Comments (16)
Likely a system problem, like full disk, OOM killer, something like that.
If you rerun that command it should hopefully continue from the stage that failed.
from icefall.
What's the output of the following command?
head -n 5 /home/maduo/miniconda3/envs/k2-fsa20210823/bin/lhotse
If the first line of the output is #!python
, you would probably need to manually modify it to #!/usr/bin/env python3
from icefall.
Thanks for your reply, you are right,
(base) maduo@pd:~$ head -n 5 /home/maduo/miniconda3/envs/k2-fsa20210823/bin/lhotse
#!python
"""
Use this script like:
$ lhotse --help
Yes, I manually modified it.
from icefall.
@pzelasko
Looks like the shebang
problem is reproducible.
from icefall.
I occurs another error in the command lhotse download librispeech
, the errror is as follows:
(k2-fsa20210823) maduo@pd:~/w2021/k2-fsa_20210823/icefall/egs/librispeech/ASR$ ./prepare.sh --stage -0 --stop_stage 3
2021-08-23 17:42:44 (prepare.sh:57:main) dl_dir: /mnt/4T/md/icefall_recipes/librispeech/download
2021-08-23 17:42:44 (prepare.sh:66:main) stage 0: Download data
Downloading LibriSpeech parts: 0%| | 0/7 [00:00<?, ?it/s]2021-08-23 17:42:45,494 INFO [librispeech.py:56] Processing split: dev-clean
2021-08-23 17:42:45,494 INFO [librispeech.py:69] Skipping dev-clean because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/dev-clean/.completed exists.
2021-08-23 17:42:45,494 INFO [librispeech.py:56] Processing split: dev-other
2021-08-23 17:42:45,494 INFO [librispeech.py:69] Skipping dev-other because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/dev-other/.completed exists.
2021-08-23 17:42:45,494 INFO [librispeech.py:56] Processing split: test-clean
2021-08-23 17:42:45,494 INFO [librispeech.py:69] Skipping test-clean because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/test-clean/.completed exists.
2021-08-23 17:42:45,494 INFO [librispeech.py:56] Processing split: test-other
2021-08-23 17:42:45,494 INFO [librispeech.py:69] Skipping test-other because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/test-other/.completed exists.
2021-08-23 17:42:45,494 INFO [librispeech.py:56] Processing split: train-clean-100
2021-08-23 17:42:45,494 INFO [librispeech.py:69] Skipping train-clean-100 because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/train-clean-100/.completed exists.
2021-08-23 17:42:45,494 INFO [librispeech.py:56] Processing split: train-clean-360
Downloading LibriSpeech parts: 71%|██████████████████████████████████████████████████████████████████████████████████████████████████████▏ | 5/7 [00:57<00:23, 11.54s/it]
Aborted!
from icefall.
If you rerun that command it should hopefully continue from the stage that failed.
Yes, rerun
./prepare.sh --stage -0 --stop_stage 3
will continue to download train-clean-360
.
from icefall.
Thanks for your reply, I follow your suggestion, but it is still fail. the the below is as follows:
Filesystem Size Used Avail Use% Mounted on
/dev/sdb 3.6T 1.4T 2.1T 41% /mnt/4T
(k2-fsa20210823) maduo@pd:~/w2021/k2-fsa_20210823/icefall/egs/librispeech/ASR$ rm -rf /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/train-clean-360
(k2-fsa20210823) maduo@pd:~/w2021/k2-fsa_20210823/icefall/egs/librispeech/ASR$ ./prepare.sh --stage -0 --stop_stage 3
2021-08-23 18:21:09 (prepare.sh:57:main) dl_dir: /mnt/4T/md/icefall_recipes/librispeech/download
2021-08-23 18:21:09 (prepare.sh:66:main) stage 0: Download data
Downloading LibriSpeech parts: 0%| | 0/7 [00:00<?, ?it/s]2021-08-23 18:21:10,792 INFO [librispeech.py:56] Processing split: dev-clean
2021-08-23 18:21:10,792 INFO [librispeech.py:69] Skipping dev-clean because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/dev-clean/.completed exists.
2021-08-23 18:21:10,792 INFO [librispeech.py:56] Processing split: dev-other
2021-08-23 18:21:10,792 INFO [librispeech.py:69] Skipping dev-other because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/dev-other/.completed exists.
