir}/wav.scp ${split_scps}
+if ${checkpoint_dir}; then
+ python utils/prepare_checkpoint.py ${model} ${checkpoint_dir} ${checkpoint_name}
+ model=${checkpoint_dir}/${model}
+fi
+
if [ $stage -le 1 ] && [ $stop_stage -ge 1 ];then
echo "Decoding ..."
gpuid_list_array=(${gpuid_list//,/ })
--
Gitblit v1.9.1