search

nanoporetech / dorado

Oxford Nanopore's Basecaller
https://nanoporetech.com/
Other
439 stars 53 forks source link

"dorado correct" failed to stop running #891

Open yzhang-github-pub opened 1 week ago

yzhang-github-pub commented 1 week ago

I have a down-sampled data of 1000 reads. Read length stat is as follows:

num_seqs sum_len min_len avg_len max_len Q1 Q2 Q3 N50 Q20(%) Q30(%) 1,000 3,484,177 30 3,484.2 24,027 258 1,222 4,174.5 10,036 74.62 54.54

The command is: "dorado correct --model-path herro-v1 reads1000.fastq > reads.fasta"

Within one minute of running the command, the status reached 100%. But the program didn't exit. Instead it kept running seemingly forever (I waited for 60 min and killed the process). The output file modification timestamp and size never changed after the status reached 100%, indicating the program actually finished correction in 1 min, and after that it did nothing. 'nvidia-smi' showed two running processes of dorado:

0   N/A  N/A     50784      C   ...n/dorado-0.7.1-linux-x64/bin/dorado    11602MiB 
1   N/A  N/A     50784      C   ...n/dorado-0.7.1-linux-x64/bin/dorado    17172MiB

BTW, the corrected reads looked good.

malton-ont commented 1 week ago

Hi @yzhang-github-pub,

Thanks for reporting this! I've been able to reproduce it here, and we'll aim to get a fix out in a future release.

yzhang-github-pub commented 6 days ago

@malton-ont Thanks for taking care of this!