2021-08-23 18:21:10,792 INFO [librispeech.py:56] Processing split: test-clean
2021-08-23 18:21:10,792 INFO [librispeech.py:69] Skipping test-clean because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/test-clean/.completed exists.
2021-08-23 18:21:10,792 INFO [librispeech.py:56] Processing split: test-other
2021-08-23 18:21:10,792 INFO [librispeech.py:69] Skipping test-other because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/test-other/.completed exists.
2021-08-23 18:21:10,792 INFO [librispeech.py:56] Processing split: train-clean-100
2021-08-23 18:21:10,792 INFO [librispeech.py:69] Skipping train-clean-100 because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/train-clean-100/.completed exists.
2021-08-23 18:21:10,792 INFO [librispeech.py:56] Processing split: train-clean-360
Downloading LibriSpeech parts: 71%|██████████████████████████████████████████████████████████████████████████████████████████████████████▏ | 5/7 [00:54<00:21, 10.89s/it]
Aborted!
from icefall.
@pzelasko do you have any ideas?
This is really a Lhotse issue.
Perhaps we need a better way to either continue partial downloads, or better debug why downloads are killed.
from icefall.
OK I will look into it. I am not sure what's the reason, maybe there is a timeout somewhere. Will check.
I'll try to take care of the shebang issue first though.
from icefall.
Was there some exception stack trace?
It looks to me like it didn't even start downloading train-clean-360, otherwise there would have been a partial download message like this:
@shanguanma could you try wrapping this whole loop:
https://github.com/lhotse-speech/lhotse/blob/master/lhotse/recipes/librispeech.py#L55
with try/except like the following:
try:
for ...
except Exception as e:
print(type(e))
print(e)
print(locals())
raise
and see what comes out? I am also not 100% sure that it is a Lhotse error, your job might be getting killed etc., but we can try to debug it.
from icefall.
@pzelasko ,I follow your suggestion, the errors is as follows:
(k2-fsa20210823) maduo@pd:~/w2021/k2-fsa_20210823/icefall/egs/librispeech/ASR$ ./prepare.sh --stage 0 --stop_stage 0
2021-08-24 19:20:54 (prepare.sh:57:main) dl_dir: /mnt/4T/md/icefall_recipes/librispeech/download
2021-08-24 19:20:54 (prepare.sh:66:main) stage 0: Download data
Downloading LibriSpeech parts: 0%| | 0/7 [00:00<?, ?it/s]2021-08-24 19:20:55,222 INFO [librispeech.py:56] Processing split: dev-clean
2021-08-24 19:20:55,222 INFO [librispeech.py:69] Skipping dev-clean because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/dev-clean/.completed exists.
2021-08-24 19:20:55,222 INFO [librispeech.py:56] Processing split: dev-other
2021-08-24 19:20:55,222 INFO [librispeech.py:69] Skipping dev-other because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/dev-other/.completed exists.
2021-08-24 19:20:55,222 INFO [librispeech.py:56] Processing split: test-clean
2021-08-24 19:20:55,222 INFO [librispeech.py:69] Skipping test-clean because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/test-clean/.completed exists.
2021-08-24 19:20:55,222 INFO [librispeech.py:56] Processing split: test-other
2021-08-24 19:20:55,222 INFO [librispeech.py:69] Skipping test-other because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/test-other/.completed exists.
2021-08-24 19:20:55,222 INFO [librispeech.py:56] Processing split: train-clean-100
2021-08-24 19:20:55,222 INFO [librispeech.py:69] Skipping train-clean-100 because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/train-clean-100/.completed exists.
