Hi Nicoara, Machine translation is a memory-intensive task, with large amounts of memory used for storing translation information and language modeling information.
>From the language packs page <https://cwiki.apache.org/confluence/display/JOSHUA/Language+Packs>: add an option to the decoder (e.g. in this case the ./joshua command) - "-m XXg" — increase the amount of memory provided to Joshua. The default is 8g, but for the larger language packs, you will want 16 or 24. In general, 50% more memory than the raw model size should be more than sufficient. Since you're using the English-German pack, which is 4.5GB, I'd suggest about 8GB of RAM. Also make sure your VM has been allocated that much RAM from the host system. Best, John On Mon, Jun 26, 2017 at 12:00 PM, Nicoara Talpes <nicoaratal...@gmail.com> wrote: > Hello Joshua Community, > > When running in an Ubuntu VM "cat example.SRC | ./prepare.sh | ./joshua" > as per instructions here https://cwiki.apache.org/ > confluence/display/JOSHUA/Language+Packs the jvm goes quickly to 4 gb > and the process crashes. > > The input is a file with one line to translate: "Hello". > > What is the amount of RAM required by Joshua? How to limit the amount of > RAM required? > > The error is this: > nicoara@ubuntu:~/Desktop/joshua/apache-joshua-en-de-2017-01-31$ cat > example.SRC | ./prepare.sh | ./joshua > Java HotSpot(TM) 64-Bit Server VM warning: INFO: > os::commit_memory(0x00000005cbf00000, 76021760, 0) failed; error='Cannot > allocate memory' (errno=12) > # > # There is insufficient memory for the Java Runtime Environment to > continue. > # Native memory allocation (mmap) failed to map 76xxxxxxxxx021760 bytes > for committing reserved memory. > # An error report file with more information is saved as: > # /home/nicoara/Desktop/joshua/apache-joshua-en-de-2017-01-31/ > hs_err_pid10488.log > > Thank you, > Nicoara >