2021-08-24 19:20:55,222 INFO [librispeech.py:56] Processing split: train-clean-360
Downloading LibriSpeech parts: 71%|██████████████████████████████████████████████████████████████████████████████████████████████████████▏ | 5/7 [00:57<00:23, 11.59s/it]<class 'EOFError'>
Compressed file ended before the end-of-stream marker was reached
{'target_dir': PosixPath('/mnt/4T/md/icefall_recipes/librispeech/download'), 'dataset_parts': ('dev-clean', 'dev-other', 'test-clean', 'test-other', 'train-clean-100', 'train-clean-360', 'train-other-500'), 'force_download': False, 'alignments': False, 'base_url': 'http://www.openslr.org/resources', 'alignments_url': 'https://drive.google.com/uc?id=1WYfgr31T-PPwMcxuAq09XZfHQO5Mw8fE', 'part': 'train-clean-360', 'url': 'http://www.openslr.org/resources/12', 'part_dir': PosixPath('/mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/train-clean-360'), 'completed_detector': PosixPath('/mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/train-clean-360/.completed'), 'tar_name': 'train-clean-360.tar.gz', 'tar_path': PosixPath('/mnt/4T/md/icefall_recipes/librispeech/download/train-clean-360.tar.gz'), 'tar': <tarfile.TarFile object at 0x7f5fb3d76f70>, 'e': EOFError('Compressed file ended before the end-of-stream marker was reached')}
Aborted!
from icefall.
You have a partially downloaded archive. If you remove train-clean-360.tar.gz and re-run it should work. I will change the code to handle this correctly.
from icefall.
@pzelasko ,I follow your suggestion( I remove train-clean-360.tar.gz and re-run it
)the errors is as follows:
Downloading LibriSpeech parts: 0%| | 0/7 [00:00<?, ?it/s]2021-08-24 19:31:09,098 INFO [librispeech.py:56] Processing split: dev-clean
2021-08-24 19:31:09,098 INFO [librispeech.py:69] Skipping dev-clean because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/dev-clean/.completed exists.
2021-08-24 19:31:09,098 INFO [librispeech.py:56] Processing split: dev-other
2021-08-24 19:31:09,098 INFO [librispeech.py:69] Skipping dev-other because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/dev-other/.completed exists.
2021-08-24 19:31:09,099 INFO [librispeech.py:56] Processing split: test-clean
2021-08-24 19:31:09,099 INFO [librispeech.py:69] Skipping test-clean because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/test-clean/.completed exists.
2021-08-24 19:31:09,099 INFO [librispeech.py:56] Processing split: test-other
2021-08-24 19:31:09,099 INFO [librispeech.py:69] Skipping test-other because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/test-other/.completed exists.
2021-08-24 19:31:09,099 INFO [librispeech.py:56] Processing split: train-clean-100
2021-08-24 19:31:09,099 INFO [librispeech.py:69] Skipping train-clean-100 because /mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/train-clean-100/.completed exists.
2021-08-24 19:31:09,099 INFO [librispeech.py:56] Processing split: train-clean-360
Downloading train-clean-360.tar.gz: 21.5GB [1:32:18, 4.16MB/s] Downloading LibriSpeech parts: 86%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ | 6/7 [1:37:25<16:14, 974.31s/it]2021-08-24 21:08:34,933 INFO [librispeech.py:56] Processing split: train-other-500
Downloading train-other-500.tar.gz: 80%|███████████████████████████████████████████████████████████████████████████████████████████████████████▊ | 22.9G/28.5G [1:50:48<26:51, 3.70MB/s]Downloading LibriSpeech parts: 86%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ | 6/7 [3:28:14<34:42, 2082.37s/it]<class 'urllib.error.ContentTooShortError'>
<urlopen error retrieval incomplete: got only 24626188545 out of 30593501606 bytes>
{'target_dir': PosixPath('/mnt/4T/md/icefall_recipes/librispeech/download'), 'dataset_parts': ('dev-clean', 'dev-other', 'test-clean', 'test-other', 'train-clean-100', 'train-clean-360', 'train-other-500'), 'force_download': False, 'alignments': False, 'base_url': 'http://www.openslr.org/resources', 'alignments_url': 'https://drive.google.com/uc?id=1WYfgr31T-PPwMcxuAq09XZfHQO5Mw8fE', 'part': 'train-other-500', 'url': 'http://www.openslr.org/resources/12', 'part_dir': PosixPath('/mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/train-other-500'), 'completed_detector': PosixPath('/mnt/4T/md/icefall_recipes/librispeech/download/LibriSpeech/train-other-500/.completed'), 'tar_name': 'train-other-500.tar.gz', 'tar_path': PosixPath('/mnt/4T/md/icefall_recipes/librispeech/download/train-other-500.tar.gz'), 'tar': <tarfile.TarFile object at 0x7f7bdff7edf0>, 'e': ContentTooShortError('retrieval incomplete: got only 24626188545 out of 30593501606 bytes')}
Traceback (most recent call last):
File "/home/maduo/miniconda3/envs/k2-fsa20210823/bin/lhotse", line 24, in <module>
cli()
File "/home/maduo/miniconda3/envs/k2-fsa20210823/lib/python3.8/site-packages/click/core.py", line 1137, in __call__
return self.main(*args, **kwargs)
File "/home/maduo/miniconda3/envs/k2-fsa20210823/lib/python3.8/site-packages/click/core.py", line 1062, in main
rv = self.invoke(ctx)
File "/home/maduo/miniconda3/envs/k2-fsa20210823/lib/python3.8/site-packages/click/core.py", line 1668, in invoke
return _process_result(sub_ctx.command.invoke(sub_ctx))
File "/home/maduo/miniconda3/envs/k2-fsa20210823/lib/python3.8/site-packages/click/core.py", line 1668, in invoke
return _process_result(sub_ctx.command.invoke(sub_ctx))
File "/home/maduo/miniconda3/envs/k2-fsa20210823/lib/python3.8/site-packages/click/core.py", line 1404, in invoke
return ctx.invoke(self.callback, **ctx.params)
File "/home/maduo/miniconda3/envs/k2-fsa20210823/lib/python3.8/site-packages/click/core.py", line 763, in invoke
return __callback(*args, **kwargs)
File "/home/maduo/miniconda3/envs/k2-fsa20210823/lib/python3.8/site-packages/lhotse/bin/modes/recipes/librispeech.py", line 32, in librispeech
download_librispeech(target_dir, dataset_parts='librispeech' if full else 'mini_librispeech')
File "/home/maduo/miniconda3/envs/k2-fsa20210823/lib/python3.8/site-packages/lhotse/recipes/librispeech.py", line 75, in download_librispeech
urlretrieve_progress(f'{url}/{tar_name}', filename=tar_path, desc=f'Downloading {tar_name}')
File "/home/maduo/miniconda3/envs/k2-fsa20210823/lib/python3.8/site-packages/lhotse/utils.py", line 335, in urlretrieve_progress
return urlretrieve(url=url, filename=filename, reporthook=reporthook, data=data)
File "/home/maduo/miniconda3/envs/k2-fsa20210823/lib/python3.8/urllib/request.py", line 286, in urlretrieve
raise ContentTooShortError(
urllib.error.ContentTooShortError: <urlopen error retrieval incomplete: got only 24626188545 out of 30593501606 bytes>
from icefall.
The download error occurred to me too. My download of train-clean-360.tar.gz
was kill twice at the same position (71%).
And this is the error of train-other-500.tar.gz.
I guess it is the problem of weak network connection, it takes too long to download these data, may be some unstable connection during these time.
from icefall.
Perhaps it would be better to use another tool for downloading, that allows continuing? E.g. wget?
Maybe that urlretrieve is only good for short files.
from icefall.
I think it’s a server side timeout… for now please use wget for these two files like Dan suggested, I might not have enough time to improve the downloading in Lhotse right away, but I definitely want it to “just work” and will work on it.
from icefall.
Related Issues (20)
- Identical Batches Across Multiple GPUs HOT 2
- CTC/AED PROBLEM IN K2 HOT 10
- append features HOT 1
- CTC/AED PROBLEMS IN EXPORTING JIT MODULE HOT 4
- Error happens with egs/librispeech/ASR/prepare_mmi.sh HOT 3
- Use CutSet.mux to effect? HOT 10
- Help with training/finetuning a zipformer based model HOT 6
- Different Training Loss with Single Node (8 GPUs) vs. Two Nodes (4 GPUs Each)
- Data cleaning HOT 3
- ONNX decode error HOT 2
- OTC with conformer librispeech/WASR isn't converage.
- ONNX bug HOT 9
- Questions about modifying prepare.sh for training ASR model on custom data HOT 2
- How to use my own dataset based on another dataset HOT 3
- kaldifeat installation error HOT 2
- Why unique lexicon is needed in Chinese ASR, but not in English ASR?
- Error during training OTC conformer_ctc2 HOT 1
- What is difference between zipformer and zipformer_ctc models? HOT 1
- ONNX and Torch models HOT 1
- how to decrease the right chunk size when using zipformer model?
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from icefall.